diff options
Diffstat (limited to 'drivers/gpu/drm/amd')
320 files changed, 14274 insertions, 7222 deletions
diff --git a/drivers/gpu/drm/amd/amdgpu/ObjectID.h b/drivers/gpu/drm/amd/amdgpu/ObjectID.h index 06192698bd96..5b393622f592 100644 --- a/drivers/gpu/drm/amd/amdgpu/ObjectID.h +++ b/drivers/gpu/drm/amd/amdgpu/ObjectID.h @@ -136,6 +136,7 @@ #define GENERIC_OBJECT_ID_PX2_NON_DRIVABLE 0x02 #define GENERIC_OBJECT_ID_MXM_OPM 0x03 #define GENERIC_OBJECT_ID_STEREO_PIN 0x04 //This object could show up from Misc Object table, it follows ATOM_OBJECT format, and contains one ATOM_OBJECT_GPIO_CNTL_RECORD for the stereo pin +#define GENERIC_OBJECT_ID_BRACKET_LAYOUT 0x05 /****************************************************/ /* Graphics Object ENUM ID Definition */ @@ -714,6 +715,13 @@ GRAPH_OBJECT_ENUM_ID1 << ENUM_ID_SHIFT |\ GENERIC_OBJECT_ID_STEREO_PIN << OBJECT_ID_SHIFT) +#define GENERICOBJECT_BRACKET_LAYOUT_ENUM_ID1 (GRAPH_OBJECT_TYPE_GENERIC << OBJECT_TYPE_SHIFT |\ + GRAPH_OBJECT_ENUM_ID1 << ENUM_ID_SHIFT |\ + GENERIC_OBJECT_ID_BRACKET_LAYOUT << OBJECT_ID_SHIFT) + +#define GENERICOBJECT_BRACKET_LAYOUT_ENUM_ID2 (GRAPH_OBJECT_TYPE_GENERIC << OBJECT_TYPE_SHIFT |\ + GRAPH_OBJECT_ENUM_ID2 << ENUM_ID_SHIFT |\ + GENERIC_OBJECT_ID_BRACKET_LAYOUT << OBJECT_ID_SHIFT) /****************************************************/ /* Object Cap definition - Shared with BIOS */ /****************************************************/ diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index 7dcbac8af9a7..447c4c7a36d6 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -73,6 +73,8 @@ #include "amdgpu_virt.h" #include "amdgpu_gart.h" #include "amdgpu_debugfs.h" +#include "amdgpu_job.h" +#include "amdgpu_bo_list.h" /* * Modules parameters. @@ -105,11 +107,8 @@ extern int amdgpu_vm_fault_stop; extern int amdgpu_vm_debug; extern int amdgpu_vm_update_mode; extern int amdgpu_dc; -extern int amdgpu_dc_log; extern int amdgpu_sched_jobs; extern int amdgpu_sched_hw_submission; -extern int amdgpu_no_evict; -extern int amdgpu_direct_gma_size; extern uint amdgpu_pcie_gen_cap; extern uint amdgpu_pcie_lane_cap; extern uint amdgpu_cg_mask; @@ -600,17 +599,6 @@ struct amdgpu_ib { 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); -int amdgpu_job_alloc_with_ib(struct amdgpu_device *adev, unsigned size, - struct amdgpu_job **job); - -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 drm_sched_entity *entity, void *owner, - struct dma_fence **f); - /* * Queue manager */ @@ -684,8 +672,8 @@ int amdgpu_ctx_ioctl(struct drm_device *dev, void *data, int amdgpu_ctx_wait_prev_fence(struct amdgpu_ctx *ctx, unsigned ring_id); void amdgpu_ctx_mgr_init(struct amdgpu_ctx_mgr *mgr); -void amdgpu_ctx_mgr_entity_cleanup(struct amdgpu_ctx_mgr *mgr); void amdgpu_ctx_mgr_entity_fini(struct amdgpu_ctx_mgr *mgr); +void amdgpu_ctx_mgr_entity_flush(struct amdgpu_ctx_mgr *mgr); void amdgpu_ctx_mgr_fini(struct amdgpu_ctx_mgr *mgr); @@ -703,37 +691,6 @@ struct amdgpu_fpriv { }; /* - * residency list - */ -struct amdgpu_bo_list_entry { - struct amdgpu_bo *robj; - struct ttm_validate_buffer tv; - struct amdgpu_bo_va *bo_va; - uint32_t priority; - struct page **user_pages; - int user_invalidated; -}; - -struct amdgpu_bo_list { - struct mutex lock; - struct rcu_head rhead; - struct kref refcount; - struct amdgpu_bo *gds_obj; - struct amdgpu_bo *gws_obj; - struct amdgpu_bo *oa_obj; - unsigned first_userptr; - unsigned num_entries; - struct amdgpu_bo_list_entry *array; -}; - -struct amdgpu_bo_list * -amdgpu_bo_list_get(struct amdgpu_fpriv *fpriv, int id); -void amdgpu_bo_list_get_list(struct amdgpu_bo_list *list, - struct list_head *validated); -void amdgpu_bo_list_put(struct amdgpu_bo_list *list); -void amdgpu_bo_list_free(struct amdgpu_bo_list *list); - -/* * GFX stuff */ #include "clearstate_defs.h" @@ -931,6 +888,11 @@ struct amdgpu_ngg { bool init; }; +struct sq_work { + struct work_struct work; + unsigned ih_data; +}; + struct amdgpu_gfx { struct mutex gpu_clock_mutex; struct amdgpu_gfx_config config; @@ -969,6 +931,10 @@ struct amdgpu_gfx { struct amdgpu_irq_src eop_irq; struct amdgpu_irq_src priv_reg_irq; struct amdgpu_irq_src priv_inst_irq; + struct amdgpu_irq_src cp_ecc_error_irq; + struct amdgpu_irq_src sq_irq; + struct sq_work sq_work; + /* gfx status */ uint32_t gfx_current_status; /* ce ram size*/ @@ -1020,6 +986,7 @@ struct amdgpu_cs_parser { /* scheduler job object */ struct amdgpu_job *job; + struct amdgpu_ring *ring; /* buffer objects */ struct ww_acquire_ctx ticket; @@ -1041,40 +1008,6 @@ struct amdgpu_cs_parser { struct drm_syncobj **post_dep_syncobjs; }; -#define AMDGPU_PREAMBLE_IB_PRESENT (1 << 0) /* bit set means command submit involves a preamble IB */ -#define AMDGPU_PREAMBLE_IB_PRESENT_FIRST (1 << 1) /* bit set means preamble IB is first presented in belonging context */ -#define AMDGPU_HAVE_CTX_SWITCH (1 << 2) /* bit set means context switch occured */ - -struct amdgpu_job { - struct drm_sched_job base; - struct amdgpu_device *adev; - struct amdgpu_vm *vm; - struct amdgpu_ring *ring; - struct amdgpu_sync sync; - struct amdgpu_sync sched_sync; - struct amdgpu_ib *ibs; - struct dma_fence *fence; /* the hw fence */ - uint32_t preamble_status; - uint32_t num_ibs; - void *owner; - uint64_t fence_ctx; /* the fence_context this job uses */ - bool vm_needs_flush; - uint64_t vm_pd_addr; - unsigned vmid; - unsigned pasid; - uint32_t gds_base, gds_size; - uint32_t gws_base, gws_size; - uint32_t oa_base, oa_size; - uint32_t vram_lost_counter; - - /* user fence handling */ - uint64_t uf_addr; - uint64_t uf_sequence; - -}; -#define to_amdgpu_job(sched_job) \ - container_of((sched_job), struct amdgpu_job, base) - static inline u32 amdgpu_get_ib_value(struct amdgpu_cs_parser *p, uint32_t ib_idx, int idx) { @@ -1389,6 +1322,7 @@ enum amd_hw_ip_block_type { PWR_HWIP, NBIF_HWIP, THM_HWIP, + CLK_HWIP, MAX_HWIP }; @@ -1579,9 +1513,9 @@ struct amdgpu_device { DECLARE_HASHTABLE(mn_hash, 7); /* tracking pinned memory */ - u64 vram_pin_size; - u64 invisible_pin_size; - u64 gart_pin_size; + atomic64_t vram_pin_size; + atomic64_t visible_pin_size; + atomic64_t gart_pin_size; /* amdkfd interface */ struct kfd_dev *kfd; @@ -1776,6 +1710,7 @@ amdgpu_get_sdma_instance(struct amdgpu_ring *ring) #define amdgpu_vm_write_pte(adev, ib, pe, value, count, incr) ((adev)->vm_manager.vm_pte_funcs->write_pte((ib), (pe), (value), (count), (incr))) #define amdgpu_vm_set_pte_pde(adev, ib, pe, addr, count, incr, flags) ((adev)->vm_manager.vm_pte_funcs->set_pte_pde((ib), (pe), (addr), (count), (incr), (flags))) #define amdgpu_ring_parse_cs(r, p, ib) ((r)->funcs->parse_cs((p), (ib))) +#define amdgpu_ring_patch_cs_in_place(r, p, ib) ((r)->funcs->patch_cs_in_place((p), (ib))) #define amdgpu_ring_test_ring(r) (r)->funcs->test_ring((r)) #define amdgpu_ring_test_ib(r, t) (r)->funcs->test_ib((r), (t)) #define amdgpu_ring_get_rptr(r) (r)->funcs->get_rptr((r)) @@ -1829,8 +1764,6 @@ void amdgpu_display_update_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_device_vram_location(struct amdgpu_device *adev, struct amdgpu_gmc *mc, u64 base); void amdgpu_device_gart_location(struct amdgpu_device *adev, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c index 0d8c3fc6eace..353993218f21 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c @@ -364,7 +364,6 @@ static int amdgpu_atif_handler(struct amdgpu_device *adev, struct acpi_bus_event *event) { struct amdgpu_atif *atif = adev->atif; - struct atif_sbios_requests req; int count; DRM_DEBUG_DRIVER("event, device_class = %s, type = %#x\n", @@ -379,42 +378,48 @@ static int amdgpu_atif_handler(struct amdgpu_device *adev, /* Not our event */ return NOTIFY_DONE; - /* Check pending SBIOS requests */ - count = amdgpu_atif_get_sbios_requests(atif, &req); + if (atif->functions.sbios_requests) { + struct atif_sbios_requests req; - if (count <= 0) - return NOTIFY_DONE; + /* Check pending SBIOS requests */ + count = amdgpu_atif_get_sbios_requests(atif, &req); + + if (count <= 0) + return NOTIFY_DONE; - DRM_DEBUG_DRIVER("ATIF: %d pending SBIOS requests\n", count); + DRM_DEBUG_DRIVER("ATIF: %d pending SBIOS requests\n", count); - if (req.pending & ATIF_PANEL_BRIGHTNESS_CHANGE_REQUEST) { - struct amdgpu_encoder *enc = atif->encoder_for_bl; + /* todo: add DC handling */ + if ((req.pending & ATIF_PANEL_BRIGHTNESS_CHANGE_REQUEST) && + !amdgpu_device_has_dc_support(adev)) { + struct amdgpu_encoder *enc = atif->encoder_for_bl; - if (enc) { - struct amdgpu_encoder_atom_dig *dig = enc->enc_priv; + if (enc) { + struct amdgpu_encoder_atom_dig *dig = enc->enc_priv; - DRM_DEBUG_DRIVER("Changing brightness to %d\n", - req.backlight_level); + DRM_DEBUG_DRIVER("Changing brightness to %d\n", + req.backlight_level); - amdgpu_display_backlight_set_level(adev, enc, req.backlight_level); + amdgpu_display_backlight_set_level(adev, enc, req.backlight_level); #if defined(CONFIG_BACKLIGHT_CLASS_DEVICE) || defined(CONFIG_BACKLIGHT_CLASS_DEVICE_MODULE) - backlight_force_update(dig->bl_dev, - BACKLIGHT_UPDATE_HOTKEY); + backlight_force_update(dig->bl_dev, + BACKLIGHT_UPDATE_HOTKEY); #endif + } } - } - if (req.pending & ATIF_DGPU_DISPLAY_EVENT) { - if ((adev->flags & AMD_IS_PX) && - amdgpu_atpx_dgpu_req_power_for_displays()) { - pm_runtime_get_sync(adev->ddev->dev); - /* Just fire off a uevent and let userspace tell us what to do */ - drm_helper_hpd_irq_event(adev->ddev); - pm_runtime_mark_last_busy(adev->ddev->dev); - pm_runtime_put_autosuspend(adev->ddev->dev); + if (req.pending & ATIF_DGPU_DISPLAY_EVENT) { + if ((adev->flags & AMD_IS_PX) && + amdgpu_atpx_dgpu_req_power_for_displays()) { + pm_runtime_get_sync(adev->ddev->dev); + /* Just fire off a uevent and let userspace tell us what to do */ + drm_helper_hpd_irq_event(adev->ddev); + pm_runtime_mark_last_busy(adev->ddev->dev); + pm_runtime_put_autosuspend(adev->ddev->dev); + } } + /* TODO: check other events */ } - /* TODO: check other events */ /* We've handled the event, stop the notifier chain. The ACPI interface * overloads ACPI_VIDEO_NOTIFY_PROBE, we don't want to send that to diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c index 305143fcc1ce..f8bbbb3a9504 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c @@ -243,6 +243,33 @@ int amdgpu_amdkfd_resume(struct amdgpu_device *adev) return r; } +int amdgpu_amdkfd_pre_reset(struct amdgpu_device *adev) +{ + int r = 0; + + if (adev->kfd) + r = kgd2kfd->pre_reset(adev->kfd); + + return r; +} + +int amdgpu_amdkfd_post_reset(struct amdgpu_device *adev) +{ + int r = 0; + + if (adev->kfd) + r = kgd2kfd->post_reset(adev->kfd); + + return r; +} + +void amdgpu_amdkfd_gpu_reset(struct kgd_dev *kgd) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)kgd; + + amdgpu_device_gpu_recover(adev, NULL, false); +} + int alloc_gtt_mem(struct kgd_dev *kgd, size_t size, void **mem_obj, uint64_t *gpu_addr, void **cpu_ptr) @@ -251,7 +278,6 @@ int alloc_gtt_mem(struct kgd_dev *kgd, size_t size, struct amdgpu_bo *bo = NULL; struct amdgpu_bo_param bp; int r; - uint64_t gpu_addr_tmp = 0; void *cpu_ptr_tmp = NULL; memset(&bp, 0, sizeof(bp)); @@ -275,13 +301,18 @@ int alloc_gtt_mem(struct kgd_dev *kgd, size_t size, goto allocate_mem_reserve_bo_failed; } - r = amdgpu_bo_pin(bo, AMDGPU_GEM_DOMAIN_GTT, - &gpu_addr_tmp); + r = amdgpu_bo_pin(bo, AMDGPU_GEM_DOMAIN_GTT); if (r) { dev_err(adev->dev, "(%d) failed to pin bo for amdkfd\n", r); goto allocate_mem_pin_bo_failed; } + r = amdgpu_ttm_alloc_gart(&bo->tbo); + if (r) { + dev_err(adev->dev, "%p bind failed\n", bo); + goto allocate_mem_kmap_bo_failed; + } + r = amdgpu_bo_kmap(bo, &cpu_ptr_tmp); if (r) { dev_err(adev->dev, @@ -290,7 +321,7 @@ int alloc_gtt_mem(struct kgd_dev *kgd, size_t size, } *mem_obj = bo; - *gpu_addr = gpu_addr_tmp; + *gpu_addr = amdgpu_bo_gpu_offset(bo); *cpu_ptr = cpu_ptr_tmp; amdgpu_bo_unreserve(bo); @@ -457,6 +488,14 @@ err: return ret; } +void amdgpu_amdkfd_set_compute_idle(struct kgd_dev *kgd, bool idle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)kgd; + + amdgpu_dpm_switch_power_profile(adev, + PP_SMC_POWER_PROFILE_COMPUTE, !idle); +} + bool amdgpu_amdkfd_is_kfd_vmid(struct amdgpu_device *adev, u32 vmid) { if (adev->kfd) { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h index a8418a3f4e9d..2f379c183ed2 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h @@ -119,6 +119,7 @@ int amdgpu_amdkfd_evict_userptr(struct kgd_mem *mem, struct mm_struct *mm); int amdgpu_amdkfd_submit_ib(struct kgd_dev *kgd, enum kgd_engine_type engine, uint32_t vmid, uint64_t gpu_addr, uint32_t *ib_cmd, uint32_t ib_len); +void amdgpu_amdkfd_set_compute_idle(struct kgd_dev *kgd, bool idle); struct kfd2kgd_calls *amdgpu_amdkfd_gfx_7_get_functions(void); struct kfd2kgd_calls *amdgpu_amdkfd_gfx_8_0_get_functions(void); @@ -126,6 +127,12 @@ struct kfd2kgd_calls *amdgpu_amdkfd_gfx_9_0_get_functions(void); bool amdgpu_amdkfd_is_kfd_vmid(struct amdgpu_device *adev, u32 vmid); +int amdgpu_amdkfd_pre_reset(struct amdgpu_device *adev); + +int amdgpu_amdkfd_post_reset(struct amdgpu_device *adev); + +void amdgpu_amdkfd_gpu_reset(struct kgd_dev *kgd); + /* Shared API */ int alloc_gtt_mem(struct kgd_dev *kgd, size_t size, void **mem_obj, uint64_t *gpu_addr, @@ -183,6 +190,9 @@ int amdgpu_amdkfd_gpuvm_map_gtt_bo_to_kernel(struct kgd_dev *kgd, int amdgpu_amdkfd_gpuvm_restore_process_bos(void *process_info, struct dma_fence **ef); +int amdgpu_amdkfd_gpuvm_get_vm_fault_info(struct kgd_dev *kgd, + struct kfd_vm_fault_info *info); + void amdgpu_amdkfd_gpuvm_init_mem_limits(void); void amdgpu_amdkfd_unreserve_system_memory_limit(struct amdgpu_bo *bo); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_fence.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_fence.c index 2c14025e5e76..574c1181ae9a 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_fence.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_fence.c @@ -173,7 +173,5 @@ static const struct dma_fence_ops amdkfd_fence_ops = { .get_driver_name = amdkfd_fence_get_driver_name, .get_timeline_name = amdkfd_fence_get_timeline_name, .enable_signaling = amdkfd_fence_enable_signaling, - .signaled = NULL, - .wait = dma_fence_default_wait, .release = amdkfd_fence_release, }; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c index ea79908dac4c..ea3f698aef5e 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c @@ -145,6 +145,7 @@ static void set_vm_context_page_table_base(struct kgd_dev *kgd, uint32_t vmid, uint32_t page_table_base); static int invalidate_tlbs(struct kgd_dev *kgd, uint16_t pasid); static int invalidate_tlbs_vmid(struct kgd_dev *kgd, uint16_t vmid); +static uint32_t read_vmid_from_vmfault_reg(struct kgd_dev *kgd); /* Because of REG_GET_FIELD() being used, we put this function in the * asic specific file. @@ -216,6 +217,10 @@ static const struct kfd2kgd_calls kfd2kgd = { .invalidate_tlbs = invalidate_tlbs, .invalidate_tlbs_vmid = invalidate_tlbs_vmid, .submit_ib = amdgpu_amdkfd_submit_ib, + .get_vm_fault_info = amdgpu_amdkfd_gpuvm_get_vm_fault_info, + .read_vmid_from_vmfault_reg = read_vmid_from_vmfault_reg, + .gpu_recover = amdgpu_amdkfd_gpu_reset, + .set_compute_idle = amdgpu_amdkfd_set_compute_idle }; struct kfd2kgd_calls *amdgpu_amdkfd_gfx_7_get_functions(void) @@ -571,6 +576,9 @@ static int kgd_hqd_destroy(struct kgd_dev *kgd, void *mqd, unsigned long flags, end_jiffies; int retry; + if (adev->in_gpu_reset) + return -EIO; + acquire_queue(kgd, pipe_id, queue_id); WREG32(mmCP_HQD_PQ_DOORBELL_CONTROL, 0); @@ -882,6 +890,9 @@ static int invalidate_tlbs(struct kgd_dev *kgd, uint16_t pasid) int vmid; unsigned int tmp; + if (adev->in_gpu_reset) + return -EIO; + for (vmid = 0; vmid < 16; vmid++) { if (!amdgpu_amdkfd_is_kfd_vmid(adev, vmid)) continue; @@ -911,3 +922,19 @@ static int invalidate_tlbs_vmid(struct kgd_dev *kgd, uint16_t vmid) RREG32(mmVM_INVALIDATE_RESPONSE); return 0; } + + /** + * read_vmid_from_vmfault_reg - read vmid from register + * + * adev: amdgpu_device pointer + * @vmid: vmid pointer + * read vmid from register (CIK). + */ +static uint32_t read_vmid_from_vmfault_reg(struct kgd_dev *kgd) +{ + struct amdgpu_device *adev = get_amdgpu_device(kgd); + + uint32_t status = RREG32(mmVM_CONTEXT1_PROTECTION_FAULT_STATUS); + + return REG_GET_FIELD(status, VM_CONTEXT1_PROTECTION_FAULT_STATUS, VMID); +} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c index 19dd665e7307..f6e53e9352bd 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c @@ -176,6 +176,9 @@ static const struct kfd2kgd_calls kfd2kgd = { .invalidate_tlbs = invalidate_tlbs, .invalidate_tlbs_vmid = invalidate_tlbs_vmid, .submit_ib = amdgpu_amdkfd_submit_ib, + .get_vm_fault_info = amdgpu_amdkfd_gpuvm_get_vm_fault_info, + .gpu_recover = amdgpu_amdkfd_gpu_reset, + .set_compute_idle = amdgpu_amdkfd_set_compute_idle }; struct kfd2kgd_calls *amdgpu_amdkfd_gfx_8_0_get_functions(void) @@ -568,6 +571,9 @@ static int kgd_hqd_destroy(struct kgd_dev *kgd, void *mqd, int retry; struct vi_mqd *m = get_mqd(mqd); + if (adev->in_gpu_reset) + return -EIO; + acquire_queue(kgd, pipe_id, queue_id); if (m->cp_hqd_vmid == 0) @@ -844,6 +850,9 @@ static int invalidate_tlbs(struct kgd_dev *kgd, uint16_t pasid) int vmid; unsigned int tmp; + if (adev->in_gpu_reset) + return -EIO; + for (vmid = 0; vmid < 16; vmid++) { if (!amdgpu_amdkfd_is_kfd_vmid(adev, vmid)) continue; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c index 1db60aa5b7f0..8efedfcb9dfc 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c @@ -213,6 +213,8 @@ static const struct kfd2kgd_calls kfd2kgd = { .invalidate_tlbs = invalidate_tlbs, .invalidate_tlbs_vmid = invalidate_tlbs_vmid, .submit_ib = amdgpu_amdkfd_submit_ib, + .gpu_recover = amdgpu_amdkfd_gpu_reset, + .set_compute_idle = amdgpu_amdkfd_set_compute_idle }; struct kfd2kgd_calls *amdgpu_amdkfd_gfx_9_0_get_functions(void) @@ -679,6 +681,9 @@ static int kgd_hqd_destroy(struct kgd_dev *kgd, void *mqd, uint32_t temp; struct v9_mqd *m = get_mqd(mqd); + if (adev->in_gpu_reset) + return -EIO; + acquire_queue(kgd, pipe_id, queue_id); if (m->cp_hqd_vmid == 0) @@ -866,6 +871,9 @@ static int invalidate_tlbs(struct kgd_dev *kgd, uint16_t pasid) int vmid; struct amdgpu_ring *ring = &adev->gfx.kiq.ring; + if (adev->in_gpu_reset) + return -EIO; + if (ring->ready) return invalidate_tlbs_with_kiq(adev, pasid); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c index ff8fd75f7ca5..8a707d8bbb1c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c @@ -334,7 +334,7 @@ static int amdgpu_amdkfd_bo_validate(struct amdgpu_bo *bo, uint32_t domain, "Called with userptr BO")) return -EINVAL; - amdgpu_ttm_placement_from_domain(bo, domain); + amdgpu_bo_placement_from_domain(bo, domain); ret = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx); if (ret) @@ -622,7 +622,7 @@ static int init_user_pages(struct kgd_mem *mem, struct mm_struct *mm, pr_err("%s: Failed to reserve BO\n", __func__); goto release_out; } - amdgpu_ttm_placement_from_domain(bo, mem->domain); + amdgpu_bo_placement_from_domain(bo, mem->domain); ret = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx); if (ret) pr_err("%s: failed to validate BO\n", __func__); @@ -1587,7 +1587,7 @@ int amdgpu_amdkfd_gpuvm_map_gtt_bo_to_kernel(struct kgd_dev *kgd, goto bo_reserve_failed; } - ret = amdgpu_bo_pin(bo, AMDGPU_GEM_DOMAIN_GTT, NULL); + ret = amdgpu_bo_pin(bo, AMDGPU_GEM_DOMAIN_GTT); if (ret) { pr_err("Failed to pin bo. ret %d\n", ret); goto pin_failed; @@ -1621,6 +1621,20 @@ bo_reserve_failed: return ret; } +int amdgpu_amdkfd_gpuvm_get_vm_fault_info(struct kgd_dev *kgd, + struct kfd_vm_fault_info *mem) +{ + struct amdgpu_device *adev; + + adev = (struct amdgpu_device *)kgd; + if (atomic_read(&adev->gmc.vm_fault_info_updated) == 1) { + *mem = *adev->gmc.vm_fault_info; + mb(); + atomic_set(&adev->gmc.vm_fault_info_updated, 0); + } + return 0; +} + /* Evict a userptr BO by stopping the queues if necessary * * Runs in MMU notifier, may be in RECLAIM_FS context. This means it @@ -1680,7 +1694,7 @@ static int update_invalid_user_pages(struct amdkfd_process_info *process_info, if (amdgpu_bo_reserve(bo, true)) return -EAGAIN; - amdgpu_ttm_placement_from_domain(bo, AMDGPU_GEM_DOMAIN_CPU); + amdgpu_bo_placement_from_domain(bo, AMDGPU_GEM_DOMAIN_CPU); ret = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx); amdgpu_bo_unreserve(bo); if (ret) { @@ -1824,7 +1838,7 @@ static int validate_invalid_user_pages(struct amdkfd_process_info *process_info) if (mem->user_pages[0]) { amdgpu_ttm_tt_set_user_pages(bo->tbo.ttm, mem->user_pages); - amdgpu_ttm_placement_from_domain(bo, mem->domain); + amdgpu_bo_placement_from_domain(bo, mem->domain); ret = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx); if (ret) { pr_err("%s: failed to validate BO\n", __func__); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c index ca8bf1c9a98e..a028661d9e20 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c @@ -32,7 +32,7 @@ struct amdgpu_atpx_functions { bool switch_start; bool switch_end; bool disp_connectors_mapping; - bool disp_detetion_ports; + bool disp_detection_ports; }; struct amdgpu_atpx { @@ -162,7 +162,7 @@ static void amdgpu_atpx_parse_functions(struct amdgpu_atpx_functions *f, u32 mas f->switch_start = mask & ATPX_GRAPHICS_DEVICE_SWITCH_START_NOTIFICATION_SUPPORTED; f->switch_end = mask & ATPX_GRAPHICS_DEVICE_SWITCH_END_NOTIFICATION_SUPPORTED; f->disp_connectors_mapping = mask & ATPX_GET_DISPLAY_CONNECTORS_MAPPING_SUPPORTED; - f->disp_detetion_ports = mask & ATPX_GET_DISPLAY_DETECTION_PORTS_SUPPORTED; + f->disp_detection_ports = mask & ATPX_GET_DISPLAY_DETECTION_PORTS_SUPPORTED; } /** diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_benchmark.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_benchmark.c index 19cfff31f2e1..3079ea8523c5 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_benchmark.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_benchmark.c @@ -95,11 +95,17 @@ static void amdgpu_benchmark_move(struct amdgpu_device *adev, unsigned size, r = amdgpu_bo_reserve(sobj, false); if (unlikely(r != 0)) goto out_cleanup; - r = amdgpu_bo_pin(sobj, sdomain, &saddr); + r = amdgpu_bo_pin(sobj, sdomain); + if (r) { + amdgpu_bo_unreserve(sobj); + goto out_cleanup; + } + r = amdgpu_ttm_alloc_gart(&sobj->tbo); amdgpu_bo_unreserve(sobj); if (r) { goto out_cleanup; } + saddr = amdgpu_bo_gpu_offset(sobj); bp.domain = ddomain; r = amdgpu_bo_create(adev, &bp, &dobj); if (r) { @@ -108,11 +114,17 @@ static void amdgpu_benchmark_move(struct amdgpu_device *adev, unsigned size, r = amdgpu_bo_reserve(dobj, false); if (unlikely(r != 0)) goto out_cleanup; - r = amdgpu_bo_pin(dobj, ddomain, &daddr); + r = amdgpu_bo_pin(dobj, ddomain); + if (r) { + amdgpu_bo_unreserve(sobj); + goto out_cleanup; + } + r = amdgpu_ttm_alloc_gart(&dobj->tbo); amdgpu_bo_unreserve(dobj); if (r) { goto out_cleanup; } + daddr = amdgpu_bo_gpu_offset(dobj); if (adev->mman.buffer_funcs) { time = amdgpu_benchmark_do_move(adev, size, saddr, daddr, n); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c index 92be7f6de197..d472a2c8399f 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c @@ -35,92 +35,53 @@ #define AMDGPU_BO_LIST_MAX_PRIORITY 32u #define AMDGPU_BO_LIST_NUM_BUCKETS (AMDGPU_BO_LIST_MAX_PRIORITY + 1) -static int amdgpu_bo_list_set(struct amdgpu_device *adev, - struct drm_file *filp, - struct amdgpu_bo_list *list, - struct drm_amdgpu_bo_list_entry *info, - unsigned num_entries); +static void amdgpu_bo_list_free_rcu(struct rcu_head *rcu) +{ + struct amdgpu_bo_list *list = container_of(rcu, struct amdgpu_bo_list, + rhead); + + kvfree(list); +} -static void amdgpu_bo_list_release_rcu(struct kref *ref) +static void amdgpu_bo_list_free(struct kref *ref) { - unsigned i; struct amdgpu_bo_list *list = container_of(ref, struct amdgpu_bo_list, refcount); + struct amdgpu_bo_list_entry *e; - for (i = 0; i < list->num_entries; ++i) - amdgpu_bo_unref(&list->array[i].robj); + amdgpu_bo_list_for_each_entry(e, list) + amdgpu_bo_unref(&e->robj); - mutex_destroy(&list->lock); - kvfree(list->array); - kfree_rcu(list, rhead); + call_rcu(&list->rhead, amdgpu_bo_list_free_rcu); } -static int amdgpu_bo_list_create(struct amdgpu_device *adev, - struct drm_file *filp, - struct drm_amdgpu_bo_list_entry *info, - unsigned num_entries, - int *id) +int amdgpu_bo_list_create(struct amdgpu_device *adev, struct drm_file *filp, + struct drm_amdgpu_bo_list_entry *info, + unsigned num_entries, struct amdgpu_bo_list **result) { - int r; - struct amdgpu_fpriv *fpriv = filp->driver_priv; + unsigned last_entry = 0, first_userptr = num_entries; + struct amdgpu_bo_list_entry *array; struct amdgpu_bo_list *list; + uint64_t total_size = 0; + size_t size; + unsigned i; + int r; + + if (num_entries > SIZE_MAX / sizeof(struct amdgpu_bo_list_entry)) + return -EINVAL; - list = kzalloc(sizeof(struct amdgpu_bo_list), GFP_KERNEL); + size = sizeof(struct amdgpu_bo_list); + size += num_entries * sizeof(struct amdgpu_bo_list_entry); + list = kvmalloc(size, GFP_KERNEL); if (!list) return -ENOMEM; - /* initialize bo list*/ - mutex_init(&list->lock); kref_init(&list->refcount); - r = amdgpu_bo_list_set(adev, filp, list, info, num_entries); - if (r) { - kfree(list); - return r; - } - - /* idr alloc should be called only after initialization of bo list. */ - mutex_lock(&fpriv->bo_list_lock); - r = idr_alloc(&fpriv->bo_list_handles, list, 1, 0, GFP_KERNEL); - mutex_unlock(&fpriv->bo_list_lock); - if (r < 0) { - amdgpu_bo_list_free(list); - return r; - } - *id = r; - - return 0; -} - -static void amdgpu_bo_list_destroy(struct amdgpu_fpriv *fpriv, int id) -{ - struct amdgpu_bo_list *list; - - mutex_lock(&fpriv->bo_list_lock); - list = idr_remove(&fpriv->bo_list_handles, id); - mutex_unlock(&fpriv->bo_list_lock); - if (list) - kref_put(&list->refcount, amdgpu_bo_list_release_rcu); -} - -static int amdgpu_bo_list_set(struct amdgpu_device *adev, - struct drm_file *filp, - struct amdgpu_bo_list *list, - struct drm_amdgpu_bo_list_entry *info, - unsigned num_entries) -{ - struct amdgpu_bo_list_entry *array; - struct amdgpu_bo *gds_obj = adev->gds.gds_gfx_bo; - struct amdgpu_bo *gws_obj = adev->gds.gws_gfx_bo; - struct amdgpu_bo *oa_obj = adev->gds.oa_gfx_bo; - - unsigned last_entry = 0, first_userptr = num_entries; - unsigned i; - int r; - unsigned long total_size = 0; + list->gds_obj = adev->gds.gds_gfx_bo; + list->gws_obj = adev->gds.gws_gfx_bo; + list->oa_obj = adev->gds.oa_gfx_bo; - array = kvmalloc_array(num_entries, sizeof(struct amdgpu_bo_list_entry), GFP_KERNEL); - if (!array) - return -ENOMEM; + array = amdgpu_bo_list_array_entry(list, 0); memset(array, 0, num_entries * sizeof(struct amdgpu_bo_list_entry)); for (i = 0; i < num_entries; ++i) { @@ -157,59 +118,56 @@ static int amdgpu_bo_list_set(struct amdgpu_device *adev, entry->tv.shared = !entry->robj->prime_shared_count; if (entry->robj->preferred_domains == AMDGPU_GEM_DOMAIN_GDS) - gds_obj = entry->robj; + list->gds_obj = entry->robj; if (entry->robj->preferred_domains == AMDGPU_GEM_DOMAIN_GWS) - gws_obj = entry->robj; + list->gws_obj = entry->robj; if (entry->robj->preferred_domains == AMDGPU_GEM_DOMAIN_OA) - oa_obj = entry->robj; + list->oa_obj = entry->robj; total_size += amdgpu_bo_size(entry->robj); trace_amdgpu_bo_list_set(list, entry->robj); } - for (i = 0; i < list->num_entries; ++i) - amdgpu_bo_unref(&list->array[i].robj); - - kvfree(list->array); - - list->gds_obj = gds_obj; - list->gws_obj = gws_obj; - list->oa_obj = oa_obj; list->first_userptr = first_userptr; - list->array = array; list->num_entries = num_entries; trace_amdgpu_cs_bo_status(list->num_entries, total_size); + + *result = list; return 0; error_free: while (i--) amdgpu_bo_unref(&array[i].robj); - kvfree(array); + kvfree(list); return r; + } -struct amdgpu_bo_list * -amdgpu_bo_list_get(struct amdgpu_fpriv *fpriv, int id) +static void amdgpu_bo_list_destroy(struct amdgpu_fpriv *fpriv, int id) { - struct amdgpu_bo_list *result; + struct amdgpu_bo_list *list; + + mutex_lock(&fpriv->bo_list_lock); + list = idr_remove(&fpriv->bo_list_handles, id); + mutex_unlock(&fpriv->bo_list_lock); + if (list) + kref_put(&list->refcount, amdgpu_bo_list_free); +} +int amdgpu_bo_list_get(struct amdgpu_fpriv *fpriv, int id, + struct amdgpu_bo_list **result) +{ rcu_read_lock(); - result = idr_find(&fpriv->bo_list_handles, id); + *result = idr_find(&fpriv->bo_list_handles, id); - if (result) { - if (kref_get_unless_zero(&result->refcount)) { - rcu_read_unlock(); - mutex_lock(&result->lock); - } else { - rcu_read_unlock(); - result = NULL; - } - } else { + if (*result && kref_get_unless_zero(&(*result)->refcount)) { rcu_read_unlock(); + return 0; } - return result; + rcu_read_unlock(); + return -ENOENT; } void amdgpu_bo_list_get_list(struct amdgpu_bo_list *list, @@ -220,6 +178,7 @@ void amdgpu_bo_list_get_list(struct amdgpu_bo_list *list, * concatenated in descending order. */ struct list_head bucket[AMDGPU_BO_LIST_NUM_BUCKETS]; + struct amdgpu_bo_list_entry *e; unsigned i; for (i = 0; i < AMDGPU_BO_LIST_NUM_BUCKETS; i++) @@ -230,14 +189,13 @@ void amdgpu_bo_list_get_list(struct amdgpu_bo_list *list, * in the list, the sort mustn't change the ordering of buffers * with the same priority, i.e. it must be stable. */ - for (i = 0; i < list->num_entries; i++) { - unsigned priority = list->array[i].priority; + amdgpu_bo_list_for_each_entry(e, list) { + unsigned priority = e->priority; - if (!list->array[i].robj->parent) - list_add_tail(&list->array[i].tv.head, - &bucket[priority]); + if (!e->robj->parent) + list_add_tail(&e->tv.head, &bucket[priority]); - list->array[i].user_pages = NULL; + e->user_pages = NULL; } /* Connect the sorted buckets in the output list. */ @@ -247,71 +205,82 @@ void amdgpu_bo_list_get_list(struct amdgpu_bo_list *list, void amdgpu_bo_list_put(struct amdgpu_bo_list *list) { - mutex_unlock(&list->lock); - kref_put(&list->refcount, amdgpu_bo_list_release_rcu); -} - -void amdgpu_bo_list_free(struct amdgpu_bo_list *list) -{ - unsigned i; - - for (i = 0; i < list->num_entries; ++i) - amdgpu_bo_unref(&list->array[i].robj); - - mutex_destroy(&list->lock); - kvfree(list->array); - kfree(list); + kref_put(&list->refcount, amdgpu_bo_list_free); } -int amdgpu_bo_list_ioctl(struct drm_device *dev, void *data, - struct drm_file *filp) +int amdgpu_bo_create_list_entry_array(struct drm_amdgpu_bo_list_in *in, + struct drm_amdgpu_bo_list_entry **info_param) { + const void __user *uptr = u64_to_user_ptr(in->bo_info_ptr); const uint32_t info_size = sizeof(struct drm_amdgpu_bo_list_entry); - - struct amdgpu_device *adev = dev->dev_private; - struct amdgpu_fpriv *fpriv = filp->driver_priv; - union drm_amdgpu_bo_list *args = data; - uint32_t handle = args->in.list_handle; - const void __user *uptr = u64_to_user_ptr(args->in.bo_info_ptr); - struct drm_amdgpu_bo_list_entry *info; - struct amdgpu_bo_list *list; - int r; - info = kvmalloc_array(args->in.bo_number, - sizeof(struct drm_amdgpu_bo_list_entry), GFP_KERNEL); + info = kvmalloc_array(in->bo_number, info_size, GFP_KERNEL); if (!info) return -ENOMEM; /* copy the handle array from userspace to a kernel buffer */ r = -EFAULT; - if (likely(info_size == args->in.bo_info_size)) { - unsigned long bytes = args->in.bo_number * - args->in.bo_info_size; + if (likely(info_size == in->bo_info_size)) { + unsigned long bytes = in->bo_number * + in->bo_info_size; if (copy_from_user(info, uptr, bytes)) goto error_free; } else { - unsigned long bytes = min(args->in.bo_info_size, info_size); + unsigned long bytes = min(in->bo_info_size, info_size); unsigned i; - memset(info, 0, args->in.bo_number * info_size); - for (i = 0; i < args->in.bo_number; ++i) { + memset(info, 0, in->bo_number * info_size); + for (i = 0; i < in->bo_number; ++i) { if (copy_from_user(&info[i], uptr, bytes)) goto error_free; - uptr += args->in.bo_info_size; + uptr += in->bo_info_size; } } + *info_param = info; + return 0; + +error_free: + kvfree(info); + return r; +} + +int amdgpu_bo_list_ioctl(struct drm_device *dev, void *data, + struct drm_file *filp) +{ + struct amdgpu_device *adev = dev->dev_private; + struct amdgpu_fpriv *fpriv = filp->driver_priv; + union drm_amdgpu_bo_list *args = data; + uint32_t handle = args->in.list_handle; + struct drm_amdgpu_bo_list_entry *info = NULL; + struct amdgpu_bo_list *list, *old; + int r; + + r = amdgpu_bo_create_list_entry_array(&args->in, &info); + if (r) + goto error_free; + switch (args->in.operation) { case AMDGPU_BO_LIST_OP_CREATE: r = amdgpu_bo_list_create(adev, filp, info, args->in.bo_number, - &handle); + &list); if (r) goto error_free; + + mutex_lock(&fpriv->bo_list_lock); + r = idr_alloc(&fpriv->bo_list_handles, list, 1, 0, GFP_KERNEL); + mutex_unlock(&fpriv->bo_list_lock); + if (r < 0) { + amdgpu_bo_list_put(list); + return r; + } + + handle = r; break; case AMDGPU_BO_LIST_OP_DESTROY: @@ -320,17 +289,22 @@ int amdgpu_bo_list_ioctl(struct drm_device *dev, void *data, break; case AMDGPU_BO_LIST_OP_UPDATE: - r = -ENOENT; - list = amdgpu_bo_list_get(fpriv, handle); - if (!list) + r = amdgpu_bo_list_create(adev, filp, info, args->in.bo_number, + &list); + if (r) goto error_free; - r = amdgpu_bo_list_set(adev, filp, list, info, - args->in.bo_number); - amdgpu_bo_list_put(list); - if (r) + mutex_lock(&fpriv->bo_list_lock); + old = idr_replace(&fpriv->bo_list_handles, list, handle); + mutex_unlock(&fpriv->bo_list_lock); + + if (IS_ERR(old)) { + amdgpu_bo_list_put(list); + r = PTR_ERR(old); goto error_free; + } + amdgpu_bo_list_put(old); break; default: @@ -345,6 +319,7 @@ int amdgpu_bo_list_ioctl(struct drm_device *dev, void *data, return 0; error_free: - kvfree(info); + if (info) + kvfree(info); return r; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.h new file mode 100644 index 000000000000..61b089768e1c --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.h @@ -0,0 +1,85 @@ +/* + * Copyright 2018 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ +#ifndef __AMDGPU_BO_LIST_H__ +#define __AMDGPU_BO_LIST_H__ + +#include <drm/ttm/ttm_execbuf_util.h> +#include <drm/amdgpu_drm.h> + +struct amdgpu_device; +struct amdgpu_bo; +struct amdgpu_bo_va; +struct amdgpu_fpriv; + +struct amdgpu_bo_list_entry { + struct amdgpu_bo *robj; + struct ttm_validate_buffer tv; + struct amdgpu_bo_va *bo_va; + uint32_t priority; + struct page **user_pages; + int user_invalidated; +}; + +struct amdgpu_bo_list { + struct rcu_head rhead; + struct kref refcount; + struct amdgpu_bo *gds_obj; + struct amdgpu_bo *gws_obj; + struct amdgpu_bo *oa_obj; + unsigned first_userptr; + unsigned num_entries; +}; + +int amdgpu_bo_list_get(struct amdgpu_fpriv *fpriv, int id, + struct amdgpu_bo_list **result); +void amdgpu_bo_list_get_list(struct amdgpu_bo_list *list, + struct list_head *validated); +void amdgpu_bo_list_put(struct amdgpu_bo_list *list); +int amdgpu_bo_create_list_entry_array(struct drm_amdgpu_bo_list_in *in, + struct drm_amdgpu_bo_list_entry **info_param); + +int amdgpu_bo_list_create(struct amdgpu_device *adev, + struct drm_file *filp, + struct drm_amdgpu_bo_list_entry *info, + unsigned num_entries, + struct amdgpu_bo_list **list); + +static inline struct amdgpu_bo_list_entry * +amdgpu_bo_list_array_entry(struct amdgpu_bo_list *list, unsigned index) +{ + struct amdgpu_bo_list_entry *array = (void *)&list[1]; + + return &array[index]; +} + +#define amdgpu_bo_list_for_each_entry(e, list) \ + for (e = amdgpu_bo_list_array_entry(list, 0); \ + e != amdgpu_bo_list_array_entry(list, (list)->num_entries); \ + ++e) + +#define amdgpu_bo_list_for_each_userptr_entry(e, list) \ + for (e = amdgpu_bo_list_array_entry(list, (list)->first_userptr); \ + e != amdgpu_bo_list_array_entry(list, (list)->num_entries); \ + ++e) + +#endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c index e950730f1933..693ec5ea4950 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c @@ -314,17 +314,17 @@ static int amdgpu_cgs_get_firmware_info(struct cgs_device *cgs_device, (adev->pdev->revision == 0x81) || (adev->pdev->device == 0x665f)) { info->is_kicker = true; - strcpy(fw_name, "radeon/bonaire_k_smc.bin"); + strcpy(fw_name, "amdgpu/bonaire_k_smc.bin"); } else { - strcpy(fw_name, "radeon/bonaire_smc.bin"); + strcpy(fw_name, "amdgpu/bonaire_smc.bin"); } break; case CHIP_HAWAII: if (adev->pdev->revision == 0x80) { info->is_kicker = true; - strcpy(fw_name, "radeon/hawaii_k_smc.bin"); + strcpy(fw_name, "amdgpu/hawaii_k_smc.bin"); } else { - strcpy(fw_name, "radeon/hawaii_smc.bin"); + strcpy(fw_name, "amdgpu/hawaii_smc.bin"); } break; case CHIP_TOPAZ: diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c index 8e66851eb427..c770d73352a7 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c @@ -212,30 +212,21 @@ static void amdgpu_connector_update_scratch_regs(struct drm_connector *connector, enum drm_connector_status status) { - struct drm_encoder *best_encoder = NULL; - struct drm_encoder *encoder = NULL; + struct drm_encoder *best_encoder; + struct drm_encoder *encoder; const struct drm_connector_helper_funcs *connector_funcs = connector->helper_private; bool connected; int i; best_encoder = connector_funcs->best_encoder(connector); - for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) { - if (connector->encoder_ids[i] == 0) - break; - - encoder = drm_encoder_find(connector->dev, NULL, - connector->encoder_ids[i]); - if (!encoder) - continue; - + drm_connector_for_each_possible_encoder(connector, encoder, i) { if ((encoder == best_encoder) && (status == connector_status_connected)) connected = true; else connected = false; amdgpu_atombios_encoder_set_bios_scratch_regs(connector, encoder, connected); - } } @@ -246,17 +237,11 @@ amdgpu_connector_find_encoder(struct drm_connector *connector, struct drm_encoder *encoder; int i; - for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) { - if (connector->encoder_ids[i] == 0) - break; - encoder = drm_encoder_find(connector->dev, NULL, - connector->encoder_ids[i]); - if (!encoder) - continue; - + drm_connector_for_each_possible_encoder(connector, encoder, i) { if (encoder->encoder_type == encoder_type) return encoder; } + return NULL; } @@ -349,22 +334,24 @@ static int amdgpu_connector_ddc_get_modes(struct drm_connector *connector) int ret; if (amdgpu_connector->edid) { - drm_mode_connector_update_edid_property(connector, amdgpu_connector->edid); + drm_connector_update_edid_property(connector, amdgpu_connector->edid); ret = drm_add_edid_modes(connector, amdgpu_connector->edid); return ret; } - drm_mode_connector_update_edid_property(connector, NULL); + drm_connector_update_edid_property(connector, NULL); return 0; } static struct drm_encoder * amdgpu_connector_best_single_encoder(struct drm_connector *connector) { - int enc_id = connector->encoder_ids[0]; + struct drm_encoder *encoder; + int i; + + /* pick the first one */ + drm_connector_for_each_possible_encoder(connector, encoder, i) + return encoder; - /* pick the encoder ids */ - if (enc_id) - return drm_encoder_find(connector->dev, NULL, enc_id); return NULL; } @@ -985,9 +972,8 @@ amdgpu_connector_dvi_detect(struct drm_connector *connector, bool force) struct drm_device *dev = connector->dev; struct amdgpu_device *adev = dev->dev_private; struct amdgpu_connector *amdgpu_connector = to_amdgpu_connector(connector); - struct drm_encoder *encoder = NULL; const struct drm_encoder_helper_funcs *encoder_funcs; - int i, r; + int r; enum drm_connector_status ret = connector_status_disconnected; bool dret = false, broken_edid = false; @@ -1077,14 +1063,10 @@ amdgpu_connector_dvi_detect(struct drm_connector *connector, bool force) /* find analog encoder */ if (amdgpu_connector->dac_load_detect) { - for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) { - if (connector->encoder_ids[i] == 0) - break; - - encoder = drm_encoder_find(connector->dev, NULL, connector->encoder_ids[i]); - if (!encoder) - continue; + struct drm_encoder *encoder; + int i; + drm_connector_for_each_possible_encoder(connector, encoder, i) { if (encoder->encoder_type != DRM_MODE_ENCODER_DAC && encoder->encoder_type != DRM_MODE_ENCODER_TVDAC) continue; @@ -1132,18 +1114,11 @@ exit: static struct drm_encoder * amdgpu_connector_dvi_encoder(struct drm_connector *connector) { - int enc_id = connector->encoder_ids[0]; struct amdgpu_connector *amdgpu_connector = to_amdgpu_connector(connector); struct drm_encoder *encoder; int i; - for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) { - if (connector->encoder_ids[i] == 0) - break; - - encoder = drm_encoder_find(connector->dev, NULL, connector->encoder_ids[i]); - if (!encoder) - continue; + drm_connector_for_each_possible_encoder(connector, encoder, i) { if (amdgpu_connector->use_digital == true) { if (encoder->encoder_type == DRM_MODE_ENCODER_TMDS) return encoder; @@ -1158,8 +1133,9 @@ amdgpu_connector_dvi_encoder(struct drm_connector *connector) /* then check use digitial */ /* pick the first one */ - if (enc_id) - return drm_encoder_find(connector->dev, NULL, enc_id); + drm_connector_for_each_possible_encoder(connector, encoder, i) + return encoder; + return NULL; } @@ -1296,15 +1272,7 @@ u16 amdgpu_connector_encoder_get_dp_bridge_encoder_id(struct drm_connector *conn struct amdgpu_encoder *amdgpu_encoder; int i; - for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) { - if (connector->encoder_ids[i] == 0) - break; - - encoder = drm_encoder_find(connector->dev, NULL, - connector->encoder_ids[i]); - if (!encoder) - continue; - + drm_connector_for_each_possible_encoder(connector, encoder, i) { amdgpu_encoder = to_amdgpu_encoder(encoder); switch (amdgpu_encoder->encoder_id) { @@ -1326,14 +1294,7 @@ static bool amdgpu_connector_encoder_is_hbr2(struct drm_connector *connector) int i; bool found = false; - for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) { - if (connector->encoder_ids[i] == 0) - break; - encoder = drm_encoder_find(connector->dev, NULL, - connector->encoder_ids[i]); - if (!encoder) - continue; - + drm_connector_for_each_possible_encoder(connector, encoder, i) { amdgpu_encoder = to_amdgpu_encoder(encoder); if (amdgpu_encoder->caps & ATOM_ENCODER_CAP_RECORD_HBR2) found = true; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c index 9c85a90be293..502b94fb116a 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c @@ -31,6 +31,7 @@ #include <drm/drm_syncobj.h> #include "amdgpu.h" #include "amdgpu_trace.h" +#include "amdgpu_gmc.h" static int amdgpu_cs_user_fence_chunk(struct amdgpu_cs_parser *p, struct drm_amdgpu_cs_chunk_fence *data, @@ -65,11 +66,35 @@ static int amdgpu_cs_user_fence_chunk(struct amdgpu_cs_parser *p, return 0; } -static int amdgpu_cs_parser_init(struct amdgpu_cs_parser *p, void *data) +static int amdgpu_cs_bo_handles_chunk(struct amdgpu_cs_parser *p, + struct drm_amdgpu_bo_list_in *data) +{ + int r; + struct drm_amdgpu_bo_list_entry *info = NULL; + + r = amdgpu_bo_create_list_entry_array(data, &info); + if (r) + return r; + + r = amdgpu_bo_list_create(p->adev, p->filp, info, data->bo_number, + &p->bo_list); + if (r) + goto error_free; + + kvfree(info); + return 0; + +error_free: + if (info) + kvfree(info); + + return r; +} + +static int amdgpu_cs_parser_init(struct amdgpu_cs_parser *p, union drm_amdgpu_cs *cs) { struct amdgpu_fpriv *fpriv = p->filp->driver_priv; struct amdgpu_vm *vm = &fpriv->vm; - union drm_amdgpu_cs *cs = data; uint64_t *chunk_array_user; uint64_t *chunk_array; unsigned size, num_ibs = 0; @@ -163,6 +188,19 @@ static int amdgpu_cs_parser_init(struct amdgpu_cs_parser *p, void *data) break; + case AMDGPU_CHUNK_ID_BO_HANDLES: + size = sizeof(struct drm_amdgpu_bo_list_in); + if (p->chunks[i].length_dw * sizeof(uint32_t) < size) { + ret = -EINVAL; + goto free_partial_kdata; + } + + ret = amdgpu_cs_bo_handles_chunk(p, p->chunks[i].kdata); + if (ret) + goto free_partial_kdata; + + break; + case AMDGPU_CHUNK_ID_DEPENDENCIES: case AMDGPU_CHUNK_ID_SYNCOBJ_IN: case AMDGPU_CHUNK_ID_SYNCOBJ_OUT: @@ -186,6 +224,10 @@ static int amdgpu_cs_parser_init(struct amdgpu_cs_parser *p, void *data) if (p->uf_entry.robj) p->job->uf_addr = uf_offset; kfree(chunk_array); + + /* Use this opportunity to fill in task info for the vm */ + amdgpu_vm_set_task_info(vm); + return 0; free_all_kdata: @@ -257,7 +299,7 @@ static void amdgpu_cs_get_threshold_for_moves(struct amdgpu_device *adev, return; } - total_vram = adev->gmc.real_vram_size - adev->vram_pin_size; + total_vram = adev->gmc.real_vram_size - atomic64_read(&adev->vram_pin_size); used_vram = amdgpu_vram_mgr_usage(&adev->mman.bdev.man[TTM_PL_VRAM]); free_vram = used_vram >= total_vram ? 0 : total_vram - used_vram; @@ -302,7 +344,7 @@ static void amdgpu_cs_get_threshold_for_moves(struct amdgpu_device *adev, *max_bytes = us_to_bytes(adev, adev->mm_stats.accum_us); /* Do the same for visible VRAM if half of it is free */ - if (adev->gmc.visible_vram_size < adev->gmc.real_vram_size) { + if (!amdgpu_gmc_vram_full_visible(&adev->gmc)) { u64 total_vis_vram = adev->gmc.visible_vram_size; u64 used_vis_vram = amdgpu_vram_mgr_vis_usage(&adev->mman.bdev.man[TTM_PL_VRAM]); @@ -359,7 +401,7 @@ static int amdgpu_cs_bo_validate(struct amdgpu_cs_parser *p, * to move it. Don't move anything if the threshold is zero. */ if (p->bytes_moved < p->bytes_moved_threshold) { - if (adev->gmc.visible_vram_size < adev->gmc.real_vram_size && + if (!amdgpu_gmc_vram_full_visible(&adev->gmc) && (bo->flags & AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED)) { /* And don't move a CPU_ACCESS_REQUIRED BO to limited * visible VRAM if we've depleted our allowance to do @@ -377,11 +419,11 @@ static int amdgpu_cs_bo_validate(struct amdgpu_cs_parser *p, } retry: - amdgpu_ttm_placement_from_domain(bo, domain); + amdgpu_bo_placement_from_domain(bo, domain); r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx); p->bytes_moved += ctx.bytes_moved; - if (adev->gmc.visible_vram_size < adev->gmc.real_vram_size && + if (!amdgpu_gmc_vram_full_visible(&adev->gmc) && amdgpu_bo_in_cpu_visible_vram(bo)) p->bytes_moved_vis += ctx.bytes_moved; @@ -434,9 +476,9 @@ static bool amdgpu_cs_try_evict(struct amdgpu_cs_parser *p, /* Good we can try to move this BO somewhere else */ update_bytes_moved_vis = - adev->gmc.visible_vram_size < adev->gmc.real_vram_size && - amdgpu_bo_in_cpu_visible_vram(bo); - amdgpu_ttm_placement_from_domain(bo, other); + !amdgpu_gmc_vram_full_visible(&adev->gmc) && + amdgpu_bo_in_cpu_visible_vram(bo); + amdgpu_bo_placement_from_domain(bo, other); r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx); p->bytes_moved += ctx.bytes_moved; if (update_bytes_moved_vis) @@ -490,8 +532,8 @@ static int amdgpu_cs_list_validate(struct amdgpu_cs_parser *p, /* Check if we have user pages and nobody bound the BO already */ if (amdgpu_ttm_tt_userptr_needs_pages(bo->tbo.ttm) && lobj->user_pages) { - amdgpu_ttm_placement_from_domain(bo, - AMDGPU_GEM_DOMAIN_CPU); + amdgpu_bo_placement_from_domain(bo, + AMDGPU_GEM_DOMAIN_CPU); r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx); if (r) return r; @@ -519,23 +561,38 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p, union drm_amdgpu_cs *cs) { struct amdgpu_fpriv *fpriv = p->filp->driver_priv; + struct amdgpu_vm *vm = &fpriv->vm; struct amdgpu_bo_list_entry *e; struct list_head duplicates; - unsigned i, tries = 10; struct amdgpu_bo *gds; struct amdgpu_bo *gws; struct amdgpu_bo *oa; + unsigned tries = 10; int r; INIT_LIST_HEAD(&p->validated); - p->bo_list = amdgpu_bo_list_get(fpriv, cs->in.bo_list_handle); - if (p->bo_list) { - amdgpu_bo_list_get_list(p->bo_list, &p->validated); - if (p->bo_list->first_userptr != p->bo_list->num_entries) - p->mn = amdgpu_mn_get(p->adev, AMDGPU_MN_TYPE_GFX); + /* p->bo_list could already be assigned if AMDGPU_CHUNK_ID_BO_HANDLES is present */ + if (cs->in.bo_list_handle) { + if (p->bo_list) + return -EINVAL; + + r = amdgpu_bo_list_get(fpriv, cs->in.bo_list_handle, + &p->bo_list); + if (r) + return r; + } else if (!p->bo_list) { + /* Create a empty bo_list when no handle is provided */ + r = amdgpu_bo_list_create(p->adev, p->filp, NULL, 0, + &p->bo_list); + if (r) + return r; } + amdgpu_bo_list_get_list(p->bo_list, &p->validated); + if (p->bo_list->first_userptr != p->bo_list->num_entries) + p->mn = amdgpu_mn_get(p->adev, AMDGPU_MN_TYPE_GFX); + INIT_LIST_HEAD(&duplicates); amdgpu_vm_get_pd_bo(&fpriv->vm, &p->validated, &p->vm_pd); @@ -544,7 +601,6 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p, while (1) { struct list_head need_pages; - unsigned i; r = ttm_eu_reserve_buffers(&p->ticket, &p->validated, true, &duplicates); @@ -554,17 +610,9 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p, goto error_free_pages; } - /* Without a BO list we don't have userptr BOs */ - if (!p->bo_list) - break; - INIT_LIST_HEAD(&need_pages); - for (i = p->bo_list->first_userptr; - i < p->bo_list->num_entries; ++i) { - struct amdgpu_bo *bo; - - e = &p->bo_list->array[i]; - bo = e->robj; + amdgpu_bo_list_for_each_userptr_entry(e, p->bo_list) { + struct amdgpu_bo *bo = e->robj; if (amdgpu_ttm_tt_userptr_invalidated(bo->tbo.ttm, &e->user_invalidated) && e->user_pages) { @@ -656,23 +704,12 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p, amdgpu_cs_report_moved_bytes(p->adev, p->bytes_moved, p->bytes_moved_vis); - if (p->bo_list) { - struct amdgpu_vm *vm = &fpriv->vm; - unsigned i; + gds = p->bo_list->gds_obj; + gws = p->bo_list->gws_obj; + oa = p->bo_list->oa_obj; - gds = p->bo_list->gds_obj; - gws = p->bo_list->gws_obj; - oa = p->bo_list->oa_obj; - for (i = 0; i < p->bo_list->num_entries; i++) { - struct amdgpu_bo *bo = p->bo_list->array[i].robj; - - p->bo_list->array[i].bo_va = amdgpu_vm_bo_find(vm, bo); - } - } else { - gds = p->adev->gds.gds_gfx_bo; - gws = p->adev->gds.gws_gfx_bo; - oa = p->adev->gds.oa_gfx_bo; - } + amdgpu_bo_list_for_each_entry(e, p->bo_list) + e->bo_va = amdgpu_vm_bo_find(vm, e->robj); if (gds) { p->job->gds_base = amdgpu_bo_gpu_offset(gds); @@ -700,18 +737,13 @@ error_validate: error_free_pages: - if (p->bo_list) { - for (i = p->bo_list->first_userptr; - i < p->bo_list->num_entries; ++i) { - e = &p->bo_list->array[i]; - - if (!e->user_pages) - continue; + amdgpu_bo_list_for_each_userptr_entry(e, p->bo_list) { + if (!e->user_pages) + continue; - release_pages(e->user_pages, - e->robj->tbo.ttm->num_pages); - kvfree(e->user_pages); - } + release_pages(e->user_pages, + e->robj->tbo.ttm->num_pages); + kvfree(e->user_pages); } return r; @@ -773,12 +805,13 @@ static void amdgpu_cs_parser_fini(struct amdgpu_cs_parser *parser, int error, static int amdgpu_bo_vm_update_pte(struct amdgpu_cs_parser *p) { - struct amdgpu_device *adev = p->adev; struct amdgpu_fpriv *fpriv = p->filp->driver_priv; + struct amdgpu_device *adev = p->adev; struct amdgpu_vm *vm = &fpriv->vm; + struct amdgpu_bo_list_entry *e; struct amdgpu_bo_va *bo_va; struct amdgpu_bo *bo; - int i, r; + int r; r = amdgpu_vm_clear_freed(adev, vm, NULL); if (r) @@ -808,29 +841,26 @@ static int amdgpu_bo_vm_update_pte(struct amdgpu_cs_parser *p) return r; } - if (p->bo_list) { - for (i = 0; i < p->bo_list->num_entries; i++) { - struct dma_fence *f; - - /* ignore duplicates */ - bo = p->bo_list->array[i].robj; - if (!bo) - continue; + amdgpu_bo_list_for_each_entry(e, p->bo_list) { + struct dma_fence *f; - bo_va = p->bo_list->array[i].bo_va; - if (bo_va == NULL) - continue; + /* ignore duplicates */ + bo = e->robj; + if (!bo) + continue; - r = amdgpu_vm_bo_update(adev, bo_va, false); - if (r) - return r; + bo_va = e->bo_va; + if (bo_va == NULL) + continue; - f = bo_va->last_pt_update; - r = amdgpu_sync_fence(adev, &p->job->sync, f, false); - if (r) - return r; - } + r = amdgpu_vm_bo_update(adev, bo_va, false); + if (r) + return r; + f = bo_va->last_pt_update; + r = amdgpu_sync_fence(adev, &p->job->sync, f, false); + if (r) + return r; } r = amdgpu_vm_handle_moved(adev, vm); @@ -845,15 +875,14 @@ static int amdgpu_bo_vm_update_pte(struct amdgpu_cs_parser *p) if (r) return r; - if (amdgpu_vm_debug && p->bo_list) { + if (amdgpu_vm_debug) { /* Invalidate all BOs to test for userspace bugs */ - for (i = 0; i < p->bo_list->num_entries; i++) { + amdgpu_bo_list_for_each_entry(e, p->bo_list) { /* ignore duplicates */ - bo = p->bo_list->array[i].robj; - if (!bo) + if (!e->robj) continue; - amdgpu_vm_bo_invalidate(adev, bo, false); + amdgpu_vm_bo_invalidate(adev, e->robj, false); } } @@ -865,11 +894,11 @@ static int amdgpu_cs_ib_vm_chunk(struct amdgpu_device *adev, { struct amdgpu_fpriv *fpriv = p->filp->driver_priv; struct amdgpu_vm *vm = &fpriv->vm; - struct amdgpu_ring *ring = p->job->ring; + struct amdgpu_ring *ring = p->ring; int r; /* Only for UVD/VCE VM emulation */ - if (p->job->ring->funcs->parse_cs) { + if (p->ring->funcs->parse_cs || p->ring->funcs->patch_cs_in_place) { unsigned i, j; for (i = 0, j = 0; i < p->nchunks && j < p->job->num_ibs; i++) { @@ -910,12 +939,20 @@ static int amdgpu_cs_ib_vm_chunk(struct amdgpu_device *adev, offset = m->start * AMDGPU_GPU_PAGE_SIZE; kptr += va_start - offset; - memcpy(ib->ptr, kptr, chunk_ib->ib_bytes); - amdgpu_bo_kunmap(aobj); - - r = amdgpu_ring_parse_cs(ring, p, j); - if (r) - return r; + if (p->ring->funcs->parse_cs) { + memcpy(ib->ptr, kptr, chunk_ib->ib_bytes); + amdgpu_bo_kunmap(aobj); + + r = amdgpu_ring_parse_cs(ring, p, j); + if (r) + return r; + } else { + ib->ptr = (uint32_t *)kptr; + r = amdgpu_ring_patch_cs_in_place(ring, p, j); + amdgpu_bo_kunmap(aobj); + if (r) + return r; + } j++; } @@ -983,10 +1020,10 @@ static int amdgpu_cs_ib_fill(struct amdgpu_device *adev, } } - if (parser->job->ring && parser->job->ring != ring) + if (parser->ring && parser->ring != ring) return -EINVAL; - parser->job->ring = ring; + parser->ring = ring; r = amdgpu_ib_get(adev, vm, ring->funcs->parse_cs ? chunk_ib->ib_bytes : 0, @@ -1005,11 +1042,11 @@ static int amdgpu_cs_ib_fill(struct amdgpu_device *adev, /* UVD & VCE fw doesn't support user fences */ if (parser->job->uf_addr && ( - parser->job->ring->funcs->type == AMDGPU_RING_TYPE_UVD || - parser->job->ring->funcs->type == AMDGPU_RING_TYPE_VCE)) + parser->ring->funcs->type == AMDGPU_RING_TYPE_UVD || + parser->ring->funcs->type == AMDGPU_RING_TYPE_VCE)) return -EINVAL; - return amdgpu_ctx_wait_prev_fence(parser->ctx, parser->job->ring->idx); + return amdgpu_ctx_wait_prev_fence(parser->ctx, parser->ring->idx); } static int amdgpu_cs_process_fence_dep(struct amdgpu_cs_parser *p, @@ -1160,31 +1197,30 @@ static void amdgpu_cs_post_dependencies(struct amdgpu_cs_parser *p) static int amdgpu_cs_submit(struct amdgpu_cs_parser *p, union drm_amdgpu_cs *cs) { - struct amdgpu_ring *ring = p->job->ring; + struct amdgpu_fpriv *fpriv = p->filp->driver_priv; + struct amdgpu_ring *ring = p->ring; struct drm_sched_entity *entity = &p->ctx->rings[ring->idx].entity; + enum drm_sched_priority priority; + struct amdgpu_bo_list_entry *e; struct amdgpu_job *job; - unsigned i; uint64_t seq; int r; amdgpu_mn_lock(p->mn); - if (p->bo_list) { - for (i = p->bo_list->first_userptr; - i < p->bo_list->num_entries; ++i) { - struct amdgpu_bo *bo = p->bo_list->array[i].robj; - - if (amdgpu_ttm_tt_userptr_needs_pages(bo->tbo.ttm)) { - amdgpu_mn_unlock(p->mn); - return -ERESTARTSYS; - } + amdgpu_bo_list_for_each_userptr_entry(e, p->bo_list) { + struct amdgpu_bo *bo = e->robj; + + if (amdgpu_ttm_tt_userptr_needs_pages(bo->tbo.ttm)) { + amdgpu_mn_unlock(p->mn); + return -ERESTARTSYS; } } job = p->job; p->job = NULL; - r = drm_sched_job_init(&job->base, &ring->sched, entity, p->filp); + r = drm_sched_job_init(&job->base, entity, p->filp); if (r) { amdgpu_job_free(job); amdgpu_mn_unlock(p->mn); @@ -1192,7 +1228,6 @@ static int amdgpu_cs_submit(struct amdgpu_cs_parser *p, } job->owner = p->filp; - job->fence_ctx = entity->fence_context; p->fence = dma_fence_get(&job->base.s_fence->finished); r = amdgpu_ctx_add_fence(p->ctx, ring, p->fence, &seq); @@ -1210,11 +1245,15 @@ static int amdgpu_cs_submit(struct amdgpu_cs_parser *p, job->uf_sequence = seq; amdgpu_job_free_resources(job); - amdgpu_ring_priority_get(job->ring, job->base.s_priority); trace_amdgpu_cs_ioctl(job); + amdgpu_vm_bo_trace_cs(&fpriv->vm, &p->ticket); + priority = job->base.s_priority; drm_sched_entity_push_job(&job->base, entity); + ring = to_amdgpu_ring(entity->rq->sched); + amdgpu_ring_priority_get(ring, priority); + ttm_eu_fence_buffer_objects(&p->ticket, &p->validated, p->fence); amdgpu_mn_unlock(p->mn); @@ -1605,7 +1644,7 @@ int amdgpu_cs_find_mapping(struct amdgpu_cs_parser *parser, if (!((*bo)->flags & AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS)) { (*bo)->flags |= AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS; - amdgpu_ttm_placement_from_domain(*bo, (*bo)->allowed_domains); + amdgpu_bo_placement_from_domain(*bo, (*bo)->allowed_domains); r = ttm_bo_validate(&(*bo)->tbo, &(*bo)->placement, &ctx); if (r) return r; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c index c5bb36275e93..df6965761046 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c @@ -90,8 +90,8 @@ static int amdgpu_ctx_init(struct amdgpu_device *adev, if (ring == &adev->gfx.kiq.ring) continue; - r = drm_sched_entity_init(&ring->sched, &ctx->rings[i].entity, - rq, &ctx->guilty); + r = drm_sched_entity_init(&ctx->rings[i].entity, + &rq, 1, &ctx->guilty); if (r) goto failed; } @@ -104,8 +104,7 @@ static int amdgpu_ctx_init(struct amdgpu_device *adev, failed: for (j = 0; j < i; j++) - drm_sched_entity_fini(&adev->rings[j]->sched, - &ctx->rings[j].entity); + drm_sched_entity_destroy(&ctx->rings[j].entity); kfree(ctx->fences); ctx->fences = NULL; return r; @@ -178,8 +177,7 @@ static void amdgpu_ctx_do_release(struct kref *ref) if (ctx->adev->rings[i] == &ctx->adev->gfx.kiq.ring) continue; - drm_sched_entity_fini(&ctx->adev->rings[i]->sched, - &ctx->rings[i].entity); + drm_sched_entity_destroy(&ctx->rings[i].entity); } amdgpu_ctx_fini(ref); @@ -444,34 +442,36 @@ void amdgpu_ctx_mgr_init(struct amdgpu_ctx_mgr *mgr) idr_init(&mgr->ctx_handles); } -void amdgpu_ctx_mgr_entity_fini(struct amdgpu_ctx_mgr *mgr) +void amdgpu_ctx_mgr_entity_flush(struct amdgpu_ctx_mgr *mgr) { struct amdgpu_ctx *ctx; struct idr *idp; uint32_t id, i; + long max_wait = MAX_WAIT_SCHED_ENTITY_Q_EMPTY; idp = &mgr->ctx_handles; + mutex_lock(&mgr->lock); idr_for_each_entry(idp, ctx, id) { - if (!ctx->adev) + if (!ctx->adev) { + mutex_unlock(&mgr->lock); return; + } for (i = 0; i < ctx->adev->num_rings; i++) { if (ctx->adev->rings[i] == &ctx->adev->gfx.kiq.ring) continue; - if (kref_read(&ctx->refcount) == 1) - drm_sched_entity_do_release(&ctx->adev->rings[i]->sched, - &ctx->rings[i].entity); - else - DRM_ERROR("ctx %p is still alive\n", ctx); + max_wait = drm_sched_entity_flush(&ctx->rings[i].entity, + max_wait); } } + mutex_unlock(&mgr->lock); } -void amdgpu_ctx_mgr_entity_cleanup(struct amdgpu_ctx_mgr *mgr) +void amdgpu_ctx_mgr_entity_fini(struct amdgpu_ctx_mgr *mgr) { struct amdgpu_ctx *ctx; struct idr *idp; @@ -490,8 +490,7 @@ void amdgpu_ctx_mgr_entity_cleanup(struct amdgpu_ctx_mgr *mgr) continue; if (kref_read(&ctx->refcount) == 1) - drm_sched_entity_cleanup(&ctx->adev->rings[i]->sched, - &ctx->rings[i].entity); + drm_sched_entity_fini(&ctx->rings[i].entity); else DRM_ERROR("ctx %p is still alive\n", ctx); } @@ -504,7 +503,7 @@ void amdgpu_ctx_mgr_fini(struct amdgpu_ctx_mgr *mgr) struct idr *idp; uint32_t id; - amdgpu_ctx_mgr_entity_cleanup(mgr); + amdgpu_ctx_mgr_entity_fini(mgr); idp = &mgr->ctx_handles; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index 2c5f093e79e3..1e66dfd0e39c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -25,6 +25,7 @@ * Alex Deucher * Jerome Glisse */ +#include <linux/power_supply.h> #include <linux/kthread.h> #include <linux/console.h> #include <linux/slab.h> @@ -675,17 +676,15 @@ void amdgpu_device_vram_location(struct amdgpu_device *adev, } /** - * amdgpu_device_gart_location - try to find GTT location + * amdgpu_device_gart_location - try to find GART location * * @adev: amdgpu device structure holding all necessary informations * @mc: memory controller structure holding memory informations * - * Function will place try to place GTT before or after VRAM. + * Function will place try to place GART before or after VRAM. * - * If GTT size is bigger than space left then we ajust GTT size. + * If GART size is bigger than space left then we ajust GART size. * Thus function will never fails. - * - * FIXME: when reducing GTT size align new size on power of 2. */ void amdgpu_device_gart_location(struct amdgpu_device *adev, struct amdgpu_gmc *mc) @@ -698,13 +697,13 @@ void amdgpu_device_gart_location(struct amdgpu_device *adev, size_bf = mc->vram_start; if (size_bf > size_af) { if (mc->gart_size > size_bf) { - dev_warn(adev->dev, "limiting GTT\n"); + dev_warn(adev->dev, "limiting GART\n"); mc->gart_size = size_bf; } mc->gart_start = 0; } else { if (mc->gart_size > size_af) { - dev_warn(adev->dev, "limiting GTT\n"); + dev_warn(adev->dev, "limiting GART\n"); mc->gart_size = size_af; } /* VCE doesn't like it when BOs cross a 4GB segment, so align @@ -713,7 +712,7 @@ void amdgpu_device_gart_location(struct amdgpu_device *adev, mc->gart_start = ALIGN(mc->vram_end + 1, 0x100000000ULL); } mc->gart_end = mc->gart_start + mc->gart_size - 1; - dev_info(adev->dev, "GTT: %lluM 0x%016llX - 0x%016llX\n", + dev_info(adev->dev, "GART: %lluM 0x%016llX - 0x%016llX\n", mc->gart_size >> 20, mc->gart_start, mc->gart_end); } @@ -1077,7 +1076,7 @@ static const struct vga_switcheroo_client_ops amdgpu_switcheroo_ops = { /** * amdgpu_device_ip_set_clockgating_state - set the CG state * - * @adev: amdgpu_device pointer + * @dev: amdgpu_device pointer * @block_type: Type of hardware IP (SMU, GFX, UVD, etc.) * @state: clockgating state (gate or ungate) * @@ -1111,7 +1110,7 @@ int amdgpu_device_ip_set_clockgating_state(void *dev, /** * amdgpu_device_ip_set_powergating_state - set the PG state * - * @adev: amdgpu_device pointer + * @dev: amdgpu_device pointer * @block_type: Type of hardware IP (SMU, GFX, UVD, etc.) * @state: powergating state (gate or ungate) * @@ -1222,7 +1221,7 @@ bool amdgpu_device_ip_is_idle(struct amdgpu_device *adev, * amdgpu_device_ip_get_ip_block - get a hw IP pointer * * @adev: amdgpu_device pointer - * @block_type: Type of hardware IP (SMU, GFX, UVD, etc.) + * @type: Type of hardware IP (SMU, GFX, UVD, etc.) * * Returns a pointer to the hardware IP block structure * if it exists for the asic, otherwise NULL. @@ -1708,10 +1707,6 @@ static int amdgpu_device_ip_late_set_cg_state(struct amdgpu_device *adev) if (amdgpu_emu_mode == 1) return 0; - r = amdgpu_ib_ring_tests(adev); - if (r) - DRM_ERROR("ib ring test failed (%d).\n", r); - for (i = 0; i < adev->num_ip_blocks; i++) { if (!adev->ip_blocks[i].status.valid) continue; @@ -1731,17 +1726,34 @@ static int amdgpu_device_ip_late_set_cg_state(struct amdgpu_device *adev) } } - if (adev->powerplay.pp_feature & PP_GFXOFF_MASK) { - /* enable gfx powergating */ - amdgpu_device_ip_set_powergating_state(adev, - AMD_IP_BLOCK_TYPE_GFX, - AMD_PG_STATE_GATE); - /* enable gfxoff */ - amdgpu_device_ip_set_powergating_state(adev, - AMD_IP_BLOCK_TYPE_SMC, - AMD_PG_STATE_GATE); - } + return 0; +} +static int amdgpu_device_ip_late_set_pg_state(struct amdgpu_device *adev) +{ + int i = 0, r; + + if (amdgpu_emu_mode == 1) + return 0; + + for (i = 0; i < adev->num_ip_blocks; i++) { + if (!adev->ip_blocks[i].status.valid) + continue; + /* skip CG for VCE/UVD, it's handled specially */ + if (adev->ip_blocks[i].version->type != AMD_IP_BLOCK_TYPE_UVD && + adev->ip_blocks[i].version->type != AMD_IP_BLOCK_TYPE_VCE && + adev->ip_blocks[i].version->type != AMD_IP_BLOCK_TYPE_VCN && + adev->ip_blocks[i].version->funcs->set_powergating_state) { + /* enable powergating to save power */ + r = adev->ip_blocks[i].version->funcs->set_powergating_state((void *)adev, + AMD_PG_STATE_GATE); + if (r) { + DRM_ERROR("set_powergating_state(gate) of IP block <%s> failed %d\n", + adev->ip_blocks[i].version->funcs->name, r); + return r; + } + } + } return 0; } @@ -1775,6 +1787,9 @@ static int amdgpu_device_ip_late_init(struct amdgpu_device *adev) } } + amdgpu_device_ip_late_set_cg_state(adev); + amdgpu_device_ip_late_set_pg_state(adev); + queue_delayed_work(system_wq, &adev->late_init_work, msecs_to_jiffies(AMDGPU_RESUME_MS)); @@ -1813,6 +1828,8 @@ static int amdgpu_device_ip_fini(struct amdgpu_device *adev) adev->ip_blocks[i].version->funcs->name, r); return r; } + if (adev->powerplay.pp_funcs->set_powergating_by_smu) + amdgpu_dpm_set_powergating_by_smu(adev, AMD_IP_BLOCK_TYPE_GFX, false); r = adev->ip_blocks[i].version->funcs->hw_fini((void *)adev); /* XXX handle errors */ if (r) { @@ -1901,11 +1918,15 @@ 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_device_ip_late_set_cg_state(adev); + int r; + + r = amdgpu_ib_ring_tests(adev); + if (r) + DRM_ERROR("ib ring test failed (%d).\n", r); } /** - * amdgpu_device_ip_suspend - run suspend for hardware IPs + * amdgpu_device_ip_suspend_phase1 - run suspend for hardware IPs (phase 1) * * @adev: amdgpu_device pointer * @@ -1915,18 +1936,60 @@ static void amdgpu_device_ip_late_init_func_handler(struct work_struct *work) * in each IP into a state suitable for suspend. * Returns 0 on success, negative error code on failure. */ -int amdgpu_device_ip_suspend(struct amdgpu_device *adev) +static int amdgpu_device_ip_suspend_phase1(struct amdgpu_device *adev) { int i, r; if (amdgpu_sriov_vf(adev)) amdgpu_virt_request_full_gpu(adev, false); - /* ungate SMC block powergating */ - if (adev->powerplay.pp_feature & PP_GFXOFF_MASK) - amdgpu_device_ip_set_powergating_state(adev, - AMD_IP_BLOCK_TYPE_SMC, - AMD_CG_STATE_UNGATE); + for (i = adev->num_ip_blocks - 1; i >= 0; i--) { + if (!adev->ip_blocks[i].status.valid) + continue; + /* displays are handled separately */ + if (adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_DCE) { + /* ungate blocks so that suspend can properly shut them down */ + if (adev->ip_blocks[i].version->funcs->set_clockgating_state) { + r = adev->ip_blocks[i].version->funcs->set_clockgating_state((void *)adev, + AMD_CG_STATE_UNGATE); + if (r) { + DRM_ERROR("set_clockgating_state(ungate) of IP block <%s> failed %d\n", + adev->ip_blocks[i].version->funcs->name, r); + } + } + /* XXX handle errors */ + r = adev->ip_blocks[i].version->funcs->suspend(adev); + /* XXX handle errors */ + if (r) { + DRM_ERROR("suspend of IP block <%s> failed %d\n", + adev->ip_blocks[i].version->funcs->name, r); + } + } + } + + if (amdgpu_sriov_vf(adev)) + amdgpu_virt_release_full_gpu(adev, false); + + return 0; +} + +/** + * amdgpu_device_ip_suspend_phase2 - run suspend for hardware IPs (phase 2) + * + * @adev: amdgpu_device pointer + * + * Main suspend function for hardware IPs. The list of all the hardware + * IPs that make up the asic is walked, clockgating is disabled and the + * suspend callbacks are run. suspend puts the hardware and software state + * in each IP into a state suitable for suspend. + * Returns 0 on success, negative error code on failure. + */ +static int amdgpu_device_ip_suspend_phase2(struct amdgpu_device *adev) +{ + int i, r; + + if (amdgpu_sriov_vf(adev)) + amdgpu_virt_request_full_gpu(adev, false); /* ungate SMC block first */ r = amdgpu_device_ip_set_clockgating_state(adev, AMD_IP_BLOCK_TYPE_SMC, @@ -1935,9 +1998,16 @@ int amdgpu_device_ip_suspend(struct amdgpu_device *adev) DRM_ERROR("set_clockgating_state(ungate) SMC failed %d\n", r); } + /* call smu to disable gfx off feature first when suspend */ + if (adev->powerplay.pp_funcs->set_powergating_by_smu) + amdgpu_dpm_set_powergating_by_smu(adev, AMD_IP_BLOCK_TYPE_GFX, false); + for (i = adev->num_ip_blocks - 1; i >= 0; i--) { if (!adev->ip_blocks[i].status.valid) continue; + /* displays are handled in phase1 */ + if (adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_DCE) + continue; /* ungate blocks so that suspend can properly shut them down */ if (adev->ip_blocks[i].version->type != AMD_IP_BLOCK_TYPE_SMC && adev->ip_blocks[i].version->funcs->set_clockgating_state) { @@ -1963,6 +2033,29 @@ int amdgpu_device_ip_suspend(struct amdgpu_device *adev) return 0; } +/** + * amdgpu_device_ip_suspend - run suspend for hardware IPs + * + * @adev: amdgpu_device pointer + * + * Main suspend function for hardware IPs. The list of all the hardware + * IPs that make up the asic is walked, clockgating is disabled and the + * suspend callbacks are run. suspend puts the hardware and software state + * in each IP into a state suitable for suspend. + * Returns 0 on success, negative error code on failure. + */ +int amdgpu_device_ip_suspend(struct amdgpu_device *adev) +{ + int r; + + r = amdgpu_device_ip_suspend_phase1(adev); + if (r) + return r; + r = amdgpu_device_ip_suspend_phase2(adev); + + return r; +} + static int amdgpu_device_ip_reinit_early_sriov(struct amdgpu_device *adev) { int i, r; @@ -1985,7 +2078,7 @@ static int amdgpu_device_ip_reinit_early_sriov(struct amdgpu_device *adev) continue; r = block->version->funcs->hw_init(adev); - DRM_INFO("RE-INIT: %s %s\n", block->version->funcs->name, r?"failed":"successed"); + DRM_INFO("RE-INIT: %s %s\n", block->version->funcs->name, r?"failed":"succeeded"); if (r) return r; } @@ -2020,7 +2113,7 @@ static int amdgpu_device_ip_reinit_late_sriov(struct amdgpu_device *adev) continue; r = block->version->funcs->hw_init(adev); - DRM_INFO("RE-INIT: %s %s\n", block->version->funcs->name, r?"failed":"successed"); + DRM_INFO("RE-INIT: %s %s\n", block->version->funcs->name, r?"failed":"succeeded"); if (r) return r; } @@ -2181,7 +2274,7 @@ bool amdgpu_device_asic_has_dc_support(enum amd_asic_type asic_type) case CHIP_VEGA10: case CHIP_VEGA12: case CHIP_VEGA20: -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 case CHIP_RAVEN: #endif return amdgpu_dc != 0; @@ -2210,7 +2303,7 @@ bool amdgpu_device_has_dc_support(struct amdgpu_device *adev) * amdgpu_device_init - initialize the driver * * @adev: amdgpu_device pointer - * @pdev: drm dev pointer + * @ddev: drm dev pointer * @pdev: pci dev pointer * @flags: driver flags * @@ -2301,6 +2394,8 @@ int amdgpu_device_init(struct amdgpu_device *adev, INIT_DELAYED_WORK(&adev->late_init_work, amdgpu_device_ip_late_init_func_handler); + adev->pm.ac_power = power_supply_is_system_supplied() > 0 ? true : false; + /* Registers mapping */ /* TODO: block userspace mapping of io register */ if (adev->asic_type >= CHIP_BONAIRE) { @@ -2581,8 +2676,9 @@ void amdgpu_device_fini(struct amdgpu_device *adev) /** * amdgpu_device_suspend - initiate device suspend * - * @pdev: drm dev pointer - * @state: suspend state + * @dev: drm dev pointer + * @suspend: suspend state + * @fbcon : notify the fbdev of suspend * * Puts the hw in the suspend state (all asics). * Returns 0 for success or an error on failure. @@ -2606,6 +2702,9 @@ int amdgpu_device_suspend(struct drm_device *dev, bool suspend, bool fbcon) drm_kms_helper_poll_disable(dev); + if (fbcon) + amdgpu_fbdev_set_suspend(adev, 1); + if (!amdgpu_device_has_dc_support(adev)) { /* turn off display hw */ drm_modeset_lock_all(dev); @@ -2613,44 +2712,46 @@ int amdgpu_device_suspend(struct drm_device *dev, bool suspend, bool fbcon) drm_helper_connector_dpms(connector, DRM_MODE_DPMS_OFF); } drm_modeset_unlock_all(dev); - } - - amdgpu_amdkfd_suspend(adev); - - /* unpin the front buffers and cursors */ - list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) { - struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc); - struct drm_framebuffer *fb = crtc->primary->fb; - struct amdgpu_bo *robj; - - if (amdgpu_crtc->cursor_bo) { - struct amdgpu_bo *aobj = gem_to_amdgpu_bo(amdgpu_crtc->cursor_bo); - r = amdgpu_bo_reserve(aobj, true); - if (r == 0) { - amdgpu_bo_unpin(aobj); - amdgpu_bo_unreserve(aobj); + /* unpin the front buffers and cursors */ + list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) { + struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc); + struct drm_framebuffer *fb = crtc->primary->fb; + struct amdgpu_bo *robj; + + if (amdgpu_crtc->cursor_bo) { + struct amdgpu_bo *aobj = gem_to_amdgpu_bo(amdgpu_crtc->cursor_bo); + r = amdgpu_bo_reserve(aobj, true); + if (r == 0) { + amdgpu_bo_unpin(aobj); + amdgpu_bo_unreserve(aobj); + } } - } - if (fb == NULL || fb->obj[0] == NULL) { - continue; - } - robj = gem_to_amdgpu_bo(fb->obj[0]); - /* don't unpin kernel fb objects */ - if (!amdgpu_fbdev_robj_is_fb(adev, robj)) { - r = amdgpu_bo_reserve(robj, true); - if (r == 0) { - amdgpu_bo_unpin(robj); - amdgpu_bo_unreserve(robj); + if (fb == NULL || fb->obj[0] == NULL) { + continue; + } + robj = gem_to_amdgpu_bo(fb->obj[0]); + /* don't unpin kernel fb objects */ + if (!amdgpu_fbdev_robj_is_fb(adev, robj)) { + r = amdgpu_bo_reserve(robj, true); + if (r == 0) { + amdgpu_bo_unpin(robj); + amdgpu_bo_unreserve(robj); + } } } } + + amdgpu_amdkfd_suspend(adev); + + r = amdgpu_device_ip_suspend_phase1(adev); + /* evict vram memory */ amdgpu_bo_evict_vram(adev); amdgpu_fence_driver_suspend(adev); - r = amdgpu_device_ip_suspend(adev); + r = amdgpu_device_ip_suspend_phase2(adev); /* evict remaining vram memory * This second call to evict vram is to evict the gart page table @@ -2669,18 +2770,15 @@ int amdgpu_device_suspend(struct drm_device *dev, bool suspend, bool fbcon) DRM_ERROR("amdgpu asic reset failed\n"); } - if (fbcon) { - console_lock(); - amdgpu_fbdev_set_suspend(adev, 1); - console_unlock(); - } return 0; } /** * amdgpu_device_resume - initiate device resume * - * @pdev: drm dev pointer + * @dev: drm dev pointer + * @resume: resume state + * @fbcon : notify the fbdev of resume * * Bring the hw back to operating state (all asics). * Returns 0 for success or an error on failure. @@ -2696,15 +2794,12 @@ int amdgpu_device_resume(struct drm_device *dev, bool resume, bool fbcon) if (dev->switch_power_state == DRM_SWITCH_POWER_OFF) return 0; - if (fbcon) - console_lock(); - if (resume) { pci_set_power_state(dev->pdev, PCI_D0); pci_restore_state(dev->pdev); r = pci_enable_device(dev->pdev); if (r) - goto unlock; + return r; } /* post card */ @@ -2717,29 +2812,30 @@ int amdgpu_device_resume(struct drm_device *dev, bool resume, bool fbcon) r = amdgpu_device_ip_resume(adev); if (r) { DRM_ERROR("amdgpu_device_ip_resume failed (%d).\n", r); - goto unlock; + return r; } amdgpu_fence_driver_resume(adev); r = amdgpu_device_ip_late_init(adev); if (r) - goto unlock; - - /* pin cursors */ - list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) { - struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc); - - if (amdgpu_crtc->cursor_bo) { - struct amdgpu_bo *aobj = gem_to_amdgpu_bo(amdgpu_crtc->cursor_bo); - r = amdgpu_bo_reserve(aobj, true); - if (r == 0) { - r = amdgpu_bo_pin(aobj, - AMDGPU_GEM_DOMAIN_VRAM, - &amdgpu_crtc->cursor_addr); - if (r != 0) - DRM_ERROR("Failed to pin cursor BO (%d)\n", r); - amdgpu_bo_unreserve(aobj); + return r; + + if (!amdgpu_device_has_dc_support(adev)) { + /* pin cursors */ + list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) { + struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc); + + if (amdgpu_crtc->cursor_bo) { + struct amdgpu_bo *aobj = gem_to_amdgpu_bo(amdgpu_crtc->cursor_bo); + r = amdgpu_bo_reserve(aobj, true); + if (r == 0) { + r = amdgpu_bo_pin(aobj, AMDGPU_GEM_DOMAIN_VRAM); + if (r != 0) + DRM_ERROR("Failed to pin cursor BO (%d)\n", r); + amdgpu_crtc->cursor_addr = amdgpu_bo_gpu_offset(aobj); + amdgpu_bo_unreserve(aobj); + } } } } @@ -2763,6 +2859,7 @@ int amdgpu_device_resume(struct drm_device *dev, bool resume, bool fbcon) } drm_modeset_unlock_all(dev); } + amdgpu_fbdev_set_suspend(adev, 0); } drm_kms_helper_poll_enable(dev); @@ -2786,15 +2883,7 @@ int amdgpu_device_resume(struct drm_device *dev, bool resume, bool fbcon) #ifdef CONFIG_PM dev->dev->power.disable_depth--; #endif - - if (fbcon) - amdgpu_fbdev_set_suspend(adev, 0); - -unlock: - if (fbcon) - console_unlock(); - - return r; + return 0; } /** @@ -3019,7 +3108,7 @@ static int amdgpu_device_handle_vram_lost(struct amdgpu_device *adev) long tmo; if (amdgpu_sriov_runtime(adev)) - tmo = msecs_to_jiffies(amdgpu_lockup_timeout); + tmo = msecs_to_jiffies(8000); else tmo = msecs_to_jiffies(100); @@ -3071,7 +3160,7 @@ static int amdgpu_device_handle_vram_lost(struct amdgpu_device *adev) * @adev: amdgpu device pointer * * attempt to do soft-reset or full-reset and reinitialize Asic - * return 0 means successed otherwise failed + * return 0 means succeeded otherwise failed */ static int amdgpu_device_reset(struct amdgpu_device *adev) { @@ -3146,9 +3235,10 @@ out: * amdgpu_device_reset_sriov - reset ASIC for SR-IOV vf * * @adev: amdgpu device pointer + * @from_hypervisor: request from hypervisor * * do VF FLR and reinitialize Asic - * return 0 means successed otherwise failed + * return 0 means succeeded otherwise failed */ static int amdgpu_device_reset_sriov(struct amdgpu_device *adev, bool from_hypervisor) @@ -3193,7 +3283,7 @@ error: * * @adev: amdgpu device pointer * @job: which job trigger hang - * @force forces reset regardless of amdgpu_gpu_recovery + * @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. @@ -3220,6 +3310,9 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev, atomic_inc(&adev->gpu_reset_counter); adev->in_gpu_reset = 1; + /* Block kfd */ + amdgpu_amdkfd_pre_reset(adev); + /* block TTM */ resched = ttm_bo_lock_delayed_workqueue(&adev->mman.bdev); @@ -3232,10 +3325,10 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev, kthread_park(ring->sched.thread); - if (job && job->ring->idx != i) + if (job && job->base.sched == &ring->sched) continue; - drm_sched_hw_job_reset(&ring->sched, &job->base); + drm_sched_hw_job_reset(&ring->sched, job ? &job->base : NULL); /* after all hw jobs are reset, hw fence is meaningless, so force_completion */ amdgpu_fence_driver_force_completion(ring); @@ -3256,7 +3349,7 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev, * or all rings (in the case @job is NULL) * after above amdgpu_reset accomplished */ - if ((!job || job->ring->idx == i) && !r) + if ((!job || job->base.sched == &ring->sched) && !r) drm_sched_job_recovery(&ring->sched); kthread_unpark(ring->sched.thread); @@ -3273,9 +3366,11 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev, dev_info(adev->dev, "GPU reset(%d) failed\n", atomic_read(&adev->gpu_reset_counter)); amdgpu_vf_error_put(adev, AMDGIM_ERROR_VF_GPU_RESET_FAIL, 0, r); } else { - dev_info(adev->dev, "GPU reset(%d) successed!\n",atomic_read(&adev->gpu_reset_counter)); + dev_info(adev->dev, "GPU reset(%d) succeeded!\n",atomic_read(&adev->gpu_reset_counter)); } + /*unlock kfd */ + amdgpu_amdkfd_post_reset(adev); amdgpu_vf_error_trans_all(adev); adev->in_gpu_reset = 0; mutex_unlock(&adev->lock_reset); @@ -3293,8 +3388,9 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev, */ static void amdgpu_device_get_pcie_info(struct amdgpu_device *adev) { - u32 mask; - int ret; + struct pci_dev *pdev; + enum pci_bus_speed speed_cap; + enum pcie_link_width link_width; if (amdgpu_pcie_gen_cap) adev->pm.pcie_gen_mask = amdgpu_pcie_gen_cap; @@ -3312,27 +3408,61 @@ static void amdgpu_device_get_pcie_info(struct amdgpu_device *adev) } if (adev->pm.pcie_gen_mask == 0) { - ret = drm_pcie_get_speed_cap_mask(adev->ddev, &mask); - if (!ret) { - adev->pm.pcie_gen_mask = (CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN1 | + /* asic caps */ + pdev = adev->pdev; + speed_cap = pcie_get_speed_cap(pdev); + if (speed_cap == PCI_SPEED_UNKNOWN) { + adev->pm.pcie_gen_mask |= (CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN1 | CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN2 | CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN3); - - if (mask & DRM_PCIE_SPEED_25) - adev->pm.pcie_gen_mask |= CAIL_PCIE_LINK_SPEED_SUPPORT_GEN1; - if (mask & DRM_PCIE_SPEED_50) - adev->pm.pcie_gen_mask |= CAIL_PCIE_LINK_SPEED_SUPPORT_GEN2; - if (mask & DRM_PCIE_SPEED_80) - adev->pm.pcie_gen_mask |= CAIL_PCIE_LINK_SPEED_SUPPORT_GEN3; } else { - adev->pm.pcie_gen_mask = AMDGPU_DEFAULT_PCIE_GEN_MASK; + if (speed_cap == PCIE_SPEED_16_0GT) + adev->pm.pcie_gen_mask |= (CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN1 | + CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN2 | + CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN3 | + CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN4); + else if (speed_cap == PCIE_SPEED_8_0GT) + adev->pm.pcie_gen_mask |= (CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN1 | + CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN2 | + CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN3); + else if (speed_cap == PCIE_SPEED_5_0GT) + adev->pm.pcie_gen_mask |= (CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN1 | + CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN2); + else + adev->pm.pcie_gen_mask |= CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN1; + } + /* platform caps */ + pdev = adev->ddev->pdev->bus->self; + speed_cap = pcie_get_speed_cap(pdev); + if (speed_cap == PCI_SPEED_UNKNOWN) { + adev->pm.pcie_gen_mask |= (CAIL_PCIE_LINK_SPEED_SUPPORT_GEN1 | + CAIL_PCIE_LINK_SPEED_SUPPORT_GEN2); + } else { + if (speed_cap == PCIE_SPEED_16_0GT) + adev->pm.pcie_gen_mask |= (CAIL_PCIE_LINK_SPEED_SUPPORT_GEN1 | + CAIL_PCIE_LINK_SPEED_SUPPORT_GEN2 | + CAIL_PCIE_LINK_SPEED_SUPPORT_GEN3 | + CAIL_PCIE_LINK_SPEED_SUPPORT_GEN4); + else if (speed_cap == PCIE_SPEED_8_0GT) + adev->pm.pcie_gen_mask |= (CAIL_PCIE_LINK_SPEED_SUPPORT_GEN1 | + CAIL_PCIE_LINK_SPEED_SUPPORT_GEN2 | + CAIL_PCIE_LINK_SPEED_SUPPORT_GEN3); + else if (speed_cap == PCIE_SPEED_5_0GT) + adev->pm.pcie_gen_mask |= (CAIL_PCIE_LINK_SPEED_SUPPORT_GEN1 | + CAIL_PCIE_LINK_SPEED_SUPPORT_GEN2); + else + adev->pm.pcie_gen_mask |= CAIL_PCIE_LINK_SPEED_SUPPORT_GEN1; + } } if (adev->pm.pcie_mlw_mask == 0) { - ret = drm_pcie_get_max_link_width(adev->ddev, &mask); - if (!ret) { - switch (mask) { - case 32: + pdev = adev->ddev->pdev->bus->self; + link_width = pcie_get_width_cap(pdev); + if (link_width == PCIE_LNK_WIDTH_UNKNOWN) { + adev->pm.pcie_mlw_mask |= AMDGPU_DEFAULT_PCIE_MLW_MASK; + } else { + switch (link_width) { + case PCIE_LNK_X32: adev->pm.pcie_mlw_mask = (CAIL_PCIE_LINK_WIDTH_SUPPORT_X32 | CAIL_PCIE_LINK_WIDTH_SUPPORT_X16 | CAIL_PCIE_LINK_WIDTH_SUPPORT_X12 | @@ -3341,7 +3471,7 @@ static void amdgpu_device_get_pcie_info(struct amdgpu_device *adev) CAIL_PCIE_LINK_WIDTH_SUPPORT_X2 | CAIL_PCIE_LINK_WIDTH_SUPPORT_X1); break; - case 16: + case PCIE_LNK_X16: adev->pm.pcie_mlw_mask = (CAIL_PCIE_LINK_WIDTH_SUPPORT_X16 | CAIL_PCIE_LINK_WIDTH_SUPPORT_X12 | CAIL_PCIE_LINK_WIDTH_SUPPORT_X8 | @@ -3349,36 +3479,34 @@ static void amdgpu_device_get_pcie_info(struct amdgpu_device *adev) CAIL_PCIE_LINK_WIDTH_SUPPORT_X2 | CAIL_PCIE_LINK_WIDTH_SUPPORT_X1); break; - case 12: + case PCIE_LNK_X12: adev->pm.pcie_mlw_mask = (CAIL_PCIE_LINK_WIDTH_SUPPORT_X12 | CAIL_PCIE_LINK_WIDTH_SUPPORT_X8 | CAIL_PCIE_LINK_WIDTH_SUPPORT_X4 | CAIL_PCIE_LINK_WIDTH_SUPPORT_X2 | CAIL_PCIE_LINK_WIDTH_SUPPORT_X1); break; - case 8: + case PCIE_LNK_X8: adev->pm.pcie_mlw_mask = (CAIL_PCIE_LINK_WIDTH_SUPPORT_X8 | CAIL_PCIE_LINK_WIDTH_SUPPORT_X4 | CAIL_PCIE_LINK_WIDTH_SUPPORT_X2 | CAIL_PCIE_LINK_WIDTH_SUPPORT_X1); break; - case 4: + case PCIE_LNK_X4: adev->pm.pcie_mlw_mask = (CAIL_PCIE_LINK_WIDTH_SUPPORT_X4 | CAIL_PCIE_LINK_WIDTH_SUPPORT_X2 | CAIL_PCIE_LINK_WIDTH_SUPPORT_X1); break; - case 2: + case PCIE_LNK_X2: adev->pm.pcie_mlw_mask = (CAIL_PCIE_LINK_WIDTH_SUPPORT_X2 | CAIL_PCIE_LINK_WIDTH_SUPPORT_X1); break; - case 1: + case PCIE_LNK_X1: adev->pm.pcie_mlw_mask = CAIL_PCIE_LINK_WIDTH_SUPPORT_X1; break; default: break; } - } else { - adev->pm.pcie_mlw_mask = AMDGPU_DEFAULT_PCIE_MLW_MASK; } } } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c index 76ee8e04ff11..6748cd7fc129 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c @@ -157,7 +157,6 @@ int amdgpu_display_crtc_page_flip_target(struct drm_crtc *crtc, struct amdgpu_bo *new_abo; unsigned long flags; u64 tiling_flags; - u64 base; int i, r; work = kzalloc(sizeof *work, GFP_KERNEL); @@ -189,12 +188,18 @@ int amdgpu_display_crtc_page_flip_target(struct drm_crtc *crtc, goto cleanup; } - r = amdgpu_bo_pin(new_abo, amdgpu_display_supported_domains(adev), &base); + r = amdgpu_bo_pin(new_abo, amdgpu_display_supported_domains(adev)); if (unlikely(r != 0)) { DRM_ERROR("failed to pin new abo buffer before flip\n"); goto unreserve; } + r = amdgpu_ttm_alloc_gart(&new_abo->tbo); + if (unlikely(r != 0)) { + DRM_ERROR("%p bind failed\n", new_abo); + goto unpin; + } + r = reservation_object_get_fences_rcu(new_abo->tbo.resv, &work->excl, &work->shared_count, &work->shared); @@ -206,7 +211,7 @@ int amdgpu_display_crtc_page_flip_target(struct drm_crtc *crtc, amdgpu_bo_get_tiling_flags(new_abo, &tiling_flags); amdgpu_bo_unreserve(new_abo); - work->base = base; + work->base = amdgpu_bo_gpu_offset(new_abo); work->target_vblank = target - (uint32_t)drm_crtc_vblank_count(crtc) + amdgpu_get_vblank_counter_kms(dev, work->crtc_id); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.c index 77ad59ade85c..1c4595562f8f 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.c @@ -28,6 +28,7 @@ #include "amdgpu_i2c.h" #include "amdgpu_dpm.h" #include "atom.h" +#include "amd_pcie.h" void amdgpu_dpm_print_class_info(u32 class, u32 class2) { @@ -936,9 +937,11 @@ enum amdgpu_pcie_gen amdgpu_get_pcie_gen_support(struct amdgpu_device *adev, case AMDGPU_PCIE_GEN3: return AMDGPU_PCIE_GEN3; default: - if ((sys_mask & DRM_PCIE_SPEED_80) && (default_gen == AMDGPU_PCIE_GEN3)) + if ((sys_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_GEN3) && + (default_gen == AMDGPU_PCIE_GEN3)) return AMDGPU_PCIE_GEN3; - else if ((sys_mask & DRM_PCIE_SPEED_50) && (default_gen == AMDGPU_PCIE_GEN2)) + else if ((sys_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_GEN2) && + (default_gen == AMDGPU_PCIE_GEN2)) return AMDGPU_PCIE_GEN2; else return AMDGPU_PCIE_GEN1; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h index dd6203a0a6b7..ff24e1cc5b65 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h @@ -287,12 +287,6 @@ enum amdgpu_pcie_gen { #define amdgpu_dpm_force_performance_level(adev, l) \ ((adev)->powerplay.pp_funcs->force_performance_level((adev)->powerplay.pp_handle, (l))) -#define amdgpu_dpm_powergate_uvd(adev, g) \ - ((adev)->powerplay.pp_funcs->powergate_uvd((adev)->powerplay.pp_handle, (g))) - -#define amdgpu_dpm_powergate_vce(adev, g) \ - ((adev)->powerplay.pp_funcs->powergate_vce((adev)->powerplay.pp_handle, (g))) - #define amdgpu_dpm_get_current_power_state(adev) \ ((adev)->powerplay.pp_funcs->get_current_power_state((adev)->powerplay.pp_handle)) @@ -347,6 +341,10 @@ enum amdgpu_pcie_gen { ((adev)->powerplay.pp_funcs->set_clockgating_by_smu(\ (adev)->powerplay.pp_handle, msg_id)) +#define amdgpu_dpm_set_powergating_by_smu(adev, block_type, gate) \ + ((adev)->powerplay.pp_funcs->set_powergating_by_smu(\ + (adev)->powerplay.pp_handle, block_type, gate)) + #define amdgpu_dpm_get_power_profile_mode(adev, buf) \ ((adev)->powerplay.pp_funcs->get_power_profile_mode(\ (adev)->powerplay.pp_handle, buf)) @@ -359,10 +357,6 @@ enum amdgpu_pcie_gen { ((adev)->powerplay.pp_funcs->odn_edit_dpm_table(\ (adev)->powerplay.pp_handle, type, parameter, size)) -#define amdgpu_dpm_set_mmhub_powergating_by_smu(adev) \ - ((adev)->powerplay.pp_funcs->set_mmhub_powergating_by_smu( \ - (adev)->powerplay.pp_handle)) - struct amdgpu_dpm { struct amdgpu_ps *ps; /* number of valid power states */ @@ -402,7 +396,6 @@ struct amdgpu_dpm { u32 tdp_adjustment; u16 load_line_slope; bool power_control; - bool ac_power; /* special states active */ bool thermal_active; bool uvd_active; @@ -439,6 +432,7 @@ struct amdgpu_pm { struct amd_pp_display_configuration pm_display_cfg;/* set by dc */ uint32_t smu_prv_buffer_size; struct amdgpu_bo *smu_prv_buffer; + bool ac_power; }; #define R600_SSTU_DFLT 0 diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c index b0bf2f24da48..8843a06360fa 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c @@ -1,10 +1,3 @@ -/** - * \file amdgpu_drv.c - * AMD Amdgpu driver - * - * \author Gareth Hughes <gareth@valinux.com> - */ - /* * Copyright 2000 VA Linux Systems, Inc., Sunnyvale, California. * All Rights Reserved. @@ -76,9 +69,10 @@ * - 3.24.0 - Add high priority compute support for gfx9 * - 3.25.0 - Add support for sensor query info (stable pstate sclk/mclk). * - 3.26.0 - GFX9: Process AMDGPU_IB_FLAG_TC_WB_NOT_INVALIDATE. + * - 3.27.0 - Add new chunk to to AMDGPU_CS to enable BO_LIST creation. */ #define KMS_DRIVER_MAJOR 3 -#define KMS_DRIVER_MINOR 26 +#define KMS_DRIVER_MINOR 27 #define KMS_DRIVER_PATCHLEVEL 0 int amdgpu_vram_limit = 0; @@ -110,11 +104,8 @@ int amdgpu_vram_page_split = 512; int amdgpu_vm_update_mode = -1; int amdgpu_exp_hw_support = 0; int amdgpu_dc = -1; -int amdgpu_dc_log = 0; int amdgpu_sched_jobs = 32; int amdgpu_sched_hw_submission = 2; -int amdgpu_no_evict = 0; -int amdgpu_direct_gma_size = 0; uint amdgpu_pcie_gen_cap = 0; uint amdgpu_pcie_lane_cap = 0; uint amdgpu_cg_mask = 0xffffffff; @@ -122,7 +113,8 @@ uint amdgpu_pg_mask = 0xffffffff; uint amdgpu_sdma_phase_quantum = 32; char *amdgpu_disable_cu = NULL; char *amdgpu_virtual_display = NULL; -uint amdgpu_pp_feature_mask = 0xffff3fff; /* gfxoff (bit 15) disabled by default */ +/* OverDrive(bit 14),gfxoff(bit 15),stutter mode(bit 17) disabled by default*/ +uint amdgpu_pp_feature_mask = 0xfffd3fff; int amdgpu_ngg = 0; int amdgpu_prim_buf_per_se = 0; int amdgpu_pos_buf_per_se = 0; @@ -135,163 +127,368 @@ int amdgpu_gpu_recovery = -1; /* auto */ int amdgpu_emu_mode = 0; uint amdgpu_smu_memory_pool_size = 0; +/** + * DOC: vramlimit (int) + * Restrict the total amount of VRAM in MiB for testing. The default is 0 (Use full VRAM). + */ MODULE_PARM_DESC(vramlimit, "Restrict VRAM for testing, in megabytes"); module_param_named(vramlimit, amdgpu_vram_limit, int, 0600); +/** + * DOC: vis_vramlimit (int) + * Restrict the amount of CPU visible VRAM in MiB for testing. The default is 0 (Use full CPU visible VRAM). + */ MODULE_PARM_DESC(vis_vramlimit, "Restrict visible VRAM for testing, in megabytes"); module_param_named(vis_vramlimit, amdgpu_vis_vram_limit, int, 0444); +/** + * DOC: gartsize (uint) + * Restrict the size of GART in Mib (32, 64, etc.) for testing. The default is -1 (The size depends on asic). + */ MODULE_PARM_DESC(gartsize, "Size of GART to setup in megabytes (32, 64, etc., -1=auto)"); module_param_named(gartsize, amdgpu_gart_size, uint, 0600); +/** + * DOC: gttsize (int) + * Restrict the size of GTT domain in MiB for testing. The default is -1 (It's VRAM size if 3GB < VRAM < 3/4 RAM, + * otherwise 3/4 RAM size). + */ MODULE_PARM_DESC(gttsize, "Size of the GTT domain in megabytes (-1 = auto)"); module_param_named(gttsize, amdgpu_gtt_size, int, 0600); +/** + * DOC: moverate (int) + * Set maximum buffer migration rate in MB/s. The default is -1 (8 MB/s). + */ MODULE_PARM_DESC(moverate, "Maximum buffer migration rate in MB/s. (32, 64, etc., -1=auto, 0=1=disabled)"); module_param_named(moverate, amdgpu_moverate, int, 0600); +/** + * DOC: benchmark (int) + * Run benchmarks. The default is 0 (Skip benchmarks). + */ MODULE_PARM_DESC(benchmark, "Run benchmark"); module_param_named(benchmark, amdgpu_benchmarking, int, 0444); +/** + * DOC: test (int) + * Test BO GTT->VRAM and VRAM->GTT GPU copies. The default is 0 (Skip test, only set 1 to run test). + */ MODULE_PARM_DESC(test, "Run tests"); module_param_named(test, amdgpu_testing, int, 0444); +/** + * DOC: audio (int) + * Set HDMI/DPAudio. Only affects non-DC display handling. The default is -1 (Enabled), set 0 to disabled it. + */ MODULE_PARM_DESC(audio, "Audio enable (-1 = auto, 0 = disable, 1 = enable)"); module_param_named(audio, amdgpu_audio, int, 0444); +/** + * DOC: disp_priority (int) + * Set display Priority (1 = normal, 2 = high). Only affects non-DC display handling. The default is 0 (auto). + */ MODULE_PARM_DESC(disp_priority, "Display Priority (0 = auto, 1 = normal, 2 = high)"); module_param_named(disp_priority, amdgpu_disp_priority, int, 0444); +/** + * DOC: hw_i2c (int) + * To enable hw i2c engine. Only affects non-DC display handling. The default is 0 (Disabled). + */ MODULE_PARM_DESC(hw_i2c, "hw i2c engine enable (0 = disable)"); module_param_named(hw_i2c, amdgpu_hw_i2c, int, 0444); +/** + * DOC: pcie_gen2 (int) + * To disable PCIE Gen2/3 mode (0 = disable, 1 = enable). The default is -1 (auto, enabled). + */ MODULE_PARM_DESC(pcie_gen2, "PCIE Gen2 mode (-1 = auto, 0 = disable, 1 = enable)"); module_param_named(pcie_gen2, amdgpu_pcie_gen2, int, 0444); +/** + * DOC: msi (int) + * To disable Message Signaled Interrupts (MSI) functionality (1 = enable, 0 = disable). The default is -1 (auto, enabled). + */ MODULE_PARM_DESC(msi, "MSI support (1 = enable, 0 = disable, -1 = auto)"); module_param_named(msi, amdgpu_msi, int, 0444); +/** + * DOC: lockup_timeout (int) + * Set GPU scheduler timeout value in ms. Value 0 is invalidated, will be adjusted to 10000. + * Negative values mean 'infinite timeout' (MAX_JIFFY_OFFSET). The default is 10000. + */ MODULE_PARM_DESC(lockup_timeout, "GPU lockup timeout in ms > 0 (default 10000)"); module_param_named(lockup_timeout, amdgpu_lockup_timeout, int, 0444); +/** + * DOC: dpm (int) + * Override for dynamic power management setting (1 = enable, 0 = disable). The default is -1 (auto). + */ MODULE_PARM_DESC(dpm, "DPM support (1 = enable, 0 = disable, -1 = auto)"); module_param_named(dpm, amdgpu_dpm, int, 0444); +/** + * DOC: fw_load_type (int) + * Set different firmware loading type for debugging (0 = direct, 1 = SMU, 2 = PSP). The default is -1 (auto). + */ MODULE_PARM_DESC(fw_load_type, "firmware loading type (0 = direct, 1 = SMU, 2 = PSP, -1 = auto)"); module_param_named(fw_load_type, amdgpu_fw_load_type, int, 0444); +/** + * DOC: aspm (int) + * To disable ASPM (1 = enable, 0 = disable). The default is -1 (auto, enabled). + */ MODULE_PARM_DESC(aspm, "ASPM support (1 = enable, 0 = disable, -1 = auto)"); module_param_named(aspm, amdgpu_aspm, int, 0444); +/** + * DOC: runpm (int) + * Override for runtime power management control for dGPUs in PX/HG laptops. The amdgpu driver can dynamically power down + * the dGPU on PX/HG laptops when it is idle. The default is -1 (auto enable). Setting the value to 0 disables this functionality. + */ MODULE_PARM_DESC(runpm, "PX runtime pm (1 = force enable, 0 = disable, -1 = PX only default)"); module_param_named(runpm, amdgpu_runtime_pm, int, 0444); +/** + * DOC: ip_block_mask (uint) + * Override what IP blocks are enabled on the GPU. Each GPU is a collection of IP blocks (gfx, display, video, etc.). + * Use this parameter to disable specific blocks. Note that the IP blocks do not have a fixed index. Some asics may not have + * some IPs or may include multiple instances of an IP so the ordering various from asic to asic. See the driver output in + * the kernel log for the list of IPs on the asic. The default is 0xffffffff (enable all blocks on a device). + */ MODULE_PARM_DESC(ip_block_mask, "IP Block Mask (all blocks enabled (default))"); module_param_named(ip_block_mask, amdgpu_ip_block_mask, uint, 0444); +/** + * DOC: bapm (int) + * Bidirectional Application Power Management (BAPM) used to dynamically share TDP between CPU and GPU. Set value 0 to disable it. + * The default -1 (auto, enabled) + */ MODULE_PARM_DESC(bapm, "BAPM support (1 = enable, 0 = disable, -1 = auto)"); module_param_named(bapm, amdgpu_bapm, int, 0444); +/** + * DOC: deep_color (int) + * Set 1 to enable Deep Color support. Only affects non-DC display handling. The default is 0 (disabled). + */ MODULE_PARM_DESC(deep_color, "Deep Color support (1 = enable, 0 = disable (default))"); module_param_named(deep_color, amdgpu_deep_color, int, 0444); +/** + * DOC: vm_size (int) + * Override the size of the GPU's per client virtual address space in GiB. The default is -1 (automatic for each asic). + */ MODULE_PARM_DESC(vm_size, "VM address space size in gigabytes (default 64GB)"); module_param_named(vm_size, amdgpu_vm_size, int, 0444); +/** + * DOC: vm_fragment_size (int) + * Override VM fragment size in bits (4, 5, etc. 4 = 64K, 9 = 2M). The default is -1 (automatic for each asic). + */ MODULE_PARM_DESC(vm_fragment_size, "VM fragment size in bits (4, 5, etc. 4 = 64K (default), Max 9 = 2M)"); module_param_named(vm_fragment_size, amdgpu_vm_fragment_size, int, 0444); +/** + * DOC: vm_block_size (int) + * Override VM page table size in bits (default depending on vm_size and hw setup). The default is -1 (automatic for each asic). + */ MODULE_PARM_DESC(vm_block_size, "VM page table size in bits (default depending on vm_size)"); module_param_named(vm_block_size, amdgpu_vm_block_size, int, 0444); +/** + * DOC: vm_fault_stop (int) + * Stop on VM fault for debugging (0 = never, 1 = print first, 2 = always). The default is 0 (No stop). + */ MODULE_PARM_DESC(vm_fault_stop, "Stop on VM fault (0 = never (default), 1 = print first, 2 = always)"); module_param_named(vm_fault_stop, amdgpu_vm_fault_stop, int, 0444); +/** + * DOC: vm_debug (int) + * Debug VM handling (0 = disabled, 1 = enabled). The default is 0 (Disabled). + */ MODULE_PARM_DESC(vm_debug, "Debug VM handling (0 = disabled (default), 1 = enabled)"); module_param_named(vm_debug, amdgpu_vm_debug, int, 0644); +/** + * DOC: vm_update_mode (int) + * Override VM update mode. VM updated by using CPU (0 = never, 1 = Graphics only, 2 = Compute only, 3 = Both). The default + * is -1 (Only in large BAR(LB) systems Compute VM tables will be updated by CPU, otherwise 0, never). + */ MODULE_PARM_DESC(vm_update_mode, "VM update using CPU (0 = never (default except for large BAR(LB)), 1 = Graphics only, 2 = Compute only (default for LB), 3 = Both"); module_param_named(vm_update_mode, amdgpu_vm_update_mode, int, 0444); +/** + * DOC: vram_page_split (int) + * Override the number of pages after we split VRAM allocations (default 512, -1 = disable). The default is 512. + */ MODULE_PARM_DESC(vram_page_split, "Number of pages after we split VRAM allocations (default 512, -1 = disable)"); module_param_named(vram_page_split, amdgpu_vram_page_split, int, 0444); +/** + * DOC: exp_hw_support (int) + * Enable experimental hw support (1 = enable). The default is 0 (disabled). + */ MODULE_PARM_DESC(exp_hw_support, "experimental hw support (1 = enable, 0 = disable (default))"); module_param_named(exp_hw_support, amdgpu_exp_hw_support, int, 0444); +/** + * DOC: dc (int) + * Disable/Enable Display Core driver for debugging (1 = enable, 0 = disable). The default is -1 (automatic for each asic). + */ MODULE_PARM_DESC(dc, "Display Core driver (1 = enable, 0 = disable, -1 = auto (default))"); module_param_named(dc, amdgpu_dc, int, 0444); -MODULE_PARM_DESC(dc_log, "Display Core Log Level (0 = minimal (default), 1 = chatty"); -module_param_named(dc_log, amdgpu_dc_log, int, 0444); - +/** + * DOC: sched_jobs (int) + * Override the max number of jobs supported in the sw queue. The default is 32. + */ MODULE_PARM_DESC(sched_jobs, "the max number of jobs supported in the sw queue (default 32)"); module_param_named(sched_jobs, amdgpu_sched_jobs, int, 0444); +/** + * DOC: sched_hw_submission (int) + * Override the max number of HW submissions. The default is 2. + */ MODULE_PARM_DESC(sched_hw_submission, "the max number of HW submissions (default 2)"); module_param_named(sched_hw_submission, amdgpu_sched_hw_submission, int, 0444); +/** + * DOC: ppfeaturemask (uint) + * Override power features enabled. See enum PP_FEATURE_MASK in drivers/gpu/drm/amd/include/amd_shared.h. + * The default is the current set of stable power features. + */ MODULE_PARM_DESC(ppfeaturemask, "all power features enabled (default))"); module_param_named(ppfeaturemask, amdgpu_pp_feature_mask, uint, 0444); -MODULE_PARM_DESC(no_evict, "Support pinning request from user space (1 = enable, 0 = disable (default))"); -module_param_named(no_evict, amdgpu_no_evict, int, 0444); - -MODULE_PARM_DESC(direct_gma_size, "Direct GMA size in megabytes (max 96MB)"); -module_param_named(direct_gma_size, amdgpu_direct_gma_size, int, 0444); - +/** + * DOC: pcie_gen_cap (uint) + * Override PCIE gen speed capabilities. See the CAIL flags in drivers/gpu/drm/amd/include/amd_pcie.h. + * The default is 0 (automatic for each asic). + */ MODULE_PARM_DESC(pcie_gen_cap, "PCIE Gen Caps (0: autodetect (default))"); module_param_named(pcie_gen_cap, amdgpu_pcie_gen_cap, uint, 0444); +/** + * DOC: pcie_lane_cap (uint) + * Override PCIE lanes capabilities. See the CAIL flags in drivers/gpu/drm/amd/include/amd_pcie.h. + * The default is 0 (automatic for each asic). + */ MODULE_PARM_DESC(pcie_lane_cap, "PCIE Lane Caps (0: autodetect (default))"); module_param_named(pcie_lane_cap, amdgpu_pcie_lane_cap, uint, 0444); +/** + * DOC: cg_mask (uint) + * Override Clockgating features enabled on GPU (0 = disable clock gating). See the AMD_CG_SUPPORT flags in + * drivers/gpu/drm/amd/include/amd_shared.h. The default is 0xffffffff (all enabled). + */ MODULE_PARM_DESC(cg_mask, "Clockgating flags mask (0 = disable clock gating)"); module_param_named(cg_mask, amdgpu_cg_mask, uint, 0444); +/** + * DOC: pg_mask (uint) + * Override Powergating features enabled on GPU (0 = disable power gating). See the AMD_PG_SUPPORT flags in + * drivers/gpu/drm/amd/include/amd_shared.h. The default is 0xffffffff (all enabled). + */ MODULE_PARM_DESC(pg_mask, "Powergating flags mask (0 = disable power gating)"); module_param_named(pg_mask, amdgpu_pg_mask, uint, 0444); +/** + * DOC: sdma_phase_quantum (uint) + * Override SDMA context switch phase quantum (x 1K GPU clock cycles, 0 = no change). The default is 32. + */ MODULE_PARM_DESC(sdma_phase_quantum, "SDMA context switch phase quantum (x 1K GPU clock cycles, 0 = no change (default 32))"); module_param_named(sdma_phase_quantum, amdgpu_sdma_phase_quantum, uint, 0444); +/** + * DOC: disable_cu (charp) + * Set to disable CUs (It's set like se.sh.cu,...). The default is NULL. + */ MODULE_PARM_DESC(disable_cu, "Disable CUs (se.sh.cu,...)"); module_param_named(disable_cu, amdgpu_disable_cu, charp, 0444); +/** + * DOC: virtual_display (charp) + * Set to enable virtual display feature. This feature provides a virtual display hardware on headless boards + * or in virtualized environments. It will be set like xxxx:xx:xx.x,x;xxxx:xx:xx.x,x. It's the pci address of + * the device, plus the number of crtcs to expose. E.g., 0000:26:00.0,4 would enable 4 virtual crtcs on the pci + * device at 26:00.0. The default is NULL. + */ MODULE_PARM_DESC(virtual_display, "Enable virtual display feature (the virtual_display will be set like xxxx:xx:xx.x,x;xxxx:xx:xx.x,x)"); module_param_named(virtual_display, amdgpu_virtual_display, charp, 0444); +/** + * DOC: ngg (int) + * Set to enable Next Generation Graphics (1 = enable). The default is 0 (disabled). + */ MODULE_PARM_DESC(ngg, "Next Generation Graphics (1 = enable, 0 = disable(default depending on gfx))"); module_param_named(ngg, amdgpu_ngg, int, 0444); +/** + * DOC: prim_buf_per_se (int) + * Override the size of Primitive Buffer per Shader Engine in Byte. The default is 0 (depending on gfx). + */ MODULE_PARM_DESC(prim_buf_per_se, "the size of Primitive Buffer per Shader Engine (default depending on gfx)"); module_param_named(prim_buf_per_se, amdgpu_prim_buf_per_se, int, 0444); +/** + * DOC: pos_buf_per_se (int) + * Override the size of Position Buffer per Shader Engine in Byte. The default is 0 (depending on gfx). + */ MODULE_PARM_DESC(pos_buf_per_se, "the size of Position Buffer per Shader Engine (default depending on gfx)"); module_param_named(pos_buf_per_se, amdgpu_pos_buf_per_se, int, 0444); +/** + * DOC: cntl_sb_buf_per_se (int) + * Override the size of Control Sideband per Shader Engine in Byte. The default is 0 (depending on gfx). + */ MODULE_PARM_DESC(cntl_sb_buf_per_se, "the size of Control Sideband per Shader Engine (default depending on gfx)"); module_param_named(cntl_sb_buf_per_se, amdgpu_cntl_sb_buf_per_se, int, 0444); +/** + * DOC: param_buf_per_se (int) + * Override the size of Off-Chip Pramater Cache per Shader Engine in Byte. The default is 0 (depending on gfx). + */ MODULE_PARM_DESC(param_buf_per_se, "the size of Off-Chip Pramater Cache per Shader Engine (default depending on gfx)"); module_param_named(param_buf_per_se, amdgpu_param_buf_per_se, int, 0444); +/** + * DOC: job_hang_limit (int) + * Set how much time allow a job hang and not drop it. The default is 0. + */ MODULE_PARM_DESC(job_hang_limit, "how much time allow a job hang and not drop it (default 0)"); module_param_named(job_hang_limit, amdgpu_job_hang_limit, int ,0444); +/** + * DOC: lbpw (int) + * Override Load Balancing Per Watt (LBPW) support (1 = enable, 0 = disable). The default is -1 (auto, enabled). + */ MODULE_PARM_DESC(lbpw, "Load Balancing Per Watt (LBPW) support (1 = enable, 0 = disable, -1 = auto)"); 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); +/** + * DOC: gpu_recovery (int) + * Set to enable GPU recovery mechanism (1 = enable, 0 = disable). The default is -1 (auto, disabled except SRIOV). + */ 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); +/** + * DOC: emu_mode (int) + * Set value 1 to enable emulation mode. This is only needed when running on an emulator. The default is 0 (disabled). + */ MODULE_PARM_DESC(emu_mode, "Emulation mode, (1 = enable, 0 = disable)"); module_param_named(emu_mode, amdgpu_emu_mode, int, 0444); +/** + * DOC: si_support (int) + * Set SI support driver. This parameter works after set config CONFIG_DRM_AMDGPU_SI. For SI asic, when radeon driver is enabled, + * set value 0 to use radeon driver, while set value 1 to use amdgpu driver. The default is using radeon driver when it available, + * otherwise using amdgpu driver. + */ #ifdef CONFIG_DRM_AMDGPU_SI #if defined(CONFIG_DRM_RADEON) || defined(CONFIG_DRM_RADEON_MODULE) @@ -305,6 +502,12 @@ MODULE_PARM_DESC(si_support, "SI support (1 = enabled (default), 0 = disabled)") module_param_named(si_support, amdgpu_si_support, int, 0444); #endif +/** + * DOC: cik_support (int) + * Set CIK support driver. This parameter works after set config CONFIG_DRM_AMDGPU_CIK. For CIK asic, when radeon driver is enabled, + * set value 0 to use radeon driver, while set value 1 to use amdgpu driver. The default is using radeon driver when it available, + * otherwise using amdgpu driver. + */ #ifdef CONFIG_DRM_AMDGPU_CIK #if defined(CONFIG_DRM_RADEON) || defined(CONFIG_DRM_RADEON_MODULE) @@ -318,6 +521,11 @@ MODULE_PARM_DESC(cik_support, "CIK support (1 = enabled (default), 0 = disabled) module_param_named(cik_support, amdgpu_cik_support, int, 0444); #endif +/** + * DOC: smu_memory_pool_size (uint) + * It is used to reserve gtt for smu debug usage, setting value 0 to disable it. The actual size is value * 256MiB. + * E.g. 0x1 = 256Mbyte, 0x2 = 512Mbyte, 0x4 = 1 Gbyte, 0x8 = 2GByte. The default is 0 (disabled). + */ MODULE_PARM_DESC(smu_memory_pool_size, "reserve gtt for smu debug usage, 0 = disable," "0x1 = 256Mbyte, 0x2 = 512Mbyte, 0x4 = 1 Gbyte, 0x8 = 2GByte"); @@ -664,7 +872,7 @@ retry_init: err_pci: pci_disable_device(pdev); err_free: - drm_dev_unref(dev); + drm_dev_put(dev); return ret; } @@ -674,7 +882,7 @@ amdgpu_pci_remove(struct pci_dev *pdev) struct drm_device *dev = pci_get_drvdata(pdev); drm_dev_unregister(dev); - drm_dev_unref(dev); + drm_dev_put(dev); pci_disable_device(pdev); pci_set_drvdata(pdev, NULL); } @@ -855,9 +1063,21 @@ static const struct dev_pm_ops amdgpu_pm_ops = { .runtime_idle = amdgpu_pmops_runtime_idle, }; +static int amdgpu_flush(struct file *f, fl_owner_t id) +{ + struct drm_file *file_priv = f->private_data; + struct amdgpu_fpriv *fpriv = file_priv->driver_priv; + + amdgpu_ctx_mgr_entity_flush(&fpriv->ctx_mgr); + + return 0; +} + + static const struct file_operations amdgpu_driver_kms_fops = { .owner = THIS_MODULE, .open = drm_open, + .flush = amdgpu_flush, .release = drm_release, .unlocked_ioctl = amdgpu_drm_ioctl, .mmap = amdgpu_mmap, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_encoders.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_encoders.c index 94138abe093b..ae8fac34f7a5 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_encoders.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_encoders.c @@ -46,7 +46,7 @@ amdgpu_link_encoder_connector(struct drm_device *dev) list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) { amdgpu_encoder = to_amdgpu_encoder(encoder); if (amdgpu_encoder->devices & amdgpu_connector->devices) { - drm_mode_connector_attach_encoder(connector, encoder); + drm_connector_attach_encoder(connector, encoder); if (amdgpu_encoder->devices & (ATOM_DEVICE_LCD_SUPPORT)) { amdgpu_atombios_encoder_init_backlight(amdgpu_encoder, connector); adev->mode_info.bl_encoder = amdgpu_encoder; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c index bc5fd8ebab5d..69c5d22f29bd 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c @@ -146,7 +146,7 @@ static int amdgpufb_create_pinned_object(struct amdgpu_fbdev *rfbdev, AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED | AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS | AMDGPU_GEM_CREATE_VRAM_CLEARED, - true, NULL, &gobj); + ttm_bo_type_kernel, NULL, &gobj); if (ret) { pr_err("failed to allocate framebuffer (%d)\n", aligned_size); return -ENOMEM; @@ -168,11 +168,19 @@ static int amdgpufb_create_pinned_object(struct amdgpu_fbdev *rfbdev, } - ret = amdgpu_bo_pin(abo, domain, NULL); + ret = amdgpu_bo_pin(abo, domain); if (ret) { amdgpu_bo_unreserve(abo); goto out_unref; } + + ret = amdgpu_ttm_alloc_gart(&abo->tbo); + if (ret) { + amdgpu_bo_unreserve(abo); + dev_err(adev->dev, "%p bind failed\n", abo); + goto out_unref; + } + ret = amdgpu_bo_kmap(abo, NULL); amdgpu_bo_unreserve(abo); if (ret) { @@ -365,8 +373,8 @@ void amdgpu_fbdev_fini(struct amdgpu_device *adev) void amdgpu_fbdev_set_suspend(struct amdgpu_device *adev, int state) { if (adev->mode_info.rfbdev) - drm_fb_helper_set_suspend(&adev->mode_info.rfbdev->helper, - state); + drm_fb_helper_set_suspend_unlocked(&adev->mode_info.rfbdev->helper, + state); } int amdgpu_fbdev_total_size(struct amdgpu_device *adev) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c index e74d620d9699..7056925eb386 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c @@ -646,7 +646,6 @@ static const struct dma_fence_ops amdgpu_fence_ops = { .get_driver_name = amdgpu_fence_get_driver_name, .get_timeline_name = amdgpu_fence_get_timeline_name, .enable_signaling = amdgpu_fence_enable_signaling, - .wait = dma_fence_default_wait, .release = amdgpu_fence_release, }; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c index dd11b7313ca0..a54d5655a191 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c @@ -143,14 +143,12 @@ int amdgpu_gart_table_vram_alloc(struct amdgpu_device *adev) */ int amdgpu_gart_table_vram_pin(struct amdgpu_device *adev) { - uint64_t gpu_addr; int r; r = amdgpu_bo_reserve(adev->gart.robj, false); if (unlikely(r != 0)) return r; - r = amdgpu_bo_pin(adev->gart.robj, - AMDGPU_GEM_DOMAIN_VRAM, &gpu_addr); + r = amdgpu_bo_pin(adev->gart.robj, AMDGPU_GEM_DOMAIN_VRAM); if (r) { amdgpu_bo_unreserve(adev->gart.robj); return r; @@ -159,7 +157,7 @@ int amdgpu_gart_table_vram_pin(struct amdgpu_device *adev) if (r) amdgpu_bo_unpin(adev->gart.robj); amdgpu_bo_unreserve(adev->gart.robj); - adev->gart.table_addr = gpu_addr; + adev->gart.table_addr = amdgpu_bo_gpu_offset(adev->gart.robj); return r; } @@ -234,7 +232,7 @@ int amdgpu_gart_unbind(struct amdgpu_device *adev, uint64_t offset, } t = offset / AMDGPU_GPU_PAGE_SIZE; - p = t / (PAGE_SIZE / AMDGPU_GPU_PAGE_SIZE); + p = t / AMDGPU_GPU_PAGES_IN_CPU_PAGE; for (i = 0; i < pages; i++, p++) { #ifdef CONFIG_DRM_AMDGPU_GART_DEBUGFS adev->gart.pages[p] = NULL; @@ -243,7 +241,7 @@ int amdgpu_gart_unbind(struct amdgpu_device *adev, uint64_t offset, if (!adev->gart.ptr) continue; - for (j = 0; j < (PAGE_SIZE / AMDGPU_GPU_PAGE_SIZE); j++, t++) { + for (j = 0; j < AMDGPU_GPU_PAGES_IN_CPU_PAGE; j++, t++) { amdgpu_gmc_set_pte_pde(adev, adev->gart.ptr, t, page_base, flags); page_base += AMDGPU_GPU_PAGE_SIZE; @@ -282,7 +280,7 @@ int amdgpu_gart_map(struct amdgpu_device *adev, uint64_t offset, for (i = 0; i < pages; i++) { page_base = dma_addr[i]; - for (j = 0; j < (PAGE_SIZE / AMDGPU_GPU_PAGE_SIZE); j++, t++) { + for (j = 0; j < AMDGPU_GPU_PAGES_IN_CPU_PAGE; j++, t++) { amdgpu_gmc_set_pte_pde(adev, dst, t, page_base, flags); page_base += AMDGPU_GPU_PAGE_SIZE; } @@ -319,7 +317,7 @@ int amdgpu_gart_bind(struct amdgpu_device *adev, uint64_t offset, #ifdef CONFIG_DRM_AMDGPU_GART_DEBUGFS t = offset / AMDGPU_GPU_PAGE_SIZE; - p = t / (PAGE_SIZE / AMDGPU_GPU_PAGE_SIZE); + p = t / AMDGPU_GPU_PAGES_IN_CPU_PAGE; for (i = 0; i < pages; i++, p++) adev->gart.pages[p] = pagelist ? pagelist[i] : NULL; #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.h index 456295c00291..9f9e9dc87da1 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.h @@ -37,6 +37,8 @@ struct amdgpu_bo; #define AMDGPU_GPU_PAGE_SHIFT 12 #define AMDGPU_GPU_PAGE_ALIGN(a) (((a) + AMDGPU_GPU_PAGE_MASK) & ~AMDGPU_GPU_PAGE_MASK) +#define AMDGPU_GPU_PAGES_IN_CPU_PAGE (PAGE_SIZE / AMDGPU_GPU_PAGE_SIZE) + struct amdgpu_gart { u64 table_addr; struct amdgpu_bo *robj; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c index 5fb156a01774..71792d820ae0 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c @@ -265,7 +265,7 @@ int amdgpu_gem_create_ioctl(struct drm_device *dev, void *data, r = amdgpu_gem_object_create(adev, size, args->in.alignment, (u32)(0xffffffff & args->in.domains), - flags, false, resv, &gobj); + flags, ttm_bo_type_device, resv, &gobj); if (flags & AMDGPU_GEM_CREATE_VM_ALWAYS_VALID) { if (!r) { struct amdgpu_bo *abo = gem_to_amdgpu_bo(gobj); @@ -317,7 +317,7 @@ int amdgpu_gem_userptr_ioctl(struct drm_device *dev, void *data, /* create a gem object to contain this object in */ r = amdgpu_gem_object_create(adev, args->size, 0, AMDGPU_GEM_DOMAIN_CPU, - 0, 0, NULL, &gobj); + 0, ttm_bo_type_device, NULL, &gobj); if (r) return r; @@ -344,7 +344,7 @@ int amdgpu_gem_userptr_ioctl(struct drm_device *dev, void *data, if (r) goto free_pages; - amdgpu_ttm_placement_from_domain(bo, AMDGPU_GEM_DOMAIN_GTT); + amdgpu_bo_placement_from_domain(bo, AMDGPU_GEM_DOMAIN_GTT); r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx); amdgpu_bo_unreserve(bo); if (r) @@ -510,7 +510,6 @@ out: * @adev: amdgpu_device pointer * @vm: vm to update * @bo_va: bo_va to update - * @list: validation list * @operation: map, unmap or clear * * Update the bo_va directly after setting its address. Errors are not @@ -519,7 +518,6 @@ out: static void amdgpu_gem_va_update_vm(struct amdgpu_device *adev, struct amdgpu_vm *vm, struct amdgpu_bo_va *bo_va, - struct list_head *list, uint32_t operation) { int r; @@ -612,7 +610,7 @@ int amdgpu_gem_va_ioctl(struct drm_device *dev, void *data, return -ENOENT; abo = gem_to_amdgpu_bo(gobj); tv.bo = &abo->tbo; - tv.shared = false; + tv.shared = !!(abo->flags & AMDGPU_GEM_CREATE_VM_ALWAYS_VALID); list_add(&tv.head, &list); } else { gobj = NULL; @@ -673,7 +671,7 @@ int amdgpu_gem_va_ioctl(struct drm_device *dev, void *data, break; } if (!r && !(args->flags & AMDGPU_VM_DELAY_UPDATE) && !amdgpu_vm_debug) - amdgpu_gem_va_update_vm(adev, &fpriv->vm, bo_va, &list, + amdgpu_gem_va_update_vm(adev, &fpriv->vm, bo_va, args->operation); error_backoff: @@ -768,7 +766,7 @@ int amdgpu_mode_dumb_create(struct drm_file *file_priv, amdgpu_display_supported_domains(adev)); r = amdgpu_gem_object_create(adev, args->size, 0, domain, AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED, - false, NULL, &gobj); + ttm_bo_type_device, NULL, &gobj); if (r) return -ENOMEM; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h index 893c2490b783..bb5a47a45790 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h @@ -105,8 +105,25 @@ struct amdgpu_gmc { /* protects concurrent invalidation */ spinlock_t invalidate_lock; bool translate_further; + struct kfd_vm_fault_info *vm_fault_info; + atomic_t vm_fault_info_updated; const struct amdgpu_gmc_funcs *gmc_funcs; }; +/** + * amdgpu_gmc_vram_full_visible - Check if full VRAM is visible through the BAR + * + * @adev: amdgpu_device pointer + * + * Returns: + * True if full VRAM is visible through the BAR + */ +static inline bool amdgpu_gmc_vram_full_visible(struct amdgpu_gmc *gmc) +{ + WARN_ON(gmc->real_vram_size < gmc->visible_vram_size); + + return (gmc->real_vram_size == gmc->visible_vram_size); +} + #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c index 7aaa263ad8c7..5518e623fed2 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c @@ -139,7 +139,7 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs, /* ring tests don't use a job */ if (job) { vm = job->vm; - fence_ctx = job->fence_ctx; + fence_ctx = job->base.s_fence->scheduled.context; } else { vm = NULL; fence_ctx = 0; @@ -353,7 +353,8 @@ int amdgpu_ib_ring_tests(struct amdgpu_device *adev) ring->funcs->type == AMDGPU_RING_TYPE_VCE || ring->funcs->type == AMDGPU_RING_TYPE_UVD_ENC || ring->funcs->type == AMDGPU_RING_TYPE_VCN_DEC || - ring->funcs->type == AMDGPU_RING_TYPE_VCN_ENC) + ring->funcs->type == AMDGPU_RING_TYPE_VCN_ENC || + ring->funcs->type == AMDGPU_RING_TYPE_VCN_JPEG) tmo = tmo_mm; else tmo = tmo_gfx; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.c index a1c78f90eadf..3a072a7a39f0 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.c @@ -578,11 +578,6 @@ void amdgpu_vmid_mgr_init(struct amdgpu_device *adev) list_add_tail(&id_mgr->ids[j].list, &id_mgr->ids_lru); } } - - adev->vm_manager.fence_context = - dma_fence_context_alloc(AMDGPU_MAX_RINGS); - for (i = 0; i < AMDGPU_MAX_RINGS; ++i) - adev->vm_manager.seqno[i] = 0; } /** diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c index 3a5ca462abf0..1abf5b5bac9e 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c @@ -25,6 +25,23 @@ * Alex Deucher * Jerome Glisse */ + +/** + * DOC: Interrupt Handling + * + * Interrupts generated within GPU hardware raise interrupt requests that are + * passed to amdgpu IRQ handler which is responsible for detecting source and + * type of the interrupt and dispatching matching handlers. If handling an + * interrupt requires calling kernel functions that may sleep processing is + * dispatched to work handlers. + * + * If MSI functionality is not disabled by module parameter then MSI + * support will be enabled. + * + * For GPU interrupt sources that may be driven by another driver, IRQ domain + * support is used (with mapping between virtual and hardware IRQs). + */ + #include <linux/irq.h> #include <drm/drmP.h> #include <drm/drm_crtc_helper.h> @@ -43,19 +60,21 @@ #define AMDGPU_WAIT_IDLE_TIMEOUT 200 -/* - * Handle hotplug events outside the interrupt handler proper. - */ /** - * amdgpu_hotplug_work_func - display hotplug work handler + * amdgpu_hotplug_work_func - work handler for display hotplug event * - * @work: work struct + * @work: work struct pointer * - * This is the hot plug event work handler (all asics). - * The work gets scheduled from the irq handler if there - * was a hot plug interrupt. It walks the connector table - * and calls the hotplug handler for each one, then sends - * a drm hotplug event to alert userspace. + * This is the hotplug event work handler (all ASICs). + * The work gets scheduled from the IRQ handler if there + * was a hotplug interrupt. It walks through the connector table + * and calls hotplug handler for each connector. After this, it sends + * a DRM hotplug event to alert userspace. + * + * This design approach is required in order to defer hotplug event handling + * from the IRQ handler to a work handler because hotplug handler has to use + * mutexes which cannot be locked in an IRQ handler (since &mutex_lock may + * sleep). */ static void amdgpu_hotplug_work_func(struct work_struct *work) { @@ -74,13 +93,12 @@ static void amdgpu_hotplug_work_func(struct work_struct *work) } /** - * amdgpu_irq_reset_work_func - execute gpu reset + * amdgpu_irq_reset_work_func - execute GPU reset * - * @work: work struct + * @work: work struct pointer * - * Execute scheduled gpu reset (cayman+). - * This function is called when the irq handler - * thinks we need a gpu reset. + * Execute scheduled GPU reset (Cayman+). + * This function is called when the IRQ handler thinks we need a GPU reset. */ static void amdgpu_irq_reset_work_func(struct work_struct *work) { @@ -91,7 +109,13 @@ static void amdgpu_irq_reset_work_func(struct work_struct *work) amdgpu_device_gpu_recover(adev, NULL, false); } -/* Disable *all* interrupts */ +/** + * amdgpu_irq_disable_all - disable *all* interrupts + * + * @adev: amdgpu device pointer + * + * Disable all types of interrupts from all sources. + */ void amdgpu_irq_disable_all(struct amdgpu_device *adev) { unsigned long irqflags; @@ -123,11 +147,15 @@ void amdgpu_irq_disable_all(struct amdgpu_device *adev) } /** - * amdgpu_irq_handler - irq handler + * amdgpu_irq_handler - IRQ handler + * + * @irq: IRQ number (unused) + * @arg: pointer to DRM device * - * @int irq, void *arg: args + * IRQ handler for amdgpu driver (all ASICs). * - * This is the irq handler for the amdgpu driver (all asics). + * Returns: + * result of handling the IRQ, as defined by &irqreturn_t */ irqreturn_t amdgpu_irq_handler(int irq, void *arg) { @@ -142,18 +170,18 @@ irqreturn_t amdgpu_irq_handler(int irq, void *arg) } /** - * amdgpu_msi_ok - asic specific msi checks + * amdgpu_msi_ok - check whether MSI functionality is enabled * - * @adev: amdgpu device pointer + * @adev: amdgpu device pointer (unused) + * + * Checks whether MSI functionality has been disabled via module parameter + * (all ASICs). * - * Handles asic specific MSI checks to determine if - * MSIs should be enabled on a particular chip (all asics). - * Returns true if MSIs should be enabled, false if MSIs - * should not be enabled. + * Returns: + * *true* if MSIs are allowed to be enabled or *false* otherwise */ static bool amdgpu_msi_ok(struct amdgpu_device *adev) { - /* force MSI on */ if (amdgpu_msi == 1) return true; else if (amdgpu_msi == 0) @@ -163,12 +191,15 @@ static bool amdgpu_msi_ok(struct amdgpu_device *adev) } /** - * amdgpu_irq_init - init driver interrupt info + * amdgpu_irq_init - initialize interrupt handling * * @adev: amdgpu device pointer * - * Sets up the work irq handlers, vblank init, MSIs, etc. (all asics). - * Returns 0 for success, error for failure. + * Sets up work functions for hotplug and reset interrupts, enables MSI + * functionality, initializes vblank, hotplug and reset interrupt handling. + * + * Returns: + * 0 on success or error code on failure */ int amdgpu_irq_init(struct amdgpu_device *adev) { @@ -176,7 +207,7 @@ int amdgpu_irq_init(struct amdgpu_device *adev) spin_lock_init(&adev->irq.lock); - /* enable msi */ + /* Enable MSI if not disabled by module parameter */ adev->irq.msi_enabled = false; if (amdgpu_msi_ok(adev)) { @@ -189,7 +220,7 @@ int amdgpu_irq_init(struct amdgpu_device *adev) if (!amdgpu_device_has_dc_support(adev)) { if (!adev->enable_virtual_display) - /* Disable vblank irqs aggressively for power-saving */ + /* Disable vblank IRQs aggressively for power-saving */ /* XXX: can this be enabled for DC? */ adev->ddev->vblank_disable_immediate = true; @@ -197,7 +228,7 @@ int amdgpu_irq_init(struct amdgpu_device *adev) if (r) return r; - /* pre DCE11 */ + /* Pre-DCE11 */ INIT_WORK(&adev->hotplug_work, amdgpu_hotplug_work_func); } @@ -220,11 +251,13 @@ int amdgpu_irq_init(struct amdgpu_device *adev) } /** - * amdgpu_irq_fini - tear down driver interrupt info + * amdgpu_irq_fini - shut down interrupt handling * * @adev: amdgpu device pointer * - * Tears down the work irq handlers, vblank handlers, MSIs, etc. (all asics). + * Tears down work functions for hotplug and reset interrupts, disables MSI + * functionality, shuts down vblank, hotplug and reset interrupt handling, + * turns off interrupts from all sources (all ASICs). */ void amdgpu_irq_fini(struct amdgpu_device *adev) { @@ -264,12 +297,17 @@ void amdgpu_irq_fini(struct amdgpu_device *adev) } /** - * amdgpu_irq_add_id - register irq source + * amdgpu_irq_add_id - register IRQ source * * @adev: amdgpu device pointer - * @src_id: source id for this source - * @source: irq source + * @client_id: client id + * @src_id: source id + * @source: IRQ source pointer + * + * Registers IRQ source on a client. * + * Returns: + * 0 on success or error code otherwise */ int amdgpu_irq_add_id(struct amdgpu_device *adev, unsigned client_id, unsigned src_id, @@ -312,12 +350,12 @@ int amdgpu_irq_add_id(struct amdgpu_device *adev, } /** - * amdgpu_irq_dispatch - dispatch irq to IP blocks + * amdgpu_irq_dispatch - dispatch IRQ to IP blocks * * @adev: amdgpu device pointer - * @entry: interrupt vector + * @entry: interrupt vector pointer * - * Dispatches the irq to the different IP blocks + * Dispatches IRQ to IP blocks. */ void amdgpu_irq_dispatch(struct amdgpu_device *adev, struct amdgpu_iv_entry *entry) @@ -361,13 +399,13 @@ void amdgpu_irq_dispatch(struct amdgpu_device *adev, } /** - * amdgpu_irq_update - update hw interrupt state + * amdgpu_irq_update - update hardware interrupt state * * @adev: amdgpu device pointer - * @src: interrupt src you want to enable - * @type: type of interrupt you want to update + * @src: interrupt source pointer + * @type: type of interrupt * - * Updates the interrupt state for a specific src (all asics). + * Updates interrupt state for the specific source (all ASICs). */ int amdgpu_irq_update(struct amdgpu_device *adev, struct amdgpu_irq_src *src, unsigned type) @@ -378,7 +416,7 @@ int amdgpu_irq_update(struct amdgpu_device *adev, spin_lock_irqsave(&adev->irq.lock, irqflags); - /* we need to determine after taking the lock, otherwise + /* We need to determine after taking the lock, otherwise we might disable just enabled interrupts again */ if (amdgpu_irq_enabled(adev, src, type)) state = AMDGPU_IRQ_STATE_ENABLE; @@ -390,6 +428,14 @@ int amdgpu_irq_update(struct amdgpu_device *adev, return r; } +/** + * amdgpu_irq_gpu_reset_resume_helper - update interrupt states on all sources + * + * @adev: amdgpu device pointer + * + * Updates state of all types of interrupts on all sources on resume after + * reset. + */ void amdgpu_irq_gpu_reset_resume_helper(struct amdgpu_device *adev) { int i, j, k; @@ -413,10 +459,13 @@ void amdgpu_irq_gpu_reset_resume_helper(struct amdgpu_device *adev) * amdgpu_irq_get - enable interrupt * * @adev: amdgpu device pointer - * @src: interrupt src you want to enable - * @type: type of interrupt you want to enable + * @src: interrupt source pointer + * @type: type of interrupt * - * Enables the interrupt type for a specific src (all asics). + * Enables specified type of interrupt on the specified source (all ASICs). + * + * Returns: + * 0 on success or error code otherwise */ int amdgpu_irq_get(struct amdgpu_device *adev, struct amdgpu_irq_src *src, unsigned type) @@ -440,10 +489,13 @@ int amdgpu_irq_get(struct amdgpu_device *adev, struct amdgpu_irq_src *src, * amdgpu_irq_put - disable interrupt * * @adev: amdgpu device pointer - * @src: interrupt src you want to disable - * @type: type of interrupt you want to disable + * @src: interrupt source pointer + * @type: type of interrupt + * + * Enables specified type of interrupt on the specified source (all ASICs). * - * Disables the interrupt type for a specific src (all asics). + * Returns: + * 0 on success or error code otherwise */ int amdgpu_irq_put(struct amdgpu_device *adev, struct amdgpu_irq_src *src, unsigned type) @@ -464,12 +516,17 @@ int amdgpu_irq_put(struct amdgpu_device *adev, struct amdgpu_irq_src *src, } /** - * amdgpu_irq_enabled - test if irq is enabled or not + * amdgpu_irq_enabled - check whether interrupt is enabled or not * * @adev: amdgpu device pointer - * @idx: interrupt src you want to test + * @src: interrupt source pointer + * @type: type of interrupt * - * Tests if the given interrupt source is enabled or not + * Checks whether the given type of interrupt is enabled on the given source. + * + * Returns: + * *true* if interrupt is enabled, *false* if interrupt is disabled or on + * invalid parameters */ bool amdgpu_irq_enabled(struct amdgpu_device *adev, struct amdgpu_irq_src *src, unsigned type) @@ -486,7 +543,7 @@ bool amdgpu_irq_enabled(struct amdgpu_device *adev, struct amdgpu_irq_src *src, return !!atomic_read(&src->enabled_types[type]); } -/* gen irq */ +/* XXX: Generic IRQ handling */ static void amdgpu_irq_mask(struct irq_data *irqd) { /* XXX */ @@ -497,12 +554,26 @@ static void amdgpu_irq_unmask(struct irq_data *irqd) /* XXX */ } +/* amdgpu hardware interrupt chip descriptor */ static struct irq_chip amdgpu_irq_chip = { .name = "amdgpu-ih", .irq_mask = amdgpu_irq_mask, .irq_unmask = amdgpu_irq_unmask, }; +/** + * amdgpu_irqdomain_map - create mapping between virtual and hardware IRQ numbers + * + * @d: amdgpu IRQ domain pointer (unused) + * @irq: virtual IRQ number + * @hwirq: hardware irq number + * + * Current implementation assigns simple interrupt handler to the given virtual + * IRQ. + * + * Returns: + * 0 on success or error code otherwise + */ static int amdgpu_irqdomain_map(struct irq_domain *d, unsigned int irq, irq_hw_number_t hwirq) { @@ -514,17 +585,21 @@ static int amdgpu_irqdomain_map(struct irq_domain *d, return 0; } +/* Implementation of methods for amdgpu IRQ domain */ static const struct irq_domain_ops amdgpu_hw_irqdomain_ops = { .map = amdgpu_irqdomain_map, }; /** - * amdgpu_irq_add_domain - create a linear irq domain + * amdgpu_irq_add_domain - create a linear IRQ domain * * @adev: amdgpu device pointer * - * Create an irq domain for GPU interrupt sources + * Creates an IRQ domain for GPU interrupt sources * that may be driven by another driver (e.g., ACP). + * + * Returns: + * 0 on success or error code otherwise */ int amdgpu_irq_add_domain(struct amdgpu_device *adev) { @@ -539,11 +614,11 @@ int amdgpu_irq_add_domain(struct amdgpu_device *adev) } /** - * amdgpu_irq_remove_domain - remove the irq domain + * amdgpu_irq_remove_domain - remove the IRQ domain * * @adev: amdgpu device pointer * - * Remove the irq domain for GPU interrupt sources + * Removes the IRQ domain for GPU interrupt sources * that may be driven by another driver (e.g., ACP). */ void amdgpu_irq_remove_domain(struct amdgpu_device *adev) @@ -555,16 +630,17 @@ void amdgpu_irq_remove_domain(struct amdgpu_device *adev) } /** - * amdgpu_irq_create_mapping - create a mapping between a domain irq and a - * Linux irq + * amdgpu_irq_create_mapping - create mapping between domain Linux IRQs * * @adev: amdgpu device pointer * @src_id: IH source id * - * Create a mapping between a domain irq (GPU IH src id) and a Linux irq + * Creates mapping between a domain IRQ (GPU IH src id) and a Linux IRQ * Use this for components that generate a GPU interrupt, but are driven * by a different driver (e.g., ACP). - * Returns the Linux irq. + * + * Returns: + * Linux IRQ */ unsigned amdgpu_irq_create_mapping(struct amdgpu_device *adev, unsigned src_id) { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c index 2bd56760c744..391e2f7c03aa 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c @@ -30,14 +30,14 @@ static void amdgpu_job_timedout(struct drm_sched_job *s_job) { - struct amdgpu_job *job = container_of(s_job, struct amdgpu_job, base); + struct amdgpu_ring *ring = to_amdgpu_ring(s_job->sched); + struct amdgpu_job *job = to_amdgpu_job(s_job); - DRM_ERROR("ring %s timeout, last signaled seq=%u, last emitted seq=%u\n", - job->base.sched->name, - atomic_read(&job->ring->fence_drv.last_seq), - job->ring->fence_drv.sync_seq); + DRM_ERROR("ring %s timeout, signaled seq=%u, emitted seq=%u\n", + job->base.sched->name, atomic_read(&ring->fence_drv.last_seq), + ring->fence_drv.sync_seq); - amdgpu_device_gpu_recover(job->adev, job, false); + amdgpu_device_gpu_recover(ring->adev, job, false); } int amdgpu_job_alloc(struct amdgpu_device *adev, unsigned num_ibs, @@ -54,7 +54,11 @@ int amdgpu_job_alloc(struct amdgpu_device *adev, unsigned num_ibs, if (!*job) return -ENOMEM; - (*job)->adev = adev; + /* + * Initialize the scheduler to at least some ring so that we always + * have a pointer to adev. + */ + (*job)->base.sched = &adev->rings[0]->sched; (*job)->vm = vm; (*job)->ibs = (void *)&(*job)[1]; (*job)->num_ibs = num_ibs; @@ -86,6 +90,7 @@ int amdgpu_job_alloc_with_ib(struct amdgpu_device *adev, unsigned size, void amdgpu_job_free_resources(struct amdgpu_job *job) { + struct amdgpu_ring *ring = to_amdgpu_ring(job->base.sched); struct dma_fence *f; unsigned i; @@ -93,14 +98,15 @@ void amdgpu_job_free_resources(struct amdgpu_job *job) f = job->base.s_fence ? &job->base.s_fence->finished : job->fence; for (i = 0; i < job->num_ibs; ++i) - amdgpu_ib_free(job->adev, &job->ibs[i], f); + amdgpu_ib_free(ring->adev, &job->ibs[i], f); } static void amdgpu_job_free_cb(struct drm_sched_job *s_job) { - struct amdgpu_job *job = container_of(s_job, struct amdgpu_job, base); + struct amdgpu_ring *ring = to_amdgpu_ring(s_job->sched); + struct amdgpu_job *job = to_amdgpu_job(s_job); - amdgpu_ring_priority_put(job->ring, s_job->s_priority); + amdgpu_ring_priority_put(ring, s_job->s_priority); dma_fence_put(job->fence); amdgpu_sync_free(&job->sync); amdgpu_sync_free(&job->sched_sync); @@ -117,50 +123,68 @@ void amdgpu_job_free(struct amdgpu_job *job) kfree(job); } -int amdgpu_job_submit(struct amdgpu_job *job, struct amdgpu_ring *ring, - struct drm_sched_entity *entity, void *owner, - struct dma_fence **f) +int amdgpu_job_submit(struct amdgpu_job *job, struct drm_sched_entity *entity, + void *owner, struct dma_fence **f) { + enum drm_sched_priority priority; + struct amdgpu_ring *ring; int r; - job->ring = ring; if (!f) return -EINVAL; - r = drm_sched_job_init(&job->base, &ring->sched, entity, owner); + r = drm_sched_job_init(&job->base, entity, owner); if (r) return r; job->owner = owner; - job->fence_ctx = entity->fence_context; *f = dma_fence_get(&job->base.s_fence->finished); amdgpu_job_free_resources(job); - amdgpu_ring_priority_get(job->ring, job->base.s_priority); + priority = job->base.s_priority; drm_sched_entity_push_job(&job->base, entity); + ring = to_amdgpu_ring(entity->rq->sched); + amdgpu_ring_priority_get(ring, priority); + + return 0; +} + +int amdgpu_job_submit_direct(struct amdgpu_job *job, struct amdgpu_ring *ring, + struct dma_fence **fence) +{ + int r; + + job->base.sched = &ring->sched; + r = amdgpu_ib_schedule(ring, job->num_ibs, job->ibs, NULL, fence); + job->fence = dma_fence_get(*fence); + if (r) + return r; + + amdgpu_job_free(job); return 0; } static struct dma_fence *amdgpu_job_dependency(struct drm_sched_job *sched_job, struct drm_sched_entity *s_entity) { + struct amdgpu_ring *ring = to_amdgpu_ring(s_entity->rq->sched); struct amdgpu_job *job = to_amdgpu_job(sched_job); struct amdgpu_vm *vm = job->vm; + struct dma_fence *fence; bool explicit = false; int r; - struct dma_fence *fence = amdgpu_sync_get_fence(&job->sync, &explicit); + fence = amdgpu_sync_get_fence(&job->sync, &explicit); if (fence && explicit) { if (drm_sched_dependency_optimized(fence, s_entity)) { - r = amdgpu_sync_fence(job->adev, &job->sched_sync, fence, false); + r = amdgpu_sync_fence(ring->adev, &job->sched_sync, + fence, false); if (r) - DRM_ERROR("Error adding fence to sync (%d)\n", r); + DRM_ERROR("Error adding fence (%d)\n", r); } } while (fence == NULL && vm && !job->vmid) { - struct amdgpu_ring *ring = job->ring; - r = amdgpu_vmid_grab(vm, ring, &job->sync, &job->base.s_fence->finished, job); @@ -175,30 +199,25 @@ static struct dma_fence *amdgpu_job_dependency(struct drm_sched_job *sched_job, static struct dma_fence *amdgpu_job_run(struct drm_sched_job *sched_job) { + struct amdgpu_ring *ring = to_amdgpu_ring(sched_job->sched); struct dma_fence *fence = NULL, *finished; - struct amdgpu_device *adev; struct amdgpu_job *job; int r; - if (!sched_job) { - DRM_ERROR("job is null\n"); - return NULL; - } job = to_amdgpu_job(sched_job); finished = &job->base.s_fence->finished; - adev = job->adev; BUG_ON(amdgpu_sync_peek_fence(&job->sync, NULL)); trace_amdgpu_sched_run_job(job); - if (job->vram_lost_counter != atomic_read(&adev->vram_lost_counter)) + if (job->vram_lost_counter != atomic_read(&ring->adev->vram_lost_counter)) dma_fence_set_error(finished, -ECANCELED);/* skip IB as well if VRAM lost */ if (finished->error < 0) { DRM_INFO("Skip scheduling IBs!\n"); } else { - r = amdgpu_ib_schedule(job->ring, job->num_ibs, job->ibs, job, + r = amdgpu_ib_schedule(ring, job->num_ibs, job->ibs, job, &fence); if (r) DRM_ERROR("Error scheduling IBs (%d)\n", r); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_job.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.h new file mode 100644 index 000000000000..57cfe78a262b --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.h @@ -0,0 +1,74 @@ +/* + * Copyright 2018 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ +#ifndef __AMDGPU_JOB_H__ +#define __AMDGPU_JOB_H__ + +/* bit set means command submit involves a preamble IB */ +#define AMDGPU_PREAMBLE_IB_PRESENT (1 << 0) +/* bit set means preamble IB is first presented in belonging context */ +#define AMDGPU_PREAMBLE_IB_PRESENT_FIRST (1 << 1) +/* bit set means context switch occured */ +#define AMDGPU_HAVE_CTX_SWITCH (1 << 2) + +#define to_amdgpu_job(sched_job) \ + container_of((sched_job), struct amdgpu_job, base) + +struct amdgpu_fence; + +struct amdgpu_job { + struct drm_sched_job base; + struct amdgpu_vm *vm; + struct amdgpu_sync sync; + struct amdgpu_sync sched_sync; + struct amdgpu_ib *ibs; + struct dma_fence *fence; /* the hw fence */ + uint32_t preamble_status; + uint32_t num_ibs; + void *owner; + bool vm_needs_flush; + uint64_t vm_pd_addr; + unsigned vmid; + unsigned pasid; + uint32_t gds_base, gds_size; + uint32_t gws_base, gws_size; + uint32_t oa_base, oa_size; + uint32_t vram_lost_counter; + + /* user fence handling */ + uint64_t uf_addr; + uint64_t uf_sequence; + +}; + +int amdgpu_job_alloc(struct amdgpu_device *adev, unsigned num_ibs, + struct amdgpu_job **job, struct amdgpu_vm *vm); +int amdgpu_job_alloc_with_ib(struct amdgpu_device *adev, unsigned size, + struct amdgpu_job **job); + +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 drm_sched_entity *entity, + void *owner, struct dma_fence **f); +int amdgpu_job_submit_direct(struct amdgpu_job *job, struct amdgpu_ring *ring, + struct dma_fence **fence); +#endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c index 91517b166a3b..bd98cc5fb97b 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c @@ -328,61 +328,71 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file case AMDGPU_HW_IP_GFX: type = AMD_IP_BLOCK_TYPE_GFX; for (i = 0; i < adev->gfx.num_gfx_rings; i++) - ring_mask |= ((adev->gfx.gfx_ring[i].ready ? 1 : 0) << i); - ib_start_alignment = AMDGPU_GPU_PAGE_SIZE; - ib_size_alignment = 8; + ring_mask |= adev->gfx.gfx_ring[i].ready << i; + ib_start_alignment = 32; + ib_size_alignment = 32; break; case AMDGPU_HW_IP_COMPUTE: type = AMD_IP_BLOCK_TYPE_GFX; for (i = 0; i < adev->gfx.num_compute_rings; i++) - ring_mask |= ((adev->gfx.compute_ring[i].ready ? 1 : 0) << i); - ib_start_alignment = AMDGPU_GPU_PAGE_SIZE; - ib_size_alignment = 8; + ring_mask |= adev->gfx.compute_ring[i].ready << i; + ib_start_alignment = 32; + ib_size_alignment = 32; break; case AMDGPU_HW_IP_DMA: type = AMD_IP_BLOCK_TYPE_SDMA; for (i = 0; i < adev->sdma.num_instances; i++) - ring_mask |= ((adev->sdma.instance[i].ring.ready ? 1 : 0) << i); - ib_start_alignment = AMDGPU_GPU_PAGE_SIZE; - ib_size_alignment = 1; + ring_mask |= adev->sdma.instance[i].ring.ready << i; + ib_start_alignment = 256; + ib_size_alignment = 4; break; case AMDGPU_HW_IP_UVD: type = AMD_IP_BLOCK_TYPE_UVD; - for (i = 0; i < adev->uvd.num_uvd_inst; i++) - ring_mask |= ((adev->uvd.inst[i].ring.ready ? 1 : 0) << i); - ib_start_alignment = AMDGPU_GPU_PAGE_SIZE; - ib_size_alignment = 16; + for (i = 0; i < adev->uvd.num_uvd_inst; i++) { + if (adev->uvd.harvest_config & (1 << i)) + continue; + ring_mask |= adev->uvd.inst[i].ring.ready; + } + ib_start_alignment = 64; + ib_size_alignment = 64; break; case AMDGPU_HW_IP_VCE: type = AMD_IP_BLOCK_TYPE_VCE; for (i = 0; i < adev->vce.num_rings; i++) - ring_mask |= ((adev->vce.ring[i].ready ? 1 : 0) << i); - ib_start_alignment = AMDGPU_GPU_PAGE_SIZE; + ring_mask |= adev->vce.ring[i].ready << i; + ib_start_alignment = 4; ib_size_alignment = 1; break; case AMDGPU_HW_IP_UVD_ENC: type = AMD_IP_BLOCK_TYPE_UVD; - for (i = 0; i < adev->uvd.num_uvd_inst; i++) + for (i = 0; i < adev->uvd.num_uvd_inst; i++) { + if (adev->uvd.harvest_config & (1 << i)) + continue; for (j = 0; j < adev->uvd.num_enc_rings; j++) - ring_mask |= - ((adev->uvd.inst[i].ring_enc[j].ready ? 1 : 0) << - (j + i * adev->uvd.num_enc_rings)); - ib_start_alignment = AMDGPU_GPU_PAGE_SIZE; - ib_size_alignment = 1; + ring_mask |= adev->uvd.inst[i].ring_enc[j].ready << j; + } + ib_start_alignment = 64; + ib_size_alignment = 64; break; case AMDGPU_HW_IP_VCN_DEC: type = AMD_IP_BLOCK_TYPE_VCN; - ring_mask = adev->vcn.ring_dec.ready ? 1 : 0; - ib_start_alignment = AMDGPU_GPU_PAGE_SIZE; + ring_mask = adev->vcn.ring_dec.ready; + ib_start_alignment = 16; ib_size_alignment = 16; break; case AMDGPU_HW_IP_VCN_ENC: type = AMD_IP_BLOCK_TYPE_VCN; for (i = 0; i < adev->vcn.num_enc_rings; i++) - ring_mask |= ((adev->vcn.ring_enc[i].ready ? 1 : 0) << i); - ib_start_alignment = AMDGPU_GPU_PAGE_SIZE; + ring_mask |= adev->vcn.ring_enc[i].ready << i; + ib_start_alignment = 64; ib_size_alignment = 1; break; + case AMDGPU_HW_IP_VCN_JPEG: + type = AMD_IP_BLOCK_TYPE_VCN; + ring_mask = adev->vcn.ring_jpeg.ready; + ib_start_alignment = 16; + ib_size_alignment = 16; + break; default: return -EINVAL; } @@ -427,6 +437,7 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file break; case AMDGPU_HW_IP_VCN_DEC: case AMDGPU_HW_IP_VCN_ENC: + case AMDGPU_HW_IP_VCN_JPEG: type = AMD_IP_BLOCK_TYPE_VCN; break; default: @@ -494,13 +505,13 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file case AMDGPU_INFO_VRAM_GTT: { struct drm_amdgpu_info_vram_gtt vram_gtt; - vram_gtt.vram_size = adev->gmc.real_vram_size; - vram_gtt.vram_size -= adev->vram_pin_size; - vram_gtt.vram_cpu_accessible_size = adev->gmc.visible_vram_size; - vram_gtt.vram_cpu_accessible_size -= (adev->vram_pin_size - adev->invisible_pin_size); + vram_gtt.vram_size = adev->gmc.real_vram_size - + atomic64_read(&adev->vram_pin_size); + vram_gtt.vram_cpu_accessible_size = adev->gmc.visible_vram_size - + atomic64_read(&adev->visible_pin_size); vram_gtt.gtt_size = adev->mman.bdev.man[TTM_PL_TT].size; vram_gtt.gtt_size *= PAGE_SIZE; - vram_gtt.gtt_size -= adev->gart_pin_size; + vram_gtt.gtt_size -= atomic64_read(&adev->gart_pin_size); return copy_to_user(out, &vram_gtt, min((size_t)size, sizeof(vram_gtt))) ? -EFAULT : 0; } @@ -509,17 +520,16 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file memset(&mem, 0, sizeof(mem)); mem.vram.total_heap_size = adev->gmc.real_vram_size; - mem.vram.usable_heap_size = - adev->gmc.real_vram_size - adev->vram_pin_size; + mem.vram.usable_heap_size = adev->gmc.real_vram_size - + atomic64_read(&adev->vram_pin_size); mem.vram.heap_usage = amdgpu_vram_mgr_usage(&adev->mman.bdev.man[TTM_PL_VRAM]); mem.vram.max_allocation = mem.vram.usable_heap_size * 3 / 4; mem.cpu_accessible_vram.total_heap_size = adev->gmc.visible_vram_size; - mem.cpu_accessible_vram.usable_heap_size = - adev->gmc.visible_vram_size - - (adev->vram_pin_size - adev->invisible_pin_size); + mem.cpu_accessible_vram.usable_heap_size = adev->gmc.visible_vram_size - + atomic64_read(&adev->visible_pin_size); mem.cpu_accessible_vram.heap_usage = amdgpu_vram_mgr_vis_usage(&adev->mman.bdev.man[TTM_PL_VRAM]); mem.cpu_accessible_vram.max_allocation = @@ -527,8 +537,8 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file mem.gtt.total_heap_size = adev->mman.bdev.man[TTM_PL_TT].size; mem.gtt.total_heap_size *= PAGE_SIZE; - mem.gtt.usable_heap_size = mem.gtt.total_heap_size - - adev->gart_pin_size; + mem.gtt.usable_heap_size = mem.gtt.total_heap_size - + atomic64_read(&adev->gart_pin_size); mem.gtt.heap_usage = amdgpu_gtt_mgr_usage(&adev->mman.bdev.man[TTM_PL_TT]); mem.gtt.max_allocation = mem.gtt.usable_heap_size * 3 / 4; @@ -930,7 +940,6 @@ void amdgpu_driver_postclose_kms(struct drm_device *dev, return; pm_runtime_get_sync(dev->dev); - amdgpu_ctx_mgr_entity_fini(&fpriv->ctx_mgr); if (adev->asic_type != CHIP_RAVEN) { amdgpu_uvd_free_handles(adev, file_priv); @@ -958,7 +967,7 @@ void amdgpu_driver_postclose_kms(struct drm_device *dev, amdgpu_bo_unref(&pd); idr_for_each_entry(&fpriv->bo_list_handles, list, handle) - amdgpu_bo_list_free(list); + amdgpu_bo_list_put(list); idr_destroy(&fpriv->bo_list_handles); mutex_destroy(&fpriv->bo_list_lock); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c index 83e344fbb50a..a365ea2383d1 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c @@ -28,6 +28,21 @@ * Christian König <christian.koenig@amd.com> */ +/** + * DOC: MMU Notifier + * + * For coherent userptr handling registers an MMU notifier to inform the driver + * about updates on the page tables of a process. + * + * When somebody tries to invalidate the page tables we block the update until + * all operations on the pages in question are completed, then those pages are + * marked as accessed and also dirty if it wasn't a read only access. + * + * New command submissions using the userptrs in question are delayed until all + * page table invalidation are completed and we once more see a coherent process + * address space. + */ + #include <linux/firmware.h> #include <linux/module.h> #include <linux/mmu_notifier.h> @@ -38,6 +53,22 @@ #include "amdgpu.h" #include "amdgpu_amdkfd.h" +/** + * struct amdgpu_mn + * + * @adev: amdgpu device pointer + * @mm: process address space + * @mn: MMU notifier structure + * @type: type of MMU notifier + * @work: destruction work item + * @node: hash table node to find structure by adev and mn + * @lock: rw semaphore protecting the notifier nodes + * @objects: interval tree containing amdgpu_mn_nodes + * @read_lock: mutex for recursive locking of @lock + * @recursion: depth of recursion + * + * Data for each amdgpu device and process address space. + */ struct amdgpu_mn { /* constant after initialisation */ struct amdgpu_device *adev; @@ -58,13 +89,21 @@ struct amdgpu_mn { atomic_t recursion; }; +/** + * struct amdgpu_mn_node + * + * @it: interval node defining start-last of the affected address range + * @bos: list of all BOs in the affected address range + * + * Manages all BOs which are affected of a certain range of address space. + */ struct amdgpu_mn_node { struct interval_tree_node it; struct list_head bos; }; /** - * amdgpu_mn_destroy - destroy the rmn + * amdgpu_mn_destroy - destroy the MMU notifier * * @work: previously sheduled work item * @@ -72,47 +111,50 @@ struct amdgpu_mn_node { */ static void amdgpu_mn_destroy(struct work_struct *work) { - struct amdgpu_mn *rmn = container_of(work, struct amdgpu_mn, work); - struct amdgpu_device *adev = rmn->adev; + struct amdgpu_mn *amn = container_of(work, struct amdgpu_mn, work); + struct amdgpu_device *adev = amn->adev; struct amdgpu_mn_node *node, *next_node; struct amdgpu_bo *bo, *next_bo; mutex_lock(&adev->mn_lock); - down_write(&rmn->lock); - hash_del(&rmn->node); + down_write(&amn->lock); + hash_del(&amn->node); rbtree_postorder_for_each_entry_safe(node, next_node, - &rmn->objects.rb_root, it.rb) { + &amn->objects.rb_root, it.rb) { list_for_each_entry_safe(bo, next_bo, &node->bos, mn_list) { bo->mn = NULL; list_del_init(&bo->mn_list); } kfree(node); } - up_write(&rmn->lock); + up_write(&amn->lock); mutex_unlock(&adev->mn_lock); - mmu_notifier_unregister_no_release(&rmn->mn, rmn->mm); - kfree(rmn); + mmu_notifier_unregister_no_release(&amn->mn, amn->mm); + kfree(amn); } /** * amdgpu_mn_release - callback to notify about mm destruction * * @mn: our notifier - * @mn: the mm this callback is about + * @mm: the mm this callback is about * * Shedule a work item to lazy destroy our notifier. */ static void amdgpu_mn_release(struct mmu_notifier *mn, struct mm_struct *mm) { - struct amdgpu_mn *rmn = container_of(mn, struct amdgpu_mn, mn); - INIT_WORK(&rmn->work, amdgpu_mn_destroy); - schedule_work(&rmn->work); + struct amdgpu_mn *amn = container_of(mn, struct amdgpu_mn, mn); + + INIT_WORK(&amn->work, amdgpu_mn_destroy); + schedule_work(&amn->work); } /** - * amdgpu_mn_lock - take the write side lock for this mn + * amdgpu_mn_lock - take the write side lock for this notifier + * + * @mn: our notifier */ void amdgpu_mn_lock(struct amdgpu_mn *mn) { @@ -121,7 +163,9 @@ void amdgpu_mn_lock(struct amdgpu_mn *mn) } /** - * amdgpu_mn_unlock - drop the write side lock for this mn + * amdgpu_mn_unlock - drop the write side lock for this notifier + * + * @mn: our notifier */ void amdgpu_mn_unlock(struct amdgpu_mn *mn) { @@ -130,40 +174,38 @@ void amdgpu_mn_unlock(struct amdgpu_mn *mn) } /** - * amdgpu_mn_read_lock - take the rmn read lock - * - * @rmn: our notifier + * amdgpu_mn_read_lock - take the read side lock for this notifier * - * Take the rmn read side lock. + * @amn: our notifier */ -static void amdgpu_mn_read_lock(struct amdgpu_mn *rmn) +static void amdgpu_mn_read_lock(struct amdgpu_mn *amn) { - mutex_lock(&rmn->read_lock); - if (atomic_inc_return(&rmn->recursion) == 1) - down_read_non_owner(&rmn->lock); - mutex_unlock(&rmn->read_lock); + mutex_lock(&amn->read_lock); + if (atomic_inc_return(&amn->recursion) == 1) + down_read_non_owner(&amn->lock); + mutex_unlock(&amn->read_lock); } /** - * amdgpu_mn_read_unlock - drop the rmn read lock + * amdgpu_mn_read_unlock - drop the read side lock for this notifier * - * @rmn: our notifier - * - * Drop the rmn read side lock. + * @amn: our notifier */ -static void amdgpu_mn_read_unlock(struct amdgpu_mn *rmn) +static void amdgpu_mn_read_unlock(struct amdgpu_mn *amn) { - if (atomic_dec_return(&rmn->recursion) == 0) - up_read_non_owner(&rmn->lock); + if (atomic_dec_return(&amn->recursion) == 0) + up_read_non_owner(&amn->lock); } /** * amdgpu_mn_invalidate_node - unmap all BOs of a node * * @node: the node with the BOs to unmap + * @start: start of address range affected + * @end: end of address range affected * - * We block for all BOs and unmap them by move them - * into system domain again. + * Block for operations on BOs to finish and mark pages as accessed and + * potentially dirty. */ static void amdgpu_mn_invalidate_node(struct amdgpu_mn_node *node, unsigned long start, @@ -190,27 +232,27 @@ static void amdgpu_mn_invalidate_node(struct amdgpu_mn_node *node, * amdgpu_mn_invalidate_range_start_gfx - callback to notify about mm change * * @mn: our notifier - * @mn: the mm this callback is about + * @mm: the mm this callback is about * @start: start of updated range * @end: end of updated range * - * We block for all BOs between start and end to be idle and - * unmap them by move them into system domain again. + * Block for operations on BOs to finish and mark pages as accessed and + * potentially dirty. */ static void amdgpu_mn_invalidate_range_start_gfx(struct mmu_notifier *mn, struct mm_struct *mm, unsigned long start, unsigned long end) { - struct amdgpu_mn *rmn = container_of(mn, struct amdgpu_mn, mn); + struct amdgpu_mn *amn = container_of(mn, struct amdgpu_mn, mn); struct interval_tree_node *it; /* notification is exclusive, but interval is inclusive */ end -= 1; - amdgpu_mn_read_lock(rmn); + amdgpu_mn_read_lock(amn); - it = interval_tree_iter_first(&rmn->objects, start, end); + it = interval_tree_iter_first(&amn->objects, start, end); while (it) { struct amdgpu_mn_node *node; @@ -225,7 +267,7 @@ static void amdgpu_mn_invalidate_range_start_gfx(struct mmu_notifier *mn, * amdgpu_mn_invalidate_range_start_hsa - callback to notify about mm change * * @mn: our notifier - * @mn: the mm this callback is about + * @mm: the mm this callback is about * @start: start of updated range * @end: end of updated range * @@ -238,15 +280,15 @@ static void amdgpu_mn_invalidate_range_start_hsa(struct mmu_notifier *mn, unsigned long start, unsigned long end) { - struct amdgpu_mn *rmn = container_of(mn, struct amdgpu_mn, mn); + struct amdgpu_mn *amn = container_of(mn, struct amdgpu_mn, mn); struct interval_tree_node *it; /* notification is exclusive, but interval is inclusive */ end -= 1; - amdgpu_mn_read_lock(rmn); + amdgpu_mn_read_lock(amn); - it = interval_tree_iter_first(&rmn->objects, start, end); + it = interval_tree_iter_first(&amn->objects, start, end); while (it) { struct amdgpu_mn_node *node; struct amdgpu_bo *bo; @@ -268,7 +310,7 @@ static void amdgpu_mn_invalidate_range_start_hsa(struct mmu_notifier *mn, * amdgpu_mn_invalidate_range_end - callback to notify about mm change * * @mn: our notifier - * @mn: the mm this callback is about + * @mm: the mm this callback is about * @start: start of updated range * @end: end of updated range * @@ -279,9 +321,9 @@ static void amdgpu_mn_invalidate_range_end(struct mmu_notifier *mn, unsigned long start, unsigned long end) { - struct amdgpu_mn *rmn = container_of(mn, struct amdgpu_mn, mn); + struct amdgpu_mn *amn = container_of(mn, struct amdgpu_mn, mn); - amdgpu_mn_read_unlock(rmn); + amdgpu_mn_read_unlock(amn); } static const struct mmu_notifier_ops amdgpu_mn_ops[] = { @@ -315,7 +357,7 @@ struct amdgpu_mn *amdgpu_mn_get(struct amdgpu_device *adev, enum amdgpu_mn_type type) { struct mm_struct *mm = current->mm; - struct amdgpu_mn *rmn; + struct amdgpu_mn *amn; unsigned long key = AMDGPU_MN_KEY(mm, type); int r; @@ -325,41 +367,41 @@ struct amdgpu_mn *amdgpu_mn_get(struct amdgpu_device *adev, return ERR_PTR(-EINTR); } - hash_for_each_possible(adev->mn_hash, rmn, node, key) - if (AMDGPU_MN_KEY(rmn->mm, rmn->type) == key) + hash_for_each_possible(adev->mn_hash, amn, node, key) + if (AMDGPU_MN_KEY(amn->mm, amn->type) == key) goto release_locks; - rmn = kzalloc(sizeof(*rmn), GFP_KERNEL); - if (!rmn) { - rmn = ERR_PTR(-ENOMEM); + amn = kzalloc(sizeof(*amn), GFP_KERNEL); + if (!amn) { + amn = ERR_PTR(-ENOMEM); goto release_locks; } - rmn->adev = adev; - rmn->mm = mm; - init_rwsem(&rmn->lock); - rmn->type = type; - rmn->mn.ops = &amdgpu_mn_ops[type]; - rmn->objects = RB_ROOT_CACHED; - mutex_init(&rmn->read_lock); - atomic_set(&rmn->recursion, 0); + amn->adev = adev; + amn->mm = mm; + init_rwsem(&amn->lock); + amn->type = type; + amn->mn.ops = &amdgpu_mn_ops[type]; + amn->objects = RB_ROOT_CACHED; + mutex_init(&amn->read_lock); + atomic_set(&amn->recursion, 0); - r = __mmu_notifier_register(&rmn->mn, mm); + r = __mmu_notifier_register(&amn->mn, mm); if (r) - goto free_rmn; + goto free_amn; - hash_add(adev->mn_hash, &rmn->node, AMDGPU_MN_KEY(mm, type)); + hash_add(adev->mn_hash, &amn->node, AMDGPU_MN_KEY(mm, type)); release_locks: up_write(&mm->mmap_sem); mutex_unlock(&adev->mn_lock); - return rmn; + return amn; -free_rmn: +free_amn: up_write(&mm->mmap_sem); mutex_unlock(&adev->mn_lock); - kfree(rmn); + kfree(amn); return ERR_PTR(r); } @@ -379,14 +421,14 @@ int amdgpu_mn_register(struct amdgpu_bo *bo, unsigned long addr) struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev); enum amdgpu_mn_type type = bo->kfd_bo ? AMDGPU_MN_TYPE_HSA : AMDGPU_MN_TYPE_GFX; - struct amdgpu_mn *rmn; + struct amdgpu_mn *amn; struct amdgpu_mn_node *node = NULL, *new_node; struct list_head bos; struct interval_tree_node *it; - rmn = amdgpu_mn_get(adev, type); - if (IS_ERR(rmn)) - return PTR_ERR(rmn); + amn = amdgpu_mn_get(adev, type); + if (IS_ERR(amn)) + return PTR_ERR(amn); new_node = kmalloc(sizeof(*new_node), GFP_KERNEL); if (!new_node) @@ -394,12 +436,12 @@ int amdgpu_mn_register(struct amdgpu_bo *bo, unsigned long addr) INIT_LIST_HEAD(&bos); - down_write(&rmn->lock); + down_write(&amn->lock); - while ((it = interval_tree_iter_first(&rmn->objects, addr, end))) { + while ((it = interval_tree_iter_first(&amn->objects, addr, end))) { kfree(node); node = container_of(it, struct amdgpu_mn_node, it); - interval_tree_remove(&node->it, &rmn->objects); + interval_tree_remove(&node->it, &amn->objects); addr = min(it->start, addr); end = max(it->last, end); list_splice(&node->bos, &bos); @@ -410,7 +452,7 @@ int amdgpu_mn_register(struct amdgpu_bo *bo, unsigned long addr) else kfree(new_node); - bo->mn = rmn; + bo->mn = amn; node->it.start = addr; node->it.last = end; @@ -418,9 +460,9 @@ int amdgpu_mn_register(struct amdgpu_bo *bo, unsigned long addr) list_splice(&bos, &node->bos); list_add(&bo->mn_list, &node->bos); - interval_tree_insert(&node->it, &rmn->objects); + interval_tree_insert(&node->it, &amn->objects); - up_write(&rmn->lock); + up_write(&amn->lock); return 0; } @@ -435,18 +477,18 @@ int amdgpu_mn_register(struct amdgpu_bo *bo, unsigned long addr) void amdgpu_mn_unregister(struct amdgpu_bo *bo) { struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev); - struct amdgpu_mn *rmn; + struct amdgpu_mn *amn; struct list_head *head; mutex_lock(&adev->mn_lock); - rmn = bo->mn; - if (rmn == NULL) { + amn = bo->mn; + if (amn == NULL) { mutex_unlock(&adev->mn_lock); return; } - down_write(&rmn->lock); + down_write(&amn->lock); /* save the next list entry for later */ head = bo->mn_list.next; @@ -456,12 +498,13 @@ void amdgpu_mn_unregister(struct amdgpu_bo *bo) if (list_empty(head)) { struct amdgpu_mn_node *node; + node = container_of(head, struct amdgpu_mn_node, bos); - interval_tree_remove(&node->it, &rmn->objects); + interval_tree_remove(&node->it, &amn->objects); kfree(node); } - up_write(&rmn->lock); + up_write(&amn->lock); mutex_unlock(&adev->mn_lock); } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c index 3526efa8960e..b0e14a3d54ef 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c @@ -38,7 +38,20 @@ #include "amdgpu_trace.h" #include "amdgpu_amdkfd.h" -static bool amdgpu_need_backup(struct amdgpu_device *adev) +/** + * DOC: amdgpu_object + * + * This defines the interfaces to operate on an &amdgpu_bo buffer object which + * represents memory used by driver (VRAM, system memory, etc.). The driver + * provides DRM/GEM APIs to userspace. DRM/GEM APIs then use these interfaces + * to create/destroy/set buffer object which are then managed by the kernel TTM + * memory manager. + * The interfaces are also used internally by kernel clients, including gfx, + * uvd, etc. for kernel managed allocations used by the GPU. + * + */ + +static bool amdgpu_bo_need_backup(struct amdgpu_device *adev) { if (adev->flags & AMD_IS_APU) return false; @@ -50,11 +63,35 @@ static bool amdgpu_need_backup(struct amdgpu_device *adev) return true; } -static void amdgpu_ttm_bo_destroy(struct ttm_buffer_object *tbo) +/** + * amdgpu_bo_subtract_pin_size - Remove BO from pin_size accounting + * + * @bo: &amdgpu_bo buffer object + * + * This function is called when a BO stops being pinned, and updates the + * &amdgpu_device pin_size values accordingly. + */ +static void amdgpu_bo_subtract_pin_size(struct amdgpu_bo *bo) +{ + struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev); + + if (bo->tbo.mem.mem_type == TTM_PL_VRAM) { + atomic64_sub(amdgpu_bo_size(bo), &adev->vram_pin_size); + atomic64_sub(amdgpu_vram_mgr_bo_visible_size(bo), + &adev->visible_pin_size); + } else if (bo->tbo.mem.mem_type == TTM_PL_TT) { + atomic64_sub(amdgpu_bo_size(bo), &adev->gart_pin_size); + } +} + +static void amdgpu_bo_destroy(struct ttm_buffer_object *tbo) { struct amdgpu_device *adev = amdgpu_ttm_adev(tbo->bdev); struct amdgpu_bo *bo = ttm_to_amdgpu_bo(tbo); + if (bo->pin_count > 0) + amdgpu_bo_subtract_pin_size(bo); + if (bo->kfd_bo) amdgpu_amdkfd_unreserve_system_memory_limit(bo); @@ -73,14 +110,32 @@ static void amdgpu_ttm_bo_destroy(struct ttm_buffer_object *tbo) kfree(bo); } -bool amdgpu_ttm_bo_is_amdgpu_bo(struct ttm_buffer_object *bo) +/** + * amdgpu_bo_is_amdgpu_bo - check if the buffer object is an &amdgpu_bo + * @bo: buffer object to be checked + * + * Uses destroy function associated with the object to determine if this is + * an &amdgpu_bo. + * + * Returns: + * true if the object belongs to &amdgpu_bo, false if not. + */ +bool amdgpu_bo_is_amdgpu_bo(struct ttm_buffer_object *bo) { - if (bo->destroy == &amdgpu_ttm_bo_destroy) + if (bo->destroy == &amdgpu_bo_destroy) return true; return false; } -void amdgpu_ttm_placement_from_domain(struct amdgpu_bo *abo, u32 domain) +/** + * amdgpu_bo_placement_from_domain - set buffer's placement + * @abo: &amdgpu_bo buffer object whose placement is to be set + * @domain: requested domain + * + * Sets buffer's placement according to requested domain and the buffer's + * flags. + */ +void amdgpu_bo_placement_from_domain(struct amdgpu_bo *abo, u32 domain) { struct amdgpu_device *adev = amdgpu_ttm_adev(abo->tbo.bdev); struct ttm_placement *placement = &abo->placement; @@ -161,6 +216,8 @@ void amdgpu_ttm_placement_from_domain(struct amdgpu_bo *abo, u32 domain) c++; } + BUG_ON(c >= AMDGPU_BO_MAX_PLACEMENTS); + placement->num_placement = c; placement->placement = places; @@ -184,7 +241,8 @@ void amdgpu_ttm_placement_from_domain(struct amdgpu_bo *abo, u32 domain) * * Note: For bo_ptr new BO is only created if bo_ptr points to NULL. * - * Returns 0 on success, negative error code otherwise. + * Returns: + * 0 on success, negative error code otherwise. */ int amdgpu_bo_create_reserved(struct amdgpu_device *adev, unsigned long size, int align, @@ -220,22 +278,33 @@ int amdgpu_bo_create_reserved(struct amdgpu_device *adev, goto error_free; } - r = amdgpu_bo_pin(*bo_ptr, domain, gpu_addr); + r = amdgpu_bo_pin(*bo_ptr, domain); if (r) { dev_err(adev->dev, "(%d) kernel bo pin failed\n", r); goto error_unreserve; } + r = amdgpu_ttm_alloc_gart(&(*bo_ptr)->tbo); + if (r) { + dev_err(adev->dev, "%p bind failed\n", *bo_ptr); + goto error_unpin; + } + + if (gpu_addr) + *gpu_addr = amdgpu_bo_gpu_offset(*bo_ptr); + if (cpu_addr) { r = amdgpu_bo_kmap(*bo_ptr, cpu_addr); if (r) { dev_err(adev->dev, "(%d) kernel bo map failed\n", r); - goto error_unreserve; + goto error_unpin; } } return 0; +error_unpin: + amdgpu_bo_unpin(*bo_ptr); error_unreserve: amdgpu_bo_unreserve(*bo_ptr); @@ -261,7 +330,8 @@ error_free: * * Note: For bo_ptr new BO is only created if bo_ptr points to NULL. * - * Returns 0 on success, negative error code otherwise. + * Returns: + * 0 on success, negative error code otherwise. */ int amdgpu_bo_create_kernel(struct amdgpu_device *adev, unsigned long size, int align, @@ -285,6 +355,8 @@ int amdgpu_bo_create_kernel(struct amdgpu_device *adev, * amdgpu_bo_free_kernel - free BO for kernel use * * @bo: amdgpu BO to free + * @gpu_addr: pointer to where the BO's GPU memory space address was stored + * @cpu_addr: pointer to where the BO's CPU memory space address was stored * * unmaps and unpin a BO for kernel internal use. */ @@ -418,17 +490,17 @@ static int amdgpu_bo_do_create(struct amdgpu_device *adev, #endif bo->tbo.bdev = &adev->mman.bdev; - amdgpu_ttm_placement_from_domain(bo, bp->domain); + amdgpu_bo_placement_from_domain(bo, bp->domain); if (bp->type == ttm_bo_type_kernel) bo->tbo.priority = 1; r = ttm_bo_init_reserved(&adev->mman.bdev, &bo->tbo, size, bp->type, &bo->placement, page_align, &ctx, acc_size, - NULL, bp->resv, &amdgpu_ttm_bo_destroy); + NULL, bp->resv, &amdgpu_bo_destroy); if (unlikely(r != 0)) return r; - if (adev->gmc.visible_vram_size < adev->gmc.real_vram_size && + if (!amdgpu_gmc_vram_full_visible(&adev->gmc) && bo->tbo.mem.mem_type == TTM_PL_VRAM && bo->tbo.mem.start < adev->gmc.visible_vram_size >> PAGE_SHIFT) amdgpu_cs_report_moved_bytes(adev, ctx.bytes_moved, @@ -498,6 +570,20 @@ static int amdgpu_bo_create_shadow(struct amdgpu_device *adev, return r; } +/** + * amdgpu_bo_create - create an &amdgpu_bo buffer object + * @adev: amdgpu device object + * @bp: parameters to be used for the buffer object + * @bo_ptr: pointer to the buffer object pointer + * + * Creates an &amdgpu_bo buffer object; and if requested, also creates a + * shadow object. + * Shadow object is used to backup the original buffer object, and is always + * in GTT. + * + * Returns: + * 0 for success or a negative error code on failure. + */ int amdgpu_bo_create(struct amdgpu_device *adev, struct amdgpu_bo_param *bp, struct amdgpu_bo **bo_ptr) @@ -510,7 +596,7 @@ int amdgpu_bo_create(struct amdgpu_device *adev, if (r) return r; - if ((flags & AMDGPU_GEM_CREATE_SHADOW) && amdgpu_need_backup(adev)) { + if ((flags & AMDGPU_GEM_CREATE_SHADOW) && amdgpu_bo_need_backup(adev)) { if (!bp->resv) WARN_ON(reservation_object_lock((*bo_ptr)->tbo.resv, NULL)); @@ -527,6 +613,21 @@ int amdgpu_bo_create(struct amdgpu_device *adev, return r; } +/** + * amdgpu_bo_backup_to_shadow - Backs up an &amdgpu_bo buffer object + * @adev: amdgpu device object + * @ring: amdgpu_ring for the engine handling the buffer operations + * @bo: &amdgpu_bo buffer to be backed up + * @resv: reservation object with embedded fence + * @fence: dma_fence associated with the operation + * @direct: whether to submit the job directly + * + * Copies an &amdgpu_bo buffer object to its shadow object. + * Not used for now. + * + * Returns: + * 0 for success or a negative error code on failure. + */ int amdgpu_bo_backup_to_shadow(struct amdgpu_device *adev, struct amdgpu_ring *ring, struct amdgpu_bo *bo, @@ -559,6 +660,18 @@ err: return r; } +/** + * amdgpu_bo_validate - validate an &amdgpu_bo buffer object + * @bo: pointer to the buffer object + * + * Sets placement according to domain; and changes placement and caching + * policy of the buffer object according to the placement. + * This is used for validating shadow bos. It calls ttm_bo_validate() to + * make sure the buffer is resident where it needs to be. + * + * Returns: + * 0 for success or a negative error code on failure. + */ int amdgpu_bo_validate(struct amdgpu_bo *bo) { struct ttm_operation_ctx ctx = { false, false }; @@ -571,7 +684,7 @@ int amdgpu_bo_validate(struct amdgpu_bo *bo) domain = bo->preferred_domains; retry: - amdgpu_ttm_placement_from_domain(bo, domain); + amdgpu_bo_placement_from_domain(bo, domain); r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx); if (unlikely(r == -ENOMEM) && domain != bo->allowed_domains) { domain = bo->allowed_domains; @@ -581,6 +694,22 @@ retry: return r; } +/** + * amdgpu_bo_restore_from_shadow - restore an &amdgpu_bo buffer object + * @adev: amdgpu device object + * @ring: amdgpu_ring for the engine handling the buffer operations + * @bo: &amdgpu_bo buffer to be restored + * @resv: reservation object with embedded fence + * @fence: dma_fence associated with the operation + * @direct: whether to submit the job directly + * + * Copies a buffer object's shadow content back to the object. + * This is used for recovering a buffer from its shadow in case of a gpu + * reset where vram context may be lost. + * + * Returns: + * 0 for success or a negative error code on failure. + */ int amdgpu_bo_restore_from_shadow(struct amdgpu_device *adev, struct amdgpu_ring *ring, struct amdgpu_bo *bo, @@ -613,6 +742,17 @@ err: return r; } +/** + * amdgpu_bo_kmap - map an &amdgpu_bo buffer object + * @bo: &amdgpu_bo buffer object to be mapped + * @ptr: kernel virtual address to be returned + * + * Calls ttm_bo_kmap() to set up the kernel virtual mapping; calls + * amdgpu_bo_kptr() to get the kernel virtual address. + * + * Returns: + * 0 for success or a negative error code on failure. + */ int amdgpu_bo_kmap(struct amdgpu_bo *bo, void **ptr) { void *kptr; @@ -643,6 +783,15 @@ int amdgpu_bo_kmap(struct amdgpu_bo *bo, void **ptr) return 0; } +/** + * amdgpu_bo_kptr - returns a kernel virtual address of the buffer object + * @bo: &amdgpu_bo buffer object + * + * Calls ttm_kmap_obj_virtual() to get the kernel virtual address + * + * Returns: + * the virtual address of a buffer object area. + */ void *amdgpu_bo_kptr(struct amdgpu_bo *bo) { bool is_iomem; @@ -650,21 +799,42 @@ void *amdgpu_bo_kptr(struct amdgpu_bo *bo) return ttm_kmap_obj_virtual(&bo->kmap, &is_iomem); } +/** + * amdgpu_bo_kunmap - unmap an &amdgpu_bo buffer object + * @bo: &amdgpu_bo buffer object to be unmapped + * + * Unmaps a kernel map set up by amdgpu_bo_kmap(). + */ void amdgpu_bo_kunmap(struct amdgpu_bo *bo) { if (bo->kmap.bo) ttm_bo_kunmap(&bo->kmap); } +/** + * amdgpu_bo_ref - reference an &amdgpu_bo buffer object + * @bo: &amdgpu_bo buffer object + * + * References the contained &ttm_buffer_object. + * + * Returns: + * a refcounted pointer to the &amdgpu_bo buffer object. + */ struct amdgpu_bo *amdgpu_bo_ref(struct amdgpu_bo *bo) { if (bo == NULL) return NULL; - ttm_bo_reference(&bo->tbo); + ttm_bo_get(&bo->tbo); return bo; } +/** + * amdgpu_bo_unref - unreference an &amdgpu_bo buffer object + * @bo: &amdgpu_bo buffer object + * + * Unreferences the contained &ttm_buffer_object and clear the pointer + */ void amdgpu_bo_unref(struct amdgpu_bo **bo) { struct ttm_buffer_object *tbo; @@ -673,14 +843,34 @@ void amdgpu_bo_unref(struct amdgpu_bo **bo) return; tbo = &((*bo)->tbo); - ttm_bo_unref(&tbo); - if (tbo == NULL) - *bo = NULL; + ttm_bo_put(tbo); + *bo = NULL; } +/** + * amdgpu_bo_pin_restricted - pin an &amdgpu_bo buffer object + * @bo: &amdgpu_bo buffer object to be pinned + * @domain: domain to be pinned to + * @min_offset: the start of requested address range + * @max_offset: the end of requested address range + * + * Pins the buffer object according to requested domain and address range. If + * the memory is unbound gart memory, binds the pages into gart table. Adjusts + * pin_count and pin_size accordingly. + * + * Pinning means to lock pages in memory along with keeping them at a fixed + * offset. It is required when a buffer can not be moved, for example, when + * a display buffer is being scanned out. + * + * Compared with amdgpu_bo_pin(), this function gives more flexibility on + * where to pin a buffer if there are specific restrictions on where a buffer + * must be located. + * + * Returns: + * 0 for success or a negative error code on failure. + */ int amdgpu_bo_pin_restricted(struct amdgpu_bo *bo, u32 domain, - u64 min_offset, u64 max_offset, - u64 *gpu_addr) + u64 min_offset, u64 max_offset) { struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev); struct ttm_operation_ctx ctx = { false, false }; @@ -712,8 +902,6 @@ int amdgpu_bo_pin_restricted(struct amdgpu_bo *bo, u32 domain, return -EINVAL; bo->pin_count++; - if (gpu_addr) - *gpu_addr = amdgpu_bo_gpu_offset(bo); if (max_offset != 0) { u64 domain_start = bo->tbo.bdev->man[mem_type].gpu_offset; @@ -728,7 +916,7 @@ int amdgpu_bo_pin_restricted(struct amdgpu_bo *bo, u32 domain, /* force to pin into visible video ram */ if (!(bo->flags & AMDGPU_GEM_CREATE_NO_CPU_ACCESS)) bo->flags |= AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED; - amdgpu_ttm_placement_from_domain(bo, domain); + amdgpu_bo_placement_from_domain(bo, domain); for (i = 0; i < bo->placement.num_placement; i++) { unsigned fpfn, lpfn; @@ -749,33 +937,48 @@ int amdgpu_bo_pin_restricted(struct amdgpu_bo *bo, u32 domain, goto error; } - r = amdgpu_ttm_alloc_gart(&bo->tbo); - if (unlikely(r)) { - dev_err(adev->dev, "%p bind failed\n", bo); - goto error; - } - bo->pin_count = 1; - if (gpu_addr != NULL) - *gpu_addr = amdgpu_bo_gpu_offset(bo); domain = amdgpu_mem_type_to_domain(bo->tbo.mem.mem_type); if (domain == AMDGPU_GEM_DOMAIN_VRAM) { - adev->vram_pin_size += amdgpu_bo_size(bo); - adev->invisible_pin_size += amdgpu_vram_mgr_bo_invisible_size(bo); + atomic64_add(amdgpu_bo_size(bo), &adev->vram_pin_size); + atomic64_add(amdgpu_vram_mgr_bo_visible_size(bo), + &adev->visible_pin_size); } else if (domain == AMDGPU_GEM_DOMAIN_GTT) { - adev->gart_pin_size += amdgpu_bo_size(bo); + atomic64_add(amdgpu_bo_size(bo), &adev->gart_pin_size); } error: return r; } -int amdgpu_bo_pin(struct amdgpu_bo *bo, u32 domain, u64 *gpu_addr) +/** + * amdgpu_bo_pin - pin an &amdgpu_bo buffer object + * @bo: &amdgpu_bo buffer object to be pinned + * @domain: domain to be pinned to + * + * A simple wrapper to amdgpu_bo_pin_restricted(). + * Provides a simpler API for buffers that do not have any strict restrictions + * on where a buffer must be located. + * + * Returns: + * 0 for success or a negative error code on failure. + */ +int amdgpu_bo_pin(struct amdgpu_bo *bo, u32 domain) { - return amdgpu_bo_pin_restricted(bo, domain, 0, 0, gpu_addr); + return amdgpu_bo_pin_restricted(bo, domain, 0, 0); } +/** + * amdgpu_bo_unpin - unpin an &amdgpu_bo buffer object + * @bo: &amdgpu_bo buffer object to be unpinned + * + * Decreases the pin_count, and clears the flags if pin_count reaches 0. + * Changes placement and pin size accordingly. + * + * Returns: + * 0 for success or a negative error code on failure. + */ int amdgpu_bo_unpin(struct amdgpu_bo *bo) { struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev); @@ -790,12 +993,7 @@ int amdgpu_bo_unpin(struct amdgpu_bo *bo) if (bo->pin_count) return 0; - if (bo->tbo.mem.mem_type == TTM_PL_VRAM) { - adev->vram_pin_size -= amdgpu_bo_size(bo); - adev->invisible_pin_size -= amdgpu_vram_mgr_bo_invisible_size(bo); - } else if (bo->tbo.mem.mem_type == TTM_PL_TT) { - adev->gart_pin_size -= amdgpu_bo_size(bo); - } + amdgpu_bo_subtract_pin_size(bo); for (i = 0; i < bo->placement.num_placement; i++) { bo->placements[i].lpfn = 0; @@ -808,6 +1006,16 @@ int amdgpu_bo_unpin(struct amdgpu_bo *bo) return r; } +/** + * amdgpu_bo_evict_vram - evict VRAM buffers + * @adev: amdgpu device object + * + * Evicts all VRAM buffers on the lru list of the memory type. + * Mainly used for evicting vram at suspend time. + * + * Returns: + * 0 for success or a negative error code on failure. + */ int amdgpu_bo_evict_vram(struct amdgpu_device *adev) { /* late 2.6.33 fix IGP hibernate - we need pm ops to do this correct */ @@ -830,6 +1038,15 @@ static const char *amdgpu_vram_names[] = { "DDR4", }; +/** + * amdgpu_bo_init - initialize memory manager + * @adev: amdgpu device object + * + * Calls amdgpu_ttm_init() to initialize amdgpu memory manager. + * + * Returns: + * 0 for success or a negative error code on failure. + */ int amdgpu_bo_init(struct amdgpu_device *adev) { /* reserve PAT memory space to WC for VRAM */ @@ -847,6 +1064,16 @@ int amdgpu_bo_init(struct amdgpu_device *adev) return amdgpu_ttm_init(adev); } +/** + * amdgpu_bo_late_init - late init + * @adev: amdgpu device object + * + * Calls amdgpu_ttm_late_init() to free resources used earlier during + * initialization. + * + * Returns: + * 0 for success or a negative error code on failure. + */ int amdgpu_bo_late_init(struct amdgpu_device *adev) { amdgpu_ttm_late_init(adev); @@ -854,6 +1081,12 @@ int amdgpu_bo_late_init(struct amdgpu_device *adev) return 0; } +/** + * amdgpu_bo_fini - tear down memory manager + * @adev: amdgpu device object + * + * Reverses amdgpu_bo_init() to tear down memory manager. + */ void amdgpu_bo_fini(struct amdgpu_device *adev) { amdgpu_ttm_fini(adev); @@ -861,12 +1094,33 @@ void amdgpu_bo_fini(struct amdgpu_device *adev) arch_io_free_memtype_wc(adev->gmc.aper_base, adev->gmc.aper_size); } +/** + * amdgpu_bo_fbdev_mmap - mmap fbdev memory + * @bo: &amdgpu_bo buffer object + * @vma: vma as input from the fbdev mmap method + * + * Calls ttm_fbdev_mmap() to mmap fbdev memory if it is backed by a bo. + * + * Returns: + * 0 for success or a negative error code on failure. + */ int amdgpu_bo_fbdev_mmap(struct amdgpu_bo *bo, struct vm_area_struct *vma) { return ttm_fbdev_mmap(vma, &bo->tbo); } +/** + * amdgpu_bo_set_tiling_flags - set tiling flags + * @bo: &amdgpu_bo buffer object + * @tiling_flags: new flags + * + * Sets buffer object's tiling flags with the new one. Used by GEM ioctl or + * kernel driver to set the tiling flags on a buffer. + * + * Returns: + * 0 for success or a negative error code on failure. + */ int amdgpu_bo_set_tiling_flags(struct amdgpu_bo *bo, u64 tiling_flags) { struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev); @@ -879,6 +1133,14 @@ int amdgpu_bo_set_tiling_flags(struct amdgpu_bo *bo, u64 tiling_flags) return 0; } +/** + * amdgpu_bo_get_tiling_flags - get tiling flags + * @bo: &amdgpu_bo buffer object + * @tiling_flags: returned flags + * + * Gets buffer object's tiling flags. Used by GEM ioctl or kernel driver to + * set the tiling flags on a buffer. + */ void amdgpu_bo_get_tiling_flags(struct amdgpu_bo *bo, u64 *tiling_flags) { lockdep_assert_held(&bo->tbo.resv->lock.base); @@ -887,6 +1149,19 @@ void amdgpu_bo_get_tiling_flags(struct amdgpu_bo *bo, u64 *tiling_flags) *tiling_flags = bo->tiling_flags; } +/** + * amdgpu_bo_set_metadata - set metadata + * @bo: &amdgpu_bo buffer object + * @metadata: new metadata + * @metadata_size: size of the new metadata + * @flags: flags of the new metadata + * + * Sets buffer object's metadata, its size and flags. + * Used via GEM ioctl. + * + * Returns: + * 0 for success or a negative error code on failure. + */ int amdgpu_bo_set_metadata (struct amdgpu_bo *bo, void *metadata, uint32_t metadata_size, uint64_t flags) { @@ -916,6 +1191,21 @@ int amdgpu_bo_set_metadata (struct amdgpu_bo *bo, void *metadata, return 0; } +/** + * amdgpu_bo_get_metadata - get metadata + * @bo: &amdgpu_bo buffer object + * @buffer: returned metadata + * @buffer_size: size of the buffer + * @metadata_size: size of the returned metadata + * @flags: flags of the returned metadata + * + * Gets buffer object's metadata, its size and flags. buffer_size shall not be + * less than metadata_size. + * Used via GEM ioctl. + * + * Returns: + * 0 for success or a negative error code on failure. + */ int amdgpu_bo_get_metadata(struct amdgpu_bo *bo, void *buffer, size_t buffer_size, uint32_t *metadata_size, uint64_t *flags) @@ -939,6 +1229,16 @@ int amdgpu_bo_get_metadata(struct amdgpu_bo *bo, void *buffer, return 0; } +/** + * amdgpu_bo_move_notify - notification about a memory move + * @bo: pointer to a buffer object + * @evict: if this move is evicting the buffer from the graphics address space + * @new_mem: new information of the bufer object + * + * Marks the corresponding &amdgpu_bo buffer object as invalid, also performs + * bookkeeping. + * TTM driver callback which is called when ttm moves a buffer. + */ void amdgpu_bo_move_notify(struct ttm_buffer_object *bo, bool evict, struct ttm_mem_reg *new_mem) @@ -947,7 +1247,7 @@ void amdgpu_bo_move_notify(struct ttm_buffer_object *bo, struct amdgpu_bo *abo; struct ttm_mem_reg *old_mem = &bo->mem; - if (!amdgpu_ttm_bo_is_amdgpu_bo(bo)) + if (!amdgpu_bo_is_amdgpu_bo(bo)) return; abo = ttm_to_amdgpu_bo(bo); @@ -964,9 +1264,20 @@ void amdgpu_bo_move_notify(struct ttm_buffer_object *bo, return; /* move_notify is called before move happens */ - trace_amdgpu_ttm_bo_move(abo, new_mem->mem_type, old_mem->mem_type); + trace_amdgpu_bo_move(abo, new_mem->mem_type, old_mem->mem_type); } +/** + * amdgpu_bo_fault_reserve_notify - notification about a memory fault + * @bo: pointer to a buffer object + * + * Notifies the driver we are taking a fault on this BO and have reserved it, + * also performs bookkeeping. + * TTM driver callback for dealing with vm faults. + * + * Returns: + * 0 for success or a negative error code on failure. + */ int amdgpu_bo_fault_reserve_notify(struct ttm_buffer_object *bo) { struct amdgpu_device *adev = amdgpu_ttm_adev(bo->bdev); @@ -975,7 +1286,7 @@ int amdgpu_bo_fault_reserve_notify(struct ttm_buffer_object *bo) unsigned long offset, size; int r; - if (!amdgpu_ttm_bo_is_amdgpu_bo(bo)) + if (!amdgpu_bo_is_amdgpu_bo(bo)) return 0; abo = ttm_to_amdgpu_bo(bo); @@ -997,8 +1308,8 @@ int amdgpu_bo_fault_reserve_notify(struct ttm_buffer_object *bo) /* hurrah the memory is not visible ! */ atomic64_inc(&adev->num_vram_cpu_page_faults); - amdgpu_ttm_placement_from_domain(abo, AMDGPU_GEM_DOMAIN_VRAM | - AMDGPU_GEM_DOMAIN_GTT); + amdgpu_bo_placement_from_domain(abo, AMDGPU_GEM_DOMAIN_VRAM | + AMDGPU_GEM_DOMAIN_GTT); /* Avoid costly evictions; only set GTT as a busy placement */ abo->placement.num_busy_placement = 1; @@ -1040,10 +1351,11 @@ void amdgpu_bo_fence(struct amdgpu_bo *bo, struct dma_fence *fence, * amdgpu_bo_gpu_offset - return GPU offset of bo * @bo: amdgpu object for which we query the offset * - * Returns current GPU offset of the object. - * * Note: object should either be pinned or reserved when calling this * function, it might be useful to add check for this for debugging. + * + * Returns: + * current GPU offset of the object. */ u64 amdgpu_bo_gpu_offset(struct amdgpu_bo *bo) { @@ -1059,6 +1371,14 @@ u64 amdgpu_bo_gpu_offset(struct amdgpu_bo *bo) return bo->tbo.offset; } +/** + * amdgpu_bo_get_preferred_pin_domain - get preferred domain for scanout + * @adev: amdgpu device object + * @domain: allowed :ref:`memory domains <amdgpu_memory_domains>` + * + * Returns: + * Which of the allowed domains is preferred for pinning the BO for scanout. + */ uint32_t amdgpu_bo_get_preferred_pin_domain(struct amdgpu_device *adev, uint32_t domain) { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h index 731748033878..18945dd6982d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h @@ -32,6 +32,7 @@ #include "amdgpu.h" #define AMDGPU_BO_INVALID_OFFSET LONG_MAX +#define AMDGPU_BO_MAX_PLACEMENTS 3 struct amdgpu_bo_param { unsigned long size; @@ -77,7 +78,7 @@ struct amdgpu_bo { /* Protected by tbo.reserved */ u32 preferred_domains; u32 allowed_domains; - struct ttm_place placements[AMDGPU_GEM_DOMAIN_MAX + 1]; + struct ttm_place placements[AMDGPU_BO_MAX_PLACEMENTS]; struct ttm_placement placement; struct ttm_buffer_object tbo; struct ttm_bo_kmap_obj kmap; @@ -234,6 +235,9 @@ static inline bool amdgpu_bo_explicit_sync(struct amdgpu_bo *bo) return bo->flags & AMDGPU_GEM_CREATE_EXPLICIT_SYNC; } +bool amdgpu_bo_is_amdgpu_bo(struct ttm_buffer_object *bo); +void amdgpu_bo_placement_from_domain(struct amdgpu_bo *abo, u32 domain); + int amdgpu_bo_create(struct amdgpu_device *adev, struct amdgpu_bo_param *bp, struct amdgpu_bo **bo_ptr); @@ -252,10 +256,9 @@ void *amdgpu_bo_kptr(struct amdgpu_bo *bo); void amdgpu_bo_kunmap(struct amdgpu_bo *bo); struct amdgpu_bo *amdgpu_bo_ref(struct amdgpu_bo *bo); void amdgpu_bo_unref(struct amdgpu_bo **bo); -int amdgpu_bo_pin(struct amdgpu_bo *bo, u32 domain, u64 *gpu_addr); +int amdgpu_bo_pin(struct amdgpu_bo *bo, u32 domain); int amdgpu_bo_pin_restricted(struct amdgpu_bo *bo, u32 domain, - u64 min_offset, u64 max_offset, - u64 *gpu_addr); + u64 min_offset, u64 max_offset); int amdgpu_bo_unpin(struct amdgpu_bo *bo); int amdgpu_bo_evict_vram(struct amdgpu_device *adev); int amdgpu_bo_init(struct amdgpu_device *adev); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c index fc818b4d849c..8f98629fbe59 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c @@ -31,7 +31,7 @@ #include <linux/power_supply.h> #include <linux/hwmon.h> #include <linux/hwmon-sysfs.h> - +#include <linux/nospec.h> static int amdgpu_debugfs_pm_init(struct amdgpu_device *adev); @@ -68,11 +68,11 @@ void amdgpu_pm_acpi_event_handler(struct amdgpu_device *adev) if (adev->pm.dpm_enabled) { mutex_lock(&adev->pm.mutex); if (power_supply_is_system_supplied() > 0) - adev->pm.dpm.ac_power = true; + adev->pm.ac_power = true; else - adev->pm.dpm.ac_power = false; + adev->pm.ac_power = false; if (adev->powerplay.pp_funcs->enable_bapm) - amdgpu_dpm_enable_bapm(adev, adev->pm.dpm.ac_power); + amdgpu_dpm_enable_bapm(adev, adev->pm.ac_power); mutex_unlock(&adev->pm.mutex); } } @@ -80,12 +80,15 @@ void amdgpu_pm_acpi_event_handler(struct amdgpu_device *adev) /** * DOC: power_dpm_state * - * This is a legacy interface and is only provided for backwards compatibility. - * The amdgpu driver provides a sysfs API for adjusting certain power - * related parameters. The file power_dpm_state is used for this. + * The power_dpm_state file is a legacy interface and is only provided for + * backwards compatibility. The amdgpu driver provides a sysfs API for adjusting + * certain power related parameters. The file power_dpm_state is used for this. * It accepts the following arguments: + * * - battery + * * - balanced + * * - performance * * battery @@ -169,14 +172,21 @@ fail: * The amdgpu driver provides a sysfs API for adjusting certain power * related parameters. The file power_dpm_force_performance_level is * used for this. It accepts the following arguments: + * * - auto + * * - low + * * - high + * * - manual - * - GPU fan + * * - profile_standard + * * - profile_min_sclk + * * - profile_min_mclk + * * - profile_peak * * auto @@ -393,6 +403,7 @@ static ssize_t amdgpu_set_pp_force_state(struct device *dev, count = -EINVAL; goto fail; } + idx = array_index_nospec(idx, ARRAY_SIZE(data.states)); amdgpu_dpm_get_pp_num_states(adev, &data); state = data.states[idx]; @@ -463,8 +474,11 @@ static ssize_t amdgpu_set_pp_table(struct device *dev, * this. * * Reading the file will display: + * * - a list of engine clock levels and voltages labeled OD_SCLK + * * - a list of memory clock levels and voltages labeled OD_MCLK + * * - a list of valid ranges for sclk, mclk, and voltage labeled OD_RANGE * * To manually adjust these settings, first select manual using @@ -593,40 +607,59 @@ static ssize_t amdgpu_get_pp_dpm_sclk(struct device *dev, return snprintf(buf, PAGE_SIZE, "\n"); } -static ssize_t amdgpu_set_pp_dpm_sclk(struct device *dev, - struct device_attribute *attr, - const char *buf, - size_t count) +/* + * Worst case: 32 bits individually specified, in octal at 12 characters + * per line (+1 for \n). + */ +#define AMDGPU_MASK_BUF_MAX (32 * 13) + +static ssize_t amdgpu_read_mask(const char *buf, size_t count, uint32_t *mask) { - struct drm_device *ddev = dev_get_drvdata(dev); - struct amdgpu_device *adev = ddev->dev_private; int ret; long level; - uint32_t mask = 0; char *sub_str = NULL; char *tmp; - char buf_cpy[count]; + char buf_cpy[AMDGPU_MASK_BUF_MAX + 1]; const char delimiter[3] = {' ', '\n', '\0'}; + size_t bytes; - memcpy(buf_cpy, buf, count+1); + *mask = 0; + + bytes = min(count, sizeof(buf_cpy) - 1); + memcpy(buf_cpy, buf, bytes); + buf_cpy[bytes] = '\0'; tmp = buf_cpy; while (tmp[0]) { - sub_str = strsep(&tmp, delimiter); + sub_str = strsep(&tmp, delimiter); if (strlen(sub_str)) { ret = kstrtol(sub_str, 0, &level); - - if (ret) { - count = -EINVAL; - goto fail; - } - mask |= 1 << level; + if (ret) + return -EINVAL; + *mask |= 1 << level; } else break; } + + return 0; +} + +static ssize_t amdgpu_set_pp_dpm_sclk(struct device *dev, + struct device_attribute *attr, + const char *buf, + size_t count) +{ + struct drm_device *ddev = dev_get_drvdata(dev); + struct amdgpu_device *adev = ddev->dev_private; + int ret; + uint32_t mask = 0; + + ret = amdgpu_read_mask(buf, count, &mask); + if (ret) + return ret; + if (adev->powerplay.pp_funcs->force_clock_level) amdgpu_dpm_force_clock_level(adev, PP_SCLK, mask); -fail: return count; } @@ -651,32 +684,15 @@ static ssize_t amdgpu_set_pp_dpm_mclk(struct device *dev, struct drm_device *ddev = dev_get_drvdata(dev); struct amdgpu_device *adev = ddev->dev_private; int ret; - long level; uint32_t mask = 0; - char *sub_str = NULL; - char *tmp; - char buf_cpy[count]; - const char delimiter[3] = {' ', '\n', '\0'}; - memcpy(buf_cpy, buf, count+1); - tmp = buf_cpy; - while (tmp[0]) { - sub_str = strsep(&tmp, delimiter); - if (strlen(sub_str)) { - ret = kstrtol(sub_str, 0, &level); + ret = amdgpu_read_mask(buf, count, &mask); + if (ret) + return ret; - if (ret) { - count = -EINVAL; - goto fail; - } - mask |= 1 << level; - } else - break; - } if (adev->powerplay.pp_funcs->force_clock_level) amdgpu_dpm_force_clock_level(adev, PP_MCLK, mask); -fail: return count; } @@ -701,33 +717,15 @@ static ssize_t amdgpu_set_pp_dpm_pcie(struct device *dev, struct drm_device *ddev = dev_get_drvdata(dev); struct amdgpu_device *adev = ddev->dev_private; int ret; - long level; uint32_t mask = 0; - char *sub_str = NULL; - char *tmp; - char buf_cpy[count]; - const char delimiter[3] = {' ', '\n', '\0'}; - memcpy(buf_cpy, buf, count+1); - tmp = buf_cpy; - - while (tmp[0]) { - sub_str = strsep(&tmp, delimiter); - if (strlen(sub_str)) { - ret = kstrtol(sub_str, 0, &level); + ret = amdgpu_read_mask(buf, count, &mask); + if (ret) + return ret; - if (ret) { - count = -EINVAL; - goto fail; - } - mask |= 1 << level; - } else - break; - } if (adev->powerplay.pp_funcs->force_clock_level) amdgpu_dpm_force_clock_level(adev, PP_PCIE, mask); -fail: return count; } @@ -905,6 +903,36 @@ fail: return -EINVAL; } +/** + * DOC: busy_percent + * + * The amdgpu driver provides a sysfs API for reading how busy the GPU + * is as a percentage. The file gpu_busy_percent is used for this. + * The SMU firmware computes a percentage of load based on the + * aggregate activity level in the IP cores. + */ +static ssize_t amdgpu_get_busy_percent(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + struct drm_device *ddev = dev_get_drvdata(dev); + struct amdgpu_device *adev = ddev->dev_private; + int r, value, size = sizeof(value); + + /* sanity check PP is enabled */ + if (!(adev->powerplay.pp_funcs && + adev->powerplay.pp_funcs->read_sensor)) + return -EINVAL; + + /* read the IP busy sensor */ + r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_GPU_LOAD, + (void *)&value, &size); + if (r) + return r; + + return snprintf(buf, PAGE_SIZE, "%d\n", value); +} + static DEVICE_ATTR(power_dpm_state, S_IRUGO | S_IWUSR, amdgpu_get_dpm_state, amdgpu_set_dpm_state); static DEVICE_ATTR(power_dpm_force_performance_level, S_IRUGO | S_IWUSR, amdgpu_get_dpm_forced_performance_level, @@ -938,6 +966,8 @@ static DEVICE_ATTR(pp_power_profile_mode, S_IRUGO | S_IWUSR, static DEVICE_ATTR(pp_od_clk_voltage, S_IRUGO | S_IWUSR, amdgpu_get_pp_od_clk_voltage, amdgpu_set_pp_od_clk_voltage); +static DEVICE_ATTR(gpu_busy_percent, S_IRUGO, + amdgpu_get_busy_percent, NULL); static ssize_t amdgpu_hwmon_show_temp(struct device *dev, struct device_attribute *attr, @@ -1156,7 +1186,7 @@ static ssize_t amdgpu_hwmon_show_vddnb(struct device *dev, int r, size = sizeof(vddnb); /* only APUs have vddnb */ - if (adev->flags & AMD_IS_APU) + if (!(adev->flags & AMD_IS_APU)) return -EINVAL; /* Can't get voltage when the card is off */ @@ -1285,35 +1315,51 @@ static ssize_t amdgpu_hwmon_set_power_cap(struct device *dev, * DOC: hwmon * * The amdgpu driver exposes the following sensor interfaces: + * * - GPU temperature (via the on-die sensor) + * * - GPU voltage + * * - Northbridge voltage (APUs only) + * * - GPU power + * * - GPU fan * * hwmon interfaces for GPU temperature: + * * - temp1_input: the on die GPU temperature in millidegrees Celsius + * * - temp1_crit: temperature critical max value in millidegrees Celsius + * * - temp1_crit_hyst: temperature hysteresis for critical limit in millidegrees Celsius * * hwmon interfaces for GPU voltage: + * * - in0_input: the voltage on the GPU in millivolts + * * - in1_input: the voltage on the Northbridge in millivolts * * hwmon interfaces for GPU power: + * * - power1_average: average power used by the GPU in microWatts + * * - power1_cap_min: minimum cap supported in microWatts + * * - power1_cap_max: maximum cap supported in microWatts + * * - power1_cap: selected power cap in microWatts * * hwmon interfaces for GPU fan: + * * - pwm1: pulse width modulation fan level (0-255) - * - pwm1_enable: pulse width modulation fan control method - * 0: no fan speed control - * 1: manual fan speed control using pwm interface - * 2: automatic fan speed control + * + * - pwm1_enable: pulse width modulation fan control method (0: no fan speed control, 1: manual fan speed control using pwm interface, 2: automatic fan speed control) + * * - pwm1_min: pulse width modulation fan control minimum level (0) + * * - pwm1_max: pulse width modulation fan control maximum level (255) + * * - fan1_input: fan speed in RPM * * You can use hwmon tools like sensors to view this information on your system. @@ -1668,10 +1714,10 @@ static void amdgpu_dpm_change_power_state_locked(struct amdgpu_device *adev) void amdgpu_dpm_enable_uvd(struct amdgpu_device *adev, bool enable) { - if (adev->powerplay.pp_funcs->powergate_uvd) { + if (adev->powerplay.pp_funcs->set_powergating_by_smu) { /* enable/disable UVD */ mutex_lock(&adev->pm.mutex); - amdgpu_dpm_powergate_uvd(adev, !enable); + amdgpu_dpm_set_powergating_by_smu(adev, AMD_IP_BLOCK_TYPE_UVD, !enable); mutex_unlock(&adev->pm.mutex); } else { if (enable) { @@ -1690,10 +1736,10 @@ void amdgpu_dpm_enable_uvd(struct amdgpu_device *adev, bool enable) void amdgpu_dpm_enable_vce(struct amdgpu_device *adev, bool enable) { - if (adev->powerplay.pp_funcs->powergate_vce) { + if (adev->powerplay.pp_funcs->set_powergating_by_smu) { /* enable/disable VCE */ mutex_lock(&adev->pm.mutex); - amdgpu_dpm_powergate_vce(adev, !enable); + amdgpu_dpm_set_powergating_by_smu(adev, AMD_IP_BLOCK_TYPE_VCE, !enable); mutex_unlock(&adev->pm.mutex); } else { if (enable) { @@ -1825,6 +1871,13 @@ int amdgpu_pm_sysfs_init(struct amdgpu_device *adev) "pp_od_clk_voltage\n"); return ret; } + ret = device_create_file(adev->dev, + &dev_attr_gpu_busy_percent); + if (ret) { + DRM_ERROR("failed to create device file " + "gpu_busy_level\n"); + return ret; + } ret = amdgpu_debugfs_pm_init(adev); if (ret) { DRM_ERROR("Failed to register debugfs file for dpm!\n"); @@ -1860,6 +1913,7 @@ void amdgpu_pm_sysfs_fini(struct amdgpu_device *adev) &dev_attr_pp_power_profile_mode); device_remove_file(adev->dev, &dev_attr_pp_od_clk_voltage); + device_remove_file(adev->dev, &dev_attr_gpu_busy_percent); } void amdgpu_pm_compute_clocks(struct amdgpu_device *adev) @@ -1878,6 +1932,14 @@ void amdgpu_pm_compute_clocks(struct amdgpu_device *adev) amdgpu_fence_wait_empty(ring); } + mutex_lock(&adev->pm.mutex); + /* update battery/ac status */ + if (power_supply_is_system_supplied() > 0) + adev->pm.ac_power = true; + else + adev->pm.ac_power = false; + mutex_unlock(&adev->pm.mutex); + if (adev->powerplay.pp_funcs->dispatch_tasks) { if (!amdgpu_device_has_dc_support(adev)) { mutex_lock(&adev->pm.mutex); @@ -1898,14 +1960,7 @@ void amdgpu_pm_compute_clocks(struct amdgpu_device *adev) } else { mutex_lock(&adev->pm.mutex); amdgpu_dpm_get_active_displays(adev); - /* update battery/ac status */ - if (power_supply_is_system_supplied() > 0) - adev->pm.dpm.ac_power = true; - else - adev->pm.dpm.ac_power = false; - amdgpu_dpm_change_power_state_locked(adev); - mutex_unlock(&adev->pm.mutex); } } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c index 4683626b065f..1c5d97f4b4dd 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c @@ -23,6 +23,14 @@ * * Authors: Alex Deucher */ + +/** + * DOC: PRIME Buffer Sharing + * + * The following callback implementations are used for :ref:`sharing GEM buffer + * objects between different devices via PRIME <prime_buffer_sharing>`. + */ + #include <drm/drmP.h> #include "amdgpu.h" @@ -32,6 +40,14 @@ static const struct dma_buf_ops amdgpu_dmabuf_ops; +/** + * amdgpu_gem_prime_get_sg_table - &drm_driver.gem_prime_get_sg_table + * implementation + * @obj: GEM buffer object + * + * Returns: + * A scatter/gather table for the pinned pages of the buffer object's memory. + */ struct sg_table *amdgpu_gem_prime_get_sg_table(struct drm_gem_object *obj) { struct amdgpu_bo *bo = gem_to_amdgpu_bo(obj); @@ -40,6 +56,15 @@ struct sg_table *amdgpu_gem_prime_get_sg_table(struct drm_gem_object *obj) return drm_prime_pages_to_sg(bo->tbo.ttm->pages, npages); } +/** + * amdgpu_gem_prime_vmap - &dma_buf_ops.vmap implementation + * @obj: GEM buffer object + * + * Sets up an in-kernel virtual mapping of the buffer object's memory. + * + * Returns: + * The virtual address of the mapping or an error pointer. + */ void *amdgpu_gem_prime_vmap(struct drm_gem_object *obj) { struct amdgpu_bo *bo = gem_to_amdgpu_bo(obj); @@ -53,6 +78,13 @@ void *amdgpu_gem_prime_vmap(struct drm_gem_object *obj) return bo->dma_buf_vmap.virtual; } +/** + * amdgpu_gem_prime_vunmap - &dma_buf_ops.vunmap implementation + * @obj: GEM buffer object + * @vaddr: virtual address (unused) + * + * Tears down the in-kernel virtual mapping of the buffer object's memory. + */ void amdgpu_gem_prime_vunmap(struct drm_gem_object *obj, void *vaddr) { struct amdgpu_bo *bo = gem_to_amdgpu_bo(obj); @@ -60,6 +92,17 @@ void amdgpu_gem_prime_vunmap(struct drm_gem_object *obj, void *vaddr) ttm_bo_kunmap(&bo->dma_buf_vmap); } +/** + * amdgpu_gem_prime_mmap - &drm_driver.gem_prime_mmap implementation + * @obj: GEM buffer object + * @vma: virtual memory area + * + * Sets up a userspace mapping of the buffer object's memory in the given + * virtual memory area. + * + * Returns: + * 0 on success or negative error code. + */ int amdgpu_gem_prime_mmap(struct drm_gem_object *obj, struct vm_area_struct *vma) { struct amdgpu_bo *bo = gem_to_amdgpu_bo(obj); @@ -94,6 +137,19 @@ int amdgpu_gem_prime_mmap(struct drm_gem_object *obj, struct vm_area_struct *vma return ret; } +/** + * amdgpu_gem_prime_import_sg_table - &drm_driver.gem_prime_import_sg_table + * implementation + * @dev: DRM device + * @attach: DMA-buf attachment + * @sg: Scatter/gather table + * + * Import shared DMA buffer memory exported by another device. + * + * Returns: + * A new GEM buffer object of the given DRM device, representing the memory + * described by the given DMA-buf attachment and scatter/gather table. + */ struct drm_gem_object * amdgpu_gem_prime_import_sg_table(struct drm_device *dev, struct dma_buf_attachment *attach, @@ -132,8 +188,19 @@ error: return ERR_PTR(ret); } +/** + * amdgpu_gem_map_attach - &dma_buf_ops.attach implementation + * @dma_buf: shared DMA buffer + * @attach: DMA-buf attachment + * + * Makes sure that the shared DMA buffer can be accessed by the target device. + * For now, simply pins it to the GTT domain, where it should be accessible by + * all DMA devices. + * + * Returns: + * 0 on success or negative error code. + */ static int amdgpu_gem_map_attach(struct dma_buf *dma_buf, - struct device *target_dev, struct dma_buf_attachment *attach) { struct drm_gem_object *obj = dma_buf->priv; @@ -141,7 +208,7 @@ static int amdgpu_gem_map_attach(struct dma_buf *dma_buf, struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev); long r; - r = drm_gem_map_attach(dma_buf, target_dev, attach); + r = drm_gem_map_attach(dma_buf, attach); if (r) return r; @@ -165,7 +232,7 @@ static int amdgpu_gem_map_attach(struct dma_buf *dma_buf, } /* pin buffer into GTT */ - r = amdgpu_bo_pin(bo, AMDGPU_GEM_DOMAIN_GTT, NULL); + r = amdgpu_bo_pin(bo, AMDGPU_GEM_DOMAIN_GTT); if (r) goto error_unreserve; @@ -181,6 +248,14 @@ error_detach: return r; } +/** + * amdgpu_gem_map_detach - &dma_buf_ops.detach implementation + * @dma_buf: shared DMA buffer + * @attach: DMA-buf attachment + * + * This is called when a shared DMA buffer no longer needs to be accessible by + * the other device. For now, simply unpins the buffer from GTT. + */ static void amdgpu_gem_map_detach(struct dma_buf *dma_buf, struct dma_buf_attachment *attach) { @@ -202,6 +277,13 @@ error: drm_gem_map_detach(dma_buf, attach); } +/** + * amdgpu_gem_prime_res_obj - &drm_driver.gem_prime_res_obj implementation + * @obj: GEM buffer object + * + * Returns: + * The buffer object's reservation object. + */ struct reservation_object *amdgpu_gem_prime_res_obj(struct drm_gem_object *obj) { struct amdgpu_bo *bo = gem_to_amdgpu_bo(obj); @@ -209,6 +291,18 @@ struct reservation_object *amdgpu_gem_prime_res_obj(struct drm_gem_object *obj) return bo->tbo.resv; } +/** + * amdgpu_gem_begin_cpu_access - &dma_buf_ops.begin_cpu_access implementation + * @dma_buf: shared DMA buffer + * @direction: direction of DMA transfer + * + * This is called before CPU access to the shared DMA buffer's memory. If it's + * a read access, the buffer is moved to the GTT domain if possible, for optimal + * CPU read performance. + * + * Returns: + * 0 on success or negative error code. + */ static int amdgpu_gem_begin_cpu_access(struct dma_buf *dma_buf, enum dma_data_direction direction) { @@ -229,7 +323,7 @@ static int amdgpu_gem_begin_cpu_access(struct dma_buf *dma_buf, return ret; if (!bo->pin_count && (bo->allowed_domains & AMDGPU_GEM_DOMAIN_GTT)) { - amdgpu_ttm_placement_from_domain(bo, AMDGPU_GEM_DOMAIN_GTT); + amdgpu_bo_placement_from_domain(bo, AMDGPU_GEM_DOMAIN_GTT); ret = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx); } @@ -245,14 +339,24 @@ static const struct dma_buf_ops amdgpu_dmabuf_ops = { .release = drm_gem_dmabuf_release, .begin_cpu_access = amdgpu_gem_begin_cpu_access, .map = drm_gem_dmabuf_kmap, - .map_atomic = drm_gem_dmabuf_kmap_atomic, .unmap = drm_gem_dmabuf_kunmap, - .unmap_atomic = drm_gem_dmabuf_kunmap_atomic, .mmap = drm_gem_dmabuf_mmap, .vmap = drm_gem_dmabuf_vmap, .vunmap = drm_gem_dmabuf_vunmap, }; +/** + * amdgpu_gem_prime_export - &drm_driver.gem_prime_export implementation + * @dev: DRM device + * @gobj: GEM buffer object + * @flags: flags like DRM_CLOEXEC and DRM_RDWR + * + * The main work is done by the &drm_gem_prime_export helper, which in turn + * uses &amdgpu_gem_prime_res_obj. + * + * Returns: + * Shared DMA buffer representing the GEM buffer object from the given device. + */ struct dma_buf *amdgpu_gem_prime_export(struct drm_device *dev, struct drm_gem_object *gobj, int flags) @@ -273,6 +377,17 @@ struct dma_buf *amdgpu_gem_prime_export(struct drm_device *dev, return buf; } +/** + * amdgpu_gem_prime_import - &drm_driver.gem_prime_import implementation + * @dev: DRM device + * @dma_buf: Shared DMA buffer + * + * The main work is done by the &drm_gem_prime_import helper, which in turn + * uses &amdgpu_gem_prime_import_sg_table. + * + * Returns: + * GEM buffer object representing the shared DMA buffer for the given device. + */ struct drm_gem_object *amdgpu_gem_prime_import(struct drm_device *dev, struct dma_buf *dma_buf) { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_queue_mgr.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_queue_mgr.c index 8af16e81c7d4..a172bba32b45 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_queue_mgr.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_queue_mgr.c @@ -66,8 +66,6 @@ static int amdgpu_identity_map(struct amdgpu_device *adev, u32 ring, struct amdgpu_ring **out_ring) { - u32 instance; - switch (mapper->hw_ip) { case AMDGPU_HW_IP_GFX: *out_ring = &adev->gfx.gfx_ring[ring]; @@ -79,16 +77,13 @@ static int amdgpu_identity_map(struct amdgpu_device *adev, *out_ring = &adev->sdma.instance[ring].ring; break; case AMDGPU_HW_IP_UVD: - instance = ring; - *out_ring = &adev->uvd.inst[instance].ring; + *out_ring = &adev->uvd.inst[0].ring; break; case AMDGPU_HW_IP_VCE: *out_ring = &adev->vce.ring[ring]; break; case AMDGPU_HW_IP_UVD_ENC: - instance = ring / adev->uvd.num_enc_rings; - *out_ring = - &adev->uvd.inst[instance].ring_enc[ring%adev->uvd.num_enc_rings]; + *out_ring = &adev->uvd.inst[0].ring_enc[ring]; break; case AMDGPU_HW_IP_VCN_DEC: *out_ring = &adev->vcn.ring_dec; @@ -96,6 +91,9 @@ static int amdgpu_identity_map(struct amdgpu_device *adev, case AMDGPU_HW_IP_VCN_ENC: *out_ring = &adev->vcn.ring_enc[ring]; break; + case AMDGPU_HW_IP_VCN_JPEG: + *out_ring = &adev->vcn.ring_jpeg; + break; default: *out_ring = NULL; DRM_ERROR("unknown HW IP type: %d\n", mapper->hw_ip); @@ -216,7 +214,7 @@ int amdgpu_queue_mgr_map(struct amdgpu_device *adev, u32 hw_ip, u32 instance, u32 ring, struct amdgpu_ring **out_ring) { - int r, ip_num_rings; + int i, r, ip_num_rings = 0; struct amdgpu_queue_mapper *mapper = &mgr->mapper[hw_ip]; if (!adev || !mgr || !out_ring) @@ -245,14 +243,21 @@ int amdgpu_queue_mgr_map(struct amdgpu_device *adev, ip_num_rings = adev->sdma.num_instances; break; case AMDGPU_HW_IP_UVD: - ip_num_rings = adev->uvd.num_uvd_inst; + for (i = 0; i < adev->uvd.num_uvd_inst; i++) { + if (!(adev->uvd.harvest_config & (1 << i))) + ip_num_rings++; + } break; case AMDGPU_HW_IP_VCE: ip_num_rings = adev->vce.num_rings; break; case AMDGPU_HW_IP_UVD_ENC: + for (i = 0; i < adev->uvd.num_uvd_inst; i++) { + if (!(adev->uvd.harvest_config & (1 << i))) + ip_num_rings++; + } ip_num_rings = - adev->uvd.num_enc_rings * adev->uvd.num_uvd_inst; + adev->uvd.num_enc_rings * ip_num_rings; break; case AMDGPU_HW_IP_VCN_DEC: ip_num_rings = 1; @@ -260,6 +265,9 @@ int amdgpu_queue_mgr_map(struct amdgpu_device *adev, case AMDGPU_HW_IP_VCN_ENC: ip_num_rings = adev->vcn.num_enc_rings; break; + case AMDGPU_HW_IP_VCN_JPEG: + ip_num_rings = 1; + break; default: DRM_DEBUG("unknown ip type: %d\n", hw_ip); return -EINVAL; @@ -287,6 +295,7 @@ int amdgpu_queue_mgr_map(struct amdgpu_device *adev, case AMDGPU_HW_IP_UVD_ENC: case AMDGPU_HW_IP_VCN_DEC: case AMDGPU_HW_IP_VCN_ENC: + case AMDGPU_HW_IP_VCN_JPEG: r = amdgpu_identity_map(adev, mapper, ring, out_ring); break; case AMDGPU_HW_IP_DMA: diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c index c6850b629d0e..93794a85f83d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c @@ -211,7 +211,8 @@ void amdgpu_ring_priority_get(struct amdgpu_ring *ring, if (!ring->funcs->set_priority) return; - atomic_inc(&ring->num_jobs[priority]); + if (atomic_inc_return(&ring->num_jobs[priority]) <= 0) + return; mutex_lock(&ring->priority_mutex); if (priority <= ring->priority) @@ -304,7 +305,7 @@ int amdgpu_ring_init(struct amdgpu_device *adev, struct amdgpu_ring *ring, 0xffffffffffffffff : ring->buf_mask; /* Allocate ring buffer */ if (ring->ring_obj == NULL) { - r = amdgpu_bo_create_kernel(adev, ring->ring_size, PAGE_SIZE, + r = amdgpu_bo_create_kernel(adev, ring->ring_size + ring->funcs->extra_dw, PAGE_SIZE, AMDGPU_GEM_DOMAIN_GTT, &ring->ring_obj, &ring->gpu_addr, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h index 1513124c5659..d242b9a51e90 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h @@ -44,6 +44,8 @@ #define AMDGPU_FENCE_FLAG_INT (1 << 1) #define AMDGPU_FENCE_FLAG_TC_WB_ONLY (1 << 2) +#define to_amdgpu_ring(s) container_of((s), struct amdgpu_ring, sched) + enum amdgpu_ring_type { AMDGPU_RING_TYPE_GFX, AMDGPU_RING_TYPE_COMPUTE, @@ -53,7 +55,8 @@ enum amdgpu_ring_type { AMDGPU_RING_TYPE_KIQ, AMDGPU_RING_TYPE_UVD_ENC, AMDGPU_RING_TYPE_VCN_DEC, - AMDGPU_RING_TYPE_VCN_ENC + AMDGPU_RING_TYPE_VCN_ENC, + AMDGPU_RING_TYPE_VCN_JPEG }; struct amdgpu_device; @@ -112,6 +115,7 @@ struct amdgpu_ring_funcs { u32 nop; bool support_64bit_ptrs; unsigned vmhub; + unsigned extra_dw; /* ring read/write ptr handling */ u64 (*get_rptr)(struct amdgpu_ring *ring); @@ -119,6 +123,7 @@ struct amdgpu_ring_funcs { void (*set_wptr)(struct amdgpu_ring *ring); /* validating and patching of IBs */ int (*parse_cs)(struct amdgpu_cs_parser *p, uint32_t ib_idx); + int (*patch_cs_in_place)(struct amdgpu_cs_parser *p, uint32_t ib_idx); /* constants to calculate how many DW are needed for an emit */ unsigned emit_frame_size; unsigned emit_ib_size; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_test.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_test.c index e3878256743a..8904e62dca7a 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_test.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_test.c @@ -1,3 +1,4 @@ +// SPDX-License-Identifier: GPL-2.0 OR MIT /* * Copyright 2009 VMware, Inc. * @@ -75,11 +76,12 @@ static void amdgpu_do_test_moves(struct amdgpu_device *adev) r = amdgpu_bo_reserve(vram_obj, false); if (unlikely(r != 0)) goto out_unref; - r = amdgpu_bo_pin(vram_obj, AMDGPU_GEM_DOMAIN_VRAM, &vram_addr); + r = amdgpu_bo_pin(vram_obj, AMDGPU_GEM_DOMAIN_VRAM); if (r) { DRM_ERROR("Failed to pin VRAM object\n"); goto out_unres; } + vram_addr = amdgpu_bo_gpu_offset(vram_obj); for (i = 0; i < n; i++) { void *gtt_map, *vram_map; void **gart_start, **gart_end; @@ -96,11 +98,17 @@ static void amdgpu_do_test_moves(struct amdgpu_device *adev) r = amdgpu_bo_reserve(gtt_obj[i], false); if (unlikely(r != 0)) goto out_lclean_unref; - r = amdgpu_bo_pin(gtt_obj[i], AMDGPU_GEM_DOMAIN_GTT, &gart_addr); + r = amdgpu_bo_pin(gtt_obj[i], AMDGPU_GEM_DOMAIN_GTT); if (r) { DRM_ERROR("Failed to pin GTT object %d\n", i); goto out_lclean_unres; } + r = amdgpu_ttm_alloc_gart(>t_obj[i]->tbo); + if (r) { + DRM_ERROR("%p bind failed\n", gtt_obj[i]); + goto out_lclean_unpin; + } + gart_addr = amdgpu_bo_gpu_offset(gtt_obj[i]); r = amdgpu_bo_kmap(gtt_obj[i], >t_map); if (r) { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h index e96e26d3f3b0..7206a0025b17 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h @@ -150,10 +150,10 @@ TRACE_EVENT(amdgpu_cs, TP_fast_assign( __entry->bo_list = p->bo_list; - __entry->ring = p->job->ring->idx; + __entry->ring = p->ring->idx; __entry->dw = p->job->ibs[i].length_dw; __entry->fences = amdgpu_fence_count_emitted( - p->job->ring); + p->ring); ), TP_printk("bo_list=%p, ring=%u, dw=%u, fences=%u", __entry->bo_list, __entry->ring, __entry->dw, @@ -178,7 +178,7 @@ TRACE_EVENT(amdgpu_cs_ioctl, __assign_str(timeline, AMDGPU_JOB_GET_TIMELINE_NAME(job)) __entry->context = job->base.s_fence->finished.context; __entry->seqno = job->base.s_fence->finished.seqno; - __entry->ring_name = job->ring->name; + __entry->ring_name = to_amdgpu_ring(job->base.sched)->name; __entry->num_ibs = job->num_ibs; ), TP_printk("sched_job=%llu, timeline=%s, context=%u, seqno=%u, ring_name=%s, num_ibs=%u", @@ -203,7 +203,7 @@ TRACE_EVENT(amdgpu_sched_run_job, __assign_str(timeline, AMDGPU_JOB_GET_TIMELINE_NAME(job)) __entry->context = job->base.s_fence->finished.context; __entry->seqno = job->base.s_fence->finished.seqno; - __entry->ring_name = job->ring->name; + __entry->ring_name = to_amdgpu_ring(job->base.sched)->name; __entry->num_ibs = job->num_ibs; ), TP_printk("sched_job=%llu, timeline=%s, context=%u, seqno=%u, ring_name=%s, num_ibs=%u", @@ -314,6 +314,11 @@ DEFINE_EVENT(amdgpu_vm_mapping, amdgpu_vm_bo_mapping, TP_ARGS(mapping) ); +DEFINE_EVENT(amdgpu_vm_mapping, amdgpu_vm_bo_cs, + TP_PROTO(struct amdgpu_bo_va_mapping *mapping), + TP_ARGS(mapping) +); + TRACE_EVENT(amdgpu_vm_set_ptes, TP_PROTO(uint64_t pe, uint64_t addr, unsigned count, uint32_t incr, uint64_t flags), @@ -436,7 +441,7 @@ TRACE_EVENT(amdgpu_cs_bo_status, __entry->total_bo, __entry->total_size) ); -TRACE_EVENT(amdgpu_ttm_bo_move, +TRACE_EVENT(amdgpu_bo_move, TP_PROTO(struct amdgpu_bo* bo, uint32_t new_placement, uint32_t old_placement), TP_ARGS(bo, new_placement, old_placement), TP_STRUCT__entry( diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c index e93a0a237dc3..fcf421263fd9 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c @@ -92,11 +92,9 @@ static void amdgpu_ttm_mem_global_release(struct drm_global_reference *ref) } /** - * amdgpu_ttm_global_init - Initialize global TTM memory reference - * structures. + * amdgpu_ttm_global_init - Initialize global TTM memory reference structures. * - * @adev: AMDGPU device for which the global structures need to be - * registered. + * @adev: AMDGPU device for which the global structures need to be registered. * * This is called as part of the AMDGPU ttm init from amdgpu_ttm_init() * during bring up. @@ -104,8 +102,6 @@ static void amdgpu_ttm_mem_global_release(struct drm_global_reference *ref) static int amdgpu_ttm_global_init(struct amdgpu_device *adev) { struct drm_global_reference *global_ref; - struct amdgpu_ring *ring; - struct drm_sched_rq *rq; int r; /* ensure reference is false in case init fails */ @@ -138,21 +134,10 @@ 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[DRM_SCHED_PRIORITY_KERNEL]; - r = drm_sched_entity_init(&ring->sched, &adev->mman.entity, - rq, NULL); - if (r) { - DRM_ERROR("Failed setting up TTM BO move run queue.\n"); - goto error_entity; - } - adev->mman.mem_global_referenced = true; return 0; -error_entity: - drm_global_item_unref(&adev->mman.bo_global_ref.ref); error_bo: drm_global_item_unref(&adev->mman.mem_global_ref); error_mem: @@ -162,8 +147,6 @@ error_mem: static void amdgpu_ttm_global_fini(struct amdgpu_device *adev) { if (adev->mman.mem_global_referenced) { - 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); drm_global_item_unref(&adev->mman.mem_global_ref); @@ -177,13 +160,12 @@ static int amdgpu_invalidate_caches(struct ttm_bo_device *bdev, uint32_t flags) } /** - * amdgpu_init_mem_type - Initialize a memory manager for a specific - * type of memory request. + * amdgpu_init_mem_type - Initialize a memory manager for a specific type of + * memory request. * - * @bdev: The TTM BO device object (contains a reference to - * amdgpu_device) - * @type: The type of memory requested - * @man: + * @bdev: The TTM BO device object (contains a reference to amdgpu_device) + * @type: The type of memory requested + * @man: The memory type manager for each domain * * This is called by ttm_bo_init_mm() when a buffer object is being * initialized. @@ -263,7 +245,7 @@ static void amdgpu_evict_flags(struct ttm_buffer_object *bo, } /* Object isn't an AMDGPU object so ignore */ - if (!amdgpu_ttm_bo_is_amdgpu_bo(bo)) { + if (!amdgpu_bo_is_amdgpu_bo(bo)) { placement->placement = &placements; placement->busy_placement = &placements; placement->num_placement = 1; @@ -276,8 +258,8 @@ static void amdgpu_evict_flags(struct ttm_buffer_object *bo, case TTM_PL_VRAM: if (!adev->mman.buffer_funcs_enabled) { /* Move to system memory */ - amdgpu_ttm_placement_from_domain(abo, AMDGPU_GEM_DOMAIN_CPU); - } else if (adev->gmc.visible_vram_size < adev->gmc.real_vram_size && + amdgpu_bo_placement_from_domain(abo, AMDGPU_GEM_DOMAIN_CPU); + } else if (!amdgpu_gmc_vram_full_visible(&adev->gmc) && !(abo->flags & AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED) && amdgpu_bo_in_cpu_visible_vram(abo)) { @@ -286,7 +268,7 @@ static void amdgpu_evict_flags(struct ttm_buffer_object *bo, * BO will be evicted to GTT rather than causing other * BOs to be evicted from VRAM */ - amdgpu_ttm_placement_from_domain(abo, AMDGPU_GEM_DOMAIN_VRAM | + amdgpu_bo_placement_from_domain(abo, AMDGPU_GEM_DOMAIN_VRAM | AMDGPU_GEM_DOMAIN_GTT); abo->placements[0].fpfn = adev->gmc.visible_vram_size >> PAGE_SHIFT; abo->placements[0].lpfn = 0; @@ -294,12 +276,12 @@ static void amdgpu_evict_flags(struct ttm_buffer_object *bo, abo->placement.num_busy_placement = 1; } else { /* Move to GTT memory */ - amdgpu_ttm_placement_from_domain(abo, AMDGPU_GEM_DOMAIN_GTT); + amdgpu_bo_placement_from_domain(abo, AMDGPU_GEM_DOMAIN_GTT); } break; case TTM_PL_TT: default: - amdgpu_ttm_placement_from_domain(abo, AMDGPU_GEM_DOMAIN_CPU); + amdgpu_bo_placement_from_domain(abo, AMDGPU_GEM_DOMAIN_CPU); } *placement = abo->placement; } @@ -307,8 +289,8 @@ static void amdgpu_evict_flags(struct ttm_buffer_object *bo, /** * amdgpu_verify_access - Verify access for a mmap call * - * @bo: The buffer object to map - * @filp: The file pointer from the process performing the mmap + * @bo: The buffer object to map + * @filp: The file pointer from the process performing the mmap * * This is called by ttm_bo_mmap() to verify whether a process * has the right to mmap a BO to their process space. @@ -333,11 +315,10 @@ static int amdgpu_verify_access(struct ttm_buffer_object *bo, struct file *filp) /** * amdgpu_move_null - Register memory for a buffer object * - * @bo: The bo to assign the memory to - * @new_mem: The memory to be assigned. + * @bo: The bo to assign the memory to + * @new_mem: The memory to be assigned. * - * Assign the memory from new_mem to the memory of the buffer object - * bo. + * Assign the memory from new_mem to the memory of the buffer object bo. */ static void amdgpu_move_null(struct ttm_buffer_object *bo, struct ttm_mem_reg *new_mem) @@ -350,8 +331,12 @@ static void amdgpu_move_null(struct ttm_buffer_object *bo, } /** - * amdgpu_mm_node_addr - Compute the GPU relative offset of a GTT - * buffer. + * amdgpu_mm_node_addr - Compute the GPU relative offset of a GTT buffer. + * + * @bo: The bo to assign the memory to. + * @mm_node: Memory manager node for drm allocator. + * @mem: The region where the bo resides. + * */ static uint64_t amdgpu_mm_node_addr(struct ttm_buffer_object *bo, struct drm_mm_node *mm_node, @@ -367,10 +352,12 @@ static uint64_t amdgpu_mm_node_addr(struct ttm_buffer_object *bo, } /** - * amdgpu_find_mm_node - Helper function finds the drm_mm_node - * corresponding to @offset. It also modifies - * the offset to be within the drm_mm_node - * returned + * amdgpu_find_mm_node - Helper function finds the drm_mm_node corresponding to + * @offset. It also modifies the offset to be within the drm_mm_node returned + * + * @mem: The region where the bo resides. + * @offset: The offset that drm_mm_node is used for finding. + * */ static struct drm_mm_node *amdgpu_find_mm_node(struct ttm_mem_reg *mem, unsigned long *offset) @@ -512,8 +499,8 @@ error: /** * amdgpu_move_blit - Copy an entire buffer to another buffer * - * This is a helper called by amdgpu_bo_move() and - * amdgpu_move_vram_ram() to help move buffers to and from VRAM. + * This is a helper called by amdgpu_bo_move() and amdgpu_move_vram_ram() to + * help move buffers to and from VRAM. */ static int amdgpu_move_blit(struct ttm_buffer_object *bo, bool evict, bool no_wait_gpu, @@ -595,7 +582,7 @@ static int amdgpu_move_vram_ram(struct ttm_buffer_object *bo, bool evict, } /* blit VRAM to GTT */ - r = amdgpu_move_blit(bo, true, ctx->no_wait_gpu, &tmp_mem, old_mem); + r = amdgpu_move_blit(bo, evict, ctx->no_wait_gpu, &tmp_mem, old_mem); if (unlikely(r)) { goto out_cleanup; } @@ -647,7 +634,7 @@ static int amdgpu_move_ram_vram(struct ttm_buffer_object *bo, bool evict, } /* copy to VRAM */ - r = amdgpu_move_blit(bo, true, ctx->no_wait_gpu, new_mem, old_mem); + r = amdgpu_move_blit(bo, evict, ctx->no_wait_gpu, new_mem, old_mem); if (unlikely(r)) { goto out_cleanup; } @@ -809,8 +796,8 @@ struct amdgpu_ttm_tt { }; /** - * amdgpu_ttm_tt_get_user_pages - Pin pages of memory pointed to - * by a USERPTR pointer to memory + * amdgpu_ttm_tt_get_user_pages - Pin pages of memory pointed to by a USERPTR + * pointer to memory * * Called by amdgpu_gem_userptr_ioctl() and amdgpu_cs_parser_bos(). * This provides a wrapper around the get_user_pages() call to provide @@ -833,8 +820,10 @@ int amdgpu_ttm_tt_get_user_pages(struct ttm_tt *ttm, struct page **pages) down_read(&mm->mmap_sem); if (gtt->userflags & AMDGPU_GEM_USERPTR_ANONONLY) { - /* check that we only use anonymous memory - to prevent problems with writeback */ + /* + * check that we only use anonymous memory to prevent problems + * with writeback + */ unsigned long end = gtt->userptr + ttm->num_pages * PAGE_SIZE; struct vm_area_struct *vma; @@ -885,10 +874,9 @@ release_pages: } /** - * amdgpu_ttm_tt_set_user_pages - Copy pages in, putting old pages - * as necessary. + * amdgpu_ttm_tt_set_user_pages - Copy pages in, putting old pages as necessary. * - * Called by amdgpu_cs_list_validate(). This creates the page list + * Called by amdgpu_cs_list_validate(). This creates the page list * that backs user memory and will ultimately be mapped into the device * address space. */ @@ -930,8 +918,7 @@ void amdgpu_ttm_tt_mark_user_pages(struct ttm_tt *ttm) } /** - * amdgpu_ttm_tt_pin_userptr - prepare the sg table with the - * user pages + * amdgpu_ttm_tt_pin_userptr - prepare the sg table with the user pages * * Called by amdgpu_ttm_backend_bind() **/ @@ -1310,8 +1297,8 @@ static void amdgpu_ttm_tt_unpopulate(struct ttm_tt *ttm) } /** - * amdgpu_ttm_tt_set_userptr - Initialize userptr GTT ttm_tt - * for the current task + * amdgpu_ttm_tt_set_userptr - Initialize userptr GTT ttm_tt for the current + * task * * @ttm: The ttm_tt object to bind this userptr object to * @addr: The address in the current tasks VM space to use @@ -1361,9 +1348,8 @@ struct mm_struct *amdgpu_ttm_tt_get_usermm(struct ttm_tt *ttm) } /** - * amdgpu_ttm_tt_affect_userptr - Determine if a ttm_tt object lays - * inside an address range for the - * current task. + * amdgpu_ttm_tt_affect_userptr - Determine if a ttm_tt object lays inside an + * address range for the current task. * */ bool amdgpu_ttm_tt_affect_userptr(struct ttm_tt *ttm, unsigned long start, @@ -1401,8 +1387,7 @@ bool amdgpu_ttm_tt_affect_userptr(struct ttm_tt *ttm, unsigned long start, } /** - * amdgpu_ttm_tt_userptr_invalidated - Has the ttm_tt object been - * invalidated? + * amdgpu_ttm_tt_userptr_invalidated - Has the ttm_tt object been invalidated? */ bool amdgpu_ttm_tt_userptr_invalidated(struct ttm_tt *ttm, int *last_invalidated) @@ -1415,10 +1400,8 @@ bool amdgpu_ttm_tt_userptr_invalidated(struct ttm_tt *ttm, } /** - * amdgpu_ttm_tt_userptr_needs_pages - Have the pages backing this - * ttm_tt object been invalidated - * since the last time they've - * been set? + * amdgpu_ttm_tt_userptr_needs_pages - Have the pages backing this ttm_tt object + * been invalidated since the last time they've been set? */ bool amdgpu_ttm_tt_userptr_needs_pages(struct ttm_tt *ttm) { @@ -1474,13 +1457,12 @@ uint64_t amdgpu_ttm_tt_pte_flags(struct amdgpu_device *adev, struct ttm_tt *ttm, } /** - * amdgpu_ttm_bo_eviction_valuable - Check to see if we can evict - * a buffer object. + * amdgpu_ttm_bo_eviction_valuable - Check to see if we can evict a buffer + * object. * - * Return true if eviction is sensible. Called by - * ttm_mem_evict_first() on behalf of ttm_bo_mem_force_space() - * which tries to evict buffer objects until it can find space - * for a new object and by ttm_bo_force_list_clean() which is + * Return true if eviction is sensible. Called by ttm_mem_evict_first() on + * behalf of ttm_bo_mem_force_space() which tries to evict buffer objects until + * it can find space for a new object and by ttm_bo_force_list_clean() which is * used to clean out a memory space. */ static bool amdgpu_ttm_bo_eviction_valuable(struct ttm_buffer_object *bo, @@ -1530,8 +1512,7 @@ static bool amdgpu_ttm_bo_eviction_valuable(struct ttm_buffer_object *bo, } /** - * amdgpu_ttm_access_memory - Read or Write memory that backs a - * buffer object. + * amdgpu_ttm_access_memory - Read or Write memory that backs a buffer object. * * @bo: The buffer object to read/write * @offset: Offset into buffer object @@ -1695,7 +1676,7 @@ static int amdgpu_ttm_fw_reserve_vram_init(struct amdgpu_device *adev) AMDGPU_GEM_DOMAIN_VRAM, adev->fw_vram_usage.start_offset, (adev->fw_vram_usage.start_offset + - adev->fw_vram_usage.size), NULL); + adev->fw_vram_usage.size)); if (r) goto error_pin; r = amdgpu_bo_kmap(adev->fw_vram_usage.reserved_bo, @@ -1719,8 +1700,8 @@ error_create: return r; } /** - * amdgpu_ttm_init - Init the memory management (ttm) as well as - * various gtt/vram related fields. + * amdgpu_ttm_init - Init the memory management (ttm) as well as various + * gtt/vram related fields. * * This initializes all of the memory space pools that the TTM layer * will need such as the GTT space (system memory mapped to the device), @@ -1871,8 +1852,7 @@ int amdgpu_ttm_init(struct amdgpu_device *adev) } /** - * amdgpu_ttm_late_init - Handle any late initialization for - * amdgpu_ttm + * amdgpu_ttm_late_init - Handle any late initialization for amdgpu_ttm */ void amdgpu_ttm_late_init(struct amdgpu_device *adev) { @@ -1921,10 +1901,30 @@ void amdgpu_ttm_set_buffer_funcs_status(struct amdgpu_device *adev, bool enable) { struct ttm_mem_type_manager *man = &adev->mman.bdev.man[TTM_PL_VRAM]; uint64_t size; + int r; - if (!adev->mman.initialized || adev->in_gpu_reset) + if (!adev->mman.initialized || adev->in_gpu_reset || + adev->mman.buffer_funcs_enabled == enable) return; + if (enable) { + struct amdgpu_ring *ring; + struct drm_sched_rq *rq; + + ring = adev->mman.buffer_funcs_ring; + rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_KERNEL]; + r = drm_sched_entity_init(&adev->mman.entity, &rq, 1, NULL); + if (r) { + DRM_ERROR("Failed setting up TTM BO move entity (%d)\n", + r); + return; + } + } else { + drm_sched_entity_destroy(&adev->mman.entity); + dma_fence_put(man->move); + man->move = NULL; + } + /* this just adjusts TTM size idea, which sets lpfn to the correct value */ if (enable) size = adev->gmc.real_vram_size; @@ -2002,7 +2002,7 @@ static int amdgpu_map_buffer(struct ttm_buffer_object *bo, if (r) goto error_free; - r = amdgpu_job_submit(job, ring, &adev->mman.entity, + r = amdgpu_job_submit(job, &adev->mman.entity, AMDGPU_FENCE_OWNER_UNDEFINED, &fence); if (r) goto error_free; @@ -2071,24 +2071,19 @@ int amdgpu_copy_buffer(struct amdgpu_ring *ring, uint64_t src_offset, amdgpu_ring_pad_ib(ring, &job->ibs[0]); WARN_ON(job->ibs[0].length_dw > num_dw); - if (direct_submit) { - r = amdgpu_ib_schedule(ring, job->num_ibs, job->ibs, - NULL, fence); - job->fence = dma_fence_get(*fence); - if (r) - DRM_ERROR("Error scheduling IBs (%d)\n", r); - amdgpu_job_free(job); - } else { - r = amdgpu_job_submit(job, ring, &adev->mman.entity, + if (direct_submit) + r = amdgpu_job_submit_direct(job, ring, fence); + else + r = amdgpu_job_submit(job, &adev->mman.entity, AMDGPU_FENCE_OWNER_UNDEFINED, fence); - if (r) - goto error_free; - } + if (r) + goto error_free; return r; error_free: amdgpu_job_free(job); + DRM_ERROR("Error scheduling IBs (%d)\n", r); return r; } @@ -2171,7 +2166,7 @@ int amdgpu_fill_buffer(struct amdgpu_bo *bo, amdgpu_ring_pad_ib(ring, &job->ibs[0]); WARN_ON(job->ibs[0].length_dw > num_dw); - r = amdgpu_job_submit(job, ring, &adev->mman.entity, + r = amdgpu_job_submit(job, &adev->mman.entity, AMDGPU_FENCE_OWNER_UNDEFINED, fence); if (r) goto error_free; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h index e5da4654b630..8b3cc6687769 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h @@ -73,7 +73,7 @@ bool amdgpu_gtt_mgr_has_gart_addr(struct ttm_mem_reg *mem); uint64_t amdgpu_gtt_mgr_usage(struct ttm_mem_type_manager *man); int amdgpu_gtt_mgr_recover(struct ttm_mem_type_manager *man); -u64 amdgpu_vram_mgr_bo_invisible_size(struct amdgpu_bo *bo); +u64 amdgpu_vram_mgr_bo_visible_size(struct amdgpu_bo *bo); uint64_t amdgpu_vram_mgr_usage(struct ttm_mem_type_manager *man); uint64_t amdgpu_vram_mgr_vis_usage(struct ttm_mem_type_manager *man); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c index 3ff08e326838..632fa5980ff4 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c @@ -53,11 +53,11 @@ /* Firmware Names */ #ifdef CONFIG_DRM_AMDGPU_CIK -#define FIRMWARE_BONAIRE "radeon/bonaire_uvd.bin" -#define FIRMWARE_KABINI "radeon/kabini_uvd.bin" -#define FIRMWARE_KAVERI "radeon/kaveri_uvd.bin" -#define FIRMWARE_HAWAII "radeon/hawaii_uvd.bin" -#define FIRMWARE_MULLINS "radeon/mullins_uvd.bin" +#define FIRMWARE_BONAIRE "amdgpu/bonaire_uvd.bin" +#define FIRMWARE_KABINI "amdgpu/kabini_uvd.bin" +#define FIRMWARE_KAVERI "amdgpu/kaveri_uvd.bin" +#define FIRMWARE_HAWAII "amdgpu/hawaii_uvd.bin" +#define FIRMWARE_MULLINS "amdgpu/mullins_uvd.bin" #endif #define FIRMWARE_TONGA "amdgpu/tonga_uvd.bin" #define FIRMWARE_CARRIZO "amdgpu/carrizo_uvd.bin" @@ -127,7 +127,7 @@ int amdgpu_uvd_sw_init(struct amdgpu_device *adev) unsigned long bo_size; const char *fw_name; const struct common_firmware_header *hdr; - unsigned version_major, version_minor, family_id; + unsigned family_id; int i, j, r; INIT_DELAYED_WORK(&adev->uvd.idle_work, amdgpu_uvd_idle_work_handler); @@ -208,29 +208,46 @@ int amdgpu_uvd_sw_init(struct amdgpu_device *adev) hdr = (const struct common_firmware_header *)adev->uvd.fw->data; family_id = le32_to_cpu(hdr->ucode_version) & 0xff; - version_major = (le32_to_cpu(hdr->ucode_version) >> 24) & 0xff; - version_minor = (le32_to_cpu(hdr->ucode_version) >> 8) & 0xff; - DRM_INFO("Found UVD firmware Version: %hu.%hu Family ID: %hu\n", - version_major, version_minor, family_id); - - /* - * Limit the number of UVD handles depending on microcode major - * and minor versions. The firmware version which has 40 UVD - * instances support is 1.80. So all subsequent versions should - * also have the same support. - */ - if ((version_major > 0x01) || - ((version_major == 0x01) && (version_minor >= 0x50))) - adev->uvd.max_handles = AMDGPU_MAX_UVD_HANDLES; - adev->uvd.fw_version = ((version_major << 24) | (version_minor << 16) | - (family_id << 8)); + if (adev->asic_type < CHIP_VEGA20) { + unsigned version_major, version_minor; - if ((adev->asic_type == CHIP_POLARIS10 || - adev->asic_type == CHIP_POLARIS11) && - (adev->uvd.fw_version < FW_1_66_16)) - DRM_ERROR("POLARIS10/11 UVD firmware version %hu.%hu is too old.\n", - version_major, version_minor); + version_major = (le32_to_cpu(hdr->ucode_version) >> 24) & 0xff; + version_minor = (le32_to_cpu(hdr->ucode_version) >> 8) & 0xff; + DRM_INFO("Found UVD firmware Version: %hu.%hu Family ID: %hu\n", + version_major, version_minor, family_id); + + /* + * Limit the number of UVD handles depending on microcode major + * and minor versions. The firmware version which has 40 UVD + * instances support is 1.80. So all subsequent versions should + * also have the same support. + */ + if ((version_major > 0x01) || + ((version_major == 0x01) && (version_minor >= 0x50))) + adev->uvd.max_handles = AMDGPU_MAX_UVD_HANDLES; + + adev->uvd.fw_version = ((version_major << 24) | (version_minor << 16) | + (family_id << 8)); + + if ((adev->asic_type == CHIP_POLARIS10 || + adev->asic_type == CHIP_POLARIS11) && + (adev->uvd.fw_version < FW_1_66_16)) + DRM_ERROR("POLARIS10/11 UVD firmware version %hu.%hu is too old.\n", + version_major, version_minor); + } else { + unsigned int enc_major, enc_minor, dec_minor; + + dec_minor = (le32_to_cpu(hdr->ucode_version) >> 8) & 0xff; + enc_minor = (le32_to_cpu(hdr->ucode_version) >> 24) & 0x3f; + enc_major = (le32_to_cpu(hdr->ucode_version) >> 30) & 0x3; + DRM_INFO("Found UVD firmware ENC: %hu.%hu DEC: .%hu Family ID: %hu\n", + enc_major, enc_minor, dec_minor, family_id); + + adev->uvd.max_handles = AMDGPU_MAX_UVD_HANDLES; + + adev->uvd.fw_version = le32_to_cpu(hdr->ucode_version); + } bo_size = AMDGPU_UVD_STACK_SIZE + AMDGPU_UVD_HEAP_SIZE + AMDGPU_UVD_SESSION_SIZE * adev->uvd.max_handles; @@ -238,7 +255,8 @@ int amdgpu_uvd_sw_init(struct amdgpu_device *adev) bo_size += AMDGPU_GPU_PAGE_ALIGN(le32_to_cpu(hdr->ucode_size_bytes) + 8); for (j = 0; j < adev->uvd.num_uvd_inst; j++) { - + if (adev->uvd.harvest_config & (1 << j)) + continue; r = amdgpu_bo_create_kernel(adev, bo_size, PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM, &adev->uvd.inst[j].vcpu_bo, &adev->uvd.inst[j].gpu_addr, &adev->uvd.inst[j].cpu_addr); @@ -246,21 +264,20 @@ int amdgpu_uvd_sw_init(struct amdgpu_device *adev) dev_err(adev->dev, "(%d) failed to allocate UVD bo\n", r); return r; } + } - ring = &adev->uvd.inst[j].ring; - rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL]; - r = drm_sched_entity_init(&ring->sched, &adev->uvd.inst[j].entity, - rq, NULL); - if (r != 0) { - DRM_ERROR("Failed setting up UVD(%d) run queue.\n", j); - return r; - } - - for (i = 0; i < adev->uvd.max_handles; ++i) { - atomic_set(&adev->uvd.inst[j].handles[i], 0); - adev->uvd.inst[j].filp[i] = NULL; - } + ring = &adev->uvd.inst[0].ring; + rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL]; + r = drm_sched_entity_init(&adev->uvd.entity, &rq, 1, NULL); + if (r) { + DRM_ERROR("Failed setting up UVD kernel entity.\n"); + return r; + } + for (i = 0; i < adev->uvd.max_handles; ++i) { + atomic_set(&adev->uvd.handles[i], 0); + adev->uvd.filp[i] = NULL; } + /* from uvd v5.0 HW addressing capacity increased to 64 bits */ if (!amdgpu_device_ip_block_version_cmp(adev, AMD_IP_BLOCK_TYPE_UVD, 5, 0)) adev->uvd.address_64_bit = true; @@ -289,11 +306,13 @@ int amdgpu_uvd_sw_fini(struct amdgpu_device *adev) { int i, j; + drm_sched_entity_destroy(&adev->uvd.entity); + for (j = 0; j < adev->uvd.num_uvd_inst; ++j) { + if (adev->uvd.harvest_config & (1 << j)) + continue; kfree(adev->uvd.inst[j].saved_bo); - drm_sched_entity_fini(&adev->uvd.inst[j].ring.sched, &adev->uvd.inst[j].entity); - amdgpu_bo_free_kernel(&adev->uvd.inst[j].vcpu_bo, &adev->uvd.inst[j].gpu_addr, (void **)&adev->uvd.inst[j].cpu_addr); @@ -316,20 +335,22 @@ int amdgpu_uvd_suspend(struct amdgpu_device *adev) cancel_delayed_work_sync(&adev->uvd.idle_work); + /* only valid for physical mode */ + if (adev->asic_type < CHIP_POLARIS10) { + for (i = 0; i < adev->uvd.max_handles; ++i) + if (atomic_read(&adev->uvd.handles[i])) + break; + + if (i == adev->uvd.max_handles) + return 0; + } + for (j = 0; j < adev->uvd.num_uvd_inst; ++j) { + if (adev->uvd.harvest_config & (1 << j)) + continue; if (adev->uvd.inst[j].vcpu_bo == NULL) continue; - /* only valid for physical mode */ - if (adev->asic_type < CHIP_POLARIS10) { - for (i = 0; i < adev->uvd.max_handles; ++i) - if (atomic_read(&adev->uvd.inst[j].handles[i])) - break; - - if (i == adev->uvd.max_handles) - continue; - } - size = amdgpu_bo_size(adev->uvd.inst[j].vcpu_bo); ptr = adev->uvd.inst[j].cpu_addr; @@ -349,6 +370,8 @@ int amdgpu_uvd_resume(struct amdgpu_device *adev) int i; for (i = 0; i < adev->uvd.num_uvd_inst; i++) { + if (adev->uvd.harvest_config & (1 << i)) + continue; if (adev->uvd.inst[i].vcpu_bo == NULL) return -EINVAL; @@ -381,30 +404,27 @@ int amdgpu_uvd_resume(struct amdgpu_device *adev) void amdgpu_uvd_free_handles(struct amdgpu_device *adev, struct drm_file *filp) { - struct amdgpu_ring *ring; - int i, j, r; - - for (j = 0; j < adev->uvd.num_uvd_inst; j++) { - ring = &adev->uvd.inst[j].ring; + struct amdgpu_ring *ring = &adev->uvd.inst[0].ring; + int i, r; - for (i = 0; i < adev->uvd.max_handles; ++i) { - uint32_t handle = atomic_read(&adev->uvd.inst[j].handles[i]); - if (handle != 0 && adev->uvd.inst[j].filp[i] == filp) { - struct dma_fence *fence; - - r = amdgpu_uvd_get_destroy_msg(ring, handle, - false, &fence); - if (r) { - DRM_ERROR("Error destroying UVD(%d) %d!\n", j, r); - continue; - } + for (i = 0; i < adev->uvd.max_handles; ++i) { + uint32_t handle = atomic_read(&adev->uvd.handles[i]); - dma_fence_wait(fence, false); - dma_fence_put(fence); + if (handle != 0 && adev->uvd.filp[i] == filp) { + struct dma_fence *fence; - adev->uvd.inst[j].filp[i] = NULL; - atomic_set(&adev->uvd.inst[j].handles[i], 0); + r = amdgpu_uvd_get_destroy_msg(ring, handle, false, + &fence); + if (r) { + DRM_ERROR("Error destroying UVD %d!\n", r); + continue; } + + dma_fence_wait(fence, false); + dma_fence_put(fence); + + adev->uvd.filp[i] = NULL; + atomic_set(&adev->uvd.handles[i], 0); } } } @@ -459,7 +479,7 @@ static int amdgpu_uvd_cs_pass1(struct amdgpu_uvd_cs_ctx *ctx) if (cmd == 0x0 || cmd == 0x3) { /* yes, force it into VRAM */ uint32_t domain = AMDGPU_GEM_DOMAIN_VRAM; - amdgpu_ttm_placement_from_domain(bo, domain); + amdgpu_bo_placement_from_domain(bo, domain); } amdgpu_uvd_force_into_uvd_segment(bo); @@ -679,16 +699,15 @@ static int amdgpu_uvd_cs_msg(struct amdgpu_uvd_cs_ctx *ctx, void *ptr; long r; int i; - uint32_t ip_instance = ctx->parser->job->ring->me; if (offset & 0x3F) { - DRM_ERROR("UVD(%d) messages must be 64 byte aligned!\n", ip_instance); + DRM_ERROR("UVD messages must be 64 byte aligned!\n"); return -EINVAL; } r = amdgpu_bo_kmap(bo, &ptr); if (r) { - DRM_ERROR("Failed mapping the UVD(%d) message (%ld)!\n", ip_instance, r); + DRM_ERROR("Failed mapping the UVD) message (%ld)!\n", r); return r; } @@ -698,7 +717,7 @@ static int amdgpu_uvd_cs_msg(struct amdgpu_uvd_cs_ctx *ctx, handle = msg[2]; if (handle == 0) { - DRM_ERROR("Invalid UVD(%d) handle!\n", ip_instance); + DRM_ERROR("Invalid UVD handle!\n"); return -EINVAL; } @@ -709,18 +728,19 @@ static int amdgpu_uvd_cs_msg(struct amdgpu_uvd_cs_ctx *ctx, /* try to alloc a new handle */ for (i = 0; i < adev->uvd.max_handles; ++i) { - if (atomic_read(&adev->uvd.inst[ip_instance].handles[i]) == handle) { - DRM_ERROR("(%d)Handle 0x%x already in use!\n", ip_instance, handle); + if (atomic_read(&adev->uvd.handles[i]) == handle) { + DRM_ERROR(")Handle 0x%x already in use!\n", + handle); return -EINVAL; } - if (!atomic_cmpxchg(&adev->uvd.inst[ip_instance].handles[i], 0, handle)) { - adev->uvd.inst[ip_instance].filp[i] = ctx->parser->filp; + if (!atomic_cmpxchg(&adev->uvd.handles[i], 0, handle)) { + adev->uvd.filp[i] = ctx->parser->filp; return 0; } } - DRM_ERROR("No more free UVD(%d) handles!\n", ip_instance); + DRM_ERROR("No more free UVD handles!\n"); return -ENOSPC; case 1: @@ -732,27 +752,27 @@ static int amdgpu_uvd_cs_msg(struct amdgpu_uvd_cs_ctx *ctx, /* validate the handle */ for (i = 0; i < adev->uvd.max_handles; ++i) { - if (atomic_read(&adev->uvd.inst[ip_instance].handles[i]) == handle) { - if (adev->uvd.inst[ip_instance].filp[i] != ctx->parser->filp) { - DRM_ERROR("UVD(%d) handle collision detected!\n", ip_instance); + if (atomic_read(&adev->uvd.handles[i]) == handle) { + if (adev->uvd.filp[i] != ctx->parser->filp) { + DRM_ERROR("UVD handle collision detected!\n"); return -EINVAL; } return 0; } } - DRM_ERROR("Invalid UVD(%d) handle 0x%x!\n", ip_instance, handle); + DRM_ERROR("Invalid UVD handle 0x%x!\n", handle); return -ENOENT; case 2: /* it's a destroy msg, free the handle */ for (i = 0; i < adev->uvd.max_handles; ++i) - atomic_cmpxchg(&adev->uvd.inst[ip_instance].handles[i], handle, 0); + atomic_cmpxchg(&adev->uvd.handles[i], handle, 0); amdgpu_bo_kunmap(bo); return 0; default: - DRM_ERROR("Illegal UVD(%d) message type (%d)!\n", ip_instance, msg_type); + DRM_ERROR("Illegal UVD message type (%d)!\n", msg_type); return -EINVAL; } BUG(); @@ -1000,7 +1020,7 @@ static int amdgpu_uvd_send_msg(struct amdgpu_ring *ring, struct amdgpu_bo *bo, if (!ring->adev->uvd.address_64_bit) { struct ttm_operation_ctx ctx = { true, false }; - amdgpu_ttm_placement_from_domain(bo, AMDGPU_GEM_DOMAIN_VRAM); + amdgpu_bo_placement_from_domain(bo, AMDGPU_GEM_DOMAIN_VRAM); amdgpu_uvd_force_into_uvd_segment(bo); r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx); if (r) @@ -1045,19 +1065,16 @@ static int amdgpu_uvd_send_msg(struct amdgpu_ring *ring, struct amdgpu_bo *bo, if (r < 0) goto err_free; - r = amdgpu_ib_schedule(ring, 1, ib, NULL, &f); - job->fence = dma_fence_get(f); + r = amdgpu_job_submit_direct(job, ring, &f); if (r) goto err_free; - - amdgpu_job_free(job); } else { r = amdgpu_sync_resv(adev, &job->sync, bo->tbo.resv, AMDGPU_FENCE_OWNER_UNDEFINED, false); if (r) goto err_free; - r = amdgpu_job_submit(job, ring, &adev->uvd.inst[ring->me].entity, + r = amdgpu_job_submit(job, &adev->uvd.entity, AMDGPU_FENCE_OWNER_UNDEFINED, &f); if (r) goto err_free; @@ -1149,6 +1166,8 @@ static void amdgpu_uvd_idle_work_handler(struct work_struct *work) unsigned fences = 0, i, j; for (i = 0; i < adev->uvd.num_uvd_inst; ++i) { + if (adev->uvd.harvest_config & (1 << i)) + continue; fences += amdgpu_fence_count_emitted(&adev->uvd.inst[i].ring); for (j = 0; j < adev->uvd.num_enc_rings; ++j) { fences += amdgpu_fence_count_emitted(&adev->uvd.inst[i].ring_enc[j]); @@ -1259,7 +1278,7 @@ uint32_t amdgpu_uvd_used_handles(struct amdgpu_device *adev) * necessarily linear. So we need to count * all non-zero handles. */ - if (atomic_read(&adev->uvd.inst->handles[i])) + if (atomic_read(&adev->uvd.handles[i])) used_handles++; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.h index 8b23a1b00c76..33c5f806f925 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.h @@ -42,26 +42,29 @@ struct amdgpu_uvd_inst { void *cpu_addr; uint64_t gpu_addr; void *saved_bo; - atomic_t handles[AMDGPU_MAX_UVD_HANDLES]; - struct drm_file *filp[AMDGPU_MAX_UVD_HANDLES]; struct amdgpu_ring ring; struct amdgpu_ring ring_enc[AMDGPU_MAX_UVD_ENC_RINGS]; struct amdgpu_irq_src irq; - struct drm_sched_entity entity; - struct drm_sched_entity entity_enc; uint32_t srbm_soft_reset; }; +#define AMDGPU_UVD_HARVEST_UVD0 (1 << 0) +#define AMDGPU_UVD_HARVEST_UVD1 (1 << 1) + struct amdgpu_uvd { const struct firmware *fw; /* UVD firmware */ unsigned fw_version; unsigned max_handles; unsigned num_enc_rings; - uint8_t num_uvd_inst; + uint8_t num_uvd_inst; bool address_64_bit; bool use_ctx_buf; - struct amdgpu_uvd_inst inst[AMDGPU_MAX_UVD_INSTANCES]; + struct amdgpu_uvd_inst inst[AMDGPU_MAX_UVD_INSTANCES]; + struct drm_file *filp[AMDGPU_MAX_UVD_HANDLES]; + atomic_t handles[AMDGPU_MAX_UVD_HANDLES]; + struct drm_sched_entity entity; struct delayed_work idle_work; + unsigned harvest_config; }; int amdgpu_uvd_sw_init(struct amdgpu_device *adev); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c index 23d960ec1cf2..b6ab4f5350c8 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c @@ -40,11 +40,11 @@ /* Firmware Names */ #ifdef CONFIG_DRM_AMDGPU_CIK -#define FIRMWARE_BONAIRE "radeon/bonaire_vce.bin" -#define FIRMWARE_KABINI "radeon/kabini_vce.bin" -#define FIRMWARE_KAVERI "radeon/kaveri_vce.bin" -#define FIRMWARE_HAWAII "radeon/hawaii_vce.bin" -#define FIRMWARE_MULLINS "radeon/mullins_vce.bin" +#define FIRMWARE_BONAIRE "amdgpu/bonaire_vce.bin" +#define FIRMWARE_KABINI "amdgpu/kabini_vce.bin" +#define FIRMWARE_KAVERI "amdgpu/kaveri_vce.bin" +#define FIRMWARE_HAWAII "amdgpu/hawaii_vce.bin" +#define FIRMWARE_MULLINS "amdgpu/mullins_vce.bin" #endif #define FIRMWARE_TONGA "amdgpu/tonga_vce.bin" #define FIRMWARE_CARRIZO "amdgpu/carrizo_vce.bin" @@ -190,8 +190,7 @@ int amdgpu_vce_sw_init(struct amdgpu_device *adev, unsigned long size) ring = &adev->vce.ring[0]; rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL]; - r = drm_sched_entity_init(&ring->sched, &adev->vce.entity, - rq, NULL); + r = drm_sched_entity_init(&adev->vce.entity, &rq, 1, NULL); if (r != 0) { DRM_ERROR("Failed setting up VCE run queue.\n"); return r; @@ -222,7 +221,7 @@ int amdgpu_vce_sw_fini(struct amdgpu_device *adev) if (adev->vce.vcpu_bo == NULL) return 0; - drm_sched_entity_fini(&adev->vce.ring[0].sched, &adev->vce.entity); + drm_sched_entity_destroy(&adev->vce.entity); amdgpu_bo_free_kernel(&adev->vce.vcpu_bo, &adev->vce.gpu_addr, (void **)&adev->vce.cpu_addr); @@ -470,12 +469,10 @@ int amdgpu_vce_get_create_msg(struct amdgpu_ring *ring, uint32_t handle, for (i = ib->length_dw; i < ib_size_dw; ++i) ib->ptr[i] = 0x0; - r = amdgpu_ib_schedule(ring, 1, ib, NULL, &f); - job->fence = dma_fence_get(f); + r = amdgpu_job_submit_direct(job, ring, &f); if (r) goto err; - amdgpu_job_free(job); if (fence) *fence = dma_fence_get(f); dma_fence_put(f); @@ -532,19 +529,13 @@ int amdgpu_vce_get_destroy_msg(struct amdgpu_ring *ring, uint32_t handle, for (i = ib->length_dw; i < ib_size_dw; ++i) ib->ptr[i] = 0x0; - if (direct) { - r = amdgpu_ib_schedule(ring, 1, ib, NULL, &f); - job->fence = dma_fence_get(f); - if (r) - goto err; - - amdgpu_job_free(job); - } else { - r = amdgpu_job_submit(job, ring, &ring->adev->vce.entity, + if (direct) + r = amdgpu_job_submit_direct(job, ring, &f); + else + r = amdgpu_job_submit(job, &ring->adev->vce.entity, AMDGPU_FENCE_OWNER_UNDEFINED, &f); - if (r) - goto err; - } + if (r) + goto err; if (fence) *fence = dma_fence_get(f); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c index 1b4ad9b2a755..798648a19710 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c @@ -140,6 +140,8 @@ int amdgpu_vcn_sw_fini(struct amdgpu_device *adev) for (i = 0; i < adev->vcn.num_enc_rings; ++i) amdgpu_ring_fini(&adev->vcn.ring_enc[i]); + amdgpu_ring_fini(&adev->vcn.ring_jpeg); + release_firmware(adev->vcn.fw); return 0; @@ -209,6 +211,8 @@ static void amdgpu_vcn_idle_work_handler(struct work_struct *work) fences += amdgpu_fence_count_emitted(&adev->vcn.ring_enc[i]); } + fences += amdgpu_fence_count_emitted(&adev->vcn.ring_jpeg); + if (fences == 0) { if (adev->pm.dpm_enabled) amdgpu_dpm_enable_uvd(adev, false); @@ -225,7 +229,7 @@ void amdgpu_vcn_ring_begin_use(struct amdgpu_ring *ring) struct amdgpu_device *adev = ring->adev; bool set_clocks = !cancel_delayed_work_sync(&adev->vcn.idle_work); - if (set_clocks && adev->pm.dpm_enabled) { + if (set_clocks) { if (adev->pm.dpm_enabled) amdgpu_dpm_enable_uvd(adev, true); else @@ -304,13 +308,10 @@ static int amdgpu_vcn_dec_send_msg(struct amdgpu_ring *ring, } ib->length_dw = 16; - r = amdgpu_ib_schedule(ring, 1, ib, NULL, &f); - job->fence = dma_fence_get(f); + r = amdgpu_job_submit_direct(job, ring, &f); if (r) goto err_free; - amdgpu_job_free(job); - amdgpu_bo_fence(bo, f, false); amdgpu_bo_unreserve(bo); amdgpu_bo_unref(&bo); @@ -495,12 +496,10 @@ static int amdgpu_vcn_enc_get_create_msg(struct amdgpu_ring *ring, uint32_t hand for (i = ib->length_dw; i < ib_size_dw; ++i) ib->ptr[i] = 0x0; - r = amdgpu_ib_schedule(ring, 1, ib, NULL, &f); - job->fence = dma_fence_get(f); + r = amdgpu_job_submit_direct(job, ring, &f); if (r) goto err; - amdgpu_job_free(job); if (fence) *fence = dma_fence_get(f); dma_fence_put(f); @@ -549,12 +548,10 @@ static int amdgpu_vcn_enc_get_destroy_msg(struct amdgpu_ring *ring, uint32_t han for (i = ib->length_dw; i < ib_size_dw; ++i) ib->ptr[i] = 0x0; - r = amdgpu_ib_schedule(ring, 1, ib, NULL, &f); - job->fence = dma_fence_get(f); + r = amdgpu_job_submit_direct(job, ring, &f); if (r) goto err; - amdgpu_job_free(job); if (fence) *fence = dma_fence_get(f); dma_fence_put(f); @@ -597,3 +594,127 @@ error: dma_fence_put(fence); return r; } + +int amdgpu_vcn_jpeg_ring_test_ring(struct amdgpu_ring *ring) +{ + struct amdgpu_device *adev = ring->adev; + uint32_t tmp = 0; + unsigned i; + int r; + + WREG32(SOC15_REG_OFFSET(UVD, 0, mmUVD_CONTEXT_ID), 0xCAFEDEAD); + r = amdgpu_ring_alloc(ring, 3); + + if (r) { + DRM_ERROR("amdgpu: cp failed to lock ring %d (%d).\n", + ring->idx, r); + return r; + } + + amdgpu_ring_write(ring, + PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_CONTEXT_ID), 0, 0, 0)); + amdgpu_ring_write(ring, 0xDEADBEEF); + amdgpu_ring_commit(ring); + + for (i = 0; i < adev->usec_timeout; i++) { + tmp = RREG32(SOC15_REG_OFFSET(UVD, 0, mmUVD_CONTEXT_ID)); + if (tmp == 0xDEADBEEF) + break; + DRM_UDELAY(1); + } + + if (i < adev->usec_timeout) { + DRM_DEBUG("ring test on %d succeeded in %d usecs\n", + ring->idx, i); + } else { + DRM_ERROR("amdgpu: ring %d test failed (0x%08X)\n", + ring->idx, tmp); + r = -EINVAL; + } + + return r; +} + +static int amdgpu_vcn_jpeg_set_reg(struct amdgpu_ring *ring, uint32_t handle, + struct dma_fence **fence) +{ + struct amdgpu_device *adev = ring->adev; + struct amdgpu_job *job; + struct amdgpu_ib *ib; + struct dma_fence *f = NULL; + const unsigned ib_size_dw = 16; + int i, r; + + r = amdgpu_job_alloc_with_ib(ring->adev, ib_size_dw * 4, &job); + if (r) + return r; + + ib = &job->ibs[0]; + + ib->ptr[0] = PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JPEG_PITCH), 0, 0, PACKETJ_TYPE0); + ib->ptr[1] = 0xDEADBEEF; + for (i = 2; i < 16; i += 2) { + ib->ptr[i] = PACKETJ(0, 0, 0, PACKETJ_TYPE6); + ib->ptr[i+1] = 0; + } + ib->length_dw = 16; + + r = amdgpu_job_submit_direct(job, ring, &f); + if (r) + goto err; + + if (fence) + *fence = dma_fence_get(f); + dma_fence_put(f); + + return 0; + +err: + amdgpu_job_free(job); + return r; +} + +int amdgpu_vcn_jpeg_ring_test_ib(struct amdgpu_ring *ring, long timeout) +{ + struct amdgpu_device *adev = ring->adev; + uint32_t tmp = 0; + unsigned i; + struct dma_fence *fence = NULL; + long r = 0; + + r = amdgpu_vcn_jpeg_set_reg(ring, 1, &fence); + if (r) { + DRM_ERROR("amdgpu: failed to set jpeg register (%ld).\n", r); + goto error; + } + + r = dma_fence_wait_timeout(fence, false, timeout); + if (r == 0) { + DRM_ERROR("amdgpu: IB test timed out.\n"); + r = -ETIMEDOUT; + goto error; + } else if (r < 0) { + DRM_ERROR("amdgpu: fence wait failed (%ld).\n", r); + goto error; + } else + r = 0; + + for (i = 0; i < adev->usec_timeout; i++) { + tmp = RREG32(SOC15_REG_OFFSET(UVD, 0, mmUVD_JPEG_PITCH)); + if (tmp == 0xDEADBEEF) + break; + DRM_UDELAY(1); + } + + if (i < adev->usec_timeout) + DRM_DEBUG("ib test on ring %d succeeded\n", ring->idx); + else { + DRM_ERROR("ib test failed (0x%08X)\n", tmp); + r = -EINVAL; + } + + dma_fence_put(fence); + +error: + return r; +} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h index 773010b9ff15..0b0b8638d73f 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h @@ -66,6 +66,7 @@ struct amdgpu_vcn { const struct firmware *fw; /* VCN firmware */ struct amdgpu_ring ring_dec; struct amdgpu_ring ring_enc[AMDGPU_VCN_MAX_ENC_RINGS]; + struct amdgpu_ring ring_jpeg; struct amdgpu_irq_src irq; unsigned num_enc_rings; }; @@ -83,4 +84,7 @@ int amdgpu_vcn_dec_ring_test_ib(struct amdgpu_ring *ring, long timeout); int amdgpu_vcn_enc_ring_test_ring(struct amdgpu_ring *ring); int amdgpu_vcn_enc_ring_test_ib(struct amdgpu_ring *ring, long timeout); +int amdgpu_vcn_jpeg_ring_test_ring(struct amdgpu_ring *ring); +int amdgpu_vcn_jpeg_ring_test_ib(struct amdgpu_ring *ring, long timeout); + #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index fdcb498f6d19..ece0ac703e27 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -33,9 +33,11 @@ #include "amdgpu.h" #include "amdgpu_trace.h" #include "amdgpu_amdkfd.h" +#include "amdgpu_gmc.h" -/* - * GPUVM +/** + * DOC: GPUVM + * * GPUVM is similar to the legacy gart on older asics, however * rather than there being a single global gart table * for the entire GPU, there are multiple VM page tables active @@ -63,37 +65,84 @@ INTERVAL_TREE_DEFINE(struct amdgpu_bo_va_mapping, rb, uint64_t, __subtree_last, #undef START #undef LAST -/* Local structure. Encapsulate some VM table update parameters to reduce +/** + * struct amdgpu_pte_update_params - Local structure + * + * Encapsulate some VM table update parameters to reduce * the number of function parameters + * */ struct amdgpu_pte_update_params { - /* amdgpu device we do this update for */ + + /** + * @adev: amdgpu device we do this update for + */ struct amdgpu_device *adev; - /* optional amdgpu_vm we do this update for */ + + /** + * @vm: optional amdgpu_vm we do this update for + */ struct amdgpu_vm *vm; - /* address where to copy page table entries from */ + + /** + * @src: address where to copy page table entries from + */ uint64_t src; - /* indirect buffer to fill with commands */ + + /** + * @ib: indirect buffer to fill with commands + */ struct amdgpu_ib *ib; - /* Function which actually does the update */ + + /** + * @func: Function which actually does the update + */ void (*func)(struct amdgpu_pte_update_params *params, struct amdgpu_bo *bo, uint64_t pe, uint64_t addr, unsigned count, uint32_t incr, uint64_t flags); - /* The next two are used during VM update by CPU - * DMA addresses to use for mapping - * Kernel pointer of PD/PT BO that needs to be updated + /** + * @pages_addr: + * + * DMA addresses to use for mapping, used during VM update by CPU */ dma_addr_t *pages_addr; + + /** + * @kptr: + * + * Kernel pointer of PD/PT BO that needs to be updated, + * used during VM update by CPU + */ void *kptr; }; -/* Helper to disable partial resident texture feature from a fence callback */ +/** + * struct amdgpu_prt_cb - Helper to disable partial resident texture feature from a fence callback + */ struct amdgpu_prt_cb { + + /** + * @adev: amdgpu device + */ struct amdgpu_device *adev; + + /** + * @cb: callback + */ struct dma_fence_cb cb; }; +/** + * amdgpu_vm_bo_base_init - Adds bo to the list of bos associated with the vm + * + * @base: base structure for tracking BO usage in a VM + * @vm: vm to which bo is to be added + * @bo: amdgpu buffer object + * + * Initialize a bo_va_base structure and add it to the appropriate lists + * + */ static void amdgpu_vm_bo_base_init(struct amdgpu_vm_bo_base *base, struct amdgpu_vm *vm, struct amdgpu_bo *bo) @@ -129,8 +178,10 @@ static void amdgpu_vm_bo_base_init(struct amdgpu_vm_bo_base *base, * amdgpu_vm_level_shift - return the addr shift for each level * * @adev: amdgpu_device pointer + * @level: VMPT level * - * Returns the number of bits the pfn needs to be right shifted for a level. + * Returns: + * The number of bits the pfn needs to be right shifted for a level. */ static unsigned amdgpu_vm_level_shift(struct amdgpu_device *adev, unsigned level) @@ -158,8 +209,10 @@ static unsigned amdgpu_vm_level_shift(struct amdgpu_device *adev, * amdgpu_vm_num_entries - return the number of entries in a PD/PT * * @adev: amdgpu_device pointer + * @level: VMPT level * - * Calculate the number of entries in a page directory or page table. + * Returns: + * The number of entries in a page directory or page table. */ static unsigned amdgpu_vm_num_entries(struct amdgpu_device *adev, unsigned level) @@ -182,8 +235,10 @@ static unsigned amdgpu_vm_num_entries(struct amdgpu_device *adev, * amdgpu_vm_bo_size - returns the size of the BOs in bytes * * @adev: amdgpu_device pointer + * @level: VMPT level * - * Calculate the size of the BO for a page directory or page table in bytes. + * Returns: + * The size of the BO for a page directory or page table in bytes. */ static unsigned amdgpu_vm_bo_size(struct amdgpu_device *adev, unsigned level) { @@ -221,6 +276,9 @@ void amdgpu_vm_get_pd_bo(struct amdgpu_vm *vm, * @param: parameter for the validation callback * * Validate the page table BOs on command submission if neccessary. + * + * Returns: + * Validation result. */ int amdgpu_vm_validate_pt_bos(struct amdgpu_device *adev, struct amdgpu_vm *vm, int (*validate)(void *p, struct amdgpu_bo *bo), @@ -276,6 +334,9 @@ int amdgpu_vm_validate_pt_bos(struct amdgpu_device *adev, struct amdgpu_vm *vm, * @vm: VM to check * * Check if all VM PDs/PTs are ready for updates + * + * Returns: + * True if eviction list is empty. */ bool amdgpu_vm_ready(struct amdgpu_vm *vm) { @@ -286,10 +347,15 @@ bool amdgpu_vm_ready(struct amdgpu_vm *vm) * amdgpu_vm_clear_bo - initially clear the PDs/PTs * * @adev: amdgpu_device pointer + * @vm: VM to clear BO from * @bo: BO to clear * @level: level this BO is at + * @pte_support_ats: indicate ATS support from PTE * * Root PD needs to be reserved when calling this. + * + * Returns: + * 0 on success, errno otherwise. */ static int amdgpu_vm_clear_bo(struct amdgpu_device *adev, struct amdgpu_vm *vm, struct amdgpu_bo *bo, @@ -321,7 +387,7 @@ static int amdgpu_vm_clear_bo(struct amdgpu_device *adev, ats_entries = 0; } - ring = container_of(vm->entity.sched, struct amdgpu_ring, sched); + ring = container_of(vm->entity.rq->sched, struct amdgpu_ring, sched); r = reservation_object_reserve_shared(bo->tbo.resv); if (r) @@ -359,8 +425,8 @@ static int amdgpu_vm_clear_bo(struct amdgpu_device *adev, if (r) goto error_free; - r = amdgpu_job_submit(job, ring, &vm->entity, - AMDGPU_FENCE_OWNER_UNDEFINED, &fence); + r = amdgpu_job_submit(job, &vm->entity, AMDGPU_FENCE_OWNER_UNDEFINED, + &fence); if (r) goto error_free; @@ -385,10 +451,16 @@ error: * * @adev: amdgpu_device pointer * @vm: requested vm + * @parent: parent PT * @saddr: start of the address range * @eaddr: end of the address range + * @level: VMPT level + * @ats: indicate ATS support from PTE * * Make sure the page directories and page tables are allocated + * + * Returns: + * 0 on success, errno otherwise. */ static int amdgpu_vm_alloc_levels(struct amdgpu_device *adev, struct amdgpu_vm *vm, @@ -423,11 +495,12 @@ static int amdgpu_vm_alloc_levels(struct amdgpu_device *adev, eaddr = eaddr & ((1 << shift) - 1); flags = AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS; + if (vm->root.base.bo->shadow) + flags |= AMDGPU_GEM_CREATE_SHADOW; if (vm->use_cpu_for_update) flags |= AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED; else - flags |= (AMDGPU_GEM_CREATE_NO_CPU_ACCESS | - AMDGPU_GEM_CREATE_SHADOW); + flags |= AMDGPU_GEM_CREATE_NO_CPU_ACCESS; /* walk over the address space and allocate the page tables */ for (pt_idx = from; pt_idx <= to; ++pt_idx) { @@ -496,6 +569,9 @@ static int amdgpu_vm_alloc_levels(struct amdgpu_device *adev, * @size: Size from start address we need. * * Make sure the page tables are allocated. + * + * Returns: + * 0 on success, errno otherwise. */ int amdgpu_vm_alloc_pts(struct amdgpu_device *adev, struct amdgpu_vm *vm, @@ -561,6 +637,15 @@ void amdgpu_vm_check_compute_bug(struct amdgpu_device *adev) } } +/** + * amdgpu_vm_need_pipeline_sync - Check if pipe sync is needed for job. + * + * @ring: ring on which the job will be submitted + * @job: job to submit + * + * Returns: + * True if sync is needed. + */ bool amdgpu_vm_need_pipeline_sync(struct amdgpu_ring *ring, struct amdgpu_job *job) { @@ -588,19 +673,17 @@ bool amdgpu_vm_need_pipeline_sync(struct amdgpu_ring *ring, return vm_flush_needed || gds_switch_needed; } -static bool amdgpu_vm_is_large_bar(struct amdgpu_device *adev) -{ - return (adev->gmc.real_vram_size == adev->gmc.visible_vram_size); -} - /** * amdgpu_vm_flush - hardware flush the vm * * @ring: ring to use for flush - * @vmid: vmid number to use - * @pd_addr: address of the page directory + * @job: related job + * @need_pipe_sync: is pipe sync needed * * Emit a VM flush when it is necessary. + * + * Returns: + * 0 on success, errno otherwise. */ int amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job, bool need_pipe_sync) { @@ -708,6 +791,9 @@ int amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job, bool need_ * Returns the found bo_va or NULL if none is found * * Object has to be reserved! + * + * Returns: + * Found bo_va or NULL. */ struct amdgpu_bo_va *amdgpu_vm_bo_find(struct amdgpu_vm *vm, struct amdgpu_bo *bo) @@ -789,7 +875,10 @@ static void amdgpu_vm_do_copy_ptes(struct amdgpu_pte_update_params *params, * @addr: the unmapped addr * * Look up the physical address of the page that the pte resolves - * to and return the pointer for the page table entry. + * to. + * + * Returns: + * The pointer for the page table entry. */ static uint64_t amdgpu_vm_map_gart(const dma_addr_t *pages_addr, uint64_t addr) { @@ -842,6 +931,17 @@ static void amdgpu_vm_cpu_set_ptes(struct amdgpu_pte_update_params *params, } } + +/** + * amdgpu_vm_wait_pd - Wait for PT BOs to be free. + * + * @adev: amdgpu_device pointer + * @vm: related vm + * @owner: fence owner + * + * Returns: + * 0 on success, errno otherwise. + */ static int amdgpu_vm_wait_pd(struct amdgpu_device *adev, struct amdgpu_vm *vm, void *owner) { @@ -895,7 +995,10 @@ static void amdgpu_vm_update_pde(struct amdgpu_pte_update_params *params, /* * amdgpu_vm_invalidate_level - mark all PD levels as invalid * + * @adev: amdgpu_device pointer + * @vm: related vm * @parent: parent PD + * @level: VMPT level * * Mark all PD level as invalid after an error. */ @@ -930,7 +1033,9 @@ static void amdgpu_vm_invalidate_level(struct amdgpu_device *adev, * @vm: requested vm * * Makes sure all directories are up to date. - * Returns 0 for success, error for failure. + * + * Returns: + * 0 for success, error for failure. */ int amdgpu_vm_update_directories(struct amdgpu_device *adev, struct amdgpu_vm *vm) @@ -980,7 +1085,7 @@ restart: struct amdgpu_vm_bo_base, vm_status); bo_base->moved = false; - list_move(&bo_base->vm_status, &vm->idle); + list_del_init(&bo_base->vm_status); bo = bo_base->bo->parent; if (!bo) @@ -1009,15 +1114,15 @@ restart: struct amdgpu_ring *ring; struct dma_fence *fence; - ring = container_of(vm->entity.sched, struct amdgpu_ring, + ring = container_of(vm->entity.rq->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); WARN_ON(params.ib->length_dw > ndw); - r = amdgpu_job_submit(job, ring, &vm->entity, - AMDGPU_FENCE_OWNER_VM, &fence); + r = amdgpu_job_submit(job, &vm->entity, AMDGPU_FENCE_OWNER_VM, + &fence); if (r) goto error; @@ -1117,14 +1222,15 @@ static void amdgpu_vm_handle_huge_pages(struct amdgpu_pte_update_params *p, * amdgpu_vm_update_ptes - make sure that page tables are valid * * @params: see amdgpu_pte_update_params definition - * @vm: requested vm * @start: start of GPU address range * @end: end of GPU address range * @dst: destination address to map to, the next dst inside the function * @flags: mapping flags * * Update the page tables in the range @start - @end. - * Returns 0 for success, -EINVAL for failure. + * + * Returns: + * 0 for success, -EINVAL for failure. */ static int amdgpu_vm_update_ptes(struct amdgpu_pte_update_params *params, uint64_t start, uint64_t end, @@ -1178,7 +1284,9 @@ static int amdgpu_vm_update_ptes(struct amdgpu_pte_update_params *params, * @end: last PTE to handle * @dst: addr those PTEs should point to * @flags: hw mapping flags - * Returns 0 for success, -EINVAL for failure. + * + * Returns: + * 0 for success, -EINVAL for failure. */ static int amdgpu_vm_frag_ptes(struct amdgpu_pte_update_params *params, uint64_t start, uint64_t end, @@ -1250,7 +1358,9 @@ static int amdgpu_vm_frag_ptes(struct amdgpu_pte_update_params *params, * @fence: optional resulting fence * * Fill in the page table entries between @start and @last. - * Returns 0 for success, -EINVAL for failure. + * + * Returns: + * 0 for success, -EINVAL for failure. */ static int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev, struct dma_fence *exclusive, @@ -1294,7 +1404,7 @@ static int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev, addr, flags); } - ring = container_of(vm->entity.sched, struct amdgpu_ring, sched); + ring = container_of(vm->entity.rq->sched, struct amdgpu_ring, sched); nptes = last - start + 1; @@ -1326,7 +1436,10 @@ static int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev, ndw += ncmds * 10; /* extra commands for begin/end fragments */ - ndw += 2 * 10 * adev->vm_manager.fragment_size; + if (vm->root.base.bo->shadow) + ndw += 2 * 10 * adev->vm_manager.fragment_size * 2; + else + ndw += 2 * 10 * adev->vm_manager.fragment_size; params.func = amdgpu_vm_do_set_ptes; } @@ -1373,8 +1486,7 @@ static int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev, amdgpu_ring_pad_ib(ring, params.ib); WARN_ON(params.ib->length_dw > ndw); - r = amdgpu_job_submit(job, ring, &vm->entity, - AMDGPU_FENCE_OWNER_VM, &f); + r = amdgpu_job_submit(job, &vm->entity, AMDGPU_FENCE_OWNER_VM, &f); if (r) goto error_free; @@ -1402,7 +1514,9 @@ error_free: * * Split the mapping into smaller chunks so that each update fits * into a SDMA IB. - * Returns 0 for success, -EINVAL for failure. + * + * Returns: + * 0 for success, -EINVAL for failure. */ static int amdgpu_vm_bo_split_mapping(struct amdgpu_device *adev, struct dma_fence *exclusive, @@ -1455,7 +1569,7 @@ static int amdgpu_vm_bo_split_mapping(struct amdgpu_device *adev, if (nodes) { addr = nodes->start << PAGE_SHIFT; max_entries = (nodes->size - pfn) * - (PAGE_SIZE / AMDGPU_GPU_PAGE_SIZE); + AMDGPU_GPU_PAGES_IN_CPU_PAGE; } else { addr = 0; max_entries = S64_MAX; @@ -1466,7 +1580,7 @@ static int amdgpu_vm_bo_split_mapping(struct amdgpu_device *adev, max_entries = min(max_entries, 16ull * 1024ull); for (count = 1; - count < max_entries / (PAGE_SIZE / AMDGPU_GPU_PAGE_SIZE); + count < max_entries / AMDGPU_GPU_PAGES_IN_CPU_PAGE; ++count) { uint64_t idx = pfn + count; @@ -1480,7 +1594,7 @@ static int amdgpu_vm_bo_split_mapping(struct amdgpu_device *adev, dma_addr = pages_addr; } else { addr = pages_addr[pfn]; - max_entries = count * (PAGE_SIZE / AMDGPU_GPU_PAGE_SIZE); + max_entries = count * AMDGPU_GPU_PAGES_IN_CPU_PAGE; } } else if (flags & AMDGPU_PTE_VALID) { @@ -1495,7 +1609,7 @@ static int amdgpu_vm_bo_split_mapping(struct amdgpu_device *adev, if (r) return r; - pfn += (last - start + 1) / (PAGE_SIZE / AMDGPU_GPU_PAGE_SIZE); + pfn += (last - start + 1) / AMDGPU_GPU_PAGES_IN_CPU_PAGE; if (nodes && nodes->size == pfn) { pfn = 0; ++nodes; @@ -1515,7 +1629,9 @@ static int amdgpu_vm_bo_split_mapping(struct amdgpu_device *adev, * @clear: if true clear the entries * * Fill in the page table entries for @bo_va. - * Returns 0 for success, -EINVAL for failure. + * + * Returns: + * 0 for success, -EINVAL for failure. */ int amdgpu_vm_bo_update(struct amdgpu_device *adev, struct amdgpu_bo_va *bo_va, @@ -1531,18 +1647,17 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev, uint64_t flags; int r; - if (clear || !bo_va->base.bo) { + if (clear || !bo) { mem = NULL; nodes = NULL; exclusive = NULL; } else { struct ttm_dma_tt *ttm; - mem = &bo_va->base.bo->tbo.mem; + mem = &bo->tbo.mem; nodes = mem->mm_node; if (mem->mem_type == TTM_PL_TT) { - ttm = container_of(bo_va->base.bo->tbo.ttm, - struct ttm_dma_tt, ttm); + ttm = container_of(bo->tbo.ttm, struct ttm_dma_tt, ttm); pages_addr = ttm->dma_address; } exclusive = reservation_object_get_excl(bo->tbo.resv); @@ -1610,6 +1725,8 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev, /** * amdgpu_vm_update_prt_state - update the global PRT state + * + * @adev: amdgpu_device pointer */ static void amdgpu_vm_update_prt_state(struct amdgpu_device *adev) { @@ -1624,6 +1741,8 @@ static void amdgpu_vm_update_prt_state(struct amdgpu_device *adev) /** * amdgpu_vm_prt_get - add a PRT user + * + * @adev: amdgpu_device pointer */ static void amdgpu_vm_prt_get(struct amdgpu_device *adev) { @@ -1636,6 +1755,8 @@ static void amdgpu_vm_prt_get(struct amdgpu_device *adev) /** * amdgpu_vm_prt_put - drop a PRT user + * + * @adev: amdgpu_device pointer */ static void amdgpu_vm_prt_put(struct amdgpu_device *adev) { @@ -1645,6 +1766,9 @@ static void amdgpu_vm_prt_put(struct amdgpu_device *adev) /** * amdgpu_vm_prt_cb - callback for updating the PRT status + * + * @fence: fence for the callback + * @_cb: the callback function */ static void amdgpu_vm_prt_cb(struct dma_fence *fence, struct dma_fence_cb *_cb) { @@ -1656,6 +1780,9 @@ static void amdgpu_vm_prt_cb(struct dma_fence *fence, struct dma_fence_cb *_cb) /** * amdgpu_vm_add_prt_cb - add callback for updating the PRT status + * + * @adev: amdgpu_device pointer + * @fence: fence for the callback */ static void amdgpu_vm_add_prt_cb(struct amdgpu_device *adev, struct dma_fence *fence) @@ -1747,9 +1874,11 @@ static void amdgpu_vm_prt_fini(struct amdgpu_device *adev, struct amdgpu_vm *vm) * or if an error occurred) * * Make sure all freed BOs are cleared in the PT. - * Returns 0 for success. - * * PTs have to be reserved and mutex must be locked! + * + * Returns: + * 0 for success. + * */ int amdgpu_vm_clear_freed(struct amdgpu_device *adev, struct amdgpu_vm *vm, @@ -1794,10 +1923,11 @@ int amdgpu_vm_clear_freed(struct amdgpu_device *adev, * * @adev: amdgpu_device pointer * @vm: requested vm - * @sync: sync object to add fences to * * Make sure all BOs which are moved are updated in the PTs. - * Returns 0 for success. + * + * Returns: + * 0 for success. * * PTs have to be reserved! */ @@ -1852,7 +1982,9 @@ int amdgpu_vm_handle_moved(struct amdgpu_device *adev, * * Add @bo into the requested vm. * Add @bo to the list of bos associated with the vm - * Returns newly added bo_va or NULL for failure + * + * Returns: + * Newly added bo_va or NULL for failure * * Object has to be reserved! */ @@ -1915,10 +2047,13 @@ static void amdgpu_vm_bo_insert_map(struct amdgpu_device *adev, * @bo_va: bo_va to store the address * @saddr: where to map the BO * @offset: requested offset in the BO + * @size: BO size in bytes * @flags: attributes of pages (read/write/valid/etc.) * * Add a mapping of the BO at the specefied addr into the VM. - * Returns 0 for success, error for failure. + * + * Returns: + * 0 for success, error for failure. * * Object has to be reserved and unreserved outside! */ @@ -1976,11 +2111,14 @@ int amdgpu_vm_bo_map(struct amdgpu_device *adev, * @bo_va: bo_va to store the address * @saddr: where to map the BO * @offset: requested offset in the BO + * @size: BO size in bytes * @flags: attributes of pages (read/write/valid/etc.) * * Add a mapping of the BO at the specefied addr into the VM. Replace existing * mappings as we do so. - * Returns 0 for success, error for failure. + * + * Returns: + * 0 for success, error for failure. * * Object has to be reserved and unreserved outside! */ @@ -2037,7 +2175,9 @@ int amdgpu_vm_bo_replace_map(struct amdgpu_device *adev, * @saddr: where to the BO is mapped * * Remove a mapping of the BO at the specefied addr from the VM. - * Returns 0 for success, error for failure. + * + * Returns: + * 0 for success, error for failure. * * Object has to be reserved and unreserved outside! */ @@ -2091,7 +2231,9 @@ int amdgpu_vm_bo_unmap(struct amdgpu_device *adev, * @size: size of the range * * Remove all mappings in a range, split them as appropriate. - * Returns 0 for success, error for failure. + * + * Returns: + * 0 for success, error for failure. */ int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev, struct amdgpu_vm *vm, @@ -2188,8 +2330,13 @@ int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev, * amdgpu_vm_bo_lookup_mapping - find mapping by address * * @vm: the requested VM + * @addr: the address * * Find a mapping by it's address. + * + * Returns: + * The amdgpu_bo_va_mapping matching for addr or NULL + * */ struct amdgpu_bo_va_mapping *amdgpu_vm_bo_lookup_mapping(struct amdgpu_vm *vm, uint64_t addr) @@ -2198,6 +2345,35 @@ struct amdgpu_bo_va_mapping *amdgpu_vm_bo_lookup_mapping(struct amdgpu_vm *vm, } /** + * amdgpu_vm_bo_trace_cs - trace all reserved mappings + * + * @vm: the requested vm + * @ticket: CS ticket + * + * Trace all mappings of BOs reserved during a command submission. + */ +void amdgpu_vm_bo_trace_cs(struct amdgpu_vm *vm, struct ww_acquire_ctx *ticket) +{ + struct amdgpu_bo_va_mapping *mapping; + + if (!trace_amdgpu_vm_bo_cs_enabled()) + return; + + for (mapping = amdgpu_vm_it_iter_first(&vm->va, 0, U64_MAX); mapping; + mapping = amdgpu_vm_it_iter_next(mapping, 0, U64_MAX)) { + if (mapping->bo_va && mapping->bo_va->base.bo) { + struct amdgpu_bo *bo; + + bo = mapping->bo_va->base.bo; + if (READ_ONCE(bo->tbo.resv->lock.ctx) != ticket) + continue; + } + + trace_amdgpu_vm_bo_cs(mapping); + } +} + +/** * amdgpu_vm_bo_rmv - remove a bo to a specific vm * * @adev: amdgpu_device pointer @@ -2241,8 +2417,8 @@ void amdgpu_vm_bo_rmv(struct amdgpu_device *adev, * amdgpu_vm_bo_invalidate - mark the bo as invalid * * @adev: amdgpu_device pointer - * @vm: requested vm * @bo: amdgpu buffer object + * @evicted: is the BO evicted * * Mark @bo as invalid. */ @@ -2282,6 +2458,14 @@ void amdgpu_vm_bo_invalidate(struct amdgpu_device *adev, } } +/** + * amdgpu_vm_get_block_size - calculate VM page table size as power of two + * + * @vm_size: VM size + * + * Returns: + * VM page table as power of two + */ static uint32_t amdgpu_vm_get_block_size(uint64_t vm_size) { /* Total bits covered by PD + PTs */ @@ -2300,6 +2484,10 @@ static uint32_t amdgpu_vm_get_block_size(uint64_t vm_size) * * @adev: amdgpu_device pointer * @vm_size: the default vm size if it's set auto + * @fragment_size_default: Default PTE fragment size + * @max_level: max VMPT level + * @max_bits: max address space size in bits + * */ void amdgpu_vm_adjust_size(struct amdgpu_device *adev, uint32_t vm_size, uint32_t fragment_size_default, unsigned max_level, @@ -2367,8 +2555,12 @@ void amdgpu_vm_adjust_size(struct amdgpu_device *adev, uint32_t vm_size, * @adev: amdgpu_device pointer * @vm: requested vm * @vm_context: Indicates if it GFX or Compute context + * @pasid: Process address space identifier * * Init @vm fields. + * + * Returns: + * 0 for success, error for failure. */ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm, int vm_context, unsigned int pasid) @@ -2400,8 +2592,7 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm, ring_instance %= adev->vm_manager.vm_pte_num_rings; ring = adev->vm_manager.vm_pte_rings[ring_instance]; rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_KERNEL]; - r = drm_sched_entity_init(&ring->sched, &vm->entity, - rq, NULL); + r = drm_sched_entity_init(&vm->entity, &rq, 1, NULL); if (r) return r; @@ -2419,14 +2610,14 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm, } DRM_DEBUG_DRIVER("VM update mode is %s\n", vm->use_cpu_for_update ? "CPU" : "SDMA"); - WARN_ONCE((vm->use_cpu_for_update & !amdgpu_vm_is_large_bar(adev)), + WARN_ONCE((vm->use_cpu_for_update & !amdgpu_gmc_vram_full_visible(&adev->gmc)), "CPU update of VM recommended only for large BAR system\n"); vm->last_update = NULL; flags = AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS; if (vm->use_cpu_for_update) flags |= AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED; - else + else if (vm_context != AMDGPU_VM_CONTEXT_COMPUTE) flags |= AMDGPU_GEM_CREATE_SHADOW; size = amdgpu_vm_bo_size(adev, adev->vm_manager.root_level); @@ -2481,7 +2672,7 @@ error_free_root: vm->root.base.bo = NULL; error_free_sched_entity: - drm_sched_entity_fini(&ring->sched, &vm->entity); + drm_sched_entity_destroy(&vm->entity); return r; } @@ -2489,6 +2680,9 @@ error_free_sched_entity: /** * amdgpu_vm_make_compute - Turn a GFX VM into a compute VM * + * @adev: amdgpu_device pointer + * @vm: requested vm + * * This only works on GFX VMs that don't have any BOs added and no * page tables allocated yet. * @@ -2498,10 +2692,10 @@ error_free_sched_entity: * - pasid (old PASID is released, because compute manages its own PASIDs) * * Reinitializes the page directory to reflect the changed ATS - * setting. May leave behind an unused shadow BO for the page - * directory when switching from SDMA updates to CPU updates. + * setting. * - * Returns 0 for success, -errno for errors. + * Returns: + * 0 for success, -errno for errors. */ int amdgpu_vm_make_compute(struct amdgpu_device *adev, struct amdgpu_vm *vm) { @@ -2535,7 +2729,7 @@ int amdgpu_vm_make_compute(struct amdgpu_device *adev, struct amdgpu_vm *vm) vm->pte_support_ats = pte_support_ats; DRM_DEBUG_DRIVER("VM update mode is %s\n", vm->use_cpu_for_update ? "CPU" : "SDMA"); - WARN_ONCE((vm->use_cpu_for_update & !amdgpu_vm_is_large_bar(adev)), + WARN_ONCE((vm->use_cpu_for_update & !amdgpu_gmc_vram_full_visible(&adev->gmc)), "CPU update of VM recommended only for large BAR system\n"); if (vm->pasid) { @@ -2548,6 +2742,9 @@ int amdgpu_vm_make_compute(struct amdgpu_device *adev, struct amdgpu_vm *vm) vm->pasid = 0; } + /* Free the shadow bo for compute VM */ + amdgpu_bo_unref(&vm->root.base.bo->shadow); + error: amdgpu_bo_unreserve(vm->root.base.bo); return r; @@ -2614,7 +2811,7 @@ void amdgpu_vm_fini(struct amdgpu_device *adev, struct amdgpu_vm *vm) spin_unlock_irqrestore(&adev->vm_manager.pasid_lock, flags); } - drm_sched_entity_fini(vm->entity.sched, &vm->entity); + drm_sched_entity_destroy(&vm->entity); if (!RB_EMPTY_ROOT(&vm->va.rb_root)) { dev_err(adev->dev, "still active bo inside vm\n"); @@ -2656,8 +2853,10 @@ void amdgpu_vm_fini(struct amdgpu_device *adev, struct amdgpu_vm *vm) * @adev: amdgpu_device pointer * @pasid: PASID do identify the VM * - * This function is expected to be called in interrupt context. Returns - * true if there was fault credit, false otherwise + * This function is expected to be called in interrupt context. + * + * Returns: + * True if there was fault credit, false otherwise */ bool amdgpu_vm_pasid_fault_credit(struct amdgpu_device *adev, unsigned int pasid) @@ -2711,7 +2910,7 @@ void amdgpu_vm_manager_init(struct amdgpu_device *adev) */ #ifdef CONFIG_X86_64 if (amdgpu_vm_update_mode == -1) { - if (amdgpu_vm_is_large_bar(adev)) + if (amdgpu_gmc_vram_full_visible(&adev->gmc)) adev->vm_manager.vm_update_mode = AMDGPU_VM_USE_CPU_FOR_COMPUTE; else @@ -2741,6 +2940,16 @@ void amdgpu_vm_manager_fini(struct amdgpu_device *adev) amdgpu_vmid_mgr_fini(adev); } +/** + * amdgpu_vm_ioctl - Manages VMID reservation for vm hubs. + * + * @dev: drm device pointer + * @data: drm_amdgpu_vm + * @filp: drm file pointer + * + * Returns: + * 0 for success, -errno for errors. + */ int amdgpu_vm_ioctl(struct drm_device *dev, void *data, struct drm_file *filp) { union drm_amdgpu_vm *args = data; @@ -2764,3 +2973,42 @@ int amdgpu_vm_ioctl(struct drm_device *dev, void *data, struct drm_file *filp) return 0; } + +/** + * amdgpu_vm_get_task_info - Extracts task info for a PASID. + * + * @dev: drm device pointer + * @pasid: PASID identifier for VM + * @task_info: task_info to fill. + */ +void amdgpu_vm_get_task_info(struct amdgpu_device *adev, unsigned int pasid, + struct amdgpu_task_info *task_info) +{ + struct amdgpu_vm *vm; + + spin_lock(&adev->vm_manager.pasid_lock); + + vm = idr_find(&adev->vm_manager.pasid_idr, pasid); + if (vm) + *task_info = vm->task_info; + + spin_unlock(&adev->vm_manager.pasid_lock); +} + +/** + * amdgpu_vm_set_task_info - Sets VMs task info. + * + * @vm: vm for which to set the info + */ +void amdgpu_vm_set_task_info(struct amdgpu_vm *vm) +{ + if (!vm->task_info.pid) { + vm->task_info.pid = current->pid; + get_task_comm(vm->task_info.task_name, current); + + if (current->group_leader->mm == current->mm) { + vm->task_info.tgid = current->group_leader->pid; + get_task_comm(vm->task_info.process_name, current->group_leader); + } + } +} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h index 061b99a18cb8..67a15d439ac0 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h @@ -164,6 +164,14 @@ struct amdgpu_vm_pt { #define AMDGPU_VM_FAULT_PASID(fault) ((u64)(fault) >> 48) #define AMDGPU_VM_FAULT_ADDR(fault) ((u64)(fault) & 0xfffffffff000ULL) + +struct amdgpu_task_info { + char process_name[TASK_COMM_LEN]; + char task_name[TASK_COMM_LEN]; + pid_t pid; + pid_t tgid; +}; + struct amdgpu_vm { /* tree of virtual addresses mapped */ struct rb_root_cached va; @@ -215,6 +223,9 @@ struct amdgpu_vm { /* Valid while the PD is reserved or fenced */ uint64_t pd_phys_addr; + + /* Some basic info about the task */ + struct amdgpu_task_info task_info; }; struct amdgpu_vm_manager { @@ -307,6 +318,7 @@ int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev, uint64_t saddr, uint64_t size); struct amdgpu_bo_va_mapping *amdgpu_vm_bo_lookup_mapping(struct amdgpu_vm *vm, uint64_t addr); +void amdgpu_vm_bo_trace_cs(struct amdgpu_vm *vm, struct ww_acquire_ctx *ticket); void amdgpu_vm_bo_rmv(struct amdgpu_device *adev, struct amdgpu_bo_va *bo_va); void amdgpu_vm_adjust_size(struct amdgpu_device *adev, uint32_t vm_size, @@ -317,4 +329,9 @@ bool amdgpu_vm_need_pipeline_sync(struct amdgpu_ring *ring, struct amdgpu_job *job); void amdgpu_vm_check_compute_bug(struct amdgpu_device *adev); +void amdgpu_vm_get_task_info(struct amdgpu_device *adev, unsigned int pasid, + struct amdgpu_task_info *task_info); + +void amdgpu_vm_set_task_info(struct amdgpu_vm *vm); + #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c index b6333f92ba45..9cfa8a9ada92 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c @@ -97,33 +97,29 @@ static u64 amdgpu_vram_mgr_vis_size(struct amdgpu_device *adev, } /** - * amdgpu_vram_mgr_bo_invisible_size - CPU invisible BO size + * amdgpu_vram_mgr_bo_visible_size - CPU visible BO size * * @bo: &amdgpu_bo buffer object (must be in VRAM) * * Returns: - * How much of the given &amdgpu_bo buffer object lies in CPU invisible VRAM. + * How much of the given &amdgpu_bo buffer object lies in CPU visible VRAM. */ -u64 amdgpu_vram_mgr_bo_invisible_size(struct amdgpu_bo *bo) +u64 amdgpu_vram_mgr_bo_visible_size(struct amdgpu_bo *bo) { struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev); struct ttm_mem_reg *mem = &bo->tbo.mem; struct drm_mm_node *nodes = mem->mm_node; unsigned pages = mem->num_pages; - u64 usage = 0; + u64 usage; - if (adev->gmc.visible_vram_size == adev->gmc.real_vram_size) - return 0; + if (amdgpu_gmc_vram_full_visible(&adev->gmc)) + return amdgpu_bo_size(bo); if (mem->start >= adev->gmc.visible_vram_size >> PAGE_SHIFT) - return amdgpu_bo_size(bo); + return 0; - while (nodes && pages) { - usage += nodes->size << PAGE_SHIFT; - usage -= amdgpu_vram_mgr_vis_size(adev, nodes); - pages -= nodes->size; - ++nodes; - } + for (usage = 0; nodes && pages; pages -= nodes->size, nodes++) + usage += amdgpu_vram_mgr_vis_size(adev, nodes); return usage; } diff --git a/drivers/gpu/drm/amd/amdgpu/ci_dpm.c b/drivers/gpu/drm/amd/amdgpu/ci_dpm.c index 7fbad2f5f0bd..d2469453dca2 100644 --- a/drivers/gpu/drm/amd/amdgpu/ci_dpm.c +++ b/drivers/gpu/drm/amd/amdgpu/ci_dpm.c @@ -49,10 +49,10 @@ #include "gmc/gmc_7_1_d.h" #include "gmc/gmc_7_1_sh_mask.h" -MODULE_FIRMWARE("radeon/bonaire_smc.bin"); -MODULE_FIRMWARE("radeon/bonaire_k_smc.bin"); -MODULE_FIRMWARE("radeon/hawaii_smc.bin"); -MODULE_FIRMWARE("radeon/hawaii_k_smc.bin"); +MODULE_FIRMWARE("amdgpu/bonaire_smc.bin"); +MODULE_FIRMWARE("amdgpu/bonaire_k_smc.bin"); +MODULE_FIRMWARE("amdgpu/hawaii_smc.bin"); +MODULE_FIRMWARE("amdgpu/hawaii_k_smc.bin"); #define MC_CG_ARB_FREQ_F0 0x0a #define MC_CG_ARB_FREQ_F1 0x0b @@ -951,12 +951,12 @@ static void ci_apply_state_adjust_rules(struct amdgpu_device *adev, else pi->battery_state = false; - if (adev->pm.dpm.ac_power) + if (adev->pm.ac_power) max_limits = &adev->pm.dpm.dyn_state.max_clock_voltage_on_ac; else max_limits = &adev->pm.dpm.dyn_state.max_clock_voltage_on_dc; - if (adev->pm.dpm.ac_power == false) { + if (adev->pm.ac_power == false) { for (i = 0; i < ps->performance_level_count; i++) { if (ps->performance_levels[i].mclk > max_limits->mclk) ps->performance_levels[i].mclk = max_limits->mclk; @@ -4078,7 +4078,7 @@ static int ci_enable_uvd_dpm(struct amdgpu_device *adev, bool enable) const struct amdgpu_clock_and_voltage_limits *max_limits; int i; - if (adev->pm.dpm.ac_power) + if (adev->pm.ac_power) max_limits = &adev->pm.dpm.dyn_state.max_clock_voltage_on_ac; else max_limits = &adev->pm.dpm.dyn_state.max_clock_voltage_on_dc; @@ -4127,7 +4127,7 @@ static int ci_enable_vce_dpm(struct amdgpu_device *adev, bool enable) const struct amdgpu_clock_and_voltage_limits *max_limits; int i; - if (adev->pm.dpm.ac_power) + if (adev->pm.ac_power) max_limits = &adev->pm.dpm.dyn_state.max_clock_voltage_on_ac; else max_limits = &adev->pm.dpm.dyn_state.max_clock_voltage_on_dc; @@ -4160,7 +4160,7 @@ static int ci_enable_samu_dpm(struct amdgpu_device *adev, bool enable) const struct amdgpu_clock_and_voltage_limits *max_limits; int i; - if (adev->pm.dpm.ac_power) + if (adev->pm.ac_power) max_limits = &adev->pm.dpm.dyn_state.max_clock_voltage_on_ac; else max_limits = &adev->pm.dpm.dyn_state.max_clock_voltage_on_dc; @@ -4191,7 +4191,7 @@ static int ci_enable_acp_dpm(struct amdgpu_device *adev, bool enable) const struct amdgpu_clock_and_voltage_limits *max_limits; int i; - if (adev->pm.dpm.ac_power) + if (adev->pm.ac_power) max_limits = &adev->pm.dpm.dyn_state.max_clock_voltage_on_ac; else max_limits = &adev->pm.dpm.dyn_state.max_clock_voltage_on_dc; @@ -5815,7 +5815,7 @@ static int ci_dpm_init_microcode(struct amdgpu_device *adev) default: BUG(); } - snprintf(fw_name, sizeof(fw_name), "radeon/%s_smc.bin", chip_name); + snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_smc.bin", chip_name); err = request_firmware(&adev->pm.fw, fw_name, adev->dev); if (err) goto out; @@ -5846,8 +5846,7 @@ static int ci_dpm_init(struct amdgpu_device *adev) adev->pm.dpm.priv = pi; pi->sys_pcie_mask = - (adev->pm.pcie_gen_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_MASK) >> - CAIL_PCIE_LINK_SPEED_SUPPORT_SHIFT; + adev->pm.pcie_gen_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_MASK; pi->force_pcie_gen = AMDGPU_PCIE_GEN_INVALID; @@ -6767,6 +6766,19 @@ static int ci_dpm_read_sensor(void *handle, int idx, } } +static int ci_set_powergating_by_smu(void *handle, + uint32_t block_type, bool gate) +{ + switch (block_type) { + case AMD_IP_BLOCK_TYPE_UVD: + ci_dpm_powergate_uvd(handle, gate); + break; + default: + break; + } + return 0; +} + static const struct amd_ip_funcs ci_dpm_ip_funcs = { .name = "ci_dpm", .early_init = ci_dpm_early_init, @@ -6804,7 +6816,7 @@ static const struct amd_pm_funcs ci_dpm_funcs = { .debugfs_print_current_performance_level = &ci_dpm_debugfs_print_current_performance_level, .force_performance_level = &ci_dpm_force_performance_level, .vblank_too_short = &ci_dpm_vblank_too_short, - .powergate_uvd = &ci_dpm_powergate_uvd, + .set_powergating_by_smu = &ci_set_powergating_by_smu, .set_fan_control_mode = &ci_dpm_set_fan_control_mode, .get_fan_control_mode = &ci_dpm_get_fan_control_mode, .set_fan_speed_percent = &ci_dpm_set_fan_speed_percent, diff --git a/drivers/gpu/drm/amd/amdgpu/cik.c b/drivers/gpu/drm/amd/amdgpu/cik.c index 8ff4c60d1b59..78ab939ae5d8 100644 --- a/drivers/gpu/drm/amd/amdgpu/cik.c +++ b/drivers/gpu/drm/amd/amdgpu/cik.c @@ -1476,7 +1476,7 @@ static void cik_pcie_gen3_enable(struct amdgpu_device *adev) tmp |= PCIE_LC_CNTL4__LC_REDO_EQ_MASK; WREG32_PCIE(ixPCIE_LC_CNTL4, tmp); - mdelay(100); + msleep(100); /* linkctl */ pci_read_config_word(root, bridge_pos + PCI_EXP_LNKCTL, &tmp16); @@ -2003,9 +2003,9 @@ int cik_set_ip_blocks(struct amdgpu_device *adev) amdgpu_device_ip_block_add(adev, &gmc_v7_0_ip_block); amdgpu_device_ip_block_add(adev, &cik_ih_ip_block); if (amdgpu_dpm == -1) - amdgpu_device_ip_block_add(adev, &ci_smu_ip_block); - else amdgpu_device_ip_block_add(adev, &pp_smu_ip_block); + else + amdgpu_device_ip_block_add(adev, &ci_smu_ip_block); if (adev->enable_virtual_display) amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block); #if defined(CONFIG_DRM_AMD_DC) @@ -2024,9 +2024,9 @@ int cik_set_ip_blocks(struct amdgpu_device *adev) amdgpu_device_ip_block_add(adev, &gmc_v7_0_ip_block); amdgpu_device_ip_block_add(adev, &cik_ih_ip_block); if (amdgpu_dpm == -1) - amdgpu_device_ip_block_add(adev, &ci_smu_ip_block); - else amdgpu_device_ip_block_add(adev, &pp_smu_ip_block); + else + amdgpu_device_ip_block_add(adev, &ci_smu_ip_block); if (adev->enable_virtual_display) amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block); #if defined(CONFIG_DRM_AMD_DC) diff --git a/drivers/gpu/drm/amd/amdgpu/cik_sdma.c b/drivers/gpu/drm/amd/amdgpu/cik_sdma.c index a7576255cc30..d0fa2aac2388 100644 --- a/drivers/gpu/drm/amd/amdgpu/cik_sdma.c +++ b/drivers/gpu/drm/amd/amdgpu/cik_sdma.c @@ -54,16 +54,16 @@ static void cik_sdma_set_buffer_funcs(struct amdgpu_device *adev); static void cik_sdma_set_vm_pte_funcs(struct amdgpu_device *adev); static int cik_sdma_soft_reset(void *handle); -MODULE_FIRMWARE("radeon/bonaire_sdma.bin"); -MODULE_FIRMWARE("radeon/bonaire_sdma1.bin"); -MODULE_FIRMWARE("radeon/hawaii_sdma.bin"); -MODULE_FIRMWARE("radeon/hawaii_sdma1.bin"); -MODULE_FIRMWARE("radeon/kaveri_sdma.bin"); -MODULE_FIRMWARE("radeon/kaveri_sdma1.bin"); -MODULE_FIRMWARE("radeon/kabini_sdma.bin"); -MODULE_FIRMWARE("radeon/kabini_sdma1.bin"); -MODULE_FIRMWARE("radeon/mullins_sdma.bin"); -MODULE_FIRMWARE("radeon/mullins_sdma1.bin"); +MODULE_FIRMWARE("amdgpu/bonaire_sdma.bin"); +MODULE_FIRMWARE("amdgpu/bonaire_sdma1.bin"); +MODULE_FIRMWARE("amdgpu/hawaii_sdma.bin"); +MODULE_FIRMWARE("amdgpu/hawaii_sdma1.bin"); +MODULE_FIRMWARE("amdgpu/kaveri_sdma.bin"); +MODULE_FIRMWARE("amdgpu/kaveri_sdma1.bin"); +MODULE_FIRMWARE("amdgpu/kabini_sdma.bin"); +MODULE_FIRMWARE("amdgpu/kabini_sdma1.bin"); +MODULE_FIRMWARE("amdgpu/mullins_sdma.bin"); +MODULE_FIRMWARE("amdgpu/mullins_sdma1.bin"); u32 amdgpu_cik_gpu_check_soft_reset(struct amdgpu_device *adev); @@ -132,9 +132,9 @@ static int cik_sdma_init_microcode(struct amdgpu_device *adev) for (i = 0; i < adev->sdma.num_instances; i++) { if (i == 0) - snprintf(fw_name, sizeof(fw_name), "radeon/%s_sdma.bin", chip_name); + snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_sdma.bin", chip_name); else - snprintf(fw_name, sizeof(fw_name), "radeon/%s_sdma1.bin", chip_name); + snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_sdma1.bin", chip_name); err = request_firmware(&adev->sdma.instance[i].fw, fw_name, adev->dev); if (err) goto out; @@ -177,9 +177,8 @@ static uint64_t cik_sdma_ring_get_rptr(struct amdgpu_ring *ring) static uint64_t cik_sdma_ring_get_wptr(struct amdgpu_ring *ring) { struct amdgpu_device *adev = ring->adev; - u32 me = (ring == &adev->sdma.instance[0].ring) ? 0 : 1; - return (RREG32(mmSDMA0_GFX_RB_WPTR + sdma_offsets[me]) & 0x3fffc) >> 2; + return (RREG32(mmSDMA0_GFX_RB_WPTR + sdma_offsets[ring->me]) & 0x3fffc) >> 2; } /** @@ -192,9 +191,8 @@ static uint64_t cik_sdma_ring_get_wptr(struct amdgpu_ring *ring) static void cik_sdma_ring_set_wptr(struct amdgpu_ring *ring) { struct amdgpu_device *adev = ring->adev; - u32 me = (ring == &adev->sdma.instance[0].ring) ? 0 : 1; - WREG32(mmSDMA0_GFX_RB_WPTR + sdma_offsets[me], + WREG32(mmSDMA0_GFX_RB_WPTR + sdma_offsets[ring->me], (lower_32_bits(ring->wptr) << 2) & 0x3fffc); } @@ -248,7 +246,7 @@ static void cik_sdma_ring_emit_hdp_flush(struct amdgpu_ring *ring) SDMA_POLL_REG_MEM_EXTRA_FUNC(3)); /* == */ u32 ref_and_mask; - if (ring == &ring->adev->sdma.instance[0].ring) + if (ring->me == 0) ref_and_mask = GPU_HDP_FLUSH_DONE__SDMA0_MASK; else ref_and_mask = GPU_HDP_FLUSH_DONE__SDMA1_MASK; @@ -1290,8 +1288,10 @@ static void cik_sdma_set_ring_funcs(struct amdgpu_device *adev) { int i; - for (i = 0; i < adev->sdma.num_instances; i++) + for (i = 0; i < adev->sdma.num_instances; i++) { adev->sdma.instance[i].ring.funcs = &cik_sdma_ring_funcs; + adev->sdma.instance[i].ring.me = i; + } } static const struct amdgpu_irq_src_funcs cik_sdma_trap_irq_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c index ada241bfeee9..308f9f238bc1 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c @@ -41,6 +41,8 @@ #include "gmc/gmc_8_1_d.h" #include "gmc/gmc_8_1_sh_mask.h" +#include "ivsrcid/ivsrcid_vislands30.h" + static void dce_v10_0_set_display_funcs(struct amdgpu_device *adev); static void dce_v10_0_set_irq_funcs(struct amdgpu_device *adev); @@ -1855,15 +1857,14 @@ static int dce_v10_0_crtc_do_set_base(struct drm_crtc *crtc, if (unlikely(r != 0)) return r; - if (atomic) { - fb_location = amdgpu_bo_gpu_offset(abo); - } else { - r = amdgpu_bo_pin(abo, AMDGPU_GEM_DOMAIN_VRAM, &fb_location); + if (!atomic) { + r = amdgpu_bo_pin(abo, AMDGPU_GEM_DOMAIN_VRAM); if (unlikely(r != 0)) { amdgpu_bo_unreserve(abo); return -EINVAL; } } + fb_location = amdgpu_bo_gpu_offset(abo); amdgpu_bo_get_tiling_flags(abo, &tiling_flags); amdgpu_bo_unreserve(abo); @@ -2370,13 +2371,14 @@ static int dce_v10_0_crtc_cursor_set2(struct drm_crtc *crtc, return ret; } - ret = amdgpu_bo_pin(aobj, AMDGPU_GEM_DOMAIN_VRAM, &amdgpu_crtc->cursor_addr); + ret = amdgpu_bo_pin(aobj, AMDGPU_GEM_DOMAIN_VRAM); amdgpu_bo_unreserve(aobj); if (ret) { DRM_ERROR("Failed to pin new cursor BO (%d)\n", ret); drm_gem_object_put_unlocked(obj); return ret; } + amdgpu_crtc->cursor_addr = amdgpu_bo_gpu_offset(aobj); dce_v10_0_lock_cursor(crtc, true); @@ -2737,14 +2739,14 @@ static int dce_v10_0_sw_init(void *handle) return r; } - for (i = 8; i < 20; i += 2) { + for (i = VISLANDS30_IV_SRCID_D1_GRPH_PFLIP; i < 20; i += 2) { r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, i, &adev->pageflip_irq); if (r) return r; } /* HPD hotplug */ - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 42, &adev->hpd_irq); + r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_HOTPLUG_DETECT_A, &adev->hpd_irq); if (r) return r; diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c index a5b96eac3033..76dfb76f7900 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c @@ -41,6 +41,8 @@ #include "gmc/gmc_8_1_d.h" #include "gmc/gmc_8_1_sh_mask.h" +#include "ivsrcid/ivsrcid_vislands30.h" + static void dce_v11_0_set_display_funcs(struct amdgpu_device *adev); static void dce_v11_0_set_irq_funcs(struct amdgpu_device *adev); @@ -1897,15 +1899,14 @@ static int dce_v11_0_crtc_do_set_base(struct drm_crtc *crtc, if (unlikely(r != 0)) return r; - if (atomic) { - fb_location = amdgpu_bo_gpu_offset(abo); - } else { - r = amdgpu_bo_pin(abo, AMDGPU_GEM_DOMAIN_VRAM, &fb_location); + if (!atomic) { + r = amdgpu_bo_pin(abo, AMDGPU_GEM_DOMAIN_VRAM); if (unlikely(r != 0)) { amdgpu_bo_unreserve(abo); return -EINVAL; } } + fb_location = amdgpu_bo_gpu_offset(abo); amdgpu_bo_get_tiling_flags(abo, &tiling_flags); amdgpu_bo_unreserve(abo); @@ -2449,13 +2450,14 @@ static int dce_v11_0_crtc_cursor_set2(struct drm_crtc *crtc, return ret; } - ret = amdgpu_bo_pin(aobj, AMDGPU_GEM_DOMAIN_VRAM, &amdgpu_crtc->cursor_addr); + ret = amdgpu_bo_pin(aobj, AMDGPU_GEM_DOMAIN_VRAM); amdgpu_bo_unreserve(aobj); if (ret) { DRM_ERROR("Failed to pin new cursor BO (%d)\n", ret); drm_gem_object_put_unlocked(obj); return ret; } + amdgpu_crtc->cursor_addr = amdgpu_bo_gpu_offset(aobj); dce_v11_0_lock_cursor(crtc, true); @@ -2858,14 +2860,14 @@ static int dce_v11_0_sw_init(void *handle) return r; } - for (i = 8; i < 20; i += 2) { + for (i = VISLANDS30_IV_SRCID_D1_GRPH_PFLIP; i < 20; i += 2) { r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, i, &adev->pageflip_irq); if (r) return r; } /* HPD hotplug */ - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 42, &adev->hpd_irq); + r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_HOTPLUG_DETECT_A, &adev->hpd_irq); if (r) return r; diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c index 394cc1e8fe20..c9adc627305d 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c @@ -1811,15 +1811,14 @@ static int dce_v6_0_crtc_do_set_base(struct drm_crtc *crtc, if (unlikely(r != 0)) return r; - if (atomic) { - fb_location = amdgpu_bo_gpu_offset(abo); - } else { - r = amdgpu_bo_pin(abo, AMDGPU_GEM_DOMAIN_VRAM, &fb_location); + if (!atomic) { + r = amdgpu_bo_pin(abo, AMDGPU_GEM_DOMAIN_VRAM); if (unlikely(r != 0)) { amdgpu_bo_unreserve(abo); return -EINVAL; } } + fb_location = amdgpu_bo_gpu_offset(abo); amdgpu_bo_get_tiling_flags(abo, &tiling_flags); amdgpu_bo_unreserve(abo); @@ -2263,13 +2262,14 @@ static int dce_v6_0_crtc_cursor_set2(struct drm_crtc *crtc, return ret; } - ret = amdgpu_bo_pin(aobj, AMDGPU_GEM_DOMAIN_VRAM, &amdgpu_crtc->cursor_addr); + ret = amdgpu_bo_pin(aobj, AMDGPU_GEM_DOMAIN_VRAM); amdgpu_bo_unreserve(aobj); if (ret) { DRM_ERROR("Failed to pin new cursor BO (%d)\n", ret); drm_gem_object_put_unlocked(obj); return ret; } + amdgpu_crtc->cursor_addr = amdgpu_bo_gpu_offset(aobj); dce_v6_0_lock_cursor(crtc, true); diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c index c9b9ab8f1b05..50cd03beac7d 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c @@ -1786,15 +1786,14 @@ static int dce_v8_0_crtc_do_set_base(struct drm_crtc *crtc, if (unlikely(r != 0)) return r; - if (atomic) { - fb_location = amdgpu_bo_gpu_offset(abo); - } else { - r = amdgpu_bo_pin(abo, AMDGPU_GEM_DOMAIN_VRAM, &fb_location); + if (!atomic) { + r = amdgpu_bo_pin(abo, AMDGPU_GEM_DOMAIN_VRAM); if (unlikely(r != 0)) { amdgpu_bo_unreserve(abo); return -EINVAL; } } + fb_location = amdgpu_bo_gpu_offset(abo); amdgpu_bo_get_tiling_flags(abo, &tiling_flags); amdgpu_bo_unreserve(abo); @@ -2274,13 +2273,14 @@ static int dce_v8_0_crtc_cursor_set2(struct drm_crtc *crtc, return ret; } - ret = amdgpu_bo_pin(aobj, AMDGPU_GEM_DOMAIN_VRAM, &amdgpu_crtc->cursor_addr); + ret = amdgpu_bo_pin(aobj, AMDGPU_GEM_DOMAIN_VRAM); amdgpu_bo_unreserve(aobj); if (ret) { DRM_ERROR("Failed to pin new cursor BO (%d)\n", ret); drm_gem_object_put_unlocked(obj); return ret; } + amdgpu_crtc->cursor_addr = amdgpu_bo_gpu_offset(aobj); dce_v8_0_lock_cursor(crtc, true); diff --git a/drivers/gpu/drm/amd/amdgpu/dce_virtual.c b/drivers/gpu/drm/amd/amdgpu/dce_virtual.c index dbf2ccd0c744..15257634a53a 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_virtual.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_virtual.c @@ -36,6 +36,7 @@ #include "dce_v10_0.h" #include "dce_v11_0.h" #include "dce_virtual.h" +#include "ivsrcid/ivsrcid_vislands30.h" #define DCE_VIRTUAL_VBLANK_PERIOD 16666666 @@ -269,25 +270,18 @@ static int dce_virtual_early_init(void *handle) static struct drm_encoder * dce_virtual_encoder(struct drm_connector *connector) { - int enc_id = connector->encoder_ids[0]; struct drm_encoder *encoder; int i; - for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) { - if (connector->encoder_ids[i] == 0) - break; - - encoder = drm_encoder_find(connector->dev, NULL, connector->encoder_ids[i]); - if (!encoder) - continue; - + drm_connector_for_each_possible_encoder(connector, encoder, i) { if (encoder->encoder_type == DRM_MODE_ENCODER_VIRTUAL) return encoder; } /* pick the first one */ - if (enc_id) - return drm_encoder_find(connector->dev, NULL, enc_id); + drm_connector_for_each_possible_encoder(connector, encoder, i) + return encoder; + return NULL; } @@ -378,7 +372,7 @@ static int dce_virtual_sw_init(void *handle) int r, i; struct amdgpu_device *adev = (struct amdgpu_device *)handle; - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 229, &adev->crtc_irq); + r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_SMU_DISP_TIMER2_TRIGGER, &adev->crtc_irq); if (r) return r; @@ -634,7 +628,7 @@ static int dce_virtual_connector_encoder_init(struct amdgpu_device *adev, drm_connector_register(connector); /* link them */ - drm_mode_connector_attach_encoder(connector, encoder); + drm_connector_attach_encoder(connector, encoder); return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c index cd6bf291a853..de184a886057 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c @@ -44,30 +44,30 @@ static void gfx_v6_0_set_ring_funcs(struct amdgpu_device *adev); static void gfx_v6_0_set_irq_funcs(struct amdgpu_device *adev); static void gfx_v6_0_get_cu_info(struct amdgpu_device *adev); -MODULE_FIRMWARE("radeon/tahiti_pfp.bin"); -MODULE_FIRMWARE("radeon/tahiti_me.bin"); -MODULE_FIRMWARE("radeon/tahiti_ce.bin"); -MODULE_FIRMWARE("radeon/tahiti_rlc.bin"); - -MODULE_FIRMWARE("radeon/pitcairn_pfp.bin"); -MODULE_FIRMWARE("radeon/pitcairn_me.bin"); -MODULE_FIRMWARE("radeon/pitcairn_ce.bin"); -MODULE_FIRMWARE("radeon/pitcairn_rlc.bin"); - -MODULE_FIRMWARE("radeon/verde_pfp.bin"); -MODULE_FIRMWARE("radeon/verde_me.bin"); -MODULE_FIRMWARE("radeon/verde_ce.bin"); -MODULE_FIRMWARE("radeon/verde_rlc.bin"); - -MODULE_FIRMWARE("radeon/oland_pfp.bin"); -MODULE_FIRMWARE("radeon/oland_me.bin"); -MODULE_FIRMWARE("radeon/oland_ce.bin"); -MODULE_FIRMWARE("radeon/oland_rlc.bin"); - -MODULE_FIRMWARE("radeon/hainan_pfp.bin"); -MODULE_FIRMWARE("radeon/hainan_me.bin"); -MODULE_FIRMWARE("radeon/hainan_ce.bin"); -MODULE_FIRMWARE("radeon/hainan_rlc.bin"); +MODULE_FIRMWARE("amdgpu/tahiti_pfp.bin"); +MODULE_FIRMWARE("amdgpu/tahiti_me.bin"); +MODULE_FIRMWARE("amdgpu/tahiti_ce.bin"); +MODULE_FIRMWARE("amdgpu/tahiti_rlc.bin"); + +MODULE_FIRMWARE("amdgpu/pitcairn_pfp.bin"); +MODULE_FIRMWARE("amdgpu/pitcairn_me.bin"); +MODULE_FIRMWARE("amdgpu/pitcairn_ce.bin"); +MODULE_FIRMWARE("amdgpu/pitcairn_rlc.bin"); + +MODULE_FIRMWARE("amdgpu/verde_pfp.bin"); +MODULE_FIRMWARE("amdgpu/verde_me.bin"); +MODULE_FIRMWARE("amdgpu/verde_ce.bin"); +MODULE_FIRMWARE("amdgpu/verde_rlc.bin"); + +MODULE_FIRMWARE("amdgpu/oland_pfp.bin"); +MODULE_FIRMWARE("amdgpu/oland_me.bin"); +MODULE_FIRMWARE("amdgpu/oland_ce.bin"); +MODULE_FIRMWARE("amdgpu/oland_rlc.bin"); + +MODULE_FIRMWARE("amdgpu/hainan_pfp.bin"); +MODULE_FIRMWARE("amdgpu/hainan_me.bin"); +MODULE_FIRMWARE("amdgpu/hainan_ce.bin"); +MODULE_FIRMWARE("amdgpu/hainan_rlc.bin"); static u32 gfx_v6_0_get_csb_size(struct amdgpu_device *adev); static void gfx_v6_0_get_csb_buffer(struct amdgpu_device *adev, volatile u32 *buffer); @@ -335,7 +335,7 @@ static int gfx_v6_0_init_microcode(struct amdgpu_device *adev) default: BUG(); } - snprintf(fw_name, sizeof(fw_name), "radeon/%s_pfp.bin", chip_name); + snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_pfp.bin", chip_name); err = request_firmware(&adev->gfx.pfp_fw, fw_name, adev->dev); if (err) goto out; @@ -346,7 +346,7 @@ static int gfx_v6_0_init_microcode(struct amdgpu_device *adev) adev->gfx.pfp_fw_version = le32_to_cpu(cp_hdr->header.ucode_version); adev->gfx.pfp_feature_version = le32_to_cpu(cp_hdr->ucode_feature_version); - snprintf(fw_name, sizeof(fw_name), "radeon/%s_me.bin", chip_name); + snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_me.bin", chip_name); err = request_firmware(&adev->gfx.me_fw, fw_name, adev->dev); if (err) goto out; @@ -357,7 +357,7 @@ static int gfx_v6_0_init_microcode(struct amdgpu_device *adev) adev->gfx.me_fw_version = le32_to_cpu(cp_hdr->header.ucode_version); adev->gfx.me_feature_version = le32_to_cpu(cp_hdr->ucode_feature_version); - snprintf(fw_name, sizeof(fw_name), "radeon/%s_ce.bin", chip_name); + snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_ce.bin", chip_name); err = request_firmware(&adev->gfx.ce_fw, fw_name, adev->dev); if (err) goto out; @@ -368,7 +368,7 @@ static int gfx_v6_0_init_microcode(struct amdgpu_device *adev) adev->gfx.ce_fw_version = le32_to_cpu(cp_hdr->header.ucode_version); adev->gfx.ce_feature_version = le32_to_cpu(cp_hdr->ucode_feature_version); - snprintf(fw_name, sizeof(fw_name), "radeon/%s_rlc.bin", chip_name); + snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_rlc.bin", chip_name); err = request_firmware(&adev->gfx.rlc_fw, fw_name, adev->dev); if (err) goto out; diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c index 42b6144c1fd5..95452c5a9df6 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c @@ -57,36 +57,36 @@ static void gfx_v7_0_set_ring_funcs(struct amdgpu_device *adev); static void gfx_v7_0_set_irq_funcs(struct amdgpu_device *adev); static void gfx_v7_0_set_gds_init(struct amdgpu_device *adev); -MODULE_FIRMWARE("radeon/bonaire_pfp.bin"); -MODULE_FIRMWARE("radeon/bonaire_me.bin"); -MODULE_FIRMWARE("radeon/bonaire_ce.bin"); -MODULE_FIRMWARE("radeon/bonaire_rlc.bin"); -MODULE_FIRMWARE("radeon/bonaire_mec.bin"); - -MODULE_FIRMWARE("radeon/hawaii_pfp.bin"); -MODULE_FIRMWARE("radeon/hawaii_me.bin"); -MODULE_FIRMWARE("radeon/hawaii_ce.bin"); -MODULE_FIRMWARE("radeon/hawaii_rlc.bin"); -MODULE_FIRMWARE("radeon/hawaii_mec.bin"); - -MODULE_FIRMWARE("radeon/kaveri_pfp.bin"); -MODULE_FIRMWARE("radeon/kaveri_me.bin"); -MODULE_FIRMWARE("radeon/kaveri_ce.bin"); -MODULE_FIRMWARE("radeon/kaveri_rlc.bin"); -MODULE_FIRMWARE("radeon/kaveri_mec.bin"); -MODULE_FIRMWARE("radeon/kaveri_mec2.bin"); - -MODULE_FIRMWARE("radeon/kabini_pfp.bin"); -MODULE_FIRMWARE("radeon/kabini_me.bin"); -MODULE_FIRMWARE("radeon/kabini_ce.bin"); -MODULE_FIRMWARE("radeon/kabini_rlc.bin"); -MODULE_FIRMWARE("radeon/kabini_mec.bin"); - -MODULE_FIRMWARE("radeon/mullins_pfp.bin"); -MODULE_FIRMWARE("radeon/mullins_me.bin"); -MODULE_FIRMWARE("radeon/mullins_ce.bin"); -MODULE_FIRMWARE("radeon/mullins_rlc.bin"); -MODULE_FIRMWARE("radeon/mullins_mec.bin"); +MODULE_FIRMWARE("amdgpu/bonaire_pfp.bin"); +MODULE_FIRMWARE("amdgpu/bonaire_me.bin"); +MODULE_FIRMWARE("amdgpu/bonaire_ce.bin"); +MODULE_FIRMWARE("amdgpu/bonaire_rlc.bin"); +MODULE_FIRMWARE("amdgpu/bonaire_mec.bin"); + +MODULE_FIRMWARE("amdgpu/hawaii_pfp.bin"); +MODULE_FIRMWARE("amdgpu/hawaii_me.bin"); +MODULE_FIRMWARE("amdgpu/hawaii_ce.bin"); +MODULE_FIRMWARE("amdgpu/hawaii_rlc.bin"); +MODULE_FIRMWARE("amdgpu/hawaii_mec.bin"); + +MODULE_FIRMWARE("amdgpu/kaveri_pfp.bin"); +MODULE_FIRMWARE("amdgpu/kaveri_me.bin"); +MODULE_FIRMWARE("amdgpu/kaveri_ce.bin"); +MODULE_FIRMWARE("amdgpu/kaveri_rlc.bin"); +MODULE_FIRMWARE("amdgpu/kaveri_mec.bin"); +MODULE_FIRMWARE("amdgpu/kaveri_mec2.bin"); + +MODULE_FIRMWARE("amdgpu/kabini_pfp.bin"); +MODULE_FIRMWARE("amdgpu/kabini_me.bin"); +MODULE_FIRMWARE("amdgpu/kabini_ce.bin"); +MODULE_FIRMWARE("amdgpu/kabini_rlc.bin"); +MODULE_FIRMWARE("amdgpu/kabini_mec.bin"); + +MODULE_FIRMWARE("amdgpu/mullins_pfp.bin"); +MODULE_FIRMWARE("amdgpu/mullins_me.bin"); +MODULE_FIRMWARE("amdgpu/mullins_ce.bin"); +MODULE_FIRMWARE("amdgpu/mullins_rlc.bin"); +MODULE_FIRMWARE("amdgpu/mullins_mec.bin"); static const struct amdgpu_gds_reg_offset amdgpu_gds_reg_offset[] = { @@ -925,7 +925,7 @@ static int gfx_v7_0_init_microcode(struct amdgpu_device *adev) default: BUG(); } - snprintf(fw_name, sizeof(fw_name), "radeon/%s_pfp.bin", chip_name); + snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_pfp.bin", chip_name); err = request_firmware(&adev->gfx.pfp_fw, fw_name, adev->dev); if (err) goto out; @@ -933,7 +933,7 @@ static int gfx_v7_0_init_microcode(struct amdgpu_device *adev) if (err) goto out; - snprintf(fw_name, sizeof(fw_name), "radeon/%s_me.bin", chip_name); + snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_me.bin", chip_name); err = request_firmware(&adev->gfx.me_fw, fw_name, adev->dev); if (err) goto out; @@ -941,7 +941,7 @@ static int gfx_v7_0_init_microcode(struct amdgpu_device *adev) if (err) goto out; - snprintf(fw_name, sizeof(fw_name), "radeon/%s_ce.bin", chip_name); + snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_ce.bin", chip_name); err = request_firmware(&adev->gfx.ce_fw, fw_name, adev->dev); if (err) goto out; @@ -949,7 +949,7 @@ static int gfx_v7_0_init_microcode(struct amdgpu_device *adev) if (err) goto out; - snprintf(fw_name, sizeof(fw_name), "radeon/%s_mec.bin", chip_name); + snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_mec.bin", chip_name); err = request_firmware(&adev->gfx.mec_fw, fw_name, adev->dev); if (err) goto out; @@ -958,7 +958,7 @@ static int gfx_v7_0_init_microcode(struct amdgpu_device *adev) goto out; if (adev->asic_type == CHIP_KAVERI) { - snprintf(fw_name, sizeof(fw_name), "radeon/%s_mec2.bin", chip_name); + snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_mec2.bin", chip_name); err = request_firmware(&adev->gfx.mec2_fw, fw_name, adev->dev); if (err) goto out; @@ -967,7 +967,7 @@ static int gfx_v7_0_init_microcode(struct amdgpu_device *adev) goto out; } - snprintf(fw_name, sizeof(fw_name), "radeon/%s_rlc.bin", chip_name); + snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_rlc.bin", chip_name); err = request_firmware(&adev->gfx.rlc_fw, fw_name, adev->dev); if (err) goto out; diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c index 818874b13c99..5cd45210113f 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c @@ -51,6 +51,8 @@ #include "smu/smu_7_1_3_d.h" +#include "ivsrcid/ivsrcid_vislands30.h" + #define GFX8_NUM_GFX_RINGS 1 #define GFX8_MEC_HPD_SIZE 2048 @@ -704,6 +706,17 @@ static const u32 stoney_mgcg_cgcg_init[] = mmCGTS_SM_CTRL_REG, 0xffffffff, 0x96940200, }; + +static const char * const sq_edc_source_names[] = { + "SQ_EDC_INFO_SOURCE_INVALID: No EDC error has occurred", + "SQ_EDC_INFO_SOURCE_INST: EDC source is Instruction Fetch", + "SQ_EDC_INFO_SOURCE_SGPR: EDC source is SGPR or SQC data return", + "SQ_EDC_INFO_SOURCE_VGPR: EDC source is VGPR", + "SQ_EDC_INFO_SOURCE_LDS: EDC source is LDS", + "SQ_EDC_INFO_SOURCE_GDS: EDC source is GDS", + "SQ_EDC_INFO_SOURCE_TA: EDC source is TA", +}; + static void gfx_v8_0_set_ring_funcs(struct amdgpu_device *adev); static void gfx_v8_0_set_irq_funcs(struct amdgpu_device *adev); static void gfx_v8_0_set_gds_init(struct amdgpu_device *adev); @@ -866,26 +879,32 @@ static int gfx_v8_0_ring_test_ib(struct amdgpu_ring *ring, long timeout) struct amdgpu_device *adev = ring->adev; struct amdgpu_ib ib; struct dma_fence *f = NULL; - uint32_t scratch; - uint32_t tmp = 0; + + unsigned int index; + uint64_t gpu_addr; + uint32_t tmp; long r; - r = amdgpu_gfx_scratch_get(adev, &scratch); + r = amdgpu_device_wb_get(adev, &index); if (r) { - DRM_ERROR("amdgpu: failed to get scratch reg (%ld).\n", r); + dev_err(adev->dev, "(%ld) failed to allocate wb slot\n", r); return r; } - WREG32(scratch, 0xCAFEDEAD); + + gpu_addr = adev->wb.gpu_addr + (index * 4); + adev->wb.wb[index] = cpu_to_le32(0xCAFEDEAD); memset(&ib, 0, sizeof(ib)); - r = amdgpu_ib_get(adev, NULL, 256, &ib); + r = amdgpu_ib_get(adev, NULL, 16, &ib); if (r) { DRM_ERROR("amdgpu: failed to get ib (%ld).\n", r); goto err1; } - ib.ptr[0] = PACKET3(PACKET3_SET_UCONFIG_REG, 1); - ib.ptr[1] = ((scratch - PACKET3_SET_UCONFIG_REG_START)); - ib.ptr[2] = 0xDEADBEEF; - ib.length_dw = 3; + ib.ptr[0] = PACKET3(PACKET3_WRITE_DATA, 3); + ib.ptr[1] = WRITE_DATA_DST_SEL(5) | WR_CONFIRM; + ib.ptr[2] = lower_32_bits(gpu_addr); + ib.ptr[3] = upper_32_bits(gpu_addr); + ib.ptr[4] = 0xDEADBEEF; + ib.length_dw = 5; r = amdgpu_ib_schedule(ring, 1, &ib, NULL, &f); if (r) @@ -900,20 +919,21 @@ static int gfx_v8_0_ring_test_ib(struct amdgpu_ring *ring, long timeout) DRM_ERROR("amdgpu: fence wait failed (%ld).\n", r); goto err2; } - tmp = RREG32(scratch); + + tmp = adev->wb.wb[index]; if (tmp == 0xDEADBEEF) { DRM_DEBUG("ib test on ring %d succeeded\n", ring->idx); r = 0; } else { - DRM_ERROR("amdgpu: ib test failed (scratch(0x%04X)=0x%08X)\n", - scratch, tmp); + DRM_ERROR("ib test on ring %d failed\n", ring->idx); r = -EINVAL; } + err2: amdgpu_ib_free(adev, &ib, NULL); dma_fence_put(f); err1: - amdgpu_gfx_scratch_free(adev, scratch); + amdgpu_device_wb_free(adev, index); return r; } @@ -1999,6 +2019,8 @@ static int gfx_v8_0_compute_ring_init(struct amdgpu_device *adev, int ring_id, return 0; } +static void gfx_v8_0_sq_irq_work_func(struct work_struct *work); + static int gfx_v8_0_sw_init(void *handle) { int i, j, k, r, ring_id; @@ -2027,27 +2049,43 @@ static int gfx_v8_0_sw_init(void *handle) adev->gfx.mec.num_queue_per_pipe = 8; /* KIQ event */ - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 178, &adev->gfx.kiq.irq); + r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_CP_INT_IB2, &adev->gfx.kiq.irq); if (r) return r; /* EOP Event */ - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 181, &adev->gfx.eop_irq); + r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_CP_END_OF_PIPE, &adev->gfx.eop_irq); if (r) return r; /* Privileged reg */ - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 184, + r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_CP_PRIV_REG_FAULT, &adev->gfx.priv_reg_irq); if (r) return r; /* Privileged inst */ - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 185, + r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_CP_PRIV_INSTR_FAULT, &adev->gfx.priv_inst_irq); if (r) return r; + /* Add CP EDC/ECC irq */ + r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_CP_ECC_ERROR, + &adev->gfx.cp_ecc_error_irq); + if (r) + return r; + + /* SQ interrupts. */ + r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_SQ_INTERRUPT_MSG, + &adev->gfx.sq_irq); + if (r) { + DRM_ERROR("amdgpu_irq_add() for SQ failed: %d\n", r); + return r; + } + + INIT_WORK(&adev->gfx.sq_work.work, gfx_v8_0_sq_irq_work_func); + adev->gfx.gfx_current_status = AMDGPU_GFX_NORMAL_MODE; gfx_v8_0_scratch_init(adev); @@ -5111,6 +5149,10 @@ static int gfx_v8_0_hw_fini(void *handle) amdgpu_irq_put(adev, &adev->gfx.priv_reg_irq, 0); amdgpu_irq_put(adev, &adev->gfx.priv_inst_irq, 0); + amdgpu_irq_put(adev, &adev->gfx.cp_ecc_error_irq, 0); + + amdgpu_irq_put(adev, &adev->gfx.sq_irq, 0); + /* disable KCQ to avoid CPC touch memory not valid anymore */ for (i = 0; i < adev->gfx.num_compute_rings; i++) gfx_v8_0_kcq_disable(&adev->gfx.kiq.ring, &adev->gfx.compute_ring[i]); @@ -5542,9 +5584,19 @@ static int gfx_v8_0_late_init(void *handle) if (r) return r; - amdgpu_device_ip_set_powergating_state(adev, - AMD_IP_BLOCK_TYPE_GFX, - AMD_PG_STATE_GATE); + r = amdgpu_irq_get(adev, &adev->gfx.cp_ecc_error_irq, 0); + if (r) { + DRM_ERROR("amdgpu_irq_get() failed to get IRQ for EDC, r: %d.\n", r); + return r; + } + + r = amdgpu_irq_get(adev, &adev->gfx.sq_irq, 0); + if (r) { + DRM_ERROR( + "amdgpu_irq_get() failed to get IRQ for SQ, r: %d.\n", + r); + return r; + } return 0; } @@ -5552,14 +5604,12 @@ static int gfx_v8_0_late_init(void *handle) static void gfx_v8_0_enable_gfx_static_mg_power_gating(struct amdgpu_device *adev, bool enable) { - if ((adev->asic_type == CHIP_POLARIS11) || + if (((adev->asic_type == CHIP_POLARIS11) || (adev->asic_type == CHIP_POLARIS12) || - (adev->asic_type == CHIP_VEGAM)) + (adev->asic_type == CHIP_VEGAM)) && + adev->powerplay.pp_funcs->set_powergating_by_smu) /* Send msg to SMU via Powerplay */ - amdgpu_device_ip_set_powergating_state(adev, - AMD_IP_BLOCK_TYPE_SMC, - enable ? - AMD_PG_STATE_GATE : AMD_PG_STATE_UNGATE); + amdgpu_dpm_set_powergating_by_smu(adev, AMD_IP_BLOCK_TYPE_GFX, enable); WREG32_FIELD(RLC_PG_CNTL, STATIC_PER_CU_PG_ENABLE, enable ? 1 : 0); } @@ -6787,6 +6837,77 @@ static int gfx_v8_0_set_eop_interrupt_state(struct amdgpu_device *adev, return 0; } +static int gfx_v8_0_set_cp_ecc_int_state(struct amdgpu_device *adev, + struct amdgpu_irq_src *source, + unsigned int type, + enum amdgpu_interrupt_state state) +{ + int enable_flag; + + switch (state) { + case AMDGPU_IRQ_STATE_DISABLE: + enable_flag = 0; + break; + + case AMDGPU_IRQ_STATE_ENABLE: + enable_flag = 1; + break; + + default: + return -EINVAL; + } + + WREG32_FIELD(CP_INT_CNTL, CP_ECC_ERROR_INT_ENABLE, enable_flag); + WREG32_FIELD(CP_INT_CNTL_RING0, CP_ECC_ERROR_INT_ENABLE, enable_flag); + WREG32_FIELD(CP_INT_CNTL_RING1, CP_ECC_ERROR_INT_ENABLE, enable_flag); + WREG32_FIELD(CP_INT_CNTL_RING2, CP_ECC_ERROR_INT_ENABLE, enable_flag); + WREG32_FIELD(CPC_INT_CNTL, CP_ECC_ERROR_INT_ENABLE, enable_flag); + WREG32_FIELD(CP_ME1_PIPE0_INT_CNTL, CP_ECC_ERROR_INT_ENABLE, + enable_flag); + WREG32_FIELD(CP_ME1_PIPE1_INT_CNTL, CP_ECC_ERROR_INT_ENABLE, + enable_flag); + WREG32_FIELD(CP_ME1_PIPE2_INT_CNTL, CP_ECC_ERROR_INT_ENABLE, + enable_flag); + WREG32_FIELD(CP_ME1_PIPE3_INT_CNTL, CP_ECC_ERROR_INT_ENABLE, + enable_flag); + WREG32_FIELD(CP_ME2_PIPE0_INT_CNTL, CP_ECC_ERROR_INT_ENABLE, + enable_flag); + WREG32_FIELD(CP_ME2_PIPE1_INT_CNTL, CP_ECC_ERROR_INT_ENABLE, + enable_flag); + WREG32_FIELD(CP_ME2_PIPE2_INT_CNTL, CP_ECC_ERROR_INT_ENABLE, + enable_flag); + WREG32_FIELD(CP_ME2_PIPE3_INT_CNTL, CP_ECC_ERROR_INT_ENABLE, + enable_flag); + + return 0; +} + +static int gfx_v8_0_set_sq_int_state(struct amdgpu_device *adev, + struct amdgpu_irq_src *source, + unsigned int type, + enum amdgpu_interrupt_state state) +{ + int enable_flag; + + switch (state) { + case AMDGPU_IRQ_STATE_DISABLE: + enable_flag = 1; + break; + + case AMDGPU_IRQ_STATE_ENABLE: + enable_flag = 0; + break; + + default: + return -EINVAL; + } + + WREG32_FIELD(SQ_INTERRUPT_MSG_CTRL, STALL, + enable_flag); + + return 0; +} + static int gfx_v8_0_eop_irq(struct amdgpu_device *adev, struct amdgpu_irq_src *source, struct amdgpu_iv_entry *entry) @@ -6837,6 +6958,114 @@ static int gfx_v8_0_priv_inst_irq(struct amdgpu_device *adev, return 0; } +static int gfx_v8_0_cp_ecc_error_irq(struct amdgpu_device *adev, + struct amdgpu_irq_src *source, + struct amdgpu_iv_entry *entry) +{ + DRM_ERROR("CP EDC/ECC error detected."); + return 0; +} + +static void gfx_v8_0_parse_sq_irq(struct amdgpu_device *adev, unsigned ih_data) +{ + u32 enc, se_id, sh_id, cu_id; + char type[20]; + int sq_edc_source = -1; + + enc = REG_GET_FIELD(ih_data, SQ_INTERRUPT_WORD_CMN, ENCODING); + se_id = REG_GET_FIELD(ih_data, SQ_INTERRUPT_WORD_CMN, SE_ID); + + switch (enc) { + case 0: + DRM_INFO("SQ general purpose intr detected:" + "se_id %d, immed_overflow %d, host_reg_overflow %d," + "host_cmd_overflow %d, cmd_timestamp %d," + "reg_timestamp %d, thread_trace_buff_full %d," + "wlt %d, thread_trace %d.\n", + se_id, + REG_GET_FIELD(ih_data, SQ_INTERRUPT_WORD_AUTO, IMMED_OVERFLOW), + REG_GET_FIELD(ih_data, SQ_INTERRUPT_WORD_AUTO, HOST_REG_OVERFLOW), + REG_GET_FIELD(ih_data, SQ_INTERRUPT_WORD_AUTO, HOST_CMD_OVERFLOW), + REG_GET_FIELD(ih_data, SQ_INTERRUPT_WORD_AUTO, CMD_TIMESTAMP), + REG_GET_FIELD(ih_data, SQ_INTERRUPT_WORD_AUTO, REG_TIMESTAMP), + REG_GET_FIELD(ih_data, SQ_INTERRUPT_WORD_AUTO, THREAD_TRACE_BUF_FULL), + REG_GET_FIELD(ih_data, SQ_INTERRUPT_WORD_AUTO, WLT), + REG_GET_FIELD(ih_data, SQ_INTERRUPT_WORD_AUTO, THREAD_TRACE) + ); + break; + case 1: + case 2: + + cu_id = REG_GET_FIELD(ih_data, SQ_INTERRUPT_WORD_WAVE, CU_ID); + sh_id = REG_GET_FIELD(ih_data, SQ_INTERRUPT_WORD_WAVE, SH_ID); + + /* + * This function can be called either directly from ISR + * or from BH in which case we can access SQ_EDC_INFO + * instance + */ + if (in_task()) { + mutex_lock(&adev->grbm_idx_mutex); + gfx_v8_0_select_se_sh(adev, se_id, sh_id, cu_id); + + sq_edc_source = REG_GET_FIELD(RREG32(mmSQ_EDC_INFO), SQ_EDC_INFO, SOURCE); + + gfx_v8_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff); + mutex_unlock(&adev->grbm_idx_mutex); + } + + if (enc == 1) + sprintf(type, "instruction intr"); + else + sprintf(type, "EDC/ECC error"); + + DRM_INFO( + "SQ %s detected: " + "se_id %d, sh_id %d, cu_id %d, simd_id %d, wave_id %d, vm_id %d " + "trap %s, sq_ed_info.source %s.\n", + type, se_id, sh_id, cu_id, + REG_GET_FIELD(ih_data, SQ_INTERRUPT_WORD_WAVE, SIMD_ID), + REG_GET_FIELD(ih_data, SQ_INTERRUPT_WORD_WAVE, WAVE_ID), + REG_GET_FIELD(ih_data, SQ_INTERRUPT_WORD_WAVE, VM_ID), + REG_GET_FIELD(ih_data, SQ_INTERRUPT_WORD_WAVE, PRIV) ? "true" : "false", + (sq_edc_source != -1) ? sq_edc_source_names[sq_edc_source] : "unavailable" + ); + break; + default: + DRM_ERROR("SQ invalid encoding type\n."); + } +} + +static void gfx_v8_0_sq_irq_work_func(struct work_struct *work) +{ + + struct amdgpu_device *adev = container_of(work, struct amdgpu_device, gfx.sq_work.work); + struct sq_work *sq_work = container_of(work, struct sq_work, work); + + gfx_v8_0_parse_sq_irq(adev, sq_work->ih_data); +} + +static int gfx_v8_0_sq_irq(struct amdgpu_device *adev, + struct amdgpu_irq_src *source, + struct amdgpu_iv_entry *entry) +{ + unsigned ih_data = entry->src_data[0]; + + /* + * Try to submit work so SQ_EDC_INFO can be accessed from + * BH. If previous work submission hasn't finished yet + * just print whatever info is possible directly from the ISR. + */ + if (work_pending(&adev->gfx.sq_work.work)) { + gfx_v8_0_parse_sq_irq(adev, ih_data); + } else { + adev->gfx.sq_work.ih_data = ih_data; + schedule_work(&adev->gfx.sq_work.work); + } + + return 0; +} + static int gfx_v8_0_kiq_set_interrupt_state(struct amdgpu_device *adev, struct amdgpu_irq_src *src, unsigned int type, @@ -7037,6 +7266,16 @@ static const struct amdgpu_irq_src_funcs gfx_v8_0_kiq_irq_funcs = { .process = gfx_v8_0_kiq_irq, }; +static const struct amdgpu_irq_src_funcs gfx_v8_0_cp_ecc_error_irq_funcs = { + .set = gfx_v8_0_set_cp_ecc_int_state, + .process = gfx_v8_0_cp_ecc_error_irq, +}; + +static const struct amdgpu_irq_src_funcs gfx_v8_0_sq_irq_funcs = { + .set = gfx_v8_0_set_sq_int_state, + .process = gfx_v8_0_sq_irq, +}; + static void gfx_v8_0_set_irq_funcs(struct amdgpu_device *adev) { adev->gfx.eop_irq.num_types = AMDGPU_CP_IRQ_LAST; @@ -7050,6 +7289,12 @@ static void gfx_v8_0_set_irq_funcs(struct amdgpu_device *adev) adev->gfx.kiq.irq.num_types = AMDGPU_CP_KIQ_IRQ_LAST; adev->gfx.kiq.irq.funcs = &gfx_v8_0_kiq_irq_funcs; + + adev->gfx.cp_ecc_error_irq.num_types = 1; + adev->gfx.cp_ecc_error_irq.funcs = &gfx_v8_0_cp_ecc_error_irq_funcs; + + adev->gfx.sq_irq.num_types = 1; + adev->gfx.sq_irq.funcs = &gfx_v8_0_sq_irq_funcs; } static void gfx_v8_0_set_rlc_funcs(struct amdgpu_device *adev) diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c index a69153435ea7..ef00d14f8645 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c @@ -38,6 +38,8 @@ #include "clearstate_gfx9.h" #include "v9_structs.h" +#include "ivsrcid/gfx/irqsrcs_gfx_9_0.h" + #define GFX9_NUM_GFX_RINGS 1 #define GFX9_MEC_HPD_SIZE 2048 #define RLCG_UCODE_LOADING_START_ADDRESS 0x00002000L @@ -102,11 +104,22 @@ static const struct soc15_reg_golden golden_settings_gc_9_0_vg10[] = { 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, 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, mmGB_ADDR_CONFIG, 0xffff77ff, 0x2a114042), SOC15_REG_GOLDEN_VALUE(GC, 0, mmGB_ADDR_CONFIG_READ, 0xffff77ff, 0x2a114042), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmIA_UTCL1_CNTL, 0x08000000, 0x08000080), SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_ENHANCE_1, 0x00008000, 0x00048000), + 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, mmRMI_UTCL1_CNTL2, 0x00030000, 0x00020000), - SOC15_REG_GOLDEN_VALUE(GC, 0, mmTD_CNTL, 0x00001800, 0x00000800) + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSPI_CONFIG_CNTL_1, 0x0000000f, 0x01000107), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmTD_CNTL, 0x00001800, 0x00000800), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmWD_UTCL1_CNTL, 0x08000000, 0x08000080) }; static const struct soc15_reg_golden golden_settings_gc_9_0_vg20[] = @@ -648,7 +661,10 @@ static int gfx_v9_0_init_microcode(struct amdgpu_device *adev) adev->firmware.fw_size += ALIGN(le32_to_cpu(header->ucode_size_bytes), PAGE_SIZE); - if (adev->gfx.rlc.is_rlc_v2_1) { + if (adev->gfx.rlc.is_rlc_v2_1 && + adev->gfx.rlc.save_restore_list_cntl_size_bytes && + adev->gfx.rlc.save_restore_list_gpm_size_bytes && + adev->gfx.rlc.save_restore_list_srm_size_bytes) { info = &adev->firmware.ucode[AMDGPU_UCODE_ID_RLC_RESTORE_LIST_CNTL]; info->ucode_id = AMDGPU_UCODE_ID_RLC_RESTORE_LIST_CNTL; info->fw = adev->gfx.rlc_fw; @@ -943,6 +959,7 @@ static int gfx_v9_0_rlc_init(struct amdgpu_device *adev) dst_ptr = adev->gfx.rlc.cs_ptr; gfx_v9_0_get_csb_buffer(adev, dst_ptr); amdgpu_bo_kunmap(adev->gfx.rlc.clear_state_obj); + amdgpu_bo_unpin(adev->gfx.rlc.clear_state_obj); amdgpu_bo_unreserve(adev->gfx.rlc.clear_state_obj); } @@ -971,6 +988,39 @@ static int gfx_v9_0_rlc_init(struct amdgpu_device *adev) return 0; } +static int gfx_v9_0_csb_vram_pin(struct amdgpu_device *adev) +{ + int r; + + r = amdgpu_bo_reserve(adev->gfx.rlc.clear_state_obj, false); + if (unlikely(r != 0)) + return r; + + r = amdgpu_bo_pin(adev->gfx.rlc.clear_state_obj, + AMDGPU_GEM_DOMAIN_VRAM); + if (!r) + adev->gfx.rlc.clear_state_gpu_addr = + amdgpu_bo_gpu_offset(adev->gfx.rlc.clear_state_obj); + + amdgpu_bo_unreserve(adev->gfx.rlc.clear_state_obj); + + return r; +} + +static void gfx_v9_0_csb_vram_unpin(struct amdgpu_device *adev) +{ + int r; + + if (!adev->gfx.rlc.clear_state_obj) + return; + + r = amdgpu_bo_reserve(adev->gfx.rlc.clear_state_obj, true); + if (likely(r == 0)) { + amdgpu_bo_unpin(adev->gfx.rlc.clear_state_obj); + amdgpu_bo_unreserve(adev->gfx.rlc.clear_state_obj); + } +} + static void gfx_v9_0_mec_fini(struct amdgpu_device *adev) { amdgpu_bo_free_kernel(&adev->gfx.mec.hpd_eop_obj, NULL, NULL); @@ -1451,23 +1501,23 @@ static int gfx_v9_0_sw_init(void *handle) adev->gfx.mec.num_queue_per_pipe = 8; /* KIQ event */ - r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_GRBM_CP, 178, &adev->gfx.kiq.irq); + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_GRBM_CP, GFX_9_0__SRCID__CP_IB2_INTERRUPT_PKT, &adev->gfx.kiq.irq); if (r) return r; /* EOP Event */ - r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_GRBM_CP, 181, &adev->gfx.eop_irq); + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_GRBM_CP, GFX_9_0__SRCID__CP_EOP_INTERRUPT, &adev->gfx.eop_irq); if (r) return r; /* Privileged reg */ - r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_GRBM_CP, 184, + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_GRBM_CP, GFX_9_0__SRCID__CP_PRIV_REG_FAULT, &adev->gfx.priv_reg_irq); if (r) return r; /* Privileged inst */ - r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_GRBM_CP, 185, + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_GRBM_CP, GFX_9_0__SRCID__CP_PRIV_INSTR_FAULT, &adev->gfx.priv_inst_irq); if (r) return r; @@ -2148,8 +2198,16 @@ static void gfx_v9_0_enable_gfx_dynamic_mg_power_gating(struct amdgpu_device *ad static void gfx_v9_0_init_pg(struct amdgpu_device *adev) { - if (!adev->gfx.rlc.is_rlc_v2_1) - return; + gfx_v9_0_init_csb(adev); + + /* + * Rlc save restore list is workable since v2_1. + * And it's needed by gfxoff feature. + */ + if (adev->gfx.rlc.is_rlc_v2_1) { + gfx_v9_1_init_rlc_save_restore_list(adev); + gfx_v9_0_enable_save_restore_machine(adev); + } if (adev->pg_flags & (AMD_PG_SUPPORT_GFX_PG | AMD_PG_SUPPORT_GFX_SMG | @@ -2157,10 +2215,6 @@ static void gfx_v9_0_init_pg(struct amdgpu_device *adev) AMD_PG_SUPPORT_CP | AMD_PG_SUPPORT_GDS | AMD_PG_SUPPORT_RLC_SMU_HS)) { - gfx_v9_0_init_csb(adev); - gfx_v9_1_init_rlc_save_restore_list(adev); - gfx_v9_0_enable_save_restore_machine(adev); - WREG32(mmRLC_JUMP_TABLE_RESTORE, adev->gfx.rlc.cp_table_gpu_addr >> 8); gfx_v9_0_init_gfx_power_gating(adev); @@ -2252,9 +2306,6 @@ static int gfx_v9_0_rlc_resume(struct amdgpu_device *adev) /* disable CG */ WREG32_SOC15(GC, 0, mmRLC_CGCG_CGLS_CTRL, 0); - /* disable PG */ - WREG32_SOC15(GC, 0, mmRLC_PG_CNTL, 0); - gfx_v9_0_rlc_reset(adev); gfx_v9_0_init_pg(adev); @@ -3116,6 +3167,10 @@ static int gfx_v9_0_hw_init(void *handle) gfx_v9_0_gpu_init(adev); + r = gfx_v9_0_csb_vram_pin(adev); + if (r) + return r; + r = gfx_v9_0_rlc_resume(adev); if (r) return r; @@ -3224,6 +3279,8 @@ static int gfx_v9_0_hw_fini(void *handle) gfx_v9_0_cp_enable(adev, false); gfx_v9_0_rlc_stop(adev); + gfx_v9_0_csb_vram_unpin(adev); + return 0; } @@ -3433,7 +3490,7 @@ static void gfx_v9_0_enter_rlc_safe_mode(struct amdgpu_device *adev) /* wait for RLC_SAFE_MODE */ for (i = 0; i < adev->usec_timeout; i++) { - if (!REG_GET_FIELD(SOC15_REG_OFFSET(GC, 0, mmRLC_SAFE_MODE), RLC_SAFE_MODE, CMD)) + if (!REG_GET_FIELD(RREG32_SOC15(GC, 0, mmRLC_SAFE_MODE), RLC_SAFE_MODE, CMD)) break; udelay(1); } @@ -3510,8 +3567,11 @@ static void gfx_v9_0_update_medium_grain_clock_gating(struct amdgpu_device *adev if (enable && (adev->cg_flags & AMD_CG_SUPPORT_GFX_MGCG)) { /* 1 - RLC_CGTT_MGCG_OVERRIDE */ def = data = RREG32_SOC15(GC, 0, mmRLC_CGTT_MGCG_OVERRIDE); - data &= ~(RLC_CGTT_MGCG_OVERRIDE__CPF_CGTT_SCLK_OVERRIDE_MASK | - RLC_CGTT_MGCG_OVERRIDE__GRBM_CGTT_SCLK_OVERRIDE_MASK | + + if (adev->asic_type != CHIP_VEGA12) + data &= ~RLC_CGTT_MGCG_OVERRIDE__CPF_CGTT_SCLK_OVERRIDE_MASK; + + data &= ~(RLC_CGTT_MGCG_OVERRIDE__GRBM_CGTT_SCLK_OVERRIDE_MASK | RLC_CGTT_MGCG_OVERRIDE__GFXIP_MGCG_OVERRIDE_MASK | RLC_CGTT_MGCG_OVERRIDE__GFXIP_MGLS_OVERRIDE_MASK); @@ -3541,11 +3601,15 @@ static void gfx_v9_0_update_medium_grain_clock_gating(struct amdgpu_device *adev } else { /* 1 - MGCG_OVERRIDE */ def = data = RREG32_SOC15(GC, 0, mmRLC_CGTT_MGCG_OVERRIDE); - data |= (RLC_CGTT_MGCG_OVERRIDE__CPF_CGTT_SCLK_OVERRIDE_MASK | - RLC_CGTT_MGCG_OVERRIDE__RLC_CGTT_SCLK_OVERRIDE_MASK | + + if (adev->asic_type != CHIP_VEGA12) + data |= RLC_CGTT_MGCG_OVERRIDE__CPF_CGTT_SCLK_OVERRIDE_MASK; + + data |= (RLC_CGTT_MGCG_OVERRIDE__RLC_CGTT_SCLK_OVERRIDE_MASK | RLC_CGTT_MGCG_OVERRIDE__GRBM_CGTT_SCLK_OVERRIDE_MASK | RLC_CGTT_MGCG_OVERRIDE__GFXIP_MGCG_OVERRIDE_MASK | RLC_CGTT_MGCG_OVERRIDE__GFXIP_MGLS_OVERRIDE_MASK); + if (def != data) WREG32_SOC15(GC, 0, mmRLC_CGTT_MGCG_OVERRIDE, data); @@ -3581,9 +3645,11 @@ static void gfx_v9_0_update_3d_clock_gating(struct amdgpu_device *adev, /* update CGCG and CGLS override bits */ if (def != data) WREG32_SOC15(GC, 0, mmRLC_CGTT_MGCG_OVERRIDE, data); - /* enable 3Dcgcg FSM(0x0020003f) */ + + /* enable 3Dcgcg FSM(0x0000363f) */ def = RREG32_SOC15(GC, 0, mmRLC_CGCG_CGLS_CTRL_3D); - data = (0x2000 << RLC_CGCG_CGLS_CTRL_3D__CGCG_GFX_IDLE_THRESHOLD__SHIFT) | + + data = (0x36 << RLC_CGCG_CGLS_CTRL_3D__CGCG_GFX_IDLE_THRESHOLD__SHIFT) | RLC_CGCG_CGLS_CTRL_3D__CGCG_EN_MASK; if (adev->cg_flags & AMD_CG_SUPPORT_GFX_3D_CGLS) data |= (0x000F << RLC_CGCG_CGLS_CTRL_3D__CGLS_REP_COMPANSAT_DELAY__SHIFT) | @@ -3630,9 +3696,10 @@ static void gfx_v9_0_update_coarse_grain_clock_gating(struct amdgpu_device *adev if (def != data) WREG32_SOC15(GC, 0, mmRLC_CGTT_MGCG_OVERRIDE, data); - /* enable cgcg FSM(0x0020003F) */ + /* enable cgcg FSM(0x0000363F) */ def = RREG32_SOC15(GC, 0, mmRLC_CGCG_CGLS_CTRL); - data = (0x2000 << RLC_CGCG_CGLS_CTRL__CGCG_GFX_IDLE_THRESHOLD__SHIFT) | + + data = (0x36 << RLC_CGCG_CGLS_CTRL__CGCG_GFX_IDLE_THRESHOLD__SHIFT) | RLC_CGCG_CGLS_CTRL__CGCG_EN_MASK; if (adev->cg_flags & AMD_CG_SUPPORT_GFX_CGLS) data |= (0x000F << RLC_CGCG_CGLS_CTRL__CGLS_REP_COMPANSAT_DELAY__SHIFT) | @@ -3714,6 +3781,15 @@ static int gfx_v9_0_set_powergating_state(void *handle, /* update mgcg state */ gfx_v9_0_update_gfx_mg_power_gating(adev, enable); + + /* set gfx off through smu */ + if (enable && adev->powerplay.pp_funcs->set_powergating_by_smu) + amdgpu_dpm_set_powergating_by_smu(adev, AMD_IP_BLOCK_TYPE_GFX, true); + break; + case CHIP_VEGA12: + /* set gfx off through smu */ + if (enable && adev->powerplay.pp_funcs->set_powergating_by_smu) + amdgpu_dpm_set_powergating_by_smu(adev, AMD_IP_BLOCK_TYPE_GFX, true); break; default: break; diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c index 79f9ac29019b..75317f283c69 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c @@ -41,11 +41,11 @@ static void gmc_v6_0_set_gmc_funcs(struct amdgpu_device *adev); static void gmc_v6_0_set_irq_funcs(struct amdgpu_device *adev); static int gmc_v6_0_wait_for_idle(void *handle); -MODULE_FIRMWARE("radeon/tahiti_mc.bin"); -MODULE_FIRMWARE("radeon/pitcairn_mc.bin"); -MODULE_FIRMWARE("radeon/verde_mc.bin"); -MODULE_FIRMWARE("radeon/oland_mc.bin"); -MODULE_FIRMWARE("radeon/si58_mc.bin"); +MODULE_FIRMWARE("amdgpu/tahiti_mc.bin"); +MODULE_FIRMWARE("amdgpu/pitcairn_mc.bin"); +MODULE_FIRMWARE("amdgpu/verde_mc.bin"); +MODULE_FIRMWARE("amdgpu/oland_mc.bin"); +MODULE_FIRMWARE("amdgpu/si58_mc.bin"); #define MC_SEQ_MISC0__MT__MASK 0xf0000000 #define MC_SEQ_MISC0__MT__GDDR1 0x10000000 @@ -134,9 +134,9 @@ static int gmc_v6_0_init_microcode(struct amdgpu_device *adev) is_58_fw = true; if (is_58_fw) - snprintf(fw_name, sizeof(fw_name), "radeon/si58_mc.bin"); + snprintf(fw_name, sizeof(fw_name), "amdgpu/si58_mc.bin"); else - snprintf(fw_name, sizeof(fw_name), "radeon/%s_mc.bin", chip_name); + snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_mc.bin", chip_name); err = request_firmware(&adev->gmc.fw, fw_name, adev->dev); if (err) goto out; diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c index 7147bfe25a23..36dc367c4b45 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c @@ -28,6 +28,7 @@ #include "cik.h" #include "gmc_v7_0.h" #include "amdgpu_ucode.h" +#include "amdgpu_amdkfd.h" #include "bif/bif_4_1_d.h" #include "bif/bif_4_1_sh_mask.h" @@ -43,12 +44,14 @@ #include "amdgpu_atombios.h" +#include "ivsrcid/ivsrcid_vislands30.h" + static void gmc_v7_0_set_gmc_funcs(struct amdgpu_device *adev); static void gmc_v7_0_set_irq_funcs(struct amdgpu_device *adev); static int gmc_v7_0_wait_for_idle(void *handle); -MODULE_FIRMWARE("radeon/bonaire_mc.bin"); -MODULE_FIRMWARE("radeon/hawaii_mc.bin"); +MODULE_FIRMWARE("amdgpu/bonaire_mc.bin"); +MODULE_FIRMWARE("amdgpu/hawaii_mc.bin"); MODULE_FIRMWARE("amdgpu/topaz_mc.bin"); static const u32 golden_settings_iceland_a11[] = @@ -147,10 +150,7 @@ static int gmc_v7_0_init_microcode(struct amdgpu_device *adev) default: BUG(); } - if (adev->asic_type == CHIP_TOPAZ) - snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_mc.bin", chip_name); - else - snprintf(fw_name, sizeof(fw_name), "radeon/%s_mc.bin", chip_name); + snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_mc.bin", chip_name); err = request_firmware(&adev->gmc.fw, fw_name, adev->dev); if (err) @@ -999,11 +999,11 @@ static int gmc_v7_0_sw_init(void *handle) adev->gmc.vram_type = gmc_v7_0_convert_vram_type(tmp); } - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 146, &adev->gmc.vm_fault); + r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_GFX_PAGE_INV_FAULT, &adev->gmc.vm_fault); if (r) return r; - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 147, &adev->gmc.vm_fault); + r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_GFX_MEM_PROT_FAULT, &adev->gmc.vm_fault); if (r) return r; @@ -1079,6 +1079,12 @@ static int gmc_v7_0_sw_init(void *handle) adev->vm_manager.vram_base_offset = 0; } + adev->gmc.vm_fault_info = kmalloc(sizeof(struct kfd_vm_fault_info), + GFP_KERNEL); + if (!adev->gmc.vm_fault_info) + return -ENOMEM; + atomic_set(&adev->gmc.vm_fault_info_updated, 0); + return 0; } @@ -1088,6 +1094,7 @@ static int gmc_v7_0_sw_fini(void *handle) amdgpu_gem_force_release(adev); amdgpu_vm_manager_fini(adev); + kfree(adev->gmc.vm_fault_info); gmc_v7_0_gart_fini(adev); amdgpu_bo_fini(adev); release_firmware(adev->gmc.fw); @@ -1277,7 +1284,7 @@ static int gmc_v7_0_process_interrupt(struct amdgpu_device *adev, struct amdgpu_irq_src *source, struct amdgpu_iv_entry *entry) { - u32 addr, status, mc_client; + u32 addr, status, mc_client, vmid; addr = RREG32(mmVM_CONTEXT1_PROTECTION_FAULT_ADDR); status = RREG32(mmVM_CONTEXT1_PROTECTION_FAULT_STATUS); @@ -1302,6 +1309,29 @@ static int gmc_v7_0_process_interrupt(struct amdgpu_device *adev, entry->pasid); } + vmid = REG_GET_FIELD(status, VM_CONTEXT1_PROTECTION_FAULT_STATUS, + VMID); + if (amdgpu_amdkfd_is_kfd_vmid(adev, vmid) + && !atomic_read(&adev->gmc.vm_fault_info_updated)) { + struct kfd_vm_fault_info *info = adev->gmc.vm_fault_info; + u32 protections = REG_GET_FIELD(status, + VM_CONTEXT1_PROTECTION_FAULT_STATUS, + PROTECTIONS); + + info->vmid = vmid; + info->mc_id = REG_GET_FIELD(status, + VM_CONTEXT1_PROTECTION_FAULT_STATUS, + MEMORY_CLIENT_ID); + info->status = status; + info->page_addr = addr; + info->prot_valid = protections & 0x7 ? true : false; + info->prot_read = protections & 0x8 ? true : false; + info->prot_write = protections & 0x10 ? true : false; + info->prot_exec = protections & 0x20 ? true : false; + mb(); + atomic_set(&adev->gmc.vm_fault_info_updated, 1); + } + return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c index 1edbe6b477b5..70fc97b59b4f 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c @@ -26,6 +26,7 @@ #include "amdgpu.h" #include "gmc_v8_0.h" #include "amdgpu_ucode.h" +#include "amdgpu_amdkfd.h" #include "gmc/gmc_8_1_d.h" #include "gmc/gmc_8_1_sh_mask.h" @@ -44,6 +45,7 @@ #include "amdgpu_atombios.h" +#include "ivsrcid/ivsrcid_vislands30.h" static void gmc_v8_0_set_gmc_funcs(struct amdgpu_device *adev); static void gmc_v8_0_set_irq_funcs(struct amdgpu_device *adev); @@ -1101,11 +1103,11 @@ static int gmc_v8_0_sw_init(void *handle) adev->gmc.vram_type = gmc_v8_0_convert_vram_type(tmp); } - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 146, &adev->gmc.vm_fault); + r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_GFX_PAGE_INV_FAULT, &adev->gmc.vm_fault); if (r) return r; - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 147, &adev->gmc.vm_fault); + r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_GFX_MEM_PROT_FAULT, &adev->gmc.vm_fault); if (r) return r; @@ -1181,6 +1183,12 @@ static int gmc_v8_0_sw_init(void *handle) adev->vm_manager.vram_base_offset = 0; } + adev->gmc.vm_fault_info = kmalloc(sizeof(struct kfd_vm_fault_info), + GFP_KERNEL); + if (!adev->gmc.vm_fault_info) + return -ENOMEM; + atomic_set(&adev->gmc.vm_fault_info_updated, 0); + return 0; } @@ -1190,6 +1198,7 @@ static int gmc_v8_0_sw_fini(void *handle) amdgpu_gem_force_release(adev); amdgpu_vm_manager_fini(adev); + kfree(adev->gmc.vm_fault_info); gmc_v8_0_gart_fini(adev); amdgpu_bo_fini(adev); release_firmware(adev->gmc.fw); @@ -1425,7 +1434,7 @@ static int gmc_v8_0_process_interrupt(struct amdgpu_device *adev, struct amdgpu_irq_src *source, struct amdgpu_iv_entry *entry) { - u32 addr, status, mc_client; + u32 addr, status, mc_client, vmid; if (amdgpu_sriov_vf(adev)) { dev_err(adev->dev, "GPU fault detected: %d 0x%08x\n", @@ -1447,8 +1456,13 @@ static int gmc_v8_0_process_interrupt(struct amdgpu_device *adev, gmc_v8_0_set_fault_enable_default(adev, false); if (printk_ratelimit()) { - dev_err(adev->dev, "GPU fault detected: %d 0x%08x\n", - entry->src_id, entry->src_data[0]); + struct amdgpu_task_info task_info = { 0 }; + + amdgpu_vm_get_task_info(adev, entry->pasid, &task_info); + + dev_err(adev->dev, "GPU fault detected: %d 0x%08x for process %s pid %d thread %s pid %d\n", + entry->src_id, entry->src_data[0], task_info.process_name, + task_info.tgid, task_info.task_name, task_info.pid); dev_err(adev->dev, " VM_CONTEXT1_PROTECTION_FAULT_ADDR 0x%08X\n", addr); dev_err(adev->dev, " VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x%08X\n", @@ -1457,6 +1471,29 @@ static int gmc_v8_0_process_interrupt(struct amdgpu_device *adev, entry->pasid); } + vmid = REG_GET_FIELD(status, VM_CONTEXT1_PROTECTION_FAULT_STATUS, + VMID); + if (amdgpu_amdkfd_is_kfd_vmid(adev, vmid) + && !atomic_read(&adev->gmc.vm_fault_info_updated)) { + struct kfd_vm_fault_info *info = adev->gmc.vm_fault_info; + u32 protections = REG_GET_FIELD(status, + VM_CONTEXT1_PROTECTION_FAULT_STATUS, + PROTECTIONS); + + info->vmid = vmid; + info->mc_id = REG_GET_FIELD(status, + VM_CONTEXT1_PROTECTION_FAULT_STATUS, + MEMORY_CLIENT_ID); + info->status = status; + info->page_addr = addr; + info->prot_valid = protections & 0x7 ? true : false; + info->prot_read = protections & 0x8 ? true : false; + info->prot_write = protections & 0x10 ? true : false; + info->prot_exec = protections & 0x20 ? true : false; + mb(); + atomic_set(&adev->gmc.vm_fault_info_updated, 1); + } + return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c index 3c0a85d4e4ab..399a5db27649 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c @@ -43,6 +43,8 @@ #include "gfxhub_v1_0.h" #include "mmhub_v1_0.h" +#include "ivsrcid/vmc/irqsrcs_vmc_1_0.h" + /* add these here since we already include dce12 headers and these are for DCN */ #define mmHUBP0_DCSURF_PRI_VIEWPORT_DIMENSION 0x055d #define mmHUBP0_DCSURF_PRI_VIEWPORT_DIMENSION_BASE_IDX 2 @@ -257,12 +259,17 @@ static int gmc_v9_0_process_interrupt(struct amdgpu_device *adev, } if (printk_ratelimit()) { + struct amdgpu_task_info task_info = { 0 }; + + amdgpu_vm_get_task_info(adev, entry->pasid, &task_info); + dev_err(adev->dev, - "[%s] VMC page fault (src_id:%u ring:%u vmid:%u pasid:%u)\n", + "[%s] VMC page fault (src_id:%u ring:%u vmid:%u pasid:%u, for process %s pid %d thread %s pid %d\n)\n", entry->vmid_src ? "mmhub" : "gfxhub", entry->src_id, entry->ring_id, entry->vmid, - entry->pasid); - dev_err(adev->dev, " at page 0x%016llx from %d\n", + entry->pasid, task_info.process_name, task_info.tgid, + task_info.task_name, task_info.pid); + dev_err(adev->dev, " at address 0x%016llx from %d\n", addr, entry->client_id); if (!amdgpu_sriov_vf(adev)) dev_err(adev->dev, @@ -872,9 +879,9 @@ static int gmc_v9_0_sw_init(void *handle) } /* This interrupt is VMC page fault.*/ - r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_VMC, 0, + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_VMC, VMC_1_0__SRCID__VM_FAULT, &adev->gmc.vm_fault); - r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_UTCL2, 0, + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_UTCL2, UTCL2_1_0__SRCID__FAULT, &adev->gmc.vm_fault); if (r) diff --git a/drivers/gpu/drm/amd/amdgpu/kv_dpm.c b/drivers/gpu/drm/amd/amdgpu/kv_dpm.c index 7a1e77c93bf1..3f57f6463dc8 100644 --- a/drivers/gpu/drm/amd/amdgpu/kv_dpm.c +++ b/drivers/gpu/drm/amd/amdgpu/kv_dpm.c @@ -1921,7 +1921,7 @@ static int kv_dpm_set_power_state(void *handle) int ret; if (pi->bapm_enable) { - ret = amdgpu_kv_smc_bapm_enable(adev, adev->pm.dpm.ac_power); + ret = amdgpu_kv_smc_bapm_enable(adev, adev->pm.ac_power); if (ret) { DRM_ERROR("amdgpu_kv_smc_bapm_enable failed\n"); return ret; @@ -3306,6 +3306,19 @@ static int kv_dpm_read_sensor(void *handle, int idx, } } +static int kv_set_powergating_by_smu(void *handle, + uint32_t block_type, bool gate) +{ + switch (block_type) { + case AMD_IP_BLOCK_TYPE_UVD: + kv_dpm_powergate_uvd(handle, gate); + break; + default: + break; + } + return 0; +} + static const struct amd_ip_funcs kv_dpm_ip_funcs = { .name = "kv_dpm", .early_init = kv_dpm_early_init, @@ -3342,7 +3355,7 @@ static const struct amd_pm_funcs kv_dpm_funcs = { .print_power_state = &kv_dpm_print_power_state, .debugfs_print_current_performance_level = &kv_dpm_debugfs_print_current_performance_level, .force_performance_level = &kv_dpm_force_performance_level, - .powergate_uvd = &kv_dpm_powergate_uvd, + .set_powergating_by_smu = kv_set_powergating_by_smu, .enable_bapm = &kv_dpm_enable_bapm, .get_vce_clock_state = amdgpu_get_vce_clock_state, .check_state_equal = kv_check_state_equal, diff --git a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c index 3d53c4413f13..e70a0d4d6db4 100644 --- a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c +++ b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c @@ -471,8 +471,8 @@ void mmhub_v1_0_update_power_gating(struct amdgpu_device *adev, RENG_EXECUTE_ON_REG_UPDATE, 1); WREG32_SOC15(MMHUB, 0, mmPCTL1_RENG_EXECUTE, pctl1_reng_execute); - if (adev->powerplay.pp_funcs->set_mmhub_powergating_by_smu) - amdgpu_dpm_set_mmhub_powergating_by_smu(adev); + if (adev->powerplay.pp_funcs->set_powergating_by_smu) + amdgpu_dpm_set_powergating_by_smu(adev, AMD_IP_BLOCK_TYPE_GMC, true); } else { pctl0_reng_execute = REG_SET_FIELD(pctl0_reng_execute, diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c b/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c index c7190c39c4f5..15ae4bc9c072 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c @@ -44,6 +44,8 @@ #include "iceland_sdma_pkt_open.h" +#include "ivsrcid/ivsrcid_vislands30.h" + static void sdma_v2_4_set_ring_funcs(struct amdgpu_device *adev); static void sdma_v2_4_set_buffer_funcs(struct amdgpu_device *adev); static void sdma_v2_4_set_vm_pte_funcs(struct amdgpu_device *adev); @@ -202,8 +204,7 @@ static uint64_t sdma_v2_4_ring_get_rptr(struct amdgpu_ring *ring) static uint64_t sdma_v2_4_ring_get_wptr(struct amdgpu_ring *ring) { struct amdgpu_device *adev = ring->adev; - int me = (ring == &ring->adev->sdma.instance[0].ring) ? 0 : 1; - u32 wptr = RREG32(mmSDMA0_GFX_RB_WPTR + sdma_offsets[me]) >> 2; + u32 wptr = RREG32(mmSDMA0_GFX_RB_WPTR + sdma_offsets[ring->me]) >> 2; return wptr; } @@ -218,9 +219,8 @@ static uint64_t sdma_v2_4_ring_get_wptr(struct amdgpu_ring *ring) static void sdma_v2_4_ring_set_wptr(struct amdgpu_ring *ring) { struct amdgpu_device *adev = ring->adev; - int me = (ring == &ring->adev->sdma.instance[0].ring) ? 0 : 1; - WREG32(mmSDMA0_GFX_RB_WPTR + sdma_offsets[me], lower_32_bits(ring->wptr) << 2); + WREG32(mmSDMA0_GFX_RB_WPTR + sdma_offsets[ring->me], lower_32_bits(ring->wptr) << 2); } static void sdma_v2_4_ring_insert_nop(struct amdgpu_ring *ring, uint32_t count) @@ -273,7 +273,7 @@ static void sdma_v2_4_ring_emit_hdp_flush(struct amdgpu_ring *ring) { u32 ref_and_mask = 0; - if (ring == &ring->adev->sdma.instance[0].ring) + if (ring->me == 0) ref_and_mask = REG_SET_FIELD(ref_and_mask, GPU_HDP_FLUSH_DONE, SDMA0, 1); else ref_and_mask = REG_SET_FIELD(ref_and_mask, GPU_HDP_FLUSH_DONE, SDMA1, 1); @@ -898,7 +898,7 @@ static int sdma_v2_4_sw_init(void *handle) struct amdgpu_device *adev = (struct amdgpu_device *)handle; /* SDMA trap event */ - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 224, + r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_SDMA_TRAP, &adev->sdma.trap_irq); if (r) return r; @@ -910,7 +910,7 @@ static int sdma_v2_4_sw_init(void *handle) return r; /* SDMA Privileged inst */ - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 247, + r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_SDMA_SRBM_WRITE, &adev->sdma.illegal_inst_irq); if (r) return r; @@ -1213,8 +1213,10 @@ static void sdma_v2_4_set_ring_funcs(struct amdgpu_device *adev) { int i; - for (i = 0; i < adev->sdma.num_instances; i++) + for (i = 0; i < adev->sdma.num_instances; i++) { adev->sdma.instance[i].ring.funcs = &sdma_v2_4_ring_funcs; + adev->sdma.instance[i].ring.me = i; + } } static const struct amdgpu_irq_src_funcs sdma_v2_4_trap_irq_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c index aa9ab299fd32..1e07ff274d73 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c @@ -44,6 +44,8 @@ #include "tonga_sdma_pkt_open.h" +#include "ivsrcid/ivsrcid_vislands30.h" + static void sdma_v3_0_set_ring_funcs(struct amdgpu_device *adev); static void sdma_v3_0_set_buffer_funcs(struct amdgpu_device *adev); static void sdma_v3_0_set_vm_pte_funcs(struct amdgpu_device *adev); @@ -365,9 +367,7 @@ static uint64_t sdma_v3_0_ring_get_wptr(struct amdgpu_ring *ring) /* XXX check if swapping is necessary on BE */ wptr = ring->adev->wb.wb[ring->wptr_offs] >> 2; } else { - int me = (ring == &ring->adev->sdma.instance[0].ring) ? 0 : 1; - - wptr = RREG32(mmSDMA0_GFX_RB_WPTR + sdma_offsets[me]) >> 2; + wptr = RREG32(mmSDMA0_GFX_RB_WPTR + sdma_offsets[ring->me]) >> 2; } return wptr; @@ -394,9 +394,7 @@ static void sdma_v3_0_ring_set_wptr(struct amdgpu_ring *ring) WRITE_ONCE(*wb, (lower_32_bits(ring->wptr) << 2)); } else { - int me = (ring == &ring->adev->sdma.instance[0].ring) ? 0 : 1; - - WREG32(mmSDMA0_GFX_RB_WPTR + sdma_offsets[me], lower_32_bits(ring->wptr) << 2); + WREG32(mmSDMA0_GFX_RB_WPTR + sdma_offsets[ring->me], lower_32_bits(ring->wptr) << 2); } } @@ -450,7 +448,7 @@ static void sdma_v3_0_ring_emit_hdp_flush(struct amdgpu_ring *ring) { u32 ref_and_mask = 0; - if (ring == &ring->adev->sdma.instance[0].ring) + if (ring->me == 0) ref_and_mask = REG_SET_FIELD(ref_and_mask, GPU_HDP_FLUSH_DONE, SDMA0, 1); else ref_and_mask = REG_SET_FIELD(ref_and_mask, GPU_HDP_FLUSH_DONE, SDMA1, 1); @@ -1179,7 +1177,7 @@ static int sdma_v3_0_sw_init(void *handle) struct amdgpu_device *adev = (struct amdgpu_device *)handle; /* SDMA trap event */ - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 224, + r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_SDMA_TRAP, &adev->sdma.trap_irq); if (r) return r; @@ -1191,7 +1189,7 @@ static int sdma_v3_0_sw_init(void *handle) return r; /* SDMA Privileged inst */ - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 247, + r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_SDMA_SRBM_WRITE, &adev->sdma.illegal_inst_irq); if (r) return r; @@ -1655,8 +1653,10 @@ static void sdma_v3_0_set_ring_funcs(struct amdgpu_device *adev) { int i; - for (i = 0; i < adev->sdma.num_instances; i++) + for (i = 0; i < adev->sdma.num_instances; i++) { adev->sdma.instance[i].ring.funcs = &sdma_v3_0_ring_funcs; + adev->sdma.instance[i].ring.me = i; + } } static const struct amdgpu_irq_src_funcs sdma_v3_0_trap_irq_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c index ca53b3fba422..e7ca4623cfb9 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c @@ -38,6 +38,9 @@ #include "soc15.h" #include "vega10_sdma_pkt_open.h" +#include "ivsrcid/sdma0/irqsrcs_sdma0_4_0.h" +#include "ivsrcid/sdma1/irqsrcs_sdma1_4_0.h" + MODULE_FIRMWARE("amdgpu/vega10_sdma.bin"); MODULE_FIRMWARE("amdgpu/vega10_sdma1.bin"); MODULE_FIRMWARE("amdgpu/vega12_sdma.bin"); @@ -296,13 +299,12 @@ static uint64_t sdma_v4_0_ring_get_wptr(struct amdgpu_ring *ring) DRM_DEBUG("wptr/doorbell before shift == 0x%016llx\n", wptr); } else { u32 lowbit, highbit; - int me = (ring == &adev->sdma.instance[0].ring) ? 0 : 1; - 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; + lowbit = RREG32(sdma_v4_0_get_reg_offset(adev, ring->me, mmSDMA0_GFX_RB_WPTR)) >> 2; + highbit = RREG32(sdma_v4_0_get_reg_offset(adev, ring->me, mmSDMA0_GFX_RB_WPTR_HI)) >> 2; DRM_DEBUG("wptr [%i]high== 0x%08x low==0x%08x\n", - me, highbit, lowbit); + ring->me, highbit, lowbit); wptr = highbit; wptr = wptr << 32; wptr |= lowbit; @@ -339,17 +341,15 @@ static void sdma_v4_0_ring_set_wptr(struct amdgpu_ring *ring) ring->doorbell_index, ring->wptr << 2); WDOORBELL64(ring->doorbell_index, ring->wptr << 2); } else { - int me = (ring == &ring->adev->sdma.instance[0].ring) ? 0 : 1; - DRM_DEBUG("Not using doorbell -- " "mmSDMA%i_GFX_RB_WPTR == 0x%08x " "mmSDMA%i_GFX_RB_WPTR_HI == 0x%08x\n", - me, + ring->me, lower_32_bits(ring->wptr << 2), - me, + ring->me, 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)); + WREG32(sdma_v4_0_get_reg_offset(adev, ring->me, mmSDMA0_GFX_RB_WPTR), lower_32_bits(ring->wptr << 2)); + WREG32(sdma_v4_0_get_reg_offset(adev, ring->me, mmSDMA0_GFX_RB_WPTR_HI), upper_32_bits(ring->wptr << 2)); } } @@ -430,7 +430,7 @@ static void sdma_v4_0_ring_emit_hdp_flush(struct amdgpu_ring *ring) u32 ref_and_mask = 0; const struct nbio_hdp_flush_reg *nbio_hf_reg = adev->nbio_funcs->hdp_flush_reg; - if (ring == &ring->adev->sdma.instance[0].ring) + if (ring->me == 0) ref_and_mask = nbio_hf_reg->ref_and_mask_sdma0; else ref_and_mask = nbio_hf_reg->ref_and_mask_sdma1; @@ -1228,13 +1228,13 @@ static int sdma_v4_0_sw_init(void *handle) struct amdgpu_device *adev = (struct amdgpu_device *)handle; /* SDMA trap event */ - r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_SDMA0, 224, + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_SDMA0, SDMA0_4_0__SRCID__SDMA_TRAP, &adev->sdma.trap_irq); if (r) return r; /* SDMA trap event */ - r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_SDMA1, 224, + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_SDMA1, SDMA1_4_0__SRCID__SDMA_TRAP, &adev->sdma.trap_irq); if (r) return r; @@ -1651,8 +1651,10 @@ static void sdma_v4_0_set_ring_funcs(struct amdgpu_device *adev) { int i; - for (i = 0; i < adev->sdma.num_instances; i++) + for (i = 0; i < adev->sdma.num_instances; i++) { adev->sdma.instance[i].ring.funcs = &sdma_v4_0_ring_funcs; + adev->sdma.instance[i].ring.me = i; + } } static const struct amdgpu_irq_src_funcs sdma_v4_0_trap_irq_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/si_dpm.c b/drivers/gpu/drm/amd/amdgpu/si_dpm.c index 5c97a3671726..db327b412562 100644 --- a/drivers/gpu/drm/amd/amdgpu/si_dpm.c +++ b/drivers/gpu/drm/amd/amdgpu/si_dpm.c @@ -56,16 +56,16 @@ #define BIOS_SCRATCH_4 0x5cd -MODULE_FIRMWARE("radeon/tahiti_smc.bin"); -MODULE_FIRMWARE("radeon/pitcairn_smc.bin"); -MODULE_FIRMWARE("radeon/pitcairn_k_smc.bin"); -MODULE_FIRMWARE("radeon/verde_smc.bin"); -MODULE_FIRMWARE("radeon/verde_k_smc.bin"); -MODULE_FIRMWARE("radeon/oland_smc.bin"); -MODULE_FIRMWARE("radeon/oland_k_smc.bin"); -MODULE_FIRMWARE("radeon/hainan_smc.bin"); -MODULE_FIRMWARE("radeon/hainan_k_smc.bin"); -MODULE_FIRMWARE("radeon/banks_k_2_smc.bin"); +MODULE_FIRMWARE("amdgpu/tahiti_smc.bin"); +MODULE_FIRMWARE("amdgpu/pitcairn_smc.bin"); +MODULE_FIRMWARE("amdgpu/pitcairn_k_smc.bin"); +MODULE_FIRMWARE("amdgpu/verde_smc.bin"); +MODULE_FIRMWARE("amdgpu/verde_k_smc.bin"); +MODULE_FIRMWARE("amdgpu/oland_smc.bin"); +MODULE_FIRMWARE("amdgpu/oland_k_smc.bin"); +MODULE_FIRMWARE("amdgpu/hainan_smc.bin"); +MODULE_FIRMWARE("amdgpu/hainan_k_smc.bin"); +MODULE_FIRMWARE("amdgpu/banks_k_2_smc.bin"); static const struct amd_pm_funcs si_dpm_funcs; @@ -3480,7 +3480,7 @@ static void si_apply_state_adjust_rules(struct amdgpu_device *adev, disable_sclk_switching = true; } - if (adev->pm.dpm.ac_power) + if (adev->pm.ac_power) max_limits = &adev->pm.dpm.dyn_state.max_clock_voltage_on_ac; else max_limits = &adev->pm.dpm.dyn_state.max_clock_voltage_on_dc; @@ -3489,7 +3489,7 @@ static void si_apply_state_adjust_rules(struct amdgpu_device *adev, if (ps->performance_levels[i].vddc > ps->performance_levels[i+1].vddc) ps->performance_levels[i].vddc = ps->performance_levels[i+1].vddc; } - if (adev->pm.dpm.ac_power == false) { + if (adev->pm.ac_power == false) { for (i = 0; i < ps->performance_level_count; i++) { if (ps->performance_levels[i].mclk > max_limits->mclk) ps->performance_levels[i].mclk = max_limits->mclk; @@ -7318,8 +7318,7 @@ static int si_dpm_init(struct amdgpu_device *adev) pi = &eg_pi->rv7xx; si_pi->sys_pcie_mask = - (adev->pm.pcie_gen_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_MASK) >> - CAIL_PCIE_LINK_SPEED_SUPPORT_SHIFT; + adev->pm.pcie_gen_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_MASK; si_pi->force_pcie_gen = AMDGPU_PCIE_GEN_INVALID; si_pi->boot_pcie_gen = si_get_current_pcie_speed(adev); @@ -7667,7 +7666,7 @@ static int si_dpm_init_microcode(struct amdgpu_device *adev) default: BUG(); } - snprintf(fw_name, sizeof(fw_name), "radeon/%s_smc.bin", chip_name); + snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_smc.bin", chip_name); err = request_firmware(&adev->pm.fw, fw_name, adev->dev); if (err) goto out; diff --git a/drivers/gpu/drm/amd/amdgpu/soc15d.h b/drivers/gpu/drm/amd/amdgpu/soc15d.h index 8dc29107228f..edfe50821cd9 100644 --- a/drivers/gpu/drm/amd/amdgpu/soc15d.h +++ b/drivers/gpu/drm/amd/amdgpu/soc15d.h @@ -53,6 +53,29 @@ #define PACKET3_COMPUTE(op, n) (PACKET3(op, n) | 1 << 1) +#define PACKETJ_CONDITION_CHECK0 0 +#define PACKETJ_CONDITION_CHECK1 1 +#define PACKETJ_CONDITION_CHECK2 2 +#define PACKETJ_CONDITION_CHECK3 3 +#define PACKETJ_CONDITION_CHECK4 4 +#define PACKETJ_CONDITION_CHECK5 5 +#define PACKETJ_CONDITION_CHECK6 6 +#define PACKETJ_CONDITION_CHECK7 7 + +#define PACKETJ_TYPE0 0 +#define PACKETJ_TYPE1 1 +#define PACKETJ_TYPE2 2 +#define PACKETJ_TYPE3 3 +#define PACKETJ_TYPE4 4 +#define PACKETJ_TYPE5 5 +#define PACKETJ_TYPE6 6 +#define PACKETJ_TYPE7 7 + +#define PACKETJ(reg, r, cond, type) ((reg & 0x3FFFF) | \ + ((r & 0x3F) << 18) | \ + ((cond & 0xF) << 24) | \ + ((type & 0xF) << 28)) + /* Packet 3 types */ #define PACKET3_NOP 0x10 #define PACKET3_SET_BASE 0x11 diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c index 341ee6d55ce8..aeaa1ca46a99 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c @@ -35,6 +35,7 @@ #include "vi.h" #include "smu/smu_7_1_2_d.h" #include "smu/smu_7_1_2_sh_mask.h" +#include "ivsrcid/ivsrcid_vislands30.h" static void uvd_v5_0_set_ring_funcs(struct amdgpu_device *adev); static void uvd_v5_0_set_irq_funcs(struct amdgpu_device *adev); @@ -104,7 +105,7 @@ static int uvd_v5_0_sw_init(void *handle) int r; /* UVD TRAP */ - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 124, &adev->uvd.inst->irq); + r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_UVD_SYSTEM_MESSAGE, &adev->uvd.inst->irq); if (r) return r; diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c index bfddf97dd13e..598dbeaba636 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c @@ -36,6 +36,7 @@ #include "bif/bif_5_1_d.h" #include "gmc/gmc_8_1_d.h" #include "vi.h" +#include "ivsrcid/ivsrcid_vislands30.h" /* Polaris10/11/12 firmware version */ #define FW_1_130_16 ((1 << 24) | (130 << 16) | (16 << 8)) @@ -247,12 +248,10 @@ static int uvd_v6_0_enc_get_create_msg(struct amdgpu_ring *ring, uint32_t handle for (i = ib->length_dw; i < ib_size_dw; ++i) ib->ptr[i] = 0x0; - r = amdgpu_ib_schedule(ring, 1, ib, NULL, &f); - job->fence = dma_fence_get(f); + r = amdgpu_job_submit_direct(job, ring, &f); if (r) goto err; - amdgpu_job_free(job); if (fence) *fence = dma_fence_get(f); dma_fence_put(f); @@ -311,19 +310,13 @@ static int uvd_v6_0_enc_get_destroy_msg(struct amdgpu_ring *ring, for (i = ib->length_dw; i < ib_size_dw; ++i) ib->ptr[i] = 0x0; - if (direct) { - r = amdgpu_ib_schedule(ring, 1, ib, NULL, &f); - job->fence = dma_fence_get(f); - if (r) - goto err; - - amdgpu_job_free(job); - } else { - r = amdgpu_job_submit(job, ring, &ring->adev->vce.entity, + if (direct) + r = amdgpu_job_submit_direct(job, ring, &f); + else + r = amdgpu_job_submit(job, &ring->adev->vce.entity, AMDGPU_FENCE_OWNER_UNDEFINED, &f); - if (r) - goto err; - } + if (r) + goto err; if (fence) *fence = dma_fence_get(f); @@ -400,14 +393,14 @@ static int uvd_v6_0_sw_init(void *handle) struct amdgpu_device *adev = (struct amdgpu_device *)handle; /* UVD TRAP */ - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 124, &adev->uvd.inst->irq); + r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_UVD_SYSTEM_MESSAGE, &adev->uvd.inst->irq); if (r) return r; /* UVD ENC TRAP */ if (uvd_v6_0_enc_support(adev)) { for (i = 0; i < adev->uvd.num_enc_rings; ++i) { - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, i + 119, &adev->uvd.inst->irq); + r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, i + VISLANDS30_IV_SRCID_UVD_ENC_GEN_PURP, &adev->uvd.inst->irq); if (r) return r; } @@ -425,16 +418,6 @@ static int uvd_v6_0_sw_init(void *handle) adev->uvd.num_enc_rings = 0; DRM_INFO("UVD ENC is disabled\n"); - } else { - struct drm_sched_rq *rq; - ring = &adev->uvd.inst->ring_enc[0]; - rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL]; - r = drm_sched_entity_init(&ring->sched, &adev->uvd.inst->entity_enc, - rq, NULL); - if (r) { - DRM_ERROR("Failed setting up UVD ENC run queue.\n"); - return r; - } } r = amdgpu_uvd_resume(adev); @@ -470,8 +453,6 @@ static int uvd_v6_0_sw_fini(void *handle) return r; if (uvd_v6_0_enc_support(adev)) { - drm_sched_entity_fini(&adev->uvd.inst->ring_enc[0].sched, &adev->uvd.inst->entity_enc); - for (i = 0; i < adev->uvd.num_enc_rings; ++i) amdgpu_ring_fini(&adev->uvd.inst->ring_enc[i]); } @@ -1569,7 +1550,6 @@ static const struct amdgpu_ring_funcs uvd_v6_0_ring_phys_funcs = { static const struct amdgpu_ring_funcs uvd_v6_0_ring_vm_funcs = { .type = AMDGPU_RING_TYPE_UVD, .align_mask = 0xf, - .nop = PACKET0(mmUVD_NO_OP, 0), .support_64bit_ptrs = false, .get_rptr = uvd_v6_0_ring_get_rptr, .get_wptr = uvd_v6_0_ring_get_wptr, @@ -1587,7 +1567,7 @@ static const struct amdgpu_ring_funcs uvd_v6_0_ring_vm_funcs = { .emit_hdp_flush = uvd_v6_0_ring_emit_hdp_flush, .test_ring = uvd_v6_0_ring_test_ring, .test_ib = amdgpu_uvd_ring_test_ib, - .insert_nop = amdgpu_ring_insert_nop, + .insert_nop = uvd_v6_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, diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c index 57d32f21b3a6..5fab3560a71d 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c @@ -39,6 +39,13 @@ #include "hdp/hdp_4_0_offset.h" #include "mmhub/mmhub_1_0_offset.h" #include "mmhub/mmhub_1_0_sh_mask.h" +#include "ivsrcid/uvd/irqsrcs_uvd_7_0.h" + +#define mmUVD_PG0_CC_UVD_HARVESTING 0x00c7 +#define mmUVD_PG0_CC_UVD_HARVESTING_BASE_IDX 1 +//UVD_PG0_CC_UVD_HARVESTING +#define UVD_PG0_CC_UVD_HARVESTING__UVD_DISABLE__SHIFT 0x1 +#define UVD_PG0_CC_UVD_HARVESTING__UVD_DISABLE_MASK 0x00000002L #define UVD7_MAX_HW_INSTANCES_VEGA20 2 @@ -249,12 +256,10 @@ static int uvd_v7_0_enc_get_create_msg(struct amdgpu_ring *ring, uint32_t handle for (i = ib->length_dw; i < ib_size_dw; ++i) ib->ptr[i] = 0x0; - r = amdgpu_ib_schedule(ring, 1, ib, NULL, &f); - job->fence = dma_fence_get(f); + r = amdgpu_job_submit_direct(job, ring, &f); if (r) goto err; - amdgpu_job_free(job); if (fence) *fence = dma_fence_get(f); dma_fence_put(f); @@ -312,19 +317,13 @@ int uvd_v7_0_enc_get_destroy_msg(struct amdgpu_ring *ring, uint32_t handle, for (i = ib->length_dw; i < ib_size_dw; ++i) ib->ptr[i] = 0x0; - if (direct) { - r = amdgpu_ib_schedule(ring, 1, ib, NULL, &f); - job->fence = dma_fence_get(f); - if (r) - goto err; - - amdgpu_job_free(job); - } else { - r = amdgpu_job_submit(job, ring, &ring->adev->vce.entity, + if (direct) + r = amdgpu_job_submit_direct(job, ring, &f); + else + r = amdgpu_job_submit(job, &ring->adev->vce.entity, AMDGPU_FENCE_OWNER_UNDEFINED, &f); - if (r) - goto err; - } + if (r) + goto err; if (fence) *fence = dma_fence_get(f); @@ -377,10 +376,25 @@ error: static int uvd_v7_0_early_init(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; - if (adev->asic_type == CHIP_VEGA20) + + if (adev->asic_type == CHIP_VEGA20) { + u32 harvest; + int i; + adev->uvd.num_uvd_inst = UVD7_MAX_HW_INSTANCES_VEGA20; - else + for (i = 0; i < adev->uvd.num_uvd_inst; i++) { + harvest = RREG32_SOC15(UVD, i, mmUVD_PG0_CC_UVD_HARVESTING); + if (harvest & UVD_PG0_CC_UVD_HARVESTING__UVD_DISABLE_MASK) { + adev->uvd.harvest_config |= 1 << i; + } + } + if (adev->uvd.harvest_config == (AMDGPU_UVD_HARVEST_UVD0 | + AMDGPU_UVD_HARVEST_UVD1)) + /* both instances are harvested, disable the block */ + return -ENOENT; + } else { adev->uvd.num_uvd_inst = 1; + } if (amdgpu_sriov_vf(adev)) adev->uvd.num_enc_rings = 1; @@ -396,19 +410,20 @@ static int uvd_v7_0_early_init(void *handle) static int uvd_v7_0_sw_init(void *handle) { struct amdgpu_ring *ring; - struct drm_sched_rq *rq; int i, j, r; struct amdgpu_device *adev = (struct amdgpu_device *)handle; for (j = 0; j < adev->uvd.num_uvd_inst; j++) { + if (adev->uvd.harvest_config & (1 << j)) + continue; /* UVD TRAP */ - r = amdgpu_irq_add_id(adev, amdgpu_ih_clientid_uvds[j], 124, &adev->uvd.inst[j].irq); + r = amdgpu_irq_add_id(adev, amdgpu_ih_clientid_uvds[j], UVD_7_0__SRCID__UVD_SYSTEM_MESSAGE_INTERRUPT, &adev->uvd.inst[j].irq); if (r) return r; /* UVD ENC TRAP */ for (i = 0; i < adev->uvd.num_enc_rings; ++i) { - r = amdgpu_irq_add_id(adev, amdgpu_ih_clientid_uvds[j], i + 119, &adev->uvd.inst[j].irq); + r = amdgpu_irq_add_id(adev, amdgpu_ih_clientid_uvds[j], i + UVD_7_0__SRCID__UVD_ENC_GEN_PURP, &adev->uvd.inst[j].irq); if (r) return r; } @@ -428,22 +443,13 @@ static int uvd_v7_0_sw_init(void *handle) DRM_INFO("PSP loading UVD firmware\n"); } - for (j = 0; j < adev->uvd.num_uvd_inst; j++) { - ring = &adev->uvd.inst[j].ring_enc[0]; - rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL]; - r = drm_sched_entity_init(&ring->sched, &adev->uvd.inst[j].entity_enc, - rq, NULL); - if (r) { - DRM_ERROR("(%d)Failed setting up UVD ENC run queue.\n", j); - return r; - } - } - r = amdgpu_uvd_resume(adev); if (r) return r; for (j = 0; j < adev->uvd.num_uvd_inst; j++) { + if (adev->uvd.harvest_config & (1 << j)) + continue; if (!amdgpu_sriov_vf(adev)) { ring = &adev->uvd.inst[j].ring; sprintf(ring->name, "uvd<%d>", j); @@ -491,8 +497,8 @@ static int uvd_v7_0_sw_fini(void *handle) return r; for (j = 0; j < adev->uvd.num_uvd_inst; ++j) { - drm_sched_entity_fini(&adev->uvd.inst[j].ring_enc[0].sched, &adev->uvd.inst[j].entity_enc); - + if (adev->uvd.harvest_config & (1 << j)) + continue; for (i = 0; i < adev->uvd.num_enc_rings; ++i) amdgpu_ring_fini(&adev->uvd.inst[j].ring_enc[i]); } @@ -521,6 +527,8 @@ static int uvd_v7_0_hw_init(void *handle) goto done; for (j = 0; j < adev->uvd.num_uvd_inst; ++j) { + if (adev->uvd.harvest_config & (1 << j)) + continue; ring = &adev->uvd.inst[j].ring; if (!amdgpu_sriov_vf(adev)) { @@ -600,8 +608,11 @@ static int uvd_v7_0_hw_fini(void *handle) DRM_DEBUG("For SRIOV client, shouldn't do anything.\n"); } - for (i = 0; i < adev->uvd.num_uvd_inst; ++i) + for (i = 0; i < adev->uvd.num_uvd_inst; ++i) { + if (adev->uvd.harvest_config & (1 << i)) + continue; adev->uvd.inst[i].ring.ready = false; + } return 0; } @@ -644,6 +655,8 @@ static void uvd_v7_0_mc_resume(struct amdgpu_device *adev) int i; for (i = 0; i < adev->uvd.num_uvd_inst; ++i) { + if (adev->uvd.harvest_config & (1 << i)) + continue; if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) { WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW, lower_32_bits(adev->firmware.ucode[AMDGPU_UCODE_ID_UVD].mc_addr)); @@ -716,6 +729,8 @@ static int uvd_v7_0_mmsch_start(struct amdgpu_device *adev, WREG32_SOC15(VCE, 0, mmVCE_MMSCH_VF_MAILBOX_RESP, 0); for (i = 0; i < adev->uvd.num_uvd_inst; ++i) { + if (adev->uvd.harvest_config & (1 << i)) + continue; WDOORBELL32(adev->uvd.inst[i].ring_enc[0].doorbell_index, 0); adev->wb.wb[adev->uvd.inst[i].ring_enc[0].wptr_offs] = 0; adev->uvd.inst[i].ring_enc[0].wptr = 0; @@ -772,6 +787,8 @@ static int uvd_v7_0_sriov_start(struct amdgpu_device *adev) init_table += header->uvd_table_offset; for (i = 0; i < adev->uvd.num_uvd_inst; ++i) { + if (adev->uvd.harvest_config & (1 << i)) + continue; ring = &adev->uvd.inst[i].ring; ring->wptr = 0; size = AMDGPU_GPU_PAGE_ALIGN(adev->uvd.fw->size + 4); @@ -911,6 +928,8 @@ static int uvd_v7_0_start(struct amdgpu_device *adev) int i, j, k, r; for (k = 0; k < adev->uvd.num_uvd_inst; ++k) { + if (adev->uvd.harvest_config & (1 << k)) + continue; /* disable DPG */ WREG32_P(SOC15_REG_OFFSET(UVD, k, mmUVD_POWER_STATUS), 0, ~UVD_POWER_STATUS__UVD_PG_MODE_MASK); @@ -923,6 +942,8 @@ static int uvd_v7_0_start(struct amdgpu_device *adev) uvd_v7_0_mc_resume(adev); for (k = 0; k < adev->uvd.num_uvd_inst; ++k) { + if (adev->uvd.harvest_config & (1 << k)) + continue; ring = &adev->uvd.inst[k].ring; /* disable clock gating */ WREG32_P(SOC15_REG_OFFSET(UVD, k, mmUVD_CGC_CTRL), 0, @@ -1090,6 +1111,8 @@ static void uvd_v7_0_stop(struct amdgpu_device *adev) uint8_t i = 0; for (i = 0; i < adev->uvd.num_uvd_inst; ++i) { + if (adev->uvd.harvest_config & (1 << i)) + continue; /* force RBC into idle state */ WREG32_SOC15(UVD, i, mmUVD_RBC_RB_CNTL, 0x11010101); @@ -1227,6 +1250,34 @@ static int uvd_v7_0_ring_test_ring(struct amdgpu_ring *ring) } /** + * uvd_v7_0_ring_patch_cs_in_place - Patch the IB for command submission. + * + * @p: the CS parser with the IBs + * @ib_idx: which IB to patch + * + */ +static int uvd_v7_0_ring_patch_cs_in_place(struct amdgpu_cs_parser *p, + uint32_t ib_idx) +{ + struct amdgpu_ib *ib = &p->job->ibs[ib_idx]; + unsigned i; + + /* No patching necessary for the first instance */ + if (!p->ring->me) + return 0; + + for (i = 0; i < ib->length_dw; i += 2) { + uint32_t reg = amdgpu_get_ib_value(p, ib_idx, i); + + reg -= p->adev->reg_offset[UVD_HWIP][0][1]; + reg += p->adev->reg_offset[UVD_HWIP][1][1]; + + amdgpu_set_ib_value(p, ib_idx, i, reg); + } + return 0; +} + +/** * uvd_v7_0_ring_emit_ib - execute indirect buffer * * @ring: amdgpu_ring pointer @@ -1718,6 +1769,7 @@ static const struct amdgpu_ring_funcs uvd_v7_0_ring_vm_funcs = { .get_rptr = uvd_v7_0_ring_get_rptr, .get_wptr = uvd_v7_0_ring_get_wptr, .set_wptr = uvd_v7_0_ring_set_wptr, + .patch_cs_in_place = uvd_v7_0_ring_patch_cs_in_place, .emit_frame_size = 6 + /* hdp invalidate */ SOC15_FLUSH_GPU_TLB_NUM_WREG * 6 + @@ -1777,6 +1829,8 @@ static void uvd_v7_0_set_ring_funcs(struct amdgpu_device *adev) int i; for (i = 0; i < adev->uvd.num_uvd_inst; i++) { + if (adev->uvd.harvest_config & (1 << i)) + continue; adev->uvd.inst[i].ring.funcs = &uvd_v7_0_ring_vm_funcs; adev->uvd.inst[i].ring.me = i; DRM_INFO("UVD(%d) is enabled in VM mode\n", i); @@ -1788,6 +1842,8 @@ static void uvd_v7_0_set_enc_ring_funcs(struct amdgpu_device *adev) int i, j; for (j = 0; j < adev->uvd.num_uvd_inst; j++) { + if (adev->uvd.harvest_config & (1 << j)) + continue; for (i = 0; i < adev->uvd.num_enc_rings; ++i) { adev->uvd.inst[j].ring_enc[i].funcs = &uvd_v7_0_enc_ring_vm_funcs; adev->uvd.inst[j].ring_enc[i].me = j; @@ -1807,6 +1863,8 @@ static void uvd_v7_0_set_irq_funcs(struct amdgpu_device *adev) int i; for (i = 0; i < adev->uvd.num_uvd_inst; i++) { + if (adev->uvd.harvest_config & (1 << i)) + continue; adev->uvd.inst[i].irq.num_types = adev->uvd.num_enc_rings + 1; adev->uvd.inst[i].irq.funcs = &uvd_v7_0_irq_funcs; } diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c index 47f70827195b..d48e877b682e 100644 --- a/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c @@ -56,7 +56,7 @@ static uint64_t vce_v2_0_ring_get_rptr(struct amdgpu_ring *ring) { struct amdgpu_device *adev = ring->adev; - if (ring == &adev->vce.ring[0]) + if (ring->me == 0) return RREG32(mmVCE_RB_RPTR); else return RREG32(mmVCE_RB_RPTR2); @@ -73,7 +73,7 @@ static uint64_t vce_v2_0_ring_get_wptr(struct amdgpu_ring *ring) { struct amdgpu_device *adev = ring->adev; - if (ring == &adev->vce.ring[0]) + if (ring->me == 0) return RREG32(mmVCE_RB_WPTR); else return RREG32(mmVCE_RB_WPTR2); @@ -90,7 +90,7 @@ static void vce_v2_0_ring_set_wptr(struct amdgpu_ring *ring) { struct amdgpu_device *adev = ring->adev; - if (ring == &adev->vce.ring[0]) + if (ring->me == 0) WREG32(mmVCE_RB_WPTR, lower_32_bits(ring->wptr)); else WREG32(mmVCE_RB_WPTR2, lower_32_bits(ring->wptr)); @@ -627,8 +627,10 @@ static void vce_v2_0_set_ring_funcs(struct amdgpu_device *adev) { int i; - for (i = 0; i < adev->vce.num_rings; i++) + for (i = 0; i < adev->vce.num_rings; i++) { adev->vce.ring[i].funcs = &vce_v2_0_ring_funcs; + adev->vce.ring[i].me = i; + } } static const struct amdgpu_irq_src_funcs vce_v2_0_irq_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c index a71b97519cc0..cc6ce6cc03f4 100644 --- a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c @@ -39,6 +39,7 @@ #include "smu/smu_7_1_2_sh_mask.h" #include "gca/gfx_8_0_d.h" #include "gca/gfx_8_0_sh_mask.h" +#include "ivsrcid/ivsrcid_vislands30.h" #define GRBM_GFX_INDEX__VCE_INSTANCE__SHIFT 0x04 @@ -86,9 +87,9 @@ static uint64_t vce_v3_0_ring_get_rptr(struct amdgpu_ring *ring) else if (adev->vce.harvest_config == AMDGPU_VCE_HARVEST_VCE0) WREG32(mmGRBM_GFX_INDEX, GET_VCE_INSTANCE(1)); - if (ring == &adev->vce.ring[0]) + if (ring->me == 0) v = RREG32(mmVCE_RB_RPTR); - else if (ring == &adev->vce.ring[1]) + else if (ring->me == 1) v = RREG32(mmVCE_RB_RPTR2); else v = RREG32(mmVCE_RB_RPTR3); @@ -118,9 +119,9 @@ static uint64_t vce_v3_0_ring_get_wptr(struct amdgpu_ring *ring) else if (adev->vce.harvest_config == AMDGPU_VCE_HARVEST_VCE0) WREG32(mmGRBM_GFX_INDEX, GET_VCE_INSTANCE(1)); - if (ring == &adev->vce.ring[0]) + if (ring->me == 0) v = RREG32(mmVCE_RB_WPTR); - else if (ring == &adev->vce.ring[1]) + else if (ring->me == 1) v = RREG32(mmVCE_RB_WPTR2); else v = RREG32(mmVCE_RB_WPTR3); @@ -149,9 +150,9 @@ static void vce_v3_0_ring_set_wptr(struct amdgpu_ring *ring) else if (adev->vce.harvest_config == AMDGPU_VCE_HARVEST_VCE0) WREG32(mmGRBM_GFX_INDEX, GET_VCE_INSTANCE(1)); - if (ring == &adev->vce.ring[0]) + if (ring->me == 0) WREG32(mmVCE_RB_WPTR, lower_32_bits(ring->wptr)); - else if (ring == &adev->vce.ring[1]) + else if (ring->me == 1) WREG32(mmVCE_RB_WPTR2, lower_32_bits(ring->wptr)); else WREG32(mmVCE_RB_WPTR3, lower_32_bits(ring->wptr)); @@ -422,7 +423,7 @@ static int vce_v3_0_sw_init(void *handle) int r, i; /* VCE */ - r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 167, &adev->vce.irq); + r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_VCE_TRAP, &adev->vce.irq); if (r) return r; @@ -942,12 +943,16 @@ static void vce_v3_0_set_ring_funcs(struct amdgpu_device *adev) int i; if (adev->asic_type >= CHIP_STONEY) { - for (i = 0; i < adev->vce.num_rings; i++) + for (i = 0; i < adev->vce.num_rings; i++) { adev->vce.ring[i].funcs = &vce_v3_0_ring_vm_funcs; + adev->vce.ring[i].me = i; + } DRM_INFO("VCE enabled in VM mode\n"); } else { - for (i = 0; i < adev->vce.num_rings; i++) + for (i = 0; i < adev->vce.num_rings; i++) { adev->vce.ring[i].funcs = &vce_v3_0_ring_phys_funcs; + adev->vce.ring[i].me = i; + } DRM_INFO("VCE enabled in physical mode\n"); } } diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c index 8fd1b742985a..65f8860169e9 100644 --- a/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c @@ -39,6 +39,8 @@ #include "mmhub/mmhub_1_0_offset.h" #include "mmhub/mmhub_1_0_sh_mask.h" +#include "ivsrcid/vce/irqsrcs_vce_4_0.h" + #define VCE_STATUS_VCPU_REPORT_FW_LOADED_MASK 0x02 #define VCE_V4_0_FW_SIZE (384 * 1024) @@ -60,9 +62,9 @@ static uint64_t vce_v4_0_ring_get_rptr(struct amdgpu_ring *ring) { struct amdgpu_device *adev = ring->adev; - if (ring == &adev->vce.ring[0]) + if (ring->me == 0) return RREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_RB_RPTR)); - else if (ring == &adev->vce.ring[1]) + else if (ring->me == 1) return RREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_RB_RPTR2)); else return RREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_RB_RPTR3)); @@ -82,9 +84,9 @@ static uint64_t vce_v4_0_ring_get_wptr(struct amdgpu_ring *ring) if (ring->use_doorbell) return adev->wb.wb[ring->wptr_offs]; - if (ring == &adev->vce.ring[0]) + if (ring->me == 0) return RREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_RB_WPTR)); - else if (ring == &adev->vce.ring[1]) + else if (ring->me == 1) return RREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_RB_WPTR2)); else return RREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_RB_WPTR3)); @@ -108,10 +110,10 @@ static void vce_v4_0_ring_set_wptr(struct amdgpu_ring *ring) return; } - if (ring == &adev->vce.ring[0]) + if (ring->me == 0) WREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_RB_WPTR), lower_32_bits(ring->wptr)); - else if (ring == &adev->vce.ring[1]) + else if (ring->me == 1) WREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_RB_WPTR2), lower_32_bits(ring->wptr)); else @@ -1088,8 +1090,10 @@ static void vce_v4_0_set_ring_funcs(struct amdgpu_device *adev) { int i; - for (i = 0; i < adev->vce.num_rings; i++) + for (i = 0; i < adev->vce.num_rings; i++) { adev->vce.ring[i].funcs = &vce_v4_0_ring_vm_funcs; + adev->vce.ring[i].me = i; + } DRM_INFO("VCE enabled in VM mode\n"); } diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c index 29684c3ea4ef..2ce91a748c40 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c @@ -35,10 +35,14 @@ #include "mmhub/mmhub_9_1_offset.h" #include "mmhub/mmhub_9_1_sh_mask.h" +#include "ivsrcid/vcn/irqsrcs_vcn_1_0.h" + static int vcn_v1_0_stop(struct amdgpu_device *adev); static void vcn_v1_0_set_dec_ring_funcs(struct amdgpu_device *adev); static void vcn_v1_0_set_enc_ring_funcs(struct amdgpu_device *adev); +static void vcn_v1_0_set_jpeg_ring_funcs(struct amdgpu_device *adev); static void vcn_v1_0_set_irq_funcs(struct amdgpu_device *adev); +static void vcn_v1_0_jpeg_ring_set_patch_ring(struct amdgpu_ring *ring, uint32_t ptr); /** * vcn_v1_0_early_init - set function pointers @@ -55,6 +59,7 @@ static int vcn_v1_0_early_init(void *handle) vcn_v1_0_set_dec_ring_funcs(adev); vcn_v1_0_set_enc_ring_funcs(adev); + vcn_v1_0_set_jpeg_ring_funcs(adev); vcn_v1_0_set_irq_funcs(adev); return 0; @@ -74,18 +79,23 @@ static int vcn_v1_0_sw_init(void *handle) struct amdgpu_device *adev = (struct amdgpu_device *)handle; /* VCN DEC TRAP */ - r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_VCN, 124, &adev->vcn.irq); + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_VCN, VCN_1_0__SRCID__UVD_SYSTEM_MESSAGE_INTERRUPT, &adev->vcn.irq); if (r) return r; /* VCN ENC TRAP */ for (i = 0; i < adev->vcn.num_enc_rings; ++i) { - r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_VCN, i + 119, + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_VCN, i + VCN_1_0__SRCID__UVD_ENC_GENERAL_PURPOSE, &adev->vcn.irq); if (r) return r; } + /* VCN JPEG TRAP */ + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_VCN, 126, &adev->vcn.irq); + if (r) + return r; + r = amdgpu_vcn_sw_init(adev); if (r) return r; @@ -108,6 +118,12 @@ static int vcn_v1_0_sw_init(void *handle) return r; } + ring = &adev->vcn.ring_jpeg; + sprintf(ring->name, "vcn_jpeg"); + r = amdgpu_ring_init(adev, ring, 512, &adev->vcn.irq, 0); + if (r) + return r; + return r; } @@ -162,6 +178,14 @@ static int vcn_v1_0_hw_init(void *handle) } } + ring = &adev->vcn.ring_jpeg; + ring->ready = true; + r = amdgpu_ring_test_ring(ring); + if (r) { + ring->ready = false; + goto done; + } + done: if (!r) DRM_INFO("VCN decode and encode initialized successfully.\n"); @@ -578,12 +602,12 @@ static int vcn_v1_0_start(struct amdgpu_device *adev) /* disable byte swapping */ lmi_swap_cntl = 0; - vcn_v1_0_mc_resume(adev); - vcn_1_0_disable_static_power_gating(adev); /* disable clock gating */ vcn_v1_0_disable_clock_gating(adev); + vcn_v1_0_mc_resume(adev); + /* disable interupt */ WREG32_P(SOC15_REG_OFFSET(UVD, 0, mmUVD_MASTINT_EN), 0, ~UVD_MASTINT_EN__VCPU_EN_MASK); @@ -729,6 +753,22 @@ static int vcn_v1_0_start(struct amdgpu_device *adev) WREG32_SOC15(UVD, 0, mmUVD_RB_BASE_HI2, upper_32_bits(ring->gpu_addr)); WREG32_SOC15(UVD, 0, mmUVD_RB_SIZE2, ring->ring_size / 4); + ring = &adev->vcn.ring_jpeg; + WREG32_SOC15(UVD, 0, mmUVD_LMI_JRBC_RB_VMID, 0); + WREG32_SOC15(UVD, 0, mmUVD_JRBC_RB_CNTL, (0x00000001L | 0x00000002L)); + WREG32_SOC15(UVD, 0, mmUVD_LMI_JRBC_RB_64BIT_BAR_LOW, lower_32_bits(ring->gpu_addr)); + WREG32_SOC15(UVD, 0, mmUVD_LMI_JRBC_RB_64BIT_BAR_HIGH, upper_32_bits(ring->gpu_addr)); + WREG32_SOC15(UVD, 0, mmUVD_JRBC_RB_RPTR, 0); + WREG32_SOC15(UVD, 0, mmUVD_JRBC_RB_WPTR, 0); + WREG32_SOC15(UVD, 0, mmUVD_JRBC_RB_CNTL, 0x00000002L); + + /* initialize wptr */ + ring->wptr = RREG32_SOC15(UVD, 0, mmUVD_JRBC_RB_WPTR); + + /* copy patch commands to the jpeg ring */ + vcn_v1_0_jpeg_ring_set_patch_ring(ring, + (ring->wptr + ring->max_dw * amdgpu_sched_hw_submission)); + return 0; } @@ -1126,6 +1166,383 @@ static void vcn_v1_0_enc_ring_emit_wreg(struct amdgpu_ring *ring, amdgpu_ring_write(ring, val); } + +/** + * vcn_v1_0_jpeg_ring_get_rptr - get read pointer + * + * @ring: amdgpu_ring pointer + * + * Returns the current hardware read pointer + */ +static uint64_t vcn_v1_0_jpeg_ring_get_rptr(struct amdgpu_ring *ring) +{ + struct amdgpu_device *adev = ring->adev; + + return RREG32_SOC15(UVD, 0, mmUVD_JRBC_RB_RPTR); +} + +/** + * vcn_v1_0_jpeg_ring_get_wptr - get write pointer + * + * @ring: amdgpu_ring pointer + * + * Returns the current hardware write pointer + */ +static uint64_t vcn_v1_0_jpeg_ring_get_wptr(struct amdgpu_ring *ring) +{ + struct amdgpu_device *adev = ring->adev; + + return RREG32_SOC15(UVD, 0, mmUVD_JRBC_RB_WPTR); +} + +/** + * vcn_v1_0_jpeg_ring_set_wptr - set write pointer + * + * @ring: amdgpu_ring pointer + * + * Commits the write pointer to the hardware + */ +static void vcn_v1_0_jpeg_ring_set_wptr(struct amdgpu_ring *ring) +{ + struct amdgpu_device *adev = ring->adev; + + WREG32_SOC15(UVD, 0, mmUVD_JRBC_RB_WPTR, lower_32_bits(ring->wptr)); +} + +/** + * vcn_v1_0_jpeg_ring_insert_start - insert a start command + * + * @ring: amdgpu_ring pointer + * + * Write a start command to the ring. + */ +static void vcn_v1_0_jpeg_ring_insert_start(struct amdgpu_ring *ring) +{ + struct amdgpu_device *adev = ring->adev; + + amdgpu_ring_write(ring, + PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_EXTERNAL_REG_BASE), 0, 0, PACKETJ_TYPE0)); + amdgpu_ring_write(ring, 0x68e04); + + amdgpu_ring_write(ring, PACKETJ(0, 0, 0, PACKETJ_TYPE0)); + amdgpu_ring_write(ring, 0x80010000); +} + +/** + * vcn_v1_0_jpeg_ring_insert_end - insert a end command + * + * @ring: amdgpu_ring pointer + * + * Write a end command to the ring. + */ +static void vcn_v1_0_jpeg_ring_insert_end(struct amdgpu_ring *ring) +{ + struct amdgpu_device *adev = ring->adev; + + amdgpu_ring_write(ring, + PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_EXTERNAL_REG_BASE), 0, 0, PACKETJ_TYPE0)); + amdgpu_ring_write(ring, 0x68e04); + + amdgpu_ring_write(ring, PACKETJ(0, 0, 0, PACKETJ_TYPE0)); + amdgpu_ring_write(ring, 0x00010000); +} + +/** + * vcn_v1_0_jpeg_ring_emit_fence - emit an fence & trap command + * + * @ring: amdgpu_ring pointer + * @fence: fence to emit + * + * Write a fence and a trap command to the ring. + */ +static void vcn_v1_0_jpeg_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, + PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JPEG_GPCOM_DATA0), 0, 0, PACKETJ_TYPE0)); + amdgpu_ring_write(ring, seq); + + amdgpu_ring_write(ring, + PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JPEG_GPCOM_DATA1), 0, 0, PACKETJ_TYPE0)); + amdgpu_ring_write(ring, seq); + + amdgpu_ring_write(ring, + PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_JRBC_RB_MEM_WR_64BIT_BAR_LOW), 0, 0, PACKETJ_TYPE0)); + amdgpu_ring_write(ring, lower_32_bits(addr)); + + amdgpu_ring_write(ring, + PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_JRBC_RB_MEM_WR_64BIT_BAR_HIGH), 0, 0, PACKETJ_TYPE0)); + amdgpu_ring_write(ring, upper_32_bits(addr)); + + amdgpu_ring_write(ring, + PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JPEG_GPCOM_CMD), 0, 0, PACKETJ_TYPE0)); + amdgpu_ring_write(ring, 0x8); + + amdgpu_ring_write(ring, + PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JPEG_GPCOM_CMD), 0, PACKETJ_CONDITION_CHECK0, PACKETJ_TYPE4)); + amdgpu_ring_write(ring, 0); + + amdgpu_ring_write(ring, + PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_RB_COND_RD_TIMER), 0, 0, PACKETJ_TYPE0)); + amdgpu_ring_write(ring, 0x01400200); + + amdgpu_ring_write(ring, + PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_RB_REF_DATA), 0, 0, PACKETJ_TYPE0)); + amdgpu_ring_write(ring, seq); + + amdgpu_ring_write(ring, + PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_JRBC_RB_MEM_RD_64BIT_BAR_LOW), 0, 0, PACKETJ_TYPE0)); + amdgpu_ring_write(ring, lower_32_bits(addr)); + + amdgpu_ring_write(ring, + PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_JRBC_RB_MEM_RD_64BIT_BAR_HIGH), 0, 0, PACKETJ_TYPE0)); + amdgpu_ring_write(ring, upper_32_bits(addr)); + + amdgpu_ring_write(ring, + PACKETJ(0, 0, PACKETJ_CONDITION_CHECK3, PACKETJ_TYPE2)); + amdgpu_ring_write(ring, 0xffffffff); + + amdgpu_ring_write(ring, + PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_EXTERNAL_REG_BASE), 0, 0, PACKETJ_TYPE0)); + amdgpu_ring_write(ring, 0x3fbc); + + amdgpu_ring_write(ring, + PACKETJ(0, 0, 0, PACKETJ_TYPE0)); + amdgpu_ring_write(ring, 0x1); +} + +/** + * vcn_v1_0_jpeg_ring_emit_ib - execute indirect buffer + * + * @ring: amdgpu_ring pointer + * @ib: indirect buffer to execute + * + * Write ring commands to execute the indirect buffer. + */ +static void vcn_v1_0_jpeg_ring_emit_ib(struct amdgpu_ring *ring, + struct amdgpu_ib *ib, + unsigned vmid, bool ctx_switch) +{ + struct amdgpu_device *adev = ring->adev; + + amdgpu_ring_write(ring, + PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_JRBC_IB_VMID), 0, 0, PACKETJ_TYPE0)); + amdgpu_ring_write(ring, (vmid | (vmid << 4))); + + amdgpu_ring_write(ring, + PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_JPEG_VMID), 0, 0, PACKETJ_TYPE0)); + amdgpu_ring_write(ring, (vmid | (vmid << 4))); + + amdgpu_ring_write(ring, + PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_JRBC_IB_64BIT_BAR_LOW), 0, 0, PACKETJ_TYPE0)); + amdgpu_ring_write(ring, lower_32_bits(ib->gpu_addr)); + + amdgpu_ring_write(ring, + PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_JRBC_IB_64BIT_BAR_HIGH), 0, 0, PACKETJ_TYPE0)); + amdgpu_ring_write(ring, upper_32_bits(ib->gpu_addr)); + + amdgpu_ring_write(ring, + PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_IB_SIZE), 0, 0, PACKETJ_TYPE0)); + amdgpu_ring_write(ring, ib->length_dw); + + amdgpu_ring_write(ring, + PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_JRBC_RB_MEM_RD_64BIT_BAR_LOW), 0, 0, PACKETJ_TYPE0)); + amdgpu_ring_write(ring, lower_32_bits(ring->gpu_addr)); + + amdgpu_ring_write(ring, + PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_JRBC_RB_MEM_RD_64BIT_BAR_HIGH), 0, 0, PACKETJ_TYPE0)); + amdgpu_ring_write(ring, upper_32_bits(ring->gpu_addr)); + + amdgpu_ring_write(ring, + PACKETJ(0, 0, PACKETJ_CONDITION_CHECK0, PACKETJ_TYPE2)); + amdgpu_ring_write(ring, 0); + + amdgpu_ring_write(ring, + PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_RB_COND_RD_TIMER), 0, 0, PACKETJ_TYPE0)); + amdgpu_ring_write(ring, 0x01400200); + + amdgpu_ring_write(ring, + PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_RB_REF_DATA), 0, 0, PACKETJ_TYPE0)); + amdgpu_ring_write(ring, 0x2); + + amdgpu_ring_write(ring, + PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_STATUS), 0, PACKETJ_CONDITION_CHECK3, PACKETJ_TYPE3)); + amdgpu_ring_write(ring, 0x2); +} + +static void vcn_v1_0_jpeg_ring_emit_reg_wait(struct amdgpu_ring *ring, + uint32_t reg, uint32_t val, + uint32_t mask) +{ + struct amdgpu_device *adev = ring->adev; + uint32_t reg_offset = (reg << 2); + + amdgpu_ring_write(ring, + PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_RB_COND_RD_TIMER), 0, 0, PACKETJ_TYPE0)); + amdgpu_ring_write(ring, 0x01400200); + + amdgpu_ring_write(ring, + PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_RB_REF_DATA), 0, 0, PACKETJ_TYPE0)); + amdgpu_ring_write(ring, val); + + amdgpu_ring_write(ring, + PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_EXTERNAL_REG_BASE), 0, 0, PACKETJ_TYPE0)); + if (((reg_offset >= 0x1f800) && (reg_offset <= 0x21fff)) || + ((reg_offset >= 0x1e000) && (reg_offset <= 0x1e1ff))) { + amdgpu_ring_write(ring, 0); + amdgpu_ring_write(ring, + PACKETJ((reg_offset >> 2), 0, 0, PACKETJ_TYPE3)); + } else { + amdgpu_ring_write(ring, reg_offset); + amdgpu_ring_write(ring, + PACKETJ(0, 0, 0, PACKETJ_TYPE3)); + } + amdgpu_ring_write(ring, mask); +} + +static void vcn_v1_0_jpeg_ring_emit_vm_flush(struct amdgpu_ring *ring, + unsigned vmid, uint64_t pd_addr) +{ + struct amdgpu_vmhub *hub = &ring->adev->vmhub[ring->funcs->vmhub]; + uint32_t data0, data1, mask; + + pd_addr = amdgpu_gmc_emit_flush_gpu_tlb(ring, vmid, pd_addr); + + /* wait for register write */ + data0 = hub->ctx0_ptb_addr_lo32 + vmid * 2; + data1 = lower_32_bits(pd_addr); + mask = 0xffffffff; + vcn_v1_0_jpeg_ring_emit_reg_wait(ring, data0, data1, mask); +} + +static void vcn_v1_0_jpeg_ring_emit_wreg(struct amdgpu_ring *ring, + uint32_t reg, uint32_t val) +{ + struct amdgpu_device *adev = ring->adev; + uint32_t reg_offset = (reg << 2); + + amdgpu_ring_write(ring, + PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_EXTERNAL_REG_BASE), 0, 0, PACKETJ_TYPE0)); + if (((reg_offset >= 0x1f800) && (reg_offset <= 0x21fff)) || + ((reg_offset >= 0x1e000) && (reg_offset <= 0x1e1ff))) { + amdgpu_ring_write(ring, 0); + amdgpu_ring_write(ring, + PACKETJ((reg_offset >> 2), 0, 0, PACKETJ_TYPE0)); + } else { + amdgpu_ring_write(ring, reg_offset); + amdgpu_ring_write(ring, + PACKETJ(0, 0, 0, PACKETJ_TYPE0)); + } + amdgpu_ring_write(ring, val); +} + +static void vcn_v1_0_jpeg_ring_nop(struct amdgpu_ring *ring, uint32_t count) +{ + int i; + + WARN_ON(ring->wptr % 2 || count % 2); + + for (i = 0; i < count / 2; i++) { + amdgpu_ring_write(ring, PACKETJ(0, 0, 0, PACKETJ_TYPE6)); + amdgpu_ring_write(ring, 0); + } +} + +static void vcn_v1_0_jpeg_ring_patch_wreg(struct amdgpu_ring *ring, uint32_t *ptr, uint32_t reg_offset, uint32_t val) +{ + struct amdgpu_device *adev = ring->adev; + ring->ring[(*ptr)++] = PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_EXTERNAL_REG_BASE), 0, 0, PACKETJ_TYPE0); + if (((reg_offset >= 0x1f800) && (reg_offset <= 0x21fff)) || + ((reg_offset >= 0x1e000) && (reg_offset <= 0x1e1ff))) { + ring->ring[(*ptr)++] = 0; + ring->ring[(*ptr)++] = PACKETJ((reg_offset >> 2), 0, 0, PACKETJ_TYPE0); + } else { + ring->ring[(*ptr)++] = reg_offset; + ring->ring[(*ptr)++] = PACKETJ(0, 0, 0, PACKETJ_TYPE0); + } + ring->ring[(*ptr)++] = val; +} + +static void vcn_v1_0_jpeg_ring_set_patch_ring(struct amdgpu_ring *ring, uint32_t ptr) +{ + struct amdgpu_device *adev = ring->adev; + + uint32_t reg, reg_offset, val, mask, i; + + // 1st: program mmUVD_LMI_JRBC_RB_MEM_RD_64BIT_BAR_LOW + reg = SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_JRBC_RB_MEM_RD_64BIT_BAR_LOW); + reg_offset = (reg << 2); + val = lower_32_bits(ring->gpu_addr); + vcn_v1_0_jpeg_ring_patch_wreg(ring, &ptr, reg_offset, val); + + // 2nd: program mmUVD_LMI_JRBC_RB_MEM_RD_64BIT_BAR_HIGH + reg = SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_JRBC_RB_MEM_RD_64BIT_BAR_HIGH); + reg_offset = (reg << 2); + val = upper_32_bits(ring->gpu_addr); + vcn_v1_0_jpeg_ring_patch_wreg(ring, &ptr, reg_offset, val); + + // 3rd to 5th: issue MEM_READ commands + for (i = 0; i <= 2; i++) { + ring->ring[ptr++] = PACKETJ(0, 0, 0, PACKETJ_TYPE2); + ring->ring[ptr++] = 0; + } + + // 6th: program mmUVD_JRBC_RB_CNTL register to enable NO_FETCH and RPTR write ability + reg = SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_RB_CNTL); + reg_offset = (reg << 2); + val = 0x13; + vcn_v1_0_jpeg_ring_patch_wreg(ring, &ptr, reg_offset, val); + + // 7th: program mmUVD_JRBC_RB_REF_DATA + reg = SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_RB_REF_DATA); + reg_offset = (reg << 2); + val = 0x1; + vcn_v1_0_jpeg_ring_patch_wreg(ring, &ptr, reg_offset, val); + + // 8th: issue conditional register read mmUVD_JRBC_RB_CNTL + reg = SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_RB_CNTL); + reg_offset = (reg << 2); + val = 0x1; + mask = 0x1; + + ring->ring[ptr++] = PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_RB_COND_RD_TIMER), 0, 0, PACKETJ_TYPE0); + ring->ring[ptr++] = 0x01400200; + ring->ring[ptr++] = PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_RB_REF_DATA), 0, 0, PACKETJ_TYPE0); + ring->ring[ptr++] = val; + ring->ring[ptr++] = PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_EXTERNAL_REG_BASE), 0, 0, PACKETJ_TYPE0); + if (((reg_offset >= 0x1f800) && (reg_offset <= 0x21fff)) || + ((reg_offset >= 0x1e000) && (reg_offset <= 0x1e1ff))) { + ring->ring[ptr++] = 0; + ring->ring[ptr++] = PACKETJ((reg_offset >> 2), 0, 0, PACKETJ_TYPE3); + } else { + ring->ring[ptr++] = reg_offset; + ring->ring[ptr++] = PACKETJ(0, 0, 0, PACKETJ_TYPE3); + } + ring->ring[ptr++] = mask; + + //9th to 21st: insert no-op + for (i = 0; i <= 12; i++) { + ring->ring[ptr++] = PACKETJ(0, 0, 0, PACKETJ_TYPE6); + ring->ring[ptr++] = 0; + } + + //22nd: reset mmUVD_JRBC_RB_RPTR + reg = SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_RB_RPTR); + reg_offset = (reg << 2); + val = 0; + vcn_v1_0_jpeg_ring_patch_wreg(ring, &ptr, reg_offset, val); + + //23rd: program mmUVD_JRBC_RB_CNTL to disable no_fetch + reg = SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_RB_CNTL); + reg_offset = (reg << 2); + val = 0x12; + vcn_v1_0_jpeg_ring_patch_wreg(ring, &ptr, reg_offset, val); +} + static int vcn_v1_0_set_interrupt_state(struct amdgpu_device *adev, struct amdgpu_irq_src *source, unsigned type, @@ -1150,6 +1567,9 @@ static int vcn_v1_0_process_interrupt(struct amdgpu_device *adev, case 120: amdgpu_fence_process(&adev->vcn.ring_enc[1]); break; + case 126: + amdgpu_fence_process(&adev->vcn.ring_jpeg); + break; default: DRM_ERROR("Unhandled interrupt: %d %d\n", entry->src_id, entry->src_data[0]); @@ -1273,6 +1693,39 @@ static const struct amdgpu_ring_funcs vcn_v1_0_enc_ring_vm_funcs = { .emit_reg_write_reg_wait = amdgpu_ring_emit_reg_write_reg_wait_helper, }; +static const struct amdgpu_ring_funcs vcn_v1_0_jpeg_ring_vm_funcs = { + .type = AMDGPU_RING_TYPE_VCN_JPEG, + .align_mask = 0xf, + .nop = PACKET0(0x81ff, 0), + .support_64bit_ptrs = false, + .vmhub = AMDGPU_MMHUB, + .extra_dw = 64, + .get_rptr = vcn_v1_0_jpeg_ring_get_rptr, + .get_wptr = vcn_v1_0_jpeg_ring_get_wptr, + .set_wptr = vcn_v1_0_jpeg_ring_set_wptr, + .emit_frame_size = + 6 + 6 + /* hdp invalidate / flush */ + SOC15_FLUSH_GPU_TLB_NUM_WREG * 6 + + SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 8 + + 8 + /* vcn_v1_0_dec_ring_emit_vm_flush */ + 14 + 14 + /* vcn_v1_0_dec_ring_emit_fence x2 vm fence */ + 6, + .emit_ib_size = 22, /* vcn_v1_0_dec_ring_emit_ib */ + .emit_ib = vcn_v1_0_jpeg_ring_emit_ib, + .emit_fence = vcn_v1_0_jpeg_ring_emit_fence, + .emit_vm_flush = vcn_v1_0_jpeg_ring_emit_vm_flush, + .test_ring = amdgpu_vcn_jpeg_ring_test_ring, + .test_ib = amdgpu_vcn_jpeg_ring_test_ib, + .insert_nop = vcn_v1_0_jpeg_ring_nop, + .insert_start = vcn_v1_0_jpeg_ring_insert_start, + .insert_end = vcn_v1_0_jpeg_ring_insert_end, + .pad_ib = amdgpu_ring_generic_pad_ib, + .begin_use = amdgpu_vcn_ring_begin_use, + .end_use = amdgpu_vcn_ring_end_use, + .emit_wreg = vcn_v1_0_jpeg_ring_emit_wreg, + .emit_reg_wait = vcn_v1_0_jpeg_ring_emit_reg_wait, +}; + static void vcn_v1_0_set_dec_ring_funcs(struct amdgpu_device *adev) { adev->vcn.ring_dec.funcs = &vcn_v1_0_dec_ring_vm_funcs; @@ -1289,6 +1742,12 @@ static void vcn_v1_0_set_enc_ring_funcs(struct amdgpu_device *adev) DRM_INFO("VCN encode is enabled in VM mode\n"); } +static void vcn_v1_0_set_jpeg_ring_funcs(struct amdgpu_device *adev) +{ + adev->vcn.ring_jpeg.funcs = &vcn_v1_0_jpeg_ring_vm_funcs; + DRM_INFO("VCN jpeg decode is enabled in VM mode\n"); +} + static const struct amdgpu_irq_src_funcs vcn_v1_0_irq_funcs = { .set = vcn_v1_0_set_interrupt_state, .process = vcn_v1_0_process_interrupt, diff --git a/drivers/gpu/drm/amd/amdgpu/vega10_reg_init.c b/drivers/gpu/drm/amd/amdgpu/vega10_reg_init.c index 45aafca7f315..c5c9b2bc190d 100644 --- a/drivers/gpu/drm/amd/amdgpu/vega10_reg_init.c +++ b/drivers/gpu/drm/amd/amdgpu/vega10_reg_init.c @@ -51,6 +51,7 @@ int vega10_reg_base_init(struct amdgpu_device *adev) adev->reg_offset[PWR_HWIP][i] = (uint32_t *)(&(PWR_BASE.instance[i])); adev->reg_offset[NBIF_HWIP][i] = (uint32_t *)(&(NBIF_BASE.instance[i])); adev->reg_offset[THM_HWIP][i] = (uint32_t *)(&(THM_BASE.instance[i])); + adev->reg_offset[CLK_HWIP][i] = (uint32_t *)(&(CLK_BASE.instance[i])); } return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/vi.c b/drivers/gpu/drm/amd/amdgpu/vi.c index 4ac1288ab7df..42c8ad105b05 100644 --- a/drivers/gpu/drm/amd/amdgpu/vi.c +++ b/drivers/gpu/drm/amd/amdgpu/vi.c @@ -1363,11 +1363,11 @@ static int vi_common_set_clockgating_state_by_smu(void *handle, if (adev->cg_flags & (AMD_CG_SUPPORT_MC_LS | AMD_CG_SUPPORT_MC_MGCG)) { if (adev->cg_flags & AMD_CG_SUPPORT_MC_LS) { - pp_support_state = AMD_CG_SUPPORT_MC_LS; + pp_support_state = PP_STATE_SUPPORT_LS; pp_state = PP_STATE_LS; } if (adev->cg_flags & AMD_CG_SUPPORT_MC_MGCG) { - pp_support_state |= AMD_CG_SUPPORT_MC_MGCG; + pp_support_state |= PP_STATE_SUPPORT_CG; pp_state |= PP_STATE_CG; } if (state == AMD_CG_STATE_UNGATE) @@ -1382,11 +1382,11 @@ static int vi_common_set_clockgating_state_by_smu(void *handle, if (adev->cg_flags & (AMD_CG_SUPPORT_SDMA_LS | AMD_CG_SUPPORT_SDMA_MGCG)) { if (adev->cg_flags & AMD_CG_SUPPORT_SDMA_LS) { - pp_support_state = AMD_CG_SUPPORT_SDMA_LS; + pp_support_state = PP_STATE_SUPPORT_LS; pp_state = PP_STATE_LS; } if (adev->cg_flags & AMD_CG_SUPPORT_SDMA_MGCG) { - pp_support_state |= AMD_CG_SUPPORT_SDMA_MGCG; + pp_support_state |= PP_STATE_SUPPORT_CG; pp_state |= PP_STATE_CG; } if (state == AMD_CG_STATE_UNGATE) @@ -1401,11 +1401,11 @@ static int vi_common_set_clockgating_state_by_smu(void *handle, if (adev->cg_flags & (AMD_CG_SUPPORT_HDP_LS | AMD_CG_SUPPORT_HDP_MGCG)) { if (adev->cg_flags & AMD_CG_SUPPORT_HDP_LS) { - pp_support_state = AMD_CG_SUPPORT_HDP_LS; + pp_support_state = PP_STATE_SUPPORT_LS; pp_state = PP_STATE_LS; } if (adev->cg_flags & AMD_CG_SUPPORT_HDP_MGCG) { - pp_support_state |= AMD_CG_SUPPORT_HDP_MGCG; + pp_support_state |= PP_STATE_SUPPORT_CG; pp_state |= PP_STATE_CG; } if (state == AMD_CG_STATE_UNGATE) diff --git a/drivers/gpu/drm/amd/amdkfd/cik_event_interrupt.c b/drivers/gpu/drm/amd/amdkfd/cik_event_interrupt.c index 49df6c791cfc..5d2475d5392c 100644 --- a/drivers/gpu/drm/amd/amdkfd/cik_event_interrupt.c +++ b/drivers/gpu/drm/amd/amdkfd/cik_event_interrupt.c @@ -25,12 +25,39 @@ #include "cik_int.h" static bool cik_event_interrupt_isr(struct kfd_dev *dev, - const uint32_t *ih_ring_entry) + const uint32_t *ih_ring_entry, + uint32_t *patched_ihre, + bool *patched_flag) { const struct cik_ih_ring_entry *ihre = (const struct cik_ih_ring_entry *)ih_ring_entry; + const struct kfd2kgd_calls *f2g = dev->kfd2kgd; unsigned int vmid, pasid; + /* This workaround is due to HW/FW limitation on Hawaii that + * VMID and PASID are not written into ih_ring_entry + */ + if ((ihre->source_id == CIK_INTSRC_GFX_PAGE_INV_FAULT || + ihre->source_id == CIK_INTSRC_GFX_MEM_PROT_FAULT) && + dev->device_info->asic_family == CHIP_HAWAII) { + struct cik_ih_ring_entry *tmp_ihre = + (struct cik_ih_ring_entry *)patched_ihre; + + *patched_flag = true; + *tmp_ihre = *ihre; + + vmid = f2g->read_vmid_from_vmfault_reg(dev->kgd); + pasid = f2g->get_atc_vmid_pasid_mapping_pasid(dev->kgd, vmid); + + tmp_ihre->ring_id &= 0x000000ff; + tmp_ihre->ring_id |= vmid << 8; + tmp_ihre->ring_id |= pasid << 16; + + return (pasid != 0) && + vmid >= dev->vm_info.first_vmid_kfd && + vmid <= dev->vm_info.last_vmid_kfd; + } + /* Only handle interrupts from KFD VMIDs */ vmid = (ihre->ring_id & 0x0000ff00) >> 8; if (vmid < dev->vm_info.first_vmid_kfd || @@ -48,18 +75,19 @@ static bool cik_event_interrupt_isr(struct kfd_dev *dev, return ihre->source_id == CIK_INTSRC_CP_END_OF_PIPE || ihre->source_id == CIK_INTSRC_SDMA_TRAP || ihre->source_id == CIK_INTSRC_SQ_INTERRUPT_MSG || - ihre->source_id == CIK_INTSRC_CP_BAD_OPCODE; + ihre->source_id == CIK_INTSRC_CP_BAD_OPCODE || + ihre->source_id == CIK_INTSRC_GFX_PAGE_INV_FAULT || + ihre->source_id == CIK_INTSRC_GFX_MEM_PROT_FAULT; } static void cik_event_interrupt_wq(struct kfd_dev *dev, const uint32_t *ih_ring_entry) { - unsigned int pasid; const struct cik_ih_ring_entry *ihre = (const struct cik_ih_ring_entry *)ih_ring_entry; uint32_t context_id = ihre->data & 0xfffffff; - - pasid = (ihre->ring_id & 0xffff0000) >> 16; + unsigned int vmid = (ihre->ring_id & 0x0000ff00) >> 8; + unsigned int pasid = (ihre->ring_id & 0xffff0000) >> 16; if (pasid == 0) return; @@ -72,6 +100,22 @@ static void cik_event_interrupt_wq(struct kfd_dev *dev, kfd_signal_event_interrupt(pasid, context_id & 0xff, 8); else if (ihre->source_id == CIK_INTSRC_CP_BAD_OPCODE) kfd_signal_hw_exception_event(pasid); + else if (ihre->source_id == CIK_INTSRC_GFX_PAGE_INV_FAULT || + ihre->source_id == CIK_INTSRC_GFX_MEM_PROT_FAULT) { + struct kfd_vm_fault_info info; + + kfd_process_vm_fault(dev->dqm, pasid); + + memset(&info, 0, sizeof(info)); + dev->kfd2kgd->get_vm_fault_info(dev->kgd, &info); + if (!info.page_addr && !info.status) + return; + + if (info.vmid == vmid) + kfd_signal_vm_fault_event(dev, pasid, &info); + else + kfd_signal_vm_fault_event(dev, pasid, NULL); + } } const struct kfd_event_interrupt_class event_interrupt_class_cik = { diff --git a/drivers/gpu/drm/amd/amdkfd/cik_int.h b/drivers/gpu/drm/amd/amdkfd/cik_int.h index 109298b9d507..76f8677a7926 100644 --- a/drivers/gpu/drm/amd/amdkfd/cik_int.h +++ b/drivers/gpu/drm/amd/amdkfd/cik_int.h @@ -20,8 +20,8 @@ * OTHER DEALINGS IN THE SOFTWARE. */ -#ifndef HSA_RADEON_CIK_INT_H_INCLUDED -#define HSA_RADEON_CIK_INT_H_INCLUDED +#ifndef CIK_INT_H_INCLUDED +#define CIK_INT_H_INCLUDED #include <linux/types.h> @@ -34,9 +34,10 @@ struct cik_ih_ring_entry { #define CIK_INTSRC_CP_END_OF_PIPE 0xB5 #define CIK_INTSRC_CP_BAD_OPCODE 0xB7 -#define CIK_INTSRC_DEQUEUE_COMPLETE 0xC6 #define CIK_INTSRC_SDMA_TRAP 0xE0 #define CIK_INTSRC_SQ_INTERRUPT_MSG 0xEF +#define CIK_INTSRC_GFX_PAGE_INV_FAULT 0x92 +#define CIK_INTSRC_GFX_MEM_PROT_FAULT 0x93 #endif diff --git a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler.h b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler.h index f68aef02fc1f..3621efbd5759 100644 --- a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler.h +++ b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler.h @@ -21,18 +21,21 @@ */ static const uint32_t cwsr_trap_gfx8_hex[] = { - 0xbf820001, 0xbf820125, + 0xbf820001, 0xbf82012b, 0xb8f4f802, 0x89748674, 0xb8f5f803, 0x8675ff75, - 0x00000400, 0xbf850011, + 0x00000400, 0xbf850017, 0xc00a1e37, 0x00000000, 0xbf8c007f, 0x87777978, - 0xbf840002, 0xb974f802, - 0xbe801d78, 0xb8f5f803, - 0x8675ff75, 0x000001ff, - 0xbf850002, 0x80708470, - 0x82718071, 0x8671ff71, - 0x0000ffff, 0xb974f802, + 0xbf840005, 0x8f728374, + 0xb972e0c2, 0xbf800002, + 0xb9740002, 0xbe801d78, + 0xb8f5f803, 0x8675ff75, + 0x000001ff, 0xbf850002, + 0x80708470, 0x82718071, + 0x8671ff71, 0x0000ffff, + 0x8f728374, 0xb972e0c2, + 0xbf800002, 0xb9740002, 0xbe801f70, 0xb8f5f803, 0x8675ff75, 0x00000100, 0xbf840006, 0xbefa0080, @@ -168,7 +171,7 @@ static const uint32_t cwsr_trap_gfx8_hex[] = { 0x807c847c, 0x806eff6e, 0x00000400, 0xbf0a757c, 0xbf85ffef, 0xbf9c0000, - 0xbf8200ca, 0xbef8007e, + 0xbf8200cd, 0xbef8007e, 0x8679ff7f, 0x0000ffff, 0x8779ff79, 0x00040000, 0xbefa0080, 0xbefb00ff, @@ -268,16 +271,18 @@ static const uint32_t cwsr_trap_gfx8_hex[] = { 0x8f739773, 0xb976f807, 0x8671ff71, 0x0000ffff, 0x86fe7e7e, 0x86ea6a6a, - 0xb974f802, 0xbf8a0000, - 0x95807370, 0xbf810000, + 0x8f768374, 0xb976e0c2, + 0xbf800002, 0xb9740002, + 0xbf8a0000, 0x95807370, + 0xbf810000, 0x00000000, }; static const uint32_t cwsr_trap_gfx9_hex[] = { - 0xbf820001, 0xbf82015a, + 0xbf820001, 0xbf82015d, 0xb8f8f802, 0x89788678, 0xb8f1f803, 0x866eff71, - 0x00000400, 0xbf850034, + 0x00000400, 0xbf850037, 0x866eff71, 0x00000800, 0xbf850003, 0x866eff71, 0x00000100, 0xbf840008, @@ -303,258 +308,261 @@ static const uint32_t cwsr_trap_gfx9_hex[] = { 0x8f6e8b77, 0x866eff6e, 0x001f8000, 0xb96ef807, 0x86fe7e7e, 0x86ea6a6a, - 0xb978f802, 0xbe801f6c, - 0x866dff6d, 0x0000ffff, - 0xbef00080, 0xb9700283, - 0xb8f02407, 0x8e709c70, - 0x876d706d, 0xb8f003c7, - 0x8e709b70, 0x876d706d, - 0xb8f0f807, 0x8670ff70, - 0x00007fff, 0xb970f807, - 0xbeee007e, 0xbeef007f, - 0xbefe0180, 0xbf900004, - 0x87708478, 0xb970f802, - 0xbf8e0002, 0xbf88fffe, - 0xb8f02a05, 0x80708170, - 0x8e708a70, 0xb8f11605, - 0x80718171, 0x8e718671, - 0x80707170, 0x80707e70, - 0x8271807f, 0x8671ff71, - 0x0000ffff, 0xc0471cb8, - 0x00000040, 0xbf8cc07f, - 0xc04b1d38, 0x00000048, - 0xbf8cc07f, 0xc0431e78, - 0x00000058, 0xbf8cc07f, - 0xc0471eb8, 0x0000005c, - 0xbf8cc07f, 0xbef4007e, - 0x8675ff7f, 0x0000ffff, - 0x8775ff75, 0x00040000, - 0xbef60080, 0xbef700ff, - 0x00807fac, 0x8670ff7f, - 0x08000000, 0x8f708370, - 0x87777077, 0x8670ff7f, - 0x70000000, 0x8f708170, - 0x87777077, 0xbefb007c, - 0xbefa0080, 0xb8fa2a05, - 0x807a817a, 0x8e7a8a7a, - 0xb8f01605, 0x80708170, - 0x8e708670, 0x807a707a, - 0xbef60084, 0xbef600ff, - 0x01000000, 0xbefe007c, - 0xbefc007a, 0xc0611efa, - 0x0000007c, 0xbf8cc07f, - 0x807a847a, 0xbefc007e, + 0x8f6e8378, 0xb96ee0c2, + 0xbf800002, 0xb9780002, + 0xbe801f6c, 0x866dff6d, + 0x0000ffff, 0xbef00080, + 0xb9700283, 0xb8f02407, + 0x8e709c70, 0x876d706d, + 0xb8f003c7, 0x8e709b70, + 0x876d706d, 0xb8f0f807, + 0x8670ff70, 0x00007fff, + 0xb970f807, 0xbeee007e, + 0xbeef007f, 0xbefe0180, + 0xbf900004, 0x87708478, + 0xb970f802, 0xbf8e0002, + 0xbf88fffe, 0xb8f02a05, + 0x80708170, 0x8e708a70, + 0xb8f11605, 0x80718171, + 0x8e718671, 0x80707170, + 0x80707e70, 0x8271807f, + 0x8671ff71, 0x0000ffff, + 0xc0471cb8, 0x00000040, + 0xbf8cc07f, 0xc04b1d38, + 0x00000048, 0xbf8cc07f, + 0xc0431e78, 0x00000058, + 0xbf8cc07f, 0xc0471eb8, + 0x0000005c, 0xbf8cc07f, + 0xbef4007e, 0x8675ff7f, + 0x0000ffff, 0x8775ff75, + 0x00040000, 0xbef60080, + 0xbef700ff, 0x00807fac, + 0x8670ff7f, 0x08000000, + 0x8f708370, 0x87777077, + 0x8670ff7f, 0x70000000, + 0x8f708170, 0x87777077, + 0xbefb007c, 0xbefa0080, + 0xb8fa2a05, 0x807a817a, + 0x8e7a8a7a, 0xb8f01605, + 0x80708170, 0x8e708670, + 0x807a707a, 0xbef60084, + 0xbef600ff, 0x01000000, 0xbefe007c, 0xbefc007a, - 0xc0611b3a, 0x0000007c, + 0xc0611efa, 0x0000007c, 0xbf8cc07f, 0x807a847a, 0xbefc007e, 0xbefe007c, - 0xbefc007a, 0xc0611b7a, + 0xbefc007a, 0xc0611b3a, 0x0000007c, 0xbf8cc07f, 0x807a847a, 0xbefc007e, 0xbefe007c, 0xbefc007a, - 0xc0611bba, 0x0000007c, + 0xc0611b7a, 0x0000007c, 0xbf8cc07f, 0x807a847a, 0xbefc007e, 0xbefe007c, - 0xbefc007a, 0xc0611bfa, + 0xbefc007a, 0xc0611bba, 0x0000007c, 0xbf8cc07f, 0x807a847a, 0xbefc007e, 0xbefe007c, 0xbefc007a, - 0xc0611e3a, 0x0000007c, - 0xbf8cc07f, 0x807a847a, - 0xbefc007e, 0xb8f1f803, - 0xbefe007c, 0xbefc007a, - 0xc0611c7a, 0x0000007c, + 0xc0611bfa, 0x0000007c, 0xbf8cc07f, 0x807a847a, 0xbefc007e, 0xbefe007c, - 0xbefc007a, 0xc0611a3a, + 0xbefc007a, 0xc0611e3a, + 0x0000007c, 0xbf8cc07f, + 0x807a847a, 0xbefc007e, + 0xb8f1f803, 0xbefe007c, + 0xbefc007a, 0xc0611c7a, 0x0000007c, 0xbf8cc07f, 0x807a847a, 0xbefc007e, 0xbefe007c, 0xbefc007a, - 0xc0611a7a, 0x0000007c, - 0xbf8cc07f, 0x807a847a, - 0xbefc007e, 0xb8fbf801, - 0xbefe007c, 0xbefc007a, - 0xc0611efa, 0x0000007c, + 0xc0611a3a, 0x0000007c, 0xbf8cc07f, 0x807a847a, - 0xbefc007e, 0x8670ff7f, - 0x04000000, 0xbeef0080, - 0x876f6f70, 0xb8fa2a05, + 0xbefc007e, 0xbefe007c, + 0xbefc007a, 0xc0611a7a, + 0x0000007c, 0xbf8cc07f, + 0x807a847a, 0xbefc007e, + 0xb8fbf801, 0xbefe007c, + 0xbefc007a, 0xc0611efa, + 0x0000007c, 0xbf8cc07f, + 0x807a847a, 0xbefc007e, + 0x8670ff7f, 0x04000000, + 0xbeef0080, 0x876f6f70, + 0xb8fa2a05, 0x807a817a, + 0x8e7a8a7a, 0xb8f11605, + 0x80718171, 0x8e718471, + 0x8e768271, 0xbef600ff, + 0x01000000, 0xbef20174, + 0x80747a74, 0x82758075, + 0xbefc0080, 0xbf800000, + 0xbe802b00, 0xbe822b02, + 0xbe842b04, 0xbe862b06, + 0xbe882b08, 0xbe8a2b0a, + 0xbe8c2b0c, 0xbe8e2b0e, + 0xc06b003a, 0x00000000, + 0xbf8cc07f, 0xc06b013a, + 0x00000010, 0xbf8cc07f, + 0xc06b023a, 0x00000020, + 0xbf8cc07f, 0xc06b033a, + 0x00000030, 0xbf8cc07f, + 0x8074c074, 0x82758075, + 0x807c907c, 0xbf0a717c, + 0xbf85ffe7, 0xbef40172, + 0xbefa0080, 0xbefe00c1, + 0xbeff00c1, 0xbee80080, + 0xbee90080, 0xbef600ff, + 0x01000000, 0xe0724000, + 0x7a1d0000, 0xe0724100, + 0x7a1d0100, 0xe0724200, + 0x7a1d0200, 0xe0724300, + 0x7a1d0300, 0xbefe00c1, + 0xbeff00c1, 0xb8f14306, + 0x8671c171, 0xbf84002c, + 0xbf8a0000, 0x8670ff6f, + 0x04000000, 0xbf840028, + 0x8e718671, 0x8e718271, + 0xbef60071, 0xb8fa2a05, 0x807a817a, 0x8e7a8a7a, - 0xb8f11605, 0x80718171, - 0x8e718471, 0x8e768271, + 0xb8f01605, 0x80708170, + 0x8e708670, 0x807a707a, + 0x807aff7a, 0x00000080, 0xbef600ff, 0x01000000, - 0xbef20174, 0x80747a74, - 0x82758075, 0xbefc0080, - 0xbf800000, 0xbe802b00, - 0xbe822b02, 0xbe842b04, - 0xbe862b06, 0xbe882b08, - 0xbe8a2b0a, 0xbe8c2b0c, - 0xbe8e2b0e, 0xc06b003a, - 0x00000000, 0xbf8cc07f, - 0xc06b013a, 0x00000010, - 0xbf8cc07f, 0xc06b023a, - 0x00000020, 0xbf8cc07f, - 0xc06b033a, 0x00000030, - 0xbf8cc07f, 0x8074c074, - 0x82758075, 0x807c907c, - 0xbf0a717c, 0xbf85ffe7, - 0xbef40172, 0xbefa0080, + 0xbefc0080, 0xd28c0002, + 0x000100c1, 0xd28d0003, + 0x000204c1, 0xd1060002, + 0x00011103, 0x7e0602ff, + 0x00000200, 0xbefc00ff, + 0x00010000, 0xbe800077, + 0x8677ff77, 0xff7fffff, + 0x8777ff77, 0x00058000, + 0xd8ec0000, 0x00000002, + 0xbf8cc07f, 0xe0765000, + 0x7a1d0002, 0x68040702, + 0xd0c9006a, 0x0000e302, + 0xbf87fff7, 0xbef70000, + 0xbefa00ff, 0x00000400, 0xbefe00c1, 0xbeff00c1, - 0xbee80080, 0xbee90080, + 0xb8f12a05, 0x80718171, + 0x8e718271, 0x8e768871, 0xbef600ff, 0x01000000, + 0xbefc0084, 0xbf0a717c, + 0xbf840015, 0xbf11017c, + 0x8071ff71, 0x00001000, + 0x7e000300, 0x7e020301, + 0x7e040302, 0x7e060303, 0xe0724000, 0x7a1d0000, 0xe0724100, 0x7a1d0100, 0xe0724200, 0x7a1d0200, 0xe0724300, 0x7a1d0300, + 0x807c847c, 0x807aff7a, + 0x00000400, 0xbf0a717c, + 0xbf85ffef, 0xbf9c0000, + 0xbf8200dc, 0xbef4007e, + 0x8675ff7f, 0x0000ffff, + 0x8775ff75, 0x00040000, + 0xbef60080, 0xbef700ff, + 0x00807fac, 0x866eff7f, + 0x08000000, 0x8f6e836e, + 0x87776e77, 0x866eff7f, + 0x70000000, 0x8f6e816e, + 0x87776e77, 0x866eff7f, + 0x04000000, 0xbf84001e, 0xbefe00c1, 0xbeff00c1, - 0xb8f14306, 0x8671c171, - 0xbf84002c, 0xbf8a0000, - 0x8670ff6f, 0x04000000, - 0xbf840028, 0x8e718671, - 0x8e718271, 0xbef60071, - 0xb8fa2a05, 0x807a817a, - 0x8e7a8a7a, 0xb8f01605, - 0x80708170, 0x8e708670, - 0x807a707a, 0x807aff7a, + 0xb8ef4306, 0x866fc16f, + 0xbf840019, 0x8e6f866f, + 0x8e6f826f, 0xbef6006f, + 0xb8f82a05, 0x80788178, + 0x8e788a78, 0xb8ee1605, + 0x806e816e, 0x8e6e866e, + 0x80786e78, 0x8078ff78, 0x00000080, 0xbef600ff, 0x01000000, 0xbefc0080, - 0xd28c0002, 0x000100c1, - 0xd28d0003, 0x000204c1, - 0xd1060002, 0x00011103, - 0x7e0602ff, 0x00000200, - 0xbefc00ff, 0x00010000, - 0xbe800077, 0x8677ff77, - 0xff7fffff, 0x8777ff77, - 0x00058000, 0xd8ec0000, - 0x00000002, 0xbf8cc07f, - 0xe0765000, 0x7a1d0002, - 0x68040702, 0xd0c9006a, - 0x0000e302, 0xbf87fff7, - 0xbef70000, 0xbefa00ff, - 0x00000400, 0xbefe00c1, - 0xbeff00c1, 0xb8f12a05, - 0x80718171, 0x8e718271, - 0x8e768871, 0xbef600ff, - 0x01000000, 0xbefc0084, - 0xbf0a717c, 0xbf840015, - 0xbf11017c, 0x8071ff71, - 0x00001000, 0x7e000300, + 0xe0510000, 0x781d0000, + 0xe0510100, 0x781d0000, + 0x807cff7c, 0x00000200, + 0x8078ff78, 0x00000200, + 0xbf0a6f7c, 0xbf85fff6, + 0xbef80080, 0xbefe00c1, + 0xbeff00c1, 0xb8ef2a05, + 0x806f816f, 0x8e6f826f, + 0x8e76886f, 0xbef600ff, + 0x01000000, 0xbeee0078, + 0x8078ff78, 0x00000400, + 0xbefc0084, 0xbf11087c, + 0x806fff6f, 0x00008000, + 0xe0524000, 0x781d0000, + 0xe0524100, 0x781d0100, + 0xe0524200, 0x781d0200, + 0xe0524300, 0x781d0300, + 0xbf8c0f70, 0x7e000300, 0x7e020301, 0x7e040302, - 0x7e060303, 0xe0724000, - 0x7a1d0000, 0xe0724100, - 0x7a1d0100, 0xe0724200, - 0x7a1d0200, 0xe0724300, - 0x7a1d0300, 0x807c847c, - 0x807aff7a, 0x00000400, - 0xbf0a717c, 0xbf85ffef, - 0xbf9c0000, 0xbf8200d9, - 0xbef4007e, 0x8675ff7f, - 0x0000ffff, 0x8775ff75, - 0x00040000, 0xbef60080, - 0xbef700ff, 0x00807fac, - 0x866eff7f, 0x08000000, - 0x8f6e836e, 0x87776e77, - 0x866eff7f, 0x70000000, - 0x8f6e816e, 0x87776e77, - 0x866eff7f, 0x04000000, - 0xbf84001e, 0xbefe00c1, - 0xbeff00c1, 0xb8ef4306, - 0x866fc16f, 0xbf840019, - 0x8e6f866f, 0x8e6f826f, - 0xbef6006f, 0xb8f82a05, + 0x7e060303, 0x807c847c, + 0x8078ff78, 0x00000400, + 0xbf0a6f7c, 0xbf85ffee, + 0xbf9c0000, 0xe0524000, + 0x6e1d0000, 0xe0524100, + 0x6e1d0100, 0xe0524200, + 0x6e1d0200, 0xe0524300, + 0x6e1d0300, 0xb8f82a05, 0x80788178, 0x8e788a78, 0xb8ee1605, 0x806e816e, 0x8e6e866e, 0x80786e78, - 0x8078ff78, 0x00000080, - 0xbef600ff, 0x01000000, - 0xbefc0080, 0xe0510000, - 0x781d0000, 0xe0510100, - 0x781d0000, 0x807cff7c, - 0x00000200, 0x8078ff78, - 0x00000200, 0xbf0a6f7c, - 0xbf85fff6, 0xbef80080, - 0xbefe00c1, 0xbeff00c1, - 0xb8ef2a05, 0x806f816f, - 0x8e6f826f, 0x8e76886f, - 0xbef600ff, 0x01000000, - 0xbeee0078, 0x8078ff78, - 0x00000400, 0xbefc0084, - 0xbf11087c, 0x806fff6f, - 0x00008000, 0xe0524000, - 0x781d0000, 0xe0524100, - 0x781d0100, 0xe0524200, - 0x781d0200, 0xe0524300, - 0x781d0300, 0xbf8c0f70, - 0x7e000300, 0x7e020301, - 0x7e040302, 0x7e060303, - 0x807c847c, 0x8078ff78, - 0x00000400, 0xbf0a6f7c, - 0xbf85ffee, 0xbf9c0000, - 0xe0524000, 0x6e1d0000, - 0xe0524100, 0x6e1d0100, - 0xe0524200, 0x6e1d0200, - 0xe0524300, 0x6e1d0300, + 0x80f8c078, 0xb8ef1605, + 0x806f816f, 0x8e6f846f, + 0x8e76826f, 0xbef600ff, + 0x01000000, 0xbefc006f, + 0xc031003a, 0x00000078, + 0x80f8c078, 0xbf8cc07f, + 0x80fc907c, 0xbf800000, + 0xbe802d00, 0xbe822d02, + 0xbe842d04, 0xbe862d06, + 0xbe882d08, 0xbe8a2d0a, + 0xbe8c2d0c, 0xbe8e2d0e, + 0xbf06807c, 0xbf84fff0, 0xb8f82a05, 0x80788178, 0x8e788a78, 0xb8ee1605, 0x806e816e, 0x8e6e866e, - 0x80786e78, 0x80f8c078, - 0xb8ef1605, 0x806f816f, - 0x8e6f846f, 0x8e76826f, + 0x80786e78, 0xbef60084, 0xbef600ff, 0x01000000, - 0xbefc006f, 0xc031003a, - 0x00000078, 0x80f8c078, - 0xbf8cc07f, 0x80fc907c, - 0xbf800000, 0xbe802d00, - 0xbe822d02, 0xbe842d04, - 0xbe862d06, 0xbe882d08, - 0xbe8a2d0a, 0xbe8c2d0c, - 0xbe8e2d0e, 0xbf06807c, - 0xbf84fff0, 0xb8f82a05, - 0x80788178, 0x8e788a78, - 0xb8ee1605, 0x806e816e, - 0x8e6e866e, 0x80786e78, - 0xbef60084, 0xbef600ff, - 0x01000000, 0xc0211bfa, + 0xc0211bfa, 0x00000078, + 0x80788478, 0xc0211b3a, 0x00000078, 0x80788478, - 0xc0211b3a, 0x00000078, - 0x80788478, 0xc0211b7a, + 0xc0211b7a, 0x00000078, + 0x80788478, 0xc0211eba, 0x00000078, 0x80788478, - 0xc0211eba, 0x00000078, - 0x80788478, 0xc0211efa, + 0xc0211efa, 0x00000078, + 0x80788478, 0xc0211c3a, 0x00000078, 0x80788478, - 0xc0211c3a, 0x00000078, - 0x80788478, 0xc0211c7a, + 0xc0211c7a, 0x00000078, + 0x80788478, 0xc0211a3a, 0x00000078, 0x80788478, - 0xc0211a3a, 0x00000078, - 0x80788478, 0xc0211a7a, + 0xc0211a7a, 0x00000078, + 0x80788478, 0xc0211cfa, 0x00000078, 0x80788478, - 0xc0211cfa, 0x00000078, - 0x80788478, 0xbf8cc07f, - 0xbefc006f, 0xbefe007a, - 0xbeff007b, 0x866f71ff, - 0x000003ff, 0xb96f4803, - 0x866f71ff, 0xfffff800, - 0x8f6f8b6f, 0xb96fa2c3, - 0xb973f801, 0xb8ee2a05, - 0x806e816e, 0x8e6e8a6e, - 0xb8ef1605, 0x806f816f, - 0x8e6f866f, 0x806e6f6e, - 0x806e746e, 0x826f8075, - 0x866fff6f, 0x0000ffff, - 0xc0071cb7, 0x00000040, - 0xc00b1d37, 0x00000048, - 0xc0031e77, 0x00000058, - 0xc0071eb7, 0x0000005c, - 0xbf8cc07f, 0x866fff6d, - 0xf0000000, 0x8f6f9c6f, - 0x8e6f906f, 0xbeee0080, - 0x876e6f6e, 0x866fff6d, - 0x08000000, 0x8f6f9b6f, - 0x8e6f8f6f, 0x876e6f6e, - 0x866fff70, 0x00800000, - 0x8f6f976f, 0xb96ef807, - 0x866dff6d, 0x0000ffff, - 0x86fe7e7e, 0x86ea6a6a, - 0xb970f802, 0xbf8a0000, + 0xbf8cc07f, 0xbefc006f, + 0xbefe007a, 0xbeff007b, + 0x866f71ff, 0x000003ff, + 0xb96f4803, 0x866f71ff, + 0xfffff800, 0x8f6f8b6f, + 0xb96fa2c3, 0xb973f801, + 0xb8ee2a05, 0x806e816e, + 0x8e6e8a6e, 0xb8ef1605, + 0x806f816f, 0x8e6f866f, + 0x806e6f6e, 0x806e746e, + 0x826f8075, 0x866fff6f, + 0x0000ffff, 0xc0071cb7, + 0x00000040, 0xc00b1d37, + 0x00000048, 0xc0031e77, + 0x00000058, 0xc0071eb7, + 0x0000005c, 0xbf8cc07f, + 0x866fff6d, 0xf0000000, + 0x8f6f9c6f, 0x8e6f906f, + 0xbeee0080, 0x876e6f6e, + 0x866fff6d, 0x08000000, + 0x8f6f9b6f, 0x8e6f8f6f, + 0x876e6f6e, 0x866fff70, + 0x00800000, 0x8f6f976f, + 0xb96ef807, 0x866dff6d, + 0x0000ffff, 0x86fe7e7e, + 0x86ea6a6a, 0x8f6e8370, + 0xb96ee0c2, 0xbf800002, + 0xb9700002, 0xbf8a0000, 0x95806f6c, 0xbf810000, }; diff --git a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx8.asm b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx8.asm index a2a04bb64096..abe1a5da29fb 100644 --- a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx8.asm +++ b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx8.asm @@ -103,6 +103,10 @@ var SQ_WAVE_STATUS_INST_ATC_SHIFT = 23 var SQ_WAVE_STATUS_INST_ATC_MASK = 0x00800000 var SQ_WAVE_STATUS_SPI_PRIO_SHIFT = 1 var SQ_WAVE_STATUS_SPI_PRIO_MASK = 0x00000006 +var SQ_WAVE_STATUS_PRE_SPI_PRIO_SHIFT = 0 +var SQ_WAVE_STATUS_PRE_SPI_PRIO_SIZE = 1 +var SQ_WAVE_STATUS_POST_SPI_PRIO_SHIFT = 3 +var SQ_WAVE_STATUS_POST_SPI_PRIO_SIZE = 29 var SQ_WAVE_LDS_ALLOC_LDS_SIZE_SHIFT = 12 var SQ_WAVE_LDS_ALLOC_LDS_SIZE_SIZE = 9 @@ -251,7 +255,7 @@ if (!EMU_RUN_HACK) s_waitcnt lgkmcnt(0) s_or_b32 ttmp7, ttmp8, ttmp9 s_cbranch_scc0 L_NO_NEXT_TRAP //next level trap handler not been set - s_setreg_b32 hwreg(HW_REG_STATUS), s_save_status //restore HW status(SCC) + set_status_without_spi_prio(s_save_status, ttmp2) //restore HW status(SCC) s_setpc_b64 [ttmp8,ttmp9] //jump to next level trap handler L_NO_NEXT_TRAP: @@ -262,7 +266,7 @@ L_NO_NEXT_TRAP: s_addc_u32 ttmp1, ttmp1, 0 L_EXCP_CASE: s_and_b32 ttmp1, ttmp1, 0xFFFF - s_setreg_b32 hwreg(HW_REG_STATUS), s_save_status //restore HW status(SCC) + set_status_without_spi_prio(s_save_status, ttmp2) //restore HW status(SCC) s_rfe_b64 [ttmp0, ttmp1] end // ********* End handling of non-CWSR traps ******************* @@ -1053,7 +1057,7 @@ end s_and_b32 s_restore_pc_hi, s_restore_pc_hi, 0x0000ffff //pc[47:32] //Do it here in order not to affect STATUS s_and_b64 exec, exec, exec // Restore STATUS.EXECZ, not writable by s_setreg_b32 s_and_b64 vcc, vcc, vcc // Restore STATUS.VCCZ, not writable by s_setreg_b32 - s_setreg_b32 hwreg(HW_REG_STATUS), s_restore_status // SCC is included, which is changed by previous salu + set_status_without_spi_prio(s_restore_status, s_restore_tmp) // SCC is included, which is changed by previous salu s_barrier //barrier to ensure the readiness of LDS before access attempts from any other wave in the same TG //FIXME not performance-optimal at this time @@ -1134,3 +1138,11 @@ end function get_hwreg_size_bytes return 128 //HWREG size 128 bytes end + +function set_status_without_spi_prio(status, tmp) + // Do not restore STATUS.SPI_PRIO since scheduler may have raised it. + s_lshr_b32 tmp, status, SQ_WAVE_STATUS_POST_SPI_PRIO_SHIFT + s_setreg_b32 hwreg(HW_REG_STATUS, SQ_WAVE_STATUS_POST_SPI_PRIO_SHIFT, SQ_WAVE_STATUS_POST_SPI_PRIO_SIZE), tmp + s_nop 0x2 // avoid S_SETREG => S_SETREG hazard + s_setreg_b32 hwreg(HW_REG_STATUS, SQ_WAVE_STATUS_PRE_SPI_PRIO_SHIFT, SQ_WAVE_STATUS_PRE_SPI_PRIO_SIZE), status +end diff --git a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx9.asm b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx9.asm index 998be96be736..0bb9c577b3a2 100644 --- a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx9.asm +++ b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx9.asm @@ -103,6 +103,10 @@ var SQ_WAVE_STATUS_INST_ATC_MASK = 0x00800000 var SQ_WAVE_STATUS_SPI_PRIO_SHIFT = 1 var SQ_WAVE_STATUS_SPI_PRIO_MASK = 0x00000006 var SQ_WAVE_STATUS_HALT_MASK = 0x2000 +var SQ_WAVE_STATUS_PRE_SPI_PRIO_SHIFT = 0 +var SQ_WAVE_STATUS_PRE_SPI_PRIO_SIZE = 1 +var SQ_WAVE_STATUS_POST_SPI_PRIO_SHIFT = 3 +var SQ_WAVE_STATUS_POST_SPI_PRIO_SIZE = 29 var SQ_WAVE_LDS_ALLOC_LDS_SIZE_SHIFT = 12 var SQ_WAVE_LDS_ALLOC_LDS_SIZE_SIZE = 9 @@ -317,7 +321,7 @@ L_EXCP_CASE: // Restore SQ_WAVE_STATUS. s_and_b64 exec, exec, exec // Restore STATUS.EXECZ, not writable by s_setreg_b32 s_and_b64 vcc, vcc, vcc // Restore STATUS.VCCZ, not writable by s_setreg_b32 - s_setreg_b32 hwreg(HW_REG_STATUS), s_save_status + set_status_without_spi_prio(s_save_status, ttmp2) s_rfe_b64 [ttmp0, ttmp1] end @@ -1120,7 +1124,7 @@ end s_and_b32 s_restore_pc_hi, s_restore_pc_hi, 0x0000ffff //pc[47:32] //Do it here in order not to affect STATUS s_and_b64 exec, exec, exec // Restore STATUS.EXECZ, not writable by s_setreg_b32 s_and_b64 vcc, vcc, vcc // Restore STATUS.VCCZ, not writable by s_setreg_b32 - s_setreg_b32 hwreg(HW_REG_STATUS), s_restore_status // SCC is included, which is changed by previous salu + set_status_without_spi_prio(s_restore_status, s_restore_tmp) // SCC is included, which is changed by previous salu s_barrier //barrier to ensure the readiness of LDS before access attempts from any other wave in the same TG //FIXME not performance-optimal at this time @@ -1212,3 +1216,11 @@ function ack_sqc_store_workaround s_waitcnt lgkmcnt(0) end end + +function set_status_without_spi_prio(status, tmp) + // Do not restore STATUS.SPI_PRIO since scheduler may have raised it. + s_lshr_b32 tmp, status, SQ_WAVE_STATUS_POST_SPI_PRIO_SHIFT + s_setreg_b32 hwreg(HW_REG_STATUS, SQ_WAVE_STATUS_POST_SPI_PRIO_SHIFT, SQ_WAVE_STATUS_POST_SPI_PRIO_SIZE), tmp + s_nop 0x2 // avoid S_SETREG => S_SETREG hazard + s_setreg_b32 hwreg(HW_REG_STATUS, SQ_WAVE_STATUS_PRE_SPI_PRIO_SHIFT, SQ_WAVE_STATUS_PRE_SPI_PRIO_SIZE), status +end diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c index f64c5551cdba..297b36c26a05 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c @@ -122,6 +122,9 @@ static int kfd_open(struct inode *inode, struct file *filep) if (IS_ERR(process)) return PTR_ERR(process); + if (kfd_is_locked()) + return -EAGAIN; + dev_dbg(kfd_device, "process %d opened, compat mode (32 bit) - %d\n", process->pasid, process->is_32bit_user_mode); @@ -389,6 +392,61 @@ static int kfd_ioctl_update_queue(struct file *filp, struct kfd_process *p, return retval; } +static int kfd_ioctl_set_cu_mask(struct file *filp, struct kfd_process *p, + void *data) +{ + int retval; + const int max_num_cus = 1024; + struct kfd_ioctl_set_cu_mask_args *args = data; + struct queue_properties properties; + uint32_t __user *cu_mask_ptr = (uint32_t __user *)args->cu_mask_ptr; + size_t cu_mask_size = sizeof(uint32_t) * (args->num_cu_mask / 32); + + if ((args->num_cu_mask % 32) != 0) { + pr_debug("num_cu_mask 0x%x must be a multiple of 32", + args->num_cu_mask); + return -EINVAL; + } + + properties.cu_mask_count = args->num_cu_mask; + if (properties.cu_mask_count == 0) { + pr_debug("CU mask cannot be 0"); + return -EINVAL; + } + + /* To prevent an unreasonably large CU mask size, set an arbitrary + * limit of max_num_cus bits. We can then just drop any CU mask bits + * past max_num_cus bits and just use the first max_num_cus bits. + */ + if (properties.cu_mask_count > max_num_cus) { + pr_debug("CU mask cannot be greater than 1024 bits"); + properties.cu_mask_count = max_num_cus; + cu_mask_size = sizeof(uint32_t) * (max_num_cus/32); + } + + properties.cu_mask = kzalloc(cu_mask_size, GFP_KERNEL); + if (!properties.cu_mask) + return -ENOMEM; + + retval = copy_from_user(properties.cu_mask, cu_mask_ptr, cu_mask_size); + if (retval) { + pr_debug("Could not copy CU mask from userspace"); + kfree(properties.cu_mask); + return -EFAULT; + } + + mutex_lock(&p->mutex); + + retval = pqm_set_cu_mask(&p->pqm, args->queue_id, &properties); + + mutex_unlock(&p->mutex); + + if (retval) + kfree(properties.cu_mask); + + return retval; +} + static int kfd_ioctl_set_memory_policy(struct file *filep, struct kfd_process *p, void *data) { @@ -754,7 +812,6 @@ static int kfd_ioctl_get_clock_counters(struct file *filep, { struct kfd_ioctl_get_clock_counters_args *args = data; struct kfd_dev *dev; - struct timespec64 time; dev = kfd_device_by_id(args->gpu_id); if (dev) @@ -766,11 +823,8 @@ static int kfd_ioctl_get_clock_counters(struct file *filep, args->gpu_clock_counter = 0; /* No access to rdtsc. Using raw monotonic time */ - getrawmonotonic64(&time); - args->cpu_clock_counter = (uint64_t)timespec64_to_ns(&time); - - get_monotonic_boottime64(&time); - args->system_clock_counter = (uint64_t)timespec64_to_ns(&time); + args->cpu_clock_counter = ktime_get_raw_ns(); + args->system_clock_counter = ktime_get_boot_ns(); /* Since the counter is in nano-seconds we use 1GHz frequency */ args->system_clock_freq = 1000000000; @@ -1558,6 +1612,9 @@ static const struct amdkfd_ioctl_desc amdkfd_ioctls[] = { AMDKFD_IOCTL_DEF(AMDKFD_IOC_UNMAP_MEMORY_FROM_GPU, kfd_ioctl_unmap_memory_from_gpu, 0), + AMDKFD_IOCTL_DEF(AMDKFD_IOC_SET_CU_MASK, + kfd_ioctl_set_cu_mask, 0), + }; #define AMDKFD_CORE_IOCTL_COUNT ARRAY_SIZE(amdkfd_ioctls) diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_crat.c b/drivers/gpu/drm/amd/amdkfd/kfd_crat.c index 296b3f230280..ee4996029a86 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_crat.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_crat.c @@ -189,6 +189,21 @@ static int kfd_parse_subtype_cu(struct crat_subtype_computeunit *cu, return 0; } +static struct kfd_mem_properties * +find_subtype_mem(uint32_t heap_type, uint32_t flags, uint32_t width, + struct kfd_topology_device *dev) +{ + struct kfd_mem_properties *props; + + list_for_each_entry(props, &dev->mem_props, list) { + if (props->heap_type == heap_type + && props->flags == flags + && props->width == width) + return props; + } + + return NULL; +} /* kfd_parse_subtype_mem - parse memory subtypes and attach it to correct * topology device present in the device_list */ @@ -197,36 +212,56 @@ static int kfd_parse_subtype_mem(struct crat_subtype_memory *mem, { struct kfd_mem_properties *props; struct kfd_topology_device *dev; + uint32_t heap_type; + uint64_t size_in_bytes; + uint32_t flags = 0; + uint32_t width; pr_debug("Found memory entry in CRAT table with proximity_domain=%d\n", mem->proximity_domain); list_for_each_entry(dev, device_list, list) { if (mem->proximity_domain == dev->proximity_domain) { - props = kfd_alloc_struct(props); - if (!props) - return -ENOMEM; - /* We're on GPU node */ if (dev->node_props.cpu_cores_count == 0) { /* APU */ if (mem->visibility_type == 0) - props->heap_type = + heap_type = HSA_MEM_HEAP_TYPE_FB_PRIVATE; /* dGPU */ else - props->heap_type = mem->visibility_type; + heap_type = mem->visibility_type; } else - props->heap_type = HSA_MEM_HEAP_TYPE_SYSTEM; + heap_type = HSA_MEM_HEAP_TYPE_SYSTEM; if (mem->flags & CRAT_MEM_FLAGS_HOT_PLUGGABLE) - props->flags |= HSA_MEM_FLAGS_HOT_PLUGGABLE; + flags |= HSA_MEM_FLAGS_HOT_PLUGGABLE; if (mem->flags & CRAT_MEM_FLAGS_NON_VOLATILE) - props->flags |= HSA_MEM_FLAGS_NON_VOLATILE; + flags |= HSA_MEM_FLAGS_NON_VOLATILE; - props->size_in_bytes = + size_in_bytes = ((uint64_t)mem->length_high << 32) + mem->length_low; - props->width = mem->width; + width = mem->width; + + /* Multiple banks of the same type are aggregated into + * one. User mode doesn't care about multiple physical + * memory segments. It's managed as a single virtual + * heap for user mode. + */ + props = find_subtype_mem(heap_type, flags, width, dev); + if (props) { + props->size_in_bytes += size_in_bytes; + break; + } + + props = kfd_alloc_struct(props); + if (!props) + return -ENOMEM; + + props->heap_type = heap_type; + props->flags = flags; + props->size_in_bytes = size_in_bytes; + props->width = width; dev->node_props.mem_banks_count++; list_add_tail(&props->list, &dev->mem_props); diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_dbgdev.c b/drivers/gpu/drm/amd/amdkfd/kfd_dbgdev.c index afb26f205d29..a3441b0e385b 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_dbgdev.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_dbgdev.c @@ -38,7 +38,6 @@ #include "kfd_dbgmgr.h" #include "kfd_dbgdev.h" #include "kfd_device_queue_manager.h" -#include "../../radeon/cik_reg.h" static void dbgdev_address_watch_disable_nodiq(struct kfd_dev *dev) { diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_dbgdev.h b/drivers/gpu/drm/amd/amdkfd/kfd_dbgdev.h index 03424c20920c..0619c777b47e 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_dbgdev.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_dbgdev.h @@ -60,6 +60,9 @@ enum { SH_REG_SIZE = SH_REG_END - SH_REG_BASE }; +/* SQ_CMD definitions */ +#define SQ_CMD 0x8DEC + enum SQ_IND_CMD_CMD { SQ_IND_CMD_CMD_NULL = 0x00000000, SQ_IND_CMD_CMD_HALT = 0x00000001, @@ -190,4 +193,38 @@ union ULARGE_INTEGER { void kfd_dbgdev_init(struct kfd_dbgdev *pdbgdev, struct kfd_dev *pdev, enum DBGDEV_TYPE type); +union TCP_WATCH_CNTL_BITS { + struct { + uint32_t mask:24; + uint32_t vmid:4; + uint32_t atc:1; + uint32_t mode:2; + uint32_t valid:1; + } bitfields, bits; + uint32_t u32All; + signed int i32All; + float f32All; +}; + +enum { + ADDRESS_WATCH_REG_CNTL_ATC_BIT = 0x10000000UL, + ADDRESS_WATCH_REG_CNTL_DEFAULT_MASK = 0x00FFFFFF, + ADDRESS_WATCH_REG_ADDLOW_MASK_EXTENSION = 0x03000000, + /* extend the mask to 26 bits in order to match the low address field */ + ADDRESS_WATCH_REG_ADDLOW_SHIFT = 6, + ADDRESS_WATCH_REG_ADDHIGH_MASK = 0xFFFF +}; + +enum { + MAX_TRAPID = 8, /* 3 bits in the bitfield. */ + MAX_WATCH_ADDRESSES = 4 +}; + +enum { + ADDRESS_WATCH_REG_ADDR_HI = 0, + ADDRESS_WATCH_REG_ADDR_LO, + ADDRESS_WATCH_REG_CNTL, + ADDRESS_WATCH_REG_MAX +}; + #endif /* KFD_DBGDEV_H_ */ diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_debugfs.c b/drivers/gpu/drm/amd/amdkfd/kfd_debugfs.c index 4bd6ebfaf425..ab37d36d9cd6 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_debugfs.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_debugfs.c @@ -21,6 +21,8 @@ */ #include <linux/debugfs.h> +#include <linux/uaccess.h> + #include "kfd_priv.h" static struct dentry *debugfs_root; @@ -32,6 +34,38 @@ static int kfd_debugfs_open(struct inode *inode, struct file *file) return single_open(file, show, NULL); } +static ssize_t kfd_debugfs_hang_hws_write(struct file *file, + const char __user *user_buf, size_t size, loff_t *ppos) +{ + struct kfd_dev *dev; + char tmp[16]; + uint32_t gpu_id; + int ret = -EINVAL; + + memset(tmp, 0, 16); + if (size >= 16) { + pr_err("Invalid input for gpu id.\n"); + goto out; + } + if (copy_from_user(tmp, user_buf, size)) { + ret = -EFAULT; + goto out; + } + if (kstrtoint(tmp, 10, &gpu_id)) { + pr_err("Invalid input for gpu id.\n"); + goto out; + } + dev = kfd_device_by_id(gpu_id); + if (dev) { + kfd_debugfs_hang_hws(dev); + ret = size; + } else + pr_err("Cannot find device %d.\n", gpu_id); + +out: + return ret; +} + static const struct file_operations kfd_debugfs_fops = { .owner = THIS_MODULE, .open = kfd_debugfs_open, @@ -40,6 +74,15 @@ static const struct file_operations kfd_debugfs_fops = { .release = single_release, }; +static const struct file_operations kfd_debugfs_hang_hws_fops = { + .owner = THIS_MODULE, + .open = kfd_debugfs_open, + .read = seq_read, + .write = kfd_debugfs_hang_hws_write, + .llseek = seq_lseek, + .release = single_release, +}; + void kfd_debugfs_init(void) { struct dentry *ent; @@ -65,6 +108,11 @@ void kfd_debugfs_init(void) ent = debugfs_create_file("rls", S_IFREG | 0444, debugfs_root, kfd_debugfs_rls_by_device, &kfd_debugfs_fops); + + ent = debugfs_create_file("hang_hws", S_IFREG | 0644, debugfs_root, + NULL, + &kfd_debugfs_hang_hws_fops); + if (!ent) pr_warn("Failed to create rls in kfd debugfs\n"); } diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device.c b/drivers/gpu/drm/amd/amdkfd/kfd_device.c index 7ee6cec2c060..1b048715ab8a 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_device.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device.c @@ -30,7 +30,13 @@ #include "kfd_iommu.h" #define MQD_SIZE_ALIGNED 768 -static atomic_t kfd_device_suspended = ATOMIC_INIT(0); + +/* + * kfd_locked is used to lock the kfd driver during suspend or reset + * once locked, kfd driver will stop any further GPU execution. + * create process (open) will return -EAGAIN. + */ +static atomic_t kfd_locked = ATOMIC_INIT(0); #ifdef KFD_SUPPORT_IOMMU_V2 static const struct kfd_device_info kaveri_device_info = { @@ -46,6 +52,7 @@ static const struct kfd_device_info kaveri_device_info = { .supports_cwsr = false, .needs_iommu_device = true, .needs_pci_atomics = false, + .num_sdma_engines = 2, }; static const struct kfd_device_info carrizo_device_info = { @@ -61,6 +68,22 @@ static const struct kfd_device_info carrizo_device_info = { .supports_cwsr = true, .needs_iommu_device = true, .needs_pci_atomics = false, + .num_sdma_engines = 2, +}; + +static const struct kfd_device_info raven_device_info = { + .asic_family = CHIP_RAVEN, + .max_pasid_bits = 16, + .max_no_of_hqd = 24, + .doorbell_size = 8, + .ih_ring_entry_size = 8 * sizeof(uint32_t), + .event_interrupt_class = &event_interrupt_class_v9, + .num_of_watch_points = 4, + .mqd_size_aligned = MQD_SIZE_ALIGNED, + .supports_cwsr = true, + .needs_iommu_device = true, + .needs_pci_atomics = true, + .num_sdma_engines = 1, }; #endif @@ -77,6 +100,7 @@ static const struct kfd_device_info hawaii_device_info = { .supports_cwsr = false, .needs_iommu_device = false, .needs_pci_atomics = false, + .num_sdma_engines = 2, }; static const struct kfd_device_info tonga_device_info = { @@ -91,6 +115,7 @@ static const struct kfd_device_info tonga_device_info = { .supports_cwsr = false, .needs_iommu_device = false, .needs_pci_atomics = true, + .num_sdma_engines = 2, }; static const struct kfd_device_info tonga_vf_device_info = { @@ -105,6 +130,7 @@ static const struct kfd_device_info tonga_vf_device_info = { .supports_cwsr = false, .needs_iommu_device = false, .needs_pci_atomics = false, + .num_sdma_engines = 2, }; static const struct kfd_device_info fiji_device_info = { @@ -119,6 +145,7 @@ static const struct kfd_device_info fiji_device_info = { .supports_cwsr = true, .needs_iommu_device = false, .needs_pci_atomics = true, + .num_sdma_engines = 2, }; static const struct kfd_device_info fiji_vf_device_info = { @@ -133,6 +160,7 @@ static const struct kfd_device_info fiji_vf_device_info = { .supports_cwsr = true, .needs_iommu_device = false, .needs_pci_atomics = false, + .num_sdma_engines = 2, }; @@ -148,6 +176,7 @@ static const struct kfd_device_info polaris10_device_info = { .supports_cwsr = true, .needs_iommu_device = false, .needs_pci_atomics = true, + .num_sdma_engines = 2, }; static const struct kfd_device_info polaris10_vf_device_info = { @@ -162,6 +191,7 @@ static const struct kfd_device_info polaris10_vf_device_info = { .supports_cwsr = true, .needs_iommu_device = false, .needs_pci_atomics = false, + .num_sdma_engines = 2, }; static const struct kfd_device_info polaris11_device_info = { @@ -176,6 +206,7 @@ static const struct kfd_device_info polaris11_device_info = { .supports_cwsr = true, .needs_iommu_device = false, .needs_pci_atomics = true, + .num_sdma_engines = 2, }; static const struct kfd_device_info vega10_device_info = { @@ -190,6 +221,7 @@ static const struct kfd_device_info vega10_device_info = { .supports_cwsr = true, .needs_iommu_device = false, .needs_pci_atomics = false, + .num_sdma_engines = 2, }; static const struct kfd_device_info vega10_vf_device_info = { @@ -204,6 +236,7 @@ static const struct kfd_device_info vega10_vf_device_info = { .supports_cwsr = true, .needs_iommu_device = false, .needs_pci_atomics = false, + .num_sdma_engines = 2, }; @@ -241,6 +274,7 @@ static const struct kfd_deviceid supported_devices[] = { { 0x9875, &carrizo_device_info }, /* Carrizo */ { 0x9876, &carrizo_device_info }, /* Carrizo */ { 0x9877, &carrizo_device_info }, /* Carrizo */ + { 0x15DD, &raven_device_info }, /* Raven */ #endif { 0x67A0, &hawaii_device_info }, /* Hawaii */ { 0x67A1, &hawaii_device_info }, /* Hawaii */ @@ -514,13 +548,54 @@ void kgd2kfd_device_exit(struct kfd_dev *kfd) kfree(kfd); } +int kgd2kfd_pre_reset(struct kfd_dev *kfd) +{ + if (!kfd->init_complete) + return 0; + kgd2kfd_suspend(kfd); + + /* hold dqm->lock to prevent further execution*/ + dqm_lock(kfd->dqm); + + kfd_signal_reset_event(kfd); + return 0; +} + +/* + * Fix me. KFD won't be able to resume existing process for now. + * We will keep all existing process in a evicted state and + * wait the process to be terminated. + */ + +int kgd2kfd_post_reset(struct kfd_dev *kfd) +{ + int ret, count; + + if (!kfd->init_complete) + return 0; + + dqm_unlock(kfd->dqm); + + ret = kfd_resume(kfd); + if (ret) + return ret; + count = atomic_dec_return(&kfd_locked); + WARN_ONCE(count != 0, "KFD reset ref. error"); + return 0; +} + +bool kfd_is_locked(void) +{ + return (atomic_read(&kfd_locked) > 0); +} + void kgd2kfd_suspend(struct kfd_dev *kfd) { if (!kfd->init_complete) return; /* For first KFD device suspend all the KFD processes */ - if (atomic_inc_return(&kfd_device_suspended) == 1) + if (atomic_inc_return(&kfd_locked) == 1) kfd_suspend_all_processes(); kfd->dqm->ops.stop(kfd->dqm); @@ -539,7 +614,7 @@ int kgd2kfd_resume(struct kfd_dev *kfd) if (ret) return ret; - count = atomic_dec_return(&kfd_device_suspended); + count = atomic_dec_return(&kfd_locked); WARN_ONCE(count < 0, "KFD suspend / resume ref. error"); if (count == 0) ret = kfd_resume_all_processes(); @@ -577,14 +652,24 @@ dqm_start_error: /* This is called directly from KGD at ISR. */ void kgd2kfd_interrupt(struct kfd_dev *kfd, const void *ih_ring_entry) { + uint32_t patched_ihre[KFD_MAX_RING_ENTRY_SIZE]; + bool is_patched = false; + if (!kfd->init_complete) return; + if (kfd->device_info->ih_ring_entry_size > sizeof(patched_ihre)) { + dev_err_once(kfd_device, "Ring entry too small\n"); + return; + } + spin_lock(&kfd->interrupt_lock); if (kfd->interrupts_active - && interrupt_is_wanted(kfd, ih_ring_entry) - && enqueue_ih_ring_entry(kfd, ih_ring_entry)) + && interrupt_is_wanted(kfd, ih_ring_entry, + patched_ihre, &is_patched) + && enqueue_ih_ring_entry(kfd, + is_patched ? patched_ihre : ih_ring_entry)) queue_work(kfd->ih_wq, &kfd->interrupt_work); spin_unlock(&kfd->interrupt_lock); @@ -739,8 +824,8 @@ int kfd_gtt_sa_allocate(struct kfd_dev *kfd, unsigned int size, if (size > kfd->gtt_sa_num_of_chunks * kfd->gtt_sa_chunk_size) return -ENOMEM; - *mem_obj = kzalloc(sizeof(struct kfd_mem_obj), GFP_NOIO); - if ((*mem_obj) == NULL) + *mem_obj = kzalloc(sizeof(struct kfd_mem_obj), GFP_KERNEL); + if (!(*mem_obj)) return -ENOMEM; pr_debug("Allocated mem_obj = %p for size = %d\n", *mem_obj, size); @@ -857,3 +942,26 @@ int kfd_gtt_sa_free(struct kfd_dev *kfd, struct kfd_mem_obj *mem_obj) kfree(mem_obj); return 0; } + +#if defined(CONFIG_DEBUG_FS) + +/* This function will send a package to HIQ to hang the HWS + * which will trigger a GPU reset and bring the HWS back to normal state + */ +int kfd_debugfs_hang_hws(struct kfd_dev *dev) +{ + int r = 0; + + if (dev->dqm->sched_policy != KFD_SCHED_POLICY_HWS) { + pr_err("HWS is not enabled"); + return -EINVAL; + } + + r = pm_debugfs_hang_hws(&dev->dqm->packets); + if (!r) + r = dqm_debugfs_execute_queues(dev->dqm); + + return r; +} + +#endif diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c index 668ad07ebe1f..ec0d62a16e53 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c @@ -61,6 +61,8 @@ static int create_sdma_queue_nocpsch(struct device_queue_manager *dqm, static void deallocate_sdma_queue(struct device_queue_manager *dqm, unsigned int sdma_queue_id); +static void kfd_process_hw_exception(struct work_struct *work); + static inline enum KFD_MQD_TYPE get_mqd_type_from_queue_type(enum kfd_queue_type type) { @@ -99,6 +101,17 @@ unsigned int get_pipes_per_mec(struct device_queue_manager *dqm) return dqm->dev->shared_resources.num_pipe_per_mec; } +static unsigned int get_num_sdma_engines(struct device_queue_manager *dqm) +{ + return dqm->dev->device_info->num_sdma_engines; +} + +unsigned int get_num_sdma_queues(struct device_queue_manager *dqm) +{ + return dqm->dev->device_info->num_sdma_engines + * KFD_SDMA_QUEUES_PER_ENGINE; +} + void program_sh_mem_settings(struct device_queue_manager *dqm, struct qcm_process_device *qpd) { @@ -240,7 +253,7 @@ static int create_queue_nocpsch(struct device_queue_manager *dqm, print_queue(q); - mutex_lock(&dqm->lock); + dqm_lock(dqm); if (dqm->total_queue_count >= max_num_of_queues_per_device) { pr_warn("Can't create new usermode queue because %d queues were already created\n", @@ -297,7 +310,7 @@ static int create_queue_nocpsch(struct device_queue_manager *dqm, dqm->total_queue_count); out_unlock: - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); return retval; } @@ -346,10 +359,10 @@ static int create_compute_queue_nocpsch(struct device_queue_manager *dqm, struct qcm_process_device *qpd) { int retval; - struct mqd_manager *mqd; + struct mqd_manager *mqd_mgr; - mqd = dqm->ops.get_mqd_manager(dqm, KFD_MQD_TYPE_COMPUTE); - if (!mqd) + mqd_mgr = dqm->ops.get_mqd_manager(dqm, KFD_MQD_TYPE_COMPUTE); + if (!mqd_mgr) return -ENOMEM; retval = allocate_hqd(dqm, q); @@ -360,7 +373,7 @@ static int create_compute_queue_nocpsch(struct device_queue_manager *dqm, if (retval) goto out_deallocate_hqd; - retval = mqd->init_mqd(mqd, &q->mqd, &q->mqd_mem_obj, + retval = mqd_mgr->init_mqd(mqd_mgr, &q->mqd, &q->mqd_mem_obj, &q->gart_mqd_addr, &q->properties); if (retval) goto out_deallocate_doorbell; @@ -374,15 +387,15 @@ static int create_compute_queue_nocpsch(struct device_queue_manager *dqm, if (!q->properties.is_active) return 0; - retval = mqd->load_mqd(mqd, q->mqd, q->pipe, q->queue, &q->properties, - q->process->mm); + retval = mqd_mgr->load_mqd(mqd_mgr, q->mqd, q->pipe, q->queue, + &q->properties, q->process->mm); if (retval) goto out_uninit_mqd; return 0; out_uninit_mqd: - mqd->uninit_mqd(mqd, q->mqd, q->mqd_mem_obj); + mqd_mgr->uninit_mqd(mqd_mgr, q->mqd, q->mqd_mem_obj); out_deallocate_doorbell: deallocate_doorbell(qpd, q); out_deallocate_hqd: @@ -399,11 +412,11 @@ static int destroy_queue_nocpsch_locked(struct device_queue_manager *dqm, struct queue *q) { int retval; - struct mqd_manager *mqd; + struct mqd_manager *mqd_mgr; - mqd = dqm->ops.get_mqd_manager(dqm, + mqd_mgr = dqm->ops.get_mqd_manager(dqm, get_mqd_type_from_queue_type(q->properties.type)); - if (!mqd) + if (!mqd_mgr) return -ENOMEM; if (q->properties.type == KFD_QUEUE_TYPE_COMPUTE) { @@ -420,14 +433,14 @@ static int destroy_queue_nocpsch_locked(struct device_queue_manager *dqm, deallocate_doorbell(qpd, q); - retval = mqd->destroy_mqd(mqd, q->mqd, + retval = mqd_mgr->destroy_mqd(mqd_mgr, q->mqd, KFD_PREEMPT_TYPE_WAVEFRONT_RESET, KFD_UNMAP_LATENCY_MS, q->pipe, q->queue); if (retval == -ETIME) qpd->reset_wavefronts = true; - mqd->uninit_mqd(mqd, q->mqd, q->mqd_mem_obj); + mqd_mgr->uninit_mqd(mqd_mgr, q->mqd, q->mqd_mem_obj); list_del(&q->list); if (list_empty(&qpd->queues_list)) { @@ -457,9 +470,9 @@ static int destroy_queue_nocpsch(struct device_queue_manager *dqm, { int retval; - mutex_lock(&dqm->lock); + dqm_lock(dqm); retval = destroy_queue_nocpsch_locked(dqm, qpd, q); - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); return retval; } @@ -467,19 +480,19 @@ static int destroy_queue_nocpsch(struct device_queue_manager *dqm, static int update_queue(struct device_queue_manager *dqm, struct queue *q) { int retval; - struct mqd_manager *mqd; + struct mqd_manager *mqd_mgr; struct kfd_process_device *pdd; bool prev_active = false; - mutex_lock(&dqm->lock); + dqm_lock(dqm); pdd = kfd_get_process_device_data(q->device, q->process); if (!pdd) { retval = -ENODEV; goto out_unlock; } - mqd = dqm->ops.get_mqd_manager(dqm, + mqd_mgr = dqm->ops.get_mqd_manager(dqm, get_mqd_type_from_queue_type(q->properties.type)); - if (!mqd) { + if (!mqd_mgr) { retval = -ENOMEM; goto out_unlock; } @@ -506,7 +519,7 @@ static int update_queue(struct device_queue_manager *dqm, struct queue *q) } else if (prev_active && (q->properties.type == KFD_QUEUE_TYPE_COMPUTE || q->properties.type == KFD_QUEUE_TYPE_SDMA)) { - retval = mqd->destroy_mqd(mqd, q->mqd, + retval = mqd_mgr->destroy_mqd(mqd_mgr, q->mqd, KFD_PREEMPT_TYPE_WAVEFRONT_DRAIN, KFD_UNMAP_LATENCY_MS, q->pipe, q->queue); if (retval) { @@ -515,7 +528,7 @@ static int update_queue(struct device_queue_manager *dqm, struct queue *q) } } - retval = mqd->update_mqd(mqd, q->mqd, &q->properties); + retval = mqd_mgr->update_mqd(mqd_mgr, q->mqd, &q->properties); /* * check active state vs. the previous state and modify @@ -533,44 +546,44 @@ static int update_queue(struct device_queue_manager *dqm, struct queue *q) else if (q->properties.is_active && (q->properties.type == KFD_QUEUE_TYPE_COMPUTE || q->properties.type == KFD_QUEUE_TYPE_SDMA)) - retval = mqd->load_mqd(mqd, q->mqd, q->pipe, q->queue, + retval = mqd_mgr->load_mqd(mqd_mgr, q->mqd, q->pipe, q->queue, &q->properties, q->process->mm); out_unlock: - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); return retval; } static struct mqd_manager *get_mqd_manager( struct device_queue_manager *dqm, enum KFD_MQD_TYPE type) { - struct mqd_manager *mqd; + struct mqd_manager *mqd_mgr; if (WARN_ON(type >= KFD_MQD_TYPE_MAX)) return NULL; pr_debug("mqd type %d\n", type); - mqd = dqm->mqds[type]; - if (!mqd) { - mqd = mqd_manager_init(type, dqm->dev); - if (!mqd) + mqd_mgr = dqm->mqd_mgrs[type]; + if (!mqd_mgr) { + mqd_mgr = mqd_manager_init(type, dqm->dev); + if (!mqd_mgr) pr_err("mqd manager is NULL"); - dqm->mqds[type] = mqd; + dqm->mqd_mgrs[type] = mqd_mgr; } - return mqd; + return mqd_mgr; } static int evict_process_queues_nocpsch(struct device_queue_manager *dqm, struct qcm_process_device *qpd) { struct queue *q; - struct mqd_manager *mqd; + struct mqd_manager *mqd_mgr; struct kfd_process_device *pdd; int retval = 0; - mutex_lock(&dqm->lock); + dqm_lock(dqm); if (qpd->evicted++ > 0) /* already evicted, do nothing */ goto out; @@ -582,16 +595,16 @@ static int evict_process_queues_nocpsch(struct device_queue_manager *dqm, list_for_each_entry(q, &qpd->queues_list, list) { if (!q->properties.is_active) continue; - mqd = dqm->ops.get_mqd_manager(dqm, + mqd_mgr = dqm->ops.get_mqd_manager(dqm, get_mqd_type_from_queue_type(q->properties.type)); - if (!mqd) { /* should not be here */ + if (!mqd_mgr) { /* should not be here */ pr_err("Cannot evict queue, mqd mgr is NULL\n"); retval = -ENOMEM; goto out; } q->properties.is_evicted = true; q->properties.is_active = false; - retval = mqd->destroy_mqd(mqd, q->mqd, + retval = mqd_mgr->destroy_mqd(mqd_mgr, q->mqd, KFD_PREEMPT_TYPE_WAVEFRONT_DRAIN, KFD_UNMAP_LATENCY_MS, q->pipe, q->queue); if (retval) @@ -600,7 +613,7 @@ static int evict_process_queues_nocpsch(struct device_queue_manager *dqm, } out: - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); return retval; } @@ -611,7 +624,7 @@ static int evict_process_queues_cpsch(struct device_queue_manager *dqm, struct kfd_process_device *pdd; int retval = 0; - mutex_lock(&dqm->lock); + dqm_lock(dqm); if (qpd->evicted++ > 0) /* already evicted, do nothing */ goto out; @@ -633,7 +646,7 @@ static int evict_process_queues_cpsch(struct device_queue_manager *dqm, KFD_UNMAP_QUEUES_FILTER_DYNAMIC_QUEUES, 0); out: - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); return retval; } @@ -641,7 +654,7 @@ static int restore_process_queues_nocpsch(struct device_queue_manager *dqm, struct qcm_process_device *qpd) { struct queue *q; - struct mqd_manager *mqd; + struct mqd_manager *mqd_mgr; struct kfd_process_device *pdd; uint32_t pd_base; int retval = 0; @@ -650,7 +663,7 @@ static int restore_process_queues_nocpsch(struct device_queue_manager *dqm, /* Retrieve PD base */ pd_base = dqm->dev->kfd2kgd->get_process_page_dir(pdd->vm); - mutex_lock(&dqm->lock); + dqm_lock(dqm); if (WARN_ON_ONCE(!qpd->evicted)) /* already restored, do nothing */ goto out; if (qpd->evicted > 1) { /* ref count still > 0, decrement & quit */ @@ -677,16 +690,16 @@ static int restore_process_queues_nocpsch(struct device_queue_manager *dqm, list_for_each_entry(q, &qpd->queues_list, list) { if (!q->properties.is_evicted) continue; - mqd = dqm->ops.get_mqd_manager(dqm, + mqd_mgr = dqm->ops.get_mqd_manager(dqm, get_mqd_type_from_queue_type(q->properties.type)); - if (!mqd) { /* should not be here */ + if (!mqd_mgr) { /* should not be here */ pr_err("Cannot restore queue, mqd mgr is NULL\n"); retval = -ENOMEM; goto out; } q->properties.is_evicted = false; q->properties.is_active = true; - retval = mqd->load_mqd(mqd, q->mqd, q->pipe, + retval = mqd_mgr->load_mqd(mqd_mgr, q->mqd, q->pipe, q->queue, &q->properties, q->process->mm); if (retval) @@ -695,7 +708,7 @@ static int restore_process_queues_nocpsch(struct device_queue_manager *dqm, } qpd->evicted = 0; out: - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); return retval; } @@ -711,7 +724,7 @@ static int restore_process_queues_cpsch(struct device_queue_manager *dqm, /* Retrieve PD base */ pd_base = dqm->dev->kfd2kgd->get_process_page_dir(pdd->vm); - mutex_lock(&dqm->lock); + dqm_lock(dqm); if (WARN_ON_ONCE(!qpd->evicted)) /* already restored, do nothing */ goto out; if (qpd->evicted > 1) { /* ref count still > 0, decrement & quit */ @@ -739,7 +752,7 @@ static int restore_process_queues_cpsch(struct device_queue_manager *dqm, if (!retval) qpd->evicted = 0; out: - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); return retval; } @@ -761,7 +774,7 @@ static int register_process(struct device_queue_manager *dqm, /* Retrieve PD base */ pd_base = dqm->dev->kfd2kgd->get_process_page_dir(pdd->vm); - mutex_lock(&dqm->lock); + dqm_lock(dqm); list_add(&n->list, &dqm->queues); /* Update PD Base in QPD */ @@ -769,9 +782,10 @@ static int register_process(struct device_queue_manager *dqm, retval = dqm->asic_ops.update_qpd(dqm, qpd); - dqm->processes_count++; + if (dqm->processes_count++ == 0) + dqm->dev->kfd2kgd->set_compute_idle(dqm->dev->kgd, false); - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); return retval; } @@ -786,20 +800,22 @@ static int unregister_process(struct device_queue_manager *dqm, list_empty(&qpd->queues_list) ? "empty" : "not empty"); retval = 0; - mutex_lock(&dqm->lock); + dqm_lock(dqm); list_for_each_entry_safe(cur, next, &dqm->queues, list) { if (qpd == cur->qpd) { list_del(&cur->list); kfree(cur); - dqm->processes_count--; + if (--dqm->processes_count == 0) + dqm->dev->kfd2kgd->set_compute_idle( + dqm->dev->kgd, true); goto out; } } /* qpd not found in dqm list */ retval = 1; out: - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); return retval; } @@ -838,7 +854,7 @@ static int initialize_nocpsch(struct device_queue_manager *dqm) if (!dqm->allocated_queues) return -ENOMEM; - mutex_init(&dqm->lock); + mutex_init(&dqm->lock_hidden); INIT_LIST_HEAD(&dqm->queues); dqm->queue_count = dqm->next_pipe_to_allocate = 0; dqm->sdma_queue_count = 0; @@ -853,7 +869,7 @@ static int initialize_nocpsch(struct device_queue_manager *dqm) } dqm->vmid_bitmap = (1 << dqm->dev->vm_info.vmid_num_kfd) - 1; - dqm->sdma_bitmap = (1 << CIK_SDMA_QUEUES) - 1; + dqm->sdma_bitmap = (1 << get_num_sdma_queues(dqm)) - 1; return 0; } @@ -866,8 +882,8 @@ static void uninitialize(struct device_queue_manager *dqm) kfree(dqm->allocated_queues); for (i = 0 ; i < KFD_MQD_TYPE_MAX ; i++) - kfree(dqm->mqds[i]); - mutex_destroy(&dqm->lock); + kfree(dqm->mqd_mgrs[i]); + mutex_destroy(&dqm->lock_hidden); kfd_gtt_sa_free(dqm->dev, dqm->pipeline_mem); } @@ -901,7 +917,7 @@ static int allocate_sdma_queue(struct device_queue_manager *dqm, static void deallocate_sdma_queue(struct device_queue_manager *dqm, unsigned int sdma_queue_id) { - if (sdma_queue_id >= CIK_SDMA_QUEUES) + if (sdma_queue_id >= get_num_sdma_queues(dqm)) return; dqm->sdma_bitmap |= (1 << sdma_queue_id); } @@ -910,19 +926,19 @@ static int create_sdma_queue_nocpsch(struct device_queue_manager *dqm, struct queue *q, struct qcm_process_device *qpd) { - struct mqd_manager *mqd; + struct mqd_manager *mqd_mgr; int retval; - mqd = dqm->ops.get_mqd_manager(dqm, KFD_MQD_TYPE_SDMA); - if (!mqd) + mqd_mgr = dqm->ops.get_mqd_manager(dqm, KFD_MQD_TYPE_SDMA); + if (!mqd_mgr) return -ENOMEM; retval = allocate_sdma_queue(dqm, &q->sdma_id); if (retval) return retval; - q->properties.sdma_queue_id = q->sdma_id / CIK_SDMA_QUEUES_PER_ENGINE; - q->properties.sdma_engine_id = q->sdma_id % CIK_SDMA_QUEUES_PER_ENGINE; + q->properties.sdma_queue_id = q->sdma_id / get_num_sdma_engines(dqm); + q->properties.sdma_engine_id = q->sdma_id % get_num_sdma_engines(dqm); retval = allocate_doorbell(qpd, q); if (retval) @@ -933,19 +949,20 @@ static int create_sdma_queue_nocpsch(struct device_queue_manager *dqm, pr_debug("SDMA engine id: %d\n", q->properties.sdma_engine_id); dqm->asic_ops.init_sdma_vm(dqm, q, qpd); - retval = mqd->init_mqd(mqd, &q->mqd, &q->mqd_mem_obj, + retval = mqd_mgr->init_mqd(mqd_mgr, &q->mqd, &q->mqd_mem_obj, &q->gart_mqd_addr, &q->properties); if (retval) goto out_deallocate_doorbell; - retval = mqd->load_mqd(mqd, q->mqd, 0, 0, &q->properties, NULL); + retval = mqd_mgr->load_mqd(mqd_mgr, q->mqd, 0, 0, &q->properties, + NULL); if (retval) goto out_uninit_mqd; return 0; out_uninit_mqd: - mqd->uninit_mqd(mqd, q->mqd, q->mqd_mem_obj); + mqd_mgr->uninit_mqd(mqd_mgr, q->mqd, q->mqd_mem_obj); out_deallocate_doorbell: deallocate_doorbell(qpd, q); out_deallocate_sdma_queue: @@ -1003,12 +1020,14 @@ static int initialize_cpsch(struct device_queue_manager *dqm) { pr_debug("num of pipes: %d\n", get_pipes_per_mec(dqm)); - mutex_init(&dqm->lock); + mutex_init(&dqm->lock_hidden); INIT_LIST_HEAD(&dqm->queues); dqm->queue_count = dqm->processes_count = 0; dqm->sdma_queue_count = 0; dqm->active_runlist = false; - dqm->sdma_bitmap = (1 << CIK_SDMA_QUEUES) - 1; + dqm->sdma_bitmap = (1 << get_num_sdma_queues(dqm)) - 1; + + INIT_WORK(&dqm->hw_exception_work, kfd_process_hw_exception); return 0; } @@ -1041,9 +1060,11 @@ static int start_cpsch(struct device_queue_manager *dqm) init_interrupts(dqm); - mutex_lock(&dqm->lock); + dqm_lock(dqm); + /* clear hang status when driver try to start the hw scheduler */ + dqm->is_hws_hang = false; execute_queues_cpsch(dqm, KFD_UNMAP_QUEUES_FILTER_DYNAMIC_QUEUES, 0); - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); return 0; fail_allocate_vidmem: @@ -1055,9 +1076,9 @@ fail_packet_manager_init: static int stop_cpsch(struct device_queue_manager *dqm) { - mutex_lock(&dqm->lock); + dqm_lock(dqm); unmap_queues_cpsch(dqm, KFD_UNMAP_QUEUES_FILTER_ALL_QUEUES, 0); - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); kfd_gtt_sa_free(dqm->dev, dqm->fence_mem); pm_uninit(&dqm->packets); @@ -1069,11 +1090,11 @@ static int create_kernel_queue_cpsch(struct device_queue_manager *dqm, struct kernel_queue *kq, struct qcm_process_device *qpd) { - mutex_lock(&dqm->lock); + dqm_lock(dqm); if (dqm->total_queue_count >= max_num_of_queues_per_device) { pr_warn("Can't create new kernel queue because %d queues were already created\n", dqm->total_queue_count); - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); return -EPERM; } @@ -1089,7 +1110,7 @@ static int create_kernel_queue_cpsch(struct device_queue_manager *dqm, dqm->queue_count++; qpd->is_debug = true; execute_queues_cpsch(dqm, KFD_UNMAP_QUEUES_FILTER_DYNAMIC_QUEUES, 0); - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); return 0; } @@ -1098,7 +1119,7 @@ static void destroy_kernel_queue_cpsch(struct device_queue_manager *dqm, struct kernel_queue *kq, struct qcm_process_device *qpd) { - mutex_lock(&dqm->lock); + dqm_lock(dqm); list_del(&kq->list); dqm->queue_count--; qpd->is_debug = false; @@ -1110,18 +1131,18 @@ static void destroy_kernel_queue_cpsch(struct device_queue_manager *dqm, dqm->total_queue_count--; pr_debug("Total of %d queues are accountable so far\n", dqm->total_queue_count); - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); } static int create_queue_cpsch(struct device_queue_manager *dqm, struct queue *q, struct qcm_process_device *qpd) { int retval; - struct mqd_manager *mqd; + struct mqd_manager *mqd_mgr; retval = 0; - mutex_lock(&dqm->lock); + dqm_lock(dqm); if (dqm->total_queue_count >= max_num_of_queues_per_device) { pr_warn("Can't create new usermode queue because %d queues were already created\n", @@ -1135,19 +1156,19 @@ static int create_queue_cpsch(struct device_queue_manager *dqm, struct queue *q, if (retval) goto out_unlock; q->properties.sdma_queue_id = - q->sdma_id / CIK_SDMA_QUEUES_PER_ENGINE; + q->sdma_id / get_num_sdma_engines(dqm); q->properties.sdma_engine_id = - q->sdma_id % CIK_SDMA_QUEUES_PER_ENGINE; + q->sdma_id % get_num_sdma_engines(dqm); } retval = allocate_doorbell(qpd, q); if (retval) goto out_deallocate_sdma_queue; - mqd = dqm->ops.get_mqd_manager(dqm, + mqd_mgr = dqm->ops.get_mqd_manager(dqm, get_mqd_type_from_queue_type(q->properties.type)); - if (!mqd) { + if (!mqd_mgr) { retval = -ENOMEM; goto out_deallocate_doorbell; } @@ -1164,7 +1185,7 @@ static int create_queue_cpsch(struct device_queue_manager *dqm, struct queue *q, q->properties.tba_addr = qpd->tba_addr; q->properties.tma_addr = qpd->tma_addr; - retval = mqd->init_mqd(mqd, &q->mqd, &q->mqd_mem_obj, + retval = mqd_mgr->init_mqd(mqd_mgr, &q->mqd, &q->mqd_mem_obj, &q->gart_mqd_addr, &q->properties); if (retval) goto out_deallocate_doorbell; @@ -1188,7 +1209,7 @@ static int create_queue_cpsch(struct device_queue_manager *dqm, struct queue *q, pr_debug("Total of %d queues are accountable so far\n", dqm->total_queue_count); - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); return retval; out_deallocate_doorbell: @@ -1197,7 +1218,8 @@ out_deallocate_sdma_queue: if (q->properties.type == KFD_QUEUE_TYPE_SDMA) deallocate_sdma_queue(dqm, q->sdma_id); out_unlock: - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); + return retval; } @@ -1210,6 +1232,13 @@ int amdkfd_fence_wait_timeout(unsigned int *fence_addr, while (*fence_addr != fence_value) { if (time_after(jiffies, end_jiffies)) { pr_err("qcm fence wait loop timeout expired\n"); + /* In HWS case, this is used to halt the driver thread + * in order not to mess up CP states before doing + * scandumps for FW debugging. + */ + while (halt_if_hws_hang) + schedule(); + return -ETIME; } schedule(); @@ -1254,6 +1283,8 @@ static int unmap_queues_cpsch(struct device_queue_manager *dqm, { int retval = 0; + if (dqm->is_hws_hang) + return -EIO; if (!dqm->active_runlist) return retval; @@ -1292,9 +1323,13 @@ static int execute_queues_cpsch(struct device_queue_manager *dqm, { int retval; + if (dqm->is_hws_hang) + return -EIO; retval = unmap_queues_cpsch(dqm, filter, filter_param); if (retval) { pr_err("The cp might be in an unrecoverable state due to an unsuccessful queues preemption\n"); + dqm->is_hws_hang = true; + schedule_work(&dqm->hw_exception_work); return retval; } @@ -1306,7 +1341,7 @@ static int destroy_queue_cpsch(struct device_queue_manager *dqm, struct queue *q) { int retval; - struct mqd_manager *mqd; + struct mqd_manager *mqd_mgr; bool preempt_all_queues; preempt_all_queues = false; @@ -1314,7 +1349,7 @@ static int destroy_queue_cpsch(struct device_queue_manager *dqm, retval = 0; /* remove queue from list to prevent rescheduling after preemption */ - mutex_lock(&dqm->lock); + dqm_lock(dqm); if (qpd->is_debug) { /* @@ -1326,9 +1361,9 @@ static int destroy_queue_cpsch(struct device_queue_manager *dqm, } - mqd = dqm->ops.get_mqd_manager(dqm, + mqd_mgr = dqm->ops.get_mqd_manager(dqm, get_mqd_type_from_queue_type(q->properties.type)); - if (!mqd) { + if (!mqd_mgr) { retval = -ENOMEM; goto failed; } @@ -1350,7 +1385,7 @@ static int destroy_queue_cpsch(struct device_queue_manager *dqm, qpd->reset_wavefronts = true; } - mqd->uninit_mqd(mqd, q->mqd, q->mqd_mem_obj); + mqd_mgr->uninit_mqd(mqd_mgr, q->mqd, q->mqd_mem_obj); /* * Unconditionally decrement this counter, regardless of the queue's @@ -1360,14 +1395,14 @@ static int destroy_queue_cpsch(struct device_queue_manager *dqm, pr_debug("Total of %d queues are accountable so far\n", dqm->total_queue_count); - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); return retval; failed: failed_try_destroy_debugged_queue: - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); return retval; } @@ -1391,7 +1426,7 @@ static bool set_cache_memory_policy(struct device_queue_manager *dqm, if (!dqm->asic_ops.set_cache_memory_policy) return retval; - mutex_lock(&dqm->lock); + dqm_lock(dqm); if (alternate_aperture_size == 0) { /* base > limit disables APE1 */ @@ -1437,7 +1472,7 @@ static bool set_cache_memory_policy(struct device_queue_manager *dqm, qpd->sh_mem_ape1_limit); out: - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); return retval; } @@ -1468,7 +1503,7 @@ static int process_termination_nocpsch(struct device_queue_manager *dqm, struct device_process_node *cur, *next_dpn; int retval = 0; - mutex_lock(&dqm->lock); + dqm_lock(dqm); /* Clear all user mode queues */ list_for_each_entry_safe(q, next, &qpd->queues_list, list) { @@ -1489,7 +1524,7 @@ static int process_termination_nocpsch(struct device_queue_manager *dqm, } } - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); return retval; } @@ -1500,14 +1535,14 @@ static int process_termination_cpsch(struct device_queue_manager *dqm, int retval; struct queue *q, *next; struct kernel_queue *kq, *kq_next; - struct mqd_manager *mqd; + struct mqd_manager *mqd_mgr; struct device_process_node *cur, *next_dpn; enum kfd_unmap_queues_filter filter = KFD_UNMAP_QUEUES_FILTER_DYNAMIC_QUEUES; retval = 0; - mutex_lock(&dqm->lock); + dqm_lock(dqm); /* Clean all kernel queues */ list_for_each_entry_safe(kq, kq_next, &qpd->priv_queue_list, list) { @@ -1542,7 +1577,7 @@ static int process_termination_cpsch(struct device_queue_manager *dqm, } retval = execute_queues_cpsch(dqm, filter, 0); - if (retval || qpd->reset_wavefronts) { + if ((!dqm->is_hws_hang) && (retval || qpd->reset_wavefronts)) { pr_warn("Resetting wave fronts (cpsch) on dev %p\n", dqm->dev); dbgdev_wave_reset_wavefronts(dqm->dev, qpd->pqm->process); qpd->reset_wavefronts = false; @@ -1550,19 +1585,19 @@ static int process_termination_cpsch(struct device_queue_manager *dqm, /* lastly, free mqd resources */ list_for_each_entry_safe(q, next, &qpd->queues_list, list) { - mqd = dqm->ops.get_mqd_manager(dqm, + mqd_mgr = dqm->ops.get_mqd_manager(dqm, get_mqd_type_from_queue_type(q->properties.type)); - if (!mqd) { + if (!mqd_mgr) { retval = -ENOMEM; goto out; } list_del(&q->list); qpd->queue_count--; - mqd->uninit_mqd(mqd, q->mqd, q->mqd_mem_obj); + mqd_mgr->uninit_mqd(mqd_mgr, q->mqd, q->mqd_mem_obj); } out: - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); return retval; } @@ -1683,6 +1718,30 @@ void device_queue_manager_uninit(struct device_queue_manager *dqm) kfree(dqm); } +int kfd_process_vm_fault(struct device_queue_manager *dqm, + unsigned int pasid) +{ + struct kfd_process_device *pdd; + struct kfd_process *p = kfd_lookup_process_by_pasid(pasid); + int ret = 0; + + if (!p) + return -EINVAL; + pdd = kfd_get_process_device_data(dqm->dev, p); + if (pdd) + ret = dqm->ops.evict_process_queues(dqm, &pdd->qpd); + kfd_unref_process(p); + + return ret; +} + +static void kfd_process_hw_exception(struct work_struct *work) +{ + struct device_queue_manager *dqm = container_of(work, + struct device_queue_manager, hw_exception_work); + dqm->dev->kfd2kgd->gpu_recover(dqm->dev->kgd); +} + #if defined(CONFIG_DEBUG_FS) static void seq_reg_dump(struct seq_file *m, @@ -1746,8 +1805,8 @@ int dqm_debugfs_hqds(struct seq_file *m, void *data) } } - for (pipe = 0; pipe < CIK_SDMA_ENGINE_NUM; pipe++) { - for (queue = 0; queue < CIK_SDMA_QUEUES_PER_ENGINE; queue++) { + for (pipe = 0; pipe < get_num_sdma_engines(dqm); pipe++) { + for (queue = 0; queue < KFD_SDMA_QUEUES_PER_ENGINE; queue++) { r = dqm->dev->kfd2kgd->hqd_sdma_dump( dqm->dev->kgd, pipe, queue, &dump, &n_regs); if (r) @@ -1764,4 +1823,16 @@ int dqm_debugfs_hqds(struct seq_file *m, void *data) return r; } +int dqm_debugfs_execute_queues(struct device_queue_manager *dqm) +{ + int r = 0; + + dqm_lock(dqm); + dqm->active_runlist = true; + r = execute_queues_cpsch(dqm, KFD_UNMAP_QUEUES_FILTER_ALL_QUEUES, 0); + dqm_unlock(dqm); + + return r; +} + #endif diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h index 59a6b1956932..00da3169a004 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h @@ -26,15 +26,14 @@ #include <linux/rwsem.h> #include <linux/list.h> +#include <linux/mutex.h> +#include <linux/sched/mm.h> #include "kfd_priv.h" #include "kfd_mqd_manager.h" #define KFD_UNMAP_LATENCY_MS (4000) #define QUEUE_PREEMPT_DEFAULT_TIMEOUT_MS (2 * KFD_UNMAP_LATENCY_MS + 1000) - -#define CIK_SDMA_QUEUES (4) -#define CIK_SDMA_QUEUES_PER_ENGINE (2) -#define CIK_SDMA_ENGINE_NUM (2) +#define KFD_SDMA_QUEUES_PER_ENGINE (2) struct device_process_node { struct qcm_process_device *qpd; @@ -170,11 +169,12 @@ struct device_queue_manager { struct device_queue_manager_ops ops; struct device_queue_manager_asic_ops asic_ops; - struct mqd_manager *mqds[KFD_MQD_TYPE_MAX]; + struct mqd_manager *mqd_mgrs[KFD_MQD_TYPE_MAX]; struct packet_manager packets; struct kfd_dev *dev; - struct mutex lock; + struct mutex lock_hidden; /* use dqm_lock/unlock(dqm) */ struct list_head queues; + unsigned int saved_flags; unsigned int processes_count; unsigned int queue_count; unsigned int sdma_queue_count; @@ -190,6 +190,10 @@ struct device_queue_manager { struct kfd_mem_obj *fence_mem; bool active_runlist; int sched_policy; + + /* hw exception */ + bool is_hws_hang; + struct work_struct hw_exception_work; }; void device_queue_manager_init_cik( @@ -207,6 +211,7 @@ void program_sh_mem_settings(struct device_queue_manager *dqm, unsigned int get_queues_num(struct device_queue_manager *dqm); unsigned int get_queues_per_pipe(struct device_queue_manager *dqm); unsigned int get_pipes_per_mec(struct device_queue_manager *dqm); +unsigned int get_num_sdma_queues(struct device_queue_manager *dqm); static inline unsigned int get_sh_mem_bases_32(struct kfd_process_device *pdd) { @@ -219,4 +224,19 @@ get_sh_mem_bases_nybble_64(struct kfd_process_device *pdd) return (pdd->lds_base >> 60) & 0x0E; } +/* The DQM lock can be taken in MMU notifiers. Make sure no reclaim-FS + * happens while holding this lock anywhere to prevent deadlocks when + * an MMU notifier runs in reclaim-FS context. + */ +static inline void dqm_lock(struct device_queue_manager *dqm) +{ + mutex_lock(&dqm->lock_hidden); + dqm->saved_flags = memalloc_nofs_save(); +} +static inline void dqm_unlock(struct device_queue_manager *dqm) +{ + memalloc_nofs_restore(dqm->saved_flags); + mutex_unlock(&dqm->lock_hidden); +} + #endif /* KFD_DEVICE_QUEUE_MANAGER_H_ */ diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager_v9.c b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager_v9.c index 79e5bcf6367c..417515332c35 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager_v9.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager_v9.c @@ -60,7 +60,7 @@ static int update_qpd_v9(struct device_queue_manager *dqm, qpd->sh_mem_config = SH_MEM_ALIGNMENT_MODE_UNALIGNED << SH_MEM_CONFIG__ALIGNMENT_MODE__SHIFT; - if (vega10_noretry && + if (noretry && !dqm->dev->device_info->needs_iommu_device) qpd->sh_mem_config |= 1 << SH_MEM_CONFIG__RETRY_DISABLE__SHIFT; diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c b/drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c index c3744d89352c..ebe79bf00145 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c @@ -188,9 +188,9 @@ void __iomem *kfd_get_kernel_doorbell(struct kfd_dev *kfd, *doorbell_off = kfd->doorbell_id_offset + inx; pr_debug("Get kernel queue doorbell\n" - " doorbell offset == 0x%08X\n" - " kernel address == %p\n", - *doorbell_off, (kfd->doorbell_kernel_ptr + inx)); + " doorbell offset == 0x%08X\n" + " doorbell index == 0x%x\n", + *doorbell_off, inx); return kfd->doorbell_kernel_ptr + inx; } @@ -199,7 +199,8 @@ void kfd_release_kernel_doorbell(struct kfd_dev *kfd, u32 __iomem *db_addr) { unsigned int inx; - inx = (unsigned int)(db_addr - kfd->doorbell_kernel_ptr); + inx = (unsigned int)(db_addr - kfd->doorbell_kernel_ptr) + * sizeof(u32) / kfd->device_info->doorbell_size; mutex_lock(&kfd->doorbell_mutex); __clear_bit(inx, kfd->doorbell_available_index); diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_events.c b/drivers/gpu/drm/amd/amdkfd/kfd_events.c index 5562e94e786a..e9f0e0a1b41c 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_events.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_events.c @@ -850,6 +850,13 @@ static void lookup_events_by_type_and_signal(struct kfd_process *p, ev->memory_exception_data = *ev_data; } + if (type == KFD_EVENT_TYPE_MEMORY) { + dev_warn(kfd_device, + "Sending SIGSEGV to HSA Process with PID %d ", + p->lead_thread->pid); + send_sig(SIGSEGV, p->lead_thread, 0); + } + /* Send SIGTERM no event of type "type" has been found*/ if (send_signal) { if (send_sigterm) { @@ -904,34 +911,41 @@ void kfd_signal_iommu_event(struct kfd_dev *dev, unsigned int pasid, memory_exception_data.failure.NotPresent = 1; memory_exception_data.failure.NoExecute = 0; memory_exception_data.failure.ReadOnly = 0; - if (vma) { - if (vma->vm_start > address) { - memory_exception_data.failure.NotPresent = 1; - memory_exception_data.failure.NoExecute = 0; + if (vma && address >= vma->vm_start) { + memory_exception_data.failure.NotPresent = 0; + + if (is_write_requested && !(vma->vm_flags & VM_WRITE)) + memory_exception_data.failure.ReadOnly = 1; + else memory_exception_data.failure.ReadOnly = 0; - } else { - memory_exception_data.failure.NotPresent = 0; - if (is_write_requested && !(vma->vm_flags & VM_WRITE)) - memory_exception_data.failure.ReadOnly = 1; - else - memory_exception_data.failure.ReadOnly = 0; - if (is_execute_requested && !(vma->vm_flags & VM_EXEC)) - memory_exception_data.failure.NoExecute = 1; - else - memory_exception_data.failure.NoExecute = 0; - } + + if (is_execute_requested && !(vma->vm_flags & VM_EXEC)) + memory_exception_data.failure.NoExecute = 1; + else + memory_exception_data.failure.NoExecute = 0; } up_read(&mm->mmap_sem); mmput(mm); - mutex_lock(&p->event_mutex); + pr_debug("notpresent %d, noexecute %d, readonly %d\n", + memory_exception_data.failure.NotPresent, + memory_exception_data.failure.NoExecute, + memory_exception_data.failure.ReadOnly); - /* Lookup events by type and signal them */ - lookup_events_by_type_and_signal(p, KFD_EVENT_TYPE_MEMORY, - &memory_exception_data); + /* Workaround on Raven to not kill the process when memory is freed + * before IOMMU is able to finish processing all the excessive PPRs + */ + if (dev->device_info->asic_family != CHIP_RAVEN) { + mutex_lock(&p->event_mutex); + + /* Lookup events by type and signal them */ + lookup_events_by_type_and_signal(p, KFD_EVENT_TYPE_MEMORY, + &memory_exception_data); + + mutex_unlock(&p->event_mutex); + } - mutex_unlock(&p->event_mutex); kfd_unref_process(p); } #endif /* KFD_SUPPORT_IOMMU_V2 */ @@ -956,3 +970,67 @@ void kfd_signal_hw_exception_event(unsigned int pasid) mutex_unlock(&p->event_mutex); kfd_unref_process(p); } + +void kfd_signal_vm_fault_event(struct kfd_dev *dev, unsigned int pasid, + struct kfd_vm_fault_info *info) +{ + struct kfd_event *ev; + uint32_t id; + struct kfd_process *p = kfd_lookup_process_by_pasid(pasid); + struct kfd_hsa_memory_exception_data memory_exception_data; + + if (!p) + return; /* Presumably process exited. */ + memset(&memory_exception_data, 0, sizeof(memory_exception_data)); + memory_exception_data.gpu_id = dev->id; + memory_exception_data.failure.imprecise = 1; + /* Set failure reason */ + if (info) { + memory_exception_data.va = (info->page_addr) << PAGE_SHIFT; + memory_exception_data.failure.NotPresent = + info->prot_valid ? 1 : 0; + memory_exception_data.failure.NoExecute = + info->prot_exec ? 1 : 0; + memory_exception_data.failure.ReadOnly = + info->prot_write ? 1 : 0; + memory_exception_data.failure.imprecise = 0; + } + mutex_lock(&p->event_mutex); + + id = KFD_FIRST_NONSIGNAL_EVENT_ID; + idr_for_each_entry_continue(&p->event_idr, ev, id) + if (ev->type == KFD_EVENT_TYPE_MEMORY) { + ev->memory_exception_data = memory_exception_data; + set_event(ev); + } + + mutex_unlock(&p->event_mutex); + kfd_unref_process(p); +} + +void kfd_signal_reset_event(struct kfd_dev *dev) +{ + struct kfd_hsa_hw_exception_data hw_exception_data; + struct kfd_process *p; + struct kfd_event *ev; + unsigned int temp; + uint32_t id, idx; + + /* Whole gpu reset caused by GPU hang and memory is lost */ + memset(&hw_exception_data, 0, sizeof(hw_exception_data)); + hw_exception_data.gpu_id = dev->id; + hw_exception_data.memory_lost = 1; + + idx = srcu_read_lock(&kfd_processes_srcu); + hash_for_each_rcu(kfd_processes_table, temp, p, kfd_processes) { + mutex_lock(&p->event_mutex); + id = KFD_FIRST_NONSIGNAL_EVENT_ID; + idr_for_each_entry_continue(&p->event_idr, ev, id) + if (ev->type == KFD_EVENT_TYPE_HW_EXCEPTION) { + ev->hw_exception_data = hw_exception_data; + set_event(ev); + } + mutex_unlock(&p->event_mutex); + } + srcu_read_unlock(&kfd_processes_srcu, idx); +} diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_events.h b/drivers/gpu/drm/amd/amdkfd/kfd_events.h index abca5bfebbff..c7ac6c73af86 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_events.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_events.h @@ -66,6 +66,7 @@ struct kfd_event { /* type specific data */ union { struct kfd_hsa_memory_exception_data memory_exception_data; + struct kfd_hsa_hw_exception_data hw_exception_data; }; }; diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c b/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c index 37029baa3346..f836897bbf58 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c @@ -26,7 +26,9 @@ static bool event_interrupt_isr_v9(struct kfd_dev *dev, - const uint32_t *ih_ring_entry) + const uint32_t *ih_ring_entry, + uint32_t *patched_ihre, + bool *patched_flag) { uint16_t source_id, client_id, pasid, vmid; const uint32_t *data = ih_ring_entry; @@ -57,7 +59,9 @@ static bool event_interrupt_isr_v9(struct kfd_dev *dev, return source_id == SOC15_INTSRC_CP_END_OF_PIPE || source_id == SOC15_INTSRC_SDMA_TRAP || source_id == SOC15_INTSRC_SQ_INTERRUPT_MSG || - source_id == SOC15_INTSRC_CP_BAD_OPCODE; + source_id == SOC15_INTSRC_CP_BAD_OPCODE || + client_id == SOC15_IH_CLIENTID_VMC || + client_id == SOC15_IH_CLIENTID_UTCL2; } static void event_interrupt_wq_v9(struct kfd_dev *dev, @@ -82,7 +86,19 @@ static void event_interrupt_wq_v9(struct kfd_dev *dev, kfd_signal_hw_exception_event(pasid); else if (client_id == SOC15_IH_CLIENTID_VMC || client_id == SOC15_IH_CLIENTID_UTCL2) { - /* TODO */ + struct kfd_vm_fault_info info = {0}; + uint16_t ring_id = SOC15_RING_ID_FROM_IH_ENTRY(ih_ring_entry); + + info.vmid = vmid; + info.mc_id = client_id; + info.page_addr = ih_ring_entry[4] | + (uint64_t)(ih_ring_entry[5] & 0xf) << 32; + info.prot_valid = ring_id & 0x08; + info.prot_read = ring_id & 0x10; + info.prot_write = ring_id & 0x20; + + kfd_process_vm_fault(dev->dqm, pasid); + kfd_signal_vm_fault_event(dev, pasid, &info); } } diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_interrupt.c b/drivers/gpu/drm/amd/amdkfd/kfd_interrupt.c index db6d9336b80d..c56ac47cd318 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_interrupt.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_interrupt.c @@ -151,13 +151,15 @@ static void interrupt_wq(struct work_struct *work) ih_ring_entry); } -bool interrupt_is_wanted(struct kfd_dev *dev, const uint32_t *ih_ring_entry) +bool interrupt_is_wanted(struct kfd_dev *dev, + const uint32_t *ih_ring_entry, + uint32_t *patched_ihre, bool *flag) { /* integer and bitwise OR so there is no boolean short-circuiting */ unsigned int wanted = 0; wanted |= dev->device_info->event_interrupt_class->interrupt_isr(dev, - ih_ring_entry); + ih_ring_entry, patched_ihre, flag); return wanted != 0; } diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_iommu.c b/drivers/gpu/drm/amd/amdkfd/kfd_iommu.c index c71817963eea..7a61f38c09e6 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_iommu.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_iommu.c @@ -190,7 +190,7 @@ static int iommu_invalid_ppr_cb(struct pci_dev *pdev, int pasid, { struct kfd_dev *dev; - dev_warn(kfd_device, + dev_warn_ratelimited(kfd_device, "Invalid PPR device %x:%x.%x pasid %d address 0x%lX flags 0x%X", PCI_BUS_NUM(pdev->devfn), PCI_SLOT(pdev->devfn), diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c index 476951d8c91c..9f84b4d9fb88 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c @@ -59,7 +59,7 @@ static bool initialize(struct kernel_queue *kq, struct kfd_dev *dev, switch (type) { case KFD_QUEUE_TYPE_DIQ: case KFD_QUEUE_TYPE_HIQ: - kq->mqd = dev->dqm->ops.get_mqd_manager(dev->dqm, + kq->mqd_mgr = dev->dqm->ops.get_mqd_manager(dev->dqm, KFD_MQD_TYPE_HIQ); break; default: @@ -67,7 +67,7 @@ static bool initialize(struct kernel_queue *kq, struct kfd_dev *dev, return false; } - if (!kq->mqd) + if (!kq->mqd_mgr) return false; prop.doorbell_ptr = kfd_get_kernel_doorbell(dev, &prop.doorbell_off); @@ -123,6 +123,7 @@ static bool initialize(struct kernel_queue *kq, struct kfd_dev *dev, prop.write_ptr = (uint32_t *) kq->wptr_gpu_addr; prop.eop_ring_buffer_address = kq->eop_gpu_addr; prop.eop_ring_buffer_size = PAGE_SIZE; + prop.cu_mask = NULL; if (init_queue(&kq->queue, &prop) != 0) goto err_init_queue; @@ -130,7 +131,7 @@ static bool initialize(struct kernel_queue *kq, struct kfd_dev *dev, kq->queue->device = dev; kq->queue->process = kfd_get_process(current); - retval = kq->mqd->init_mqd(kq->mqd, &kq->queue->mqd, + retval = kq->mqd_mgr->init_mqd(kq->mqd_mgr, &kq->queue->mqd, &kq->queue->mqd_mem_obj, &kq->queue->gart_mqd_addr, &kq->queue->properties); @@ -142,9 +143,9 @@ static bool initialize(struct kernel_queue *kq, struct kfd_dev *dev, pr_debug("Assigning hiq to hqd\n"); kq->queue->pipe = KFD_CIK_HIQ_PIPE; kq->queue->queue = KFD_CIK_HIQ_QUEUE; - kq->mqd->load_mqd(kq->mqd, kq->queue->mqd, kq->queue->pipe, - kq->queue->queue, &kq->queue->properties, - NULL); + kq->mqd_mgr->load_mqd(kq->mqd_mgr, kq->queue->mqd, + kq->queue->pipe, kq->queue->queue, + &kq->queue->properties, NULL); } else { /* allocate fence for DIQ */ @@ -182,7 +183,7 @@ err_get_kernel_doorbell: static void uninitialize(struct kernel_queue *kq) { if (kq->queue->properties.type == KFD_QUEUE_TYPE_HIQ) - kq->mqd->destroy_mqd(kq->mqd, + kq->mqd_mgr->destroy_mqd(kq->mqd_mgr, kq->queue->mqd, KFD_PREEMPT_TYPE_WAVEFRONT_RESET, KFD_UNMAP_LATENCY_MS, @@ -191,7 +192,8 @@ static void uninitialize(struct kernel_queue *kq) else if (kq->queue->properties.type == KFD_QUEUE_TYPE_DIQ) kfd_gtt_sa_free(kq->dev, kq->fence_mem_obj); - kq->mqd->uninit_mqd(kq->mqd, kq->queue->mqd, kq->queue->mqd_mem_obj); + kq->mqd_mgr->uninit_mqd(kq->mqd_mgr, kq->queue->mqd, + kq->queue->mqd_mem_obj); kfd_gtt_sa_free(kq->dev, kq->rptr_mem); kfd_gtt_sa_free(kq->dev, kq->wptr_mem); diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.h b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.h index 97aff2041a5d..a7116a939029 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.h @@ -70,7 +70,7 @@ struct kernel_queue { /* data */ struct kfd_dev *dev; - struct mqd_manager *mqd; + struct mqd_manager *mqd_mgr; struct queue *queue; uint64_t pending_wptr64; uint32_t pending_wptr; diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_module.c b/drivers/gpu/drm/amd/amdkfd/kfd_module.c index 76bf2dc8aec4..6e1f5c7c2d4b 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_module.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_module.c @@ -47,6 +47,8 @@ static const struct kgd2kfd_calls kgd2kfd = { .resume_mm = kgd2kfd_resume_mm, .schedule_evict_and_restore_process = kgd2kfd_schedule_evict_and_restore_process, + .pre_reset = kgd2kfd_pre_reset, + .post_reset = kgd2kfd_post_reset, }; int sched_policy = KFD_SCHED_POLICY_HWS; @@ -61,7 +63,7 @@ MODULE_PARM_DESC(hws_max_conc_proc, int cwsr_enable = 1; module_param(cwsr_enable, int, 0444); -MODULE_PARM_DESC(cwsr_enable, "CWSR enable (0 = Off, 1 = On (Default))"); +MODULE_PARM_DESC(cwsr_enable, "CWSR enable (0 = off, 1 = on (default))"); int max_num_of_queues_per_device = KFD_MAX_NUM_OF_QUEUES_PER_DEVICE_DEFAULT; module_param(max_num_of_queues_per_device, int, 0444); @@ -83,13 +85,19 @@ module_param(ignore_crat, int, 0444); MODULE_PARM_DESC(ignore_crat, "Ignore CRAT table during KFD initialization (0 = use CRAT (default), 1 = ignore CRAT)"); -int vega10_noretry; -module_param_named(noretry, vega10_noretry, int, 0644); +int noretry; +module_param(noretry, int, 0644); MODULE_PARM_DESC(noretry, - "Set sh_mem_config.retry_disable on Vega10 (0 = retry enabled (default), 1 = retry disabled)"); + "Set sh_mem_config.retry_disable on GFXv9+ dGPUs (0 = retry enabled (default), 1 = retry disabled)"); + +int halt_if_hws_hang; +module_param(halt_if_hws_hang, int, 0644); +MODULE_PARM_DESC(halt_if_hws_hang, "Halt if HWS hang is detected (0 = off (default), 1 = on)"); + static int amdkfd_init_completed; + int kgd2kfd_init(unsigned int interface_version, const struct kgd2kfd_calls **g2f) { diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.c index 4b8eb506642b..3bc25ab84f34 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.c @@ -21,7 +21,7 @@ * */ -#include "kfd_priv.h" +#include "kfd_mqd_manager.h" struct mqd_manager *mqd_manager_init(enum KFD_MQD_TYPE type, struct kfd_dev *dev) @@ -48,3 +48,42 @@ struct mqd_manager *mqd_manager_init(enum KFD_MQD_TYPE type, return NULL; } + +void mqd_symmetrically_map_cu_mask(struct mqd_manager *mm, + const uint32_t *cu_mask, uint32_t cu_mask_count, + uint32_t *se_mask) +{ + struct kfd_cu_info cu_info; + uint32_t cu_per_sh[4] = {0}; + int i, se, cu = 0; + + mm->dev->kfd2kgd->get_cu_info(mm->dev->kgd, &cu_info); + + if (cu_mask_count > cu_info.cu_active_number) + cu_mask_count = cu_info.cu_active_number; + + for (se = 0; se < cu_info.num_shader_engines; se++) + for (i = 0; i < 4; i++) + cu_per_sh[se] += hweight32(cu_info.cu_bitmap[se][i]); + + /* Symmetrically map cu_mask to all SEs: + * cu_mask[0] bit0 -> se_mask[0] bit0; + * cu_mask[0] bit1 -> se_mask[1] bit0; + * ... (if # SE is 4) + * cu_mask[0] bit4 -> se_mask[0] bit1; + * ... + */ + se = 0; + for (i = 0; i < cu_mask_count; i++) { + if (cu_mask[i / 32] & (1 << (i % 32))) + se_mask[se] |= 1 << cu; + + do { + se++; + if (se == cu_info.num_shader_engines) { + se = 0; + cu++; + } + } while (cu >= cu_per_sh[se] && cu < 32); + } +} diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.h b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.h index 8972bcfbf701..4e84052d4e21 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.h @@ -93,4 +93,8 @@ struct mqd_manager { struct kfd_dev *dev; }; +void mqd_symmetrically_map_cu_mask(struct mqd_manager *mm, + const uint32_t *cu_mask, uint32_t cu_mask_count, + uint32_t *se_mask); + #endif /* KFD_MQD_MANAGER_H_ */ diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c index 06eaa218eba6..47243165a082 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c @@ -41,6 +41,31 @@ static inline struct cik_sdma_rlc_registers *get_sdma_mqd(void *mqd) return (struct cik_sdma_rlc_registers *)mqd; } +static void update_cu_mask(struct mqd_manager *mm, void *mqd, + struct queue_properties *q) +{ + struct cik_mqd *m; + uint32_t se_mask[4] = {0}; /* 4 is the max # of SEs */ + + if (q->cu_mask_count == 0) + return; + + mqd_symmetrically_map_cu_mask(mm, + q->cu_mask, q->cu_mask_count, se_mask); + + m = get_mqd(mqd); + m->compute_static_thread_mgmt_se0 = se_mask[0]; + m->compute_static_thread_mgmt_se1 = se_mask[1]; + m->compute_static_thread_mgmt_se2 = se_mask[2]; + m->compute_static_thread_mgmt_se3 = se_mask[3]; + + pr_debug("Update cu mask to %#x %#x %#x %#x\n", + m->compute_static_thread_mgmt_se0, + m->compute_static_thread_mgmt_se1, + m->compute_static_thread_mgmt_se2, + m->compute_static_thread_mgmt_se3); +} + static int init_mqd(struct mqd_manager *mm, void **mqd, struct kfd_mem_obj **mqd_mem_obj, uint64_t *gart_addr, struct queue_properties *q) @@ -196,6 +221,8 @@ static int __update_mqd(struct mqd_manager *mm, void *mqd, if (q->format == KFD_QUEUE_FORMAT_AQL) m->cp_hqd_pq_control |= NO_UPDATE_RPTR; + update_cu_mask(mm, mqd, q); + q->is_active = (q->queue_size > 0 && q->queue_address != 0 && q->queue_percent > 0 && @@ -408,7 +435,7 @@ struct mqd_manager *mqd_manager_init_cik(enum KFD_MQD_TYPE type, if (WARN_ON(type >= KFD_MQD_TYPE_MAX)) return NULL; - mqd = kzalloc(sizeof(*mqd), GFP_NOIO); + mqd = kzalloc(sizeof(*mqd), GFP_KERNEL); if (!mqd) return NULL; diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c index 684054ff02cd..f5fc3675f21e 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c @@ -41,6 +41,31 @@ static inline struct v9_sdma_mqd *get_sdma_mqd(void *mqd) return (struct v9_sdma_mqd *)mqd; } +static void update_cu_mask(struct mqd_manager *mm, void *mqd, + struct queue_properties *q) +{ + struct v9_mqd *m; + uint32_t se_mask[4] = {0}; /* 4 is the max # of SEs */ + + if (q->cu_mask_count == 0) + return; + + mqd_symmetrically_map_cu_mask(mm, + q->cu_mask, q->cu_mask_count, se_mask); + + m = get_mqd(mqd); + m->compute_static_thread_mgmt_se0 = se_mask[0]; + m->compute_static_thread_mgmt_se1 = se_mask[1]; + m->compute_static_thread_mgmt_se2 = se_mask[2]; + m->compute_static_thread_mgmt_se3 = se_mask[3]; + + pr_debug("update cu mask to %#x %#x %#x %#x\n", + m->compute_static_thread_mgmt_se0, + m->compute_static_thread_mgmt_se1, + m->compute_static_thread_mgmt_se2, + m->compute_static_thread_mgmt_se3); +} + static int init_mqd(struct mqd_manager *mm, void **mqd, struct kfd_mem_obj **mqd_mem_obj, uint64_t *gart_addr, struct queue_properties *q) @@ -55,7 +80,7 @@ static int init_mqd(struct mqd_manager *mm, void **mqd, * instead of sub-allocation function. */ if (kfd->cwsr_enabled && (q->type == KFD_QUEUE_TYPE_COMPUTE)) { - *mqd_mem_obj = kzalloc(sizeof(struct kfd_mem_obj), GFP_NOIO); + *mqd_mem_obj = kzalloc(sizeof(struct kfd_mem_obj), GFP_KERNEL); if (!*mqd_mem_obj) return -ENOMEM; retval = kfd->kfd2kgd->init_gtt_mem_allocation(kfd->kgd, @@ -198,6 +223,8 @@ static int update_mqd(struct mqd_manager *mm, void *mqd, if (mm->dev->cwsr_enabled && q->ctx_save_restore_area_address) m->cp_hqd_ctx_save_control = 0; + update_cu_mask(mm, mqd, q); + q->is_active = (q->queue_size > 0 && q->queue_address != 0 && q->queue_percent > 0 && @@ -393,7 +420,7 @@ struct mqd_manager *mqd_manager_init_v9(enum KFD_MQD_TYPE type, if (WARN_ON(type >= KFD_MQD_TYPE_MAX)) return NULL; - mqd = kzalloc(sizeof(*mqd), GFP_NOIO); + mqd = kzalloc(sizeof(*mqd), GFP_KERNEL); if (!mqd) return NULL; diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c index 481307b8b4db..b81fda3754da 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c @@ -43,6 +43,31 @@ static inline struct vi_sdma_mqd *get_sdma_mqd(void *mqd) return (struct vi_sdma_mqd *)mqd; } +static void update_cu_mask(struct mqd_manager *mm, void *mqd, + struct queue_properties *q) +{ + struct vi_mqd *m; + uint32_t se_mask[4] = {0}; /* 4 is the max # of SEs */ + + if (q->cu_mask_count == 0) + return; + + mqd_symmetrically_map_cu_mask(mm, + q->cu_mask, q->cu_mask_count, se_mask); + + m = get_mqd(mqd); + m->compute_static_thread_mgmt_se0 = se_mask[0]; + m->compute_static_thread_mgmt_se1 = se_mask[1]; + m->compute_static_thread_mgmt_se2 = se_mask[2]; + m->compute_static_thread_mgmt_se3 = se_mask[3]; + + pr_debug("Update cu mask to %#x %#x %#x %#x\n", + m->compute_static_thread_mgmt_se0, + m->compute_static_thread_mgmt_se1, + m->compute_static_thread_mgmt_se2, + m->compute_static_thread_mgmt_se3); +} + static int init_mqd(struct mqd_manager *mm, void **mqd, struct kfd_mem_obj **mqd_mem_obj, uint64_t *gart_addr, struct queue_properties *q) @@ -196,6 +221,8 @@ static int __update_mqd(struct mqd_manager *mm, void *mqd, atc_bit << CP_HQD_CTX_SAVE_CONTROL__ATC__SHIFT | mtype << CP_HQD_CTX_SAVE_CONTROL__MTYPE__SHIFT; + update_cu_mask(mm, mqd, q); + q->is_active = (q->queue_size > 0 && q->queue_address != 0 && q->queue_percent > 0 && @@ -394,7 +421,7 @@ struct mqd_manager *mqd_manager_init_vi(enum KFD_MQD_TYPE type, if (WARN_ON(type >= KFD_MQD_TYPE_MAX)) return NULL; - mqd = kzalloc(sizeof(*mqd), GFP_NOIO); + mqd = kzalloc(sizeof(*mqd), GFP_KERNEL); if (!mqd) return NULL; diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c index c317feb43f69..1092631765cb 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c @@ -418,4 +418,30 @@ out: return 0; } +int pm_debugfs_hang_hws(struct packet_manager *pm) +{ + uint32_t *buffer, size; + int r = 0; + + size = pm->pmf->query_status_size; + mutex_lock(&pm->lock); + pm->priv_queue->ops.acquire_packet_buffer(pm->priv_queue, + size / sizeof(uint32_t), (unsigned int **)&buffer); + if (!buffer) { + pr_err("Failed to allocate buffer on kernel queue\n"); + r = -ENOMEM; + goto out; + } + memset(buffer, 0x55, size); + pm->priv_queue->ops.submit_packet(pm->priv_queue); + + pr_info("Submitting %x %x %x %x %x %x %x to HIQ to hang the HWS.", + buffer[0], buffer[1], buffer[2], buffer[3], + buffer[4], buffer[5], buffer[6]); +out: + mutex_unlock(&pm->lock); + return r; +} + + #endif diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h index 5e3990bb4c4b..f971710f1c91 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h @@ -73,7 +73,7 @@ /* * When working with cp scheduler we should assign the HIQ manually or via - * the radeon driver to a fixed hqd slot, here are the fixed HIQ hqd slot + * the amdgpu driver to a fixed hqd slot, here are the fixed HIQ hqd slot * definitions for Kaveri. In Kaveri only the first ME queues participates * in the cp scheduling taking that in mind we set the HIQ slot in the * second ME. @@ -142,7 +142,12 @@ extern int ignore_crat; /* * Set sh_mem_config.retry_disable on Vega10 */ -extern int vega10_noretry; +extern int noretry; + +/* + * Halt if HWS hang is detected + */ +extern int halt_if_hws_hang; /** * enum kfd_sched_policy @@ -180,9 +185,10 @@ enum cache_policy { struct kfd_event_interrupt_class { bool (*interrupt_isr)(struct kfd_dev *dev, - const uint32_t *ih_ring_entry); + const uint32_t *ih_ring_entry, uint32_t *patched_ihre, + bool *patched_flag); void (*interrupt_wq)(struct kfd_dev *dev, - const uint32_t *ih_ring_entry); + const uint32_t *ih_ring_entry); }; struct kfd_device_info { @@ -197,6 +203,7 @@ struct kfd_device_info { bool supports_cwsr; bool needs_iommu_device; bool needs_pci_atomics; + unsigned int num_sdma_engines; }; struct kfd_mem_obj { @@ -415,6 +422,9 @@ struct queue_properties { uint32_t ctl_stack_size; uint64_t tba_addr; uint64_t tma_addr; + /* Relevant for CU */ + uint32_t cu_mask_count; /* Must be a multiple of 32 */ + uint32_t *cu_mask; }; /** @@ -806,12 +816,18 @@ int kfd_interrupt_init(struct kfd_dev *dev); void kfd_interrupt_exit(struct kfd_dev *dev); void kgd2kfd_interrupt(struct kfd_dev *kfd, const void *ih_ring_entry); bool enqueue_ih_ring_entry(struct kfd_dev *kfd, const void *ih_ring_entry); -bool interrupt_is_wanted(struct kfd_dev *dev, const uint32_t *ih_ring_entry); +bool interrupt_is_wanted(struct kfd_dev *dev, + const uint32_t *ih_ring_entry, + uint32_t *patched_ihre, bool *flag); /* Power Management */ void kgd2kfd_suspend(struct kfd_dev *kfd); int kgd2kfd_resume(struct kfd_dev *kfd); +/* GPU reset */ +int kgd2kfd_pre_reset(struct kfd_dev *kfd); +int kgd2kfd_post_reset(struct kfd_dev *kfd); + /* amdkfd Apertures */ int kfd_init_apertures(struct kfd_process *process); @@ -838,6 +854,7 @@ void device_queue_manager_uninit(struct device_queue_manager *dqm); struct kernel_queue *kernel_queue_init(struct kfd_dev *dev, enum kfd_queue_type type); void kernel_queue_uninit(struct kernel_queue *kq); +int kfd_process_vm_fault(struct device_queue_manager *dqm, unsigned int pasid); /* Process Queue Manager */ struct process_queue_node { @@ -858,6 +875,8 @@ int pqm_create_queue(struct process_queue_manager *pqm, int pqm_destroy_queue(struct process_queue_manager *pqm, unsigned int qid); int pqm_update_queue(struct process_queue_manager *pqm, unsigned int qid, struct queue_properties *p); +int pqm_set_cu_mask(struct process_queue_manager *pqm, unsigned int qid, + struct queue_properties *p); struct kernel_queue *pqm_get_kernel_queue(struct process_queue_manager *pqm, unsigned int qid); @@ -964,10 +983,17 @@ int kfd_event_create(struct file *devkfd, struct kfd_process *p, uint64_t *event_page_offset, uint32_t *event_slot_index); int kfd_event_destroy(struct kfd_process *p, uint32_t event_id); +void kfd_signal_vm_fault_event(struct kfd_dev *dev, unsigned int pasid, + struct kfd_vm_fault_info *info); + +void kfd_signal_reset_event(struct kfd_dev *dev); + void kfd_flush_tlb(struct kfd_process_device *pdd); int dbgdev_wave_reset_wavefronts(struct kfd_dev *dev, struct kfd_process *p); +bool kfd_is_locked(void); + /* Debugfs */ #if defined(CONFIG_DEBUG_FS) @@ -980,6 +1006,10 @@ int dqm_debugfs_hqds(struct seq_file *m, void *data); int kfd_debugfs_rls_by_device(struct seq_file *m, void *data); int pm_debugfs_runlist(struct seq_file *m, void *data); +int kfd_debugfs_hang_hws(struct kfd_dev *dev); +int pm_debugfs_hang_hws(struct packet_manager *pm); +int dqm_debugfs_execute_queues(struct device_queue_manager *dqm); + #else static inline void kfd_debugfs_init(void) {} diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process.c b/drivers/gpu/drm/amd/amdkfd/kfd_process.c index 1d80b4f7c681..4694386cc623 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_process.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_process.c @@ -244,6 +244,8 @@ struct kfd_process *kfd_get_process(const struct task_struct *thread) return ERR_PTR(-EINVAL); process = find_process(thread); + if (!process) + return ERR_PTR(-EINVAL); return process; } diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c index d65ce0436b31..c8cad9c078ae 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c @@ -186,8 +186,7 @@ int pqm_create_queue(struct process_queue_manager *pqm, switch (type) { case KFD_QUEUE_TYPE_SDMA: - if (dev->dqm->queue_count >= - CIK_SDMA_QUEUES_PER_ENGINE * CIK_SDMA_ENGINE_NUM) { + if (dev->dqm->queue_count >= get_num_sdma_queues(dev->dqm)) { pr_err("Over-subscription is not allowed for SDMA.\n"); retval = -EPERM; goto err_create_queue; @@ -209,7 +208,7 @@ int pqm_create_queue(struct process_queue_manager *pqm, KFD_SCHED_POLICY_HWS_NO_OVERSUBSCRIPTION) && ((dev->dqm->processes_count >= dev->vm_info.vmid_num_kfd) || (dev->dqm->queue_count >= get_queues_num(dev->dqm)))) { - pr_err("Over-subscription is not allowed in radeon_kfd.sched_policy == 1\n"); + pr_debug("Over-subscription is not allowed when amdkfd.sched_policy == 1\n"); retval = -EPERM; goto err_create_queue; } @@ -326,6 +325,8 @@ int pqm_destroy_queue(struct process_queue_manager *pqm, unsigned int qid) if (retval != -ETIME) goto err_destroy_queue; } + kfree(pqn->q->properties.cu_mask); + pqn->q->properties.cu_mask = NULL; uninit_queue(pqn->q); } @@ -366,6 +367,34 @@ int pqm_update_queue(struct process_queue_manager *pqm, unsigned int qid, return 0; } +int pqm_set_cu_mask(struct process_queue_manager *pqm, unsigned int qid, + struct queue_properties *p) +{ + int retval; + struct process_queue_node *pqn; + + pqn = get_queue_by_qid(pqm, qid); + if (!pqn) { + pr_debug("No queue %d exists for update operation\n", qid); + return -EFAULT; + } + + /* Free the old CU mask memory if it is already allocated, then + * allocate memory for the new CU mask. + */ + kfree(pqn->q->properties.cu_mask); + + pqn->q->properties.cu_mask_count = p->cu_mask_count; + pqn->q->properties.cu_mask = p->cu_mask; + + retval = pqn->q->device->dqm->ops.update_queue(pqn->q->device->dqm, + pqn->q); + if (retval != 0) + return retval; + + return 0; +} + struct kernel_queue *pqm_get_kernel_queue( struct process_queue_manager *pqm, unsigned int qid) @@ -387,7 +416,7 @@ int pqm_debugfs_mqds(struct seq_file *m, void *data) struct process_queue_node *pqn; struct queue *q; enum KFD_MQD_TYPE mqd_type; - struct mqd_manager *mqd_manager; + struct mqd_manager *mqd_mgr; int r = 0; list_for_each_entry(pqn, &pqm->queues, process_queue_list) { @@ -410,11 +439,11 @@ int pqm_debugfs_mqds(struct seq_file *m, void *data) q->properties.type, q->device->id); continue; } - mqd_manager = q->device->dqm->ops.get_mqd_manager( + mqd_mgr = q->device->dqm->ops.get_mqd_manager( q->device->dqm, mqd_type); } else if (pqn->kq) { q = pqn->kq->queue; - mqd_manager = pqn->kq->mqd; + mqd_mgr = pqn->kq->mqd_mgr; switch (q->properties.type) { case KFD_QUEUE_TYPE_DIQ: seq_printf(m, " DIQ on device %x\n", @@ -434,7 +463,7 @@ int pqm_debugfs_mqds(struct seq_file *m, void *data) continue; } - r = mqd_manager->debugfs_show_mqd(m, q->mqd); + r = mqd_mgr->debugfs_show_mqd(m, q->mqd); if (r != 0) break; } diff --git a/drivers/gpu/drm/amd/display/Kconfig b/drivers/gpu/drm/amd/display/Kconfig index d5d4586e6176..325083b0297e 100644 --- a/drivers/gpu/drm/amd/display/Kconfig +++ b/drivers/gpu/drm/amd/display/Kconfig @@ -9,23 +9,6 @@ config DRM_AMD_DC support for AMDGPU. This adds required support for Vega and Raven ASICs. -config DRM_AMD_DC_FBC - bool "AMD FBC - Enable Frame Buffer Compression" - depends on DRM_AMD_DC - help - Choose this option if you want to use frame buffer compression - support. - This is a power optimisation feature, check its availability - on your hardware before enabling this option. - - -config DRM_AMD_DC_DCN1_0 - bool "DCN 1.0 Raven family" - depends on DRM_AMD_DC && X86 - help - Choose this option if you want to have - RV family for display engine - config DEBUG_KERNEL_DC bool "Enable kgdb break in DC" depends on DRM_AMD_DC diff --git a/drivers/gpu/drm/amd/display/TODO b/drivers/gpu/drm/amd/display/TODO index 357d59648401..a8a6c106e8c7 100644 --- a/drivers/gpu/drm/amd/display/TODO +++ b/drivers/gpu/drm/amd/display/TODO @@ -97,10 +97,10 @@ share it with drivers. But that's a very long term goal, and by far not just an issue with DC - other drivers, especially around DP sink handling, are equally guilty. -19. The DC logger is still a rather sore thing, but I know that the DRM_DEBUG -stuff just isn't up to the challenges either. We need to figure out something -that integrates better with DRM and linux debug printing, while not being -useless with filtering output. dynamic debug printing might be an option. +19. DONE - The DC logger is still a rather sore thing, but I know that the +DRM_DEBUG stuff just isn't up to the challenges either. We need to figure out +something that integrates better with DRM and linux debug printing, while not +being 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 diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile b/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile index af16973f2c41..94911871eb9b 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile @@ -28,11 +28,11 @@ AMDGPUDM = amdgpu_dm.o amdgpu_dm_irq.o amdgpu_dm_mst_types.o amdgpu_dm_color.o ifneq ($(CONFIG_DRM_AMD_DC),) -AMDGPUDM += amdgpu_dm_services.o amdgpu_dm_helpers.o +AMDGPUDM += amdgpu_dm_services.o amdgpu_dm_helpers.o amdgpu_dm_pp_smu.o endif ifneq ($(CONFIG_DEBUG_FS),) -AMDGPUDM += amdgpu_dm_crc.o +AMDGPUDM += amdgpu_dm_crc.o amdgpu_dm_debugfs.o endif subdir-ccflags-y += -I$(FULL_AMD_DISPLAY_PATH)/dc diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c index 770c6b24be0b..34f34823bab5 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -39,6 +39,9 @@ #include "dm_helpers.h" #include "dm_services_types.h" #include "amdgpu_dm_mst_types.h" +#if defined(CONFIG_DEBUG_FS) +#include "amdgpu_dm_debugfs.h" +#endif #include "ivsrcid/ivsrcid_vislands30.h" @@ -57,7 +60,7 @@ #include "modules/inc/mod_freesync.h" -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 #include "ivsrcid/irqsrcs_dcn_1_0.h" #include "dcn/dcn_1_0_offset.h" @@ -347,7 +350,6 @@ static void hotplug_notify_work_func(struct work_struct *work) drm_kms_helper_hotplug_event(dev); } -#if defined(CONFIG_DRM_AMD_DC_FBC) /* Allocate memory for FBC compressed data */ static void amdgpu_dm_fbc_init(struct drm_connector *connector) { @@ -388,7 +390,6 @@ static void amdgpu_dm_fbc_init(struct drm_connector *connector) } } -#endif /* Init display KMS @@ -902,14 +903,14 @@ amdgpu_dm_update_connector_after_detect(struct amdgpu_dm_connector *aconnector) (struct edid *) sink->dc_edid.raw_edid; - drm_mode_connector_update_edid_property(connector, + drm_connector_update_edid_property(connector, aconnector->edid); } amdgpu_dm_add_sink_to_freesync_module(connector, aconnector->edid); } else { amdgpu_dm_remove_sink_from_freesync_module(connector); - drm_mode_connector_update_edid_property(connector, NULL); + drm_connector_update_edid_property(connector, NULL); aconnector->num_modes = 0; aconnector->dc_sink = NULL; aconnector->edid = NULL; @@ -1040,7 +1041,7 @@ static void handle_hpd_rx_irq(void *param) if (dc_link->type != dc_connection_mst_branch) mutex_lock(&aconnector->hpd_lock); - if (dc_link_handle_hpd_rx_irq(dc_link, NULL) && + if (dc_link_handle_hpd_rx_irq(dc_link, NULL, NULL) && !is_mst_root_connector) { /* Downstream Port status changed. */ if (dc_link_detect(dc_link, DETECT_REASON_HPDRX)) { @@ -1191,7 +1192,7 @@ static int dce110_register_irq_handlers(struct amdgpu_device *adev) return 0; } -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 /* Register IRQ sources and initialize IRQ callbacks */ static int dcn10_register_irq_handlers(struct amdgpu_device *adev) { @@ -1319,7 +1320,12 @@ static int amdgpu_dm_backlight_update_status(struct backlight_device *bd) static int amdgpu_dm_backlight_get_brightness(struct backlight_device *bd) { - return bd->props.brightness; + struct amdgpu_display_manager *dm = bl_get_data(bd); + int ret = dc_link_get_backlight_level(dm->backlight_link); + + if (ret == DC_ERROR_UNEXPECTED) + return bd->props.brightness; + return ret; } static const struct backlight_ops amdgpu_dm_backlight_ops = { @@ -1334,6 +1340,7 @@ amdgpu_dm_register_backlight_device(struct amdgpu_display_manager *dm) struct backlight_properties props = { 0 }; props.max_brightness = AMDGPU_MAX_BL_LEVEL; + props.brightness = AMDGPU_MAX_BL_LEVEL; props.type = BACKLIGHT_RAW; snprintf(bl_name, sizeof(bl_name), "amdgpu_bl%d", @@ -1525,16 +1532,12 @@ static int amdgpu_dm_initialize_drm_device(struct amdgpu_device *adev) goto fail; } break; -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 case CHIP_RAVEN: if (dcn10_register_irq_handlers(dm->adev)) { DRM_ERROR("DM: Failed to initialize IRQ\n"); goto fail; } - /* - * Temporary disable until pplib/smu interaction is implemented - */ - dm->dc->debug.disable_stutter = true; break; #endif default: @@ -1542,6 +1545,9 @@ static int amdgpu_dm_initialize_drm_device(struct amdgpu_device *adev) goto fail; } + if (adev->asic_type != CHIP_CARRIZO && adev->asic_type != CHIP_STONEY) + dm->dc->debug.disable_stutter = amdgpu_pp_feature_mask & PP_STUTTER_MODE ? false : true; + return 0; fail: kfree(aencoder); @@ -1573,18 +1579,6 @@ static void dm_bandwidth_update(struct amdgpu_device *adev) /* TODO: implement later */ } -static void dm_set_backlight_level(struct amdgpu_encoder *amdgpu_encoder, - u8 level) -{ - /* TODO: translate amdgpu_encoder to display_index and call DAL */ -} - -static u8 dm_get_backlight_level(struct amdgpu_encoder *amdgpu_encoder) -{ - /* TODO: translate amdgpu_encoder to display_index and call DAL */ - return 0; -} - static int amdgpu_notify_freesync(struct drm_device *dev, void *data, struct drm_file *filp) { @@ -1613,10 +1607,8 @@ static int amdgpu_notify_freesync(struct drm_device *dev, void *data, static const struct amdgpu_display_funcs dm_display_funcs = { .bandwidth_update = dm_bandwidth_update, /* called unconditionally */ .vblank_get_counter = dm_vblank_get_counter,/* called unconditionally */ - .backlight_set_level = - dm_set_backlight_level,/* called unconditionally */ - .backlight_get_level = - dm_get_backlight_level,/* called unconditionally */ + .backlight_set_level = NULL, /* never called for DC */ + .backlight_get_level = NULL, /* never called for DC */ .hpd_sense = NULL,/* called unconditionally */ .hpd_set_polarity = NULL, /* called unconditionally */ .hpd_get_gpio_reg = NULL, /* VBIOS parsing. DAL does it. */ @@ -1724,7 +1716,7 @@ static int dm_early_init(void *handle) adev->mode_info.num_dig = 6; adev->mode_info.plane_type = dm_plane_type_default; break; -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 case CHIP_RAVEN: adev->mode_info.num_crtc = 4; adev->mode_info.num_hpd = 4; @@ -2123,13 +2115,8 @@ convert_color_depth_from_display_info(const struct drm_connector *connector) static enum dc_aspect_ratio get_aspect_ratio(const struct drm_display_mode *mode_in) { - int32_t width = mode_in->crtc_hdisplay * 9; - int32_t height = mode_in->crtc_vdisplay * 16; - - if ((width - height) < 10 && (width - height) > -10) - return ASPECT_RATIO_16_9; - else - return ASPECT_RATIO_4_3; + /* 1-1 mapping, since both enums follow the HDMI spec. */ + return (enum dc_aspect_ratio) mode_in->picture_aspect_ratio; } static enum dc_color_space @@ -3093,15 +3080,25 @@ static int dm_plane_helper_prepare_fb(struct drm_plane *plane, else domain = AMDGPU_GEM_DOMAIN_VRAM; - r = amdgpu_bo_pin(rbo, domain, &afb->address); - amdgpu_bo_unreserve(rbo); - + r = amdgpu_bo_pin(rbo, domain); if (unlikely(r != 0)) { if (r != -ERESTARTSYS) DRM_ERROR("Failed to pin framebuffer with error %d\n", r); + amdgpu_bo_unreserve(rbo); return r; } + r = amdgpu_ttm_alloc_gart(&rbo->tbo); + if (unlikely(r != 0)) { + amdgpu_bo_unpin(rbo); + amdgpu_bo_unreserve(rbo); + DRM_ERROR("%p bind failed\n", rbo); + return r; + } + amdgpu_bo_unreserve(rbo); + + afb->address = amdgpu_bo_gpu_offset(rbo); + amdgpu_bo_ref(rbo); if (dm_plane_state_new->dc_state && @@ -3471,12 +3468,15 @@ static int amdgpu_dm_connector_get_modes(struct drm_connector *connector) struct edid *edid = amdgpu_dm_connector->edid; encoder = helper->best_encoder(connector); - amdgpu_dm_connector_ddc_get_modes(connector, edid); - amdgpu_dm_connector_add_common_modes(encoder, connector); -#if defined(CONFIG_DRM_AMD_DC_FBC) + if (!edid || !drm_edid_is_valid(edid)) { + drm_add_modes_noedid(connector, 640, 480); + } else { + amdgpu_dm_connector_ddc_get_modes(connector, edid); + amdgpu_dm_connector_add_common_modes(encoder, connector); + } amdgpu_dm_fbc_init(connector); -#endif + return amdgpu_dm_connector->num_modes; } @@ -3495,7 +3495,6 @@ void amdgpu_dm_connector_init_helper(struct amdgpu_display_manager *dm, aconnector->base.stereo_allowed = false; aconnector->base.dpms = DRM_MODE_DPMS_OFF; aconnector->hpd.hpd = AMDGPU_HPD_NONE; /* not used */ - mutex_init(&aconnector->hpd_lock); /* configure support HPD hot plug connector_>polled default value is 0 @@ -3504,9 +3503,13 @@ void amdgpu_dm_connector_init_helper(struct amdgpu_display_manager *dm, switch (connector_type) { case DRM_MODE_CONNECTOR_HDMIA: aconnector->base.polled = DRM_CONNECTOR_POLL_HPD; + aconnector->base.ycbcr_420_allowed = + link->link_enc->features.ycbcr420_supported ? true : false; break; case DRM_MODE_CONNECTOR_DisplayPort: aconnector->base.polled = DRM_CONNECTOR_POLL_HPD; + aconnector->base.ycbcr_420_allowed = + link->link_enc->features.ycbcr420_supported ? true : false; break; case DRM_MODE_CONNECTOR_DVID: aconnector->base.polled = DRM_CONNECTOR_POLL_HPD; @@ -3659,10 +3662,17 @@ static int amdgpu_dm_connector_init(struct amdgpu_display_manager *dm, link, link_index); - drm_mode_connector_attach_encoder( + drm_connector_attach_encoder( &aconnector->base, &aencoder->base); drm_connector_register(&aconnector->base); +#if defined(CONFIG_DEBUG_FS) + res = connector_debugfs_init(aconnector); + if (res) { + DRM_ERROR("Failed to create debugfs for connector"); + goto out_free; + } +#endif if (connector_type == DRM_MODE_CONNECTOR_DisplayPort || connector_type == DRM_MODE_CONNECTOR_eDP) @@ -3959,8 +3969,6 @@ static void amdgpu_dm_do_flip(struct drm_crtc *crtc, /* Flip */ spin_lock_irqsave(&crtc->dev->event_lock, flags); - /* update crtc fb */ - crtc->primary->fb = fb; WARN_ON(acrtc->pflip_status != AMDGPU_FLIP_NONE); WARN_ON(!acrtc_state->stream); diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h index d5aa89ad5571..a29dc35954c9 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h @@ -72,13 +72,11 @@ struct irq_list_head { struct work_struct work; }; -#if defined(CONFIG_DRM_AMD_DC_FBC) struct dm_comressor_info { void *cpu_addr; struct amdgpu_bo *bo_ptr; uint64_t gpu_addr; }; -#endif struct amdgpu_display_manager { @@ -129,9 +127,8 @@ struct amdgpu_display_manager { * Caches device atomic state for suspend/resume */ struct drm_atomic_state *cached_state; -#if defined(CONFIG_DRM_AMD_DC_FBC) + struct dm_comressor_info compressor; -#endif }; struct amdgpu_dm_connector { diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c index b329393307e5..326f6fb7e0bc 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c @@ -231,18 +231,21 @@ void amdgpu_dm_set_ctm(struct dm_crtc_state *crtc) * preparation for hardware commit. If no lut is specified by user, we default * to SRGB degamma. * - * Currently, we only support degamma bypass, or preprogrammed SRGB degamma. - * Programmable degamma is not supported, and an attempt to do so will return - * -EINVAL. + * We support degamma bypass, predefined SRGB, and custom degamma * * RETURNS: - * 0 on success, -EINVAL if custom degamma curve is given. + * 0 on success + * -EINVAL if crtc_state has a degamma_lut of invalid size + * -ENOMEM if gamma allocation fails */ int amdgpu_dm_set_degamma_lut(struct drm_crtc_state *crtc_state, struct dc_plane_state *dc_plane_state) { struct drm_property_blob *blob = crtc_state->degamma_lut; struct drm_color_lut *lut; + uint32_t lut_size; + struct dc_gamma *gamma; + bool ret; if (!blob) { /* Default to SRGB */ @@ -258,11 +261,30 @@ int amdgpu_dm_set_degamma_lut(struct drm_crtc_state *crtc_state, return 0; } - /* Otherwise, assume SRGB, since programmable degamma is not - * supported. - */ - dc_plane_state->in_transfer_func->type = TF_TYPE_PREDEFINED; - dc_plane_state->in_transfer_func->tf = TRANSFER_FUNCTION_SRGB; - return -EINVAL; + gamma = dc_create_gamma(); + if (!gamma) + return -ENOMEM; + + lut_size = blob->length / sizeof(struct drm_color_lut); + gamma->num_entries = lut_size; + if (gamma->num_entries == MAX_COLOR_LUT_ENTRIES) + gamma->type = GAMMA_CUSTOM; + else { + dc_gamma_release(&gamma); + return -EINVAL; + } + + __drm_lut_to_dc_gamma(lut, gamma, false); + + dc_plane_state->in_transfer_func->type = TF_TYPE_DISTRIBUTED_POINTS; + ret = mod_color_calculate_degamma_params(dc_plane_state->in_transfer_func, gamma, true); + dc_gamma_release(&gamma); + if (!ret) { + dc_plane_state->in_transfer_func->type = TF_TYPE_BYPASS; + DRM_ERROR("Out of memory when calculating degamma params\n"); + return -ENOMEM; + } + + return 0; } diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c new file mode 100644 index 000000000000..0d9e410ca01e --- /dev/null +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c @@ -0,0 +1,722 @@ +/* + * Copyright 2018 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: AMD + * + */ + +#include <linux/debugfs.h> + +#include "dc.h" +#include "amdgpu.h" +#include "amdgpu_dm.h" +#include "amdgpu_dm_debugfs.h" + +/* function description + * get/ set DP configuration: lane_count, link_rate, spread_spectrum + * + * valid lane count value: 1, 2, 4 + * valid link rate value: + * 06h = 1.62Gbps per lane + * 0Ah = 2.7Gbps per lane + * 0Ch = 3.24Gbps per lane + * 14h = 5.4Gbps per lane + * 1Eh = 8.1Gbps per lane + * + * debugfs is located at /sys/kernel/debug/dri/0/DP-x/link_settings + * + * --- to get dp configuration + * + * cat link_settings + * + * It will list current, verified, reported, preferred dp configuration. + * current -- for current video mode + * verified --- maximum configuration which pass link training + * reported --- DP rx report caps (DPCD register offset 0, 1 2) + * preferred --- user force settings + * + * --- set (or force) dp configuration + * + * echo <lane_count> <link_rate> > link_settings + * + * for example, to force to 2 lane, 2.7GHz, + * echo 4 0xa > link_settings + * + * spread_spectrum could not be changed dynamically. + * + * in case invalid lane count, link rate are force, no hw programming will be + * done. please check link settings after force operation to see if HW get + * programming. + * + * cat link_settings + * + * check current and preferred settings. + * + */ +static ssize_t dp_link_settings_read(struct file *f, char __user *buf, + size_t size, loff_t *pos) +{ + struct amdgpu_dm_connector *connector = file_inode(f)->i_private; + struct dc_link *link = connector->dc_link; + char *rd_buf = NULL; + char *rd_buf_ptr = NULL; + const uint32_t rd_buf_size = 100; + uint32_t result = 0; + uint8_t str_len = 0; + int r; + + if (*pos & 3 || size & 3) + return -EINVAL; + + rd_buf = kcalloc(rd_buf_size, sizeof(char), GFP_KERNEL); + if (!rd_buf) + return 0; + + rd_buf_ptr = rd_buf; + + str_len = strlen("Current: %d %d %d "); + snprintf(rd_buf_ptr, str_len, "Current: %d %d %d ", + link->cur_link_settings.lane_count, + link->cur_link_settings.link_rate, + link->cur_link_settings.link_spread); + rd_buf_ptr += str_len; + + str_len = strlen("Verified: %d %d %d "); + snprintf(rd_buf_ptr, str_len, "Verified: %d %d %d ", + link->verified_link_cap.lane_count, + link->verified_link_cap.link_rate, + link->verified_link_cap.link_spread); + rd_buf_ptr += str_len; + + str_len = strlen("Reported: %d %d %d "); + snprintf(rd_buf_ptr, str_len, "Reported: %d %d %d ", + link->reported_link_cap.lane_count, + link->reported_link_cap.link_rate, + link->reported_link_cap.link_spread); + rd_buf_ptr += str_len; + + str_len = strlen("Preferred: %d %d %d "); + snprintf(rd_buf_ptr, str_len, "Preferred: %d %d %d\n", + link->preferred_link_setting.lane_count, + link->preferred_link_setting.link_rate, + link->preferred_link_setting.link_spread); + + while (size) { + if (*pos >= rd_buf_size) + break; + + r = put_user(*(rd_buf + result), buf); + if (r) + return r; /* r = -EFAULT */ + + buf += 1; + size -= 1; + *pos += 1; + result += 1; + } + + kfree(rd_buf); + return result; +} + +static ssize_t dp_link_settings_write(struct file *f, const char __user *buf, + size_t size, loff_t *pos) +{ + struct amdgpu_dm_connector *connector = file_inode(f)->i_private; + struct dc_link *link = connector->dc_link; + struct dc *dc = (struct dc *)link->dc; + struct dc_link_settings prefer_link_settings; + char *wr_buf = NULL; + char *wr_buf_ptr = NULL; + const uint32_t wr_buf_size = 40; + int r; + int bytes_from_user; + char *sub_str; + /* 0: lane_count; 1: link_rate */ + uint8_t param_index = 0; + long param[2]; + const char delimiter[3] = {' ', '\n', '\0'}; + bool valid_input = false; + + if (size == 0) + return -EINVAL; + + wr_buf = kcalloc(wr_buf_size, sizeof(char), GFP_KERNEL); + if (!wr_buf) + return -EINVAL; + wr_buf_ptr = wr_buf; + + r = copy_from_user(wr_buf_ptr, buf, wr_buf_size); + + /* r is bytes not be copied */ + if (r >= wr_buf_size) { + kfree(wr_buf); + DRM_DEBUG_DRIVER("user data not read\n"); + return -EINVAL; + } + + bytes_from_user = wr_buf_size - r; + + while (isspace(*wr_buf_ptr)) + wr_buf_ptr++; + + while ((*wr_buf_ptr != '\0') && (param_index < 2)) { + + sub_str = strsep(&wr_buf_ptr, delimiter); + + r = kstrtol(sub_str, 16, ¶m[param_index]); + + if (r) + DRM_DEBUG_DRIVER("string to int convert error code: %d\n", r); + + param_index++; + while (isspace(*wr_buf_ptr)) + wr_buf_ptr++; + } + + switch (param[0]) { + case LANE_COUNT_ONE: + case LANE_COUNT_TWO: + case LANE_COUNT_FOUR: + valid_input = true; + break; + default: + break; + } + + switch (param[1]) { + case LINK_RATE_LOW: + case LINK_RATE_HIGH: + case LINK_RATE_RBR2: + case LINK_RATE_HIGH2: + case LINK_RATE_HIGH3: + valid_input = true; + break; + default: + break; + } + + if (!valid_input) { + kfree(wr_buf); + DRM_DEBUG_DRIVER("Invalid Input value No HW will be programmed\n"); + return bytes_from_user; + } + + /* save user force lane_count, link_rate to preferred settings + * spread spectrum will not be changed + */ + prefer_link_settings.link_spread = link->cur_link_settings.link_spread; + prefer_link_settings.lane_count = param[0]; + prefer_link_settings.link_rate = param[1]; + + dc_link_set_preferred_link_settings(dc, &prefer_link_settings, link); + + kfree(wr_buf); + return bytes_from_user; +} + +/* function: get current DP PHY settings: voltage swing, pre-emphasis, + * post-cursor2 (defined by VESA DP specification) + * + * valid values + * voltage swing: 0,1,2,3 + * pre-emphasis : 0,1,2,3 + * post cursor2 : 0,1,2,3 + * + * + * how to use this debugfs + * + * debugfs is located at /sys/kernel/debug/dri/0/DP-x + * + * there will be directories, like DP-1, DP-2,DP-3, etc. for DP display + * + * To figure out which DP-x is the display for DP to be check, + * cd DP-x + * ls -ll + * There should be debugfs file, like link_settings, phy_settings. + * cat link_settings + * from lane_count, link_rate to figure which DP-x is for display to be worked + * on + * + * To get current DP PHY settings, + * cat phy_settings + * + * To change DP PHY settings, + * echo <voltage_swing> <pre-emphasis> <post_cursor2> > phy_settings + * for examle, to change voltage swing to 2, pre-emphasis to 3, post_cursor2 to + * 0, + * echo 2 3 0 > phy_settings + * + * To check if change be applied, get current phy settings by + * cat phy_settings + * + * In case invalid values are set by user, like + * echo 1 4 0 > phy_settings + * + * HW will NOT be programmed by these settings. + * cat phy_settings will show the previous valid settings. + */ +static ssize_t dp_phy_settings_read(struct file *f, char __user *buf, + size_t size, loff_t *pos) +{ + struct amdgpu_dm_connector *connector = file_inode(f)->i_private; + struct dc_link *link = connector->dc_link; + char *rd_buf = NULL; + const uint32_t rd_buf_size = 20; + uint32_t result = 0; + int r; + + if (*pos & 3 || size & 3) + return -EINVAL; + + rd_buf = kcalloc(rd_buf_size, sizeof(char), GFP_KERNEL); + if (!rd_buf) + return -EINVAL; + + snprintf(rd_buf, rd_buf_size, " %d %d %d ", + link->cur_lane_setting.VOLTAGE_SWING, + link->cur_lane_setting.PRE_EMPHASIS, + link->cur_lane_setting.POST_CURSOR2); + + while (size) { + if (*pos >= rd_buf_size) + break; + + r = put_user((*(rd_buf + result)), buf); + if (r) + return r; /* r = -EFAULT */ + + buf += 1; + size -= 1; + *pos += 1; + result += 1; + } + + kfree(rd_buf); + return result; +} + +static ssize_t dp_phy_settings_write(struct file *f, const char __user *buf, + size_t size, loff_t *pos) +{ + struct amdgpu_dm_connector *connector = file_inode(f)->i_private; + struct dc_link *link = connector->dc_link; + struct dc *dc = (struct dc *)link->dc; + char *wr_buf = NULL; + char *wr_buf_ptr = NULL; + uint32_t wr_buf_size = 40; + int r; + int bytes_from_user; + char *sub_str; + uint8_t param_index = 0; + long param[3]; + const char delimiter[3] = {' ', '\n', '\0'}; + bool use_prefer_link_setting; + struct link_training_settings link_lane_settings; + + if (size == 0) + return 0; + + wr_buf = kcalloc(wr_buf_size, sizeof(char), GFP_KERNEL); + if (!wr_buf) + return 0; + wr_buf_ptr = wr_buf; + + r = copy_from_user(wr_buf_ptr, buf, wr_buf_size); + + /* r is bytes not be copied */ + if (r >= wr_buf_size) { + kfree(wr_buf); + DRM_DEBUG_DRIVER("user data not be read\n"); + return 0; + } + + bytes_from_user = wr_buf_size - r; + + while (isspace(*wr_buf_ptr)) + wr_buf_ptr++; + + while ((*wr_buf_ptr != '\0') && (param_index < 3)) { + + sub_str = strsep(&wr_buf_ptr, delimiter); + + r = kstrtol(sub_str, 16, ¶m[param_index]); + + if (r) + DRM_DEBUG_DRIVER("string to int convert error code: %d\n", r); + + param_index++; + while (isspace(*wr_buf_ptr)) + wr_buf_ptr++; + } + + if ((param[0] > VOLTAGE_SWING_MAX_LEVEL) || + (param[1] > PRE_EMPHASIS_MAX_LEVEL) || + (param[2] > POST_CURSOR2_MAX_LEVEL)) { + kfree(wr_buf); + DRM_DEBUG_DRIVER("Invalid Input No HW will be programmed\n"); + return bytes_from_user; + } + + /* get link settings: lane count, link rate */ + use_prefer_link_setting = + ((link->preferred_link_setting.link_rate != LINK_RATE_UNKNOWN) && + (link->test_pattern_enabled)); + + memset(&link_lane_settings, 0, sizeof(link_lane_settings)); + + if (use_prefer_link_setting) { + link_lane_settings.link_settings.lane_count = + link->preferred_link_setting.lane_count; + link_lane_settings.link_settings.link_rate = + link->preferred_link_setting.link_rate; + link_lane_settings.link_settings.link_spread = + link->preferred_link_setting.link_spread; + } else { + link_lane_settings.link_settings.lane_count = + link->cur_link_settings.lane_count; + link_lane_settings.link_settings.link_rate = + link->cur_link_settings.link_rate; + link_lane_settings.link_settings.link_spread = + link->cur_link_settings.link_spread; + } + + /* apply phy settings from user */ + for (r = 0; r < link_lane_settings.link_settings.lane_count; r++) { + link_lane_settings.lane_settings[r].VOLTAGE_SWING = + (enum dc_voltage_swing) (param[0]); + link_lane_settings.lane_settings[r].PRE_EMPHASIS = + (enum dc_pre_emphasis) (param[1]); + link_lane_settings.lane_settings[r].POST_CURSOR2 = + (enum dc_post_cursor2) (param[2]); + } + + /* program ASIC registers and DPCD registers */ + dc_link_set_drive_settings(dc, &link_lane_settings, link); + + kfree(wr_buf); + return bytes_from_user; +} + +/* function description + * + * set PHY layer or Link layer test pattern + * PHY test pattern is used for PHY SI check. + * Link layer test will not affect PHY SI. + * + * Reset Test Pattern: + * 0 = DP_TEST_PATTERN_VIDEO_MODE + * + * PHY test pattern supported: + * 1 = DP_TEST_PATTERN_D102 + * 2 = DP_TEST_PATTERN_SYMBOL_ERROR + * 3 = DP_TEST_PATTERN_PRBS7 + * 4 = DP_TEST_PATTERN_80BIT_CUSTOM + * 5 = DP_TEST_PATTERN_CP2520_1 + * 6 = DP_TEST_PATTERN_CP2520_2 = DP_TEST_PATTERN_HBR2_COMPLIANCE_EYE + * 7 = DP_TEST_PATTERN_CP2520_3 + * + * DP PHY Link Training Patterns + * 8 = DP_TEST_PATTERN_TRAINING_PATTERN1 + * 9 = DP_TEST_PATTERN_TRAINING_PATTERN2 + * a = DP_TEST_PATTERN_TRAINING_PATTERN3 + * b = DP_TEST_PATTERN_TRAINING_PATTERN4 + * + * DP Link Layer Test pattern + * c = DP_TEST_PATTERN_COLOR_SQUARES + * d = DP_TEST_PATTERN_COLOR_SQUARES_CEA + * e = DP_TEST_PATTERN_VERTICAL_BARS + * f = DP_TEST_PATTERN_HORIZONTAL_BARS + * 10= DP_TEST_PATTERN_COLOR_RAMP + * + * debugfs phy_test_pattern is located at /syskernel/debug/dri/0/DP-x + * + * --- set test pattern + * echo <test pattern #> > test_pattern + * + * If test pattern # is not supported, NO HW programming will be done. + * for DP_TEST_PATTERN_80BIT_CUSTOM, it needs extra 10 bytes of data + * for the user pattern. input 10 bytes data are separated by space + * + * echo 0x4 0x11 0x22 0x33 0x44 0x55 0x66 0x77 0x88 0x99 0xaa > test_pattern + * + * --- reset test pattern + * echo 0 > test_pattern + * + * --- HPD detection is disabled when set PHY test pattern + * + * when PHY test pattern (pattern # within [1,7]) is set, HPD pin of HW ASIC + * is disable. User could unplug DP display from DP connected and plug scope to + * check test pattern PHY SI. + * If there is need unplug scope and plug DP display back, do steps below: + * echo 0 > phy_test_pattern + * unplug scope + * plug DP display. + * + * "echo 0 > phy_test_pattern" will re-enable HPD pin again so that video sw + * driver could detect "unplug scope" and "plug DP display" + */ +static ssize_t dp_phy_test_pattern_debugfs_write(struct file *f, const char __user *buf, + size_t size, loff_t *pos) +{ + struct amdgpu_dm_connector *connector = file_inode(f)->i_private; + struct dc_link *link = connector->dc_link; + char *wr_buf = NULL; + char *wr_buf_ptr = NULL; + uint32_t wr_buf_size = 100; + uint32_t wr_buf_count = 0; + int r; + int bytes_from_user; + char *sub_str = NULL; + uint8_t param_index = 0; + uint8_t param_nums = 0; + long param[11] = {0x0}; + const char delimiter[3] = {' ', '\n', '\0'}; + enum dp_test_pattern test_pattern = DP_TEST_PATTERN_UNSUPPORTED; + bool disable_hpd = false; + bool valid_test_pattern = false; + /* init with defalut 80bit custom pattern */ + uint8_t custom_pattern[10] = { + 0x1f, 0x7c, 0xf0, 0xc1, 0x07, + 0x1f, 0x7c, 0xf0, 0xc1, 0x07 + }; + struct dc_link_settings prefer_link_settings = {LANE_COUNT_UNKNOWN, + LINK_RATE_UNKNOWN, LINK_SPREAD_DISABLED}; + struct dc_link_settings cur_link_settings = {LANE_COUNT_UNKNOWN, + LINK_RATE_UNKNOWN, LINK_SPREAD_DISABLED}; + struct link_training_settings link_training_settings; + int i; + + if (size == 0) + return 0; + + wr_buf = kcalloc(wr_buf_size, sizeof(char), GFP_KERNEL); + if (!wr_buf) + return 0; + wr_buf_ptr = wr_buf; + + r = copy_from_user(wr_buf_ptr, buf, wr_buf_size); + + /* r is bytes not be copied */ + if (r >= wr_buf_size) { + kfree(wr_buf); + DRM_DEBUG_DRIVER("user data not be read\n"); + return 0; + } + + bytes_from_user = wr_buf_size - r; + + /* check number of parameters. isspace could not differ space and \n */ + while ((*wr_buf_ptr != 0xa) && (wr_buf_count < wr_buf_size)) { + /* skip space*/ + while (isspace(*wr_buf_ptr) && (wr_buf_count < wr_buf_size)) { + wr_buf_ptr++; + wr_buf_count++; + } + + if (wr_buf_count == wr_buf_size) + break; + + /* skip non-space*/ + while ((!isspace(*wr_buf_ptr)) && (wr_buf_count < wr_buf_size)) { + wr_buf_ptr++; + wr_buf_count++; + } + + param_nums++; + + if (wr_buf_count == wr_buf_size) + break; + } + + /* max 11 parameters */ + if (param_nums > 11) + param_nums = 11; + + wr_buf_ptr = wr_buf; /* reset buf pinter */ + wr_buf_count = 0; /* number of char already checked */ + + while (isspace(*wr_buf_ptr) && (wr_buf_count < wr_buf_size)) { + wr_buf_ptr++; + wr_buf_count++; + } + + while (param_index < param_nums) { + /* after strsep, wr_buf_ptr will be moved to after space */ + sub_str = strsep(&wr_buf_ptr, delimiter); + + r = kstrtol(sub_str, 16, ¶m[param_index]); + + if (r) + DRM_DEBUG_DRIVER("string to int convert error code: %d\n", r); + + param_index++; + } + + test_pattern = param[0]; + + switch (test_pattern) { + case DP_TEST_PATTERN_VIDEO_MODE: + case DP_TEST_PATTERN_COLOR_SQUARES: + case DP_TEST_PATTERN_COLOR_SQUARES_CEA: + case DP_TEST_PATTERN_VERTICAL_BARS: + case DP_TEST_PATTERN_HORIZONTAL_BARS: + case DP_TEST_PATTERN_COLOR_RAMP: + valid_test_pattern = true; + break; + + case DP_TEST_PATTERN_D102: + case DP_TEST_PATTERN_SYMBOL_ERROR: + case DP_TEST_PATTERN_PRBS7: + case DP_TEST_PATTERN_80BIT_CUSTOM: + case DP_TEST_PATTERN_HBR2_COMPLIANCE_EYE: + case DP_TEST_PATTERN_TRAINING_PATTERN4: + disable_hpd = true; + valid_test_pattern = true; + break; + + default: + valid_test_pattern = false; + test_pattern = DP_TEST_PATTERN_UNSUPPORTED; + break; + } + + if (!valid_test_pattern) { + kfree(wr_buf); + DRM_DEBUG_DRIVER("Invalid Test Pattern Parameters\n"); + return bytes_from_user; + } + + if (test_pattern == DP_TEST_PATTERN_80BIT_CUSTOM) { + for (i = 0; i < 10; i++) { + if ((uint8_t) param[i + 1] != 0x0) + break; + } + + if (i < 10) { + /* not use default value */ + for (i = 0; i < 10; i++) + custom_pattern[i] = (uint8_t) param[i + 1]; + } + } + + /* Usage: set DP physical test pattern using debugfs with normal DP + * panel. Then plug out DP panel and connect a scope to measure + * For normal video mode and test pattern generated from CRCT, + * they are visibile to user. So do not disable HPD. + * Video Mode is also set to clear the test pattern, so enable HPD + * because it might have been disabled after a test pattern was set. + * AUX depends on HPD * sequence dependent, do not move! + */ + if (!disable_hpd) + dc_link_enable_hpd(link); + + prefer_link_settings.lane_count = link->verified_link_cap.lane_count; + prefer_link_settings.link_rate = link->verified_link_cap.link_rate; + prefer_link_settings.link_spread = link->verified_link_cap.link_spread; + + cur_link_settings.lane_count = link->cur_link_settings.lane_count; + cur_link_settings.link_rate = link->cur_link_settings.link_rate; + cur_link_settings.link_spread = link->cur_link_settings.link_spread; + + link_training_settings.link_settings = cur_link_settings; + + + if (test_pattern != DP_TEST_PATTERN_VIDEO_MODE) { + if (prefer_link_settings.lane_count != LANE_COUNT_UNKNOWN && + prefer_link_settings.link_rate != LINK_RATE_UNKNOWN && + (prefer_link_settings.lane_count != cur_link_settings.lane_count || + prefer_link_settings.link_rate != cur_link_settings.link_rate)) + link_training_settings.link_settings = prefer_link_settings; + } + + for (i = 0; i < (unsigned int)(link_training_settings.link_settings.lane_count); i++) + link_training_settings.lane_settings[i] = link->cur_lane_setting; + + dc_link_set_test_pattern( + link, + test_pattern, + &link_training_settings, + custom_pattern, + 10); + + /* Usage: Set DP physical test pattern using AMDDP with normal DP panel + * Then plug out DP panel and connect a scope to measure DP PHY signal. + * Need disable interrupt to avoid SW driver disable DP output. This is + * done after the test pattern is set. + */ + if (valid_test_pattern && disable_hpd) + dc_link_disable_hpd(link); + + kfree(wr_buf); + + return bytes_from_user; +} + +static const struct file_operations dp_link_settings_debugfs_fops = { + .owner = THIS_MODULE, + .read = dp_link_settings_read, + .write = dp_link_settings_write, + .llseek = default_llseek +}; + +static const struct file_operations dp_phy_settings_debugfs_fop = { + .owner = THIS_MODULE, + .read = dp_phy_settings_read, + .write = dp_phy_settings_write, + .llseek = default_llseek +}; + +static const struct file_operations dp_phy_test_pattern_fops = { + .owner = THIS_MODULE, + .write = dp_phy_test_pattern_debugfs_write, + .llseek = default_llseek +}; + +static const struct { + char *name; + const struct file_operations *fops; +} dp_debugfs_entries[] = { + {"link_settings", &dp_link_settings_debugfs_fops}, + {"phy_settings", &dp_phy_settings_debugfs_fop}, + {"test_pattern", &dp_phy_test_pattern_fops} +}; + +int connector_debugfs_init(struct amdgpu_dm_connector *connector) +{ + int i; + struct dentry *ent, *dir = connector->base.debugfs_entry; + + if (connector->base.connector_type == DRM_MODE_CONNECTOR_DisplayPort) { + for (i = 0; i < ARRAY_SIZE(dp_debugfs_entries); i++) { + ent = debugfs_create_file(dp_debugfs_entries[i].name, + 0644, + dir, + connector, + dp_debugfs_entries[i].fops); + if (IS_ERR(ent)) + return PTR_ERR(ent); + } + } + + return 0; +} + diff --git a/drivers/gpu/drm/amd/powerplay/inc/pp_power_source.h b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.h index b43315cc5d58..d9ed1b2aa811 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/pp_power_source.h +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.h @@ -1,5 +1,5 @@ /* - * Copyright 2015 Advanced Micro Devices, Inc. + * Copyright 2018 Advanced Micro Devices, Inc. * * Permission is hereby granted, free of charge, to any person obtaining a * copy of this software and associated documentation files (the "Software"), @@ -19,18 +19,16 @@ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR * OTHER DEALINGS IN THE SOFTWARE. * + * Authors: AMD + * */ -#ifndef PP_POWERSOURCE_H -#define PP_POWERSOURCE_H +#ifndef __AMDGPU_DM_DEBUGFS_H__ +#define __AMDGPU_DM_DEBUGFS_H__ -enum pp_power_source { - PP_PowerSource_AC = 0, - PP_PowerSource_DC, - PP_PowerSource_LimitedPower, - PP_PowerSource_LimitedPower_2, - PP_PowerSource_Max -}; +#include "amdgpu.h" +#include "amdgpu_dm.h" +int connector_debugfs_init(struct amdgpu_dm_connector *connector); #endif diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c index ec304b1a5973..8403b6a9a77b 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c @@ -169,6 +169,11 @@ static void get_payload_table( mutex_unlock(&mst_mgr->payload_lock); } +void dm_helpers_dp_update_branch_info( + struct dc_context *ctx, + const struct dc_link *link) +{} + /* * Writes payload allocation table in immediate downstream device. */ @@ -454,6 +459,22 @@ bool dm_helpers_submit_i2c( return result; } +bool dm_helpers_is_dp_sink_present(struct dc_link *link) +{ + bool dp_sink_present; + struct amdgpu_dm_connector *aconnector = link->priv; + + if (!aconnector) { + BUG_ON("Failed to found connector for link!"); + return true; + } + + mutex_lock(&aconnector->dm_dp_aux.aux.hw_mutex); + dp_sink_present = dc_link_is_dp_sink_present(link); + mutex_unlock(&aconnector->dm_dp_aux.aux.hw_mutex); + return dp_sink_present; +} + enum dc_edid_status dm_helpers_read_local_edid( struct dc_context *ctx, struct dc_link *link, @@ -497,6 +518,34 @@ enum dc_edid_status dm_helpers_read_local_edid( DRM_ERROR("EDID err: %d, on connector: %s", edid_status, aconnector->base.name); + if (link->aux_mode) { + union test_request test_request = { {0} }; + union test_response test_response = { {0} }; + + dm_helpers_dp_read_dpcd(ctx, + link, + DP_TEST_REQUEST, + &test_request.raw, + sizeof(union test_request)); + + if (!test_request.bits.EDID_READ) + return edid_status; + + test_response.bits.EDID_CHECKSUM_WRITE = 1; + + dm_helpers_dp_write_dpcd(ctx, + link, + DP_TEST_EDID_CHECKSUM, + &sink->dc_edid.raw_edid[sink->dc_edid.length-1], + 1); + + dm_helpers_dp_write_dpcd(ctx, + link, + DP_TEST_RESPONSE, + &test_response.raw, + sizeof(test_response)); + + } return edid_status; } diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c index ace9ad578ca0..9a300732ba37 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c @@ -80,53 +80,72 @@ static void log_dpcd(uint8_t type, static ssize_t dm_dp_aux_transfer(struct drm_dp_aux *aux, struct drm_dp_aux_msg *msg) { - enum i2c_mot_mode mot = (msg->request & DP_AUX_I2C_MOT) ? - I2C_MOT_TRUE : I2C_MOT_FALSE; - enum ddc_result res; - ssize_t read_bytes; + ssize_t result = 0; + enum i2caux_transaction_action action; + enum aux_transaction_type type; if (WARN_ON(msg->size > 16)) return -E2BIG; switch (msg->request & ~DP_AUX_I2C_MOT) { case DP_AUX_NATIVE_READ: - read_bytes = dal_ddc_service_read_dpcd_data( - TO_DM_AUX(aux)->ddc_service, - false, - I2C_MOT_UNDEF, - msg->address, - msg->buffer, - msg->size); - return read_bytes; + type = AUX_TRANSACTION_TYPE_DP; + action = I2CAUX_TRANSACTION_ACTION_DP_READ; + + result = dc_link_aux_transfer(TO_DM_AUX(aux)->ddc_service, + msg->address, + &msg->reply, + msg->buffer, + msg->size, + type, + action); + break; case DP_AUX_NATIVE_WRITE: - res = dal_ddc_service_write_dpcd_data( - TO_DM_AUX(aux)->ddc_service, - false, - I2C_MOT_UNDEF, - msg->address, - msg->buffer, - msg->size); + type = AUX_TRANSACTION_TYPE_DP; + action = I2CAUX_TRANSACTION_ACTION_DP_WRITE; + + dc_link_aux_transfer(TO_DM_AUX(aux)->ddc_service, + msg->address, + &msg->reply, + msg->buffer, + msg->size, + type, + action); + result = msg->size; break; case DP_AUX_I2C_READ: - read_bytes = dal_ddc_service_read_dpcd_data( - TO_DM_AUX(aux)->ddc_service, - true, - mot, - msg->address, - msg->buffer, - msg->size); - return read_bytes; + type = AUX_TRANSACTION_TYPE_I2C; + if (msg->request & DP_AUX_I2C_MOT) + action = I2CAUX_TRANSACTION_ACTION_I2C_READ_MOT; + else + action = I2CAUX_TRANSACTION_ACTION_I2C_READ; + + result = dc_link_aux_transfer(TO_DM_AUX(aux)->ddc_service, + msg->address, + &msg->reply, + msg->buffer, + msg->size, + type, + action); + break; case DP_AUX_I2C_WRITE: - res = dal_ddc_service_write_dpcd_data( - TO_DM_AUX(aux)->ddc_service, - true, - mot, - msg->address, - msg->buffer, - msg->size); + type = AUX_TRANSACTION_TYPE_I2C; + if (msg->request & DP_AUX_I2C_MOT) + action = I2CAUX_TRANSACTION_ACTION_I2C_WRITE_MOT; + else + action = I2CAUX_TRANSACTION_ACTION_I2C_WRITE; + + dc_link_aux_transfer(TO_DM_AUX(aux)->ddc_service, + msg->address, + &msg->reply, + msg->buffer, + msg->size, + type, + action); + result = msg->size; break; default: - return 0; + return -EINVAL; } #ifdef TRACE_DPCD @@ -137,7 +156,10 @@ static ssize_t dm_dp_aux_transfer(struct drm_dp_aux *aux, r == DDC_RESULT_SUCESSFULL); #endif - return msg->size; + if (result < 0) /* DC doesn't know about kernel error codes */ + result = -EIO; + + return result; } static enum drm_connector_status @@ -229,7 +251,7 @@ static int dm_dp_mst_get_modes(struct drm_connector *connector) edid = drm_dp_mst_get_edid(connector, &aconnector->mst_port->mst_mgr, aconnector->port); if (!edid) { - drm_mode_connector_update_edid_property( + drm_connector_update_edid_property( &aconnector->base, NULL); return ret; @@ -257,7 +279,7 @@ static int dm_dp_mst_get_modes(struct drm_connector *connector) connector, aconnector->edid); } - drm_mode_connector_update_edid_property( + drm_connector_update_edid_property( &aconnector->base, aconnector->edid); ret = drm_add_edid_modes(connector, aconnector->edid); @@ -341,7 +363,7 @@ dm_dp_add_mst_connector(struct drm_dp_mst_topology_mgr *mgr, aconnector, connector->base.id, aconnector->mst_port); aconnector->port = port; - drm_mode_connector_set_path_property(connector, pathprop); + drm_connector_set_path_property(connector, pathprop); drm_connector_list_iter_end(&conn_iter); aconnector->mst_connected = true; @@ -389,7 +411,7 @@ dm_dp_add_mst_connector(struct drm_dp_mst_topology_mgr *mgr, dev->mode_config.tile_property, 0); - drm_mode_connector_set_path_property(connector, pathprop); + drm_connector_set_path_property(connector, pathprop); /* * Initialize connector state before adding the connectror to drm and @@ -437,7 +459,7 @@ static void dm_dp_mst_hotplug(struct drm_dp_mst_topology_mgr *mgr) 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); + drm_connector_set_link_status_property(connector, DRM_MODE_LINK_STATUS_BAD); mutex_unlock(&connector->dev->mode_config.mutex); } diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_pp_smu.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_pp_smu.c new file mode 100644 index 000000000000..fbe878ae1e8c --- /dev/null +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_pp_smu.c @@ -0,0 +1,562 @@ +/* + * Copyright 2018 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: AMD + */ +#include <linux/string.h> +#include <linux/acpi.h> + +#include <drm/drmP.h> +#include <drm/drm_crtc_helper.h> +#include <drm/amdgpu_drm.h> +#include "dm_services.h" +#include "amdgpu.h" +#include "amdgpu_dm.h" +#include "amdgpu_dm_irq.h" +#include "amdgpu_pm.h" +#include "dm_pp_smu.h" + + +bool dm_pp_apply_display_requirements( + const struct dc_context *ctx, + const struct dm_pp_display_configuration *pp_display_cfg) +{ + struct amdgpu_device *adev = ctx->driver_context; + int i; + + if (adev->pm.dpm_enabled) { + + memset(&adev->pm.pm_display_cfg, 0, + sizeof(adev->pm.pm_display_cfg)); + + adev->pm.pm_display_cfg.cpu_cc6_disable = + pp_display_cfg->cpu_cc6_disable; + + adev->pm.pm_display_cfg.cpu_pstate_disable = + pp_display_cfg->cpu_pstate_disable; + + adev->pm.pm_display_cfg.cpu_pstate_separation_time = + pp_display_cfg->cpu_pstate_separation_time; + + adev->pm.pm_display_cfg.nb_pstate_switch_disable = + pp_display_cfg->nb_pstate_switch_disable; + + adev->pm.pm_display_cfg.num_display = + pp_display_cfg->display_count; + adev->pm.pm_display_cfg.num_path_including_non_display = + pp_display_cfg->display_count; + + adev->pm.pm_display_cfg.min_core_set_clock = + pp_display_cfg->min_engine_clock_khz/10; + adev->pm.pm_display_cfg.min_core_set_clock_in_sr = + pp_display_cfg->min_engine_clock_deep_sleep_khz/10; + adev->pm.pm_display_cfg.min_mem_set_clock = + pp_display_cfg->min_memory_clock_khz/10; + + adev->pm.pm_display_cfg.min_dcef_deep_sleep_set_clk = + pp_display_cfg->min_engine_clock_deep_sleep_khz/10; + adev->pm.pm_display_cfg.min_dcef_set_clk = + pp_display_cfg->min_dcfclock_khz/10; + + adev->pm.pm_display_cfg.multi_monitor_in_sync = + pp_display_cfg->all_displays_in_sync; + adev->pm.pm_display_cfg.min_vblank_time = + pp_display_cfg->avail_mclk_switch_time_us; + + adev->pm.pm_display_cfg.display_clk = + pp_display_cfg->disp_clk_khz/10; + + adev->pm.pm_display_cfg.dce_tolerable_mclk_in_active_latency = + pp_display_cfg->avail_mclk_switch_time_in_disp_active_us; + + adev->pm.pm_display_cfg.crtc_index = pp_display_cfg->crtc_index; + adev->pm.pm_display_cfg.line_time_in_us = + pp_display_cfg->line_time_in_us; + + adev->pm.pm_display_cfg.vrefresh = pp_display_cfg->disp_configs[0].v_refresh; + adev->pm.pm_display_cfg.crossfire_display_index = -1; + adev->pm.pm_display_cfg.min_bus_bandwidth = 0; + + for (i = 0; i < pp_display_cfg->display_count; i++) { + const struct dm_pp_single_disp_config *dc_cfg = + &pp_display_cfg->disp_configs[i]; + adev->pm.pm_display_cfg.displays[i].controller_id = dc_cfg->pipe_idx + 1; + } + + /* TODO: complete implementation of + * pp_display_configuration_change(). + * Follow example of: + * PHM_StoreDALConfigurationData - powerplay\hwmgr\hardwaremanager.c + * PP_IRI_DisplayConfigurationChange - powerplay\eventmgr\iri.c */ + if (adev->powerplay.pp_funcs->display_configuration_change) + adev->powerplay.pp_funcs->display_configuration_change( + adev->powerplay.pp_handle, + &adev->pm.pm_display_cfg); + + /* TODO: replace by a separate call to 'apply display cfg'? */ + amdgpu_pm_compute_clocks(adev); + } + + return true; +} + +static void get_default_clock_levels( + enum dm_pp_clock_type clk_type, + struct dm_pp_clock_levels *clks) +{ + uint32_t disp_clks_in_khz[6] = { + 300000, 400000, 496560, 626090, 685720, 757900 }; + uint32_t sclks_in_khz[6] = { + 300000, 360000, 423530, 514290, 626090, 720000 }; + uint32_t mclks_in_khz[2] = { 333000, 800000 }; + + switch (clk_type) { + case DM_PP_CLOCK_TYPE_DISPLAY_CLK: + clks->num_levels = 6; + memmove(clks->clocks_in_khz, disp_clks_in_khz, + sizeof(disp_clks_in_khz)); + break; + case DM_PP_CLOCK_TYPE_ENGINE_CLK: + clks->num_levels = 6; + memmove(clks->clocks_in_khz, sclks_in_khz, + sizeof(sclks_in_khz)); + break; + case DM_PP_CLOCK_TYPE_MEMORY_CLK: + clks->num_levels = 2; + memmove(clks->clocks_in_khz, mclks_in_khz, + sizeof(mclks_in_khz)); + break; + default: + clks->num_levels = 0; + break; + } +} + +static enum amd_pp_clock_type dc_to_pp_clock_type( + enum dm_pp_clock_type dm_pp_clk_type) +{ + enum amd_pp_clock_type amd_pp_clk_type = 0; + + switch (dm_pp_clk_type) { + case DM_PP_CLOCK_TYPE_DISPLAY_CLK: + amd_pp_clk_type = amd_pp_disp_clock; + break; + case DM_PP_CLOCK_TYPE_ENGINE_CLK: + amd_pp_clk_type = amd_pp_sys_clock; + break; + case DM_PP_CLOCK_TYPE_MEMORY_CLK: + amd_pp_clk_type = amd_pp_mem_clock; + break; + case DM_PP_CLOCK_TYPE_DCEFCLK: + amd_pp_clk_type = amd_pp_dcef_clock; + break; + case DM_PP_CLOCK_TYPE_DCFCLK: + amd_pp_clk_type = amd_pp_dcf_clock; + break; + case DM_PP_CLOCK_TYPE_PIXELCLK: + amd_pp_clk_type = amd_pp_pixel_clock; + break; + case DM_PP_CLOCK_TYPE_FCLK: + amd_pp_clk_type = amd_pp_f_clock; + break; + case DM_PP_CLOCK_TYPE_DISPLAYPHYCLK: + amd_pp_clk_type = amd_pp_phy_clock; + break; + case DM_PP_CLOCK_TYPE_DPPCLK: + amd_pp_clk_type = amd_pp_dpp_clock; + break; + default: + DRM_ERROR("DM_PPLIB: invalid clock type: %d!\n", + dm_pp_clk_type); + break; + } + + return amd_pp_clk_type; +} + +static enum dm_pp_clocks_state pp_to_dc_powerlevel_state( + enum PP_DAL_POWERLEVEL max_clocks_state) +{ + switch (max_clocks_state) { + case PP_DAL_POWERLEVEL_0: + return DM_PP_CLOCKS_DPM_STATE_LEVEL_0; + case PP_DAL_POWERLEVEL_1: + return DM_PP_CLOCKS_DPM_STATE_LEVEL_1; + case PP_DAL_POWERLEVEL_2: + return DM_PP_CLOCKS_DPM_STATE_LEVEL_2; + case PP_DAL_POWERLEVEL_3: + return DM_PP_CLOCKS_DPM_STATE_LEVEL_3; + case PP_DAL_POWERLEVEL_4: + return DM_PP_CLOCKS_DPM_STATE_LEVEL_4; + case PP_DAL_POWERLEVEL_5: + return DM_PP_CLOCKS_DPM_STATE_LEVEL_5; + case PP_DAL_POWERLEVEL_6: + return DM_PP_CLOCKS_DPM_STATE_LEVEL_6; + case PP_DAL_POWERLEVEL_7: + return DM_PP_CLOCKS_DPM_STATE_LEVEL_7; + default: + DRM_ERROR("DM_PPLIB: invalid powerlevel state: %d!\n", + max_clocks_state); + return DM_PP_CLOCKS_STATE_INVALID; + } +} + +static void pp_to_dc_clock_levels( + const struct amd_pp_clocks *pp_clks, + struct dm_pp_clock_levels *dc_clks, + enum dm_pp_clock_type dc_clk_type) +{ + uint32_t i; + + if (pp_clks->count > DM_PP_MAX_CLOCK_LEVELS) { + DRM_INFO("DM_PPLIB: Warning: %s clock: number of levels %d exceeds maximum of %d!\n", + DC_DECODE_PP_CLOCK_TYPE(dc_clk_type), + pp_clks->count, + DM_PP_MAX_CLOCK_LEVELS); + + dc_clks->num_levels = DM_PP_MAX_CLOCK_LEVELS; + } else + dc_clks->num_levels = pp_clks->count; + + DRM_INFO("DM_PPLIB: values for %s clock\n", + DC_DECODE_PP_CLOCK_TYPE(dc_clk_type)); + + for (i = 0; i < dc_clks->num_levels; i++) { + DRM_INFO("DM_PPLIB:\t %d\n", pp_clks->clock[i]); + dc_clks->clocks_in_khz[i] = pp_clks->clock[i]; + } +} + +static void pp_to_dc_clock_levels_with_latency( + const struct pp_clock_levels_with_latency *pp_clks, + struct dm_pp_clock_levels_with_latency *clk_level_info, + enum dm_pp_clock_type dc_clk_type) +{ + uint32_t i; + + if (pp_clks->num_levels > DM_PP_MAX_CLOCK_LEVELS) { + DRM_INFO("DM_PPLIB: Warning: %s clock: number of levels %d exceeds maximum of %d!\n", + DC_DECODE_PP_CLOCK_TYPE(dc_clk_type), + pp_clks->num_levels, + DM_PP_MAX_CLOCK_LEVELS); + + clk_level_info->num_levels = DM_PP_MAX_CLOCK_LEVELS; + } else + clk_level_info->num_levels = pp_clks->num_levels; + + DRM_DEBUG("DM_PPLIB: values for %s clock\n", + DC_DECODE_PP_CLOCK_TYPE(dc_clk_type)); + + for (i = 0; i < clk_level_info->num_levels; i++) { + DRM_DEBUG("DM_PPLIB:\t %d in kHz\n", pp_clks->data[i].clocks_in_khz); + clk_level_info->data[i].clocks_in_khz = pp_clks->data[i].clocks_in_khz; + clk_level_info->data[i].latency_in_us = pp_clks->data[i].latency_in_us; + } +} + +static void pp_to_dc_clock_levels_with_voltage( + const struct pp_clock_levels_with_voltage *pp_clks, + struct dm_pp_clock_levels_with_voltage *clk_level_info, + enum dm_pp_clock_type dc_clk_type) +{ + uint32_t i; + + if (pp_clks->num_levels > DM_PP_MAX_CLOCK_LEVELS) { + DRM_INFO("DM_PPLIB: Warning: %s clock: number of levels %d exceeds maximum of %d!\n", + DC_DECODE_PP_CLOCK_TYPE(dc_clk_type), + pp_clks->num_levels, + DM_PP_MAX_CLOCK_LEVELS); + + clk_level_info->num_levels = DM_PP_MAX_CLOCK_LEVELS; + } else + clk_level_info->num_levels = pp_clks->num_levels; + + DRM_INFO("DM_PPLIB: values for %s clock\n", + DC_DECODE_PP_CLOCK_TYPE(dc_clk_type)); + + for (i = 0; i < clk_level_info->num_levels; i++) { + DRM_INFO("DM_PPLIB:\t %d in kHz\n", pp_clks->data[i].clocks_in_khz); + clk_level_info->data[i].clocks_in_khz = pp_clks->data[i].clocks_in_khz; + clk_level_info->data[i].voltage_in_mv = pp_clks->data[i].voltage_in_mv; + } +} + +bool dm_pp_get_clock_levels_by_type( + const struct dc_context *ctx, + enum dm_pp_clock_type clk_type, + struct dm_pp_clock_levels *dc_clks) +{ + struct amdgpu_device *adev = ctx->driver_context; + void *pp_handle = adev->powerplay.pp_handle; + struct amd_pp_clocks pp_clks = { 0 }; + struct amd_pp_simple_clock_info validation_clks = { 0 }; + uint32_t i; + + if (adev->powerplay.pp_funcs->get_clock_by_type) { + if (adev->powerplay.pp_funcs->get_clock_by_type(pp_handle, + dc_to_pp_clock_type(clk_type), &pp_clks)) { + /* Error in pplib. Provide default values. */ + get_default_clock_levels(clk_type, dc_clks); + return true; + } + } + + pp_to_dc_clock_levels(&pp_clks, dc_clks, clk_type); + + if (adev->powerplay.pp_funcs->get_display_mode_validation_clocks) { + if (adev->powerplay.pp_funcs->get_display_mode_validation_clocks( + pp_handle, &validation_clks)) { + /* Error in pplib. Provide default values. */ + DRM_INFO("DM_PPLIB: Warning: using default validation clocks!\n"); + validation_clks.engine_max_clock = 72000; + validation_clks.memory_max_clock = 80000; + validation_clks.level = 0; + } + } + + DRM_INFO("DM_PPLIB: Validation clocks:\n"); + DRM_INFO("DM_PPLIB: engine_max_clock: %d\n", + validation_clks.engine_max_clock); + DRM_INFO("DM_PPLIB: memory_max_clock: %d\n", + validation_clks.memory_max_clock); + DRM_INFO("DM_PPLIB: level : %d\n", + validation_clks.level); + + /* Translate 10 kHz to kHz. */ + validation_clks.engine_max_clock *= 10; + validation_clks.memory_max_clock *= 10; + + /* Determine the highest non-boosted level from the Validation Clocks */ + if (clk_type == DM_PP_CLOCK_TYPE_ENGINE_CLK) { + for (i = 0; i < dc_clks->num_levels; i++) { + if (dc_clks->clocks_in_khz[i] > validation_clks.engine_max_clock) { + /* This clock is higher the validation clock. + * Than means the previous one is the highest + * non-boosted one. */ + DRM_INFO("DM_PPLIB: reducing engine clock level from %d to %d\n", + dc_clks->num_levels, i); + dc_clks->num_levels = i > 0 ? i : 1; + break; + } + } + } else if (clk_type == DM_PP_CLOCK_TYPE_MEMORY_CLK) { + for (i = 0; i < dc_clks->num_levels; i++) { + if (dc_clks->clocks_in_khz[i] > validation_clks.memory_max_clock) { + DRM_INFO("DM_PPLIB: reducing memory clock level from %d to %d\n", + dc_clks->num_levels, i); + dc_clks->num_levels = i > 0 ? i : 1; + break; + } + } + } + + return true; +} + +bool dm_pp_get_clock_levels_by_type_with_latency( + const struct dc_context *ctx, + enum dm_pp_clock_type clk_type, + struct dm_pp_clock_levels_with_latency *clk_level_info) +{ + struct amdgpu_device *adev = ctx->driver_context; + void *pp_handle = adev->powerplay.pp_handle; + struct pp_clock_levels_with_latency pp_clks = { 0 }; + const struct amd_pm_funcs *pp_funcs = adev->powerplay.pp_funcs; + + if (!pp_funcs || !pp_funcs->get_clock_by_type_with_latency) + return false; + + if (pp_funcs->get_clock_by_type_with_latency(pp_handle, + dc_to_pp_clock_type(clk_type), + &pp_clks)) + return false; + + pp_to_dc_clock_levels_with_latency(&pp_clks, clk_level_info, clk_type); + + return true; +} + +bool dm_pp_get_clock_levels_by_type_with_voltage( + const struct dc_context *ctx, + enum dm_pp_clock_type clk_type, + struct dm_pp_clock_levels_with_voltage *clk_level_info) +{ + struct amdgpu_device *adev = ctx->driver_context; + void *pp_handle = adev->powerplay.pp_handle; + struct pp_clock_levels_with_voltage pp_clk_info = {0}; + const struct amd_pm_funcs *pp_funcs = adev->powerplay.pp_funcs; + + if (pp_funcs->get_clock_by_type_with_voltage(pp_handle, + dc_to_pp_clock_type(clk_type), + &pp_clk_info)) + return false; + + pp_to_dc_clock_levels_with_voltage(&pp_clk_info, clk_level_info, clk_type); + + return true; +} + +bool dm_pp_notify_wm_clock_changes( + const struct dc_context *ctx, + struct dm_pp_wm_sets_with_clock_ranges *wm_with_clock_ranges) +{ + /* TODO: to be implemented */ + return false; +} + +bool dm_pp_apply_power_level_change_request( + const struct dc_context *ctx, + struct dm_pp_power_level_change_request *level_change_req) +{ + /* TODO: to be implemented */ + return false; +} + +bool dm_pp_apply_clock_for_voltage_request( + const struct dc_context *ctx, + struct dm_pp_clock_for_voltage_req *clock_for_voltage_req) +{ + struct amdgpu_device *adev = ctx->driver_context; + struct pp_display_clock_request pp_clock_request = {0}; + int ret = 0; + + pp_clock_request.clock_type = dc_to_pp_clock_type(clock_for_voltage_req->clk_type); + pp_clock_request.clock_freq_in_khz = clock_for_voltage_req->clocks_in_khz; + + if (!pp_clock_request.clock_type) + return false; + + if (adev->powerplay.pp_funcs->display_clock_voltage_request) + ret = adev->powerplay.pp_funcs->display_clock_voltage_request( + adev->powerplay.pp_handle, + &pp_clock_request); + if (ret) + return false; + return true; +} + +bool dm_pp_get_static_clocks( + const struct dc_context *ctx, + struct dm_pp_static_clock_info *static_clk_info) +{ + struct amdgpu_device *adev = ctx->driver_context; + struct amd_pp_clock_info pp_clk_info = {0}; + int ret = 0; + + if (adev->powerplay.pp_funcs->get_current_clocks) + ret = adev->powerplay.pp_funcs->get_current_clocks( + adev->powerplay.pp_handle, + &pp_clk_info); + if (ret) + return false; + + static_clk_info->max_clocks_state = pp_to_dc_powerlevel_state(pp_clk_info.max_clocks_state); + static_clk_info->max_mclk_khz = pp_clk_info.max_memory_clock * 10; + static_clk_info->max_sclk_khz = pp_clk_info.max_engine_clock * 10; + + return true; +} + +void pp_rv_set_display_requirement(struct pp_smu *pp, + struct pp_smu_display_requirement_rv *req) +{ + struct dc_context *ctx = pp->ctx; + struct amdgpu_device *adev = ctx->driver_context; + const struct amd_pm_funcs *pp_funcs = adev->powerplay.pp_funcs; + + if (!pp_funcs || !pp_funcs->display_configuration_changed) + return; + + amdgpu_dpm_display_configuration_changed(adev); +} + +void pp_rv_set_wm_ranges(struct pp_smu *pp, + struct pp_smu_wm_range_sets *ranges) +{ + struct dc_context *ctx = pp->ctx; + struct amdgpu_device *adev = ctx->driver_context; + void *pp_handle = adev->powerplay.pp_handle; + const struct amd_pm_funcs *pp_funcs = adev->powerplay.pp_funcs; + struct dm_pp_wm_sets_with_clock_ranges_soc15 wm_with_clock_ranges; + struct dm_pp_clock_range_for_dmif_wm_set_soc15 *wm_dce_clocks = wm_with_clock_ranges.wm_dmif_clocks_ranges; + struct dm_pp_clock_range_for_mcif_wm_set_soc15 *wm_soc_clocks = wm_with_clock_ranges.wm_mcif_clocks_ranges; + int32_t i; + + wm_with_clock_ranges.num_wm_dmif_sets = ranges->num_reader_wm_sets; + wm_with_clock_ranges.num_wm_mcif_sets = ranges->num_writer_wm_sets; + + for (i = 0; i < wm_with_clock_ranges.num_wm_dmif_sets; i++) { + if (ranges->reader_wm_sets[i].wm_inst > 3) + wm_dce_clocks[i].wm_set_id = WM_SET_A; + else + wm_dce_clocks[i].wm_set_id = + ranges->reader_wm_sets[i].wm_inst; + wm_dce_clocks[i].wm_max_dcfclk_clk_in_khz = + ranges->reader_wm_sets[i].max_drain_clk_khz; + wm_dce_clocks[i].wm_min_dcfclk_clk_in_khz = + ranges->reader_wm_sets[i].min_drain_clk_khz; + wm_dce_clocks[i].wm_max_mem_clk_in_khz = + ranges->reader_wm_sets[i].max_fill_clk_khz; + wm_dce_clocks[i].wm_min_mem_clk_in_khz = + ranges->reader_wm_sets[i].min_fill_clk_khz; + } + + for (i = 0; i < wm_with_clock_ranges.num_wm_mcif_sets; i++) { + if (ranges->writer_wm_sets[i].wm_inst > 3) + wm_soc_clocks[i].wm_set_id = WM_SET_A; + else + wm_soc_clocks[i].wm_set_id = + ranges->writer_wm_sets[i].wm_inst; + wm_soc_clocks[i].wm_max_socclk_clk_in_khz = + ranges->writer_wm_sets[i].max_fill_clk_khz; + wm_soc_clocks[i].wm_min_socclk_clk_in_khz = + ranges->writer_wm_sets[i].min_fill_clk_khz; + wm_soc_clocks[i].wm_max_mem_clk_in_khz = + ranges->writer_wm_sets[i].max_drain_clk_khz; + wm_soc_clocks[i].wm_min_mem_clk_in_khz = + ranges->writer_wm_sets[i].min_drain_clk_khz; + } + + pp_funcs->set_watermarks_for_clocks_ranges(pp_handle, &wm_with_clock_ranges); +} + +void pp_rv_set_pme_wa_enable(struct pp_smu *pp) +{ + struct dc_context *ctx = pp->ctx; + struct amdgpu_device *adev = ctx->driver_context; + void *pp_handle = adev->powerplay.pp_handle; + const struct amd_pm_funcs *pp_funcs = adev->powerplay.pp_funcs; + + if (!pp_funcs || !pp_funcs->notify_smu_enable_pwe) + return; + + pp_funcs->notify_smu_enable_pwe(pp_handle); +} + +void dm_pp_get_funcs_rv( + struct dc_context *ctx, + struct pp_smu_funcs_rv *funcs) +{ + funcs->pp_smu.ctx = ctx; + funcs->set_display_requirement = pp_rv_set_display_requirement; + funcs->set_wm_ranges = pp_rv_set_wm_ranges; + funcs->set_pme_wa_enable = pp_rv_set_pme_wa_enable; +} diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c index 5a2e952c5bea..516795342dd2 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c @@ -35,13 +35,7 @@ #include "amdgpu_dm_irq.h" #include "amdgpu_pm.h" -unsigned long long dm_get_timestamp(struct dc_context *ctx) -{ - struct timespec64 time; - getrawmonotonic64(&time); - return timespec64_to_ns(&time); -} unsigned long long dm_get_elapse_time_in_ns(struct dc_context *ctx, unsigned long long current_time_stamp, @@ -80,327 +74,3 @@ bool dm_read_persistent_data(struct dc_context *ctx, /**** power component interfaces ****/ -bool dm_pp_apply_display_requirements( - const struct dc_context *ctx, - const struct dm_pp_display_configuration *pp_display_cfg) -{ - struct amdgpu_device *adev = ctx->driver_context; - - if (adev->pm.dpm_enabled) { - - memset(&adev->pm.pm_display_cfg, 0, - sizeof(adev->pm.pm_display_cfg)); - - adev->pm.pm_display_cfg.cpu_cc6_disable = - pp_display_cfg->cpu_cc6_disable; - - adev->pm.pm_display_cfg.cpu_pstate_disable = - pp_display_cfg->cpu_pstate_disable; - - adev->pm.pm_display_cfg.cpu_pstate_separation_time = - pp_display_cfg->cpu_pstate_separation_time; - - adev->pm.pm_display_cfg.nb_pstate_switch_disable = - pp_display_cfg->nb_pstate_switch_disable; - - adev->pm.pm_display_cfg.num_display = - pp_display_cfg->display_count; - adev->pm.pm_display_cfg.num_path_including_non_display = - pp_display_cfg->display_count; - - adev->pm.pm_display_cfg.min_core_set_clock = - pp_display_cfg->min_engine_clock_khz/10; - adev->pm.pm_display_cfg.min_core_set_clock_in_sr = - pp_display_cfg->min_engine_clock_deep_sleep_khz/10; - adev->pm.pm_display_cfg.min_mem_set_clock = - pp_display_cfg->min_memory_clock_khz/10; - - adev->pm.pm_display_cfg.multi_monitor_in_sync = - pp_display_cfg->all_displays_in_sync; - adev->pm.pm_display_cfg.min_vblank_time = - pp_display_cfg->avail_mclk_switch_time_us; - - adev->pm.pm_display_cfg.display_clk = - pp_display_cfg->disp_clk_khz/10; - - adev->pm.pm_display_cfg.dce_tolerable_mclk_in_active_latency = - pp_display_cfg->avail_mclk_switch_time_in_disp_active_us; - - adev->pm.pm_display_cfg.crtc_index = pp_display_cfg->crtc_index; - adev->pm.pm_display_cfg.line_time_in_us = - pp_display_cfg->line_time_in_us; - - adev->pm.pm_display_cfg.vrefresh = pp_display_cfg->disp_configs[0].v_refresh; - adev->pm.pm_display_cfg.crossfire_display_index = -1; - adev->pm.pm_display_cfg.min_bus_bandwidth = 0; - - /* TODO: complete implementation of - * pp_display_configuration_change(). - * Follow example of: - * PHM_StoreDALConfigurationData - powerplay\hwmgr\hardwaremanager.c - * PP_IRI_DisplayConfigurationChange - powerplay\eventmgr\iri.c */ - if (adev->powerplay.pp_funcs->display_configuration_change) - adev->powerplay.pp_funcs->display_configuration_change( - adev->powerplay.pp_handle, - &adev->pm.pm_display_cfg); - - /* TODO: replace by a separate call to 'apply display cfg'? */ - amdgpu_pm_compute_clocks(adev); - } - - return true; -} - -static void get_default_clock_levels( - enum dm_pp_clock_type clk_type, - struct dm_pp_clock_levels *clks) -{ - uint32_t disp_clks_in_khz[6] = { - 300000, 400000, 496560, 626090, 685720, 757900 }; - uint32_t sclks_in_khz[6] = { - 300000, 360000, 423530, 514290, 626090, 720000 }; - uint32_t mclks_in_khz[2] = { 333000, 800000 }; - - switch (clk_type) { - case DM_PP_CLOCK_TYPE_DISPLAY_CLK: - clks->num_levels = 6; - memmove(clks->clocks_in_khz, disp_clks_in_khz, - sizeof(disp_clks_in_khz)); - break; - case DM_PP_CLOCK_TYPE_ENGINE_CLK: - clks->num_levels = 6; - memmove(clks->clocks_in_khz, sclks_in_khz, - sizeof(sclks_in_khz)); - break; - case DM_PP_CLOCK_TYPE_MEMORY_CLK: - clks->num_levels = 2; - memmove(clks->clocks_in_khz, mclks_in_khz, - sizeof(mclks_in_khz)); - break; - default: - clks->num_levels = 0; - break; - } -} - -static enum amd_pp_clock_type dc_to_pp_clock_type( - enum dm_pp_clock_type dm_pp_clk_type) -{ - enum amd_pp_clock_type amd_pp_clk_type = 0; - - switch (dm_pp_clk_type) { - case DM_PP_CLOCK_TYPE_DISPLAY_CLK: - amd_pp_clk_type = amd_pp_disp_clock; - break; - case DM_PP_CLOCK_TYPE_ENGINE_CLK: - amd_pp_clk_type = amd_pp_sys_clock; - break; - case DM_PP_CLOCK_TYPE_MEMORY_CLK: - amd_pp_clk_type = amd_pp_mem_clock; - break; - default: - DRM_ERROR("DM_PPLIB: invalid clock type: %d!\n", - dm_pp_clk_type); - break; - } - - return amd_pp_clk_type; -} - -static void pp_to_dc_clock_levels( - const struct amd_pp_clocks *pp_clks, - struct dm_pp_clock_levels *dc_clks, - enum dm_pp_clock_type dc_clk_type) -{ - uint32_t i; - - if (pp_clks->count > DM_PP_MAX_CLOCK_LEVELS) { - DRM_INFO("DM_PPLIB: Warning: %s clock: number of levels %d exceeds maximum of %d!\n", - DC_DECODE_PP_CLOCK_TYPE(dc_clk_type), - pp_clks->count, - DM_PP_MAX_CLOCK_LEVELS); - - dc_clks->num_levels = DM_PP_MAX_CLOCK_LEVELS; - } else - dc_clks->num_levels = pp_clks->count; - - DRM_INFO("DM_PPLIB: values for %s clock\n", - DC_DECODE_PP_CLOCK_TYPE(dc_clk_type)); - - for (i = 0; i < dc_clks->num_levels; i++) { - DRM_INFO("DM_PPLIB:\t %d\n", pp_clks->clock[i]); - /* translate 10kHz to kHz */ - dc_clks->clocks_in_khz[i] = pp_clks->clock[i] * 10; - } -} - -static void pp_to_dc_clock_levels_with_latency( - const struct pp_clock_levels_with_latency *pp_clks, - struct dm_pp_clock_levels_with_latency *clk_level_info, - enum dm_pp_clock_type dc_clk_type) -{ - uint32_t i; - - if (pp_clks->num_levels > DM_PP_MAX_CLOCK_LEVELS) { - DRM_INFO("DM_PPLIB: Warning: %s clock: number of levels %d exceeds maximum of %d!\n", - DC_DECODE_PP_CLOCK_TYPE(dc_clk_type), - pp_clks->num_levels, - DM_PP_MAX_CLOCK_LEVELS); - - clk_level_info->num_levels = DM_PP_MAX_CLOCK_LEVELS; - } else - clk_level_info->num_levels = pp_clks->num_levels; - - DRM_DEBUG("DM_PPLIB: values for %s clock\n", - DC_DECODE_PP_CLOCK_TYPE(dc_clk_type)); - - for (i = 0; i < clk_level_info->num_levels; i++) { - DRM_DEBUG("DM_PPLIB:\t %d in 10kHz\n", pp_clks->data[i].clocks_in_khz); - /* translate 10kHz to kHz */ - clk_level_info->data[i].clocks_in_khz = pp_clks->data[i].clocks_in_khz * 10; - clk_level_info->data[i].latency_in_us = pp_clks->data[i].latency_in_us; - } -} - -bool dm_pp_get_clock_levels_by_type( - const struct dc_context *ctx, - enum dm_pp_clock_type clk_type, - struct dm_pp_clock_levels *dc_clks) -{ - struct amdgpu_device *adev = ctx->driver_context; - void *pp_handle = adev->powerplay.pp_handle; - struct amd_pp_clocks pp_clks = { 0 }; - struct amd_pp_simple_clock_info validation_clks = { 0 }; - uint32_t i; - - if (adev->powerplay.pp_funcs->get_clock_by_type) { - if (adev->powerplay.pp_funcs->get_clock_by_type(pp_handle, - dc_to_pp_clock_type(clk_type), &pp_clks)) { - /* Error in pplib. Provide default values. */ - get_default_clock_levels(clk_type, dc_clks); - return true; - } - } - - pp_to_dc_clock_levels(&pp_clks, dc_clks, clk_type); - - if (adev->powerplay.pp_funcs->get_display_mode_validation_clocks) { - if (adev->powerplay.pp_funcs->get_display_mode_validation_clocks( - pp_handle, &validation_clks)) { - /* Error in pplib. Provide default values. */ - DRM_INFO("DM_PPLIB: Warning: using default validation clocks!\n"); - validation_clks.engine_max_clock = 72000; - validation_clks.memory_max_clock = 80000; - validation_clks.level = 0; - } - } - - DRM_INFO("DM_PPLIB: Validation clocks:\n"); - DRM_INFO("DM_PPLIB: engine_max_clock: %d\n", - validation_clks.engine_max_clock); - DRM_INFO("DM_PPLIB: memory_max_clock: %d\n", - validation_clks.memory_max_clock); - DRM_INFO("DM_PPLIB: level : %d\n", - validation_clks.level); - - /* Translate 10 kHz to kHz. */ - validation_clks.engine_max_clock *= 10; - validation_clks.memory_max_clock *= 10; - - /* Determine the highest non-boosted level from the Validation Clocks */ - if (clk_type == DM_PP_CLOCK_TYPE_ENGINE_CLK) { - for (i = 0; i < dc_clks->num_levels; i++) { - if (dc_clks->clocks_in_khz[i] > validation_clks.engine_max_clock) { - /* This clock is higher the validation clock. - * Than means the previous one is the highest - * non-boosted one. */ - DRM_INFO("DM_PPLIB: reducing engine clock level from %d to %d\n", - dc_clks->num_levels, i); - dc_clks->num_levels = i > 0 ? i : 1; - break; - } - } - } else if (clk_type == DM_PP_CLOCK_TYPE_MEMORY_CLK) { - for (i = 0; i < dc_clks->num_levels; i++) { - if (dc_clks->clocks_in_khz[i] > validation_clks.memory_max_clock) { - DRM_INFO("DM_PPLIB: reducing memory clock level from %d to %d\n", - dc_clks->num_levels, i); - dc_clks->num_levels = i > 0 ? i : 1; - break; - } - } - } - - return true; -} - -bool dm_pp_get_clock_levels_by_type_with_latency( - const struct dc_context *ctx, - enum dm_pp_clock_type clk_type, - struct dm_pp_clock_levels_with_latency *clk_level_info) -{ - struct amdgpu_device *adev = ctx->driver_context; - void *pp_handle = adev->powerplay.pp_handle; - struct pp_clock_levels_with_latency pp_clks = { 0 }; - const struct amd_pm_funcs *pp_funcs = adev->powerplay.pp_funcs; - - if (!pp_funcs || !pp_funcs->get_clock_by_type_with_latency) - return false; - - if (pp_funcs->get_clock_by_type_with_latency(pp_handle, - dc_to_pp_clock_type(clk_type), - &pp_clks)) - return false; - - pp_to_dc_clock_levels_with_latency(&pp_clks, clk_level_info, clk_type); - - return true; -} - -bool dm_pp_get_clock_levels_by_type_with_voltage( - const struct dc_context *ctx, - enum dm_pp_clock_type clk_type, - struct dm_pp_clock_levels_with_voltage *clk_level_info) -{ - /* TODO: to be implemented */ - return false; -} - -bool dm_pp_notify_wm_clock_changes( - const struct dc_context *ctx, - struct dm_pp_wm_sets_with_clock_ranges *wm_with_clock_ranges) -{ - /* TODO: to be implemented */ - return false; -} - -bool dm_pp_apply_power_level_change_request( - const struct dc_context *ctx, - struct dm_pp_power_level_change_request *level_change_req) -{ - /* TODO: to be implemented */ - return false; -} - -bool dm_pp_apply_clock_for_voltage_request( - const struct dc_context *ctx, - struct dm_pp_clock_for_voltage_req *clock_for_voltage_req) -{ - /* TODO: to be implemented */ - return false; -} - -bool dm_pp_get_static_clocks( - const struct dc_context *ctx, - struct dm_pp_static_clock_info *static_clk_info) -{ - /* TODO: to be implemented */ - return false; -} - -void dm_pp_get_funcs_rv( - struct dc_context *ctx, - struct pp_smu_funcs_rv *funcs) -{} - -/**** end of power component interfaces ****/ diff --git a/drivers/gpu/drm/amd/display/dc/Makefile b/drivers/gpu/drm/amd/display/dc/Makefile index aed538a4d1ba..532a515fda9a 100644 --- a/drivers/gpu/drm/amd/display/dc/Makefile +++ b/drivers/gpu/drm/amd/display/dc/Makefile @@ -25,7 +25,7 @@ DC_LIBS = basics bios calcs dce gpio i2caux irq virtual -ifdef CONFIG_DRM_AMD_DC_DCN1_0 +ifdef CONFIG_X86 DC_LIBS += dcn10 dml endif diff --git a/drivers/gpu/drm/amd/display/dc/basics/Makefile b/drivers/gpu/drm/amd/display/dc/basics/Makefile index b49ea96b5dae..a50a76471107 100644 --- a/drivers/gpu/drm/amd/display/dc/basics/Makefile +++ b/drivers/gpu/drm/amd/display/dc/basics/Makefile @@ -25,7 +25,7 @@ # subcomponents. BASICS = conversion.o fixpt31_32.o \ - logger.o log_helpers.o vector.o + log_helpers.o vector.o AMD_DAL_BASICS = $(addprefix $(AMDDALPATH)/dc/basics/,$(BASICS)) diff --git a/drivers/gpu/drm/amd/display/dc/basics/log_helpers.c b/drivers/gpu/drm/amd/display/dc/basics/log_helpers.c index 021451549ff7..26583f346c39 100644 --- a/drivers/gpu/drm/amd/display/dc/basics/log_helpers.c +++ b/drivers/gpu/drm/amd/display/dc/basics/log_helpers.c @@ -28,75 +28,12 @@ #include "include/logger_interface.h" #include "dm_helpers.h" -#define NUM_ELEMENTS(a) (sizeof(a) / sizeof((a)[0])) - -struct dc_signal_type_info { - enum signal_type type; - char name[MAX_NAME_LEN]; -}; - -static const struct dc_signal_type_info signal_type_info_tbl[] = { - {SIGNAL_TYPE_NONE, "NC"}, - {SIGNAL_TYPE_DVI_SINGLE_LINK, "DVI"}, - {SIGNAL_TYPE_DVI_DUAL_LINK, "DDVI"}, - {SIGNAL_TYPE_HDMI_TYPE_A, "HDMIA"}, - {SIGNAL_TYPE_LVDS, "LVDS"}, - {SIGNAL_TYPE_RGB, "VGA"}, - {SIGNAL_TYPE_DISPLAY_PORT, "DP"}, - {SIGNAL_TYPE_DISPLAY_PORT_MST, "MST"}, - {SIGNAL_TYPE_EDP, "eDP"}, - {SIGNAL_TYPE_VIRTUAL, "Virtual"} -}; - -void dc_conn_log(struct dc_context *ctx, - const struct dc_link *link, - uint8_t *hex_data, - int hex_data_count, - enum dc_log_type event, - const char *msg, - ...) +void dc_conn_log_hex_linux(const uint8_t *hex_data, int hex_data_count) { int i; - va_list args; - struct log_entry entry = { 0 }; - enum signal_type signal; - - if (link->local_sink) - signal = link->local_sink->sink_signal; - else - signal = link->connector_signal; - - if (link->type == dc_connection_mst_branch) - signal = SIGNAL_TYPE_DISPLAY_PORT_MST; - - dm_logger_open(ctx->logger, &entry, event); - - for (i = 0; i < NUM_ELEMENTS(signal_type_info_tbl); i++) - if (signal == signal_type_info_tbl[i].type) - break; - - if (i == NUM_ELEMENTS(signal_type_info_tbl)) - goto fail; - - dm_logger_append(&entry, "[%s][ConnIdx:%d] ", - signal_type_info_tbl[i].name, - link->link_index); - - va_start(args, msg); - dm_logger_append_va(&entry, msg, args); - - if (entry.buf_offset > 0 && - entry.buf[entry.buf_offset - 1] == '\n') - entry.buf_offset--; if (hex_data) for (i = 0; i < hex_data_count; i++) - dm_logger_append(&entry, "%2.2X ", hex_data[i]); - - dm_logger_append(&entry, "^\n"); - -fail: - dm_logger_close(&entry); - - va_end(args); + DC_LOG_DEBUG("%2.2X ", hex_data[i]); } + diff --git a/drivers/gpu/drm/amd/display/dc/basics/logger.c b/drivers/gpu/drm/amd/display/dc/basics/logger.c deleted file mode 100644 index 0866874ae8c6..000000000000 --- a/drivers/gpu/drm/amd/display/dc/basics/logger.c +++ /dev/null @@ -1,405 +0,0 @@ -/* - * Copyright 2012-15 Advanced Micro Devices, Inc. - * - * Permission is hereby granted, free of charge, to any person obtaining a - * copy of this software and associated documentation files (the "Software"), - * to deal in the Software without restriction, including without limitation - * the rights to use, copy, modify, merge, publish, distribute, sublicense, - * and/or sell copies of the Software, and to permit persons to whom the - * Software is furnished to do so, subject to the following conditions: - * - * The above copyright notice and this permission notice shall be included in - * all copies or substantial portions of the Software. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR - * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, - * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL - * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR - * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, - * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR - * OTHER DEALINGS IN THE SOFTWARE. - * - * Authors: AMD - * - */ -#include "dm_services.h" -#include "include/logger_interface.h" -#include "logger.h" - - -#define NUM_ELEMENTS(a) (sizeof(a) / sizeof((a)[0])) - -static const struct dc_log_type_info log_type_info_tbl[] = { - {LOG_ERROR, "Error"}, - {LOG_WARNING, "Warning"}, - {LOG_DEBUG, "Debug"}, - {LOG_DC, "DC_Interface"}, - {LOG_SURFACE, "Surface"}, - {LOG_HW_HOTPLUG, "HW_Hotplug"}, - {LOG_HW_LINK_TRAINING, "HW_LKTN"}, - {LOG_HW_SET_MODE, "HW_Mode"}, - {LOG_HW_RESUME_S3, "HW_Resume"}, - {LOG_HW_AUDIO, "HW_Audio"}, - {LOG_HW_HPD_IRQ, "HW_HPDIRQ"}, - {LOG_MST, "MST"}, - {LOG_SCALER, "Scaler"}, - {LOG_BIOS, "BIOS"}, - {LOG_BANDWIDTH_CALCS, "BWCalcs"}, - {LOG_BANDWIDTH_VALIDATION, "BWValidation"}, - {LOG_I2C_AUX, "I2C_AUX"}, - {LOG_SYNC, "Sync"}, - {LOG_BACKLIGHT, "Backlight"}, - {LOG_FEATURE_OVERRIDE, "Override"}, - {LOG_DETECTION_EDID_PARSER, "Edid"}, - {LOG_DETECTION_DP_CAPS, "DP_Caps"}, - {LOG_RESOURCE, "Resource"}, - {LOG_DML, "DML"}, - {LOG_EVENT_MODE_SET, "Mode"}, - {LOG_EVENT_DETECTION, "Detect"}, - {LOG_EVENT_LINK_TRAINING, "LKTN"}, - {LOG_EVENT_LINK_LOSS, "LinkLoss"}, - {LOG_EVENT_UNDERFLOW, "Underflow"}, - {LOG_IF_TRACE, "InterfaceTrace"}, - {LOG_DTN, "DTN"}, - {LOG_DISPLAYSTATS, "DisplayStats"} -}; - - -/* ----------- Object init and destruction ----------- */ -static bool construct(struct dc_context *ctx, struct dal_logger *logger, - uint32_t log_mask) -{ - /* malloc buffer and init offsets */ - logger->log_buffer_size = DAL_LOGGER_BUFFER_MAX_SIZE; - logger->log_buffer = kcalloc(logger->log_buffer_size, sizeof(char), - GFP_KERNEL); - if (!logger->log_buffer) - return false; - - /* Initialize both offsets to start of buffer (empty) */ - logger->buffer_read_offset = 0; - logger->buffer_write_offset = 0; - - logger->open_count = 0; - - logger->flags.bits.ENABLE_CONSOLE = 1; - logger->flags.bits.ENABLE_BUFFER = 0; - - logger->ctx = ctx; - - logger->mask = log_mask; - - return true; -} - -static void destruct(struct dal_logger *logger) -{ - if (logger->log_buffer) { - kfree(logger->log_buffer); - logger->log_buffer = NULL; - } -} - -struct dal_logger *dal_logger_create(struct dc_context *ctx, uint32_t log_mask) -{ - /* malloc struct */ - struct dal_logger *logger = kzalloc(sizeof(struct dal_logger), - GFP_KERNEL); - - if (!logger) - return NULL; - if (!construct(ctx, logger, log_mask)) { - kfree(logger); - return NULL; - } - - return logger; -} - -uint32_t dal_logger_destroy(struct dal_logger **logger) -{ - if (logger == NULL || *logger == NULL) - return 1; - destruct(*logger); - kfree(*logger); - *logger = NULL; - - return 0; -} - -/* ------------------------------------------------------------------------ */ - - -static bool dal_logger_should_log( - struct dal_logger *logger, - enum dc_log_type log_type) -{ - if (logger->mask & (1 << log_type)) - return true; - - return false; -} - -static void log_to_debug_console(struct log_entry *entry) -{ - struct dal_logger *logger = entry->logger; - - if (logger->flags.bits.ENABLE_CONSOLE == 0) - return; - - if (entry->buf_offset) { - switch (entry->type) { - case LOG_ERROR: - dm_error("%s", entry->buf); - break; - default: - dm_output_to_console("%s", entry->buf); - break; - } - } -} - -/* Print everything unread existing in log_buffer to debug console*/ -void dm_logger_flush_buffer(struct dal_logger *logger, bool should_warn) -{ - char *string_start = &logger->log_buffer[logger->buffer_read_offset]; - - if (should_warn) - dm_output_to_console( - "---------------- FLUSHING LOG BUFFER ----------------\n"); - while (logger->buffer_read_offset < logger->buffer_write_offset) { - - if (logger->log_buffer[logger->buffer_read_offset] == '\0') { - dm_output_to_console("%s", string_start); - string_start = logger->log_buffer + logger->buffer_read_offset + 1; - } - logger->buffer_read_offset++; - } - if (should_warn) - dm_output_to_console( - "-------------- END FLUSHING LOG BUFFER --------------\n\n"); -} - -static void log_to_internal_buffer(struct log_entry *entry) -{ - - uint32_t size = entry->buf_offset; - struct dal_logger *logger = entry->logger; - - if (logger->flags.bits.ENABLE_BUFFER == 0) - return; - - if (logger->log_buffer == NULL) - return; - - if (size > 0 && size < logger->log_buffer_size) { - - int buffer_space = logger->log_buffer_size - - logger->buffer_write_offset; - - if (logger->buffer_write_offset == logger->buffer_read_offset) { - /* Buffer is empty, start writing at beginning */ - buffer_space = logger->log_buffer_size; - logger->buffer_write_offset = 0; - logger->buffer_read_offset = 0; - } - - if (buffer_space > size) { - /* No wrap around, copy 'size' bytes - * from 'entry->buf' to 'log_buffer' - */ - memmove(logger->log_buffer + - logger->buffer_write_offset, - entry->buf, size); - logger->buffer_write_offset += size; - - } else { - /* Not enough room remaining, we should flush - * existing logs */ - - /* Flush existing unread logs to console */ - dm_logger_flush_buffer(logger, true); - - /* Start writing to beginning of buffer */ - memmove(logger->log_buffer, entry->buf, size); - logger->buffer_write_offset = size; - logger->buffer_read_offset = 0; - } - - } -} - -static void log_heading(struct log_entry *entry) -{ - int j; - - for (j = 0; j < NUM_ELEMENTS(log_type_info_tbl); j++) { - - const struct dc_log_type_info *info = &log_type_info_tbl[j]; - - if (info->type == entry->type) - dm_logger_append(entry, "[%s]\t", info->name); - } -} - -static void append_entry( - struct log_entry *entry, - char *buffer, - uint32_t buf_size) -{ - if (!entry->buf || - entry->buf_offset + buf_size > entry->max_buf_bytes - ) { - BREAK_TO_DEBUGGER(); - return; - } - - /* Todo: check if off by 1 byte due to \0 anywhere */ - memmove(entry->buf + entry->buf_offset, buffer, buf_size); - entry->buf_offset += buf_size; -} - -/* ------------------------------------------------------------------------ */ - -/* Warning: Be careful that 'msg' is null terminated and the total size is - * less than DAL_LOGGER_BUFFER_MAX_LOG_LINE_SIZE (256) including '\0' - */ -void dm_logger_write( - struct dal_logger *logger, - enum dc_log_type log_type, - const char *msg, - ...) -{ - if (logger && dal_logger_should_log(logger, log_type)) { - uint32_t size; - va_list args; - char buffer[LOG_MAX_LINE_SIZE]; - struct log_entry entry; - - va_start(args, msg); - - entry.logger = logger; - - entry.buf = buffer; - - entry.buf_offset = 0; - entry.max_buf_bytes = DAL_LOGGER_BUFFER_MAX_SIZE * sizeof(char); - - entry.type = log_type; - - log_heading(&entry); - - size = dm_log_to_buffer( - buffer, LOG_MAX_LINE_SIZE - 1, msg, args); - - buffer[entry.buf_offset + size] = '\0'; - entry.buf_offset += size + 1; - - /* --Flush log_entry buffer-- */ - /* print to kernel console */ - log_to_debug_console(&entry); - /* log internally for dsat */ - log_to_internal_buffer(&entry); - - va_end(args); - } -} - -/* Same as dm_logger_write, except without open() and close(), which must - * be done separately. - */ -void dm_logger_append( - struct log_entry *entry, - const char *msg, - ...) -{ - va_list args; - - va_start(args, msg); - dm_logger_append_va(entry, msg, args); - va_end(args); -} - -void dm_logger_append_va( - struct log_entry *entry, - const char *msg, - va_list args) -{ - struct dal_logger *logger; - - if (!entry) { - BREAK_TO_DEBUGGER(); - return; - } - - logger = entry->logger; - - if (logger && logger->open_count > 0 && - dal_logger_should_log(logger, entry->type)) { - - uint32_t size; - char buffer[LOG_MAX_LINE_SIZE]; - - size = dm_log_to_buffer( - buffer, LOG_MAX_LINE_SIZE, msg, args); - - if (size < LOG_MAX_LINE_SIZE - 1) { - append_entry(entry, buffer, size); - } else { - append_entry(entry, "LOG_ERROR, line too long\n", 27); - } - } -} - -void dm_logger_open( - struct dal_logger *logger, - struct log_entry *entry, /* out */ - enum dc_log_type log_type) -{ - if (!entry) { - BREAK_TO_DEBUGGER(); - return; - } - - entry->type = log_type; - entry->logger = logger; - - entry->buf = kzalloc(DAL_LOGGER_BUFFER_MAX_SIZE, - GFP_KERNEL); - - entry->buf_offset = 0; - entry->max_buf_bytes = DAL_LOGGER_BUFFER_MAX_SIZE * sizeof(char); - - logger->open_count++; - - log_heading(entry); -} - -void dm_logger_close(struct log_entry *entry) -{ - struct dal_logger *logger = entry->logger; - - if (logger && logger->open_count > 0) { - logger->open_count--; - } else { - BREAK_TO_DEBUGGER(); - goto cleanup; - } - - /* --Flush log_entry buffer-- */ - /* print to kernel console */ - log_to_debug_console(entry); - /* log internally for dsat */ - log_to_internal_buffer(entry); - - /* TODO: Write end heading */ - -cleanup: - if (entry->buf) { - kfree(entry->buf); - entry->buf = NULL; - entry->buf_offset = 0; - entry->max_buf_bytes = 0; - } -} - diff --git a/drivers/gpu/drm/amd/display/dc/bios/bios_parser.c b/drivers/gpu/drm/amd/display/dc/bios/bios_parser.c index c7f0b27e457e..be8a2494355a 100644 --- a/drivers/gpu/drm/amd/display/dc/bios/bios_parser.c +++ b/drivers/gpu/drm/amd/display/dc/bios/bios_parser.c @@ -3762,6 +3762,200 @@ static struct integrated_info *bios_parser_create_integrated_info( return NULL; } +enum bp_result update_slot_layout_info( + struct dc_bios *dcb, + unsigned int i, + struct slot_layout_info *slot_layout_info, + unsigned int record_offset) +{ + unsigned int j; + struct bios_parser *bp; + ATOM_BRACKET_LAYOUT_RECORD *record; + ATOM_COMMON_RECORD_HEADER *record_header; + enum bp_result result = BP_RESULT_NORECORD; + + bp = BP_FROM_DCB(dcb); + record = NULL; + record_header = NULL; + + for (;;) { + + record_header = (ATOM_COMMON_RECORD_HEADER *) + GET_IMAGE(ATOM_COMMON_RECORD_HEADER, record_offset); + if (record_header == NULL) { + result = BP_RESULT_BADBIOSTABLE; + break; + } + + /* the end of the list */ + if (record_header->ucRecordType == 0xff || + record_header->ucRecordSize == 0) { + break; + } + + if (record_header->ucRecordType == + ATOM_BRACKET_LAYOUT_RECORD_TYPE && + sizeof(ATOM_BRACKET_LAYOUT_RECORD) + <= record_header->ucRecordSize) { + record = (ATOM_BRACKET_LAYOUT_RECORD *) + (record_header); + result = BP_RESULT_OK; + break; + } + + record_offset += record_header->ucRecordSize; + } + + /* return if the record not found */ + if (result != BP_RESULT_OK) + return result; + + /* get slot sizes */ + slot_layout_info->length = record->ucLength; + slot_layout_info->width = record->ucWidth; + + /* get info for each connector in the slot */ + slot_layout_info->num_of_connectors = record->ucConnNum; + for (j = 0; j < slot_layout_info->num_of_connectors; ++j) { + slot_layout_info->connectors[j].connector_type = + (enum connector_layout_type) + (record->asConnInfo[j].ucConnectorType); + switch (record->asConnInfo[j].ucConnectorType) { + case CONNECTOR_TYPE_DVI_D: + slot_layout_info->connectors[j].connector_type = + CONNECTOR_LAYOUT_TYPE_DVI_D; + slot_layout_info->connectors[j].length = + CONNECTOR_SIZE_DVI; + break; + + case CONNECTOR_TYPE_HDMI: + slot_layout_info->connectors[j].connector_type = + CONNECTOR_LAYOUT_TYPE_HDMI; + slot_layout_info->connectors[j].length = + CONNECTOR_SIZE_HDMI; + break; + + case CONNECTOR_TYPE_DISPLAY_PORT: + slot_layout_info->connectors[j].connector_type = + CONNECTOR_LAYOUT_TYPE_DP; + slot_layout_info->connectors[j].length = + CONNECTOR_SIZE_DP; + break; + + case CONNECTOR_TYPE_MINI_DISPLAY_PORT: + slot_layout_info->connectors[j].connector_type = + CONNECTOR_LAYOUT_TYPE_MINI_DP; + slot_layout_info->connectors[j].length = + CONNECTOR_SIZE_MINI_DP; + break; + + default: + slot_layout_info->connectors[j].connector_type = + CONNECTOR_LAYOUT_TYPE_UNKNOWN; + slot_layout_info->connectors[j].length = + CONNECTOR_SIZE_UNKNOWN; + } + + slot_layout_info->connectors[j].position = + record->asConnInfo[j].ucPosition; + slot_layout_info->connectors[j].connector_id = + object_id_from_bios_object_id( + record->asConnInfo[j].usConnectorObjectId); + } + return result; +} + + +enum bp_result get_bracket_layout_record( + struct dc_bios *dcb, + unsigned int bracket_layout_id, + struct slot_layout_info *slot_layout_info) +{ + unsigned int i; + unsigned int record_offset; + struct bios_parser *bp; + enum bp_result result; + ATOM_OBJECT *object; + ATOM_OBJECT_TABLE *object_table; + unsigned int genericTableOffset; + + bp = BP_FROM_DCB(dcb); + object = NULL; + if (slot_layout_info == NULL) { + DC_LOG_DETECTION_EDID_PARSER("Invalid slot_layout_info\n"); + return BP_RESULT_BADINPUT; + } + + + genericTableOffset = bp->object_info_tbl_offset + + bp->object_info_tbl.v1_3->usMiscObjectTableOffset; + object_table = (ATOM_OBJECT_TABLE *) + GET_IMAGE(ATOM_OBJECT_TABLE, genericTableOffset); + if (!object_table) + return BP_RESULT_FAILURE; + + result = BP_RESULT_NORECORD; + for (i = 0; i < object_table->ucNumberOfObjects; ++i) { + + if (bracket_layout_id == + object_table->asObjects[i].usObjectID) { + + object = &object_table->asObjects[i]; + record_offset = object->usRecordOffset + + bp->object_info_tbl_offset; + + result = update_slot_layout_info(dcb, i, + slot_layout_info, record_offset); + break; + } + } + return result; +} + +static enum bp_result bios_get_board_layout_info( + struct dc_bios *dcb, + struct board_layout_info *board_layout_info) +{ + unsigned int i; + struct bios_parser *bp; + enum bp_result record_result; + + const unsigned int slot_index_to_vbios_id[MAX_BOARD_SLOTS] = { + GENERICOBJECT_BRACKET_LAYOUT_ENUM_ID1, + GENERICOBJECT_BRACKET_LAYOUT_ENUM_ID2, + 0, 0 + }; + + bp = BP_FROM_DCB(dcb); + if (board_layout_info == NULL) { + DC_LOG_DETECTION_EDID_PARSER("Invalid board_layout_info\n"); + return BP_RESULT_BADINPUT; + } + + board_layout_info->num_of_slots = 0; + + for (i = 0; i < MAX_BOARD_SLOTS; ++i) { + record_result = get_bracket_layout_record(dcb, + slot_index_to_vbios_id[i], + &board_layout_info->slots[i]); + + if (record_result == BP_RESULT_NORECORD && i > 0) + break; /* no more slots present in bios */ + else if (record_result != BP_RESULT_OK) + return record_result; /* fail */ + + ++board_layout_info->num_of_slots; + } + + /* all data is valid */ + board_layout_info->is_number_of_slots_valid = 1; + board_layout_info->is_slots_size_valid = 1; + board_layout_info->is_connector_offsets_valid = 1; + board_layout_info->is_connector_lengths_valid = 1; + + return BP_RESULT_OK; +} + /******************************************************************************/ static const struct dc_vbios_funcs vbios_funcs = { @@ -3836,6 +4030,8 @@ static const struct dc_vbios_funcs vbios_funcs = { .post_init = bios_parser_post_init, /* patch vbios table for mxm module by reading i2c */ .bios_parser_destroy = bios_parser_destroy, + + .get_board_layout_info = bios_get_board_layout_info, }; static bool bios_parser_construct( diff --git a/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c b/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c index b8cef7af3c4a..eab007e1793c 100644 --- a/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c +++ b/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c @@ -43,6 +43,29 @@ #include "bios_parser_interface.h" #include "bios_parser_common.h" + +/* Temporarily add in defines until ObjectID.h patch is updated in a few days */ +#ifndef GENERIC_OBJECT_ID_BRACKET_LAYOUT +#define GENERIC_OBJECT_ID_BRACKET_LAYOUT 0x05 +#endif /* GENERIC_OBJECT_ID_BRACKET_LAYOUT */ + +#ifndef GENERICOBJECT_BRACKET_LAYOUT_ENUM_ID1 +#define GENERICOBJECT_BRACKET_LAYOUT_ENUM_ID1 \ + (GRAPH_OBJECT_TYPE_GENERIC << OBJECT_TYPE_SHIFT |\ + GRAPH_OBJECT_ENUM_ID1 << ENUM_ID_SHIFT |\ + GENERIC_OBJECT_ID_BRACKET_LAYOUT << OBJECT_ID_SHIFT) +#endif /* GENERICOBJECT_BRACKET_LAYOUT_ENUM_ID1 */ + +#ifndef GENERICOBJECT_BRACKET_LAYOUT_ENUM_ID2 +#define GENERICOBJECT_BRACKET_LAYOUT_ENUM_ID2 \ + (GRAPH_OBJECT_TYPE_GENERIC << OBJECT_TYPE_SHIFT |\ + GRAPH_OBJECT_ENUM_ID2 << ENUM_ID_SHIFT |\ + GENERIC_OBJECT_ID_BRACKET_LAYOUT << OBJECT_ID_SHIFT) +#endif /* GENERICOBJECT_BRACKET_LAYOUT_ENUM_ID2 */ + +#define DC_LOGGER \ + bp->base.ctx->logger + #define LAST_RECORD_TYPE 0xff #define SMU9_SYSPLL0_ID 0 @@ -86,7 +109,6 @@ static struct atom_encoder_caps_record *get_encoder_cap_record( #define DATA_TABLES(table) (bp->master_data_tbl->listOfdatatables.table) - static void destruct(struct bios_parser *bp) { kfree(bp->base.bios_local_image); @@ -656,7 +678,7 @@ static enum bp_result bios_parser_get_gpio_pin_info( return BP_RESULT_BADBIOSTABLE; if (sizeof(struct atom_common_table_header) + - sizeof(struct atom_gpio_pin_lut_v2_1) + sizeof(struct atom_gpio_pin_assignment) > le16_to_cpu(header->table_header.structuresize)) return BP_RESULT_BADBIOSTABLE; @@ -1854,6 +1876,198 @@ static struct integrated_info *bios_parser_create_integrated_info( return NULL; } +static enum bp_result update_slot_layout_info( + struct dc_bios *dcb, + unsigned int i, + struct slot_layout_info *slot_layout_info) +{ + unsigned int record_offset; + unsigned int j; + struct atom_display_object_path_v2 *object; + struct atom_bracket_layout_record *record; + struct atom_common_record_header *record_header; + enum bp_result result; + struct bios_parser *bp; + struct object_info_table *tbl; + struct display_object_info_table_v1_4 *v1_4; + + record = NULL; + record_header = NULL; + result = BP_RESULT_NORECORD; + + bp = BP_FROM_DCB(dcb); + tbl = &bp->object_info_tbl; + v1_4 = tbl->v1_4; + + object = &v1_4->display_path[i]; + record_offset = (unsigned int) + (object->disp_recordoffset) + + (unsigned int)(bp->object_info_tbl_offset); + + for (;;) { + + record_header = (struct atom_common_record_header *) + GET_IMAGE(struct atom_common_record_header, + record_offset); + if (record_header == NULL) { + result = BP_RESULT_BADBIOSTABLE; + break; + } + + /* the end of the list */ + if (record_header->record_type == 0xff || + record_header->record_size == 0) { + break; + } + + if (record_header->record_type == + ATOM_BRACKET_LAYOUT_RECORD_TYPE && + sizeof(struct atom_bracket_layout_record) + <= record_header->record_size) { + record = (struct atom_bracket_layout_record *) + (record_header); + result = BP_RESULT_OK; + break; + } + + record_offset += record_header->record_size; + } + + /* return if the record not found */ + if (result != BP_RESULT_OK) + return result; + + /* get slot sizes */ + slot_layout_info->length = record->bracketlen; + slot_layout_info->width = record->bracketwidth; + + /* get info for each connector in the slot */ + slot_layout_info->num_of_connectors = record->conn_num; + for (j = 0; j < slot_layout_info->num_of_connectors; ++j) { + slot_layout_info->connectors[j].connector_type = + (enum connector_layout_type) + (record->conn_info[j].connector_type); + switch (record->conn_info[j].connector_type) { + case CONNECTOR_TYPE_DVI_D: + slot_layout_info->connectors[j].connector_type = + CONNECTOR_LAYOUT_TYPE_DVI_D; + slot_layout_info->connectors[j].length = + CONNECTOR_SIZE_DVI; + break; + + case CONNECTOR_TYPE_HDMI: + slot_layout_info->connectors[j].connector_type = + CONNECTOR_LAYOUT_TYPE_HDMI; + slot_layout_info->connectors[j].length = + CONNECTOR_SIZE_HDMI; + break; + + case CONNECTOR_TYPE_DISPLAY_PORT: + slot_layout_info->connectors[j].connector_type = + CONNECTOR_LAYOUT_TYPE_DP; + slot_layout_info->connectors[j].length = + CONNECTOR_SIZE_DP; + break; + + case CONNECTOR_TYPE_MINI_DISPLAY_PORT: + slot_layout_info->connectors[j].connector_type = + CONNECTOR_LAYOUT_TYPE_MINI_DP; + slot_layout_info->connectors[j].length = + CONNECTOR_SIZE_MINI_DP; + break; + + default: + slot_layout_info->connectors[j].connector_type = + CONNECTOR_LAYOUT_TYPE_UNKNOWN; + slot_layout_info->connectors[j].length = + CONNECTOR_SIZE_UNKNOWN; + } + + slot_layout_info->connectors[j].position = + record->conn_info[j].position; + slot_layout_info->connectors[j].connector_id = + object_id_from_bios_object_id( + record->conn_info[j].connectorobjid); + } + return result; +} + + +static enum bp_result get_bracket_layout_record( + struct dc_bios *dcb, + unsigned int bracket_layout_id, + struct slot_layout_info *slot_layout_info) +{ + unsigned int i; + struct bios_parser *bp = BP_FROM_DCB(dcb); + enum bp_result result; + struct object_info_table *tbl; + struct display_object_info_table_v1_4 *v1_4; + + if (slot_layout_info == NULL) { + DC_LOG_DETECTION_EDID_PARSER("Invalid slot_layout_info\n"); + return BP_RESULT_BADINPUT; + } + tbl = &bp->object_info_tbl; + v1_4 = tbl->v1_4; + + result = BP_RESULT_NORECORD; + for (i = 0; i < v1_4->number_of_path; ++i) { + + if (bracket_layout_id == + v1_4->display_path[i].display_objid) { + result = update_slot_layout_info(dcb, i, + slot_layout_info); + break; + } + } + return result; +} + +static enum bp_result bios_get_board_layout_info( + struct dc_bios *dcb, + struct board_layout_info *board_layout_info) +{ + unsigned int i; + struct bios_parser *bp; + enum bp_result record_result; + + const unsigned int slot_index_to_vbios_id[MAX_BOARD_SLOTS] = { + GENERICOBJECT_BRACKET_LAYOUT_ENUM_ID1, + GENERICOBJECT_BRACKET_LAYOUT_ENUM_ID2, + 0, 0 + }; + + bp = BP_FROM_DCB(dcb); + if (board_layout_info == NULL) { + DC_LOG_DETECTION_EDID_PARSER("Invalid board_layout_info\n"); + return BP_RESULT_BADINPUT; + } + + board_layout_info->num_of_slots = 0; + + for (i = 0; i < MAX_BOARD_SLOTS; ++i) { + record_result = get_bracket_layout_record(dcb, + slot_index_to_vbios_id[i], + &board_layout_info->slots[i]); + + if (record_result == BP_RESULT_NORECORD && i > 0) + break; /* no more slots present in bios */ + else if (record_result != BP_RESULT_OK) + return record_result; /* fail */ + + ++board_layout_info->num_of_slots; + } + + /* all data is valid */ + board_layout_info->is_number_of_slots_valid = 1; + board_layout_info->is_slots_size_valid = 1; + board_layout_info->is_connector_offsets_valid = 1; + board_layout_info->is_connector_lengths_valid = 1; + + return BP_RESULT_OK; +} + static const struct dc_vbios_funcs vbios_funcs = { .get_connectors_number = bios_parser_get_connectors_number, @@ -1925,6 +2139,8 @@ static const struct dc_vbios_funcs vbios_funcs = { .bios_parser_destroy = firmware_parser_destroy, .get_smu_clock_info = bios_parser_get_smu_clock_info, + + .get_board_layout_info = bios_get_board_layout_info, }; static bool bios_parser_construct( diff --git a/drivers/gpu/drm/amd/display/dc/bios/command_table.c b/drivers/gpu/drm/amd/display/dc/bios/command_table.c index 651e1fd4622f..a558bfaa0c46 100644 --- a/drivers/gpu/drm/amd/display/dc/bios/command_table.c +++ b/drivers/gpu/drm/amd/display/dc/bios/command_table.c @@ -808,6 +808,24 @@ static enum bp_result transmitter_control_v1_5( * (=1: 8bpp, =1.25: 10bpp, =1.5:12bpp, =2: 16bpp) * LVDS mode: usPixelClock = pixel clock */ + if (cntl->signal == SIGNAL_TYPE_HDMI_TYPE_A) { + switch (cntl->color_depth) { + case COLOR_DEPTH_101010: + params.usSymClock = + cpu_to_le16((le16_to_cpu(params.usSymClock) * 30) / 24); + break; + case COLOR_DEPTH_121212: + params.usSymClock = + cpu_to_le16((le16_to_cpu(params.usSymClock) * 36) / 24); + break; + case COLOR_DEPTH_161616: + params.usSymClock = + cpu_to_le16((le16_to_cpu(params.usSymClock) * 48) / 24); + break; + default: + break; + } + } if (EXEC_BIOS_CMD_TABLE(UNIPHYTransmitterControl, params)) result = BP_RESULT_OK; diff --git a/drivers/gpu/drm/amd/display/dc/bios/command_table2.c b/drivers/gpu/drm/amd/display/dc/bios/command_table2.c index 752b08a42d3e..2b5dc499a35e 100644 --- a/drivers/gpu/drm/amd/display/dc/bios/command_table2.c +++ b/drivers/gpu/drm/amd/display/dc/bios/command_table2.c @@ -59,36 +59,7 @@ bios_cmd_table_para_revision(bp->base.ctx->driver_context, \ GET_INDEX_INTO_MASTER_TABLE(command, fname)) -static void init_dig_encoder_control(struct bios_parser *bp); -static void init_transmitter_control(struct bios_parser *bp); -static void init_set_pixel_clock(struct bios_parser *bp); -static void init_set_crtc_timing(struct bios_parser *bp); - -static void init_select_crtc_source(struct bios_parser *bp); -static void init_enable_crtc(struct bios_parser *bp); - -static void init_external_encoder_control(struct bios_parser *bp); -static void init_enable_disp_power_gating(struct bios_parser *bp); -static void init_set_dce_clock(struct bios_parser *bp); -static void init_get_smu_clock_info(struct bios_parser *bp); - -void dal_firmware_parser_init_cmd_tbl(struct bios_parser *bp) -{ - init_dig_encoder_control(bp); - init_transmitter_control(bp); - init_set_pixel_clock(bp); - - init_set_crtc_timing(bp); - - init_select_crtc_source(bp); - init_enable_crtc(bp); - - init_external_encoder_control(bp); - init_enable_disp_power_gating(bp); - init_set_dce_clock(bp); - init_get_smu_clock_info(bp); -} static uint32_t bios_cmd_table_para_revision(void *dev, uint32_t index) @@ -829,3 +800,20 @@ static unsigned int get_smu_clock_info_v3_1(struct bios_parser *bp, uint8_t id) return 0; } +void dal_firmware_parser_init_cmd_tbl(struct bios_parser *bp) +{ + init_dig_encoder_control(bp); + init_transmitter_control(bp); + init_set_pixel_clock(bp); + + init_set_crtc_timing(bp); + + init_select_crtc_source(bp); + init_enable_crtc(bp); + + init_external_encoder_control(bp); + init_enable_disp_power_gating(bp); + init_set_dce_clock(bp); + init_get_smu_clock_info(bp); + +} diff --git a/drivers/gpu/drm/amd/display/dc/bios/command_table_helper2.c b/drivers/gpu/drm/amd/display/dc/bios/command_table_helper2.c index bbbcef566c55..770ff89ba7e1 100644 --- a/drivers/gpu/drm/amd/display/dc/bios/command_table_helper2.c +++ b/drivers/gpu/drm/amd/display/dc/bios/command_table_helper2.c @@ -55,7 +55,7 @@ bool dal_bios_parser_init_cmd_tbl_helper2( case DCE_VERSION_11_22: *h = dal_cmd_tbl_helper_dce112_get_table2(); return true; -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 case DCN_VERSION_1_0: *h = dal_cmd_tbl_helper_dce112_get_table2(); return true; diff --git a/drivers/gpu/drm/amd/display/dc/calcs/Makefile b/drivers/gpu/drm/amd/display/dc/calcs/Makefile index 95f332ee3e7e..416500e51b8d 100644 --- a/drivers/gpu/drm/amd/display/dc/calcs/Makefile +++ b/drivers/gpu/drm/amd/display/dc/calcs/Makefile @@ -38,7 +38,7 @@ CFLAGS_dcn_calc_math.o := $(calcs_ccflags) -Wno-tautological-compare BW_CALCS = dce_calcs.o bw_fixed.o custom_float.o -ifdef CONFIG_DRM_AMD_DC_DCN1_0 +ifdef CONFIG_X86 BW_CALCS += dcn_calcs.o dcn_calc_math.o dcn_calc_auto.o endif diff --git a/drivers/gpu/drm/amd/display/dc/calcs/calcs_logger.h b/drivers/gpu/drm/amd/display/dc/calcs/calcs_logger.h index fc3f98fb09ea..62435bfc274d 100644 --- a/drivers/gpu/drm/amd/display/dc/calcs/calcs_logger.h +++ b/drivers/gpu/drm/amd/display/dc/calcs/calcs_logger.h @@ -25,10 +25,9 @@ #ifndef _CALCS_CALCS_LOGGER_H_ #define _CALCS_CALCS_LOGGER_H_ -#define DC_LOGGER \ - logger +#define DC_LOGGER ctx->logger -static void print_bw_calcs_dceip(struct dal_logger *logger, const struct bw_calcs_dceip *dceip) +static void print_bw_calcs_dceip(struct dc_context *ctx, const struct bw_calcs_dceip *dceip) { DC_LOG_BANDWIDTH_CALCS("#####################################################################"); @@ -122,7 +121,7 @@ static void print_bw_calcs_dceip(struct dal_logger *logger, const struct bw_calc } -static void print_bw_calcs_vbios(struct dal_logger *logger, const struct bw_calcs_vbios *vbios) +static void print_bw_calcs_vbios(struct dc_context *ctx, const struct bw_calcs_vbios *vbios) { DC_LOG_BANDWIDTH_CALCS("#####################################################################"); @@ -181,7 +180,7 @@ static void print_bw_calcs_vbios(struct dal_logger *logger, const struct bw_calc } -static void print_bw_calcs_data(struct dal_logger *logger, struct bw_calcs_data *data) +static void print_bw_calcs_data(struct dc_context *ctx, struct bw_calcs_data *data) { int i, j, k; diff --git a/drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c b/drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c index 2c4e8f0cb2dc..160d11a15eac 100644 --- a/drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c +++ b/drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c @@ -3010,9 +3010,9 @@ bool bw_calcs(struct dc_context *ctx, struct bw_fixed low_yclk = vbios->low_yclk; if (ctx->dc->debug.bandwidth_calcs_trace) { - print_bw_calcs_dceip(ctx->logger, dceip); - print_bw_calcs_vbios(ctx->logger, vbios); - print_bw_calcs_data(ctx->logger, data); + print_bw_calcs_dceip(ctx, dceip); + print_bw_calcs_vbios(ctx, vbios); + print_bw_calcs_data(ctx, data); } calculate_bandwidth(dceip, vbios, data); diff --git a/drivers/gpu/drm/amd/display/dc/calcs/dcn_calcs.c b/drivers/gpu/drm/amd/display/dc/calcs/dcn_calcs.c index 49a4ea45466d..bd039322f697 100644 --- a/drivers/gpu/drm/amd/display/dc/calcs/dcn_calcs.c +++ b/drivers/gpu/drm/amd/display/dc/calcs/dcn_calcs.c @@ -31,6 +31,8 @@ #include "resource.h" #include "dcn10/dcn10_resource.h" +#include "dcn10/dcn10_hubbub.h" + #include "dcn_calc_math.h" #define DC_LOGGER \ @@ -248,7 +250,24 @@ static void pipe_ctx_to_e2e_pipe_params ( else if (pipe->bottom_pipe != NULL && pipe->bottom_pipe->plane_state == pipe->plane_state) input->src.is_hsplit = true; - input->src.dcc = pipe->plane_state->dcc.enable; + if (pipe->plane_res.dpp->ctx->dc->debug.optimized_watermark) { + /* + * this method requires us to always re-calculate watermark when dcc change + * between flip. + */ + input->src.dcc = pipe->plane_state->dcc.enable ? 1 : 0; + } else { + /* + * allow us to disable dcc on the fly without re-calculating WM + * + * extra overhead for DCC is quite small. for 1080p WM without + * DCC is only 0.417us lower (urgent goes from 6.979us to 6.562us) + */ + unsigned int bpe; + + input->src.dcc = pipe->plane_res.dpp->ctx->dc->res_pool->hubbub->funcs-> + dcc_support_pixel_format(pipe->plane_state->format, &bpe) ? 1 : 0; + } input->src.dcc_rate = 1; input->src.meta_pitch = pipe->plane_state->dcc.grph.meta_pitch; input->src.source_scan = dm_horz; @@ -423,6 +442,10 @@ static void dcn_bw_calc_rq_dlg_ttu( int total_flip_bytes = 0; int i; + memset(dlg_regs, 0, sizeof(*dlg_regs)); + memset(ttu_regs, 0, sizeof(*ttu_regs)); + memset(rq_regs, 0, sizeof(*rq_regs)); + for (i = 0; i < number_of_planes; i++) { total_active_bw += v->read_bandwidth[i]; total_prefetch_bw += v->prefetch_bandwidth[i]; @@ -501,6 +524,7 @@ static void split_stream_across_pipes( resource_build_scaling_params(secondary_pipe); } +#if 0 static void calc_wm_sets_and_perf_params( struct dc_state *context, struct dcn_bw_internal_vars *v) @@ -582,6 +606,7 @@ static void calc_wm_sets_and_perf_params( if (v->voltage_level >= 3) context->bw.dcn.watermarks.d = context->bw.dcn.watermarks.a; } +#endif static bool dcn_bw_apply_registry_override(struct dc *dc) { @@ -651,7 +676,7 @@ static void hack_force_pipe_split(struct dcn_bw_internal_vars *v, } static void hack_bounding_box(struct dcn_bw_internal_vars *v, - struct dc_debug *dbg, + struct dc_debug_options *dbg, struct dc_state *context) { if (dbg->pipe_split_policy == MPC_SPLIT_AVOID) @@ -883,7 +908,26 @@ bool dcn_validate_bandwidth( ASSERT(pipe->plane_res.scl_data.ratios.vert.value != dc_fixpt_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; + + if (dc->debug.optimized_watermark) { + /* + * this method requires us to always re-calculate watermark when dcc change + * between flip. + */ + v->dcc_enable[input_idx] = pipe->plane_state->dcc.enable ? dcn_bw_yes : dcn_bw_no; + } else { + /* + * allow us to disable dcc on the fly without re-calculating WM + * + * extra overhead for DCC is quite small. for 1080p WM without + * DCC is only 0.417us lower (urgent goes from 6.979us to 6.562us) + */ + unsigned int bpe; + + v->dcc_enable[input_idx] = dc->res_pool->hubbub->funcs->dcc_support_pixel_format( + pipe->plane_state->format, &bpe) ? dcn_bw_yes : dcn_bw_no; + } + v->source_pixel_format[input_idx] = tl_pixel_format_to_bw_defs( pipe->plane_state->format); v->source_surface_mode[input_idx] = tl_sw_mode_to_bw_defs( @@ -976,43 +1020,60 @@ bool dcn_validate_bandwidth( bw_consumed = v->fabric_and_dram_bandwidth; display_pipe_configuration(v); - calc_wm_sets_and_perf_params(context, v); - context->bw.dcn.calc_clk.fclk_khz = (int)(bw_consumed * 1000000 / + /*calc_wm_sets_and_perf_params(context, v);*/ + /* Only 1 set is used by dcn since no noticeable + * performance improvement was measured and due to hw bug DEGVIDCN10-254 + */ + dispclkdppclkdcfclk_deep_sleep_prefetch_parameters_watermarks_and_performance_calculation(v); + + context->bw.dcn.watermarks.a.cstate_pstate.cstate_exit_ns = + v->stutter_exit_watermark * 1000; + context->bw.dcn.watermarks.a.cstate_pstate.cstate_enter_plus_exit_ns = + v->stutter_enter_plus_exit_watermark * 1000; + context->bw.dcn.watermarks.a.cstate_pstate.pstate_change_ns = + v->dram_clock_change_watermark * 1000; + context->bw.dcn.watermarks.a.pte_meta_urgent_ns = v->ptemeta_urgent_watermark * 1000; + context->bw.dcn.watermarks.a.urgent_ns = v->urgent_watermark * 1000; + context->bw.dcn.watermarks.b = context->bw.dcn.watermarks.a; + context->bw.dcn.watermarks.c = context->bw.dcn.watermarks.a; + context->bw.dcn.watermarks.d = context->bw.dcn.watermarks.a; + + context->bw.dcn.clk.fclk_khz = (int)(bw_consumed * 1000000 / (ddr4_dram_factor_single_Channel * v->number_of_channels)); if (bw_consumed == v->fabric_and_dram_bandwidth_vmin0p65) { - context->bw.dcn.calc_clk.fclk_khz = (int)(bw_consumed * 1000000 / 32); + context->bw.dcn.clk.fclk_khz = (int)(bw_consumed * 1000000 / 32); } - context->bw.dcn.calc_clk.dcfclk_deep_sleep_khz = (int)(v->dcf_clk_deep_sleep * 1000); - context->bw.dcn.calc_clk.dcfclk_khz = (int)(v->dcfclk * 1000); + context->bw.dcn.clk.dcfclk_deep_sleep_khz = (int)(v->dcf_clk_deep_sleep * 1000); + context->bw.dcn.clk.dcfclk_khz = (int)(v->dcfclk * 1000); - context->bw.dcn.calc_clk.dispclk_khz = (int)(v->dispclk * 1000); + context->bw.dcn.clk.dispclk_khz = (int)(v->dispclk * 1000); if (dc->debug.max_disp_clk == true) - context->bw.dcn.calc_clk.dispclk_khz = (int)(dc->dcn_soc->max_dispclk_vmax0p9 * 1000); + context->bw.dcn.clk.dispclk_khz = (int)(dc->dcn_soc->max_dispclk_vmax0p9 * 1000); - if (context->bw.dcn.calc_clk.dispclk_khz < + if (context->bw.dcn.clk.dispclk_khz < dc->debug.min_disp_clk_khz) { - context->bw.dcn.calc_clk.dispclk_khz = + context->bw.dcn.clk.dispclk_khz = dc->debug.min_disp_clk_khz; } - context->bw.dcn.calc_clk.dppclk_khz = context->bw.dcn.calc_clk.dispclk_khz / v->dispclk_dppclk_ratio; - + context->bw.dcn.clk.dppclk_khz = context->bw.dcn.clk.dispclk_khz / v->dispclk_dppclk_ratio; + context->bw.dcn.clk.phyclk_khz = v->phyclk_per_state[v->voltage_level]; switch (v->voltage_level) { case 0: - context->bw.dcn.calc_clk.max_supported_dppclk_khz = + context->bw.dcn.clk.max_supported_dppclk_khz = (int)(dc->dcn_soc->max_dppclk_vmin0p65 * 1000); break; case 1: - context->bw.dcn.calc_clk.max_supported_dppclk_khz = + context->bw.dcn.clk.max_supported_dppclk_khz = (int)(dc->dcn_soc->max_dppclk_vmid0p72 * 1000); break; case 2: - context->bw.dcn.calc_clk.max_supported_dppclk_khz = + context->bw.dcn.clk.max_supported_dppclk_khz = (int)(dc->dcn_soc->max_dppclk_vnom0p8 * 1000); break; default: - context->bw.dcn.calc_clk.max_supported_dppclk_khz = + context->bw.dcn.clk.max_supported_dppclk_khz = (int)(dc->dcn_soc->max_dppclk_vmax0p9 * 1000); break; } @@ -1225,27 +1286,27 @@ static unsigned int dcn_find_normalized_clock_vdd_Level( unsigned int dcn_find_dcfclk_suits_all( const struct dc *dc, - struct clocks_value *clocks) + struct dc_clocks *clocks) { unsigned vdd_level, vdd_level_temp; unsigned dcf_clk; /*find a common supported voltage level*/ vdd_level = dcn_find_normalized_clock_vdd_Level( - dc, DM_PP_CLOCK_TYPE_DISPLAY_CLK, clocks->dispclk_in_khz); + dc, DM_PP_CLOCK_TYPE_DISPLAY_CLK, clocks->dispclk_khz); vdd_level_temp = dcn_find_normalized_clock_vdd_Level( - dc, DM_PP_CLOCK_TYPE_DISPLAYPHYCLK, clocks->phyclk_in_khz); + dc, DM_PP_CLOCK_TYPE_DISPLAYPHYCLK, clocks->phyclk_khz); vdd_level = dcn_bw_max(vdd_level, vdd_level_temp); vdd_level_temp = dcn_find_normalized_clock_vdd_Level( - dc, DM_PP_CLOCK_TYPE_DPPCLK, clocks->dppclk_in_khz); + dc, DM_PP_CLOCK_TYPE_DPPCLK, clocks->dppclk_khz); vdd_level = dcn_bw_max(vdd_level, vdd_level_temp); vdd_level_temp = dcn_find_normalized_clock_vdd_Level( - dc, DM_PP_CLOCK_TYPE_MEMORY_CLK, clocks->dcfclock_in_khz); + dc, DM_PP_CLOCK_TYPE_MEMORY_CLK, clocks->fclk_khz); vdd_level = dcn_bw_max(vdd_level, vdd_level_temp); vdd_level_temp = dcn_find_normalized_clock_vdd_Level( - dc, DM_PP_CLOCK_TYPE_DCFCLK, clocks->dcfclock_in_khz); + dc, DM_PP_CLOCK_TYPE_DCFCLK, clocks->dcfclk_khz); /*find that level conresponding dcfclk*/ vdd_level = dcn_bw_max(vdd_level, vdd_level_temp); @@ -1331,21 +1392,14 @@ void dcn_bw_notify_pplib_of_wm_ranges(struct dc *dc) { struct pp_smu_funcs_rv *pp = dc->res_pool->pp_smu; struct pp_smu_wm_range_sets ranges = {0}; - int max_fclk_khz, nom_fclk_khz, mid_fclk_khz, min_fclk_khz; - int max_dcfclk_khz, min_dcfclk_khz; - int socclk_khz; + int min_fclk_khz, min_dcfclk_khz, socclk_khz; const int overdrive = 5000000; /* 5 GHz to cover Overdrive */ - unsigned factor = (ddr4_dram_factor_single_Channel * dc->dcn_soc->number_of_channels); if (!pp->set_wm_ranges) return; kernel_fpu_begin(); - max_fclk_khz = dc->dcn_soc->fabric_and_dram_bandwidth_vmax0p9 * 1000000 / factor; - nom_fclk_khz = dc->dcn_soc->fabric_and_dram_bandwidth_vnom0p8 * 1000000 / factor; - mid_fclk_khz = dc->dcn_soc->fabric_and_dram_bandwidth_vmid0p72 * 1000000 / factor; min_fclk_khz = dc->dcn_soc->fabric_and_dram_bandwidth_vmin0p65 * 1000000 / 32; - max_dcfclk_khz = dc->dcn_soc->dcfclkv_max0p9 * 1000; min_dcfclk_khz = dc->dcn_soc->dcfclkv_min0p65 * 1000; socclk_khz = dc->dcn_soc->socclk * 1000; kernel_fpu_end(); @@ -1353,105 +1407,46 @@ void dcn_bw_notify_pplib_of_wm_ranges(struct dc *dc) /* Now notify PPLib/SMU about which Watermarks sets they should select * depending on DPM state they are in. And update BW MGR GFX Engine and * Memory clock member variables for Watermarks calculations for each - * Watermark Set + * Watermark Set. Only one watermark set for dcn1 due to hw bug DEGVIDCN10-254. */ /* SOCCLK does not affect anytihng but writeback for DCN so for now we dont * care what the value is, hence min to overdrive level */ - ranges.num_reader_wm_sets = WM_COUNT; - ranges.num_writer_wm_sets = WM_COUNT; + ranges.num_reader_wm_sets = WM_SET_COUNT; + ranges.num_writer_wm_sets = WM_SET_COUNT; ranges.reader_wm_sets[0].wm_inst = WM_A; ranges.reader_wm_sets[0].min_drain_clk_khz = min_dcfclk_khz; - ranges.reader_wm_sets[0].max_drain_clk_khz = max_dcfclk_khz; + ranges.reader_wm_sets[0].max_drain_clk_khz = overdrive; ranges.reader_wm_sets[0].min_fill_clk_khz = min_fclk_khz; - ranges.reader_wm_sets[0].max_fill_clk_khz = min_fclk_khz; + ranges.reader_wm_sets[0].max_fill_clk_khz = overdrive; ranges.writer_wm_sets[0].wm_inst = WM_A; ranges.writer_wm_sets[0].min_fill_clk_khz = socclk_khz; ranges.writer_wm_sets[0].max_fill_clk_khz = overdrive; ranges.writer_wm_sets[0].min_drain_clk_khz = min_fclk_khz; - ranges.writer_wm_sets[0].max_drain_clk_khz = min_fclk_khz; - - ranges.reader_wm_sets[1].wm_inst = WM_B; - ranges.reader_wm_sets[1].min_drain_clk_khz = min_fclk_khz; - ranges.reader_wm_sets[1].max_drain_clk_khz = max_dcfclk_khz; - ranges.reader_wm_sets[1].min_fill_clk_khz = mid_fclk_khz; - ranges.reader_wm_sets[1].max_fill_clk_khz = mid_fclk_khz; - ranges.writer_wm_sets[1].wm_inst = WM_B; - ranges.writer_wm_sets[1].min_fill_clk_khz = socclk_khz; - ranges.writer_wm_sets[1].max_fill_clk_khz = overdrive; - ranges.writer_wm_sets[1].min_drain_clk_khz = mid_fclk_khz; - ranges.writer_wm_sets[1].max_drain_clk_khz = mid_fclk_khz; - - - ranges.reader_wm_sets[2].wm_inst = WM_C; - ranges.reader_wm_sets[2].min_drain_clk_khz = min_fclk_khz; - ranges.reader_wm_sets[2].max_drain_clk_khz = max_dcfclk_khz; - ranges.reader_wm_sets[2].min_fill_clk_khz = nom_fclk_khz; - ranges.reader_wm_sets[2].max_fill_clk_khz = nom_fclk_khz; - ranges.writer_wm_sets[2].wm_inst = WM_C; - ranges.writer_wm_sets[2].min_fill_clk_khz = socclk_khz; - ranges.writer_wm_sets[2].max_fill_clk_khz = overdrive; - ranges.writer_wm_sets[2].min_drain_clk_khz = nom_fclk_khz; - ranges.writer_wm_sets[2].max_drain_clk_khz = nom_fclk_khz; - - ranges.reader_wm_sets[3].wm_inst = WM_D; - ranges.reader_wm_sets[3].min_drain_clk_khz = min_fclk_khz; - ranges.reader_wm_sets[3].max_drain_clk_khz = max_dcfclk_khz; - ranges.reader_wm_sets[3].min_fill_clk_khz = max_fclk_khz; - ranges.reader_wm_sets[3].max_fill_clk_khz = max_fclk_khz; - ranges.writer_wm_sets[3].wm_inst = WM_D; - ranges.writer_wm_sets[3].min_fill_clk_khz = socclk_khz; - ranges.writer_wm_sets[3].max_fill_clk_khz = overdrive; - ranges.writer_wm_sets[3].min_drain_clk_khz = max_fclk_khz; - ranges.writer_wm_sets[3].max_drain_clk_khz = max_fclk_khz; + ranges.writer_wm_sets[0].max_drain_clk_khz = overdrive; if (dc->debug.pplib_wm_report_mode == WM_REPORT_OVERRIDE) { ranges.reader_wm_sets[0].wm_inst = WM_A; ranges.reader_wm_sets[0].min_drain_clk_khz = 300000; - ranges.reader_wm_sets[0].max_drain_clk_khz = 654000; + ranges.reader_wm_sets[0].max_drain_clk_khz = 5000000; ranges.reader_wm_sets[0].min_fill_clk_khz = 800000; - ranges.reader_wm_sets[0].max_fill_clk_khz = 800000; + ranges.reader_wm_sets[0].max_fill_clk_khz = 5000000; ranges.writer_wm_sets[0].wm_inst = WM_A; ranges.writer_wm_sets[0].min_fill_clk_khz = 200000; - ranges.writer_wm_sets[0].max_fill_clk_khz = 757000; + ranges.writer_wm_sets[0].max_fill_clk_khz = 5000000; ranges.writer_wm_sets[0].min_drain_clk_khz = 800000; - ranges.writer_wm_sets[0].max_drain_clk_khz = 800000; - - ranges.reader_wm_sets[1].wm_inst = WM_B; - ranges.reader_wm_sets[1].min_drain_clk_khz = 300000; - ranges.reader_wm_sets[1].max_drain_clk_khz = 654000; - ranges.reader_wm_sets[1].min_fill_clk_khz = 933000; - ranges.reader_wm_sets[1].max_fill_clk_khz = 933000; - ranges.writer_wm_sets[1].wm_inst = WM_B; - ranges.writer_wm_sets[1].min_fill_clk_khz = 200000; - ranges.writer_wm_sets[1].max_fill_clk_khz = 757000; - ranges.writer_wm_sets[1].min_drain_clk_khz = 933000; - ranges.writer_wm_sets[1].max_drain_clk_khz = 933000; - - - ranges.reader_wm_sets[2].wm_inst = WM_C; - ranges.reader_wm_sets[2].min_drain_clk_khz = 300000; - ranges.reader_wm_sets[2].max_drain_clk_khz = 654000; - ranges.reader_wm_sets[2].min_fill_clk_khz = 1067000; - ranges.reader_wm_sets[2].max_fill_clk_khz = 1067000; - ranges.writer_wm_sets[2].wm_inst = WM_C; - ranges.writer_wm_sets[2].min_fill_clk_khz = 200000; - ranges.writer_wm_sets[2].max_fill_clk_khz = 757000; - ranges.writer_wm_sets[2].min_drain_clk_khz = 1067000; - ranges.writer_wm_sets[2].max_drain_clk_khz = 1067000; - - ranges.reader_wm_sets[3].wm_inst = WM_D; - ranges.reader_wm_sets[3].min_drain_clk_khz = 300000; - ranges.reader_wm_sets[3].max_drain_clk_khz = 654000; - ranges.reader_wm_sets[3].min_fill_clk_khz = 1200000; - ranges.reader_wm_sets[3].max_fill_clk_khz = 1200000; - ranges.writer_wm_sets[3].wm_inst = WM_D; - ranges.writer_wm_sets[3].min_fill_clk_khz = 200000; - ranges.writer_wm_sets[3].max_fill_clk_khz = 757000; - ranges.writer_wm_sets[3].min_drain_clk_khz = 1200000; - ranges.writer_wm_sets[3].max_drain_clk_khz = 1200000; + ranges.writer_wm_sets[0].max_drain_clk_khz = 5000000; } + ranges.reader_wm_sets[1] = ranges.writer_wm_sets[0]; + ranges.reader_wm_sets[1].wm_inst = WM_B; + + ranges.reader_wm_sets[2] = ranges.writer_wm_sets[0]; + ranges.reader_wm_sets[2].wm_inst = WM_C; + + ranges.reader_wm_sets[3] = ranges.writer_wm_sets[0]; + ranges.reader_wm_sets[3].wm_inst = WM_D; + /* Notify PP Lib/SMU which Watermarks to use for which clock ranges */ pp->set_wm_ranges(&pp->pp_smu, &ranges); } diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c index 644b2187507b..733ac224e7fd 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc.c @@ -169,6 +169,22 @@ failed_alloc: return false; } +/** + ***************************************************************************** + * Function: dc_stream_adjust_vmin_vmax + * + * @brief + * Looks up the pipe context of dc_stream_state and updates the + * vertical_total_min and vertical_total_max of the DRR, Dynamic Refresh + * Rate, which is a power-saving feature that targets reducing panel + * refresh rate while the screen is static + * + * @param [in] dc: dc reference + * @param [in] stream: Initial dc stream state + * @param [in] adjust: Updated parameters for vertical_total_min and + * vertical_total_max + ***************************************************************************** + */ bool dc_stream_adjust_vmin_vmax(struct dc *dc, struct dc_stream_state **streams, int num_streams, int vmin, int vmax) @@ -368,6 +384,71 @@ void dc_stream_set_static_screen_events(struct dc *dc, dc->hwss.set_static_screen_control(pipes_affected, num_pipes_affected, events); } +void dc_link_set_drive_settings(struct dc *dc, + struct link_training_settings *lt_settings, + const struct dc_link *link) +{ + + int i; + + for (i = 0; i < dc->link_count; i++) { + if (dc->links[i] == link) + break; + } + + if (i >= dc->link_count) + ASSERT_CRITICAL(false); + + dc_link_dp_set_drive_settings(dc->links[i], lt_settings); +} + +void dc_link_perform_link_training(struct dc *dc, + struct dc_link_settings *link_setting, + bool skip_video_pattern) +{ + int i; + + for (i = 0; i < dc->link_count; i++) + dc_link_dp_perform_link_training( + dc->links[i], + link_setting, + skip_video_pattern); +} + +void dc_link_set_preferred_link_settings(struct dc *dc, + struct dc_link_settings *link_setting, + struct dc_link *link) +{ + link->preferred_link_setting = *link_setting; + dp_retrain_link_dp_test(link, link_setting, false); +} + +void dc_link_enable_hpd(const struct dc_link *link) +{ + dc_link_dp_enable_hpd(link); +} + +void dc_link_disable_hpd(const struct dc_link *link) +{ + dc_link_dp_disable_hpd(link); +} + + +void dc_link_set_test_pattern(struct dc_link *link, + enum dp_test_pattern test_pattern, + const struct link_training_settings *p_link_settings, + const unsigned char *p_custom_pattern, + unsigned int cust_pattern_size) +{ + if (link != NULL) + dc_link_dp_set_test_pattern( + link, + test_pattern, + p_link_settings, + p_custom_pattern, + cust_pattern_size); +} + static void destruct(struct dc *dc) { dc_release_state(dc->current_state); @@ -386,9 +467,6 @@ static void destruct(struct dc *dc) if (dc->ctx->created_bios) dal_bios_parser_destroy(&dc->ctx->dc_bios); - if (dc->ctx->logger) - dal_logger_destroy(&dc->ctx->logger); - kfree(dc->ctx); dc->ctx = NULL; @@ -398,7 +476,7 @@ static void destruct(struct dc *dc) kfree(dc->bw_dceip); dc->bw_dceip = NULL; -#ifdef CONFIG_DRM_AMD_DC_DCN1_0 +#ifdef CONFIG_X86 kfree(dc->dcn_soc); dc->dcn_soc = NULL; @@ -411,11 +489,10 @@ static void destruct(struct dc *dc) static bool construct(struct dc *dc, const struct dc_init_data *init_params) { - struct dal_logger *logger; struct dc_context *dc_ctx; struct bw_calcs_dceip *dc_dceip; struct bw_calcs_vbios *dc_vbios; -#ifdef CONFIG_DRM_AMD_DC_DCN1_0 +#ifdef CONFIG_X86 struct dcn_soc_bounding_box *dcn_soc; struct dcn_ip_params *dcn_ip; #endif @@ -437,7 +514,7 @@ static bool construct(struct dc *dc, } dc->bw_vbios = dc_vbios; -#ifdef CONFIG_DRM_AMD_DC_DCN1_0 +#ifdef CONFIG_X86 dcn_soc = kzalloc(sizeof(*dcn_soc), GFP_KERNEL); if (!dcn_soc) { dm_error("%s: failed to create dcn_soc\n", __func__); @@ -465,6 +542,7 @@ static bool construct(struct dc *dc, dc_ctx->driver_context = init_params->driver; dc_ctx->dc = dc; dc_ctx->asic_id = init_params->asic_id; + dc_ctx->dc_sink_id_count = 0; dc->ctx = dc_ctx; dc->current_state = dc_create_state(); @@ -475,14 +553,7 @@ static bool construct(struct dc *dc, } /* Create logger */ - logger = dal_logger_create(dc_ctx, init_params->log_mask); - if (!logger) { - /* can *not* call logger. call base driver 'print error' */ - dm_error("%s: failed to create Logger!\n", __func__); - goto fail; - } - dc_ctx->logger = logger; dc_ctx->dce_environment = init_params->dce_environment; dc_version = resource_parse_asic_id(init_params->asic_id); @@ -901,9 +972,7 @@ bool dc_commit_state(struct dc *dc, struct dc_state *context) for (i = 0; i < context->stream_count; i++) { struct dc_stream_state *stream = context->streams[i]; - dc_stream_log(stream, - dc->ctx->logger, - LOG_DC); + dc_stream_log(dc, stream); } result = dc_commit_state_no_check(dc, context); @@ -927,12 +996,7 @@ bool dc_post_update_surfaces_to_stream(struct dc *dc) 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); -#else dc->hwss.set_bandwidth(dc, context, true); -#endif return true; } @@ -1548,7 +1612,7 @@ struct dc_sink *dc_link_add_remote_sink( struct dc_sink *dc_sink; enum dc_edid_status edid_status; - if (len > MAX_EDID_BUFFER_SIZE) { + if (len > DC_MAX_EDID_BUFFER_SIZE) { dm_error("Max EDID buffer size breached!\n"); return NULL; } diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_debug.c b/drivers/gpu/drm/amd/display/dc/core/dc_debug.c index 267c76766dea..caece7c13bc6 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_debug.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_debug.c @@ -348,23 +348,23 @@ void context_clock_trace( struct dc *dc, struct dc_state *context) { -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 DC_LOGGER_INIT(dc->ctx->logger); CLOCK_TRACE("Current: dispclk_khz:%d max_dppclk_khz:%d dcfclk_khz:%d\n" "dcfclk_deep_sleep_khz:%d fclk_khz:%d socclk_khz:%d\n", - context->bw.dcn.calc_clk.dispclk_khz, - context->bw.dcn.calc_clk.dppclk_khz, - context->bw.dcn.calc_clk.dcfclk_khz, - context->bw.dcn.calc_clk.dcfclk_deep_sleep_khz, - context->bw.dcn.calc_clk.fclk_khz, - context->bw.dcn.calc_clk.socclk_khz); + context->bw.dcn.clk.dispclk_khz, + context->bw.dcn.clk.dppclk_khz, + context->bw.dcn.clk.dcfclk_khz, + context->bw.dcn.clk.dcfclk_deep_sleep_khz, + context->bw.dcn.clk.fclk_khz, + context->bw.dcn.clk.socclk_khz); CLOCK_TRACE("Calculated: dispclk_khz:%d max_dppclk_khz:%d dcfclk_khz:%d\n" "dcfclk_deep_sleep_khz:%d fclk_khz:%d socclk_khz:%d\n", - context->bw.dcn.calc_clk.dispclk_khz, - context->bw.dcn.calc_clk.dppclk_khz, - context->bw.dcn.calc_clk.dcfclk_khz, - context->bw.dcn.calc_clk.dcfclk_deep_sleep_khz, - context->bw.dcn.calc_clk.fclk_khz, - context->bw.dcn.calc_clk.socclk_khz); + context->bw.dcn.clk.dispclk_khz, + context->bw.dcn.clk.dppclk_khz, + context->bw.dcn.clk.dcfclk_khz, + context->bw.dcn.clk.dcfclk_deep_sleep_khz, + context->bw.dcn.clk.fclk_khz, + context->bw.dcn.clk.socclk_khz); #endif } diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link.c b/drivers/gpu/drm/amd/display/dc/core/dc_link.c index 2fa521812d23..a38e7ad36a7e 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_link.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_link.c @@ -33,6 +33,7 @@ #include "dc_link_dp.h" #include "dc_link_ddc.h" #include "link_hwss.h" +#include "opp.h" #include "link_encoder.h" #include "hw_sequencer.h" @@ -59,7 +60,14 @@ enum { LINK_RATE_REF_FREQ_IN_MHZ = 27, - PEAK_FACTOR_X1000 = 1006 + PEAK_FACTOR_X1000 = 1006, + /* + * Some receivers fail to train on first try and are good + * on subsequent tries. 2 retries should be plenty. If we + * don't have a successful training then we don't expect to + * ever get one. + */ + LINK_TRAINING_MAX_VERIFY_RETRY = 2 }; /******************************************************************************* @@ -312,7 +320,7 @@ static enum signal_type get_basic_signal_type( * @brief * Check whether there is a dongle on DP connector */ -static bool is_dp_sink_present(struct dc_link *link) +bool dc_link_is_dp_sink_present(struct dc_link *link) { enum gpio_result gpio_result; uint32_t clock_pin = 0; @@ -405,7 +413,7 @@ static enum signal_type link_detect_sink( * we assume signal is DVI; it could be corrected * to HDMI after dongle detection */ - if (!is_dp_sink_present(link)) + if (!dm_helpers_is_dp_sink_present(link)) result = SIGNAL_TYPE_DVI_SINGLE_LINK; } } @@ -497,6 +505,10 @@ static bool detect_dp( sink_caps->signal = SIGNAL_TYPE_DISPLAY_PORT_MST; link->type = dc_connection_mst_branch; + dal_ddc_service_set_transaction_type( + link->ddc, + sink_caps->transaction_type); + /* * This call will initiate MST topology discovery. Which * will detect MST ports and add new DRM connector DRM @@ -524,6 +536,10 @@ static bool detect_dp( if (reason == DETECT_REASON_BOOT) boot = true; + dm_helpers_dp_update_branch_info( + link->ctx, + link); + if (!dm_helpers_dp_mst_start_top_mgr( link->ctx, link, boot)) { @@ -728,6 +744,18 @@ bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason) break; case EDID_NO_RESPONSE: DC_LOG_ERROR("No EDID read.\n"); + + /* + * Abort detection for non-DP connectors if we have + * no EDID + * + * DP needs to report as connected if HDP is high + * even if we have no EDID in order to go to + * fail-safe mode + */ + if (dc_is_hdmi_signal(link->connector_signal) || + dc_is_dvi_signal(link->connector_signal)) + return false; default: break; } @@ -736,30 +764,41 @@ bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason) if ((prev_sink != NULL) && ((edid_status == EDID_THE_SAME) || (edid_status == EDID_OK))) same_edid = is_same_edid(&prev_sink->dc_edid, &sink->dc_edid); - // If both edid and dpcd are the same, then discard new sink and revert back to original sink - if ((same_edid) && (same_dpcd)) { - link_disconnect_remap(prev_sink, link); - sink = prev_sink; - prev_sink = NULL; - } else { - if (link->connector_signal == SIGNAL_TYPE_DISPLAY_PORT && - sink_caps.transaction_type == - DDC_TRANSACTION_TYPE_I2C_OVER_AUX) { - /* - * TODO debug why Dell 2413 doesn't like - * two link trainings - */ + if (link->connector_signal == SIGNAL_TYPE_DISPLAY_PORT && + sink_caps.transaction_type == DDC_TRANSACTION_TYPE_I2C_OVER_AUX && + reason != DETECT_REASON_HPDRX) { + /* + * TODO debug why Dell 2413 doesn't like + * two link trainings + */ + + /* deal with non-mst cases */ + for (i = 0; i < LINK_TRAINING_MAX_VERIFY_RETRY; i++) { + int fail_count = 0; + + dp_verify_link_cap(link, + &link->reported_link_cap, + &fail_count); - /* deal with non-mst cases */ - dp_hbr_verify_link_cap(link, &link->reported_link_cap); + if (fail_count == 0) + break; } - /* HDMI-DVI Dongle */ - if (sink->sink_signal == SIGNAL_TYPE_HDMI_TYPE_A && - !sink->edid_caps.edid_hdmi) - sink->sink_signal = SIGNAL_TYPE_DVI_SINGLE_LINK; + } else { + // If edid is the same, then discard new sink and revert back to original sink + if (same_edid) { + link_disconnect_remap(prev_sink, link); + sink = prev_sink; + prev_sink = NULL; + + } } + /* HDMI-DVI Dongle */ + if (sink->sink_signal == SIGNAL_TYPE_HDMI_TYPE_A && + !sink->edid_caps.edid_hdmi) + sink->sink_signal = SIGNAL_TYPE_DVI_SINGLE_LINK; + /* Connectivity log: detection */ for (i = 0; i < sink->dc_edid.length / EDID_BLOCK_SIZE; i++) { CONN_DATA_DETECT(link, @@ -1000,6 +1039,9 @@ static bool construct( link->link_id = bios->funcs->get_connector_id(bios, init_params->connector_index); + if (dc_ctx->dc_bios->integrated_info) + link->dp_ss_off = !!dc_ctx->dc_bios->integrated_info->dp_ss_control; + if (link->link_id.type != OBJECT_TYPE_CONNECTOR) { dm_error("%s: Invalid Connector ObjectID from Adapter Service for connector index:%d! type %d expected %d\n", __func__, init_params->connector_index, @@ -1284,29 +1326,15 @@ static enum dc_status enable_link_dp( max_link_rate = LINK_RATE_HIGH3; if (link_settings.link_rate == max_link_rate) { - if (state->dis_clk->funcs->set_min_clocks_state) { - if (state->dis_clk->cur_min_clks_state < DM_PP_CLOCKS_STATE_NOMINAL) - state->dis_clk->funcs->set_min_clocks_state( - state->dis_clk, DM_PP_CLOCKS_STATE_NOMINAL); - } else { - uint32_t dp_phyclk_in_khz; - const struct clocks_value clocks_value = - state->dis_clk->cur_clocks_value; - - /* 27mhz = 27000000hz= 27000khz */ - dp_phyclk_in_khz = link_settings.link_rate * 27000; - - if (((clocks_value.max_non_dp_phyclk_in_khz != 0) && - (dp_phyclk_in_khz > clocks_value.max_non_dp_phyclk_in_khz)) || - (dp_phyclk_in_khz > clocks_value.max_dp_phyclk_in_khz)) { - state->dis_clk->funcs->apply_clock_voltage_request( - state->dis_clk, - DM_PP_CLOCK_TYPE_DISPLAYPHYCLK, - dp_phyclk_in_khz, - false, - true); - } - } + struct dc_clocks clocks = state->bw.dcn.clk; + + /* dce/dcn compat, do not update dispclk */ + clocks.dispclk_khz = 0; + /* 27mhz = 27000000hz= 27000khz */ + clocks.phyclk_khz = link_settings.link_rate * 27000; + + state->dis_clk->funcs->update_clocks( + state->dis_clk, &clocks, false); } dp_enable_link_phy( @@ -1861,28 +1889,6 @@ static enum dc_status enable_link( break; } - if (pipe_ctx->stream_res.audio && status == DC_OK) { - struct dc *core_dc = pipe_ctx->stream->ctx->dc; - /* notify audio driver for audio modes of monitor */ - struct pp_smu_funcs_rv *pp_smu = core_dc->res_pool->pp_smu; - unsigned int i, num_audio = 1; - for (i = 0; i < MAX_PIPES; i++) { - /*current_state not updated yet*/ - if (core_dc->current_state->res_ctx.pipe_ctx[i].stream_res.audio != NULL) - num_audio++; - } - - pipe_ctx->stream_res.audio->funcs->az_enable(pipe_ctx->stream_res.audio); - - if (num_audio == 1 && pp_smu != NULL && pp_smu->set_pme_wa_enable != NULL) - /*this is the first audio. apply the PME w/a in order to wake AZ from D3*/ - pp_smu->set_pme_wa_enable(&pp_smu->pp_smu); - /* un-mute audio */ - /* TODO: audio should be per stream rather than per link */ - pipe_ctx->stream_res.stream_enc->funcs->audio_mute_control( - pipe_ctx->stream_res.stream_enc, false); - } - return status; } @@ -2023,6 +2029,15 @@ enum dc_status dc_link_validate_mode_timing( return DC_OK; } +int dc_link_get_backlight_level(const struct dc_link *link) +{ + struct abm *abm = link->ctx->dc->res_pool->abm; + + if (abm == NULL || abm->funcs->get_current_backlight_8_bit == NULL) + return DC_ERROR_UNEXPECTED; + + return (int) abm->funcs->get_current_backlight_8_bit(abm); +} bool dc_link_set_backlight_level(const struct dc_link *link, uint32_t level, uint32_t frame_ramp, const struct dc_stream_state *stream) @@ -2415,10 +2430,13 @@ void core_link_enable_stream( } } + core_dc->hwss.enable_audio_stream(pipe_ctx); + /* turn off otg test pattern if enable */ - pipe_ctx->stream_res.tg->funcs->set_test_pattern(pipe_ctx->stream_res.tg, - CONTROLLER_DP_TEST_PATTERN_VIDEOMODE, - COLOR_DEPTH_UNDEFINED); + if (pipe_ctx->stream_res.tg->funcs->set_test_pattern) + pipe_ctx->stream_res.tg->funcs->set_test_pattern(pipe_ctx->stream_res.tg, + CONTROLLER_DP_TEST_PATTERN_VIDEOMODE, + COLOR_DEPTH_UNDEFINED); core_dc->hwss.enable_stream(pipe_ctx); @@ -2453,6 +2471,22 @@ void core_link_set_avmute(struct pipe_ctx *pipe_ctx, bool enable) core_dc->hwss.set_avmute(pipe_ctx, enable); } +/** + ***************************************************************************** + * Function: dc_link_enable_hpd_filter + * + * @brief + * If enable is true, programs HPD filter on associated HPD line using + * delay_on_disconnect/delay_on_connect values dependent on + * link->connector_signal + * + * If enable is false, programs HPD filter on associated HPD line with no + * delays on connect or disconnect + * + * @param [in] link: pointer to the dc link + * @param [in] enable: boolean specifying whether to enable hbd + ***************************************************************************** + */ void dc_link_enable_hpd_filter(struct dc_link *link, bool enable) { struct gpio *hpd; diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c index 49c2face1e7a..8def0d9fa0ff 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c @@ -33,6 +33,7 @@ #include "include/vector.h" #include "core_types.h" #include "dc_link_ddc.h" +#include "aux_engine.h" #define AUX_POWER_UP_WA_DELAY 500 #define I2C_OVER_AUX_DEFER_WA_DELAY 70 @@ -629,79 +630,61 @@ bool dal_ddc_service_query_ddc_data( return ret; } -ssize_t dal_ddc_service_read_dpcd_data( - struct ddc_service *ddc, - bool i2c, - enum i2c_mot_mode mot, - uint32_t address, - uint8_t *data, - uint32_t len) +int dc_link_aux_transfer(struct ddc_service *ddc, + unsigned int address, + uint8_t *reply, + void *buffer, + unsigned int size, + enum aux_transaction_type type, + enum i2caux_transaction_action action) { - struct aux_payload read_payload = { - .i2c_over_aux = i2c, - .write = false, - .address = address, - .length = len, - .data = data, - }; - struct aux_command command = { - .payloads = &read_payload, - .number_of_payloads = 1, - .defer_delay = 0, - .max_defer_write_retry = 0, - .mot = mot - }; - - if (len > DEFAULT_AUX_MAX_DATA_SIZE) { - BREAK_TO_DEBUGGER(); - return DDC_RESULT_FAILED_INVALID_OPERATION; - } + struct ddc *ddc_pin = ddc->ddc_pin; + struct aux_engine *aux_engine; + enum aux_channel_operation_result operation_result; + struct aux_request_transaction_data aux_req; + struct aux_reply_transaction_data aux_rep; + uint8_t returned_bytes = 0; + int res = -1; + uint32_t status; - if (dal_i2caux_submit_aux_command( - ddc->ctx->i2caux, - ddc->ddc_pin, - &command)) { - return (ssize_t)command.payloads->length; - } + memset(&aux_req, 0, sizeof(aux_req)); + memset(&aux_rep, 0, sizeof(aux_rep)); - return DDC_RESULT_FAILED_OPERATION; -} + aux_engine = ddc->ctx->dc->res_pool->engines[ddc_pin->pin_data->en]; + aux_engine->funcs->acquire(aux_engine, ddc_pin); -enum ddc_result dal_ddc_service_write_dpcd_data( - struct ddc_service *ddc, - bool i2c, - enum i2c_mot_mode mot, - uint32_t address, - const uint8_t *data, - uint32_t len) -{ - struct aux_payload write_payload = { - .i2c_over_aux = i2c, - .write = true, - .address = address, - .length = len, - .data = (uint8_t *)data, - }; - struct aux_command command = { - .payloads = &write_payload, - .number_of_payloads = 1, - .defer_delay = 0, - .max_defer_write_retry = 0, - .mot = mot - }; - - if (len > DEFAULT_AUX_MAX_DATA_SIZE) { - BREAK_TO_DEBUGGER(); - return DDC_RESULT_FAILED_INVALID_OPERATION; - } + aux_req.type = type; + aux_req.action = action; + + aux_req.address = address; + aux_req.delay = 0; + aux_req.length = size; + aux_req.data = buffer; - if (dal_i2caux_submit_aux_command( - ddc->ctx->i2caux, - ddc->ddc_pin, - &command)) - return DDC_RESULT_SUCESSFULL; + aux_engine->funcs->submit_channel_request(aux_engine, &aux_req); + operation_result = aux_engine->funcs->get_channel_status(aux_engine, &returned_bytes); - return DDC_RESULT_FAILED_OPERATION; + switch (operation_result) { + case AUX_CHANNEL_OPERATION_SUCCEEDED: + res = returned_bytes; + + if (res <= size && res >= 0) + res = aux_engine->funcs->read_channel_reply(aux_engine, size, + buffer, reply, + &status); + + break; + case AUX_CHANNEL_OPERATION_FAILED_HPD_DISCON: + res = 0; + break; + case AUX_CHANNEL_OPERATION_FAILED_REASON_UNKNOWN: + case AUX_CHANNEL_OPERATION_FAILED_INVALID_REPLY: + case AUX_CHANNEL_OPERATION_FAILED_TIMEOUT: + res = -1; + break; + } + aux_engine->funcs->release_engine(aux_engine); + return res; } /*test only function*/ diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c index bdd121485cbc..a7553b6d59c2 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c @@ -3,6 +3,7 @@ #include "dc.h" #include "dc_link_dp.h" #include "dm_helpers.h" +#include "opp.h" #include "inc/core_types.h" #include "link_hwss.h" @@ -38,7 +39,7 @@ static bool decide_fallback_link_setting( struct dc_link_settings initial_link_settings, struct dc_link_settings *current_link_setting, enum link_training_result training_result); -static struct dc_link_settings get_common_supported_link_settings ( +static struct dc_link_settings get_common_supported_link_settings( struct dc_link_settings link_setting_a, struct dc_link_settings link_setting_b); @@ -93,8 +94,8 @@ static void dpcd_set_link_settings( uint8_t rate = (uint8_t) (lt_settings->link_settings.link_rate); - union down_spread_ctrl downspread = {{0}}; - union lane_count_set lane_count_set = {{0}}; + union down_spread_ctrl downspread = { {0} }; + union lane_count_set lane_count_set = { {0} }; uint8_t link_set_buffer[2]; downspread.raw = (uint8_t) @@ -164,11 +165,11 @@ static void dpcd_set_lt_pattern_and_lane_settings( const struct link_training_settings *lt_settings, enum hw_dp_training_pattern pattern) { - union dpcd_training_lane dpcd_lane[LANE_COUNT_DP_MAX] = {{{0}}}; + union dpcd_training_lane dpcd_lane[LANE_COUNT_DP_MAX] = { { {0} } }; const uint32_t dpcd_base_lt_offset = DP_TRAINING_PATTERN_SET; uint8_t dpcd_lt_buffer[5] = {0}; - union dpcd_training_pattern dpcd_pattern = {{0}}; + union dpcd_training_pattern dpcd_pattern = { {0} }; uint32_t lane; uint32_t size_in_bytes; bool edp_workaround = false; /* TODO link_prop.INTERNAL */ @@ -232,7 +233,7 @@ static void dpcd_set_lt_pattern_and_lane_settings( link, DP_TRAINING_PATTERN_SET, &dpcd_pattern.raw, - sizeof(dpcd_pattern.raw) ); + sizeof(dpcd_pattern.raw)); core_link_write_dpcd( link, @@ -246,7 +247,7 @@ static void dpcd_set_lt_pattern_and_lane_settings( link, dpcd_base_lt_offset, dpcd_lt_buffer, - size_in_bytes + sizeof(dpcd_pattern.raw) ); + size_in_bytes + sizeof(dpcd_pattern.raw)); link->cur_lane_setting = lt_settings->lane_settings[0]; } @@ -428,8 +429,8 @@ static void get_lane_status_and_drive_settings( struct link_training_settings *req_settings) { uint8_t dpcd_buf[6] = {0}; - union lane_adjust dpcd_lane_adjust[LANE_COUNT_DP_MAX] = {{{0}}}; - struct link_training_settings request_settings = {{0}}; + union lane_adjust dpcd_lane_adjust[LANE_COUNT_DP_MAX] = { { {0} } }; + struct link_training_settings request_settings = { {0} }; uint32_t lane; memset(req_settings, '\0', sizeof(struct link_training_settings)); @@ -651,7 +652,7 @@ static bool perform_post_lt_adj_req_sequence( if (req_drv_setting_changed) { update_drive_settings( - lt_settings,req_settings); + lt_settings, req_settings); dc_link_dp_set_drive_settings(link, lt_settings); @@ -724,8 +725,8 @@ static enum link_training_result perform_channel_equalization_sequence( enum hw_dp_training_pattern hw_tr_pattern; uint32_t retries_ch_eq; enum dc_lane_count lane_count = lt_settings->link_settings.lane_count; - union lane_align_status_updated dpcd_lane_status_updated = {{0}}; - union lane_status dpcd_lane_status[LANE_COUNT_DP_MAX] = {{{0}}}; + union lane_align_status_updated dpcd_lane_status_updated = { {0} }; + union lane_status dpcd_lane_status[LANE_COUNT_DP_MAX] = { { {0} } }; hw_tr_pattern = get_supported_tp(link); @@ -952,7 +953,10 @@ enum link_training_result dc_link_dp_perform_link_training( * LINK_SPREAD_05_DOWNSPREAD_30KHZ : * LINK_SPREAD_DISABLED; */ - lt_settings.link_settings.link_spread = LINK_SPREAD_05_DOWNSPREAD_30KHZ; + if (link->dp_ss_off) + lt_settings.link_settings.link_spread = LINK_SPREAD_DISABLED; + else + lt_settings.link_settings.link_spread = LINK_SPREAD_05_DOWNSPREAD_30KHZ; /* 1. set link rate, lane count and spread*/ dpcd_set_link_settings(link, <_settings); @@ -1027,6 +1031,9 @@ enum link_training_result dc_link_dp_perform_link_training( lt_settings.lane_settings[0].VOLTAGE_SWING, lt_settings.lane_settings[0].PRE_EMPHASIS); + if (status != LINK_TRAINING_SUCCESS) + link->ctx->dc->debug_data.ltFailCount++; + return status; } @@ -1082,9 +1089,10 @@ static struct dc_link_settings get_max_link_cap(struct dc_link *link) return max_link_cap; } -bool dp_hbr_verify_link_cap( +bool dp_verify_link_cap( struct dc_link *link, - struct dc_link_settings *known_limit_link_setting) + struct dc_link_settings *known_limit_link_setting, + int *fail_count) { struct dc_link_settings max_link_cap = {0}; struct dc_link_settings cur_link_setting = {0}; @@ -1097,6 +1105,11 @@ bool dp_hbr_verify_link_cap( enum clock_source_id dp_cs_id = CLOCK_SOURCE_ID_EXTERNAL; enum link_training_result status; + if (link->dc->debug.skip_detection_link_training) { + link->verified_link_cap = *known_limit_link_setting; + return true; + } + success = false; skip_link_training = false; @@ -1151,6 +1164,8 @@ bool dp_hbr_verify_link_cap( skip_video_pattern); if (status == LINK_TRAINING_SUCCESS) success = true; + else + (*fail_count)++; } if (success) @@ -1182,7 +1197,7 @@ bool dp_hbr_verify_link_cap( return success; } -static struct dc_link_settings get_common_supported_link_settings ( +static struct dc_link_settings get_common_supported_link_settings( struct dc_link_settings link_setting_a, struct dc_link_settings link_setting_b) { @@ -1428,6 +1443,7 @@ static uint32_t bandwidth_in_kbps_from_link_settings( uint32_t lane_count = link_setting->lane_count; uint32_t kbps = link_rate_in_kbps; + kbps *= lane_count; kbps *= 8; /* 8 bits per byte*/ @@ -1445,9 +1461,9 @@ bool dp_validate_mode_timing( const struct dc_link_settings *link_setting; /*always DP fail safe mode*/ - if (timing->pix_clk_khz == (uint32_t)25175 && - timing->h_addressable == (uint32_t)640 && - timing->v_addressable == (uint32_t)480) + if (timing->pix_clk_khz == (uint32_t) 25175 && + timing->h_addressable == (uint32_t) 640 && + timing->v_addressable == (uint32_t) 480) return true; /* We always use verified link settings */ @@ -1647,22 +1663,26 @@ static enum dc_status read_hpd_rx_irq_data( irq_data->raw, sizeof(union hpd_irq_data)); else { - /* Read 2 bytes at this location,... */ + /* Read 14 bytes in a single read and then copy only the required fields. + * This is more efficient than doing it in two separate AUX reads. */ + + uint8_t tmp[DP_SINK_STATUS_ESI - DP_SINK_COUNT_ESI + 1]; + retval = core_link_read_dpcd( link, DP_SINK_COUNT_ESI, - irq_data->raw, - 2); + tmp, + sizeof(tmp)); if (retval != DC_OK) return retval; - /* ... then read remaining 4 at the other location */ - retval = core_link_read_dpcd( - link, - DP_LANE0_1_STATUS_ESI, - &irq_data->raw[2], - 4); + irq_data->bytes.sink_cnt.raw = tmp[DP_SINK_COUNT_ESI - DP_SINK_COUNT_ESI]; + irq_data->bytes.device_service_irq.raw = tmp[DP_DEVICE_SERVICE_IRQ_VECTOR_ESI0 - DP_SINK_COUNT_ESI]; + irq_data->bytes.lane01_status.raw = tmp[DP_LANE0_1_STATUS_ESI - DP_SINK_COUNT_ESI]; + irq_data->bytes.lane23_status.raw = tmp[DP_LANE2_3_STATUS_ESI - DP_SINK_COUNT_ESI]; + irq_data->bytes.lane_status_updated.raw = tmp[DP_LANE_ALIGN_STATUS_UPDATED_ESI - DP_SINK_COUNT_ESI]; + irq_data->bytes.sink_status.raw = tmp[DP_SINK_STATUS_ESI - DP_SINK_COUNT_ESI]; } return retval; @@ -1989,12 +2009,16 @@ static void handle_automated_test(struct dc_link *link) sizeof(test_response)); } -bool dc_link_handle_hpd_rx_irq(struct dc_link *link, union hpd_irq_data *out_hpd_irq_dpcd_data) +bool dc_link_handle_hpd_rx_irq(struct dc_link *link, union hpd_irq_data *out_hpd_irq_dpcd_data, bool *out_link_loss) { - union hpd_irq_data hpd_irq_dpcd_data = {{{{0}}}}; + union hpd_irq_data hpd_irq_dpcd_data = { { { {0} } } }; union device_service_irq device_service_clear = { { 0 } }; - enum dc_status result = DDC_RESULT_UNKNOWN; + enum dc_status result; + bool status = false; + + if (out_link_loss) + *out_link_loss = false; /* For use cases related to down stream connection status change, * PSR and device auto test, refer to function handle_sst_hpd_irq * in DAL2.1*/ @@ -2069,6 +2093,8 @@ bool dc_link_handle_hpd_rx_irq(struct dc_link *link, union hpd_irq_data *out_hpd true, LINK_TRAINING_ATTEMPTS); status = false; + if (out_link_loss) + *out_link_loss = true; } if (link->type == dc_connection_active_dongle && @@ -2255,6 +2281,11 @@ static void get_active_converter_info( link->dpcd_caps.branch_hw_revision = dp_hw_fw_revision.ieee_hw_rev; + + memmove( + link->dpcd_caps.branch_fw_revision, + dp_hw_fw_revision.ieee_fw_rev, + sizeof(dp_hw_fw_revision.ieee_fw_rev)); } } @@ -2303,12 +2334,14 @@ static bool retrieve_link_cap(struct dc_link *link) { uint8_t dpcd_data[DP_ADAPTER_CAP - DP_DPCD_REV + 1]; + struct dp_device_vendor_id sink_id; union down_stream_port_count down_strm_port_count; union edp_configuration_cap edp_config_cap; union dp_downstream_port_present ds_port = { 0 }; enum dc_status status = DC_ERROR_UNEXPECTED; uint32_t read_dpcd_retry_cnt = 3; int i; + struct dp_sink_hw_fw_revision dp_hw_fw_revision; memset(dpcd_data, '\0', sizeof(dpcd_data)); memset(&down_strm_port_count, @@ -2389,6 +2422,36 @@ static bool retrieve_link_cap(struct dc_link *link) &link->dpcd_caps.sink_count.raw, sizeof(link->dpcd_caps.sink_count.raw)); + /* read sink ieee oui */ + core_link_read_dpcd(link, + DP_SINK_OUI, + (uint8_t *)(&sink_id), + sizeof(sink_id)); + + link->dpcd_caps.sink_dev_id = + (sink_id.ieee_oui[0] << 16) + + (sink_id.ieee_oui[1] << 8) + + (sink_id.ieee_oui[2]); + + memmove( + link->dpcd_caps.sink_dev_id_str, + sink_id.ieee_device_id, + sizeof(sink_id.ieee_device_id)); + + core_link_read_dpcd( + link, + DP_SINK_HW_REVISION_START, + (uint8_t *)&dp_hw_fw_revision, + sizeof(dp_hw_fw_revision)); + + link->dpcd_caps.sink_hw_revision = + dp_hw_fw_revision.ieee_hw_rev; + + memmove( + link->dpcd_caps.sink_fw_revision, + dp_hw_fw_revision.ieee_fw_rev, + sizeof(dp_hw_fw_revision.ieee_fw_rev)); + /* Connectivity log: detection */ CONN_DATA_DETECT(link, dpcd_data, sizeof(dpcd_data), "Rx Caps: "); @@ -2493,8 +2556,8 @@ static void set_crtc_test_pattern(struct dc_link *link, pipe_ctx->stream->bit_depth_params = params; pipe_ctx->stream_res.opp->funcs-> opp_program_bit_depth_reduction(pipe_ctx->stream_res.opp, ¶ms); - - pipe_ctx->stream_res.tg->funcs->set_test_pattern(pipe_ctx->stream_res.tg, + if (pipe_ctx->stream_res.tg->funcs->set_test_pattern) + pipe_ctx->stream_res.tg->funcs->set_test_pattern(pipe_ctx->stream_res.tg, controller_test_pattern, color_depth); } break; @@ -2506,8 +2569,8 @@ static void set_crtc_test_pattern(struct dc_link *link, pipe_ctx->stream->bit_depth_params = params; pipe_ctx->stream_res.opp->funcs-> opp_program_bit_depth_reduction(pipe_ctx->stream_res.opp, ¶ms); - - pipe_ctx->stream_res.tg->funcs->set_test_pattern(pipe_ctx->stream_res.tg, + if (pipe_ctx->stream_res.tg->funcs->set_test_pattern) + pipe_ctx->stream_res.tg->funcs->set_test_pattern(pipe_ctx->stream_res.tg, CONTROLLER_DP_TEST_PATTERN_VIDEOMODE, color_depth); } diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c index 751f3ac9d921..4ca41d6e3bcf 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c @@ -41,7 +41,7 @@ #include "dce100/dce100_resource.h" #include "dce110/dce110_resource.h" #include "dce112/dce112_resource.h" -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 #include "dcn10/dcn10_resource.h" #endif #include "dce120/dce120_resource.h" @@ -85,7 +85,7 @@ enum dce_version resource_parse_asic_id(struct hw_asic_id asic_id) case FAMILY_AI: dc_version = DCE_VERSION_12_0; break; -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 case FAMILY_RV: dc_version = DCN_VERSION_1_0; break; @@ -136,7 +136,7 @@ struct resource_pool *dc_create_resource_pool( num_virtual_links, dc); break; -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 case DCN_VERSION_1_0: res_pool = dcn10_create_resource_pool( num_virtual_links, dc); @@ -330,6 +330,9 @@ bool resource_are_streams_timing_synchronizable( != stream2->timing.pix_clk_khz) return false; + if (stream1->clamping.c_depth != stream2->clamping.c_depth) + return false; + if (stream1->phy_pix_clk != stream2->phy_pix_clk && (!dc_is_dp_signal(stream1->signal) || !dc_is_dp_signal(stream2->signal))) @@ -337,6 +340,20 @@ bool resource_are_streams_timing_synchronizable( return true; } +static bool is_dp_and_hdmi_sharable( + struct dc_stream_state *stream1, + struct dc_stream_state *stream2) +{ + if (stream1->ctx->dc->caps.disable_dp_clk_share) + return false; + + if (stream1->clamping.c_depth != COLOR_DEPTH_888 || + stream2->clamping.c_depth != COLOR_DEPTH_888) + return false; + + return true; + +} static bool is_sharable_clk_src( const struct pipe_ctx *pipe_with_clk_src, @@ -348,7 +365,10 @@ static bool is_sharable_clk_src( if (pipe_with_clk_src->stream->signal == SIGNAL_TYPE_VIRTUAL) return false; - if (dc_is_dp_signal(pipe_with_clk_src->stream->signal)) + if (dc_is_dp_signal(pipe_with_clk_src->stream->signal) || + (dc_is_dp_signal(pipe->stream->signal) && + !is_dp_and_hdmi_sharable(pipe_with_clk_src->stream, + pipe->stream))) return false; if (dc_is_hdmi_signal(pipe_with_clk_src->stream->signal) @@ -522,13 +542,12 @@ static void calculate_viewport(struct pipe_ctx *pipe_ctx) } } -static void calculate_recout(struct pipe_ctx *pipe_ctx, struct view *recout_skip) +static void calculate_recout(struct pipe_ctx *pipe_ctx, struct rect *recout_full) { const struct dc_plane_state *plane_state = pipe_ctx->plane_state; const struct dc_stream_state *stream = pipe_ctx->stream; 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 && @@ -597,20 +616,22 @@ static void calculate_recout(struct pipe_ctx *pipe_ctx, struct view *recout_skip } } /* 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) + * * 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; - recout_skip->width = pipe_ctx->plane_res.scl_data.recout.x - recout_full_x; - recout_skip->height = pipe_ctx->plane_res.scl_data.recout.y - recout_full_y; + recout_full->width = plane_state->dst_rect.width + * stream->dst.width / stream->src.width; + recout_full->height = plane_state->dst_rect.height + * stream->dst.height / stream->src.height; } static void calculate_scaling_ratios(struct pipe_ctx *pipe_ctx) @@ -662,7 +683,7 @@ static void calculate_scaling_ratios(struct pipe_ctx *pipe_ctx) pipe_ctx->plane_res.scl_data.ratios.vert_c, 19); } -static void calculate_inits_and_adj_vp(struct pipe_ctx *pipe_ctx, struct view *recout_skip) +static void calculate_inits_and_adj_vp(struct pipe_ctx *pipe_ctx, struct rect *recout_full) { struct scaler_data *data = &pipe_ctx->plane_res.scl_data; struct rect src = pipe_ctx->plane_state->src_rect; @@ -680,15 +701,14 @@ static void calculate_inits_and_adj_vp(struct pipe_ctx *pipe_ctx, struct view *r 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) { rect_swap_helper(&src); rect_swap_helper(&data->viewport_c); rect_swap_helper(&data->viewport); - } + } else if (pipe_ctx->plane_state->horizontal_mirror) + flip_horz_scan_dir = !flip_horz_scan_dir; /* * Init calculated according to formula: @@ -708,127 +728,286 @@ static void calculate_inits_and_adj_vp(struct pipe_ctx *pipe_ctx, struct view *r data->inits.v_c = dc_fixpt_truncate(dc_fixpt_add(data->inits.v_c, dc_fixpt_div_int( dc_fixpt_add_int(data->ratios.vert_c, data->taps.v_taps_c + 1), 2)), 19); + if (!flip_horz_scan_dir) { + /* Adjust for viewport end clip-off */ + if ((data->viewport.x + data->viewport.width) < (src.x + src.width)) { + int vp_clip = src.x + src.width - data->viewport.width - data->viewport.x; + int int_part = dc_fixpt_floor( + dc_fixpt_sub(data->inits.h, data->ratios.horz)); + int_part = int_part > 0 ? int_part : 0; + data->viewport.width += int_part < vp_clip ? int_part : vp_clip; + } + if ((data->viewport_c.x + data->viewport_c.width) < (src.x + src.width) / vpc_div) { + int vp_clip = (src.x + src.width) / vpc_div - + data->viewport_c.width - data->viewport_c.x; + int int_part = dc_fixpt_floor( + dc_fixpt_sub(data->inits.h_c, data->ratios.horz_c)); + + int_part = int_part > 0 ? int_part : 0; + data->viewport_c.width += int_part < vp_clip ? int_part : vp_clip; + } + + /* Adjust for non-0 viewport offset */ + if (data->viewport.x) { + int int_part; + + data->inits.h = dc_fixpt_add(data->inits.h, dc_fixpt_mul_int( + data->ratios.horz, data->recout.x - recout_full->x)); + int_part = dc_fixpt_floor(data->inits.h) - data->viewport.x; + if (int_part < data->taps.h_taps) { + int int_adj = data->viewport.x >= (data->taps.h_taps - int_part) ? + (data->taps.h_taps - int_part) : data->viewport.x; + data->viewport.x -= int_adj; + data->viewport.width += int_adj; + int_part += int_adj; + } else if (int_part > data->taps.h_taps) { + data->viewport.x += int_part - data->taps.h_taps; + data->viewport.width -= int_part - data->taps.h_taps; + int_part = data->taps.h_taps; + } + data->inits.h.value &= 0xffffffff; + data->inits.h = dc_fixpt_add_int(data->inits.h, int_part); + } - /* Adjust for viewport end clip-off */ - if ((data->viewport.x + data->viewport.width) < (src.x + src.width) && !flip_horz_scan_dir) { - int vp_clip = src.x + src.width - data->viewport.width - data->viewport.x; - int int_part = dc_fixpt_floor( - dc_fixpt_sub(data->inits.h, data->ratios.horz)); - - int_part = int_part > 0 ? int_part : 0; - data->viewport.width += int_part < vp_clip ? int_part : vp_clip; - } - if ((data->viewport.y + data->viewport.height) < (src.y + src.height) && !flip_vert_scan_dir) { - int vp_clip = src.y + src.height - data->viewport.height - data->viewport.y; - int int_part = dc_fixpt_floor( - dc_fixpt_sub(data->inits.v, data->ratios.vert)); - - int_part = int_part > 0 ? int_part : 0; - data->viewport.height += int_part < vp_clip ? int_part : vp_clip; - } - if ((data->viewport_c.x + data->viewport_c.width) < (src.x + src.width) / vpc_div && !flip_horz_scan_dir) { - int vp_clip = (src.x + src.width) / vpc_div - - data->viewport_c.width - data->viewport_c.x; - int int_part = dc_fixpt_floor( - dc_fixpt_sub(data->inits.h_c, data->ratios.horz_c)); - - int_part = int_part > 0 ? int_part : 0; - data->viewport_c.width += int_part < vp_clip ? int_part : vp_clip; - } - if ((data->viewport_c.y + data->viewport_c.height) < (src.y + src.height) / vpc_div && !flip_vert_scan_dir) { - int vp_clip = (src.y + src.height) / vpc_div - - data->viewport_c.height - data->viewport_c.y; - int int_part = dc_fixpt_floor( - dc_fixpt_sub(data->inits.v_c, data->ratios.vert_c)); - - int_part = int_part > 0 ? int_part : 0; - data->viewport_c.height += int_part < vp_clip ? int_part : vp_clip; - } - - /* Adjust for non-0 viewport offset */ - if (data->viewport.x && !flip_horz_scan_dir) { - int int_part; - - data->inits.h = dc_fixpt_add(data->inits.h, dc_fixpt_mul_int( - data->ratios.horz, recout_skip->width)); - int_part = dc_fixpt_floor(data->inits.h) - data->viewport.x; - if (int_part < data->taps.h_taps) { - int int_adj = data->viewport.x >= (data->taps.h_taps - int_part) ? - (data->taps.h_taps - int_part) : data->viewport.x; - data->viewport.x -= int_adj; - data->viewport.width += int_adj; - int_part += int_adj; - } else if (int_part > data->taps.h_taps) { - data->viewport.x += int_part - data->taps.h_taps; - data->viewport.width -= int_part - data->taps.h_taps; - int_part = data->taps.h_taps; + if (data->viewport_c.x) { + int int_part; + + data->inits.h_c = dc_fixpt_add(data->inits.h_c, dc_fixpt_mul_int( + data->ratios.horz_c, data->recout.x - recout_full->x)); + int_part = dc_fixpt_floor(data->inits.h_c) - data->viewport_c.x; + if (int_part < data->taps.h_taps_c) { + int int_adj = data->viewport_c.x >= (data->taps.h_taps_c - int_part) ? + (data->taps.h_taps_c - int_part) : data->viewport_c.x; + data->viewport_c.x -= int_adj; + data->viewport_c.width += int_adj; + int_part += int_adj; + } else if (int_part > data->taps.h_taps_c) { + data->viewport_c.x += int_part - data->taps.h_taps_c; + data->viewport_c.width -= int_part - data->taps.h_taps_c; + int_part = data->taps.h_taps_c; + } + data->inits.h_c.value &= 0xffffffff; + data->inits.h_c = dc_fixpt_add_int(data->inits.h_c, int_part); } - data->inits.h.value &= 0xffffffff; - data->inits.h = dc_fixpt_add_int(data->inits.h, int_part); - } - - if (data->viewport_c.x && !flip_horz_scan_dir) { - int int_part; - - data->inits.h_c = dc_fixpt_add(data->inits.h_c, dc_fixpt_mul_int( - data->ratios.horz_c, recout_skip->width)); - int_part = dc_fixpt_floor(data->inits.h_c) - data->viewport_c.x; - if (int_part < data->taps.h_taps_c) { - int int_adj = data->viewport_c.x >= (data->taps.h_taps_c - int_part) ? - (data->taps.h_taps_c - int_part) : data->viewport_c.x; - data->viewport_c.x -= int_adj; - data->viewport_c.width += int_adj; - int_part += int_adj; - } else if (int_part > data->taps.h_taps_c) { - data->viewport_c.x += int_part - data->taps.h_taps_c; - data->viewport_c.width -= int_part - data->taps.h_taps_c; - int_part = data->taps.h_taps_c; + } else { + /* Adjust for non-0 viewport offset */ + if (data->viewport.x) { + int int_part = dc_fixpt_floor( + dc_fixpt_sub(data->inits.h, data->ratios.horz)); + + int_part = int_part > 0 ? int_part : 0; + data->viewport.width += int_part < data->viewport.x ? int_part : data->viewport.x; + data->viewport.x -= int_part < data->viewport.x ? int_part : data->viewport.x; } - data->inits.h_c.value &= 0xffffffff; - data->inits.h_c = dc_fixpt_add_int(data->inits.h_c, int_part); - } - - if (data->viewport.y && !flip_vert_scan_dir) { - int int_part; - - data->inits.v = dc_fixpt_add(data->inits.v, dc_fixpt_mul_int( - data->ratios.vert, recout_skip->height)); - int_part = dc_fixpt_floor(data->inits.v) - data->viewport.y; - if (int_part < data->taps.v_taps) { - int int_adj = data->viewport.y >= (data->taps.v_taps - int_part) ? - (data->taps.v_taps - int_part) : data->viewport.y; - data->viewport.y -= int_adj; - data->viewport.height += int_adj; - int_part += int_adj; - } else if (int_part > data->taps.v_taps) { - data->viewport.y += int_part - data->taps.v_taps; - data->viewport.height -= int_part - data->taps.v_taps; - int_part = data->taps.v_taps; + if (data->viewport_c.x) { + int int_part = dc_fixpt_floor( + dc_fixpt_sub(data->inits.h_c, data->ratios.horz_c)); + + int_part = int_part > 0 ? int_part : 0; + data->viewport_c.width += int_part < data->viewport_c.x ? int_part : data->viewport_c.x; + data->viewport_c.x -= int_part < data->viewport_c.x ? int_part : data->viewport_c.x; } - data->inits.v.value &= 0xffffffff; - data->inits.v = dc_fixpt_add_int(data->inits.v, int_part); - } - - if (data->viewport_c.y && !flip_vert_scan_dir) { - int int_part; - - data->inits.v_c = dc_fixpt_add(data->inits.v_c, dc_fixpt_mul_int( - data->ratios.vert_c, recout_skip->height)); - int_part = dc_fixpt_floor(data->inits.v_c) - data->viewport_c.y; - if (int_part < data->taps.v_taps_c) { - int int_adj = data->viewport_c.y >= (data->taps.v_taps_c - int_part) ? - (data->taps.v_taps_c - int_part) : data->viewport_c.y; - data->viewport_c.y -= int_adj; - data->viewport_c.height += int_adj; - int_part += int_adj; - } else if (int_part > data->taps.v_taps_c) { - data->viewport_c.y += int_part - data->taps.v_taps_c; - data->viewport_c.height -= int_part - data->taps.v_taps_c; - int_part = data->taps.v_taps_c; + + /* Adjust for viewport end clip-off */ + if ((data->viewport.x + data->viewport.width) < (src.x + src.width)) { + int int_part; + int end_offset = src.x + src.width + - data->viewport.x - data->viewport.width; + + /* + * this is init if vp had no offset, keep in mind this is from the + * right side of vp due to scan direction + */ + data->inits.h = dc_fixpt_add(data->inits.h, dc_fixpt_mul_int( + data->ratios.horz, data->recout.x - recout_full->x)); + /* + * this is the difference between first pixel of viewport available to read + * and init position, takning into account scan direction + */ + int_part = dc_fixpt_floor(data->inits.h) - end_offset; + if (int_part < data->taps.h_taps) { + int int_adj = end_offset >= (data->taps.h_taps - int_part) ? + (data->taps.h_taps - int_part) : end_offset; + data->viewport.width += int_adj; + int_part += int_adj; + } else if (int_part > data->taps.h_taps) { + data->viewport.width += int_part - data->taps.h_taps; + int_part = data->taps.h_taps; + } + data->inits.h.value &= 0xffffffff; + data->inits.h = dc_fixpt_add_int(data->inits.h, int_part); + } + + if ((data->viewport_c.x + data->viewport_c.width) < (src.x + src.width) / vpc_div) { + int int_part; + int end_offset = (src.x + src.width) / vpc_div + - data->viewport_c.x - data->viewport_c.width; + + /* + * this is init if vp had no offset, keep in mind this is from the + * right side of vp due to scan direction + */ + data->inits.h_c = dc_fixpt_add(data->inits.h_c, dc_fixpt_mul_int( + data->ratios.horz_c, data->recout.x - recout_full->x)); + /* + * this is the difference between first pixel of viewport available to read + * and init position, takning into account scan direction + */ + int_part = dc_fixpt_floor(data->inits.h_c) - end_offset; + if (int_part < data->taps.h_taps_c) { + int int_adj = end_offset >= (data->taps.h_taps_c - int_part) ? + (data->taps.h_taps_c - int_part) : end_offset; + data->viewport_c.width += int_adj; + int_part += int_adj; + } else if (int_part > data->taps.h_taps_c) { + data->viewport_c.width += int_part - data->taps.h_taps_c; + int_part = data->taps.h_taps_c; + } + data->inits.h_c.value &= 0xffffffff; + data->inits.h_c = dc_fixpt_add_int(data->inits.h_c, int_part); + } + + } + if (!flip_vert_scan_dir) { + /* Adjust for viewport end clip-off */ + if ((data->viewport.y + data->viewport.height) < (src.y + src.height)) { + int vp_clip = src.y + src.height - data->viewport.height - data->viewport.y; + int int_part = dc_fixpt_floor( + dc_fixpt_sub(data->inits.v, data->ratios.vert)); + + int_part = int_part > 0 ? int_part : 0; + data->viewport.height += int_part < vp_clip ? int_part : vp_clip; + } + if ((data->viewport_c.y + data->viewport_c.height) < (src.y + src.height) / vpc_div) { + int vp_clip = (src.y + src.height) / vpc_div - + data->viewport_c.height - data->viewport_c.y; + int int_part = dc_fixpt_floor( + dc_fixpt_sub(data->inits.v_c, data->ratios.vert_c)); + + int_part = int_part > 0 ? int_part : 0; + data->viewport_c.height += int_part < vp_clip ? int_part : vp_clip; + } + + /* Adjust for non-0 viewport offset */ + if (data->viewport.y) { + int int_part; + + data->inits.v = dc_fixpt_add(data->inits.v, dc_fixpt_mul_int( + data->ratios.vert, data->recout.y - recout_full->y)); + int_part = dc_fixpt_floor(data->inits.v) - data->viewport.y; + if (int_part < data->taps.v_taps) { + int int_adj = data->viewport.y >= (data->taps.v_taps - int_part) ? + (data->taps.v_taps - int_part) : data->viewport.y; + data->viewport.y -= int_adj; + data->viewport.height += int_adj; + int_part += int_adj; + } else if (int_part > data->taps.v_taps) { + data->viewport.y += int_part - data->taps.v_taps; + data->viewport.height -= int_part - data->taps.v_taps; + int_part = data->taps.v_taps; + } + data->inits.v.value &= 0xffffffff; + data->inits.v = dc_fixpt_add_int(data->inits.v, int_part); + } + + if (data->viewport_c.y) { + int int_part; + + data->inits.v_c = dc_fixpt_add(data->inits.v_c, dc_fixpt_mul_int( + data->ratios.vert_c, data->recout.y - recout_full->y)); + int_part = dc_fixpt_floor(data->inits.v_c) - data->viewport_c.y; + if (int_part < data->taps.v_taps_c) { + int int_adj = data->viewport_c.y >= (data->taps.v_taps_c - int_part) ? + (data->taps.v_taps_c - int_part) : data->viewport_c.y; + data->viewport_c.y -= int_adj; + data->viewport_c.height += int_adj; + int_part += int_adj; + } else if (int_part > data->taps.v_taps_c) { + data->viewport_c.y += int_part - data->taps.v_taps_c; + data->viewport_c.height -= int_part - data->taps.v_taps_c; + int_part = data->taps.v_taps_c; + } + data->inits.v_c.value &= 0xffffffff; + data->inits.v_c = dc_fixpt_add_int(data->inits.v_c, int_part); + } + } else { + /* Adjust for non-0 viewport offset */ + if (data->viewport.y) { + int int_part = dc_fixpt_floor( + dc_fixpt_sub(data->inits.v, data->ratios.vert)); + + int_part = int_part > 0 ? int_part : 0; + data->viewport.height += int_part < data->viewport.y ? int_part : data->viewport.y; + data->viewport.y -= int_part < data->viewport.y ? int_part : data->viewport.y; + } + if (data->viewport_c.y) { + int int_part = dc_fixpt_floor( + dc_fixpt_sub(data->inits.v_c, data->ratios.vert_c)); + + int_part = int_part > 0 ? int_part : 0; + data->viewport_c.height += int_part < data->viewport_c.y ? int_part : data->viewport_c.y; + data->viewport_c.y -= int_part < data->viewport_c.y ? int_part : data->viewport_c.y; + } + + /* Adjust for viewport end clip-off */ + if ((data->viewport.y + data->viewport.height) < (src.y + src.height)) { + int int_part; + int end_offset = src.y + src.height + - data->viewport.y - data->viewport.height; + + /* + * this is init if vp had no offset, keep in mind this is from the + * right side of vp due to scan direction + */ + data->inits.v = dc_fixpt_add(data->inits.v, dc_fixpt_mul_int( + data->ratios.vert, data->recout.y - recout_full->y)); + /* + * this is the difference between first pixel of viewport available to read + * and init position, taking into account scan direction + */ + int_part = dc_fixpt_floor(data->inits.v) - end_offset; + if (int_part < data->taps.v_taps) { + int int_adj = end_offset >= (data->taps.v_taps - int_part) ? + (data->taps.v_taps - int_part) : end_offset; + data->viewport.height += int_adj; + int_part += int_adj; + } else if (int_part > data->taps.v_taps) { + data->viewport.height += int_part - data->taps.v_taps; + int_part = data->taps.v_taps; + } + data->inits.v.value &= 0xffffffff; + data->inits.v = dc_fixpt_add_int(data->inits.v, int_part); + } + + if ((data->viewport_c.y + data->viewport_c.height) < (src.y + src.height) / vpc_div) { + int int_part; + int end_offset = (src.y + src.height) / vpc_div + - data->viewport_c.y - data->viewport_c.height; + + /* + * this is init if vp had no offset, keep in mind this is from the + * right side of vp due to scan direction + */ + data->inits.v_c = dc_fixpt_add(data->inits.v_c, dc_fixpt_mul_int( + data->ratios.vert_c, data->recout.y - recout_full->y)); + /* + * this is the difference between first pixel of viewport available to read + * and init position, taking into account scan direction + */ + int_part = dc_fixpt_floor(data->inits.v_c) - end_offset; + if (int_part < data->taps.v_taps_c) { + int int_adj = end_offset >= (data->taps.v_taps_c - int_part) ? + (data->taps.v_taps_c - int_part) : end_offset; + data->viewport_c.height += int_adj; + int_part += int_adj; + } else if (int_part > data->taps.v_taps_c) { + data->viewport_c.height += int_part - data->taps.v_taps_c; + int_part = data->taps.v_taps_c; + } + data->inits.v_c.value &= 0xffffffff; + data->inits.v_c = dc_fixpt_add_int(data->inits.v_c, int_part); } - data->inits.v_c.value &= 0xffffffff; - data->inits.v_c = dc_fixpt_add_int(data->inits.v_c, int_part); } /* Interlaced inits based on final vert inits */ @@ -846,7 +1025,7 @@ bool resource_build_scaling_params(struct pipe_ctx *pipe_ctx) { const struct dc_plane_state *plane_state = pipe_ctx->plane_state; struct dc_crtc_timing *timing = &pipe_ctx->stream->timing; - struct view recout_skip = { 0 }; + struct rect recout_full = { 0 }; bool res = false; DC_LOGGER_INIT(pipe_ctx->stream->ctx->logger); /* Important: scaling ratio calculation requires pixel format, @@ -866,7 +1045,7 @@ bool resource_build_scaling_params(struct pipe_ctx *pipe_ctx) if (pipe_ctx->plane_res.scl_data.viewport.height < 16 || pipe_ctx->plane_res.scl_data.viewport.width < 16) return false; - calculate_recout(pipe_ctx, &recout_skip); + calculate_recout(pipe_ctx, &recout_full); /** * Setting line buffer pixel depth to 24bpp yields banding @@ -910,7 +1089,7 @@ bool resource_build_scaling_params(struct pipe_ctx *pipe_ctx) if (res) /* May need to re-check lb size after this in some obscure scenario */ - calculate_inits_and_adj_vp(pipe_ctx, &recout_skip); + calculate_inits_and_adj_vp(pipe_ctx, &recout_full); DC_LOG_SCALER( "%s: Viewport:\nheight:%d width:%d x:%d " @@ -1054,7 +1233,7 @@ static struct pipe_ctx *acquire_free_pipe_for_stream( } -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 static int acquire_first_split_pipe( struct resource_context *res_ctx, const struct resource_pool *pool, @@ -1125,7 +1304,7 @@ bool dc_add_plane_to_context( free_pipe = acquire_free_pipe_for_stream(context, pool, stream); -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 if (!free_pipe) { int pipe_idx = acquire_first_split_pipe(&context->res_ctx, pool, stream); if (pipe_idx >= 0) @@ -1546,8 +1725,8 @@ enum dc_status dc_add_stream_to_ctx( struct dc_context *dc_ctx = dc->ctx; enum dc_status res; - if (new_ctx->stream_count >= dc->res_pool->pipe_count) { - DC_ERROR("Max streams reached, can add stream %p !\n", stream); + if (new_ctx->stream_count >= dc->res_pool->timing_generator_count) { + DC_ERROR("Max streams reached, can't add stream %p !\n", stream); return DC_ERROR_UNEXPECTED; } @@ -1723,7 +1902,7 @@ enum dc_status resource_map_pool_resources( /* acquire new resources */ pipe_idx = acquire_first_free_pipe(&context->res_ctx, pool, stream); -#ifdef CONFIG_DRM_AMD_DC_DCN1_0 +#ifdef CONFIG_X86 if (pipe_idx < 0) pipe_idx = acquire_first_split_pipe(&context->res_ctx, pool, stream); #endif @@ -1789,7 +1968,7 @@ void dc_resource_state_construct( const struct dc *dc, struct dc_state *dst_ctx) { - dst_ctx->dis_clk = dc->res_pool->display_clock; + dst_ctx->dis_clk = dc->res_pool->dccg; } enum dc_status dc_validate_global_state( @@ -2347,7 +2526,8 @@ static void set_hdr_static_info_packet( { /* HDR Static Metadata info packet for HDR10 */ - if (!stream->hdr_static_metadata.valid) + if (!stream->hdr_static_metadata.valid || + stream->use_dynamic_meta) return; *info_packet = stream->hdr_static_metadata; diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_sink.c b/drivers/gpu/drm/amd/display/dc/core/dc_sink.c index 25fae38409ab..9971b515c3eb 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_sink.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_sink.c @@ -53,6 +53,10 @@ static bool construct(struct dc_sink *sink, const struct dc_sink_init_data *init sink->dongle_max_pix_clk = init_params->dongle_max_pix_clk; sink->converter_disable_audio = init_params->converter_disable_audio; sink->dc_container_id = NULL; + sink->sink_id = init_params->link->ctx->dc_sink_id_count; + // increment dc_sink_id_count because we don't want two sinks with same ID + // unless they are actually the same + init_params->link->ctx->dc_sink_id_count++; return true; } diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_stream.c b/drivers/gpu/drm/amd/display/dc/core/dc_stream.c index 3732a1de9d6c..fdcc8ab19bf3 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_stream.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_stream.c @@ -30,6 +30,8 @@ #include "ipp.h" #include "timing_generator.h" +#define DC_LOGGER dc->ctx->logger + /******************************************************************************* * Private functions ******************************************************************************/ @@ -212,6 +214,8 @@ bool dc_stream_set_cursor_attributes( } core_dc->hwss.set_cursor_attribute(pipe_ctx); + if (core_dc->hwss.set_cursor_sdr_white_level) + core_dc->hwss.set_cursor_sdr_white_level(pipe_ctx); } if (pipe_to_program) @@ -317,16 +321,10 @@ bool dc_stream_get_scanoutpos(const struct dc_stream_state *stream, return ret; } - -void dc_stream_log( - const struct dc_stream_state *stream, - struct dal_logger *dm_logger, - enum dc_log_type log_type) +void dc_stream_log(const struct dc *dc, const struct dc_stream_state *stream) { - - dm_logger_write(dm_logger, - log_type, - "core_stream 0x%x: src: %d, %d, %d, %d; dst: %d, %d, %d, %d, colorSpace:%d\n", + DC_LOG_DC( + "core_stream 0x%p: src: %d, %d, %d, %d; dst: %d, %d, %d, %d, colorSpace:%d\n", stream, stream->src.x, stream->src.y, @@ -337,21 +335,18 @@ void dc_stream_log( stream->dst.width, stream->dst.height, stream->output_color_space); - dm_logger_write(dm_logger, - log_type, + DC_LOG_DC( "\tpix_clk_khz: %d, h_total: %d, v_total: %d, pixelencoder:%d, displaycolorDepth:%d\n", stream->timing.pix_clk_khz, stream->timing.h_total, stream->timing.v_total, stream->timing.pixel_encoding, stream->timing.display_color_depth); - dm_logger_write(dm_logger, - log_type, + DC_LOG_DC( "\tsink name: %s, serial: %d\n", stream->sink->edid_caps.display_name, stream->sink->edid_caps.serial_number); - dm_logger_write(dm_logger, - log_type, + DC_LOG_DC( "\tlink: %d\n", stream->sink->link->link_index); } diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_surface.c b/drivers/gpu/drm/amd/display/dc/core/dc_surface.c index 68a71adeb12e..8fb3aefd195c 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_surface.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_surface.c @@ -84,6 +84,17 @@ struct dc_plane_state *dc_create_plane_state(struct dc *dc) return plane_state; } +/** + ***************************************************************************** + * Function: dc_plane_get_status + * + * @brief + * Looks up the pipe context of plane_state and updates the pending status + * of the pipe context. Then returns plane_state->status + * + * @param [in] plane_state: pointer to the plane_state to get the status of + ***************************************************************************** + */ const struct dc_plane_status *dc_plane_get_status( const struct dc_plane_state *plane_state) { @@ -181,7 +192,7 @@ void dc_transfer_func_release(struct dc_transfer_func *tf) kref_put(&tf->refcount, dc_transfer_func_free); } -struct dc_transfer_func *dc_create_transfer_func() +struct dc_transfer_func *dc_create_transfer_func(void) { struct dc_transfer_func *tf = kvzalloc(sizeof(*tf), GFP_KERNEL); diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h index 53c71296f3dd..e2f033d420a0 100644 --- a/drivers/gpu/drm/amd/display/dc/dc.h +++ b/drivers/gpu/drm/amd/display/dc/dc.h @@ -38,7 +38,7 @@ #include "inc/compressor.h" #include "dml/display_mode_lib.h" -#define DC_VER "3.1.44" +#define DC_VER "3.1.59" #define MAX_SURFACES 3 #define MAX_STREAMS 6 @@ -68,6 +68,7 @@ struct dc_caps { uint32_t max_planes; uint32_t max_downscale_ratio; uint32_t i2c_speed_in_khz; + uint32_t dmdata_alloc_size; unsigned int max_cursor_size; unsigned int max_video_width; int linear_pitch_alignment; @@ -77,6 +78,8 @@ struct dc_caps { bool dual_link_dvi; bool post_blend_color_processing; bool force_dp_tps4_for_cp2520; + bool disable_dp_clk_share; + bool psp_setup_panel_mode; }; struct dc_dcc_surface_param { @@ -169,6 +172,12 @@ struct dc_config { bool disable_disp_pll_sharing; }; +enum visual_confirm { + VISUAL_CONFIRM_DISABLE = 0, + VISUAL_CONFIRM_SURFACE = 1, + VISUAL_CONFIRM_HDR = 2, +}; + enum dcc_option { DCC_ENABLE = 0, DCC_DISABLE = 1, @@ -186,6 +195,10 @@ enum wm_report_mode { WM_REPORT_OVERRIDE = 1, }; +/* + * For any clocks that may differ per pipe + * only the max is stored in this structure + */ struct dc_clocks { int dispclk_khz; int max_supported_dppclk_khz; @@ -194,10 +207,11 @@ struct dc_clocks { int socclk_khz; int dcfclk_deep_sleep_khz; int fclk_khz; + int phyclk_khz; }; -struct dc_debug { - bool surface_visual_confirm; +struct dc_debug_options { + enum visual_confirm visual_confirm; bool sanity_checks; bool max_disp_clk; bool surface_trace; @@ -228,6 +242,7 @@ struct dc_debug { int urgent_latency_ns; int percent_of_ideal_drambw; int dram_clock_change_latency_ns; + bool optimized_watermark; int always_scale; bool disable_pplib_clock_request; bool disable_clock_gate; @@ -243,8 +258,19 @@ struct dc_debug { bool always_use_regamma; bool p010_mpo_support; bool recovery_enabled; + bool avoid_vbios_exec_table; + bool scl_reset_length10; + bool hdmi20_disable; + bool skip_detection_link_training; +}; +struct dc_debug_data { + uint32_t ltFailCount; + uint32_t i2cErrorCount; + uint32_t auxErrorCount; }; + + struct dc_state; struct resource_pool; struct dce_hwseq; @@ -253,8 +279,7 @@ struct dc { struct dc_caps caps; struct dc_cap_funcs cap_funcs; struct dc_config config; - struct dc_debug debug; - + struct dc_debug_options debug; struct dc_context *ctx; uint8_t link_count; @@ -269,7 +294,7 @@ struct dc { /* Inputs into BW and WM calculations. */ struct bw_calcs_dceip *bw_dceip; struct bw_calcs_vbios *bw_vbios; -#ifdef CONFIG_DRM_AMD_DC_DCN1_0 +#ifdef CONFIG_X86 struct dcn_soc_bounding_box *dcn_soc; struct dcn_ip_params *dcn_ip; struct display_mode_lib dml; @@ -289,9 +314,9 @@ struct dc { bool apply_edp_fast_boot_optimization; /* FBC compressor */ -#if defined(CONFIG_DRM_AMD_DC_FBC) struct compressor *fbc_compressor; -#endif + + struct dc_debug_data debug_data; }; enum frame_buffer_mode { @@ -359,6 +384,7 @@ enum dc_transfer_func_type { TF_TYPE_PREDEFINED, TF_TYPE_DISTRIBUTED_POINTS, TF_TYPE_BYPASS, + TF_TYPE_HWPWL }; struct dc_transfer_func_distributed_points { @@ -378,16 +404,22 @@ enum dc_transfer_func_predefined { TRANSFER_FUNCTION_PQ, TRANSFER_FUNCTION_LINEAR, TRANSFER_FUNCTION_UNITY, + TRANSFER_FUNCTION_HLG, + TRANSFER_FUNCTION_HLG12, + TRANSFER_FUNCTION_GAMMA22 }; struct dc_transfer_func { struct kref refcount; - struct dc_transfer_func_distributed_points tf_pts; enum dc_transfer_func_type type; enum dc_transfer_func_predefined tf; /* FP16 1.0 reference level in nits, default is 80 nits, only for PQ*/ uint32_t sdr_ref_white_level; struct dc_context *ctx; + union { + struct pwl_params pwl; + struct dc_transfer_func_distributed_points tf_pts; + }; }; /* @@ -617,9 +649,14 @@ struct dpcd_caps { struct dc_dongle_caps dongle_caps; uint32_t sink_dev_id; + int8_t sink_dev_id_str[6]; + int8_t sink_hw_revision; + int8_t sink_fw_revision[2]; + uint32_t branch_dev_id; int8_t branch_dev_name[6]; int8_t branch_hw_revision; + int8_t branch_fw_revision[2]; bool allow_invalid_MSA_timing_param; bool panel_mode_edp; @@ -662,9 +699,13 @@ struct dc_sink { struct dc_link *link; struct dc_context *ctx; + uint32_t sink_id; + /* private to dc_sink.c */ + // refcount must be the last member in dc_sink, since we want the + // sink structure to be logically cloneable up to (but not including) + // refcount struct kref refcount; - }; void dc_sink_retain(struct dc_sink *sink); diff --git a/drivers/gpu/drm/amd/display/dc/dc_bios_types.h b/drivers/gpu/drm/amd/display/dc/dc_bios_types.h index d9b84ec7954c..90082bab71f0 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_bios_types.h +++ b/drivers/gpu/drm/amd/display/dc/dc_bios_types.h @@ -198,6 +198,10 @@ struct dc_vbios_funcs { void (*post_init)(struct dc_bios *bios); void (*bios_parser_destroy)(struct dc_bios **dcb); + + enum bp_result (*get_board_layout_info)( + struct dc_bios *dcb, + struct board_layout_info *board_layout_info); }; struct bios_registers { diff --git a/drivers/gpu/drm/amd/display/dc/dc_ddc_types.h b/drivers/gpu/drm/amd/display/dc/dc_ddc_types.h index e1affeb5cc51..05c8c31d8b31 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_ddc_types.h +++ b/drivers/gpu/drm/amd/display/dc/dc_ddc_types.h @@ -25,6 +25,65 @@ #ifndef DC_DDC_TYPES_H_ #define DC_DDC_TYPES_H_ +enum aux_transaction_type { + AUX_TRANSACTION_TYPE_DP, + AUX_TRANSACTION_TYPE_I2C +}; + + +enum i2caux_transaction_action { + I2CAUX_TRANSACTION_ACTION_I2C_WRITE = 0x00, + I2CAUX_TRANSACTION_ACTION_I2C_READ = 0x10, + I2CAUX_TRANSACTION_ACTION_I2C_STATUS_REQUEST = 0x20, + + I2CAUX_TRANSACTION_ACTION_I2C_WRITE_MOT = 0x40, + I2CAUX_TRANSACTION_ACTION_I2C_READ_MOT = 0x50, + I2CAUX_TRANSACTION_ACTION_I2C_STATUS_REQUEST_MOT = 0x60, + + I2CAUX_TRANSACTION_ACTION_DP_WRITE = 0x80, + I2CAUX_TRANSACTION_ACTION_DP_READ = 0x90 +}; + +enum aux_channel_operation_result { + AUX_CHANNEL_OPERATION_SUCCEEDED, + AUX_CHANNEL_OPERATION_FAILED_REASON_UNKNOWN, + AUX_CHANNEL_OPERATION_FAILED_INVALID_REPLY, + AUX_CHANNEL_OPERATION_FAILED_TIMEOUT, + AUX_CHANNEL_OPERATION_FAILED_HPD_DISCON +}; + + +struct aux_request_transaction_data { + enum aux_transaction_type type; + enum i2caux_transaction_action action; + /* 20-bit AUX channel transaction address */ + uint32_t address; + /* delay, in 100-microsecond units */ + uint8_t delay; + uint32_t length; + uint8_t *data; +}; + +enum aux_transaction_reply { + AUX_TRANSACTION_REPLY_AUX_ACK = 0x00, + AUX_TRANSACTION_REPLY_AUX_NACK = 0x01, + AUX_TRANSACTION_REPLY_AUX_DEFER = 0x02, + + AUX_TRANSACTION_REPLY_I2C_ACK = 0x00, + AUX_TRANSACTION_REPLY_I2C_NACK = 0x10, + AUX_TRANSACTION_REPLY_I2C_DEFER = 0x20, + + AUX_TRANSACTION_REPLY_HPD_DISCON = 0x40, + + AUX_TRANSACTION_REPLY_INVALID = 0xFF +}; + +struct aux_reply_transaction_data { + enum aux_transaction_reply status; + uint32_t length; + uint8_t *data; +}; + struct i2c_payload { bool write; uint8_t address; @@ -109,7 +168,7 @@ struct ddc_service { uint32_t address; uint32_t edid_buf_len; - uint8_t edid_buf[MAX_EDID_BUFFER_SIZE]; + uint8_t edid_buf[DC_MAX_EDID_BUFFER_SIZE]; }; #endif /* DC_DDC_TYPES_H_ */ diff --git a/drivers/gpu/drm/amd/display/dc/dc_dp_types.h b/drivers/gpu/drm/amd/display/dc/dc_dp_types.h index 90bccd5ccaa2..da93ab43f2d8 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_dp_types.h +++ b/drivers/gpu/drm/amd/display/dc/dc_dp_types.h @@ -430,7 +430,7 @@ union test_request { struct { uint8_t LINK_TRAINING :1; uint8_t LINK_TEST_PATTRN :1; - uint8_t EDID_REAT :1; + uint8_t EDID_READ :1; uint8_t PHY_TEST_PATTERN :1; uint8_t AUDIO_TEST_PATTERN :1; uint8_t RESERVED :1; @@ -443,7 +443,8 @@ union test_response { struct { uint8_t ACK :1; uint8_t NO_ACK :1; - uint8_t RESERVED :6; + uint8_t EDID_CHECKSUM_WRITE:1; + uint8_t RESERVED :5; } bits; uint8_t raw; }; diff --git a/drivers/gpu/drm/amd/display/dc/dc_helper.c b/drivers/gpu/drm/amd/display/dc/dc_helper.c index bd0fda0ceb91..e68077e65565 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_helper.c +++ b/drivers/gpu/drm/amd/display/dc/dc_helper.c @@ -255,3 +255,54 @@ uint32_t generic_reg_wait(const struct dc_context *ctx, return reg_val; } + +void generic_write_indirect_reg(const struct dc_context *ctx, + uint32_t addr_index, uint32_t addr_data, + uint32_t index, uint32_t data) +{ + dm_write_reg(ctx, addr_index, index); + dm_write_reg(ctx, addr_data, data); +} + +uint32_t generic_read_indirect_reg(const struct dc_context *ctx, + uint32_t addr_index, uint32_t addr_data, + uint32_t index) +{ + uint32_t value = 0; + + dm_write_reg(ctx, addr_index, index); + value = dm_read_reg(ctx, addr_data); + + return value; +} + + +uint32_t generic_indirect_reg_update_ex(const struct dc_context *ctx, + uint32_t addr_index, uint32_t addr_data, + uint32_t index, uint32_t reg_val, int n, + uint8_t shift1, uint32_t mask1, uint32_t field_value1, + ...) +{ + uint32_t shift, mask, field_value; + int i = 1; + + va_list ap; + + va_start(ap, field_value1); + + reg_val = set_reg_field_value_ex(reg_val, field_value1, mask1, shift1); + + while (i < n) { + shift = va_arg(ap, uint32_t); + mask = va_arg(ap, uint32_t); + field_value = va_arg(ap, uint32_t); + + reg_val = set_reg_field_value_ex(reg_val, field_value, mask, shift); + i++; + } + + generic_write_indirect_reg(ctx, addr_index, addr_data, index, reg_val); + va_end(ap); + + return reg_val; +} diff --git a/drivers/gpu/drm/amd/display/dc/dc_hw_types.h b/drivers/gpu/drm/amd/display/dc/dc_hw_types.h index b1f70579d61b..b789cb2b354b 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_hw_types.h +++ b/drivers/gpu/drm/amd/display/dc/dc_hw_types.h @@ -192,13 +192,14 @@ enum surface_pixel_format { /*swaped & float*/ SURFACE_PIXEL_FORMAT_GRPH_ABGR16161616F, /*grow graphics here if necessary */ - + SURFACE_PIXEL_FORMAT_VIDEO_AYCrCb8888, SURFACE_PIXEL_FORMAT_VIDEO_BEGIN, SURFACE_PIXEL_FORMAT_VIDEO_420_YCbCr = SURFACE_PIXEL_FORMAT_VIDEO_BEGIN, SURFACE_PIXEL_FORMAT_VIDEO_420_YCrCb, SURFACE_PIXEL_FORMAT_VIDEO_420_10bpc_YCbCr, SURFACE_PIXEL_FORMAT_VIDEO_420_10bpc_YCrCb, + SURFACE_PIXEL_FORMAT_SUBSAMPLE_END, SURFACE_PIXEL_FORMAT_INVALID /*grow 444 video here if necessary */ @@ -403,9 +404,11 @@ struct dc_cursor_position { struct dc_cursor_mi_param { unsigned int pixel_clk_khz; unsigned int ref_clk_khz; - unsigned int viewport_x_start; - unsigned int viewport_width; + struct rect viewport; struct fixed31_32 h_scale_ratio; + struct fixed31_32 v_scale_ratio; + enum dc_rotation_angle rotation; + bool mirror; }; /* IPP related types */ @@ -414,6 +417,7 @@ enum { GAMMA_RGB_256_ENTRIES = 256, GAMMA_RGB_FLOAT_1024_ENTRIES = 1024, GAMMA_CS_TFM_1D_ENTRIES = 4096, + GAMMA_CUSTOM_ENTRIES = 4096, GAMMA_MAX_ENTRIES = 4096 }; @@ -421,6 +425,7 @@ enum dc_gamma_type { GAMMA_RGB_256 = 1, GAMMA_RGB_FLOAT_1024 = 2, GAMMA_CS_TFM_1D = 3, + GAMMA_CUSTOM = 4, }; struct dc_csc_transform { @@ -489,6 +494,7 @@ struct dc_cursor_attributes { uint32_t height; enum dc_cursor_color_format color_format; + uint32_t sdr_white_level; // for boosting (SDR) cursor in HDR mode /* In case we support HW Cursor rotation in the future */ enum dc_rotation_angle rotation_angle; @@ -496,6 +502,11 @@ struct dc_cursor_attributes { union dc_cursor_attribute_flags attribute_flags; }; +struct dpp_cursor_attributes { + int bias; + int scale; +}; + /* OPP */ enum dc_color_space { @@ -567,25 +578,25 @@ struct scaling_taps { }; enum dc_timing_standard { - TIMING_STANDARD_UNDEFINED, - TIMING_STANDARD_DMT, - TIMING_STANDARD_GTF, - TIMING_STANDARD_CVT, - TIMING_STANDARD_CVT_RB, - TIMING_STANDARD_CEA770, - TIMING_STANDARD_CEA861, - TIMING_STANDARD_HDMI, - TIMING_STANDARD_TV_NTSC, - TIMING_STANDARD_TV_NTSC_J, - TIMING_STANDARD_TV_PAL, - TIMING_STANDARD_TV_PAL_M, - TIMING_STANDARD_TV_PAL_CN, - TIMING_STANDARD_TV_SECAM, - TIMING_STANDARD_EXPLICIT, + DC_TIMING_STANDARD_UNDEFINED, + DC_TIMING_STANDARD_DMT, + DC_TIMING_STANDARD_GTF, + DC_TIMING_STANDARD_CVT, + DC_TIMING_STANDARD_CVT_RB, + DC_TIMING_STANDARD_CEA770, + DC_TIMING_STANDARD_CEA861, + DC_TIMING_STANDARD_HDMI, + DC_TIMING_STANDARD_TV_NTSC, + DC_TIMING_STANDARD_TV_NTSC_J, + DC_TIMING_STANDARD_TV_PAL, + DC_TIMING_STANDARD_TV_PAL_M, + DC_TIMING_STANDARD_TV_PAL_CN, + DC_TIMING_STANDARD_TV_SECAM, + DC_TIMING_STANDARD_EXPLICIT, /*!< For explicit timings from EDID, VBIOS, etc.*/ - TIMING_STANDARD_USER_OVERRIDE, + DC_TIMING_STANDARD_USER_OVERRIDE, /*!< For mode timing override by user*/ - TIMING_STANDARD_MAX + DC_TIMING_STANDARD_MAX }; enum dc_color_depth { diff --git a/drivers/gpu/drm/amd/display/dc/dc_link.h b/drivers/gpu/drm/amd/display/dc/dc_link.h index 8a716baa1203..d43cefbc43d3 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_link.h +++ b/drivers/gpu/drm/amd/display/dc/dc_link.h @@ -73,6 +73,7 @@ struct dc_link { enum dc_irq_source irq_source_hpd; enum dc_irq_source irq_source_hpd_rx;/* aka DP Short Pulse */ bool is_hpd_filter_disabled; + bool dp_ss_off; /* caps is the same as reported_link_cap. link_traing use * reported_link_cap. Will clean up. TODO @@ -141,6 +142,8 @@ static inline struct dc_link *dc_get_link_at_index(struct dc *dc, uint32_t link_ bool dc_link_set_backlight_level(const struct dc_link *dc_link, uint32_t level, uint32_t frame_ramp, const struct dc_stream_state *stream); +int dc_link_get_backlight_level(const struct dc_link *dc_link); + bool dc_link_set_abm_disable(const struct dc_link *dc_link); bool dc_link_set_psr_enable(const struct dc_link *dc_link, bool enable, bool wait); @@ -172,7 +175,7 @@ bool dc_link_detect(struct dc_link *dc_link, enum dc_detect_reason reason); * false - no change in Downstream port status. No further action required * from DM. */ bool dc_link_handle_hpd_rx_irq(struct dc_link *dc_link, - union hpd_irq_data *hpd_irq_dpcd_data); + union hpd_irq_data *hpd_irq_dpcd_data, bool *out_link_loss); struct dc_sink_init_data; @@ -210,10 +213,29 @@ bool dc_link_dp_set_test_pattern( void dc_link_enable_hpd_filter(struct dc_link *link, bool enable); +bool dc_link_is_dp_sink_present(struct dc_link *link); + /* * DPCD access interfaces */ +void dc_link_set_drive_settings(struct dc *dc, + struct link_training_settings *lt_settings, + const struct dc_link *link); +void dc_link_perform_link_training(struct dc *dc, + struct dc_link_settings *link_setting, + bool skip_video_pattern); +void dc_link_set_preferred_link_settings(struct dc *dc, + struct dc_link_settings *link_setting, + struct dc_link *link); +void dc_link_enable_hpd(const struct dc_link *link); +void dc_link_disable_hpd(const struct dc_link *link); +void dc_link_set_test_pattern(struct dc_link *link, + enum dp_test_pattern test_pattern, + const struct link_training_settings *p_link_settings, + const unsigned char *p_custom_pattern, + unsigned int cust_pattern_size); + bool dc_submit_i2c( struct dc *dc, uint32_t link_index, diff --git a/drivers/gpu/drm/amd/display/dc/dc_stream.h b/drivers/gpu/drm/amd/display/dc/dc_stream.h index d7e6d53bb383..cbfe418006cb 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_stream.h +++ b/drivers/gpu/drm/amd/display/dc/dc_stream.h @@ -59,6 +59,9 @@ struct dc_stream_state { struct freesync_context freesync_ctx; struct dc_info_packet hdr_static_metadata; + PHYSICAL_ADDRESS_LOC dmdata_address; + bool use_dynamic_meta; + struct dc_transfer_func *out_transfer_func; struct colorspace_transform gamut_remap_matrix; struct dc_csc_transform csc_color_matrix; @@ -97,6 +100,7 @@ struct dc_stream_state { struct dc_cursor_attributes cursor_attributes; struct dc_cursor_position cursor_position; + uint32_t sdr_white_level; // for boosting (SDR) cursor in HDR mode /* from stream struct */ struct kref refcount; @@ -144,10 +148,7 @@ void dc_commit_updates_for_stream(struct dc *dc, /* * Log the current stream state. */ -void dc_stream_log( - const struct dc_stream_state *stream, - struct dal_logger *dc_logger, - enum dc_log_type log_type); +void dc_stream_log(const struct dc *dc, const struct dc_stream_state *stream); uint8_t dc_get_current_stream_count(struct dc *dc); struct dc_stream_state *dc_get_stream_at_index(struct dc *dc, uint8_t i); @@ -255,6 +256,7 @@ bool dc_stream_set_cursor_position( struct dc_stream_state *stream, const struct dc_cursor_position *position); + bool dc_stream_adjust_vmin_vmax(struct dc *dc, struct dc_stream_state **stream, int num_streams, @@ -299,9 +301,4 @@ bool dc_stream_get_crtc_position(struct dc *dc, unsigned int *v_pos, unsigned int *nom_v_pos); -void dc_stream_set_static_screen_events(struct dc *dc, - struct dc_stream_state **stream, - int num_streams, - const struct dc_static_screen_events *events); - #endif /* DC_STREAM_H_ */ diff --git a/drivers/gpu/drm/amd/display/dc/dc_types.h b/drivers/gpu/drm/amd/display/dc/dc_types.h index 76df2534c4a4..8c6eb78b0c3b 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_types.h +++ b/drivers/gpu/drm/amd/display/dc/dc_types.h @@ -77,8 +77,6 @@ struct dc_context { struct dc *dc; void *driver_context; /* e.g. amdgpu_device */ - - struct dal_logger *logger; void *cgs_device; enum dce_environment dce_environment; @@ -92,13 +90,12 @@ struct dc_context { bool created_bios; struct gpio_service *gpio_service; struct i2caux *i2caux; -#if defined(CONFIG_DRM_AMD_DC_FBC) + uint32_t dc_sink_id_count; uint64_t fbc_gpu_addr; -#endif }; -#define MAX_EDID_BUFFER_SIZE 512 +#define DC_MAX_EDID_BUFFER_SIZE 512 #define EDID_BLOCK_SIZE 128 #define MAX_SURFACE_NUM 4 #define NUM_PIXEL_FORMATS 10 @@ -137,13 +134,13 @@ enum plane_stereo_format { */ enum dc_edid_connector_type { - EDID_CONNECTOR_UNKNOWN = 0, - EDID_CONNECTOR_ANALOG = 1, - EDID_CONNECTOR_DIGITAL = 10, - EDID_CONNECTOR_DVI = 11, - EDID_CONNECTOR_HDMIA = 12, - EDID_CONNECTOR_MDDI = 14, - EDID_CONNECTOR_DISPLAYPORT = 15 + DC_EDID_CONNECTOR_UNKNOWN = 0, + DC_EDID_CONNECTOR_ANALOG = 1, + DC_EDID_CONNECTOR_DIGITAL = 10, + DC_EDID_CONNECTOR_DVI = 11, + DC_EDID_CONNECTOR_HDMIA = 12, + DC_EDID_CONNECTOR_MDDI = 14, + DC_EDID_CONNECTOR_DISPLAYPORT = 15 }; enum dc_edid_status { @@ -169,7 +166,7 @@ struct dc_cea_audio_mode { struct dc_edid { uint32_t length; - uint8_t raw_edid[MAX_EDID_BUFFER_SIZE]; + uint8_t raw_edid[DC_MAX_EDID_BUFFER_SIZE]; }; /* When speaker location data block is not available, DEFAULT_SPEAKER_LOCATION @@ -195,6 +192,7 @@ union display_content_support { struct dc_panel_patch { unsigned int dppowerup_delay; + unsigned int extra_t12_ms; }; struct dc_edid_caps { diff --git a/drivers/gpu/drm/amd/display/dc/dce/Makefile b/drivers/gpu/drm/amd/display/dc/dce/Makefile index 11401fd8e535..825537bd4545 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/Makefile +++ b/drivers/gpu/drm/amd/display/dc/dce/Makefile @@ -28,7 +28,7 @@ DCE = dce_audio.o dce_stream_encoder.o dce_link_encoder.o dce_hwseq.o \ dce_mem_input.o dce_clock_source.o dce_scl_filters.o dce_transform.o \ -dce_clocks.o dce_opp.o dce_dmcu.o dce_abm.o dce_ipp.o +dce_clocks.o dce_opp.o dce_dmcu.o dce_abm.o dce_ipp.o dce_aux.o AMD_DAL_DCE = $(addprefix $(AMDDALPATH)/dc/dce/,$(DCE)) diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_aux.c b/drivers/gpu/drm/amd/display/dc/dce/dce_aux.c new file mode 100644 index 000000000000..3f5b2e6f7553 --- /dev/null +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_aux.c @@ -0,0 +1,937 @@ +/* + * Copyright 2012-15 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: AMD + * + */ + +#include "dm_services.h" +#include "dce_aux.h" +#include "dce/dce_11_0_sh_mask.h" + +#define CTX \ + aux110->base.ctx +#define REG(reg_name)\ + (aux110->regs->reg_name) + +#define DC_LOGGER \ + engine->ctx->logger + +#include "reg_helper.h" + +#define FROM_AUX_ENGINE(ptr) \ + container_of((ptr), struct aux_engine_dce110, base) + +#define FROM_ENGINE(ptr) \ + FROM_AUX_ENGINE(container_of((ptr), struct aux_engine, base)) + +#define FROM_AUX_ENGINE_ENGINE(ptr) \ + container_of((ptr), struct aux_engine, base) +enum { + AUX_INVALID_REPLY_RETRY_COUNTER = 1, + AUX_TIMED_OUT_RETRY_COUNTER = 2, + AUX_DEFER_RETRY_COUNTER = 6 +}; +static void release_engine( + struct aux_engine *engine) +{ + struct aux_engine_dce110 *aux110 = FROM_AUX_ENGINE(engine); + + dal_ddc_close(engine->ddc); + + engine->ddc = NULL; + + REG_UPDATE(AUX_ARB_CONTROL, AUX_SW_DONE_USING_AUX_REG, 1); +} + +#define SW_CAN_ACCESS_AUX 1 +#define DMCU_CAN_ACCESS_AUX 2 + +static bool is_engine_available( + struct aux_engine *engine) +{ + struct aux_engine_dce110 *aux110 = FROM_AUX_ENGINE(engine); + + uint32_t value = REG_READ(AUX_ARB_CONTROL); + uint32_t field = get_reg_field_value( + value, + AUX_ARB_CONTROL, + AUX_REG_RW_CNTL_STATUS); + + return (field != DMCU_CAN_ACCESS_AUX); +} +static bool acquire_engine( + struct aux_engine *engine) +{ + struct aux_engine_dce110 *aux110 = FROM_AUX_ENGINE(engine); + + uint32_t value = REG_READ(AUX_ARB_CONTROL); + uint32_t field = get_reg_field_value( + value, + AUX_ARB_CONTROL, + AUX_REG_RW_CNTL_STATUS); + if (field == DMCU_CAN_ACCESS_AUX) + return false; + /* enable AUX before request SW to access AUX */ + value = REG_READ(AUX_CONTROL); + field = get_reg_field_value(value, + AUX_CONTROL, + AUX_EN); + + if (field == 0) { + set_reg_field_value( + value, + 1, + AUX_CONTROL, + AUX_EN); + + if (REG(AUX_RESET_MASK)) { + /*DP_AUX block as part of the enable sequence*/ + set_reg_field_value( + value, + 1, + AUX_CONTROL, + AUX_RESET); + } + + REG_WRITE(AUX_CONTROL, value); + + if (REG(AUX_RESET_MASK)) { + /*poll HW to make sure reset it done*/ + + REG_WAIT(AUX_CONTROL, AUX_RESET_DONE, 1, + 1, 11); + + set_reg_field_value( + value, + 0, + AUX_CONTROL, + AUX_RESET); + + REG_WRITE(AUX_CONTROL, value); + + REG_WAIT(AUX_CONTROL, AUX_RESET_DONE, 0, + 1, 11); + } + } /*if (field)*/ + + /* request SW to access AUX */ + REG_UPDATE(AUX_ARB_CONTROL, AUX_SW_USE_AUX_REG_REQ, 1); + + value = REG_READ(AUX_ARB_CONTROL); + field = get_reg_field_value( + value, + AUX_ARB_CONTROL, + AUX_REG_RW_CNTL_STATUS); + + return (field == SW_CAN_ACCESS_AUX); +} + +#define COMPOSE_AUX_SW_DATA_16_20(command, address) \ + ((command) | ((0xF0000 & (address)) >> 16)) + +#define COMPOSE_AUX_SW_DATA_8_15(address) \ + ((0xFF00 & (address)) >> 8) + +#define COMPOSE_AUX_SW_DATA_0_7(address) \ + (0xFF & (address)) + +static void submit_channel_request( + struct aux_engine *engine, + struct aux_request_transaction_data *request) +{ + struct aux_engine_dce110 *aux110 = FROM_AUX_ENGINE(engine); + uint32_t value; + uint32_t length; + + bool is_write = + ((request->type == AUX_TRANSACTION_TYPE_DP) && + (request->action == I2CAUX_TRANSACTION_ACTION_DP_WRITE)) || + ((request->type == AUX_TRANSACTION_TYPE_I2C) && + ((request->action == I2CAUX_TRANSACTION_ACTION_I2C_WRITE) || + (request->action == I2CAUX_TRANSACTION_ACTION_I2C_WRITE_MOT))); + if (REG(AUXN_IMPCAL)) { + /* clear_aux_error */ + REG_UPDATE_SEQ(AUXN_IMPCAL, AUXN_CALOUT_ERROR_AK, + 1, + 0); + + REG_UPDATE_SEQ(AUXP_IMPCAL, AUXP_CALOUT_ERROR_AK, + 1, + 0); + + /* force_default_calibrate */ + REG_UPDATE_1BY1_2(AUXN_IMPCAL, + AUXN_IMPCAL_ENABLE, 1, + AUXN_IMPCAL_OVERRIDE_ENABLE, 0); + + /* bug? why AUXN update EN and OVERRIDE_EN 1 by 1 while AUX P toggles OVERRIDE? */ + + REG_UPDATE_SEQ(AUXP_IMPCAL, AUXP_IMPCAL_OVERRIDE_ENABLE, + 1, + 0); + } + /* set the delay and the number of bytes to write */ + + /* The length include + * the 4 bit header and the 20 bit address + * (that is 3 byte). + * If the requested length is non zero this means + * an addition byte specifying the length is required. + */ + + length = request->length ? 4 : 3; + if (is_write) + length += request->length; + + REG_UPDATE_2(AUX_SW_CONTROL, + AUX_SW_START_DELAY, request->delay, + AUX_SW_WR_BYTES, length); + + /* program action and address and payload data (if 'is_write') */ + value = REG_UPDATE_4(AUX_SW_DATA, + AUX_SW_INDEX, 0, + AUX_SW_DATA_RW, 0, + AUX_SW_AUTOINCREMENT_DISABLE, 1, + AUX_SW_DATA, COMPOSE_AUX_SW_DATA_16_20(request->action, request->address)); + + value = REG_SET_2(AUX_SW_DATA, value, + AUX_SW_AUTOINCREMENT_DISABLE, 0, + AUX_SW_DATA, COMPOSE_AUX_SW_DATA_8_15(request->address)); + + value = REG_SET(AUX_SW_DATA, value, + AUX_SW_DATA, COMPOSE_AUX_SW_DATA_0_7(request->address)); + + if (request->length) { + value = REG_SET(AUX_SW_DATA, value, + AUX_SW_DATA, request->length - 1); + } + + if (is_write) { + /* Load the HW buffer with the Data to be sent. + * This is relevant for write operation. + * For read, the data recived data will be + * processed in process_channel_reply(). + */ + uint32_t i = 0; + + while (i < request->length) { + value = REG_SET(AUX_SW_DATA, value, + AUX_SW_DATA, request->data[i]); + + ++i; + } + } + + REG_UPDATE(AUX_INTERRUPT_CONTROL, AUX_SW_DONE_ACK, 1); + REG_WAIT(AUX_SW_STATUS, AUX_SW_DONE, 0, + 10, aux110->timeout_period/10); + REG_UPDATE(AUX_SW_CONTROL, AUX_SW_GO, 1); +} + +static int read_channel_reply(struct aux_engine *engine, uint32_t size, + uint8_t *buffer, uint8_t *reply_result, + uint32_t *sw_status) +{ + struct aux_engine_dce110 *aux110 = FROM_AUX_ENGINE(engine); + uint32_t bytes_replied; + uint32_t reply_result_32; + + *sw_status = REG_GET(AUX_SW_STATUS, AUX_SW_REPLY_BYTE_COUNT, + &bytes_replied); + + /* In case HPD is LOW, exit AUX transaction */ + if ((*sw_status & AUX_SW_STATUS__AUX_SW_HPD_DISCON_MASK)) + return -1; + + /* Need at least the status byte */ + if (!bytes_replied) + return -1; + + REG_UPDATE_1BY1_3(AUX_SW_DATA, + AUX_SW_INDEX, 0, + AUX_SW_AUTOINCREMENT_DISABLE, 1, + AUX_SW_DATA_RW, 1); + + REG_GET(AUX_SW_DATA, AUX_SW_DATA, &reply_result_32); + reply_result_32 = reply_result_32 >> 4; + *reply_result = (uint8_t)reply_result_32; + + if (reply_result_32 == 0) { /* ACK */ + uint32_t i = 0; + + /* First byte was already used to get the command status */ + --bytes_replied; + + /* Do not overflow buffer */ + if (bytes_replied > size) + return -1; + + while (i < bytes_replied) { + uint32_t aux_sw_data_val; + + REG_GET(AUX_SW_DATA, AUX_SW_DATA, &aux_sw_data_val); + buffer[i] = aux_sw_data_val; + ++i; + } + + return i; + } + + return 0; +} + +static void process_channel_reply( + struct aux_engine *engine, + struct aux_reply_transaction_data *reply) +{ + int bytes_replied; + uint8_t reply_result; + uint32_t sw_status; + + bytes_replied = read_channel_reply(engine, reply->length, reply->data, + &reply_result, &sw_status); + + /* in case HPD is LOW, exit AUX transaction */ + if ((sw_status & AUX_SW_STATUS__AUX_SW_HPD_DISCON_MASK)) { + reply->status = AUX_CHANNEL_OPERATION_FAILED_HPD_DISCON; + return; + } + + if (bytes_replied < 0) { + /* Need to handle an error case... + * Hopefully, upper layer function won't call this function if + * the number of bytes in the reply was 0, because there was + * surely an error that was asserted that should have been + * handled for hot plug case, this could happens + */ + if (!(sw_status & AUX_SW_STATUS__AUX_SW_HPD_DISCON_MASK)) { + reply->status = AUX_TRANSACTION_REPLY_INVALID; + ASSERT_CRITICAL(false); + return; + } + } else { + + switch (reply_result) { + case 0: /* ACK */ + reply->status = AUX_TRANSACTION_REPLY_AUX_ACK; + break; + case 1: /* NACK */ + reply->status = AUX_TRANSACTION_REPLY_AUX_NACK; + break; + case 2: /* DEFER */ + reply->status = AUX_TRANSACTION_REPLY_AUX_DEFER; + break; + case 4: /* AUX ACK / I2C NACK */ + reply->status = AUX_TRANSACTION_REPLY_I2C_NACK; + break; + case 8: /* AUX ACK / I2C DEFER */ + reply->status = AUX_TRANSACTION_REPLY_I2C_DEFER; + break; + default: + reply->status = AUX_TRANSACTION_REPLY_INVALID; + } + } +} + +static enum aux_channel_operation_result get_channel_status( + struct aux_engine *engine, + uint8_t *returned_bytes) +{ + struct aux_engine_dce110 *aux110 = FROM_AUX_ENGINE(engine); + + uint32_t value; + + if (returned_bytes == NULL) { + /*caller pass NULL pointer*/ + ASSERT_CRITICAL(false); + return AUX_CHANNEL_OPERATION_FAILED_REASON_UNKNOWN; + } + *returned_bytes = 0; + + /* poll to make sure that SW_DONE is asserted */ + value = REG_WAIT(AUX_SW_STATUS, AUX_SW_DONE, 1, + 10, aux110->timeout_period/10); + + /* in case HPD is LOW, exit AUX transaction */ + if ((value & AUX_SW_STATUS__AUX_SW_HPD_DISCON_MASK)) + return AUX_CHANNEL_OPERATION_FAILED_HPD_DISCON; + + /* Note that the following bits are set in 'status.bits' + * during CTS 4.2.1.2 (FW 3.3.1): + * AUX_SW_RX_MIN_COUNT_VIOL, AUX_SW_RX_INVALID_STOP, + * AUX_SW_RX_RECV_NO_DET, AUX_SW_RX_RECV_INVALID_H. + * + * AUX_SW_RX_MIN_COUNT_VIOL is an internal, + * HW debugging bit and should be ignored. + */ + if (value & AUX_SW_STATUS__AUX_SW_DONE_MASK) { + if ((value & AUX_SW_STATUS__AUX_SW_RX_TIMEOUT_STATE_MASK) || + (value & AUX_SW_STATUS__AUX_SW_RX_TIMEOUT_MASK)) + return AUX_CHANNEL_OPERATION_FAILED_TIMEOUT; + + else if ((value & AUX_SW_STATUS__AUX_SW_RX_INVALID_STOP_MASK) || + (value & AUX_SW_STATUS__AUX_SW_RX_RECV_NO_DET_MASK) || + (value & + AUX_SW_STATUS__AUX_SW_RX_RECV_INVALID_H_MASK) || + (value & AUX_SW_STATUS__AUX_SW_RX_RECV_INVALID_L_MASK)) + return AUX_CHANNEL_OPERATION_FAILED_INVALID_REPLY; + + *returned_bytes = get_reg_field_value(value, + AUX_SW_STATUS, + AUX_SW_REPLY_BYTE_COUNT); + + if (*returned_bytes == 0) + return + AUX_CHANNEL_OPERATION_FAILED_INVALID_REPLY; + else { + *returned_bytes -= 1; + return AUX_CHANNEL_OPERATION_SUCCEEDED; + } + } else { + /*time_elapsed >= aux_engine->timeout_period + * AUX_SW_STATUS__AUX_SW_HPD_DISCON = at this point + */ + ASSERT_CRITICAL(false); + return AUX_CHANNEL_OPERATION_FAILED_TIMEOUT; + } +} +static void process_read_reply( + struct aux_engine *engine, + struct read_command_context *ctx) +{ + engine->funcs->process_channel_reply(engine, &ctx->reply); + + switch (ctx->reply.status) { + case AUX_TRANSACTION_REPLY_AUX_ACK: + ctx->defer_retry_aux = 0; + if (ctx->returned_byte > ctx->current_read_length) { + ctx->status = + I2CAUX_TRANSACTION_STATUS_FAILED_PROTOCOL_ERROR; + ctx->operation_succeeded = false; + } else if (ctx->returned_byte < ctx->current_read_length) { + ctx->current_read_length -= ctx->returned_byte; + + ctx->offset += ctx->returned_byte; + + ++ctx->invalid_reply_retry_aux_on_ack; + + if (ctx->invalid_reply_retry_aux_on_ack > + AUX_INVALID_REPLY_RETRY_COUNTER) { + ctx->status = + I2CAUX_TRANSACTION_STATUS_FAILED_PROTOCOL_ERROR; + ctx->operation_succeeded = false; + } + } else { + ctx->status = I2CAUX_TRANSACTION_STATUS_SUCCEEDED; + ctx->transaction_complete = true; + ctx->operation_succeeded = true; + } + break; + case AUX_TRANSACTION_REPLY_AUX_NACK: + ctx->status = I2CAUX_TRANSACTION_STATUS_FAILED_NACK; + ctx->operation_succeeded = false; + break; + case AUX_TRANSACTION_REPLY_AUX_DEFER: + ++ctx->defer_retry_aux; + + if (ctx->defer_retry_aux > AUX_DEFER_RETRY_COUNTER) { + ctx->status = I2CAUX_TRANSACTION_STATUS_FAILED_TIMEOUT; + ctx->operation_succeeded = false; + } + break; + case AUX_TRANSACTION_REPLY_I2C_DEFER: + ctx->defer_retry_aux = 0; + + ++ctx->defer_retry_i2c; + + if (ctx->defer_retry_i2c > AUX_DEFER_RETRY_COUNTER) { + ctx->status = I2CAUX_TRANSACTION_STATUS_FAILED_TIMEOUT; + ctx->operation_succeeded = false; + } + break; + case AUX_TRANSACTION_REPLY_HPD_DISCON: + ctx->status = I2CAUX_TRANSACTION_STATUS_FAILED_HPD_DISCON; + ctx->operation_succeeded = false; + break; + default: + ctx->status = I2CAUX_TRANSACTION_STATUS_UNKNOWN; + ctx->operation_succeeded = false; + } +} +static void process_read_request( + struct aux_engine *engine, + struct read_command_context *ctx) +{ + enum aux_channel_operation_result operation_result; + + engine->funcs->submit_channel_request(engine, &ctx->request); + + operation_result = engine->funcs->get_channel_status( + engine, &ctx->returned_byte); + + switch (operation_result) { + case AUX_CHANNEL_OPERATION_SUCCEEDED: + if (ctx->returned_byte > ctx->current_read_length) { + ctx->status = + I2CAUX_TRANSACTION_STATUS_FAILED_PROTOCOL_ERROR; + ctx->operation_succeeded = false; + } else { + ctx->timed_out_retry_aux = 0; + ctx->invalid_reply_retry_aux = 0; + + ctx->reply.length = ctx->returned_byte; + ctx->reply.data = ctx->buffer; + + process_read_reply(engine, ctx); + } + break; + case AUX_CHANNEL_OPERATION_FAILED_INVALID_REPLY: + ++ctx->invalid_reply_retry_aux; + + if (ctx->invalid_reply_retry_aux > + AUX_INVALID_REPLY_RETRY_COUNTER) { + ctx->status = + I2CAUX_TRANSACTION_STATUS_FAILED_PROTOCOL_ERROR; + ctx->operation_succeeded = false; + } else + udelay(400); + break; + case AUX_CHANNEL_OPERATION_FAILED_TIMEOUT: + ++ctx->timed_out_retry_aux; + + if (ctx->timed_out_retry_aux > AUX_TIMED_OUT_RETRY_COUNTER) { + ctx->status = I2CAUX_TRANSACTION_STATUS_FAILED_TIMEOUT; + ctx->operation_succeeded = false; + } else { + /* DP 1.2a, table 2-58: + * "S3: AUX Request CMD PENDING: + * retry 3 times, with 400usec wait on each" + * The HW timeout is set to 550usec, + * so we should not wait here + */ + } + break; + case AUX_CHANNEL_OPERATION_FAILED_HPD_DISCON: + ctx->status = I2CAUX_TRANSACTION_STATUS_FAILED_HPD_DISCON; + ctx->operation_succeeded = false; + break; + default: + ctx->status = I2CAUX_TRANSACTION_STATUS_UNKNOWN; + ctx->operation_succeeded = false; + } +} +static bool read_command( + struct aux_engine *engine, + struct i2caux_transaction_request *request, + bool middle_of_transaction) +{ + struct read_command_context ctx; + + ctx.buffer = request->payload.data; + ctx.current_read_length = request->payload.length; + ctx.offset = 0; + ctx.timed_out_retry_aux = 0; + ctx.invalid_reply_retry_aux = 0; + ctx.defer_retry_aux = 0; + ctx.defer_retry_i2c = 0; + ctx.invalid_reply_retry_aux_on_ack = 0; + ctx.transaction_complete = false; + ctx.operation_succeeded = true; + + if (request->payload.address_space == + I2CAUX_TRANSACTION_ADDRESS_SPACE_DPCD) { + ctx.request.type = AUX_TRANSACTION_TYPE_DP; + ctx.request.action = I2CAUX_TRANSACTION_ACTION_DP_READ; + ctx.request.address = request->payload.address; + } else if (request->payload.address_space == + I2CAUX_TRANSACTION_ADDRESS_SPACE_I2C) { + ctx.request.type = AUX_TRANSACTION_TYPE_I2C; + ctx.request.action = middle_of_transaction ? + I2CAUX_TRANSACTION_ACTION_I2C_READ_MOT : + I2CAUX_TRANSACTION_ACTION_I2C_READ; + ctx.request.address = request->payload.address >> 1; + } else { + /* in DAL2, there was no return in such case */ + BREAK_TO_DEBUGGER(); + return false; + } + + ctx.request.delay = 0; + + do { + memset(ctx.buffer + ctx.offset, 0, ctx.current_read_length); + + ctx.request.data = ctx.buffer + ctx.offset; + ctx.request.length = ctx.current_read_length; + + process_read_request(engine, &ctx); + + request->status = ctx.status; + + if (ctx.operation_succeeded && !ctx.transaction_complete) + if (ctx.request.type == AUX_TRANSACTION_TYPE_I2C) + msleep(engine->delay); + } while (ctx.operation_succeeded && !ctx.transaction_complete); + + if (request->payload.address_space == + I2CAUX_TRANSACTION_ADDRESS_SPACE_DPCD) { + DC_LOG_I2C_AUX("READ: addr:0x%x value:0x%x Result:%d", + request->payload.address, + request->payload.data[0], + ctx.operation_succeeded); + } + + return ctx.operation_succeeded; +} + +static void process_write_reply( + struct aux_engine *engine, + struct write_command_context *ctx) +{ + engine->funcs->process_channel_reply(engine, &ctx->reply); + + switch (ctx->reply.status) { + case AUX_TRANSACTION_REPLY_AUX_ACK: + ctx->operation_succeeded = true; + + if (ctx->returned_byte) { + ctx->request.action = ctx->mot ? + I2CAUX_TRANSACTION_ACTION_I2C_STATUS_REQUEST_MOT : + I2CAUX_TRANSACTION_ACTION_I2C_STATUS_REQUEST; + + ctx->current_write_length = 0; + + ++ctx->ack_m_retry; + + if (ctx->ack_m_retry > AUX_DEFER_RETRY_COUNTER) { + ctx->status = + I2CAUX_TRANSACTION_STATUS_FAILED_TIMEOUT; + ctx->operation_succeeded = false; + } else + udelay(300); + } else { + ctx->status = I2CAUX_TRANSACTION_STATUS_SUCCEEDED; + ctx->defer_retry_aux = 0; + ctx->ack_m_retry = 0; + ctx->transaction_complete = true; + } + break; + case AUX_TRANSACTION_REPLY_AUX_NACK: + ctx->status = I2CAUX_TRANSACTION_STATUS_FAILED_NACK; + ctx->operation_succeeded = false; + break; + case AUX_TRANSACTION_REPLY_AUX_DEFER: + ++ctx->defer_retry_aux; + + if (ctx->defer_retry_aux > ctx->max_defer_retry) { + ctx->status = I2CAUX_TRANSACTION_STATUS_FAILED_TIMEOUT; + ctx->operation_succeeded = false; + } + break; + case AUX_TRANSACTION_REPLY_I2C_DEFER: + ctx->defer_retry_aux = 0; + ctx->current_write_length = 0; + + ctx->request.action = ctx->mot ? + I2CAUX_TRANSACTION_ACTION_I2C_STATUS_REQUEST_MOT : + I2CAUX_TRANSACTION_ACTION_I2C_STATUS_REQUEST; + + ++ctx->defer_retry_i2c; + + if (ctx->defer_retry_i2c > ctx->max_defer_retry) { + ctx->status = I2CAUX_TRANSACTION_STATUS_FAILED_TIMEOUT; + ctx->operation_succeeded = false; + } + break; + case AUX_TRANSACTION_REPLY_HPD_DISCON: + ctx->status = I2CAUX_TRANSACTION_STATUS_FAILED_HPD_DISCON; + ctx->operation_succeeded = false; + break; + default: + ctx->status = I2CAUX_TRANSACTION_STATUS_UNKNOWN; + ctx->operation_succeeded = false; + } +} +static void process_write_request( + struct aux_engine *engine, + struct write_command_context *ctx) +{ + enum aux_channel_operation_result operation_result; + + engine->funcs->submit_channel_request(engine, &ctx->request); + + operation_result = engine->funcs->get_channel_status( + engine, &ctx->returned_byte); + + switch (operation_result) { + case AUX_CHANNEL_OPERATION_SUCCEEDED: + ctx->timed_out_retry_aux = 0; + ctx->invalid_reply_retry_aux = 0; + + ctx->reply.length = ctx->returned_byte; + ctx->reply.data = ctx->reply_data; + + process_write_reply(engine, ctx); + break; + case AUX_CHANNEL_OPERATION_FAILED_INVALID_REPLY: + ++ctx->invalid_reply_retry_aux; + + if (ctx->invalid_reply_retry_aux > + AUX_INVALID_REPLY_RETRY_COUNTER) { + ctx->status = + I2CAUX_TRANSACTION_STATUS_FAILED_PROTOCOL_ERROR; + ctx->operation_succeeded = false; + } else + udelay(400); + break; + case AUX_CHANNEL_OPERATION_FAILED_TIMEOUT: + ++ctx->timed_out_retry_aux; + + if (ctx->timed_out_retry_aux > AUX_TIMED_OUT_RETRY_COUNTER) { + ctx->status = I2CAUX_TRANSACTION_STATUS_FAILED_TIMEOUT; + ctx->operation_succeeded = false; + } else { + /* DP 1.2a, table 2-58: + * "S3: AUX Request CMD PENDING: + * retry 3 times, with 400usec wait on each" + * The HW timeout is set to 550usec, + * so we should not wait here + */ + } + break; + case AUX_CHANNEL_OPERATION_FAILED_HPD_DISCON: + ctx->status = I2CAUX_TRANSACTION_STATUS_FAILED_HPD_DISCON; + ctx->operation_succeeded = false; + break; + default: + ctx->status = I2CAUX_TRANSACTION_STATUS_UNKNOWN; + ctx->operation_succeeded = false; + } +} +static bool write_command( + struct aux_engine *engine, + struct i2caux_transaction_request *request, + bool middle_of_transaction) +{ + struct write_command_context ctx; + + ctx.mot = middle_of_transaction; + ctx.buffer = request->payload.data; + ctx.current_write_length = request->payload.length; + ctx.timed_out_retry_aux = 0; + ctx.invalid_reply_retry_aux = 0; + ctx.defer_retry_aux = 0; + ctx.defer_retry_i2c = 0; + ctx.ack_m_retry = 0; + ctx.transaction_complete = false; + ctx.operation_succeeded = true; + + if (request->payload.address_space == + I2CAUX_TRANSACTION_ADDRESS_SPACE_DPCD) { + ctx.request.type = AUX_TRANSACTION_TYPE_DP; + ctx.request.action = I2CAUX_TRANSACTION_ACTION_DP_WRITE; + ctx.request.address = request->payload.address; + } else if (request->payload.address_space == + I2CAUX_TRANSACTION_ADDRESS_SPACE_I2C) { + ctx.request.type = AUX_TRANSACTION_TYPE_I2C; + ctx.request.action = middle_of_transaction ? + I2CAUX_TRANSACTION_ACTION_I2C_WRITE_MOT : + I2CAUX_TRANSACTION_ACTION_I2C_WRITE; + ctx.request.address = request->payload.address >> 1; + } else { + /* in DAL2, there was no return in such case */ + BREAK_TO_DEBUGGER(); + return false; + } + + ctx.request.delay = 0; + + ctx.max_defer_retry = + (engine->max_defer_write_retry > AUX_DEFER_RETRY_COUNTER) ? + engine->max_defer_write_retry : AUX_DEFER_RETRY_COUNTER; + + do { + ctx.request.data = ctx.buffer; + ctx.request.length = ctx.current_write_length; + + process_write_request(engine, &ctx); + + request->status = ctx.status; + + if (ctx.operation_succeeded && !ctx.transaction_complete) + if (ctx.request.type == AUX_TRANSACTION_TYPE_I2C) + msleep(engine->delay); + } while (ctx.operation_succeeded && !ctx.transaction_complete); + + if (request->payload.address_space == + I2CAUX_TRANSACTION_ADDRESS_SPACE_DPCD) { + DC_LOG_I2C_AUX("WRITE: addr:0x%x value:0x%x Result:%d", + request->payload.address, + request->payload.data[0], + ctx.operation_succeeded); + } + + return ctx.operation_succeeded; +} +static bool end_of_transaction_command( + struct aux_engine *engine, + struct i2caux_transaction_request *request) +{ + struct i2caux_transaction_request dummy_request; + uint8_t dummy_data; + + /* [tcheng] We only need to send the stop (read with MOT = 0) + * for I2C-over-Aux, not native AUX + */ + + if (request->payload.address_space != + I2CAUX_TRANSACTION_ADDRESS_SPACE_I2C) + return false; + + dummy_request.operation = request->operation; + dummy_request.payload.address_space = request->payload.address_space; + dummy_request.payload.address = request->payload.address; + + /* + * Add a dummy byte due to some receiver quirk + * where one byte is sent along with MOT = 0. + * Ideally this should be 0. + */ + + dummy_request.payload.length = 0; + dummy_request.payload.data = &dummy_data; + + if (request->operation == I2CAUX_TRANSACTION_READ) + return read_command(engine, &dummy_request, false); + else + return write_command(engine, &dummy_request, false); + + /* according Syed, it does not need now DoDummyMOT */ +} +static bool submit_request( + struct aux_engine *engine, + struct i2caux_transaction_request *request, + bool middle_of_transaction) +{ + + bool result; + bool mot_used = true; + + switch (request->operation) { + case I2CAUX_TRANSACTION_READ: + result = read_command(engine, request, mot_used); + break; + case I2CAUX_TRANSACTION_WRITE: + result = write_command(engine, request, mot_used); + break; + default: + result = false; + } + + /* [tcheng] + * need to send stop for the last transaction to free up the AUX + * if the above command fails, this would be the last transaction + */ + + if (!middle_of_transaction || !result) + end_of_transaction_command(engine, request); + + /* mask AUX interrupt */ + + return result; +} +enum i2caux_engine_type get_engine_type( + const struct aux_engine *engine) +{ + return I2CAUX_ENGINE_TYPE_AUX; +} + +static bool acquire( + struct aux_engine *engine, + struct ddc *ddc) +{ + + enum gpio_result result; + + if (engine->funcs->is_engine_available) { + /*check whether SW could use the engine*/ + if (!engine->funcs->is_engine_available(engine)) + return false; + } + + result = dal_ddc_open(ddc, GPIO_MODE_HARDWARE, + GPIO_DDC_CONFIG_TYPE_MODE_AUX); + + if (result != GPIO_RESULT_OK) + return false; + + if (!engine->funcs->acquire_engine(engine)) { + dal_ddc_close(ddc); + return false; + } + + engine->ddc = ddc; + + return true; +} + +static const struct aux_engine_funcs aux_engine_funcs = { + .acquire_engine = acquire_engine, + .submit_channel_request = submit_channel_request, + .process_channel_reply = process_channel_reply, + .read_channel_reply = read_channel_reply, + .get_channel_status = get_channel_status, + .is_engine_available = is_engine_available, + .release_engine = release_engine, + .destroy_engine = dce110_engine_destroy, + .submit_request = submit_request, + .get_engine_type = get_engine_type, + .acquire = acquire, +}; + +void dce110_engine_destroy(struct aux_engine **engine) +{ + + struct aux_engine_dce110 *engine110 = FROM_AUX_ENGINE(*engine); + + kfree(engine110); + *engine = NULL; + +} +struct aux_engine *dce110_aux_engine_construct(struct aux_engine_dce110 *aux_engine110, + struct dc_context *ctx, + uint32_t inst, + uint32_t timeout_period, + const struct dce110_aux_registers *regs) +{ + aux_engine110->base.ddc = NULL; + aux_engine110->base.ctx = ctx; + aux_engine110->base.delay = 0; + aux_engine110->base.max_defer_write_retry = 0; + aux_engine110->base.funcs = &aux_engine_funcs; + aux_engine110->base.inst = inst; + aux_engine110->timeout_period = timeout_period; + aux_engine110->regs = regs; + + return &aux_engine110->base; +} + diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_aux.h b/drivers/gpu/drm/amd/display/dc/dce/dce_aux.h new file mode 100644 index 000000000000..f7caab85dc80 --- /dev/null +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_aux.h @@ -0,0 +1,111 @@ +/* + * Copyright 2012-15 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: AMD + * + */ + +#ifndef __DAL_AUX_ENGINE_DCE110_H__ +#define __DAL_AUX_ENGINE_DCE110_H__ +#include "aux_engine.h" + +#define AUX_COMMON_REG_LIST(id)\ + SRI(AUX_CONTROL, DP_AUX, id), \ + SRI(AUX_ARB_CONTROL, DP_AUX, id), \ + SRI(AUX_SW_DATA, DP_AUX, id), \ + SRI(AUX_SW_CONTROL, DP_AUX, id), \ + SRI(AUX_INTERRUPT_CONTROL, DP_AUX, id), \ + SRI(AUX_SW_STATUS, DP_AUX, id), \ + SR(AUXN_IMPCAL), \ + SR(AUXP_IMPCAL) + +struct dce110_aux_registers { + uint32_t AUX_CONTROL; + uint32_t AUX_ARB_CONTROL; + uint32_t AUX_SW_DATA; + uint32_t AUX_SW_CONTROL; + uint32_t AUX_INTERRUPT_CONTROL; + uint32_t AUX_SW_STATUS; + uint32_t AUXN_IMPCAL; + uint32_t AUXP_IMPCAL; + + uint32_t AUX_RESET_MASK; +}; + +enum { /* This is the timeout as defined in DP 1.2a, + * 2.3.4 "Detailed uPacket TX AUX CH State Description". + */ + AUX_TIMEOUT_PERIOD = 400, + + /* Ideally, the SW timeout should be just above 550usec + * which is programmed in HW. + * But the SW timeout of 600usec is not reliable, + * because on some systems, delay_in_microseconds() + * returns faster than it should. + * EPR #379763: by trial-and-error on different systems, + * 700usec is the minimum reliable SW timeout for polling + * the AUX_SW_STATUS.AUX_SW_DONE bit. + * This timeout expires *only* when there is + * AUX Error or AUX Timeout conditions - not during normal operation. + * During normal operation, AUX_SW_STATUS.AUX_SW_DONE bit is set + * at most within ~240usec. That means, + * increasing this timeout will not affect normal operation, + * and we'll timeout after + * SW_AUX_TIMEOUT_PERIOD_MULTIPLIER * AUX_TIMEOUT_PERIOD = 1600usec. + * This timeout is especially important for + * resume from S3 and CTS. + */ + SW_AUX_TIMEOUT_PERIOD_MULTIPLIER = 4 +}; +struct aux_engine_dce110 { + struct aux_engine base; + const struct dce110_aux_registers *regs; + struct { + uint32_t aux_control; + uint32_t aux_arb_control; + uint32_t aux_sw_data; + uint32_t aux_sw_control; + uint32_t aux_interrupt_control; + uint32_t aux_sw_status; + } addr; + uint32_t timeout_period; +}; + +struct aux_engine_dce110_init_data { + uint32_t engine_id; + uint32_t timeout_period; + struct dc_context *ctx; + const struct dce110_aux_registers *regs; +}; + +struct aux_engine *dce110_aux_engine_construct( + struct aux_engine_dce110 *aux_engine110, + struct dc_context *ctx, + uint32_t inst, + uint32_t timeout_period, + const struct dce110_aux_registers *regs); + +void dce110_engine_destroy(struct aux_engine **engine); + +bool dce110_aux_engine_acquire( + struct aux_engine *aux_engine, + struct ddc *ddc); +#endif diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c b/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c index 88b09dd758ba..439dcf3b596c 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c @@ -133,7 +133,7 @@ static bool calculate_fb_and_fractional_fb_divider( uint64_t feedback_divider; feedback_divider = - (uint64_t)(target_pix_clk_khz * ref_divider * post_divider); + (uint64_t)target_pix_clk_khz * ref_divider * post_divider; feedback_divider *= 10; /* additional factor, since we divide by 10 afterwards */ feedback_divider *= (uint64_t)(calc_pll_cs->fract_fb_divider_factor); @@ -145,8 +145,8 @@ static bool calculate_fb_and_fractional_fb_divider( * of fractional feedback decimal point and the fractional FB Divider precision * is 2 then the equation becomes (ullfeedbackDivider + 5*100) / (10*100))*/ - feedback_divider += (uint64_t) - (5 * calc_pll_cs->fract_fb_divider_precision_factor); + feedback_divider += 5ULL * + calc_pll_cs->fract_fb_divider_precision_factor; feedback_divider = div_u64(feedback_divider, calc_pll_cs->fract_fb_divider_precision_factor * 10); @@ -203,8 +203,8 @@ static bool calc_fb_divider_checking_tolerance( &fract_feedback_divider); /*Actual calculated value*/ - actual_calc_clk_khz = (uint64_t)(feedback_divider * - calc_pll_cs->fract_fb_divider_factor) + + actual_calc_clk_khz = (uint64_t)feedback_divider * + calc_pll_cs->fract_fb_divider_factor + fract_feedback_divider; actual_calc_clk_khz *= calc_pll_cs->ref_freq_khz; actual_calc_clk_khz = @@ -592,7 +592,7 @@ static uint32_t dce110_get_pix_clk_dividers( case DCE_VERSION_11_2: case DCE_VERSION_11_22: case DCE_VERSION_12_0: -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 case DCN_VERSION_1_0: #endif @@ -909,7 +909,7 @@ static bool dce110_program_pix_clk( struct dce110_clk_src *clk_src = TO_DCE110_CLK_SRC(clock_source); struct bp_pixel_clock_parameters bp_pc_params = {0}; -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 if (IS_FPGA_MAXIMUS_DC(clock_source->ctx->dce_environment)) { unsigned int inst = pix_clk_params->controller_id - CONTROLLER_ID_D0; unsigned dp_dto_ref_kHz = 700000; @@ -982,7 +982,7 @@ static bool dce110_program_pix_clk( case DCE_VERSION_11_2: case DCE_VERSION_11_22: case DCE_VERSION_12_0: -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 case DCN_VERSION_1_0: #endif diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.h b/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.h index c45e2f76189e..801bb65707b3 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.h +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.h @@ -55,7 +55,7 @@ CS_SF(PHYPLLA_PIXCLK_RESYNC_CNTL, PHYPLLA_DCCG_DEEP_COLOR_CNTL, mask_sh),\ CS_SF(PHYPLLA_PIXCLK_RESYNC_CNTL, PHYPLLA_PIXCLK_DOUBLE_RATE_ENABLE, mask_sh) -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 #define CS_COMMON_REG_LIST_DCN1_0(index, pllid) \ SRI(PIXCLK_RESYNC_CNTL, PHYPLL, pllid),\ diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_clocks.c b/drivers/gpu/drm/amd/display/dc/dce/dce_clocks.c index 8a581c67bf2d..684da3db7568 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_clocks.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_clocks.c @@ -30,7 +30,7 @@ #include "bios_parser_interface.h" #include "dc.h" #include "dmcu.h" -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 #include "dcn_calcs.h" #endif #include "core_types.h" @@ -38,7 +38,7 @@ #include "dal_asic_id.h" #define TO_DCE_CLOCKS(clocks)\ - container_of(clocks, struct dce_disp_clk, base) + container_of(clocks, struct dce_dccg, base) #define REG(reg) \ (clk_dce->regs->reg) @@ -101,99 +101,78 @@ static const struct state_dependent_clocks dce120_max_clks_by_state[] = { /*ClocksStatePerformance*/ { .display_clk_khz = 1133000, .pixel_clk_khz = 600000 } }; -/* Starting point for each divider range.*/ -enum dce_divider_range_start { - DIVIDER_RANGE_01_START = 200, /* 2.00*/ - DIVIDER_RANGE_02_START = 1600, /* 16.00*/ - DIVIDER_RANGE_03_START = 3200, /* 32.00*/ - DIVIDER_RANGE_SCALE_FACTOR = 100 /* Results are scaled up by 100.*/ +/* Starting DID for each range */ +enum dentist_base_divider_id { + DENTIST_BASE_DID_1 = 0x08, + DENTIST_BASE_DID_2 = 0x40, + DENTIST_BASE_DID_3 = 0x60, + DENTIST_MAX_DID = 0x80 }; -/* Ranges for divider identifiers (Divider ID or DID) - mmDENTIST_DISPCLK_CNTL.DENTIST_DISPCLK_WDIVIDER*/ -enum dce_divider_id_register_setting { - DIVIDER_RANGE_01_BASE_DIVIDER_ID = 0X08, - DIVIDER_RANGE_02_BASE_DIVIDER_ID = 0X40, - DIVIDER_RANGE_03_BASE_DIVIDER_ID = 0X60, - DIVIDER_RANGE_MAX_DIVIDER_ID = 0X80 +/* Starting point and step size for each divider range.*/ +enum dentist_divider_range { + DENTIST_DIVIDER_RANGE_1_START = 8, /* 2.00 */ + DENTIST_DIVIDER_RANGE_1_STEP = 1, /* 0.25 */ + DENTIST_DIVIDER_RANGE_2_START = 64, /* 16.00 */ + DENTIST_DIVIDER_RANGE_2_STEP = 2, /* 0.50 */ + DENTIST_DIVIDER_RANGE_3_START = 128, /* 32.00 */ + DENTIST_DIVIDER_RANGE_3_STEP = 4, /* 1.00 */ + DENTIST_DIVIDER_RANGE_SCALE_FACTOR = 4 }; -/* Step size between each divider within a range. - Incrementing the DENTIST_DISPCLK_WDIVIDER by one - will increment the divider by this much.*/ -enum dce_divider_range_step_size { - DIVIDER_RANGE_01_STEP_SIZE = 25, /* 0.25*/ - DIVIDER_RANGE_02_STEP_SIZE = 50, /* 0.50*/ - DIVIDER_RANGE_03_STEP_SIZE = 100 /* 1.00 */ -}; - -static bool dce_divider_range_construct( - struct dce_divider_range *div_range, - int range_start, - int range_step, - int did_min, - int did_max) +static int dentist_get_divider_from_did(int did) { - div_range->div_range_start = range_start; - div_range->div_range_step = range_step; - div_range->did_min = did_min; - div_range->did_max = did_max; - - if (div_range->div_range_step == 0) { - div_range->div_range_step = 1; - /*div_range_step cannot be zero*/ - BREAK_TO_DEBUGGER(); + if (did < DENTIST_BASE_DID_1) + did = DENTIST_BASE_DID_1; + if (did > DENTIST_MAX_DID) + did = DENTIST_MAX_DID; + + if (did < DENTIST_BASE_DID_2) { + return DENTIST_DIVIDER_RANGE_1_START + DENTIST_DIVIDER_RANGE_1_STEP + * (did - DENTIST_BASE_DID_1); + } else if (did < DENTIST_BASE_DID_3) { + return DENTIST_DIVIDER_RANGE_2_START + DENTIST_DIVIDER_RANGE_2_STEP + * (did - DENTIST_BASE_DID_2); + } else { + return DENTIST_DIVIDER_RANGE_3_START + DENTIST_DIVIDER_RANGE_3_STEP + * (did - DENTIST_BASE_DID_3); } - /* Calculate this based on the other inputs.*/ - /* See DividerRange.h for explanation of */ - /* the relationship between divider id (DID) and a divider.*/ - /* Number of Divider IDs = (Maximum Divider ID - Minimum Divider ID)*/ - /* Maximum divider identified in this range = - * (Number of Divider IDs)*Step size between dividers - * + The start of this range.*/ - div_range->div_range_end = (did_max - did_min) * range_step - + range_start; - return true; -} - -static int dce_divider_range_calc_divider( - struct dce_divider_range *div_range, - int did) -{ - /* Is this DID within our range?*/ - if ((did < div_range->did_min) || (did >= div_range->did_max)) - return INVALID_DIVIDER; - - return ((did - div_range->did_min) * div_range->div_range_step) - + div_range->div_range_start; - } -static int dce_divider_range_get_divider( - struct dce_divider_range *div_range, - int ranges_num, - int did) +/* SW will adjust DP REF Clock average value for all purposes + * (DP DTO / DP Audio DTO and DP GTC) + if clock is spread for all cases: + -if SS enabled on DP Ref clock and HW de-spreading enabled with SW + calculations for DS_INCR/DS_MODULO (this is planned to be default case) + -if SS enabled on DP Ref clock and HW de-spreading enabled with HW + calculations (not planned to be used, but average clock should still + be valid) + -if SS enabled on DP Ref clock and HW de-spreading disabled + (should not be case with CIK) then SW should program all rates + generated according to average value (case as with previous ASICs) + */ +static int dccg_adjust_dp_ref_freq_for_ss(struct dce_dccg *clk_dce, int dp_ref_clk_khz) { - int div = INVALID_DIVIDER; - int i; + if (clk_dce->ss_on_dprefclk && clk_dce->dprefclk_ss_divider != 0) { + struct fixed31_32 ss_percentage = dc_fixpt_div_int( + dc_fixpt_from_fraction(clk_dce->dprefclk_ss_percentage, + clk_dce->dprefclk_ss_divider), 200); + struct fixed31_32 adj_dp_ref_clk_khz; - for (i = 0; i < ranges_num; i++) { - /* Calculate divider with given divider ID*/ - div = dce_divider_range_calc_divider(&div_range[i], did); - /* Found a valid return divider*/ - if (div != INVALID_DIVIDER) - break; + ss_percentage = dc_fixpt_sub(dc_fixpt_one, ss_percentage); + adj_dp_ref_clk_khz = dc_fixpt_mul_int(ss_percentage, dp_ref_clk_khz); + dp_ref_clk_khz = dc_fixpt_floor(adj_dp_ref_clk_khz); } - return div; + return dp_ref_clk_khz; } -static int dce_clocks_get_dp_ref_freq(struct display_clock *clk) +static int dce_get_dp_ref_freq_khz(struct dccg *clk) { - struct dce_disp_clk *clk_dce = TO_DCE_CLOCKS(clk); + struct dce_dccg *clk_dce = TO_DCE_CLOCKS(clk); int dprefclk_wdivider; int dprefclk_src_sel; int dp_ref_clk_khz = 600000; - int target_div = INVALID_DIVIDER; + int target_div; /* ASSERT DP Reference Clock source is from DFS*/ REG_GET(DPREFCLK_CNTL, DPREFCLK_SRC_SEL, &dprefclk_src_sel); @@ -204,80 +183,27 @@ static int dce_clocks_get_dp_ref_freq(struct display_clock *clk) REG_GET(DENTIST_DISPCLK_CNTL, DENTIST_DPREFCLK_WDIVIDER, &dprefclk_wdivider); /* Convert DENTIST_DPREFCLK_WDIVIDERto actual divider*/ - target_div = dce_divider_range_get_divider( - clk_dce->divider_ranges, - DIVIDER_RANGE_MAX, - dprefclk_wdivider); - - if (target_div != INVALID_DIVIDER) { - /* Calculate the current DFS clock, in kHz.*/ - dp_ref_clk_khz = (DIVIDER_RANGE_SCALE_FACTOR - * clk_dce->dentist_vco_freq_khz) / target_div; - } + target_div = dentist_get_divider_from_did(dprefclk_wdivider); - /* SW will adjust DP REF Clock average value for all purposes - * (DP DTO / DP Audio DTO and DP GTC) - if clock is spread for all cases: - -if SS enabled on DP Ref clock and HW de-spreading enabled with SW - calculations for DS_INCR/DS_MODULO (this is planned to be default case) - -if SS enabled on DP Ref clock and HW de-spreading enabled with HW - calculations (not planned to be used, but average clock should still - be valid) - -if SS enabled on DP Ref clock and HW de-spreading disabled - (should not be case with CIK) then SW should program all rates - generated according to average value (case as with previous ASICs) - */ - if (clk_dce->ss_on_dprefclk && clk_dce->dprefclk_ss_divider != 0) { - struct fixed31_32 ss_percentage = dc_fixpt_div_int( - dc_fixpt_from_fraction( - clk_dce->dprefclk_ss_percentage, - clk_dce->dprefclk_ss_divider), 200); - struct fixed31_32 adj_dp_ref_clk_khz; + /* Calculate the current DFS clock, in kHz.*/ + dp_ref_clk_khz = (DENTIST_DIVIDER_RANGE_SCALE_FACTOR + * clk_dce->dentist_vco_freq_khz) / target_div; - ss_percentage = dc_fixpt_sub(dc_fixpt_one, - ss_percentage); - adj_dp_ref_clk_khz = - dc_fixpt_mul_int( - ss_percentage, - dp_ref_clk_khz); - dp_ref_clk_khz = dc_fixpt_floor(adj_dp_ref_clk_khz); - } - - return dp_ref_clk_khz; + return dccg_adjust_dp_ref_freq_for_ss(clk_dce, dp_ref_clk_khz); } -/* TODO: This is DCN DPREFCLK: it could be program by DENTIST by VBIOS - * or CLK0_CLK11 by SMU. For DCE120, it is wlays 600Mhz. Will re-visit - * clock implementation - */ -static int dce_clocks_get_dp_ref_freq_wrkaround(struct display_clock *clk) +static int dce12_get_dp_ref_freq_khz(struct dccg *clk) { - struct dce_disp_clk *clk_dce = TO_DCE_CLOCKS(clk); - int dp_ref_clk_khz = 600000; - - if (clk_dce->ss_on_dprefclk && clk_dce->dprefclk_ss_divider != 0) { - struct fixed31_32 ss_percentage = dc_fixpt_div_int( - dc_fixpt_from_fraction( - clk_dce->dprefclk_ss_percentage, - clk_dce->dprefclk_ss_divider), 200); - struct fixed31_32 adj_dp_ref_clk_khz; + struct dce_dccg *clk_dce = TO_DCE_CLOCKS(clk); - ss_percentage = dc_fixpt_sub(dc_fixpt_one, - ss_percentage); - adj_dp_ref_clk_khz = - dc_fixpt_mul_int( - ss_percentage, - dp_ref_clk_khz); - dp_ref_clk_khz = dc_fixpt_floor(adj_dp_ref_clk_khz); - } - - return dp_ref_clk_khz; + return dccg_adjust_dp_ref_freq_for_ss(clk_dce, 600000); } + static enum dm_pp_clocks_state dce_get_required_clocks_state( - struct display_clock *clk, - struct state_dependent_clocks *req_clocks) + struct dccg *clk, + struct dc_clocks *req_clocks) { - struct dce_disp_clk *clk_dce = TO_DCE_CLOCKS(clk); + struct dce_dccg *clk_dce = TO_DCE_CLOCKS(clk); int i; enum dm_pp_clocks_state low_req_clk; @@ -286,53 +212,30 @@ static enum dm_pp_clocks_state dce_get_required_clocks_state( * all required clocks */ for (i = clk->max_clks_state; i >= DM_PP_CLOCKS_STATE_ULTRA_LOW; i--) - if (req_clocks->display_clk_khz > + if (req_clocks->dispclk_khz > clk_dce->max_clks_by_state[i].display_clk_khz - || req_clocks->pixel_clk_khz > + || req_clocks->phyclk_khz > clk_dce->max_clks_by_state[i].pixel_clk_khz) break; low_req_clk = i + 1; if (low_req_clk > clk->max_clks_state) { - DC_LOG_WARNING("%s: clocks unsupported disp_clk %d pix_clk %d", - __func__, - req_clocks->display_clk_khz, - req_clocks->pixel_clk_khz); - low_req_clk = DM_PP_CLOCKS_STATE_INVALID; + /* set max clock state for high phyclock, invalid on exceeding display clock */ + if (clk_dce->max_clks_by_state[clk->max_clks_state].display_clk_khz + < req_clocks->dispclk_khz) + low_req_clk = DM_PP_CLOCKS_STATE_INVALID; + else + low_req_clk = clk->max_clks_state; } return low_req_clk; } -static bool dce_clock_set_min_clocks_state( - struct display_clock *clk, - enum dm_pp_clocks_state clocks_state) -{ - struct dm_pp_power_level_change_request level_change_req = { - clocks_state }; - - if (clocks_state > clk->max_clks_state) { - /*Requested state exceeds max supported state.*/ - DC_LOG_WARNING("Requested state exceeds max supported state"); - return false; - } else if (clocks_state == clk->cur_min_clks_state) { - /*if we're trying to set the same state, we can just return - * since nothing needs to be done*/ - return true; - } - - /* get max clock state from PPLIB */ - if (dm_pp_apply_power_level_change_request(clk->ctx, &level_change_req)) - clk->cur_min_clks_state = clocks_state; - - return true; -} - static int dce_set_clock( - struct display_clock *clk, + struct dccg *clk, int requested_clk_khz) { - struct dce_disp_clk *clk_dce = TO_DCE_CLOCKS(clk); + struct dce_dccg *clk_dce = TO_DCE_CLOCKS(clk); struct bp_pixel_clock_parameters pxl_clk_params = { 0 }; struct dc_bios *bp = clk->ctx->dc_bios; int actual_clock = requested_clk_khz; @@ -364,10 +267,10 @@ static int dce_set_clock( } static int dce_psr_set_clock( - struct display_clock *clk, + struct dccg *clk, int requested_clk_khz) { - struct dce_disp_clk *clk_dce = TO_DCE_CLOCKS(clk); + struct dce_dccg *clk_dce = TO_DCE_CLOCKS(clk); struct dc_context *ctx = clk_dce->base.ctx; struct dc *core_dc = ctx->dc; struct dmcu *dmcu = core_dc->res_pool->dmcu; @@ -380,10 +283,10 @@ static int dce_psr_set_clock( } static int dce112_set_clock( - struct display_clock *clk, + struct dccg *clk, int requested_clk_khz) { - struct dce_disp_clk *clk_dce = TO_DCE_CLOCKS(clk); + struct dce_dccg *clk_dce = TO_DCE_CLOCKS(clk); struct bp_set_dce_clock_parameters dce_clk_params; struct dc_bios *bp = clk->ctx->dc_bios; struct dc *core_dc = clk->ctx->dc; @@ -432,9 +335,9 @@ static int dce112_set_clock( return actual_clock; } -static void dce_clock_read_integrated_info(struct dce_disp_clk *clk_dce) +static void dce_clock_read_integrated_info(struct dce_dccg *clk_dce) { - struct dc_debug *debug = &clk_dce->base.ctx->dc->debug; + struct dc_debug_options *debug = &clk_dce->base.ctx->dc->debug; struct dc_bios *bp = clk_dce->base.ctx->dc_bios; struct integrated_info info = { { { 0 } } }; struct dc_firmware_info fw_info = { { 0 } }; @@ -488,11 +391,9 @@ static void dce_clock_read_integrated_info(struct dce_disp_clk *clk_dce) if (!debug->disable_dfs_bypass && bp->integrated_info) if (bp->integrated_info->gpu_cap_info & DFS_BYPASS_ENABLE) clk_dce->dfs_bypass_enabled = true; - - clk_dce->use_max_disp_clk = debug->max_disp_clk; } -static void dce_clock_read_ss_info(struct dce_disp_clk *clk_dce) +static void dce_clock_read_ss_info(struct dce_dccg *clk_dce) { struct dc_bios *bp = clk_dce->base.ctx->dc_bios; int ss_info_num = bp->funcs->get_ss_entry_number( @@ -548,139 +449,265 @@ static void dce_clock_read_ss_info(struct dce_disp_clk *clk_dce) } } -static bool dce_apply_clock_voltage_request( - struct display_clock *clk, - enum dm_pp_clock_type clocks_type, - int clocks_in_khz, - bool pre_mode_set, - bool update_dp_phyclk) +static inline bool should_set_clock(bool safe_to_lower, int calc_clk, int cur_clk) +{ + return ((safe_to_lower && calc_clk < cur_clk) || calc_clk > cur_clk); +} + +static void dce12_update_clocks(struct dccg *dccg, + struct dc_clocks *new_clocks, + bool safe_to_lower) { - bool send_request = false; struct dm_pp_clock_for_voltage_req clock_voltage_req = {0}; - switch (clocks_type) { - case DM_PP_CLOCK_TYPE_DISPLAY_CLK: - case DM_PP_CLOCK_TYPE_PIXELCLK: - case DM_PP_CLOCK_TYPE_DISPLAYPHYCLK: - break; - default: - BREAK_TO_DEBUGGER(); - return false; + if (should_set_clock(safe_to_lower, new_clocks->dispclk_khz, dccg->clks.dispclk_khz)) { + clock_voltage_req.clk_type = DM_PP_CLOCK_TYPE_DISPLAY_CLK; + clock_voltage_req.clocks_in_khz = new_clocks->dispclk_khz; + new_clocks->dispclk_khz = dccg->funcs->set_dispclk(dccg, new_clocks->dispclk_khz); + dccg->clks.dispclk_khz = new_clocks->dispclk_khz; + + dm_pp_apply_clock_for_voltage_request(dccg->ctx, &clock_voltage_req); } - clock_voltage_req.clk_type = clocks_type; - clock_voltage_req.clocks_in_khz = clocks_in_khz; - - /* to pplib */ - if (pre_mode_set) { - switch (clocks_type) { - case DM_PP_CLOCK_TYPE_DISPLAY_CLK: - if (clocks_in_khz > clk->cur_clocks_value.dispclk_in_khz) { - clk->cur_clocks_value.dispclk_notify_pplib_done = true; - send_request = true; - } else - clk->cur_clocks_value.dispclk_notify_pplib_done = false; - /* no matter incrase or decrase clock, update current clock value */ - clk->cur_clocks_value.dispclk_in_khz = clocks_in_khz; - break; - case DM_PP_CLOCK_TYPE_PIXELCLK: - if (clocks_in_khz > clk->cur_clocks_value.max_pixelclk_in_khz) { - clk->cur_clocks_value.pixelclk_notify_pplib_done = true; - send_request = true; - } else - clk->cur_clocks_value.pixelclk_notify_pplib_done = false; - /* no matter incrase or decrase clock, update current clock value */ - clk->cur_clocks_value.max_pixelclk_in_khz = clocks_in_khz; - break; - case DM_PP_CLOCK_TYPE_DISPLAYPHYCLK: - if (clocks_in_khz > clk->cur_clocks_value.max_non_dp_phyclk_in_khz) { - clk->cur_clocks_value.phyclk_notigy_pplib_done = true; - send_request = true; - } else - clk->cur_clocks_value.phyclk_notigy_pplib_done = false; - /* no matter incrase or decrase clock, update current clock value */ - clk->cur_clocks_value.max_non_dp_phyclk_in_khz = clocks_in_khz; - break; - default: - ASSERT(0); - break; - } + if (should_set_clock(safe_to_lower, new_clocks->phyclk_khz, dccg->clks.phyclk_khz)) { + clock_voltage_req.clk_type = DM_PP_CLOCK_TYPE_DISPLAYPHYCLK; + clock_voltage_req.clocks_in_khz = new_clocks->phyclk_khz; + dccg->clks.phyclk_khz = new_clocks->phyclk_khz; + + dm_pp_apply_clock_for_voltage_request(dccg->ctx, &clock_voltage_req); + } +} + +#ifdef CONFIG_X86 +static int dcn1_determine_dppclk_threshold(struct dccg *dccg, struct dc_clocks *new_clocks) +{ + bool request_dpp_div = new_clocks->dispclk_khz > new_clocks->dppclk_khz; + bool dispclk_increase = new_clocks->dispclk_khz > dccg->clks.dispclk_khz; + int disp_clk_threshold = new_clocks->max_supported_dppclk_khz; + bool cur_dpp_div = dccg->clks.dispclk_khz > dccg->clks.dppclk_khz; + + /* increase clock, looking for div is 0 for current, request div is 1*/ + if (dispclk_increase) { + /* already divided by 2, no need to reach target clk with 2 steps*/ + if (cur_dpp_div) + return new_clocks->dispclk_khz; + + /* request disp clk is lower than maximum supported dpp clk, + * no need to reach target clk with two steps. + */ + if (new_clocks->dispclk_khz <= disp_clk_threshold) + return new_clocks->dispclk_khz; + + /* target dpp clk not request divided by 2, still within threshold */ + if (!request_dpp_div) + return new_clocks->dispclk_khz; } else { - switch (clocks_type) { - case DM_PP_CLOCK_TYPE_DISPLAY_CLK: - if (!clk->cur_clocks_value.dispclk_notify_pplib_done) - send_request = true; - break; - case DM_PP_CLOCK_TYPE_PIXELCLK: - if (!clk->cur_clocks_value.pixelclk_notify_pplib_done) - send_request = true; - break; - case DM_PP_CLOCK_TYPE_DISPLAYPHYCLK: - if (!clk->cur_clocks_value.phyclk_notigy_pplib_done) - send_request = true; - break; - default: - ASSERT(0); - break; - } + /* decrease clock, looking for current dppclk divided by 2, + * request dppclk not divided by 2. + */ + + /* current dpp clk not divided by 2, no need to ramp*/ + if (!cur_dpp_div) + return new_clocks->dispclk_khz; + + /* current disp clk is lower than current maximum dpp clk, + * no need to ramp + */ + if (dccg->clks.dispclk_khz <= disp_clk_threshold) + return new_clocks->dispclk_khz; + + /* request dpp clk need to be divided by 2 */ + if (request_dpp_div) + return new_clocks->dispclk_khz; } - if (send_request) { -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) - if (clk->ctx->dce_version >= DCN_VERSION_1_0) { - struct dc *core_dc = clk->ctx->dc; - /*use dcfclk request voltage*/ - clock_voltage_req.clk_type = DM_PP_CLOCK_TYPE_DCFCLK; - clock_voltage_req.clocks_in_khz = - dcn_find_dcfclk_suits_all(core_dc, &clk->cur_clocks_value); - } + + return disp_clk_threshold; +} + +static void dcn1_ramp_up_dispclk_with_dpp(struct dccg *dccg, struct dc_clocks *new_clocks) +{ + struct dc *dc = dccg->ctx->dc; + int dispclk_to_dpp_threshold = dcn1_determine_dppclk_threshold(dccg, new_clocks); + bool request_dpp_div = new_clocks->dispclk_khz > new_clocks->dppclk_khz; + int i; + + /* set disp clk to dpp clk threshold */ + dccg->funcs->set_dispclk(dccg, dispclk_to_dpp_threshold); + + /* update request dpp clk division option */ + for (i = 0; i < dc->res_pool->pipe_count; i++) { + struct pipe_ctx *pipe_ctx = &dc->current_state->res_ctx.pipe_ctx[i]; + + if (!pipe_ctx->plane_state) + continue; + + pipe_ctx->plane_res.dpp->funcs->dpp_dppclk_control( + pipe_ctx->plane_res.dpp, + request_dpp_div, + true); + } + + /* If target clk not same as dppclk threshold, set to target clock */ + if (dispclk_to_dpp_threshold != new_clocks->dispclk_khz) + dccg->funcs->set_dispclk(dccg, new_clocks->dispclk_khz); + + dccg->clks.dispclk_khz = new_clocks->dispclk_khz; + dccg->clks.dppclk_khz = new_clocks->dppclk_khz; + dccg->clks.max_supported_dppclk_khz = new_clocks->max_supported_dppclk_khz; +} + +static void dcn1_update_clocks(struct dccg *dccg, + struct dc_clocks *new_clocks, + bool safe_to_lower) +{ + struct dc *dc = dccg->ctx->dc; + struct pp_smu_display_requirement_rv *smu_req_cur = + &dc->res_pool->pp_smu_req; + struct pp_smu_display_requirement_rv smu_req = *smu_req_cur; + struct pp_smu_funcs_rv *pp_smu = dc->res_pool->pp_smu; + struct dm_pp_clock_for_voltage_req clock_voltage_req = {0}; + bool send_request_to_increase = false; + bool send_request_to_lower = false; + + if (new_clocks->phyclk_khz) + smu_req.display_count = 1; + else + smu_req.display_count = 0; + + if (new_clocks->dispclk_khz > dccg->clks.dispclk_khz + || new_clocks->phyclk_khz > dccg->clks.phyclk_khz + || new_clocks->fclk_khz > dccg->clks.fclk_khz + || new_clocks->dcfclk_khz > dccg->clks.dcfclk_khz) + send_request_to_increase = true; + + if (should_set_clock(safe_to_lower, new_clocks->phyclk_khz, dccg->clks.phyclk_khz)) { + dccg->clks.phyclk_khz = new_clocks->phyclk_khz; + + send_request_to_lower = true; + } + + if (should_set_clock(safe_to_lower, new_clocks->fclk_khz, dccg->clks.fclk_khz)) { + dccg->clks.fclk_khz = new_clocks->fclk_khz; + clock_voltage_req.clk_type = DM_PP_CLOCK_TYPE_FCLK; + clock_voltage_req.clocks_in_khz = new_clocks->fclk_khz; + smu_req.hard_min_fclk_khz = new_clocks->fclk_khz; + + dm_pp_apply_clock_for_voltage_request(dccg->ctx, &clock_voltage_req); + send_request_to_lower = true; + } + + if (should_set_clock(safe_to_lower, new_clocks->dcfclk_khz, dccg->clks.dcfclk_khz)) { + dccg->clks.dcfclk_khz = new_clocks->dcfclk_khz; + smu_req.hard_min_dcefclk_khz = new_clocks->dcfclk_khz; + + send_request_to_lower = true; + } + + if (should_set_clock(safe_to_lower, + new_clocks->dcfclk_deep_sleep_khz, dccg->clks.dcfclk_deep_sleep_khz)) { + dccg->clks.dcfclk_deep_sleep_khz = new_clocks->dcfclk_deep_sleep_khz; + smu_req.min_deep_sleep_dcefclk_mhz = new_clocks->dcfclk_deep_sleep_khz; + + send_request_to_lower = true; + } + + /* make sure dcf clk is before dpp clk to + * make sure we have enough voltage to run dpp clk + */ + if (send_request_to_increase) { + /*use dcfclk to request voltage*/ + clock_voltage_req.clk_type = DM_PP_CLOCK_TYPE_DCFCLK; + clock_voltage_req.clocks_in_khz = dcn_find_dcfclk_suits_all(dc, new_clocks); + dm_pp_apply_clock_for_voltage_request(dccg->ctx, &clock_voltage_req); + if (pp_smu->set_display_requirement) + pp_smu->set_display_requirement(&pp_smu->pp_smu, &smu_req); + } + + /* dcn1 dppclk is tied to dispclk */ + /* program dispclk on = as a w/a for sleep resume clock ramping issues */ + if (should_set_clock(safe_to_lower, new_clocks->dispclk_khz, dccg->clks.dispclk_khz) + || new_clocks->dispclk_khz == dccg->clks.dispclk_khz) { + dcn1_ramp_up_dispclk_with_dpp(dccg, new_clocks); + dccg->clks.dispclk_khz = new_clocks->dispclk_khz; + + send_request_to_lower = true; + } + + if (!send_request_to_increase && send_request_to_lower) { + /*use dcfclk to request voltage*/ + clock_voltage_req.clk_type = DM_PP_CLOCK_TYPE_DCFCLK; + clock_voltage_req.clocks_in_khz = dcn_find_dcfclk_suits_all(dc, new_clocks); + dm_pp_apply_clock_for_voltage_request(dccg->ctx, &clock_voltage_req); + if (pp_smu->set_display_requirement) + pp_smu->set_display_requirement(&pp_smu->pp_smu, &smu_req); + } + + + *smu_req_cur = smu_req; +} #endif - dm_pp_apply_clock_for_voltage_request( - clk->ctx, &clock_voltage_req); + +static void dce_update_clocks(struct dccg *dccg, + struct dc_clocks *new_clocks, + bool safe_to_lower) +{ + struct dm_pp_power_level_change_request level_change_req; + + level_change_req.power_level = dce_get_required_clocks_state(dccg, new_clocks); + /* get max clock state from PPLIB */ + if ((level_change_req.power_level < dccg->cur_min_clks_state && safe_to_lower) + || level_change_req.power_level > dccg->cur_min_clks_state) { + if (dm_pp_apply_power_level_change_request(dccg->ctx, &level_change_req)) + dccg->cur_min_clks_state = level_change_req.power_level; } - if (update_dp_phyclk && (clocks_in_khz > - clk->cur_clocks_value.max_dp_phyclk_in_khz)) - clk->cur_clocks_value.max_dp_phyclk_in_khz = clocks_in_khz; - return true; + if (should_set_clock(safe_to_lower, new_clocks->dispclk_khz, dccg->clks.dispclk_khz)) { + new_clocks->dispclk_khz = dccg->funcs->set_dispclk(dccg, new_clocks->dispclk_khz); + dccg->clks.dispclk_khz = new_clocks->dispclk_khz; + } } +#ifdef CONFIG_X86 +static const struct display_clock_funcs dcn1_funcs = { + .get_dp_ref_clk_frequency = dce12_get_dp_ref_freq_khz, + .set_dispclk = dce112_set_clock, + .update_clocks = dcn1_update_clocks +}; +#endif static const struct display_clock_funcs dce120_funcs = { - .get_dp_ref_clk_frequency = dce_clocks_get_dp_ref_freq_wrkaround, - .apply_clock_voltage_request = dce_apply_clock_voltage_request, - .set_clock = dce112_set_clock + .get_dp_ref_clk_frequency = dce12_get_dp_ref_freq_khz, + .set_dispclk = dce112_set_clock, + .update_clocks = dce12_update_clocks }; static const struct display_clock_funcs dce112_funcs = { - .get_dp_ref_clk_frequency = dce_clocks_get_dp_ref_freq, - .get_required_clocks_state = dce_get_required_clocks_state, - .set_min_clocks_state = dce_clock_set_min_clocks_state, - .set_clock = dce112_set_clock + .get_dp_ref_clk_frequency = dce_get_dp_ref_freq_khz, + .set_dispclk = dce112_set_clock, + .update_clocks = dce_update_clocks }; static const struct display_clock_funcs dce110_funcs = { - .get_dp_ref_clk_frequency = dce_clocks_get_dp_ref_freq, - .get_required_clocks_state = dce_get_required_clocks_state, - .set_min_clocks_state = dce_clock_set_min_clocks_state, - .set_clock = dce_psr_set_clock + .get_dp_ref_clk_frequency = dce_get_dp_ref_freq_khz, + .set_dispclk = dce_psr_set_clock, + .update_clocks = dce_update_clocks }; static const struct display_clock_funcs dce_funcs = { - .get_dp_ref_clk_frequency = dce_clocks_get_dp_ref_freq, - .get_required_clocks_state = dce_get_required_clocks_state, - .set_min_clocks_state = dce_clock_set_min_clocks_state, - .set_clock = dce_set_clock + .get_dp_ref_clk_frequency = dce_get_dp_ref_freq_khz, + .set_dispclk = dce_set_clock, + .update_clocks = dce_update_clocks }; -static void dce_disp_clk_construct( - struct dce_disp_clk *clk_dce, +static void dce_dccg_construct( + struct dce_dccg *clk_dce, struct dc_context *ctx, - const struct dce_disp_clk_registers *regs, - const struct dce_disp_clk_shift *clk_shift, - const struct dce_disp_clk_mask *clk_mask) + const struct dccg_registers *regs, + const struct dccg_shift *clk_shift, + const struct dccg_mask *clk_mask) { - struct display_clock *base = &clk_dce->base; + struct dccg *base = &clk_dce->base; base->ctx = ctx; base->funcs = &dce_funcs; @@ -700,34 +727,15 @@ static void dce_disp_clk_construct( dce_clock_read_integrated_info(clk_dce); dce_clock_read_ss_info(clk_dce); - - dce_divider_range_construct( - &clk_dce->divider_ranges[DIVIDER_RANGE_01], - DIVIDER_RANGE_01_START, - DIVIDER_RANGE_01_STEP_SIZE, - DIVIDER_RANGE_01_BASE_DIVIDER_ID, - DIVIDER_RANGE_02_BASE_DIVIDER_ID); - dce_divider_range_construct( - &clk_dce->divider_ranges[DIVIDER_RANGE_02], - DIVIDER_RANGE_02_START, - DIVIDER_RANGE_02_STEP_SIZE, - DIVIDER_RANGE_02_BASE_DIVIDER_ID, - DIVIDER_RANGE_03_BASE_DIVIDER_ID); - dce_divider_range_construct( - &clk_dce->divider_ranges[DIVIDER_RANGE_03], - DIVIDER_RANGE_03_START, - DIVIDER_RANGE_03_STEP_SIZE, - DIVIDER_RANGE_03_BASE_DIVIDER_ID, - DIVIDER_RANGE_MAX_DIVIDER_ID); } -struct display_clock *dce_disp_clk_create( +struct dccg *dce_dccg_create( struct dc_context *ctx, - const struct dce_disp_clk_registers *regs, - const struct dce_disp_clk_shift *clk_shift, - const struct dce_disp_clk_mask *clk_mask) + const struct dccg_registers *regs, + const struct dccg_shift *clk_shift, + const struct dccg_mask *clk_mask) { - struct dce_disp_clk *clk_dce = kzalloc(sizeof(*clk_dce), GFP_KERNEL); + struct dce_dccg *clk_dce = kzalloc(sizeof(*clk_dce), GFP_KERNEL); if (clk_dce == NULL) { BREAK_TO_DEBUGGER(); @@ -738,19 +746,19 @@ struct display_clock *dce_disp_clk_create( dce80_max_clks_by_state, sizeof(dce80_max_clks_by_state)); - dce_disp_clk_construct( + dce_dccg_construct( clk_dce, ctx, regs, clk_shift, clk_mask); return &clk_dce->base; } -struct display_clock *dce110_disp_clk_create( +struct dccg *dce110_dccg_create( struct dc_context *ctx, - const struct dce_disp_clk_registers *regs, - const struct dce_disp_clk_shift *clk_shift, - const struct dce_disp_clk_mask *clk_mask) + const struct dccg_registers *regs, + const struct dccg_shift *clk_shift, + const struct dccg_mask *clk_mask) { - struct dce_disp_clk *clk_dce = kzalloc(sizeof(*clk_dce), GFP_KERNEL); + struct dce_dccg *clk_dce = kzalloc(sizeof(*clk_dce), GFP_KERNEL); if (clk_dce == NULL) { BREAK_TO_DEBUGGER(); @@ -761,7 +769,7 @@ struct display_clock *dce110_disp_clk_create( dce110_max_clks_by_state, sizeof(dce110_max_clks_by_state)); - dce_disp_clk_construct( + dce_dccg_construct( clk_dce, ctx, regs, clk_shift, clk_mask); clk_dce->base.funcs = &dce110_funcs; @@ -769,13 +777,13 @@ struct display_clock *dce110_disp_clk_create( return &clk_dce->base; } -struct display_clock *dce112_disp_clk_create( +struct dccg *dce112_dccg_create( struct dc_context *ctx, - const struct dce_disp_clk_registers *regs, - const struct dce_disp_clk_shift *clk_shift, - const struct dce_disp_clk_mask *clk_mask) + const struct dccg_registers *regs, + const struct dccg_shift *clk_shift, + const struct dccg_mask *clk_mask) { - struct dce_disp_clk *clk_dce = kzalloc(sizeof(*clk_dce), GFP_KERNEL); + struct dce_dccg *clk_dce = kzalloc(sizeof(*clk_dce), GFP_KERNEL); if (clk_dce == NULL) { BREAK_TO_DEBUGGER(); @@ -786,7 +794,7 @@ struct display_clock *dce112_disp_clk_create( dce112_max_clks_by_state, sizeof(dce112_max_clks_by_state)); - dce_disp_clk_construct( + dce_dccg_construct( clk_dce, ctx, regs, clk_shift, clk_mask); clk_dce->base.funcs = &dce112_funcs; @@ -794,10 +802,9 @@ struct display_clock *dce112_disp_clk_create( return &clk_dce->base; } -struct display_clock *dce120_disp_clk_create(struct dc_context *ctx) +struct dccg *dce120_dccg_create(struct dc_context *ctx) { - struct dce_disp_clk *clk_dce = kzalloc(sizeof(*clk_dce), GFP_KERNEL); - struct dm_pp_clock_levels_with_voltage clk_level_info = {0}; + struct dce_dccg *clk_dce = kzalloc(sizeof(*clk_dce), GFP_KERNEL); if (clk_dce == NULL) { BREAK_TO_DEBUGGER(); @@ -808,28 +815,59 @@ struct display_clock *dce120_disp_clk_create(struct dc_context *ctx) dce120_max_clks_by_state, sizeof(dce120_max_clks_by_state)); - dce_disp_clk_construct( + dce_dccg_construct( clk_dce, ctx, NULL, NULL, NULL); clk_dce->base.funcs = &dce120_funcs; - /* new in dce120 */ - if (!ctx->dc->debug.disable_pplib_clock_request && - dm_pp_get_clock_levels_by_type_with_voltage( - ctx, DM_PP_CLOCK_TYPE_DISPLAY_CLK, &clk_level_info) - && clk_level_info.num_levels) - clk_dce->max_displ_clk_in_khz = - clk_level_info.data[clk_level_info.num_levels - 1].clocks_in_khz; - else - clk_dce->max_displ_clk_in_khz = 1133000; + return &clk_dce->base; +} + +#ifdef CONFIG_X86 +struct dccg *dcn1_dccg_create(struct dc_context *ctx) +{ + struct dc_debug_options *debug = &ctx->dc->debug; + struct dc_bios *bp = ctx->dc_bios; + struct dc_firmware_info fw_info = { { 0 } }; + struct dce_dccg *clk_dce = kzalloc(sizeof(*clk_dce), GFP_KERNEL); + + if (clk_dce == NULL) { + BREAK_TO_DEBUGGER(); + return NULL; + } + + clk_dce->base.ctx = ctx; + clk_dce->base.funcs = &dcn1_funcs; + + clk_dce->dfs_bypass_disp_clk = 0; + + clk_dce->dprefclk_ss_percentage = 0; + clk_dce->dprefclk_ss_divider = 1000; + clk_dce->ss_on_dprefclk = false; + + if (bp->integrated_info) + clk_dce->dentist_vco_freq_khz = bp->integrated_info->dentist_vco_freq; + if (clk_dce->dentist_vco_freq_khz == 0) { + bp->funcs->get_firmware_info(bp, &fw_info); + clk_dce->dentist_vco_freq_khz = fw_info.smu_gpu_pll_output_freq; + if (clk_dce->dentist_vco_freq_khz == 0) + clk_dce->dentist_vco_freq_khz = 3600000; + } + + if (!debug->disable_dfs_bypass && bp->integrated_info) + if (bp->integrated_info->gpu_cap_info & DFS_BYPASS_ENABLE) + clk_dce->dfs_bypass_enabled = true; + + dce_clock_read_ss_info(clk_dce); return &clk_dce->base; } +#endif -void dce_disp_clk_destroy(struct display_clock **disp_clk) +void dce_dccg_destroy(struct dccg **dccg) { - struct dce_disp_clk *clk_dce = TO_DCE_CLOCKS(*disp_clk); + struct dce_dccg *clk_dce = TO_DCE_CLOCKS(*dccg); kfree(clk_dce); - *disp_clk = NULL; + *dccg = NULL; } diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_clocks.h b/drivers/gpu/drm/amd/display/dc/dce/dce_clocks.h index 0e717e0dc8f0..e5e44adc6c27 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_clocks.h +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_clocks.h @@ -33,6 +33,9 @@ .DPREFCLK_CNTL = mmDPREFCLK_CNTL, \ .DENTIST_DISPCLK_CNTL = mmDENTIST_DISPCLK_CNTL +#define CLK_COMMON_REG_LIST_DCN_BASE() \ + SR(DENTIST_DISPCLK_CNTL) + #define CLK_SF(reg_name, field_name, post_fix)\ .field_name = reg_name ## __ ## field_name ## post_fix @@ -40,58 +43,37 @@ CLK_SF(DPREFCLK_CNTL, DPREFCLK_SRC_SEL, mask_sh), \ CLK_SF(DENTIST_DISPCLK_CNTL, DENTIST_DPREFCLK_WDIVIDER, mask_sh) +#define CLK_COMMON_MASK_SH_LIST_DCN_COMMON_BASE(mask_sh) \ + CLK_SF(DENTIST_DISPCLK_CNTL, DENTIST_DISPCLK_WDIVIDER, mask_sh),\ + CLK_SF(DENTIST_DISPCLK_CNTL, DENTIST_DISPCLK_CHG_DONE, mask_sh) + #define CLK_REG_FIELD_LIST(type) \ type DPREFCLK_SRC_SEL; \ - type DENTIST_DPREFCLK_WDIVIDER; + type DENTIST_DPREFCLK_WDIVIDER; \ + type DENTIST_DISPCLK_WDIVIDER; \ + type DENTIST_DISPCLK_CHG_DONE; -struct dce_disp_clk_shift { +struct dccg_shift { CLK_REG_FIELD_LIST(uint8_t) }; -struct dce_disp_clk_mask { +struct dccg_mask { CLK_REG_FIELD_LIST(uint32_t) }; -struct dce_disp_clk_registers { +struct dccg_registers { uint32_t DPREFCLK_CNTL; uint32_t DENTIST_DISPCLK_CNTL; }; -/* Array identifiers and count for the divider ranges.*/ -enum dce_divider_range_count { - DIVIDER_RANGE_01 = 0, - DIVIDER_RANGE_02, - DIVIDER_RANGE_03, - DIVIDER_RANGE_MAX /* == 3*/ -}; - -enum dce_divider_error_types { - INVALID_DID = 0, - INVALID_DIVIDER = 1 -}; - -struct dce_divider_range { - int div_range_start; - /* The end of this range of dividers.*/ - int div_range_end; - /* The distance between each divider in this range.*/ - int div_range_step; - /* The divider id for the lowest divider.*/ - int did_min; - /* The divider id for the highest divider.*/ - int did_max; -}; - -struct dce_disp_clk { - struct display_clock base; - const struct dce_disp_clk_registers *regs; - const struct dce_disp_clk_shift *clk_shift; - const struct dce_disp_clk_mask *clk_mask; +struct dce_dccg { + struct dccg base; + const struct dccg_registers *regs; + const struct dccg_shift *clk_shift; + const struct dccg_mask *clk_mask; struct state_dependent_clocks max_clks_by_state[DM_PP_CLOCKS_MAX_STATES]; - struct dce_divider_range divider_ranges[DIVIDER_RANGE_MAX]; - bool use_max_disp_clk; int dentist_vco_freq_khz; /* Cache the status of DFS-bypass feature*/ @@ -106,32 +88,33 @@ struct dce_disp_clk { int dprefclk_ss_percentage; /* DPREFCLK SS percentage Divider (100 or 1000) */ int dprefclk_ss_divider; - - /* max disp_clk from PPLIB for max validation display clock*/ - int max_displ_clk_in_khz; }; -struct display_clock *dce_disp_clk_create( +struct dccg *dce_dccg_create( struct dc_context *ctx, - const struct dce_disp_clk_registers *regs, - const struct dce_disp_clk_shift *clk_shift, - const struct dce_disp_clk_mask *clk_mask); + const struct dccg_registers *regs, + const struct dccg_shift *clk_shift, + const struct dccg_mask *clk_mask); -struct display_clock *dce110_disp_clk_create( +struct dccg *dce110_dccg_create( struct dc_context *ctx, - const struct dce_disp_clk_registers *regs, - const struct dce_disp_clk_shift *clk_shift, - const struct dce_disp_clk_mask *clk_mask); + const struct dccg_registers *regs, + const struct dccg_shift *clk_shift, + const struct dccg_mask *clk_mask); -struct display_clock *dce112_disp_clk_create( +struct dccg *dce112_dccg_create( struct dc_context *ctx, - const struct dce_disp_clk_registers *regs, - const struct dce_disp_clk_shift *clk_shift, - const struct dce_disp_clk_mask *clk_mask); + const struct dccg_registers *regs, + const struct dccg_shift *clk_shift, + const struct dccg_mask *clk_mask); + +struct dccg *dce120_dccg_create(struct dc_context *ctx); -struct display_clock *dce120_disp_clk_create(struct dc_context *ctx); +#ifdef CONFIG_X86 +struct dccg *dcn1_dccg_create(struct dc_context *ctx); +#endif -void dce_disp_clk_destroy(struct display_clock **disp_clk); +void dce_dccg_destroy(struct dccg **dccg); #endif /* _DCE_CLOCKS_H_ */ diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.c b/drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.c index a576b8bbb3cd..ca7989e4932b 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.c @@ -150,7 +150,7 @@ static void dce_dmcu_set_psr_enable(struct dmcu *dmcu, bool enable, bool wait) } } -static void dce_dmcu_setup_psr(struct dmcu *dmcu, +static bool dce_dmcu_setup_psr(struct dmcu *dmcu, struct dc_link *link, struct psr_context *psr_context) { @@ -261,6 +261,8 @@ static void dce_dmcu_setup_psr(struct dmcu *dmcu, /* notifyDMCUMsg */ REG_UPDATE(MASTER_COMM_CNTL_REG, MASTER_COMM_INTERRUPT, 1); + + return true; } static bool dce_is_dmcu_initialized(struct dmcu *dmcu) @@ -314,7 +316,7 @@ static void dce_get_psr_wait_loop( return; } -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 static void dcn10_get_dmcu_state(struct dmcu *dmcu) { struct dce_dmcu *dmcu_dce = TO_DCE_DMCU(dmcu); @@ -545,24 +547,25 @@ static void dcn10_dmcu_set_psr_enable(struct dmcu *dmcu, bool enable, bool wait) * least a few frames. Should never hit the max retry assert below. */ if (wait == true) { - for (retryCount = 0; retryCount <= 1000; retryCount++) { - dcn10_get_dmcu_psr_state(dmcu, &psr_state); - if (enable) { - if (psr_state != 0) - break; - } else { - if (psr_state == 0) - break; + for (retryCount = 0; retryCount <= 1000; retryCount++) { + dcn10_get_dmcu_psr_state(dmcu, &psr_state); + if (enable) { + if (psr_state != 0) + break; + } else { + if (psr_state == 0) + break; + } + udelay(500); } - udelay(500); - } - /* assert if max retry hit */ - ASSERT(retryCount <= 1000); + /* assert if max retry hit */ + if (retryCount >= 1000) + ASSERT(0); } } -static void dcn10_dmcu_setup_psr(struct dmcu *dmcu, +static bool dcn10_dmcu_setup_psr(struct dmcu *dmcu, struct dc_link *link, struct psr_context *psr_context) { @@ -577,7 +580,7 @@ static void dcn10_dmcu_setup_psr(struct dmcu *dmcu, /* If microcontroller is not running, do nothing */ if (dmcu->dmcu_state != DMCU_RUNNING) - return; + return false; link->link_enc->funcs->psr_program_dp_dphy_fast_training(link->link_enc, psr_context->psrExitLinkTrainingRequired); @@ -677,6 +680,11 @@ static void dcn10_dmcu_setup_psr(struct dmcu *dmcu, /* notifyDMCUMsg */ REG_UPDATE(MASTER_COMM_CNTL_REG, MASTER_COMM_INTERRUPT, 1); + + /* waitDMCUReadyForCmd */ + REG_WAIT(MASTER_COMM_CNTL_REG, MASTER_COMM_INTERRUPT, 0, 1, 10000); + + return true; } static void dcn10_psr_wait_loop( @@ -735,7 +743,7 @@ static const struct dmcu_funcs dce_funcs = { .is_dmcu_initialized = dce_is_dmcu_initialized }; -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 static const struct dmcu_funcs dcn10_funcs = { .dmcu_init = dcn10_dmcu_init, .load_iram = dcn10_dmcu_load_iram, @@ -787,7 +795,7 @@ struct dmcu *dce_dmcu_create( return &dmcu_dce->base; } -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 struct dmcu *dcn10_dmcu_create( struct dc_context *ctx, const struct dce_dmcu_registers *regs, diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_hwseq.h b/drivers/gpu/drm/amd/display/dc/dce/dce_hwseq.h index 057b8afd74bc..64dc75378541 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_hwseq.h +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_hwseq.h @@ -147,6 +147,7 @@ SR(DCCG_GATE_DISABLE_CNTL2), \ SR(DCFCLK_CNTL),\ SR(DCFCLK_CNTL), \ + SR(DC_MEM_GLOBAL_PWR_REQ_CNTL), \ /* todo: get these from GVM instead of reading registers ourselves */\ MMHUB_SR(VM_CONTEXT0_PAGE_TABLE_BASE_ADDR_HI32),\ MMHUB_SR(VM_CONTEXT0_PAGE_TABLE_BASE_ADDR_LO32),\ @@ -249,7 +250,6 @@ struct dce_hwseq_registers { uint32_t DISPCLK_FREQ_CHANGE_CNTL; uint32_t RBBMIF_TIMEOUT_DIS; uint32_t RBBMIF_TIMEOUT_DIS_2; - uint32_t DENTIST_DISPCLK_CNTL; uint32_t DCHUBBUB_CRC_CTRL; uint32_t DPP_TOP0_DPP_CRC_CTRL; uint32_t DPP_TOP0_DPP_CRC_VAL_R_G; @@ -276,6 +276,8 @@ struct dce_hwseq_registers { uint32_t MC_VM_SYSTEM_APERTURE_DEFAULT_ADDR_LSB; uint32_t MC_VM_SYSTEM_APERTURE_LOW_ADDR; uint32_t MC_VM_SYSTEM_APERTURE_HIGH_ADDR; + uint32_t AZALIA_AUDIO_DTO; + uint32_t AZALIA_CONTROLLER_CLOCK_GATING; }; /* set field name */ #define HWS_SF(blk_name, reg_name, field_name, post_fix)\ @@ -362,7 +364,8 @@ struct dce_hwseq_registers { HWSEQ_PIXEL_RATE_MASK_SH_LIST(mask_sh, OTG0_),\ HWS_SF1(OTG0_, PHYPLL_PIXEL_RATE_CNTL, PHYPLL_PIXEL_RATE_SOURCE, mask_sh), \ HWS_SF(, DCHUBBUB_GLOBAL_TIMER_CNTL, DCHUBBUB_GLOBAL_TIMER_ENABLE, mask_sh), \ - HWS_SF(, DCFCLK_CNTL, DCFCLK_GATE_DIS, mask_sh) + HWS_SF(, DCFCLK_CNTL, DCFCLK_GATE_DIS, mask_sh), \ + HWS_SF(, DC_MEM_GLOBAL_PWR_REQ_CNTL, DC_MEM_GLOBAL_PWR_REQ_DIS, mask_sh) #define HWSEQ_DCN1_MASK_SH_LIST(mask_sh)\ HWSEQ_DCN_MASK_SH_LIST(mask_sh), \ @@ -496,14 +499,13 @@ struct dce_hwseq_registers { type DOMAIN7_PGFSM_PWR_STATUS; \ type DCFCLK_GATE_DIS; \ type DCHUBBUB_GLOBAL_TIMER_REFDIV; \ - type DENTIST_DPPCLK_WDIVIDER; \ - type DENTIST_DISPCLK_WDIVIDER; \ type VGA_TEST_ENABLE; \ type VGA_TEST_RENDER_START; \ type D1VGA_MODE_ENABLE; \ type D2VGA_MODE_ENABLE; \ type D3VGA_MODE_ENABLE; \ - type D4VGA_MODE_ENABLE; + type D4VGA_MODE_ENABLE; \ + type AZALIA_AUDIO_DTO_MODULE; struct dce_hwseq_shift { HWSEQ_REG_FIELD_LIST(uint8_t) diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_link_encoder.c b/drivers/gpu/drm/amd/display/dc/dce/dce_link_encoder.c index dbe3b26b6d9e..752b3d62e793 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_link_encoder.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_link_encoder.c @@ -256,6 +256,11 @@ static void setup_panel_mode( enum dp_panel_mode panel_mode) { uint32_t value; + struct dc_context *ctx = enc110->base.ctx; + + /* if psp set panel mode, dal should be program it */ + if (ctx->dc->caps.psp_setup_panel_mode) + return; ASSERT(REG(DP_DPHY_INTERNAL_CTRL)); value = REG_READ(DP_DPHY_INTERNAL_CTRL); @@ -646,6 +651,9 @@ static bool dce110_link_encoder_validate_hdmi_output( if (!enc110->base.features.flags.bits.HDMI_6GB_EN && adjusted_pix_clk_khz >= 300000) return false; + if (enc110->base.ctx->dc->debug.hdmi20_disable && + crtc_timing->pixel_encoding == PIXEL_ENCODING_YCBCR420) + return false; return true; } @@ -773,6 +781,9 @@ void dce110_link_encoder_construct( __func__, result); } + if (enc110->base.ctx->dc->debug.hdmi20_disable) { + enc110->base.features.flags.bits.HDMI_6GB_EN = 0; + } } bool dce110_link_encoder_validate_output_with_stream( diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.c b/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.c index bae752332a9f..85686d917636 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.c @@ -729,7 +729,7 @@ static bool dce_mi_program_surface_flip_and_addr( return true; } -static struct mem_input_funcs dce_mi_funcs = { +static const struct mem_input_funcs dce_mi_funcs = { .mem_input_program_display_marks = dce_mi_program_display_marks, .allocate_mem_input = dce_mi_allocate_dmif, .free_mem_input = dce_mi_free_dmif, @@ -741,7 +741,7 @@ static struct mem_input_funcs dce_mi_funcs = { .mem_input_is_flip_pending = dce_mi_is_flip_pending }; -static struct mem_input_funcs dce112_mi_funcs = { +static const struct mem_input_funcs dce112_mi_funcs = { .mem_input_program_display_marks = dce112_mi_program_display_marks, .allocate_mem_input = dce_mi_allocate_dmif, .free_mem_input = dce_mi_free_dmif, @@ -753,7 +753,7 @@ static struct mem_input_funcs dce112_mi_funcs = { .mem_input_is_flip_pending = dce_mi_is_flip_pending }; -static struct mem_input_funcs dce120_mi_funcs = { +static const struct mem_input_funcs dce120_mi_funcs = { .mem_input_program_display_marks = dce120_mi_program_display_marks, .allocate_mem_input = dce_mi_allocate_dmif, .free_mem_input = dce_mi_free_dmif, diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c index c0e813c7ddd4..b139b4017820 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c @@ -135,7 +135,7 @@ static void dce110_update_generic_info_packet( AFMT_GENERIC0_UPDATE, (packet_index == 0), AFMT_GENERIC2_UPDATE, (packet_index == 2)); } -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 if (REG(AFMT_VBI_PACKET_CONTROL1)) { switch (packet_index) { case 0: @@ -229,7 +229,7 @@ static void dce110_update_hdmi_info_packet( HDMI_GENERIC1_SEND, send, HDMI_GENERIC1_LINE, line); break; -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 case 4: if (REG(HDMI_GENERIC_PACKET_CONTROL2)) REG_UPDATE_3(HDMI_GENERIC_PACKET_CONTROL2, @@ -274,7 +274,7 @@ static void dce110_stream_encoder_dp_set_stream_attribute( struct dc_crtc_timing *crtc_timing, enum dc_color_space output_color_space) { -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 uint32_t h_active_start; uint32_t v_active_start; uint32_t misc0 = 0; @@ -289,11 +289,6 @@ static void dce110_stream_encoder_dp_set_stream_attribute( struct dce110_stream_encoder *enc110 = DCE110STRENC_FROM_STRENC(enc); -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) - if (REG(DP_DB_CNTL)) - REG_UPDATE(DP_DB_CNTL, DP_DB_DISABLE, 1); -#endif - /* set pixel encoding */ switch (crtc_timing->pixel_encoding) { case PIXEL_ENCODING_YCBCR422: @@ -322,7 +317,7 @@ static void dce110_stream_encoder_dp_set_stream_attribute( if (enc110->se_mask->DP_VID_M_DOUBLE_VALUE_EN) REG_UPDATE(DP_VID_TIMING, DP_VID_M_DOUBLE_VALUE_EN, 1); -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 if (enc110->se_mask->DP_VID_N_MUL) REG_UPDATE(DP_VID_TIMING, DP_VID_N_MUL, 1); #endif @@ -333,7 +328,7 @@ static void dce110_stream_encoder_dp_set_stream_attribute( break; } -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 if (REG(DP_MSA_MISC)) misc1 = REG_READ(DP_MSA_MISC); #endif @@ -367,7 +362,7 @@ static void dce110_stream_encoder_dp_set_stream_attribute( /* set dynamic range and YCbCr range */ -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 switch (crtc_timing->display_color_depth) { case COLOR_DEPTH_666: colorimetry_bpc = 0; @@ -446,7 +441,7 @@ static void dce110_stream_encoder_dp_set_stream_attribute( DP_DYN_RANGE, dynamic_range_rgb, DP_YCBCR_RANGE, dynamic_range_ycbcr); -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 if (REG(DP_MSA_COLORIMETRY)) REG_SET(DP_MSA_COLORIMETRY, 0, DP_MSA_MISC0, misc0); @@ -481,7 +476,7 @@ static void dce110_stream_encoder_dp_set_stream_attribute( crtc_timing->v_front_porch; -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 /* start at begining of left border */ if (REG(DP_MSA_TIMING_PARAM2)) REG_SET_2(DP_MSA_TIMING_PARAM2, 0, @@ -756,7 +751,7 @@ static void dce110_stream_encoder_update_hdmi_info_packets( dce110_update_hdmi_info_packet(enc110, 3, &info_frame->hdrsmd); } -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 if (enc110->se_mask->HDMI_DB_DISABLE) { /* for bring up, disable dp double TODO */ if (REG(HDMI_DB_CONTROL)) @@ -794,7 +789,7 @@ static void dce110_stream_encoder_stop_hdmi_info_packets( HDMI_GENERIC1_LINE, 0, HDMI_GENERIC1_SEND, 0); -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 /* stop generic packets 2 & 3 on HDMI */ if (REG(HDMI_GENERIC_PACKET_CONTROL2)) REG_SET_6(HDMI_GENERIC_PACKET_CONTROL2, 0, diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_transform.c b/drivers/gpu/drm/amd/display/dc/dce/dce_transform.c index a02e719d7794..ab63d0d0304c 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_transform.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_transform.c @@ -155,7 +155,7 @@ static void program_overscan( int overscan_bottom = data->v_active - data->recout.y - data->recout.height; - if (xfm_dce->base.ctx->dc->debug.surface_visual_confirm) { + if (xfm_dce->base.ctx->dc->debug.visual_confirm != VISUAL_CONFIRM_DISABLE) { overscan_bottom += 2; overscan_right += 2; } diff --git a/drivers/gpu/drm/amd/display/dc/dce100/dce100_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dce100/dce100_hw_sequencer.c index 41f83ecd7469..74c05e878807 100644 --- a/drivers/gpu/drm/amd/display/dc/dce100/dce100_hw_sequencer.c +++ b/drivers/gpu/drm/amd/display/dc/dce100/dce100_hw_sequencer.c @@ -125,17 +125,50 @@ static void dce100_pplib_apply_display_requirements( dc->prev_display_config = *pp_display_cfg; } +/* unit: in_khz before mode set, get pixel clock from context. ASIC register + * may not be programmed yet + */ +static uint32_t get_max_pixel_clock_for_all_paths( + struct dc *dc, + struct dc_state *context) +{ + uint32_t max_pix_clk = 0; + int i; + + for (i = 0; i < MAX_PIPES; i++) { + struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i]; + + if (pipe_ctx->stream == NULL) + continue; + + /* do not check under lay */ + if (pipe_ctx->top_pipe) + continue; + + if (pipe_ctx->stream_res.pix_clk_params.requested_pix_clk > max_pix_clk) + max_pix_clk = + pipe_ctx->stream_res.pix_clk_params.requested_pix_clk; + } + return max_pix_clk; +} + void dce100_set_bandwidth( struct dc *dc, struct dc_state *context, bool decrease_allowed) { - if (decrease_allowed || context->bw.dce.dispclk_khz > dc->current_state->bw.dce.dispclk_khz) { - dc->res_pool->display_clock->funcs->set_clock( - dc->res_pool->display_clock, - context->bw.dce.dispclk_khz * 115 / 100); - dc->current_state->bw.dce.dispclk_khz = context->bw.dce.dispclk_khz; - } + struct dc_clocks req_clks; + + req_clks.dispclk_khz = context->bw.dce.dispclk_khz * 115 / 100; + req_clks.phyclk_khz = get_max_pixel_clock_for_all_paths(dc, context); + + dce110_set_safe_displaymarks(&context->res_ctx, dc->res_pool); + + dc->res_pool->dccg->funcs->update_clocks( + dc->res_pool->dccg, + &req_clks, + decrease_allowed); + dce100_pplib_apply_display_requirements(dc, context); } diff --git a/drivers/gpu/drm/amd/display/dc/dce100/dce100_resource.c b/drivers/gpu/drm/amd/display/dc/dce100/dce100_resource.c index 344dd2e69e7c..3f76e6019546 100644 --- a/drivers/gpu/drm/amd/display/dc/dce100/dce100_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dce100/dce100_resource.c @@ -52,6 +52,7 @@ #include "dce/dce_10_0_sh_mask.h" #include "dce/dce_dmcu.h" +#include "dce/dce_aux.h" #include "dce/dce_abm.h" #ifndef mmMC_HUB_RDREQ_DMIF_LIMIT @@ -135,15 +136,15 @@ static const struct dce110_timing_generator_offsets dce100_tg_offsets[] = { .reg_name = mm ## block ## id ## _ ## reg_name -static const struct dce_disp_clk_registers disp_clk_regs = { +static const struct dccg_registers disp_clk_regs = { CLK_COMMON_REG_LIST_DCE_BASE() }; -static const struct dce_disp_clk_shift disp_clk_shift = { +static const struct dccg_shift disp_clk_shift = { CLK_COMMON_MASK_SH_LIST_DCE_COMMON_BASE(__SHIFT) }; -static const struct dce_disp_clk_mask disp_clk_mask = { +static const struct dccg_mask disp_clk_mask = { CLK_COMMON_MASK_SH_LIST_DCE_COMMON_BASE(_MASK) }; @@ -279,7 +280,20 @@ static const struct dce_opp_shift opp_shift = { static const struct dce_opp_mask opp_mask = { OPP_COMMON_MASK_SH_LIST_DCE_100(_MASK) }; +#define aux_engine_regs(id)\ +[id] = {\ + AUX_COMMON_REG_LIST(id), \ + .AUX_RESET_MASK = 0 \ +} +static const struct dce110_aux_registers aux_engine_regs[] = { + aux_engine_regs(0), + aux_engine_regs(1), + aux_engine_regs(2), + aux_engine_regs(3), + aux_engine_regs(4), + aux_engine_regs(5) +}; #define audio_regs(id)\ [id] = {\ @@ -572,6 +586,23 @@ struct output_pixel_processor *dce100_opp_create( return &opp->base; } +struct aux_engine *dce100_aux_engine_create( + struct dc_context *ctx, + uint32_t inst) +{ + struct aux_engine_dce110 *aux_engine = + kzalloc(sizeof(struct aux_engine_dce110), GFP_KERNEL); + + if (!aux_engine) + return NULL; + + dce110_aux_engine_construct(aux_engine, ctx, inst, + SW_AUX_TIMEOUT_PERIOD_MULTIPLIER * AUX_TIMEOUT_PERIOD, + &aux_engine_regs[inst]); + + return &aux_engine->base; +} + struct clock_source *dce100_clock_source_create( struct dc_context *ctx, struct dc_bios *bios, @@ -624,6 +655,10 @@ static void destruct(struct dce110_resource_pool *pool) kfree(DCE110TG_FROM_TG(pool->base.timing_generators[i])); pool->base.timing_generators[i] = NULL; } + + if (pool->base.engines[i] != NULL) + dce110_engine_destroy(&pool->base.engines[i]); + } for (i = 0; i < pool->base.stream_enc_count; i++) { @@ -644,8 +679,8 @@ static void destruct(struct dce110_resource_pool *pool) dce_aud_destroy(&pool->base.audios[i]); } - if (pool->base.display_clock != NULL) - dce_disp_clk_destroy(&pool->base.display_clock); + if (pool->base.dccg != NULL) + dce_dccg_destroy(&pool->base.dccg); if (pool->base.abm != NULL) dce_abm_destroy(&pool->base.abm); @@ -830,11 +865,11 @@ static bool construct( } } - pool->base.display_clock = dce_disp_clk_create(ctx, + pool->base.dccg = dce_dccg_create(ctx, &disp_clk_regs, &disp_clk_shift, &disp_clk_mask); - if (pool->base.display_clock == NULL) { + if (pool->base.dccg == NULL) { dm_error("DC: failed to create display clock!\n"); BREAK_TO_DEBUGGER(); goto res_create_fail; @@ -864,7 +899,7 @@ static bool construct( * max_clock_state */ if (dm_pp_get_static_clocks(ctx, &static_clk_info)) - pool->base.display_clock->max_clks_state = + pool->base.dccg->max_clks_state = static_clk_info.max_clocks_state; { struct irq_service_init_data init_data; @@ -884,7 +919,7 @@ static bool construct( dc->caps.i2c_speed_in_khz = 40; dc->caps.max_cursor_size = 128; dc->caps.dual_link_dvi = true; - + dc->caps.disable_dp_clk_share = true; for (i = 0; i < pool->base.pipe_count; i++) { pool->base.timing_generators[i] = dce100_timing_generator_create( @@ -928,6 +963,13 @@ static bool construct( "DC: failed to create output pixel processor!\n"); goto res_create_fail; } + pool->base.engines[i] = dce100_aux_engine_create(ctx, i); + if (pool->base.engines[i] == NULL) { + BREAK_TO_DEBUGGER(); + dm_error( + "DC:failed to create aux engine!!\n"); + goto res_create_fail; + } } dc->caps.max_planes = pool->base.pipe_count; diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_compressor.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_compressor.c index e2994d337044..1f7f25013217 100644 --- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_compressor.c +++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_compressor.c @@ -143,7 +143,7 @@ static void wait_for_fbc_state_changed( struct dce110_compressor *cp110, bool enabled) { - uint8_t counter = 0; + uint32_t counter = 0; uint32_t addr = mmFBC_STATUS; uint32_t value; @@ -158,7 +158,7 @@ static void wait_for_fbc_state_changed( counter++; } - if (counter == 10) { + if (counter == 1000) { DC_LOG_WARNING("%s: wait counter exceeded, changes to HW not applied", __func__); } else { @@ -551,9 +551,7 @@ void dce110_compressor_construct(struct dce110_compressor *compressor, compressor->base.lpt_channels_num = 0; compressor->base.attached_inst = 0; compressor->base.is_enabled = false; -#if defined(CONFIG_DRM_AMD_DC_FBC) compressor->base.funcs = &dce110_compressor_funcs; -#endif } diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c index c29052b6da5a..1d98e3678b04 100644 --- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c +++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c @@ -34,9 +34,7 @@ #include "dce/dce_hwseq.h" #include "gpio_service_interface.h" -#if defined(CONFIG_DRM_AMD_DC_FBC) #include "dce110_compressor.h" -#endif #include "bios/bios_parser_helper.h" #include "timing_generator.h" @@ -667,16 +665,25 @@ static enum dc_status bios_parser_crtc_source_select( void dce110_update_info_frame(struct pipe_ctx *pipe_ctx) { + bool is_hdmi; + bool is_dp; + ASSERT(pipe_ctx->stream); if (pipe_ctx->stream_res.stream_enc == NULL) return; /* this is not root pipe */ - if (dc_is_hdmi_signal(pipe_ctx->stream->signal)) + is_hdmi = dc_is_hdmi_signal(pipe_ctx->stream->signal); + is_dp = dc_is_dp_signal(pipe_ctx->stream->signal); + + if (!is_hdmi && !is_dp) + return; + + if (is_hdmi) pipe_ctx->stream_res.stream_enc->funcs->update_hdmi_info_packets( pipe_ctx->stream_res.stream_enc, &pipe_ctx->stream_res.encoder_info_frame); - else if (dc_is_dp_signal(pipe_ctx->stream->signal)) + else pipe_ctx->stream_res.stream_enc->funcs->update_dp_info_packets( pipe_ctx->stream_res.stream_enc, &pipe_ctx->stream_res.encoder_info_frame); @@ -857,17 +864,22 @@ void hwss_edp_power_control( if (power_up) { unsigned long long current_ts = dm_get_timestamp(ctx); unsigned long long duration_in_ms = - dm_get_elapse_time_in_ns( + div64_u64(dm_get_elapse_time_in_ns( ctx, current_ts, - div64_u64(link->link_trace.time_stamp.edp_poweroff, 1000000)); + link->link_trace.time_stamp.edp_poweroff), 1000000); unsigned long long wait_time_ms = 0; /* max 500ms from LCDVDD off to on */ + unsigned long long edp_poweroff_time_ms = 500; + + if (link->local_sink != NULL) + edp_poweroff_time_ms = + 500 + link->local_sink->edid_caps.panel_patch.extra_t12_ms; if (link->link_trace.time_stamp.edp_poweroff == 0) - wait_time_ms = 500; - else if (duration_in_ms < 500) - wait_time_ms = 500 - duration_in_ms; + wait_time_ms = edp_poweroff_time_ms; + else if (duration_in_ms < edp_poweroff_time_ms) + wait_time_ms = edp_poweroff_time_ms - duration_in_ms; if (wait_time_ms) { msleep(wait_time_ms); @@ -972,19 +984,35 @@ void hwss_edp_backlight_control( edp_receiver_ready_T9(link); } -void dce110_disable_stream(struct pipe_ctx *pipe_ctx, int option) +void dce110_enable_audio_stream(struct pipe_ctx *pipe_ctx) { - struct dc_stream_state *stream = pipe_ctx->stream; - struct dc_link *link = stream->sink->link; - struct dc *dc = pipe_ctx->stream->ctx->dc; + struct dc *core_dc = pipe_ctx->stream->ctx->dc; + /* notify audio driver for audio modes of monitor */ + struct pp_smu_funcs_rv *pp_smu = core_dc->res_pool->pp_smu; + unsigned int i, num_audio = 1; - if (dc_is_hdmi_signal(pipe_ctx->stream->signal)) - pipe_ctx->stream_res.stream_enc->funcs->stop_hdmi_info_packets( - pipe_ctx->stream_res.stream_enc); + if (pipe_ctx->stream_res.audio) { + for (i = 0; i < MAX_PIPES; i++) { + /*current_state not updated yet*/ + if (core_dc->current_state->res_ctx.pipe_ctx[i].stream_res.audio != NULL) + num_audio++; + } - if (dc_is_dp_signal(pipe_ctx->stream->signal)) - pipe_ctx->stream_res.stream_enc->funcs->stop_dp_info_packets( - pipe_ctx->stream_res.stream_enc); + pipe_ctx->stream_res.audio->funcs->az_enable(pipe_ctx->stream_res.audio); + + if (num_audio == 1 && pp_smu != NULL && pp_smu->set_pme_wa_enable != NULL) + /*this is the first audio. apply the PME w/a in order to wake AZ from D3*/ + pp_smu->set_pme_wa_enable(&pp_smu->pp_smu); + /* un-mute audio */ + /* TODO: audio should be per stream rather than per link */ + pipe_ctx->stream_res.stream_enc->funcs->audio_mute_control( + pipe_ctx->stream_res.stream_enc, false); + } +} + +void dce110_disable_audio_stream(struct pipe_ctx *pipe_ctx, int option) +{ + struct dc *dc = pipe_ctx->stream->ctx->dc; pipe_ctx->stream_res.stream_enc->funcs->audio_mute_control( pipe_ctx->stream_res.stream_enc, true); @@ -1015,7 +1043,23 @@ void dce110_disable_stream(struct pipe_ctx *pipe_ctx, int option) * stream->stream_engine_id); */ } +} +void dce110_disable_stream(struct pipe_ctx *pipe_ctx, int option) +{ + struct dc_stream_state *stream = pipe_ctx->stream; + struct dc_link *link = stream->sink->link; + struct dc *dc = pipe_ctx->stream->ctx->dc; + + if (dc_is_hdmi_signal(pipe_ctx->stream->signal)) + pipe_ctx->stream_res.stream_enc->funcs->stop_hdmi_info_packets( + pipe_ctx->stream_res.stream_enc); + + if (dc_is_dp_signal(pipe_ctx->stream->signal)) + pipe_ctx->stream_res.stream_enc->funcs->stop_dp_info_packets( + pipe_ctx->stream_res.stream_enc); + + dc->hwss.disable_audio_stream(pipe_ctx, option); link->link_enc->funcs->connect_dig_be_to_fe( link->link_enc, @@ -1206,13 +1250,13 @@ static void program_scaler(const struct dc *dc, { struct tg_color color = {0}; -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 /* TOFPGA */ if (pipe_ctx->plane_res.xfm->funcs->transform_set_pixel_storage_depth == NULL) return; #endif - if (dc->debug.surface_visual_confirm) + if (dc->debug.visual_confirm == VISUAL_CONFIRM_SURFACE) get_surface_visual_confirm_color(pipe_ctx, &color); else color_space_to_black_color(dc, @@ -1298,6 +1342,30 @@ static enum dc_status apply_single_controller_ctx_to_hw( struct pipe_ctx *pipe_ctx_old = &dc->current_state->res_ctx. pipe_ctx[pipe_ctx->pipe_idx]; + if (pipe_ctx->stream_res.audio != NULL) { + struct audio_output audio_output; + + build_audio_output(context, pipe_ctx, &audio_output); + + if (dc_is_dp_signal(pipe_ctx->stream->signal)) + pipe_ctx->stream_res.stream_enc->funcs->dp_audio_setup( + pipe_ctx->stream_res.stream_enc, + pipe_ctx->stream_res.audio->inst, + &pipe_ctx->stream->audio_info); + else + pipe_ctx->stream_res.stream_enc->funcs->hdmi_audio_setup( + pipe_ctx->stream_res.stream_enc, + pipe_ctx->stream_res.audio->inst, + &pipe_ctx->stream->audio_info, + &audio_output.crtc_info); + + pipe_ctx->stream_res.audio->funcs->az_configure( + pipe_ctx->stream_res.audio, + pipe_ctx->stream->signal, + &audio_output.crtc_info, + &pipe_ctx->stream->audio_info); + } + /* */ dc->hwss.enable_stream_timing(pipe_ctx, context, dc); @@ -1412,7 +1480,7 @@ static void power_down_controllers(struct dc *dc) { int i; - for (i = 0; i < dc->res_pool->pipe_count; i++) { + for (i = 0; i < dc->res_pool->timing_generator_count; i++) { dc->res_pool->timing_generators[i]->funcs->disable_crtc( dc->res_pool->timing_generators[i]); } @@ -1441,10 +1509,8 @@ static void power_down_all_hw_blocks(struct dc *dc) power_down_clock_sources(dc); -#if defined(CONFIG_DRM_AMD_DC_FBC) if (dc->fbc_compressor) dc->fbc_compressor->funcs->disable_fbc(dc->fbc_compressor); -#endif } static void disable_vga_and_power_gate_all_controllers( @@ -1454,12 +1520,13 @@ static void disable_vga_and_power_gate_all_controllers( struct timing_generator *tg; struct dc_context *ctx = dc->ctx; - for (i = 0; i < dc->res_pool->pipe_count; i++) { + for (i = 0; i < dc->res_pool->timing_generator_count; i++) { tg = dc->res_pool->timing_generators[i]; if (tg->funcs->disable_vga) tg->funcs->disable_vga(tg); - + } + for (i = 0; i < dc->res_pool->pipe_count; i++) { /* Enable CLOCK gating for each pipe BEFORE controller * powergating. */ enable_display_pipe_clock_gating(ctx, @@ -1602,7 +1669,7 @@ static void dce110_set_displaymarks( } } -static void set_safe_displaymarks( +void dce110_set_safe_displaymarks( struct resource_context *res_ctx, const struct resource_pool *pool) { @@ -1686,9 +1753,7 @@ static void set_static_screen_control(struct pipe_ctx **pipe_ctx, if (events->force_trigger) value |= 0x1; -#if defined(CONFIG_DRM_AMD_DC_FBC) value |= 0x84; -#endif for (i = 0; i < num_pipes; i++) pipe_ctx[i]->stream_res.tg->funcs-> @@ -1696,23 +1761,15 @@ static void set_static_screen_control(struct pipe_ctx **pipe_ctx, } /* unit: in_khz before mode set, get pixel clock from context. ASIC register - * may not be programmed yet. - * TODO: after mode set, pre_mode_set = false, - * may read PLL register to get pixel clock + * may not be programmed yet */ static uint32_t get_max_pixel_clock_for_all_paths( struct dc *dc, - struct dc_state *context, - bool pre_mode_set) + struct dc_state *context) { uint32_t max_pix_clk = 0; int i; - if (!pre_mode_set) { - /* TODO: read ASIC register to get pixel clock */ - ASSERT(0); - } - for (i = 0; i < MAX_PIPES; i++) { struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i]; @@ -1728,97 +1785,10 @@ static uint32_t get_max_pixel_clock_for_all_paths( pipe_ctx->stream_res.pix_clk_params.requested_pix_clk; } - if (max_pix_clk == 0) - ASSERT(0); - return max_pix_clk; } /* - * Find clock state based on clock requested. if clock value is 0, simply - * set clock state as requested without finding clock state by clock value - */ - -static void apply_min_clocks( - struct dc *dc, - struct dc_state *context, - enum dm_pp_clocks_state *clocks_state, - bool pre_mode_set) -{ - struct state_dependent_clocks req_clocks = {0}; - - if (!pre_mode_set) { - /* set clock_state without verification */ - if (context->dis_clk->funcs->set_min_clocks_state) { - context->dis_clk->funcs->set_min_clocks_state( - context->dis_clk, *clocks_state); - return; - } - - /* TODO: This is incorrect. Figure out how to fix. */ - context->dis_clk->funcs->apply_clock_voltage_request( - context->dis_clk, - DM_PP_CLOCK_TYPE_DISPLAY_CLK, - context->dis_clk->cur_clocks_value.dispclk_in_khz, - pre_mode_set, - false); - - context->dis_clk->funcs->apply_clock_voltage_request( - context->dis_clk, - DM_PP_CLOCK_TYPE_PIXELCLK, - context->dis_clk->cur_clocks_value.max_pixelclk_in_khz, - pre_mode_set, - false); - - context->dis_clk->funcs->apply_clock_voltage_request( - context->dis_clk, - DM_PP_CLOCK_TYPE_DISPLAYPHYCLK, - context->dis_clk->cur_clocks_value.max_non_dp_phyclk_in_khz, - pre_mode_set, - false); - return; - } - - /* get the required state based on state dependent clocks: - * display clock and pixel clock - */ - req_clocks.display_clk_khz = context->bw.dce.dispclk_khz; - - req_clocks.pixel_clk_khz = get_max_pixel_clock_for_all_paths( - dc, context, true); - - if (context->dis_clk->funcs->get_required_clocks_state) { - *clocks_state = context->dis_clk->funcs->get_required_clocks_state( - context->dis_clk, &req_clocks); - context->dis_clk->funcs->set_min_clocks_state( - context->dis_clk, *clocks_state); - } else { - context->dis_clk->funcs->apply_clock_voltage_request( - context->dis_clk, - DM_PP_CLOCK_TYPE_DISPLAY_CLK, - req_clocks.display_clk_khz, - pre_mode_set, - false); - - context->dis_clk->funcs->apply_clock_voltage_request( - context->dis_clk, - DM_PP_CLOCK_TYPE_PIXELCLK, - req_clocks.pixel_clk_khz, - pre_mode_set, - false); - - context->dis_clk->funcs->apply_clock_voltage_request( - context->dis_clk, - DM_PP_CLOCK_TYPE_DISPLAYPHYCLK, - req_clocks.pixel_clk_khz, - pre_mode_set, - false); - } -} - -#if defined(CONFIG_DRM_AMD_DC_FBC) - -/* * Check if FBC can be enabled */ static bool should_enable_fbc(struct dc *dc, @@ -1896,7 +1866,6 @@ static void enable_fbc(struct dc *dc, compr->funcs->enable_fbc(compr, ¶ms); } } -#endif static void dce110_reset_hw_ctx_wrap( struct dc *dc, @@ -1949,97 +1918,12 @@ static void dce110_reset_hw_ctx_wrap( } } - -enum dc_status dce110_apply_ctx_to_hw( +static void dce110_setup_audio_dto( struct dc *dc, struct dc_state *context) { - struct dc_bios *dcb = dc->ctx->dc_bios; - enum dc_status status; int i; - enum dm_pp_clocks_state clocks_state = DM_PP_CLOCKS_STATE_INVALID; - - /* Reset old context */ - /* look up the targets that have been removed since last commit */ - dc->hwss.reset_hw_ctx_wrap(dc, context); - - /* Skip applying if no targets */ - if (context->stream_count <= 0) - return DC_OK; - - /* Apply new context */ - dcb->funcs->set_scratch_critical_state(dcb, true); - /* below is for real asic only */ - for (i = 0; i < dc->res_pool->pipe_count; i++) { - struct pipe_ctx *pipe_ctx_old = - &dc->current_state->res_ctx.pipe_ctx[i]; - struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i]; - - if (pipe_ctx->stream == NULL || pipe_ctx->top_pipe) - continue; - - if (pipe_ctx->stream == pipe_ctx_old->stream) { - if (pipe_ctx_old->clock_source != pipe_ctx->clock_source) - dce_crtc_switch_to_clk_src(dc->hwseq, - pipe_ctx->clock_source, i); - continue; - } - - dc->hwss.enable_display_power_gating( - dc, i, dc->ctx->dc_bios, - PIPE_GATING_CONTROL_DISABLE); - } - - set_safe_displaymarks(&context->res_ctx, dc->res_pool); - -#if defined(CONFIG_DRM_AMD_DC_FBC) - if (dc->fbc_compressor) - dc->fbc_compressor->funcs->disable_fbc(dc->fbc_compressor); -#endif - /*TODO: when pplib works*/ - apply_min_clocks(dc, context, &clocks_state, true); - -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) - if (dc->ctx->dce_version >= DCN_VERSION_1_0) { - if (context->bw.dcn.calc_clk.fclk_khz - > dc->current_state->bw.dcn.cur_clk.fclk_khz) { - struct dm_pp_clock_for_voltage_req clock; - - clock.clk_type = DM_PP_CLOCK_TYPE_FCLK; - clock.clocks_in_khz = context->bw.dcn.calc_clk.fclk_khz; - dm_pp_apply_clock_for_voltage_request(dc->ctx, &clock); - dc->current_state->bw.dcn.cur_clk.fclk_khz = clock.clocks_in_khz; - context->bw.dcn.cur_clk.fclk_khz = clock.clocks_in_khz; - } - if (context->bw.dcn.calc_clk.dcfclk_khz - > dc->current_state->bw.dcn.cur_clk.dcfclk_khz) { - struct dm_pp_clock_for_voltage_req clock; - - clock.clk_type = DM_PP_CLOCK_TYPE_DCFCLK; - clock.clocks_in_khz = context->bw.dcn.calc_clk.dcfclk_khz; - dm_pp_apply_clock_for_voltage_request(dc->ctx, &clock); - dc->current_state->bw.dcn.cur_clk.dcfclk_khz = clock.clocks_in_khz; - context->bw.dcn.cur_clk.dcfclk_khz = clock.clocks_in_khz; - } - if (context->bw.dcn.calc_clk.dispclk_khz - > dc->current_state->bw.dcn.cur_clk.dispclk_khz) { - dc->res_pool->display_clock->funcs->set_clock( - dc->res_pool->display_clock, - context->bw.dcn.calc_clk.dispclk_khz); - dc->current_state->bw.dcn.cur_clk.dispclk_khz = - context->bw.dcn.calc_clk.dispclk_khz; - context->bw.dcn.cur_clk.dispclk_khz = - context->bw.dcn.calc_clk.dispclk_khz; - } - } else -#endif - if (context->bw.dce.dispclk_khz - > dc->current_state->bw.dce.dispclk_khz) { - dc->res_pool->display_clock->funcs->set_clock( - dc->res_pool->display_clock, - context->bw.dce.dispclk_khz * 115 / 100); - } /* program audio wall clock. use HDMI as clock source if HDMI * audio active. Otherwise, use DP as clock source * first, loop to find any HDMI audio, if not, loop find DP audio @@ -2113,6 +1997,52 @@ enum dc_status dce110_apply_ctx_to_hw( } } } +} + +enum dc_status dce110_apply_ctx_to_hw( + struct dc *dc, + struct dc_state *context) +{ + struct dc_bios *dcb = dc->ctx->dc_bios; + enum dc_status status; + int i; + + /* Reset old context */ + /* look up the targets that have been removed since last commit */ + dc->hwss.reset_hw_ctx_wrap(dc, context); + + /* Skip applying if no targets */ + if (context->stream_count <= 0) + return DC_OK; + + /* Apply new context */ + dcb->funcs->set_scratch_critical_state(dcb, true); + + /* below is for real asic only */ + for (i = 0; i < dc->res_pool->pipe_count; i++) { + struct pipe_ctx *pipe_ctx_old = + &dc->current_state->res_ctx.pipe_ctx[i]; + struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i]; + + if (pipe_ctx->stream == NULL || pipe_ctx->top_pipe) + continue; + + if (pipe_ctx->stream == pipe_ctx_old->stream) { + if (pipe_ctx_old->clock_source != pipe_ctx->clock_source) + dce_crtc_switch_to_clk_src(dc->hwseq, + pipe_ctx->clock_source, i); + continue; + } + + dc->hwss.enable_display_power_gating( + dc, i, dc->ctx->dc_bios, + PIPE_GATING_CONTROL_DISABLE); + } + + if (dc->fbc_compressor) + dc->fbc_compressor->funcs->disable_fbc(dc->fbc_compressor); + + dce110_setup_audio_dto(dc, context); for (i = 0; i < dc->res_pool->pipe_count; i++) { struct pipe_ctx *pipe_ctx_old = @@ -2131,31 +2061,6 @@ enum dc_status dce110_apply_ctx_to_hw( if (pipe_ctx->top_pipe) continue; - if (context->res_ctx.pipe_ctx[i].stream_res.audio != NULL) { - - struct audio_output audio_output; - - build_audio_output(context, pipe_ctx, &audio_output); - - if (dc_is_dp_signal(pipe_ctx->stream->signal)) - pipe_ctx->stream_res.stream_enc->funcs->dp_audio_setup( - pipe_ctx->stream_res.stream_enc, - pipe_ctx->stream_res.audio->inst, - &pipe_ctx->stream->audio_info); - else - pipe_ctx->stream_res.stream_enc->funcs->hdmi_audio_setup( - pipe_ctx->stream_res.stream_enc, - pipe_ctx->stream_res.audio->inst, - &pipe_ctx->stream->audio_info, - &audio_output.crtc_info); - - pipe_ctx->stream_res.audio->funcs->az_configure( - pipe_ctx->stream_res.audio, - pipe_ctx->stream->signal, - &audio_output.crtc_info, - &pipe_ctx->stream->audio_info); - } - status = apply_single_controller_ctx_to_hw( pipe_ctx, context, @@ -2165,17 +2070,11 @@ enum dc_status dce110_apply_ctx_to_hw( return status; } - /* to save power */ - apply_min_clocks(dc, context, &clocks_state, false); - dcb->funcs->set_scratch_critical_state(dcb, false); -#if defined(CONFIG_DRM_AMD_DC_FBC) if (dc->fbc_compressor) enable_fbc(dc, context); -#endif - return DC_OK; } @@ -2490,10 +2389,9 @@ static void init_hw(struct dc *dc) abm->funcs->init_backlight(abm); abm->funcs->abm_init(abm); } -#if defined(CONFIG_DRM_AMD_DC_FBC) + if (dc->fbc_compressor) dc->fbc_compressor->funcs->power_up_fbc(dc->fbc_compressor); -#endif } @@ -2632,7 +2530,7 @@ static void pplib_apply_display_requirements( /* TODO: dce11.2*/ pp_display_cfg->avail_mclk_switch_time_in_disp_active_us = 0; - pp_display_cfg->disp_clk_khz = context->bw.dce.dispclk_khz; + pp_display_cfg->disp_clk_khz = dc->res_pool->dccg->clks.dispclk_khz; dce110_fill_display_configs(context, pp_display_cfg); @@ -2654,20 +2552,25 @@ static void pplib_apply_display_requirements( dc->prev_display_config = *pp_display_cfg; } -static void dce110_set_bandwidth( +void dce110_set_bandwidth( struct dc *dc, struct dc_state *context, bool decrease_allowed) { - dce110_set_displaymarks(dc, context); + struct dc_clocks req_clks; - if (decrease_allowed || context->bw.dce.dispclk_khz > dc->current_state->bw.dce.dispclk_khz) { - dc->res_pool->display_clock->funcs->set_clock( - dc->res_pool->display_clock, - context->bw.dce.dispclk_khz * 115 / 100); - dc->current_state->bw.dce.dispclk_khz = context->bw.dce.dispclk_khz; - } + req_clks.dispclk_khz = context->bw.dce.dispclk_khz; + req_clks.phyclk_khz = get_max_pixel_clock_for_all_paths(dc, context); + + if (decrease_allowed) + dce110_set_displaymarks(dc, context); + else + dce110_set_safe_displaymarks(&context->res_ctx, dc->res_pool); + dc->res_pool->dccg->funcs->update_clocks( + dc->res_pool->dccg, + &req_clks, + decrease_allowed); pplib_apply_display_requirements(dc, context); } @@ -2679,9 +2582,7 @@ static void dce110_program_front_end_for_pipe( struct dc_plane_state *plane_state = pipe_ctx->plane_state; struct xfm_grph_csc_adjustment adjust; struct out_csc_color_matrix tbl_entry; -#if defined(CONFIG_DRM_AMD_DC_FBC) unsigned int underlay_idx = dc->res_pool->underlay_pipe_index; -#endif unsigned int i; DC_LOGGER_INIT(); memset(&tbl_entry, 0, sizeof(tbl_entry)); @@ -2722,7 +2623,6 @@ static void dce110_program_front_end_for_pipe( program_scaler(dc, pipe_ctx); -#if defined(CONFIG_DRM_AMD_DC_FBC) /* fbc not applicable on Underlay pipe */ if (dc->fbc_compressor && old_pipe->stream && pipe_ctx->pipe_idx != underlay_idx) { @@ -2731,7 +2631,6 @@ static void dce110_program_front_end_for_pipe( else enable_fbc(dc, dc->current_state); } -#endif mi->funcs->mem_input_program_surface_config( mi, @@ -2907,9 +2806,11 @@ void dce110_set_cursor_position(struct pipe_ctx *pipe_ctx) struct dc_cursor_mi_param param = { .pixel_clk_khz = pipe_ctx->stream->timing.pix_clk_khz, .ref_clk_khz = pipe_ctx->stream->ctx->dc->res_pool->ref_clock_inKhz, - .viewport_x_start = pipe_ctx->plane_res.scl_data.viewport.x, - .viewport_width = pipe_ctx->plane_res.scl_data.viewport.width, - .h_scale_ratio = pipe_ctx->plane_res.scl_data.ratios.horz + .viewport = pipe_ctx->plane_res.scl_data.viewport, + .h_scale_ratio = pipe_ctx->plane_res.scl_data.ratios.horz, + .v_scale_ratio = pipe_ctx->plane_res.scl_data.ratios.vert, + .rotation = pipe_ctx->plane_state->rotation, + .mirror = pipe_ctx->plane_state->horizontal_mirror }; if (pipe_ctx->plane_state->address.type @@ -2968,6 +2869,8 @@ static const struct hw_sequencer_funcs dce110_funcs = { .disable_stream = dce110_disable_stream, .unblank_stream = dce110_unblank_stream, .blank_stream = dce110_blank_stream, + .enable_audio_stream = dce110_enable_audio_stream, + .disable_audio_stream = dce110_disable_audio_stream, .enable_display_pipe_clock_gating = enable_display_pipe_clock_gating, .enable_display_power_gating = dce110_enable_display_power_gating, .disable_plane = dce110_power_down_fe, diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.h b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.h index 5d7e9f516827..e4c5db75c4c6 100644 --- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.h +++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.h @@ -49,6 +49,10 @@ void dce110_unblank_stream(struct pipe_ctx *pipe_ctx, struct dc_link_settings *link_settings); void dce110_blank_stream(struct pipe_ctx *pipe_ctx); + +void dce110_enable_audio_stream(struct pipe_ctx *pipe_ctx); +void dce110_disable_audio_stream(struct pipe_ctx *pipe_ctx, int option); + void dce110_update_info_frame(struct pipe_ctx *pipe_ctx); void dce110_set_avmute(struct pipe_ctx *pipe_ctx, bool enable); @@ -56,10 +60,19 @@ void dce110_enable_accelerated_mode(struct dc *dc, struct dc_state *context); void dce110_power_down(struct dc *dc); +void dce110_set_safe_displaymarks( + struct resource_context *res_ctx, + const struct resource_pool *pool); + void dce110_fill_display_configs( const struct dc_state *context, struct dm_pp_display_configuration *pp_display_cfg); +void dce110_set_bandwidth( + struct dc *dc, + struct dc_state *context, + bool decrease_allowed); + uint32_t dce110_get_min_vblank_time_us(const struct dc_state *context); void dp_receiver_power_ctrl(struct dc_link *link, bool on); diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_mem_input_v.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_mem_input_v.c index 0564c8e31252..9b9fc3d96c07 100644 --- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_mem_input_v.c +++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_mem_input_v.c @@ -1011,7 +1011,7 @@ void dce110_free_mem_input_v( { } -static struct mem_input_funcs dce110_mem_input_v_funcs = { +static const struct mem_input_funcs dce110_mem_input_v_funcs = { .mem_input_program_display_marks = dce_mem_input_v_program_display_marks, .mem_input_program_chroma_display_marks = diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_resource.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_resource.c index ee33786bdef6..e5e9e92521e9 100644 --- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_resource.c @@ -49,14 +49,14 @@ #include "dce/dce_clock_source.h" #include "dce/dce_hwseq.h" #include "dce110/dce110_hw_sequencer.h" +#include "dce/dce_aux.h" #include "dce/dce_abm.h" #include "dce/dce_dmcu.h" #define DC_LOGGER \ dc->ctx->logger -#if defined(CONFIG_DRM_AMD_DC_FBC) + #include "dce110/dce110_compressor.h" -#endif #include "reg_helper.h" @@ -147,15 +147,15 @@ static const struct dce110_timing_generator_offsets dce110_tg_offsets[] = { #define SRI(reg_name, block, id)\ .reg_name = mm ## block ## id ## _ ## reg_name -static const struct dce_disp_clk_registers disp_clk_regs = { +static const struct dccg_registers disp_clk_regs = { CLK_COMMON_REG_LIST_DCE_BASE() }; -static const struct dce_disp_clk_shift disp_clk_shift = { +static const struct dccg_shift disp_clk_shift = { CLK_COMMON_MASK_SH_LIST_DCE_COMMON_BASE(__SHIFT) }; -static const struct dce_disp_clk_mask disp_clk_mask = { +static const struct dccg_mask disp_clk_mask = { CLK_COMMON_MASK_SH_LIST_DCE_COMMON_BASE(_MASK) }; @@ -307,6 +307,21 @@ static const struct dce_opp_mask opp_mask = { OPP_COMMON_MASK_SH_LIST_DCE_110(_MASK) }; +#define aux_engine_regs(id)\ +[id] = {\ + AUX_COMMON_REG_LIST(id), \ + .AUX_RESET_MASK = 0 \ +} + +static const struct dce110_aux_registers aux_engine_regs[] = { + aux_engine_regs(0), + aux_engine_regs(1), + aux_engine_regs(2), + aux_engine_regs(3), + aux_engine_regs(4), + aux_engine_regs(5) +}; + #define audio_regs(id)\ [id] = {\ AUD_COMMON_REG_LIST(id)\ @@ -589,6 +604,23 @@ static struct output_pixel_processor *dce110_opp_create( return &opp->base; } +struct aux_engine *dce110_aux_engine_create( + struct dc_context *ctx, + uint32_t inst) +{ + struct aux_engine_dce110 *aux_engine = + kzalloc(sizeof(struct aux_engine_dce110), GFP_KERNEL); + + if (!aux_engine) + return NULL; + + dce110_aux_engine_construct(aux_engine, ctx, inst, + SW_AUX_TIMEOUT_PERIOD_MULTIPLIER * AUX_TIMEOUT_PERIOD, + &aux_engine_regs[inst]); + + return &aux_engine->base; +} + struct clock_source *dce110_clock_source_create( struct dc_context *ctx, struct dc_bios *bios, @@ -652,6 +684,10 @@ static void destruct(struct dce110_resource_pool *pool) kfree(DCE110TG_FROM_TG(pool->base.timing_generators[i])); pool->base.timing_generators[i] = NULL; } + + if (pool->base.engines[i] != NULL) + dce110_engine_destroy(&pool->base.engines[i]); + } for (i = 0; i < pool->base.stream_enc_count; i++) { @@ -680,8 +716,8 @@ static void destruct(struct dce110_resource_pool *pool) if (pool->base.dmcu != NULL) dce_dmcu_destroy(&pool->base.dmcu); - if (pool->base.display_clock != NULL) - dce_disp_clk_destroy(&pool->base.display_clock); + if (pool->base.dccg != NULL) + dce_dccg_destroy(&pool->base.dccg); if (pool->base.irqs != NULL) { dal_irq_service_destroy(&pool->base.irqs); @@ -795,43 +831,38 @@ static bool dce110_validate_bandwidth( if (memcmp(&dc->current_state->bw.dce, &context->bw.dce, sizeof(context->bw.dce))) { - struct log_entry log_entry; - dm_logger_open( - dc->ctx->logger, - &log_entry, - LOG_BANDWIDTH_CALCS); - dm_logger_append(&log_entry, "%s: finish,\n" + + DC_LOG_BANDWIDTH_CALCS( + "%s: finish,\n" + "nbpMark_b: %d nbpMark_a: %d urgentMark_b: %d urgentMark_a: %d\n" + "stutMark_b: %d stutMark_a: %d\n" + "nbpMark_b: %d nbpMark_a: %d urgentMark_b: %d urgentMark_a: %d\n" + "stutMark_b: %d stutMark_a: %d\n" "nbpMark_b: %d nbpMark_a: %d urgentMark_b: %d urgentMark_a: %d\n" - "stutMark_b: %d stutMark_a: %d\n", + "stutMark_b: %d stutMark_a: %d stutter_mode_enable: %d\n" + "cstate: %d pstate: %d nbpstate: %d sync: %d dispclk: %d\n" + "sclk: %d sclk_sleep: %d yclk: %d blackout_recovery_time_us: %d\n" + , __func__, context->bw.dce.nbp_state_change_wm_ns[0].b_mark, context->bw.dce.nbp_state_change_wm_ns[0].a_mark, context->bw.dce.urgent_wm_ns[0].b_mark, context->bw.dce.urgent_wm_ns[0].a_mark, context->bw.dce.stutter_exit_wm_ns[0].b_mark, - context->bw.dce.stutter_exit_wm_ns[0].a_mark); - dm_logger_append(&log_entry, - "nbpMark_b: %d nbpMark_a: %d urgentMark_b: %d urgentMark_a: %d\n" - "stutMark_b: %d stutMark_a: %d\n", + context->bw.dce.stutter_exit_wm_ns[0].a_mark, context->bw.dce.nbp_state_change_wm_ns[1].b_mark, context->bw.dce.nbp_state_change_wm_ns[1].a_mark, context->bw.dce.urgent_wm_ns[1].b_mark, context->bw.dce.urgent_wm_ns[1].a_mark, context->bw.dce.stutter_exit_wm_ns[1].b_mark, - context->bw.dce.stutter_exit_wm_ns[1].a_mark); - dm_logger_append(&log_entry, - "nbpMark_b: %d nbpMark_a: %d urgentMark_b: %d urgentMark_a: %d\n" - "stutMark_b: %d stutMark_a: %d stutter_mode_enable: %d\n", + context->bw.dce.stutter_exit_wm_ns[1].a_mark, context->bw.dce.nbp_state_change_wm_ns[2].b_mark, context->bw.dce.nbp_state_change_wm_ns[2].a_mark, context->bw.dce.urgent_wm_ns[2].b_mark, context->bw.dce.urgent_wm_ns[2].a_mark, context->bw.dce.stutter_exit_wm_ns[2].b_mark, context->bw.dce.stutter_exit_wm_ns[2].a_mark, - context->bw.dce.stutter_mode_enable); - dm_logger_append(&log_entry, - "cstate: %d pstate: %d nbpstate: %d sync: %d dispclk: %d\n" - "sclk: %d sclk_sleep: %d yclk: %d blackout_recovery_time_us: %d\n", + context->bw.dce.stutter_mode_enable, context->bw.dce.cpuc_state_change_enable, context->bw.dce.cpup_state_change_enable, context->bw.dce.nbp_state_change_enable, @@ -841,7 +872,6 @@ static bool dce110_validate_bandwidth( context->bw.dce.sclk_deep_sleep_khz, context->bw.dce.yclk_khz, context->bw.dce.blackout_recovery_time_us); - dm_logger_close(&log_entry); } return result; } @@ -1180,11 +1210,11 @@ static bool construct( } } - pool->base.display_clock = dce110_disp_clk_create(ctx, + pool->base.dccg = dce110_dccg_create(ctx, &disp_clk_regs, &disp_clk_shift, &disp_clk_mask); - if (pool->base.display_clock == NULL) { + if (pool->base.dccg == NULL) { dm_error("DC: failed to create display clock!\n"); BREAK_TO_DEBUGGER(); goto res_create_fail; @@ -1214,7 +1244,7 @@ static bool construct( * max_clock_state */ if (dm_pp_get_static_clocks(ctx, &static_clk_info)) - pool->base.display_clock->max_clks_state = + pool->base.dccg->max_clks_state = static_clk_info.max_clocks_state; { @@ -1265,14 +1295,18 @@ static bool construct( "DC: failed to create output pixel processor!\n"); goto res_create_fail; } + + pool->base.engines[i] = dce110_aux_engine_create(ctx, i); + if (pool->base.engines[i] == NULL) { + BREAK_TO_DEBUGGER(); + dm_error( + "DC:failed to create aux engine!!\n"); + goto res_create_fail; + } } -#if defined(CONFIG_DRM_AMD_DC_FBC) dc->fbc_compressor = dce110_compressor_create(ctx); - - -#endif if (!underlay_create(ctx, &pool->base)) goto res_create_fail; diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_transform_v.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_transform_v.c index a7dce060204f..aa8d6b10d2c3 100644 --- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_transform_v.c +++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_transform_v.c @@ -235,7 +235,7 @@ static void program_overscan( int overscan_right = data->h_active - data->recout.x - data->recout.width; int overscan_bottom = data->v_active - data->recout.y - data->recout.height; - if (xfm_dce->base.ctx->dc->debug.surface_visual_confirm) { + if (xfm_dce->base.ctx->dc->debug.visual_confirm != VISUAL_CONFIRM_DISABLE) { overscan_bottom += 2; overscan_right += 2; } diff --git a/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.c b/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.c index 00c0a1ef15eb..288129343c77 100644 --- a/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.c @@ -49,6 +49,7 @@ #include "dce112/dce112_hw_sequencer.h" #include "dce/dce_abm.h" #include "dce/dce_dmcu.h" +#include "dce/dce_aux.h" #include "reg_helper.h" @@ -146,15 +147,15 @@ static const struct dce110_timing_generator_offsets dce112_tg_offsets[] = { .reg_name = mm ## block ## id ## _ ## reg_name -static const struct dce_disp_clk_registers disp_clk_regs = { +static const struct dccg_registers disp_clk_regs = { CLK_COMMON_REG_LIST_DCE_BASE() }; -static const struct dce_disp_clk_shift disp_clk_shift = { +static const struct dccg_shift disp_clk_shift = { CLK_COMMON_MASK_SH_LIST_DCE_COMMON_BASE(__SHIFT) }; -static const struct dce_disp_clk_mask disp_clk_mask = { +static const struct dccg_mask disp_clk_mask = { CLK_COMMON_MASK_SH_LIST_DCE_COMMON_BASE(_MASK) }; @@ -314,6 +315,21 @@ static const struct dce_opp_mask opp_mask = { OPP_COMMON_MASK_SH_LIST_DCE_112(_MASK) }; +#define aux_engine_regs(id)\ +[id] = {\ + AUX_COMMON_REG_LIST(id), \ + .AUX_RESET_MASK = 0 \ +} + +static const struct dce110_aux_registers aux_engine_regs[] = { + aux_engine_regs(0), + aux_engine_regs(1), + aux_engine_regs(2), + aux_engine_regs(3), + aux_engine_regs(4), + aux_engine_regs(5) +}; + #define audio_regs(id)\ [id] = {\ AUD_COMMON_REG_LIST(id)\ @@ -588,6 +604,23 @@ struct output_pixel_processor *dce112_opp_create( return &opp->base; } +struct aux_engine *dce112_aux_engine_create( + struct dc_context *ctx, + uint32_t inst) +{ + struct aux_engine_dce110 *aux_engine = + kzalloc(sizeof(struct aux_engine_dce110), GFP_KERNEL); + + if (!aux_engine) + return NULL; + + dce110_aux_engine_construct(aux_engine, ctx, inst, + SW_AUX_TIMEOUT_PERIOD_MULTIPLIER * AUX_TIMEOUT_PERIOD, + &aux_engine_regs[inst]); + + return &aux_engine->base; +} + struct clock_source *dce112_clock_source_create( struct dc_context *ctx, struct dc_bios *bios, @@ -625,6 +658,9 @@ static void destruct(struct dce110_resource_pool *pool) if (pool->base.opps[i] != NULL) dce110_opp_destroy(&pool->base.opps[i]); + if (pool->base.engines[i] != NULL) + dce110_engine_destroy(&pool->base.engines[i]); + if (pool->base.transforms[i] != NULL) dce112_transform_destroy(&pool->base.transforms[i]); @@ -640,6 +676,7 @@ static void destruct(struct dce110_resource_pool *pool) kfree(DCE110TG_FROM_TG(pool->base.timing_generators[i])); pool->base.timing_generators[i] = NULL; } + } for (i = 0; i < pool->base.stream_enc_count; i++) { @@ -668,8 +705,8 @@ static void destruct(struct dce110_resource_pool *pool) if (pool->base.dmcu != NULL) dce_dmcu_destroy(&pool->base.dmcu); - if (pool->base.display_clock != NULL) - dce_disp_clk_destroy(&pool->base.display_clock); + if (pool->base.dccg != NULL) + dce_dccg_destroy(&pool->base.dccg); if (pool->base.irqs != NULL) { dal_irq_service_destroy(&pool->base.irqs); @@ -744,43 +781,38 @@ bool dce112_validate_bandwidth( if (memcmp(&dc->current_state->bw.dce, &context->bw.dce, sizeof(context->bw.dce))) { - struct log_entry log_entry; - dm_logger_open( - dc->ctx->logger, - &log_entry, - LOG_BANDWIDTH_CALCS); - dm_logger_append(&log_entry, "%s: finish,\n" + + DC_LOG_BANDWIDTH_CALCS( + "%s: finish,\n" + "nbpMark_b: %d nbpMark_a: %d urgentMark_b: %d urgentMark_a: %d\n" + "stutMark_b: %d stutMark_a: %d\n" "nbpMark_b: %d nbpMark_a: %d urgentMark_b: %d urgentMark_a: %d\n" - "stutMark_b: %d stutMark_a: %d\n", + "stutMark_b: %d stutMark_a: %d\n" + "nbpMark_b: %d nbpMark_a: %d urgentMark_b: %d urgentMark_a: %d\n" + "stutMark_b: %d stutMark_a: %d stutter_mode_enable: %d\n" + "cstate: %d pstate: %d nbpstate: %d sync: %d dispclk: %d\n" + "sclk: %d sclk_sleep: %d yclk: %d blackout_recovery_time_us: %d\n" + , __func__, context->bw.dce.nbp_state_change_wm_ns[0].b_mark, context->bw.dce.nbp_state_change_wm_ns[0].a_mark, context->bw.dce.urgent_wm_ns[0].b_mark, context->bw.dce.urgent_wm_ns[0].a_mark, context->bw.dce.stutter_exit_wm_ns[0].b_mark, - context->bw.dce.stutter_exit_wm_ns[0].a_mark); - dm_logger_append(&log_entry, - "nbpMark_b: %d nbpMark_a: %d urgentMark_b: %d urgentMark_a: %d\n" - "stutMark_b: %d stutMark_a: %d\n", + context->bw.dce.stutter_exit_wm_ns[0].a_mark, context->bw.dce.nbp_state_change_wm_ns[1].b_mark, context->bw.dce.nbp_state_change_wm_ns[1].a_mark, context->bw.dce.urgent_wm_ns[1].b_mark, context->bw.dce.urgent_wm_ns[1].a_mark, context->bw.dce.stutter_exit_wm_ns[1].b_mark, - context->bw.dce.stutter_exit_wm_ns[1].a_mark); - dm_logger_append(&log_entry, - "nbpMark_b: %d nbpMark_a: %d urgentMark_b: %d urgentMark_a: %d\n" - "stutMark_b: %d stutMark_a: %d stutter_mode_enable: %d\n", + context->bw.dce.stutter_exit_wm_ns[1].a_mark, context->bw.dce.nbp_state_change_wm_ns[2].b_mark, context->bw.dce.nbp_state_change_wm_ns[2].a_mark, context->bw.dce.urgent_wm_ns[2].b_mark, context->bw.dce.urgent_wm_ns[2].a_mark, context->bw.dce.stutter_exit_wm_ns[2].b_mark, context->bw.dce.stutter_exit_wm_ns[2].a_mark, - context->bw.dce.stutter_mode_enable); - dm_logger_append(&log_entry, - "cstate: %d pstate: %d nbpstate: %d sync: %d dispclk: %d\n" - "sclk: %d sclk_sleep: %d yclk: %d blackout_recovery_time_us: %d\n", + context->bw.dce.stutter_mode_enable, context->bw.dce.cpuc_state_change_enable, context->bw.dce.cpup_state_change_enable, context->bw.dce.nbp_state_change_enable, @@ -790,7 +822,6 @@ bool dce112_validate_bandwidth( context->bw.dce.sclk_deep_sleep_khz, context->bw.dce.yclk_khz, context->bw.dce.blackout_recovery_time_us); - dm_logger_close(&log_entry); } return result; } @@ -1000,7 +1031,7 @@ static void bw_calcs_data_update_from_pplib(struct dc *dc) eng_clks.data[0].clocks_in_khz; clk_ranges.wm_clk_ranges[0].wm_max_eng_clk_in_khz = eng_clks.data[eng_clks.num_levels*3/8].clocks_in_khz - 1; - clk_ranges.wm_clk_ranges[0].wm_min_memg_clk_in_khz = + clk_ranges.wm_clk_ranges[0].wm_min_mem_clk_in_khz = mem_clks.data[0].clocks_in_khz; clk_ranges.wm_clk_ranges[0].wm_max_mem_clk_in_khz = mem_clks.data[mem_clks.num_levels>>1].clocks_in_khz - 1; @@ -1010,7 +1041,7 @@ static void bw_calcs_data_update_from_pplib(struct dc *dc) eng_clks.data[eng_clks.num_levels*3/8].clocks_in_khz; /* 5 GHz instead of data[7].clockInKHz to cover Overdrive */ clk_ranges.wm_clk_ranges[1].wm_max_eng_clk_in_khz = 5000000; - clk_ranges.wm_clk_ranges[1].wm_min_memg_clk_in_khz = + clk_ranges.wm_clk_ranges[1].wm_min_mem_clk_in_khz = mem_clks.data[0].clocks_in_khz; clk_ranges.wm_clk_ranges[1].wm_max_mem_clk_in_khz = mem_clks.data[mem_clks.num_levels>>1].clocks_in_khz - 1; @@ -1020,7 +1051,7 @@ static void bw_calcs_data_update_from_pplib(struct dc *dc) eng_clks.data[0].clocks_in_khz; clk_ranges.wm_clk_ranges[2].wm_max_eng_clk_in_khz = eng_clks.data[eng_clks.num_levels*3/8].clocks_in_khz - 1; - clk_ranges.wm_clk_ranges[2].wm_min_memg_clk_in_khz = + clk_ranges.wm_clk_ranges[2].wm_min_mem_clk_in_khz = mem_clks.data[mem_clks.num_levels>>1].clocks_in_khz; /* 5 GHz instead of data[2].clockInKHz to cover Overdrive */ clk_ranges.wm_clk_ranges[2].wm_max_mem_clk_in_khz = 5000000; @@ -1030,7 +1061,7 @@ static void bw_calcs_data_update_from_pplib(struct dc *dc) eng_clks.data[eng_clks.num_levels*3/8].clocks_in_khz; /* 5 GHz instead of data[7].clockInKHz to cover Overdrive */ clk_ranges.wm_clk_ranges[3].wm_max_eng_clk_in_khz = 5000000; - clk_ranges.wm_clk_ranges[3].wm_min_memg_clk_in_khz = + clk_ranges.wm_clk_ranges[3].wm_min_mem_clk_in_khz = mem_clks.data[mem_clks.num_levels>>1].clocks_in_khz; /* 5 GHz instead of data[2].clockInKHz to cover Overdrive */ clk_ranges.wm_clk_ranges[3].wm_max_mem_clk_in_khz = 5000000; @@ -1124,11 +1155,11 @@ static bool construct( } } - pool->base.display_clock = dce112_disp_clk_create(ctx, + pool->base.dccg = dce112_dccg_create(ctx, &disp_clk_regs, &disp_clk_shift, &disp_clk_mask); - if (pool->base.display_clock == NULL) { + if (pool->base.dccg == NULL) { dm_error("DC: failed to create display clock!\n"); BREAK_TO_DEBUGGER(); goto res_create_fail; @@ -1158,7 +1189,7 @@ static bool construct( * max_clock_state */ if (dm_pp_get_static_clocks(ctx, &static_clk_info)) - pool->base.display_clock->max_clks_state = + pool->base.dccg->max_clks_state = static_clk_info.max_clocks_state; { @@ -1214,6 +1245,13 @@ static bool construct( "DC:failed to create output pixel processor!\n"); goto res_create_fail; } + pool->base.engines[i] = dce112_aux_engine_create(ctx, i); + if (pool->base.engines[i] == NULL) { + BREAK_TO_DEBUGGER(); + dm_error( + "DC:failed to create aux engine!!\n"); + goto res_create_fail; + } } if (!resource_construct(num_virtual_links, dc, &pool->base, diff --git a/drivers/gpu/drm/amd/display/dc/dce120/dce120_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dce120/dce120_hw_sequencer.c index e96ff86d2fc3..5853522a6182 100644 --- a/drivers/gpu/drm/amd/display/dc/dce120/dce120_hw_sequencer.c +++ b/drivers/gpu/drm/amd/display/dc/dce120/dce120_hw_sequencer.c @@ -244,7 +244,16 @@ static void dce120_update_dchub( dh_data->dchub_info_valid = false; } +static void dce120_set_bandwidth( + struct dc *dc, + struct dc_state *context, + bool decrease_allowed) +{ + if (context->stream_count <= 0) + return; + dce110_set_bandwidth(dc, context, decrease_allowed); +} void dce120_hw_sequencer_construct(struct dc *dc) { @@ -254,5 +263,6 @@ void dce120_hw_sequencer_construct(struct dc *dc) dce110_hw_sequencer_construct(dc); dc->hwss.enable_display_power_gating = dce120_enable_display_power_gating; dc->hwss.update_dchub = dce120_update_dchub; + dc->hwss.set_bandwidth = dce120_set_bandwidth; } diff --git a/drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c b/drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c index 2d58daccc005..d43f37d99c7d 100644 --- a/drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c @@ -53,6 +53,7 @@ #include "dce/dce_hwseq.h" #include "dce/dce_abm.h" #include "dce/dce_dmcu.h" +#include "dce/dce_aux.h" #include "dce/dce_12_0_offset.h" #include "dce/dce_12_0_sh_mask.h" @@ -297,6 +298,20 @@ static const struct dce_opp_shift opp_shift = { static const struct dce_opp_mask opp_mask = { OPP_COMMON_MASK_SH_LIST_DCE_120(_MASK) }; + #define aux_engine_regs(id)\ +[id] = {\ + AUX_COMMON_REG_LIST(id), \ + .AUX_RESET_MASK = 0 \ +} + +static const struct dce110_aux_registers aux_engine_regs[] = { + aux_engine_regs(0), + aux_engine_regs(1), + aux_engine_regs(2), + aux_engine_regs(3), + aux_engine_regs(4), + aux_engine_regs(5) +}; #define audio_regs(id)\ [id] = {\ @@ -361,6 +376,22 @@ struct output_pixel_processor *dce120_opp_create( ctx, inst, &opp_regs[inst], &opp_shift, &opp_mask); return &opp->base; } +struct aux_engine *dce120_aux_engine_create( + struct dc_context *ctx, + uint32_t inst) +{ + struct aux_engine_dce110 *aux_engine = + kzalloc(sizeof(struct aux_engine_dce110), GFP_KERNEL); + + if (!aux_engine) + return NULL; + + dce110_aux_engine_construct(aux_engine, ctx, inst, + SW_AUX_TIMEOUT_PERIOD_MULTIPLIER * AUX_TIMEOUT_PERIOD, + &aux_engine_regs[inst]); + + return &aux_engine->base; +} static const struct bios_registers bios_regs = { .BIOS_SCRATCH_6 = mmBIOS_SCRATCH_6 + NBIO_BASE(mmBIOS_SCRATCH_6_BASE_IDX) @@ -373,7 +404,7 @@ static const struct resource_caps res_cap = { .num_pll = 6, }; -static const struct dc_debug debug_defaults = { +static const struct dc_debug_options debug_defaults = { .disable_clock_gate = true, }; @@ -467,6 +498,10 @@ static void destruct(struct dce110_resource_pool *pool) kfree(DCE110TG_FROM_TG(pool->base.timing_generators[i])); pool->base.timing_generators[i] = NULL; } + + if (pool->base.engines[i] != NULL) + dce110_engine_destroy(&pool->base.engines[i]); + } for (i = 0; i < pool->base.audio_count; i++) { @@ -494,8 +529,8 @@ static void destruct(struct dce110_resource_pool *pool) if (pool->base.dmcu != NULL) dce_dmcu_destroy(&pool->base.dmcu); - if (pool->base.display_clock != NULL) - dce_disp_clk_destroy(&pool->base.display_clock); + if (pool->base.dccg != NULL) + dce_dccg_destroy(&pool->base.dccg); } static void read_dce_straps( @@ -775,7 +810,7 @@ static void bw_calcs_data_update_from_pplib(struct dc *dc) eng_clks.data[0].clocks_in_khz; clk_ranges.wm_clk_ranges[0].wm_max_eng_clk_in_khz = eng_clks.data[eng_clks.num_levels*3/8].clocks_in_khz - 1; - clk_ranges.wm_clk_ranges[0].wm_min_memg_clk_in_khz = + clk_ranges.wm_clk_ranges[0].wm_min_mem_clk_in_khz = mem_clks.data[0].clocks_in_khz; clk_ranges.wm_clk_ranges[0].wm_max_mem_clk_in_khz = mem_clks.data[mem_clks.num_levels>>1].clocks_in_khz - 1; @@ -785,7 +820,7 @@ static void bw_calcs_data_update_from_pplib(struct dc *dc) eng_clks.data[eng_clks.num_levels*3/8].clocks_in_khz; /* 5 GHz instead of data[7].clockInKHz to cover Overdrive */ clk_ranges.wm_clk_ranges[1].wm_max_eng_clk_in_khz = 5000000; - clk_ranges.wm_clk_ranges[1].wm_min_memg_clk_in_khz = + clk_ranges.wm_clk_ranges[1].wm_min_mem_clk_in_khz = mem_clks.data[0].clocks_in_khz; clk_ranges.wm_clk_ranges[1].wm_max_mem_clk_in_khz = mem_clks.data[mem_clks.num_levels>>1].clocks_in_khz - 1; @@ -795,7 +830,7 @@ static void bw_calcs_data_update_from_pplib(struct dc *dc) eng_clks.data[0].clocks_in_khz; clk_ranges.wm_clk_ranges[2].wm_max_eng_clk_in_khz = eng_clks.data[eng_clks.num_levels*3/8].clocks_in_khz - 1; - clk_ranges.wm_clk_ranges[2].wm_min_memg_clk_in_khz = + clk_ranges.wm_clk_ranges[2].wm_min_mem_clk_in_khz = mem_clks.data[mem_clks.num_levels>>1].clocks_in_khz; /* 5 GHz instead of data[2].clockInKHz to cover Overdrive */ clk_ranges.wm_clk_ranges[2].wm_max_mem_clk_in_khz = 5000000; @@ -805,7 +840,7 @@ static void bw_calcs_data_update_from_pplib(struct dc *dc) eng_clks.data[eng_clks.num_levels*3/8].clocks_in_khz; /* 5 GHz instead of data[7].clockInKHz to cover Overdrive */ clk_ranges.wm_clk_ranges[3].wm_max_eng_clk_in_khz = 5000000; - clk_ranges.wm_clk_ranges[3].wm_min_memg_clk_in_khz = + clk_ranges.wm_clk_ranges[3].wm_min_mem_clk_in_khz = mem_clks.data[mem_clks.num_levels>>1].clocks_in_khz; /* 5 GHz instead of data[2].clockInKHz to cover Overdrive */ clk_ranges.wm_clk_ranges[3].wm_max_mem_clk_in_khz = 5000000; @@ -848,6 +883,7 @@ static bool construct( dc->caps.i2c_speed_in_khz = 100; dc->caps.max_cursor_size = 128; dc->caps.dual_link_dvi = true; + dc->caps.psp_setup_panel_mode = true; dc->debug = debug_defaults; @@ -894,11 +930,11 @@ static bool construct( } } - pool->base.display_clock = dce120_disp_clk_create(ctx); - if (pool->base.display_clock == NULL) { + pool->base.dccg = dce120_dccg_create(ctx); + if (pool->base.dccg == NULL) { dm_error("DC: failed to create display clock!\n"); BREAK_TO_DEBUGGER(); - goto disp_clk_create_fail; + goto dccg_create_fail; } pool->base.dmcu = dce_dmcu_create(ctx, @@ -984,6 +1020,13 @@ static bool construct( dm_error( "DC: failed to create output pixel processor!\n"); } + pool->base.engines[i] = dce120_aux_engine_create(ctx, i); + if (pool->base.engines[i] == NULL) { + BREAK_TO_DEBUGGER(); + dm_error( + "DC:failed to create aux engine!!\n"); + goto res_create_fail; + } /* check next valid pipe */ j++; @@ -1011,7 +1054,7 @@ static bool construct( irqs_create_fail: controller_create_fail: -disp_clk_create_fail: +dccg_create_fail: clk_src_create_fail: res_create_fail: diff --git a/drivers/gpu/drm/amd/display/dc/dce80/dce80_resource.c b/drivers/gpu/drm/amd/display/dc/dce80/dce80_resource.c index 48a068964722..604c62969ead 100644 --- a/drivers/gpu/drm/amd/display/dc/dce80/dce80_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dce80/dce80_resource.c @@ -54,6 +54,7 @@ #include "reg_helper.h" #include "dce/dce_dmcu.h" +#include "dce/dce_aux.h" #include "dce/dce_abm.h" /* TODO remove this include */ @@ -153,15 +154,15 @@ static const struct dce110_timing_generator_offsets dce80_tg_offsets[] = { .reg_name = mm ## block ## id ## _ ## reg_name -static const struct dce_disp_clk_registers disp_clk_regs = { +static const struct dccg_registers disp_clk_regs = { CLK_COMMON_REG_LIST_DCE_BASE() }; -static const struct dce_disp_clk_shift disp_clk_shift = { +static const struct dccg_shift disp_clk_shift = { CLK_COMMON_MASK_SH_LIST_DCE_COMMON_BASE(__SHIFT) }; -static const struct dce_disp_clk_mask disp_clk_mask = { +static const struct dccg_mask disp_clk_mask = { CLK_COMMON_MASK_SH_LIST_DCE_COMMON_BASE(_MASK) }; @@ -298,6 +299,21 @@ static const struct dce_opp_mask opp_mask = { OPP_COMMON_MASK_SH_LIST_DCE_80(_MASK) }; +#define aux_engine_regs(id)\ +[id] = {\ + AUX_COMMON_REG_LIST(id), \ + .AUX_RESET_MASK = 0 \ +} + +static const struct dce110_aux_registers aux_engine_regs[] = { + aux_engine_regs(0), + aux_engine_regs(1), + aux_engine_regs(2), + aux_engine_regs(3), + aux_engine_regs(4), + aux_engine_regs(5) +}; + #define audio_regs(id)\ [id] = {\ AUD_COMMON_REG_LIST(id)\ @@ -448,6 +464,23 @@ static struct output_pixel_processor *dce80_opp_create( return &opp->base; } +struct aux_engine *dce80_aux_engine_create( + struct dc_context *ctx, + uint32_t inst) +{ + struct aux_engine_dce110 *aux_engine = + kzalloc(sizeof(struct aux_engine_dce110), GFP_KERNEL); + + if (!aux_engine) + return NULL; + + dce110_aux_engine_construct(aux_engine, ctx, inst, + SW_AUX_TIMEOUT_PERIOD_MULTIPLIER * AUX_TIMEOUT_PERIOD, + &aux_engine_regs[inst]); + + return &aux_engine->base; +} + static struct stream_encoder *dce80_stream_encoder_create( enum engine_id eng_id, struct dc_context *ctx) @@ -655,6 +688,9 @@ static void destruct(struct dce110_resource_pool *pool) kfree(DCE110TG_FROM_TG(pool->base.timing_generators[i])); pool->base.timing_generators[i] = NULL; } + + if (pool->base.engines[i] != NULL) + dce110_engine_destroy(&pool->base.engines[i]); } for (i = 0; i < pool->base.stream_enc_count; i++) { @@ -683,8 +719,8 @@ static void destruct(struct dce110_resource_pool *pool) } } - if (pool->base.display_clock != NULL) - dce_disp_clk_destroy(&pool->base.display_clock); + if (pool->base.dccg != NULL) + dce_dccg_destroy(&pool->base.dccg); if (pool->base.irqs != NULL) { dal_irq_service_destroy(&pool->base.irqs); @@ -822,11 +858,11 @@ static bool dce80_construct( } } - pool->base.display_clock = dce_disp_clk_create(ctx, + pool->base.dccg = dce_dccg_create(ctx, &disp_clk_regs, &disp_clk_shift, &disp_clk_mask); - if (pool->base.display_clock == NULL) { + if (pool->base.dccg == NULL) { dm_error("DC: failed to create display clock!\n"); BREAK_TO_DEBUGGER(); goto res_create_fail; @@ -852,7 +888,7 @@ static bool dce80_construct( goto res_create_fail; } if (dm_pp_get_static_clocks(ctx, &static_clk_info)) - pool->base.display_clock->max_clks_state = + pool->base.dccg->max_clks_state = static_clk_info.max_clocks_state; { @@ -899,9 +935,18 @@ static bool dce80_construct( dm_error("DC: failed to create output pixel processor!\n"); goto res_create_fail; } + + pool->base.engines[i] = dce80_aux_engine_create(ctx, i); + if (pool->base.engines[i] == NULL) { + BREAK_TO_DEBUGGER(); + dm_error( + "DC:failed to create aux engine!!\n"); + goto res_create_fail; + } } dc->caps.max_planes = pool->base.pipe_count; + dc->caps.disable_dp_clk_share = true; if (!resource_construct(num_virtual_links, dc, &pool->base, &res_create_funcs)) @@ -1006,11 +1051,11 @@ static bool dce81_construct( } } - pool->base.display_clock = dce_disp_clk_create(ctx, + pool->base.dccg = dce_dccg_create(ctx, &disp_clk_regs, &disp_clk_shift, &disp_clk_mask); - if (pool->base.display_clock == NULL) { + if (pool->base.dccg == NULL) { dm_error("DC: failed to create display clock!\n"); BREAK_TO_DEBUGGER(); goto res_create_fail; @@ -1037,7 +1082,7 @@ static bool dce81_construct( } if (dm_pp_get_static_clocks(ctx, &static_clk_info)) - pool->base.display_clock->max_clks_state = + pool->base.dccg->max_clks_state = static_clk_info.max_clocks_state; { @@ -1087,6 +1132,7 @@ static bool dce81_construct( } dc->caps.max_planes = pool->base.pipe_count; + dc->caps.disable_dp_clk_share = true; if (!resource_construct(num_virtual_links, dc, &pool->base, &res_create_funcs)) @@ -1187,11 +1233,11 @@ static bool dce83_construct( } } - pool->base.display_clock = dce_disp_clk_create(ctx, + pool->base.dccg = dce_dccg_create(ctx, &disp_clk_regs, &disp_clk_shift, &disp_clk_mask); - if (pool->base.display_clock == NULL) { + if (pool->base.dccg == NULL) { dm_error("DC: failed to create display clock!\n"); BREAK_TO_DEBUGGER(); goto res_create_fail; @@ -1218,7 +1264,7 @@ static bool dce83_construct( } if (dm_pp_get_static_clocks(ctx, &static_clk_info)) - pool->base.display_clock->max_clks_state = + pool->base.dccg->max_clks_state = static_clk_info.max_clocks_state; { @@ -1268,6 +1314,7 @@ static bool dce83_construct( } dc->caps.max_planes = pool->base.pipe_count; + dc->caps.disable_dp_clk_share = true; if (!resource_construct(num_virtual_links, dc, &pool->base, &res_create_funcs)) diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c index c69fa4bfab0a..bf8b68f8db4f 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c @@ -145,10 +145,10 @@ static bool dpp_get_optimal_number_of_taps( pixel_width = scl_data->viewport.width; /* Some ASICs does not support FP16 scaling, so we reject modes require this*/ - if (scl_data->viewport.width != scl_data->h_active && - scl_data->viewport.height != scl_data->v_active && + if (scl_data->format == PIXEL_FORMAT_FP16 && dpp->caps->dscl_data_proc_format == DSCL_DATA_PRCESSING_FIXED_FORMAT && - scl_data->format == PIXEL_FORMAT_FP16) + scl_data->ratios.horz.value != dc_fixpt_one.value && + scl_data->ratios.vert.value != dc_fixpt_one.value) return false; if (scl_data->viewport.width > scl_data->h_active && @@ -445,10 +445,10 @@ void dpp1_set_cursor_position( uint32_t width) { struct dcn10_dpp *dpp = TO_DCN10_DPP(dpp_base); - int src_x_offset = pos->x - pos->x_hotspot - param->viewport_x_start; + int src_x_offset = pos->x - pos->x_hotspot - param->viewport.x; uint32_t cur_en = pos->enable ? 1 : 0; - if (src_x_offset >= (int)param->viewport_width) + if (src_x_offset >= (int)param->viewport.width) cur_en = 0; /* not visible beyond right edge*/ if (src_x_offset + (int)width <= 0) @@ -459,6 +459,18 @@ void dpp1_set_cursor_position( } +void dpp1_cnv_set_optional_cursor_attributes( + struct dpp *dpp_base, + struct dpp_cursor_attributes *attr) +{ + struct dcn10_dpp *dpp = TO_DCN10_DPP(dpp_base); + + if (attr) { + REG_UPDATE(CURSOR0_FP_SCALE_BIAS, CUR0_FP_BIAS, attr->bias); + REG_UPDATE(CURSOR0_FP_SCALE_BIAS, CUR0_FP_SCALE, attr->scale); + } +} + void dpp1_dppclk_control( struct dpp *dpp_base, bool dppclk_div, @@ -499,6 +511,7 @@ static const struct dpp_funcs dcn10_dpp_funcs = { .dpp_full_bypass = dpp1_full_bypass, .set_cursor_attributes = dpp1_set_cursor_attributes, .set_cursor_position = dpp1_set_cursor_position, + .set_optional_cursor_attributes = dpp1_cnv_set_optional_cursor_attributes, .dpp_dppclk_control = dpp1_dppclk_control, .dpp_set_hdr_multiplier = dpp1_set_hdr_multiplier, }; diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h index e862cafa6501..e2889e61b18c 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h @@ -119,6 +119,7 @@ SRI(CURSOR0_CONTROL, CNVC_CUR, id), \ SRI(CURSOR0_COLOR0, CNVC_CUR, id), \ SRI(CURSOR0_COLOR1, CNVC_CUR, id), \ + SRI(CURSOR0_FP_SCALE_BIAS, CNVC_CUR, id), \ SRI(DPP_CONTROL, DPP_TOP, id), \ SRI(CM_HDR_MULT_COEF, CM, id) @@ -324,6 +325,8 @@ TF_SF(CNVC_CUR0_CURSOR0_CONTROL, CUR0_ENABLE, mask_sh), \ TF_SF(CNVC_CUR0_CURSOR0_COLOR0, CUR0_COLOR0, mask_sh), \ TF_SF(CNVC_CUR0_CURSOR0_COLOR1, CUR0_COLOR1, mask_sh), \ + TF_SF(CNVC_CUR0_CURSOR0_FP_SCALE_BIAS, CUR0_FP_BIAS, mask_sh), \ + TF_SF(CNVC_CUR0_CURSOR0_FP_SCALE_BIAS, CUR0_FP_SCALE, mask_sh), \ TF_SF(DPP_TOP0_DPP_CONTROL, DPP_CLOCK_ENABLE, mask_sh), \ TF_SF(CM0_CM_HDR_MULT_COEF, CM_HDR_MULT_COEF, mask_sh) @@ -1076,7 +1079,9 @@ type CUR0_COLOR1; \ type DPPCLK_RATE_CONTROL; \ type DPP_CLOCK_ENABLE; \ - type CM_HDR_MULT_COEF; + type CM_HDR_MULT_COEF; \ + type CUR0_FP_BIAS; \ + type CUR0_FP_SCALE; struct dcn_dpp_shift { TF_REG_FIELD_LIST(uint8_t) @@ -1329,7 +1334,8 @@ struct dcn_dpp_mask { uint32_t CURSOR0_COLOR0; \ uint32_t CURSOR0_COLOR1; \ uint32_t DPP_CONTROL; \ - uint32_t CM_HDR_MULT_COEF; + uint32_t CM_HDR_MULT_COEF; \ + uint32_t CURSOR0_FP_SCALE_BIAS; struct dcn_dpp_registers { DPP_COMMON_REG_VARIABLE_LIST @@ -1370,6 +1376,10 @@ void dpp1_set_cursor_position( const struct dc_cursor_mi_param *param, uint32_t width); +void dpp1_cnv_set_optional_cursor_attributes( + struct dpp *dpp_base, + struct dpp_cursor_attributes *attr); + bool dpp1_dscl_is_lb_conf_valid( int ceil_vratio, int num_partitions, diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_dscl.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_dscl.c index f862fd148cca..4a863a5dab41 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_dscl.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_dscl.c @@ -621,6 +621,10 @@ static void dpp1_dscl_set_manual_ratio_init( static void dpp1_dscl_set_recout( struct dcn10_dpp *dpp, const struct rect *recout) { + int visual_confirm_on = 0; + if (dpp->base.ctx->dc->debug.visual_confirm != VISUAL_CONFIRM_DISABLE) + visual_confirm_on = 1; + REG_SET_2(RECOUT_START, 0, /* First pixel of RECOUT */ RECOUT_START_X, recout->x, @@ -632,8 +636,7 @@ static void dpp1_dscl_set_recout( RECOUT_WIDTH, recout->width, /* Number of RECOUT vertical lines */ RECOUT_HEIGHT, recout->height - - dpp->base.ctx->dc->debug.surface_visual_confirm * 4 * - (dpp->base.inst + 1)); + - visual_confirm_on * 4 * (dpp->base.inst + 1)); } /* Main function to program scaler and line buffer in manual scaling mode */ @@ -655,6 +658,12 @@ void dpp1_dscl_set_scaler_manual_scale( dpp->scl_data = *scl_data; + /* Autocal off */ + REG_SET_3(DSCL_AUTOCAL, 0, + AUTOCAL_MODE, AUTOCAL_MODE_OFF, + AUTOCAL_NUM_PIPE, 0, + AUTOCAL_PIPE_ID, 0); + /* Recout */ dpp1_dscl_set_recout(dpp, &scl_data->recout); @@ -678,12 +687,6 @@ void dpp1_dscl_set_scaler_manual_scale( if (dscl_mode == DSCL_MODE_SCALING_444_BYPASS) return; - /* Autocal off */ - REG_SET_3(DSCL_AUTOCAL, 0, - AUTOCAL_MODE, AUTOCAL_MODE_OFF, - AUTOCAL_NUM_PIPE, 0, - AUTOCAL_PIPE_ID, 0); - /* Black offsets */ if (ycbcr) REG_SET_2(SCL_BLACK_OFFSET, 0, diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.c index 943143efbb82..1ea91e153d3a 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.c @@ -190,10 +190,17 @@ static uint32_t convert_and_clamp( } +void hubbub1_wm_change_req_wa(struct hubbub *hubbub) +{ + REG_UPDATE_SEQ(DCHUBBUB_ARB_WATERMARK_CHANGE_CNTL, + DCHUBBUB_ARB_WATERMARK_CHANGE_REQUEST, 0, 1); +} + void hubbub1_program_watermarks( struct hubbub *hubbub, struct dcn_watermark_set *watermarks, - unsigned int refclk_mhz) + unsigned int refclk_mhz, + bool safe_to_lower) { uint32_t force_en = hubbub->ctx->dc->debug.disable_stutter ? 1 : 0; /* @@ -202,191 +209,259 @@ void hubbub1_program_watermarks( */ uint32_t prog_wm_value; - REG_UPDATE(DCHUBBUB_ARB_WATERMARK_CHANGE_CNTL, - DCHUBBUB_ARB_WATERMARK_CHANGE_REQUEST, 0); /* Repeat for water mark set A, B, C and D. */ /* clock state A */ - prog_wm_value = convert_and_clamp(watermarks->a.urgent_ns, - refclk_mhz, 0x1fffff); - REG_WRITE(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_A, prog_wm_value); - - DC_LOG_BANDWIDTH_CALCS("URGENCY_WATERMARK_A calculated =%d\n" - "HW register value = 0x%x\n", - watermarks->a.urgent_ns, prog_wm_value); + if (safe_to_lower || watermarks->a.urgent_ns > hubbub->watermarks.a.urgent_ns) { + hubbub->watermarks.a.urgent_ns = watermarks->a.urgent_ns; + prog_wm_value = convert_and_clamp(watermarks->a.urgent_ns, + refclk_mhz, 0x1fffff); + REG_WRITE(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_A, prog_wm_value); - prog_wm_value = convert_and_clamp(watermarks->a.pte_meta_urgent_ns, - refclk_mhz, 0x1fffff); - REG_WRITE(DCHUBBUB_ARB_PTE_META_URGENCY_WATERMARK_A, prog_wm_value); - DC_LOG_BANDWIDTH_CALCS("PTE_META_URGENCY_WATERMARK_A calculated =%d\n" - "HW register value = 0x%x\n", - watermarks->a.pte_meta_urgent_ns, prog_wm_value); + DC_LOG_BANDWIDTH_CALCS("URGENCY_WATERMARK_A calculated =%d\n" + "HW register value = 0x%x\n", + watermarks->a.urgent_ns, prog_wm_value); + } - if (REG(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_A)) { - prog_wm_value = convert_and_clamp( - watermarks->a.cstate_pstate.cstate_enter_plus_exit_ns, + if (safe_to_lower || watermarks->a.pte_meta_urgent_ns > hubbub->watermarks.a.pte_meta_urgent_ns) { + hubbub->watermarks.a.pte_meta_urgent_ns = watermarks->a.pte_meta_urgent_ns; + prog_wm_value = convert_and_clamp(watermarks->a.pte_meta_urgent_ns, refclk_mhz, 0x1fffff); - REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_A, prog_wm_value); - DC_LOG_BANDWIDTH_CALCS("SR_ENTER_EXIT_WATERMARK_A calculated =%d\n" + REG_WRITE(DCHUBBUB_ARB_PTE_META_URGENCY_WATERMARK_A, prog_wm_value); + DC_LOG_BANDWIDTH_CALCS("PTE_META_URGENCY_WATERMARK_A calculated =%d\n" "HW register value = 0x%x\n", - watermarks->a.cstate_pstate.cstate_enter_plus_exit_ns, prog_wm_value); + watermarks->a.pte_meta_urgent_ns, prog_wm_value); + } + + if (REG(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_A)) { + if (safe_to_lower || watermarks->a.cstate_pstate.cstate_enter_plus_exit_ns + > hubbub->watermarks.a.cstate_pstate.cstate_enter_plus_exit_ns) { + hubbub->watermarks.a.cstate_pstate.cstate_enter_plus_exit_ns = + watermarks->a.cstate_pstate.cstate_enter_plus_exit_ns; + prog_wm_value = convert_and_clamp( + watermarks->a.cstate_pstate.cstate_enter_plus_exit_ns, + refclk_mhz, 0x1fffff); + REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_A, prog_wm_value); + DC_LOG_BANDWIDTH_CALCS("SR_ENTER_EXIT_WATERMARK_A calculated =%d\n" + "HW register value = 0x%x\n", + watermarks->a.cstate_pstate.cstate_enter_plus_exit_ns, prog_wm_value); + } + if (safe_to_lower || watermarks->a.cstate_pstate.cstate_exit_ns + > hubbub->watermarks.a.cstate_pstate.cstate_exit_ns) { + hubbub->watermarks.a.cstate_pstate.cstate_exit_ns = + watermarks->a.cstate_pstate.cstate_exit_ns; + prog_wm_value = convert_and_clamp( + watermarks->a.cstate_pstate.cstate_exit_ns, + refclk_mhz, 0x1fffff); + REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_A, prog_wm_value); + DC_LOG_BANDWIDTH_CALCS("SR_EXIT_WATERMARK_A calculated =%d\n" + "HW register value = 0x%x\n", + watermarks->a.cstate_pstate.cstate_exit_ns, prog_wm_value); + } + } + if (safe_to_lower || watermarks->a.cstate_pstate.pstate_change_ns + > hubbub->watermarks.a.cstate_pstate.pstate_change_ns) { + hubbub->watermarks.a.cstate_pstate.pstate_change_ns = + watermarks->a.cstate_pstate.pstate_change_ns; prog_wm_value = convert_and_clamp( - watermarks->a.cstate_pstate.cstate_exit_ns, + watermarks->a.cstate_pstate.pstate_change_ns, refclk_mhz, 0x1fffff); - REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_A, prog_wm_value); - DC_LOG_BANDWIDTH_CALCS("SR_EXIT_WATERMARK_A calculated =%d\n" - "HW register value = 0x%x\n", - watermarks->a.cstate_pstate.cstate_exit_ns, prog_wm_value); + REG_WRITE(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_A, prog_wm_value); + DC_LOG_BANDWIDTH_CALCS("DRAM_CLK_CHANGE_WATERMARK_A calculated =%d\n" + "HW register value = 0x%x\n\n", + watermarks->a.cstate_pstate.pstate_change_ns, prog_wm_value); } - prog_wm_value = convert_and_clamp( - watermarks->a.cstate_pstate.pstate_change_ns, - refclk_mhz, 0x1fffff); - REG_WRITE(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_A, prog_wm_value); - DC_LOG_BANDWIDTH_CALCS("DRAM_CLK_CHANGE_WATERMARK_A calculated =%d\n" - "HW register value = 0x%x\n\n", - watermarks->a.cstate_pstate.pstate_change_ns, prog_wm_value); - - /* clock state B */ - prog_wm_value = convert_and_clamp( - watermarks->b.urgent_ns, refclk_mhz, 0x1fffff); - REG_WRITE(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_B, prog_wm_value); - DC_LOG_BANDWIDTH_CALCS("URGENCY_WATERMARK_B calculated =%d\n" - "HW register value = 0x%x\n", - watermarks->b.urgent_ns, prog_wm_value); - - - prog_wm_value = convert_and_clamp( - watermarks->b.pte_meta_urgent_ns, - refclk_mhz, 0x1fffff); - REG_WRITE(DCHUBBUB_ARB_PTE_META_URGENCY_WATERMARK_B, prog_wm_value); - DC_LOG_BANDWIDTH_CALCS("PTE_META_URGENCY_WATERMARK_B calculated =%d\n" - "HW register value = 0x%x\n", - watermarks->b.pte_meta_urgent_ns, prog_wm_value); + if (safe_to_lower || watermarks->b.urgent_ns > hubbub->watermarks.b.urgent_ns) { + hubbub->watermarks.b.urgent_ns = watermarks->b.urgent_ns; + prog_wm_value = convert_and_clamp(watermarks->b.urgent_ns, + refclk_mhz, 0x1fffff); + REG_WRITE(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_B, prog_wm_value); + DC_LOG_BANDWIDTH_CALCS("URGENCY_WATERMARK_B calculated =%d\n" + "HW register value = 0x%x\n", + watermarks->b.urgent_ns, prog_wm_value); + } - if (REG(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_B)) { - prog_wm_value = convert_and_clamp( - watermarks->b.cstate_pstate.cstate_enter_plus_exit_ns, + if (safe_to_lower || watermarks->b.pte_meta_urgent_ns > hubbub->watermarks.b.pte_meta_urgent_ns) { + hubbub->watermarks.b.pte_meta_urgent_ns = watermarks->b.pte_meta_urgent_ns; + prog_wm_value = convert_and_clamp(watermarks->b.pte_meta_urgent_ns, refclk_mhz, 0x1fffff); - REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_B, prog_wm_value); - DC_LOG_BANDWIDTH_CALCS("SR_ENTER_WATERMARK_B calculated =%d\n" + REG_WRITE(DCHUBBUB_ARB_PTE_META_URGENCY_WATERMARK_B, prog_wm_value); + DC_LOG_BANDWIDTH_CALCS("PTE_META_URGENCY_WATERMARK_B calculated =%d\n" "HW register value = 0x%x\n", - watermarks->b.cstate_pstate.cstate_enter_plus_exit_ns, prog_wm_value); + watermarks->b.pte_meta_urgent_ns, prog_wm_value); + } + + if (REG(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_B)) { + if (safe_to_lower || watermarks->b.cstate_pstate.cstate_enter_plus_exit_ns + > hubbub->watermarks.b.cstate_pstate.cstate_enter_plus_exit_ns) { + hubbub->watermarks.b.cstate_pstate.cstate_enter_plus_exit_ns = + watermarks->b.cstate_pstate.cstate_enter_plus_exit_ns; + prog_wm_value = convert_and_clamp( + watermarks->b.cstate_pstate.cstate_enter_plus_exit_ns, + refclk_mhz, 0x1fffff); + REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_B, prog_wm_value); + DC_LOG_BANDWIDTH_CALCS("SR_ENTER_EXIT_WATERMARK_B calculated =%d\n" + "HW register value = 0x%x\n", + watermarks->b.cstate_pstate.cstate_enter_plus_exit_ns, prog_wm_value); + } + if (safe_to_lower || watermarks->b.cstate_pstate.cstate_exit_ns + > hubbub->watermarks.b.cstate_pstate.cstate_exit_ns) { + hubbub->watermarks.b.cstate_pstate.cstate_exit_ns = + watermarks->b.cstate_pstate.cstate_exit_ns; + prog_wm_value = convert_and_clamp( + watermarks->b.cstate_pstate.cstate_exit_ns, + refclk_mhz, 0x1fffff); + REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_B, prog_wm_value); + DC_LOG_BANDWIDTH_CALCS("SR_EXIT_WATERMARK_B calculated =%d\n" + "HW register value = 0x%x\n", + watermarks->b.cstate_pstate.cstate_exit_ns, prog_wm_value); + } + } + if (safe_to_lower || watermarks->b.cstate_pstate.pstate_change_ns + > hubbub->watermarks.b.cstate_pstate.pstate_change_ns) { + hubbub->watermarks.b.cstate_pstate.pstate_change_ns = + watermarks->b.cstate_pstate.pstate_change_ns; prog_wm_value = convert_and_clamp( - watermarks->b.cstate_pstate.cstate_exit_ns, + watermarks->b.cstate_pstate.pstate_change_ns, refclk_mhz, 0x1fffff); - REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_B, prog_wm_value); - DC_LOG_BANDWIDTH_CALCS("SR_EXIT_WATERMARK_B calculated =%d\n" - "HW register value = 0x%x\n", - watermarks->b.cstate_pstate.cstate_exit_ns, prog_wm_value); + REG_WRITE(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_B, prog_wm_value); + DC_LOG_BANDWIDTH_CALCS("DRAM_CLK_CHANGE_WATERMARK_B calculated =%d\n" + "HW register value = 0x%x\n\n", + watermarks->b.cstate_pstate.pstate_change_ns, prog_wm_value); } - prog_wm_value = convert_and_clamp( - watermarks->b.cstate_pstate.pstate_change_ns, - refclk_mhz, 0x1fffff); - REG_WRITE(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_B, prog_wm_value); - DC_LOG_BANDWIDTH_CALCS("DRAM_CLK_CHANGE_WATERMARK_B calculated =%d\n\n" - "HW register value = 0x%x\n", - watermarks->b.cstate_pstate.pstate_change_ns, prog_wm_value); - /* clock state C */ - prog_wm_value = convert_and_clamp( - watermarks->c.urgent_ns, refclk_mhz, 0x1fffff); - REG_WRITE(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_C, prog_wm_value); - DC_LOG_BANDWIDTH_CALCS("URGENCY_WATERMARK_C calculated =%d\n" - "HW register value = 0x%x\n", - watermarks->c.urgent_ns, prog_wm_value); - - - prog_wm_value = convert_and_clamp( - watermarks->c.pte_meta_urgent_ns, - refclk_mhz, 0x1fffff); - REG_WRITE(DCHUBBUB_ARB_PTE_META_URGENCY_WATERMARK_C, prog_wm_value); - DC_LOG_BANDWIDTH_CALCS("PTE_META_URGENCY_WATERMARK_C calculated =%d\n" - "HW register value = 0x%x\n", - watermarks->c.pte_meta_urgent_ns, prog_wm_value); + if (safe_to_lower || watermarks->c.urgent_ns > hubbub->watermarks.c.urgent_ns) { + hubbub->watermarks.c.urgent_ns = watermarks->c.urgent_ns; + prog_wm_value = convert_and_clamp(watermarks->c.urgent_ns, + refclk_mhz, 0x1fffff); + REG_WRITE(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_C, prog_wm_value); + DC_LOG_BANDWIDTH_CALCS("URGENCY_WATERMARK_C calculated =%d\n" + "HW register value = 0x%x\n", + watermarks->c.urgent_ns, prog_wm_value); + } - if (REG(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_C)) { - prog_wm_value = convert_and_clamp( - watermarks->c.cstate_pstate.cstate_enter_plus_exit_ns, + if (safe_to_lower || watermarks->c.pte_meta_urgent_ns > hubbub->watermarks.c.pte_meta_urgent_ns) { + hubbub->watermarks.c.pte_meta_urgent_ns = watermarks->c.pte_meta_urgent_ns; + prog_wm_value = convert_and_clamp(watermarks->c.pte_meta_urgent_ns, refclk_mhz, 0x1fffff); - REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_C, prog_wm_value); - DC_LOG_BANDWIDTH_CALCS("SR_ENTER_WATERMARK_C calculated =%d\n" + REG_WRITE(DCHUBBUB_ARB_PTE_META_URGENCY_WATERMARK_C, prog_wm_value); + DC_LOG_BANDWIDTH_CALCS("PTE_META_URGENCY_WATERMARK_C calculated =%d\n" "HW register value = 0x%x\n", - watermarks->c.cstate_pstate.cstate_enter_plus_exit_ns, prog_wm_value); + watermarks->c.pte_meta_urgent_ns, prog_wm_value); + } + if (REG(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_C)) { + if (safe_to_lower || watermarks->c.cstate_pstate.cstate_enter_plus_exit_ns + > hubbub->watermarks.c.cstate_pstate.cstate_enter_plus_exit_ns) { + hubbub->watermarks.c.cstate_pstate.cstate_enter_plus_exit_ns = + watermarks->c.cstate_pstate.cstate_enter_plus_exit_ns; + prog_wm_value = convert_and_clamp( + watermarks->c.cstate_pstate.cstate_enter_plus_exit_ns, + refclk_mhz, 0x1fffff); + REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_C, prog_wm_value); + DC_LOG_BANDWIDTH_CALCS("SR_ENTER_EXIT_WATERMARK_C calculated =%d\n" + "HW register value = 0x%x\n", + watermarks->c.cstate_pstate.cstate_enter_plus_exit_ns, prog_wm_value); + } + + if (safe_to_lower || watermarks->c.cstate_pstate.cstate_exit_ns + > hubbub->watermarks.c.cstate_pstate.cstate_exit_ns) { + hubbub->watermarks.c.cstate_pstate.cstate_exit_ns = + watermarks->c.cstate_pstate.cstate_exit_ns; + prog_wm_value = convert_and_clamp( + watermarks->c.cstate_pstate.cstate_exit_ns, + refclk_mhz, 0x1fffff); + REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_C, prog_wm_value); + DC_LOG_BANDWIDTH_CALCS("SR_EXIT_WATERMARK_C calculated =%d\n" + "HW register value = 0x%x\n", + watermarks->c.cstate_pstate.cstate_exit_ns, prog_wm_value); + } + } + if (safe_to_lower || watermarks->c.cstate_pstate.pstate_change_ns + > hubbub->watermarks.c.cstate_pstate.pstate_change_ns) { + hubbub->watermarks.c.cstate_pstate.pstate_change_ns = + watermarks->c.cstate_pstate.pstate_change_ns; prog_wm_value = convert_and_clamp( - watermarks->c.cstate_pstate.cstate_exit_ns, + watermarks->c.cstate_pstate.pstate_change_ns, refclk_mhz, 0x1fffff); - REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_C, prog_wm_value); - DC_LOG_BANDWIDTH_CALCS("SR_EXIT_WATERMARK_C calculated =%d\n" - "HW register value = 0x%x\n", - watermarks->c.cstate_pstate.cstate_exit_ns, prog_wm_value); + REG_WRITE(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_C, prog_wm_value); + DC_LOG_BANDWIDTH_CALCS("DRAM_CLK_CHANGE_WATERMARK_C calculated =%d\n" + "HW register value = 0x%x\n\n", + watermarks->c.cstate_pstate.pstate_change_ns, prog_wm_value); } - prog_wm_value = convert_and_clamp( - watermarks->c.cstate_pstate.pstate_change_ns, - refclk_mhz, 0x1fffff); - REG_WRITE(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_C, prog_wm_value); - DC_LOG_BANDWIDTH_CALCS("DRAM_CLK_CHANGE_WATERMARK_C calculated =%d\n\n" - "HW register value = 0x%x\n", - watermarks->c.cstate_pstate.pstate_change_ns, prog_wm_value); - /* clock state D */ - prog_wm_value = convert_and_clamp( - watermarks->d.urgent_ns, refclk_mhz, 0x1fffff); - REG_WRITE(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_D, prog_wm_value); - DC_LOG_BANDWIDTH_CALCS("URGENCY_WATERMARK_D calculated =%d\n" - "HW register value = 0x%x\n", - watermarks->d.urgent_ns, prog_wm_value); - - prog_wm_value = convert_and_clamp( - watermarks->d.pte_meta_urgent_ns, - refclk_mhz, 0x1fffff); - REG_WRITE(DCHUBBUB_ARB_PTE_META_URGENCY_WATERMARK_D, prog_wm_value); - DC_LOG_BANDWIDTH_CALCS("PTE_META_URGENCY_WATERMARK_D calculated =%d\n" - "HW register value = 0x%x\n", - watermarks->d.pte_meta_urgent_ns, prog_wm_value); - - - if (REG(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_D)) { - prog_wm_value = convert_and_clamp( - watermarks->d.cstate_pstate.cstate_enter_plus_exit_ns, + if (safe_to_lower || watermarks->d.urgent_ns > hubbub->watermarks.d.urgent_ns) { + hubbub->watermarks.d.urgent_ns = watermarks->d.urgent_ns; + prog_wm_value = convert_and_clamp(watermarks->d.urgent_ns, refclk_mhz, 0x1fffff); - REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_D, prog_wm_value); - DC_LOG_BANDWIDTH_CALCS("SR_ENTER_WATERMARK_D calculated =%d\n" - "HW register value = 0x%x\n", - watermarks->d.cstate_pstate.cstate_enter_plus_exit_ns, prog_wm_value); + REG_WRITE(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_D, prog_wm_value); + DC_LOG_BANDWIDTH_CALCS("URGENCY_WATERMARK_D calculated =%d\n" + "HW register value = 0x%x\n", + watermarks->d.urgent_ns, prog_wm_value); + } - prog_wm_value = convert_and_clamp( - watermarks->d.cstate_pstate.cstate_exit_ns, + if (safe_to_lower || watermarks->d.pte_meta_urgent_ns > hubbub->watermarks.d.pte_meta_urgent_ns) { + hubbub->watermarks.d.pte_meta_urgent_ns = watermarks->d.pte_meta_urgent_ns; + prog_wm_value = convert_and_clamp(watermarks->d.pte_meta_urgent_ns, refclk_mhz, 0x1fffff); - REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_D, prog_wm_value); - DC_LOG_BANDWIDTH_CALCS("SR_EXIT_WATERMARK_D calculated =%d\n" + REG_WRITE(DCHUBBUB_ARB_PTE_META_URGENCY_WATERMARK_D, prog_wm_value); + DC_LOG_BANDWIDTH_CALCS("PTE_META_URGENCY_WATERMARK_D calculated =%d\n" "HW register value = 0x%x\n", - watermarks->d.cstate_pstate.cstate_exit_ns, prog_wm_value); + watermarks->d.pte_meta_urgent_ns, prog_wm_value); } + if (REG(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_D)) { + if (safe_to_lower || watermarks->d.cstate_pstate.cstate_enter_plus_exit_ns + > hubbub->watermarks.d.cstate_pstate.cstate_enter_plus_exit_ns) { + hubbub->watermarks.d.cstate_pstate.cstate_enter_plus_exit_ns = + watermarks->d.cstate_pstate.cstate_enter_plus_exit_ns; + prog_wm_value = convert_and_clamp( + watermarks->d.cstate_pstate.cstate_enter_plus_exit_ns, + refclk_mhz, 0x1fffff); + REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_D, prog_wm_value); + DC_LOG_BANDWIDTH_CALCS("SR_ENTER_EXIT_WATERMARK_D calculated =%d\n" + "HW register value = 0x%x\n", + watermarks->d.cstate_pstate.cstate_enter_plus_exit_ns, prog_wm_value); + } - prog_wm_value = convert_and_clamp( - watermarks->d.cstate_pstate.pstate_change_ns, - refclk_mhz, 0x1fffff); - REG_WRITE(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_D, prog_wm_value); - DC_LOG_BANDWIDTH_CALCS("DRAM_CLK_CHANGE_WATERMARK_D calculated =%d\n" - "HW register value = 0x%x\n\n", - watermarks->d.cstate_pstate.pstate_change_ns, prog_wm_value); + if (safe_to_lower || watermarks->d.cstate_pstate.cstate_exit_ns + > hubbub->watermarks.d.cstate_pstate.cstate_exit_ns) { + hubbub->watermarks.d.cstate_pstate.cstate_exit_ns = + watermarks->d.cstate_pstate.cstate_exit_ns; + prog_wm_value = convert_and_clamp( + watermarks->d.cstate_pstate.cstate_exit_ns, + refclk_mhz, 0x1fffff); + REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_D, prog_wm_value); + DC_LOG_BANDWIDTH_CALCS("SR_EXIT_WATERMARK_D calculated =%d\n" + "HW register value = 0x%x\n", + watermarks->d.cstate_pstate.cstate_exit_ns, prog_wm_value); + } + } - REG_UPDATE(DCHUBBUB_ARB_WATERMARK_CHANGE_CNTL, - DCHUBBUB_ARB_WATERMARK_CHANGE_REQUEST, 1); + if (safe_to_lower || watermarks->d.cstate_pstate.pstate_change_ns + > hubbub->watermarks.d.cstate_pstate.pstate_change_ns) { + hubbub->watermarks.d.cstate_pstate.pstate_change_ns = + watermarks->d.cstate_pstate.pstate_change_ns; + prog_wm_value = convert_and_clamp( + watermarks->d.cstate_pstate.pstate_change_ns, + refclk_mhz, 0x1fffff); + REG_WRITE(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_D, prog_wm_value); + DC_LOG_BANDWIDTH_CALCS("DRAM_CLK_CHANGE_WATERMARK_D calculated =%d\n" + "HW register value = 0x%x\n\n", + watermarks->d.cstate_pstate.pstate_change_ns, prog_wm_value); + } REG_UPDATE(DCHUBBUB_ARB_SAT_LEVEL, DCHUBBUB_ARB_SAT_LEVEL, 60 * refclk_mhz); @@ -408,6 +483,11 @@ void hubbub1_update_dchub( struct hubbub *hubbub, struct dchub_init_data *dh_data) { + if (REG(DCHUBBUB_SDPIF_FB_TOP) == 0) { + ASSERT(false); + /*should not come here*/ + return; + } /* TODO: port code from dal2 */ switch (dh_data->fb_mode) { case FRAME_BUFFER_MODE_ZFB_ONLY: diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.h index 6315a0e6b0d6..d6e596eef4c5 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.h +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.h @@ -185,6 +185,7 @@ struct hubbub { const struct dcn_hubbub_shift *shifts; const struct dcn_hubbub_mask *masks; unsigned int debug_test_index_pstate; + struct dcn_watermark_set watermarks; }; void hubbub1_update_dchub( @@ -194,10 +195,13 @@ void hubbub1_update_dchub( bool hubbub1_verify_allow_pstate_change_high( struct hubbub *hubbub); +void hubbub1_wm_change_req_wa(struct hubbub *hubbub); + void hubbub1_program_watermarks( struct hubbub *hubbub, struct dcn_watermark_set *watermarks, - unsigned int refclk_mhz); + unsigned int refclk_mhz, + bool safe_to_lower); void hubbub1_toggle_watermark_change_req( struct hubbub *hubbub); diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c index c28085be39ff..2138cd3c5d1d 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c @@ -152,21 +152,19 @@ void hubp1_program_tiling( PIPE_ALIGNED, info->gfx9.pipe_aligned); } -void hubp1_program_size_and_rotation( +void hubp1_program_size( struct hubp *hubp, - enum dc_rotation_angle rotation, enum surface_pixel_format format, const union plane_size *plane_size, - struct dc_plane_dcc_param *dcc, - bool horizontal_mirror) + struct dc_plane_dcc_param *dcc) { struct dcn10_hubp *hubp1 = TO_DCN10_HUBP(hubp); - uint32_t pitch, meta_pitch, pitch_c, meta_pitch_c, mirror; + uint32_t pitch, meta_pitch, pitch_c, meta_pitch_c; /* Program data and meta surface pitch (calculation from addrlib) * 444 or 420 luma */ - if (format >= SURFACE_PIXEL_FORMAT_VIDEO_BEGIN) { + if (format >= SURFACE_PIXEL_FORMAT_VIDEO_BEGIN && format < SURFACE_PIXEL_FORMAT_SUBSAMPLE_END) { ASSERT(plane_size->video.chroma_pitch != 0); /* Chroma pitch zero can cause system hang! */ @@ -192,13 +190,22 @@ void hubp1_program_size_and_rotation( if (format >= SURFACE_PIXEL_FORMAT_VIDEO_BEGIN) REG_UPDATE_2(DCSURF_SURFACE_PITCH_C, PITCH_C, pitch_c, META_PITCH_C, meta_pitch_c); +} + +void hubp1_program_rotation( + struct hubp *hubp, + enum dc_rotation_angle rotation, + bool horizontal_mirror) +{ + struct dcn10_hubp *hubp1 = TO_DCN10_HUBP(hubp); + uint32_t mirror; + if (horizontal_mirror) mirror = 1; else mirror = 0; - /* Program rotation angle and horz mirror - no mirror */ if (rotation == ROTATION_ANGLE_0) REG_UPDATE_2(DCSURF_SURFACE_CONFIG, @@ -287,6 +294,10 @@ void hubp1_program_pixel_format( REG_UPDATE(DCSURF_SURFACE_CONFIG, SURFACE_PIXEL_FORMAT, 66); break; + case SURFACE_PIXEL_FORMAT_VIDEO_AYCrCb8888: + REG_UPDATE(DCSURF_SURFACE_CONFIG, + SURFACE_PIXEL_FORMAT, 12); + break; default: BREAK_TO_DEBUGGER(); break; @@ -450,9 +461,6 @@ bool hubp1_program_surface_flip_and_addr( hubp->request_address = *address; - if (flip_immediate) - hubp->current_address = *address; - return true; } @@ -481,8 +489,8 @@ void hubp1_program_surface_config( { hubp1_dcc_control(hubp, dcc->enable, dcc->grph.independent_64b_blks); hubp1_program_tiling(hubp, tiling_info, format); - hubp1_program_size_and_rotation( - hubp, rotation, format, plane_size, dcc, horizontal_mirror); + hubp1_program_size(hubp, format, plane_size, dcc); + hubp1_program_rotation(hubp, rotation, horizontal_mirror); hubp1_program_pixel_format(hubp, format); } @@ -688,7 +696,6 @@ bool hubp1_is_flip_pending(struct hubp *hubp) if (earliest_inuse_address.grph.addr.quad_part != hubp->request_address.grph.addr.quad_part) return true; - hubp->current_address = hubp->request_address; return false; } @@ -1061,9 +1068,11 @@ void hubp1_cursor_set_position( const struct dc_cursor_mi_param *param) { struct dcn10_hubp *hubp1 = TO_DCN10_HUBP(hubp); - int src_x_offset = pos->x - pos->x_hotspot - param->viewport_x_start; + int src_x_offset = pos->x - pos->x_hotspot - param->viewport.x; + int x_hotspot = pos->x_hotspot; + int y_hotspot = pos->y_hotspot; + uint32_t dst_x_offset; uint32_t cur_en = pos->enable ? 1 : 0; - uint32_t dst_x_offset = (src_x_offset >= 0) ? src_x_offset : 0; /* * Guard aganst cursor_set_position() from being called with invalid @@ -1075,6 +1084,18 @@ void hubp1_cursor_set_position( if (hubp->curs_attr.address.quad_part == 0) return; + if (param->rotation == ROTATION_ANGLE_90 || param->rotation == ROTATION_ANGLE_270) { + src_x_offset = pos->y - pos->y_hotspot - param->viewport.x; + y_hotspot = pos->x_hotspot; + x_hotspot = pos->y_hotspot; + } + + if (param->mirror) { + x_hotspot = param->viewport.width - x_hotspot; + src_x_offset = param->viewport.x + param->viewport.width - src_x_offset; + } + + dst_x_offset = (src_x_offset >= 0) ? src_x_offset : 0; dst_x_offset *= param->ref_clk_khz; dst_x_offset /= param->pixel_clk_khz; @@ -1085,7 +1106,7 @@ void hubp1_cursor_set_position( dc_fixpt_from_int(dst_x_offset), param->h_scale_ratio)); - if (src_x_offset >= (int)param->viewport_width) + if (src_x_offset >= (int)param->viewport.width) cur_en = 0; /* not visible beyond right edge*/ if (src_x_offset + (int)hubp->curs_attr.width <= 0) @@ -1102,8 +1123,8 @@ void hubp1_cursor_set_position( CURSOR_Y_POSITION, pos->y); REG_SET_2(CURSOR_HOT_SPOT, 0, - CURSOR_HOT_SPOT_X, pos->x_hotspot, - CURSOR_HOT_SPOT_Y, pos->y_hotspot); + CURSOR_HOT_SPOT_X, x_hotspot, + CURSOR_HOT_SPOT_Y, y_hotspot); REG_SET(CURSOR_DST_OFFSET, 0, CURSOR_DST_X_OFFSET, dst_x_offset); @@ -1125,7 +1146,7 @@ void hubp1_vtg_sel(struct hubp *hubp, uint32_t otg_inst) REG_UPDATE(DCHUBP_CNTL, HUBP_VTG_SEL, otg_inst); } -static struct hubp_funcs dcn10_hubp_funcs = { +static const struct hubp_funcs dcn10_hubp_funcs = { .hubp_program_surface_flip_and_addr = hubp1_program_surface_flip_and_addr, .hubp_program_surface_config = diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h index d901d5092969..f689feace82d 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h @@ -268,8 +268,6 @@ HUBP_SF(HUBPREQ0_DCSURF_SURFACE_PITCH, META_PITCH, mask_sh),\ HUBP_SF(HUBPREQ0_DCSURF_SURFACE_PITCH_C, PITCH_C, mask_sh),\ HUBP_SF(HUBPREQ0_DCSURF_SURFACE_PITCH_C, META_PITCH_C, mask_sh),\ - HUBP_SF(HUBP0_DCSURF_SURFACE_CONFIG, ROTATION_ANGLE, mask_sh),\ - HUBP_SF(HUBP0_DCSURF_SURFACE_CONFIG, H_MIRROR_EN, mask_sh),\ HUBP_SF(HUBP0_DCSURF_SURFACE_CONFIG, SURFACE_PIXEL_FORMAT, mask_sh),\ HUBP_SF(HUBPREQ0_DCSURF_FLIP_CONTROL, SURFACE_FLIP_TYPE, mask_sh),\ HUBP_SF(HUBPREQ0_DCSURF_FLIP_CONTROL, SURFACE_FLIP_PENDING, mask_sh),\ @@ -388,6 +386,8 @@ #define HUBP_MASK_SH_LIST_DCN10(mask_sh)\ HUBP_MASK_SH_LIST_DCN(mask_sh),\ HUBP_MASK_SH_LIST_DCN_VM(mask_sh),\ + HUBP_SF(HUBP0_DCSURF_SURFACE_CONFIG, ROTATION_ANGLE, mask_sh),\ + HUBP_SF(HUBP0_DCSURF_SURFACE_CONFIG, H_MIRROR_EN, mask_sh),\ HUBP_SF(HUBPREQ0_PREFETCH_SETTINS, DST_Y_PREFETCH, mask_sh),\ HUBP_SF(HUBPREQ0_PREFETCH_SETTINS, VRATIO_PREFETCH, mask_sh),\ HUBP_SF(HUBPREQ0_PREFETCH_SETTINS_C, VRATIO_PREFETCH_C, mask_sh),\ @@ -679,12 +679,15 @@ void hubp1_program_pixel_format( struct hubp *hubp, enum surface_pixel_format format); -void hubp1_program_size_and_rotation( +void hubp1_program_size( struct hubp *hubp, - enum dc_rotation_angle rotation, enum surface_pixel_format format, const union plane_size *plane_size, - struct dc_plane_dcc_param *dcc, + struct dc_plane_dcc_param *dcc); + +void hubp1_program_rotation( + struct hubp *hubp, + enum dc_rotation_angle rotation, bool horizontal_mirror); void hubp1_program_tiling( diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c index f8e0576af6e0..cfcc54f2ce65 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c @@ -337,13 +337,13 @@ void dcn10_log_hw_state(struct dc *dc) DTN_INFO("\nCALCULATED Clocks: dcfclk_khz:%d dcfclk_deep_sleep_khz:%d dispclk_khz:%d\n" "dppclk_khz:%d max_supported_dppclk_khz:%d fclk_khz:%d socclk_khz:%d\n\n", - dc->current_state->bw.dcn.calc_clk.dcfclk_khz, - dc->current_state->bw.dcn.calc_clk.dcfclk_deep_sleep_khz, - dc->current_state->bw.dcn.calc_clk.dispclk_khz, - dc->current_state->bw.dcn.calc_clk.dppclk_khz, - dc->current_state->bw.dcn.calc_clk.max_supported_dppclk_khz, - dc->current_state->bw.dcn.calc_clk.fclk_khz, - dc->current_state->bw.dcn.calc_clk.socclk_khz); + dc->current_state->bw.dcn.clk.dcfclk_khz, + dc->current_state->bw.dcn.clk.dcfclk_deep_sleep_khz, + dc->current_state->bw.dcn.clk.dispclk_khz, + dc->current_state->bw.dcn.clk.dppclk_khz, + dc->current_state->bw.dcn.clk.max_supported_dppclk_khz, + dc->current_state->bw.dcn.clk.fclk_khz, + dc->current_state->bw.dcn.clk.socclk_khz); log_mpc_crc(dc); @@ -415,6 +415,8 @@ static void dpp_pg_control( if (hws->ctx->dc->debug.disable_dpp_power_gate) return; + if (REG(DOMAIN1_PG_CONFIG) == 0) + return; switch (dpp_inst) { case 0: /* DPP0 */ @@ -465,6 +467,8 @@ static void hubp_pg_control( if (hws->ctx->dc->debug.disable_hubp_power_gate) return; + if (REG(DOMAIN0_PG_CONFIG) == 0) + return; switch (hubp_inst) { case 0: /* DCHUBP0 */ @@ -719,19 +723,7 @@ static void reset_back_end_for_pipe( if (!pipe_ctx->stream->dpms_off) core_link_disable_stream(pipe_ctx, FREE_ACQUIRED_RESOURCE); else if (pipe_ctx->stream_res.audio) { - /* - * if stream is already disabled outside of commit streams path, - * audio disable was skipped. Need to do it here - */ - pipe_ctx->stream_res.audio->funcs->az_disable(pipe_ctx->stream_res.audio); - - if (dc->caps.dynamic_audio == true) { - /*we have to dynamic arbitrate the audio endpoints*/ - pipe_ctx->stream_res.audio = NULL; - /*we free the resource, need reset is_audio_acquired*/ - update_audio_usage(&dc->current_state->res_ctx, dc->res_pool, pipe_ctx->stream_res.audio, false); - } - + dc->hwss.disable_audio_stream(pipe_ctx, FREE_ACQUIRED_RESOURCE); } } @@ -842,7 +834,7 @@ static bool dcn10_hw_wa_force_recovery(struct dc *dc) } -static void dcn10_verify_allow_pstate_change_high(struct dc *dc) +void dcn10_verify_allow_pstate_change_high(struct dc *dc) { static bool should_log_hw_state; /* prevent hw state log by default */ @@ -877,7 +869,8 @@ void hwss1_plane_atomic_disconnect(struct dc *dc, struct pipe_ctx *pipe_ctx) return; mpc->funcs->remove_mpcc(mpc, mpc_tree_params, mpcc_to_remove); - opp->mpcc_disconnect_pending[pipe_ctx->plane_res.mpcc_inst] = true; + if (opp != NULL) + opp->mpcc_disconnect_pending[pipe_ctx->plane_res.mpcc_inst] = true; dc->optimized_required = true; @@ -1022,7 +1015,7 @@ static void dcn10_init_hw(struct dc *dc) /* Reset all MPCC muxes */ dc->res_pool->mpc->funcs->mpc_init(dc->res_pool->mpc); - for (i = 0; i < dc->res_pool->pipe_count; i++) { + for (i = 0; i < dc->res_pool->timing_generator_count; i++) { struct timing_generator *tg = dc->res_pool->timing_generators[i]; struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i]; struct hubp *hubp = dc->res_pool->hubps[i]; @@ -1096,6 +1089,8 @@ static void dcn10_init_hw(struct dc *dc) } enable_power_gating_plane(dc->hwseq, true); + + memset(&dc->res_pool->dccg->clks, 0, sizeof(dc->res_pool->dccg->clks)); } static void reset_hw_ctx_wrap( @@ -1164,12 +1159,19 @@ static void dcn10_update_plane_addr(const struct dc *dc, struct pipe_ctx *pipe_c if (plane_state == NULL) return; + addr_patched = patch_address_for_sbs_tb_stereo(pipe_ctx, &addr); + pipe_ctx->plane_res.hubp->funcs->hubp_program_surface_flip_and_addr( pipe_ctx->plane_res.hubp, &plane_state->address, plane_state->flip_immediate); + plane_state->status.requested_address = plane_state->address; + + if (plane_state->flip_immediate) + plane_state->status.current_address = plane_state->address; + if (addr_patched) pipe_ctx->plane_state->address.grph_stereo.left_addr = addr; } @@ -1213,8 +1215,11 @@ static bool dcn10_set_input_transfer_func(struct pipe_ctx *pipe_ctx, } else if (tf->type == TF_TYPE_BYPASS) { dpp_base->funcs->dpp_set_degamma(dpp_base, IPP_DEGAMMA_MODE_BYPASS); } else { - /*TF_TYPE_DISTRIBUTED_POINTS*/ - result = false; + cm_helper_translate_curve_to_degamma_hw_format(tf, + &dpp_base->degamma_params); + dpp_base->funcs->dpp_program_degamma_pwl(dpp_base, + &dpp_base->degamma_params); + result = true; } return result; @@ -1355,10 +1360,11 @@ static void dcn10_enable_per_frame_crtc_position_reset( DC_SYNC_INFO("Setting up\n"); for (i = 0; i < group_size; i++) - grouped_pipes[i]->stream_res.tg->funcs->enable_crtc_reset( - grouped_pipes[i]->stream_res.tg, - grouped_pipes[i]->stream->triggered_crtc_reset.event_source->status.primary_otg_inst, - &grouped_pipes[i]->stream->triggered_crtc_reset); + if (grouped_pipes[i]->stream_res.tg->funcs->enable_crtc_reset) + grouped_pipes[i]->stream_res.tg->funcs->enable_crtc_reset( + grouped_pipes[i]->stream_res.tg, + grouped_pipes[i]->stream->triggered_crtc_reset.event_source->status.primary_otg_inst, + &grouped_pipes[i]->stream->triggered_crtc_reset); DC_SYNC_INFO("Waiting for trigger\n"); @@ -1774,6 +1780,43 @@ static void dcn10_get_surface_visual_confirm_color( } } +static void dcn10_get_hdr_visual_confirm_color( + struct pipe_ctx *pipe_ctx, + struct tg_color *color) +{ + uint32_t color_value = MAX_TG_COLOR_VALUE; + + // Determine the overscan color based on the top-most (desktop) plane's context + struct pipe_ctx *top_pipe_ctx = pipe_ctx; + + while (top_pipe_ctx->top_pipe != NULL) + top_pipe_ctx = top_pipe_ctx->top_pipe; + + switch (top_pipe_ctx->plane_res.scl_data.format) { + case PIXEL_FORMAT_ARGB2101010: + if (top_pipe_ctx->stream->out_transfer_func->tf == TRANSFER_FUNCTION_UNITY) { + /* HDR10, ARGB2101010 - set boarder color to red */ + color->color_r_cr = color_value; + } + break; + case PIXEL_FORMAT_FP16: + if (top_pipe_ctx->stream->out_transfer_func->tf == TRANSFER_FUNCTION_PQ) { + /* HDR10, FP16 - set boarder color to blue */ + color->color_b_cb = color_value; + } else if (top_pipe_ctx->stream->out_transfer_func->tf == TRANSFER_FUNCTION_GAMMA22) { + /* FreeSync 2 HDR - set boarder color to green */ + color->color_g_y = color_value; + } + break; + default: + /* SDR - set boarder color to Gray */ + color->color_r_cr = color_value/2; + color->color_b_cb = color_value/2; + color->color_g_y = color_value/2; + break; + } +} + static uint16_t fixed_point_to_int_frac( struct fixed31_32 arg, uint8_t integer_bits, @@ -1854,11 +1897,10 @@ static void update_dpp(struct dpp *dpp, struct dc_plane_state *plane_state) dpp->funcs->dpp_program_bias_and_scale(dpp, &bns_params); } - -static void update_mpcc(struct dc *dc, struct pipe_ctx *pipe_ctx) +static void dcn10_update_mpcc(struct dc *dc, struct pipe_ctx *pipe_ctx) { struct hubp *hubp = pipe_ctx->plane_res.hubp; - struct mpcc_blnd_cfg blnd_cfg; + struct mpcc_blnd_cfg blnd_cfg = {0}; bool per_pixel_alpha = pipe_ctx->plane_state->per_pixel_alpha && pipe_ctx->bottom_pipe; int mpcc_id; struct mpcc *new_mpcc; @@ -1869,13 +1911,17 @@ static void update_mpcc(struct dc *dc, struct pipe_ctx *pipe_ctx) /* TODO: proper fix once fpga works */ - if (dc->debug.surface_visual_confirm) + if (dc->debug.visual_confirm == VISUAL_CONFIRM_HDR) { + dcn10_get_hdr_visual_confirm_color( + pipe_ctx, &blnd_cfg.black_color); + } else if (dc->debug.visual_confirm == VISUAL_CONFIRM_SURFACE) { dcn10_get_surface_visual_confirm_color( pipe_ctx, &blnd_cfg.black_color); - else + } else { color_space_to_black_color( - dc, pipe_ctx->stream->output_color_space, - &blnd_cfg.black_color); + dc, pipe_ctx->stream->output_color_space, + &blnd_cfg.black_color); + } if (per_pixel_alpha) blnd_cfg.alpha_mode = MPCC_ALPHA_BLEND_MODE_PER_PIXEL_ALPHA; @@ -1964,18 +2010,17 @@ static void update_dchubp_dpp( * divided by 2 */ if (plane_state->update_flags.bits.full_update) { - bool should_divided_by_2 = context->bw.dcn.calc_clk.dppclk_khz <= - context->bw.dcn.cur_clk.dispclk_khz / 2; + bool should_divided_by_2 = context->bw.dcn.clk.dppclk_khz <= + dc->res_pool->dccg->clks.dispclk_khz / 2; dpp->funcs->dpp_dppclk_control( dpp, should_divided_by_2, true); - dc->current_state->bw.dcn.cur_clk.dppclk_khz = - should_divided_by_2 ? - context->bw.dcn.cur_clk.dispclk_khz / 2 : - context->bw.dcn.cur_clk.dispclk_khz; + dc->res_pool->dccg->clks.dppclk_khz = should_divided_by_2 ? + dc->res_pool->dccg->clks.dispclk_khz / 2 : + dc->res_pool->dccg->clks.dispclk_khz; } /* TODO: Need input parameter to tell current DCHUB pipe tie to which OTG @@ -2001,7 +2046,7 @@ static void update_dchubp_dpp( if (plane_state->update_flags.bits.full_update || plane_state->update_flags.bits.per_pixel_alpha_change) - update_mpcc(dc, pipe_ctx); + dc->hwss.update_mpcc(dc, pipe_ctx); if (plane_state->update_flags.bits.full_update || plane_state->update_flags.bits.per_pixel_alpha_change || @@ -2063,12 +2108,13 @@ static void update_dchubp_dpp( static void dcn10_blank_pixel_data( struct dc *dc, - struct stream_resource *stream_res, - struct dc_stream_state *stream, + struct pipe_ctx *pipe_ctx, bool blank) { enum dc_color_space color_space; struct tg_color black_color = {0}; + struct stream_resource *stream_res = &pipe_ctx->stream_res; + struct dc_stream_state *stream = pipe_ctx->stream; /* program otg blank color */ color_space = stream->output_color_space; @@ -2110,6 +2156,33 @@ static void set_hdr_multiplier(struct pipe_ctx *pipe_ctx) pipe_ctx->plane_res.dpp, hw_mult); } +void dcn10_program_pipe( + struct dc *dc, + struct pipe_ctx *pipe_ctx, + struct dc_state *context) +{ + if (pipe_ctx->plane_state->update_flags.bits.full_update) + dcn10_enable_plane(dc, pipe_ctx, context); + + update_dchubp_dpp(dc, pipe_ctx, context); + + set_hdr_multiplier(pipe_ctx); + + if (pipe_ctx->plane_state->update_flags.bits.full_update || + pipe_ctx->plane_state->update_flags.bits.in_transfer_func_change || + pipe_ctx->plane_state->update_flags.bits.gamma_change) + dc->hwss.set_input_transfer_func(pipe_ctx, pipe_ctx->plane_state); + + /* dcn10_translate_regamma_to_hw_format takes 750us to finish + * only do gamma programming for full update. + * TODO: This can be further optimized/cleaned up + * Always call this for now since it does memcmp inside before + * doing heavy calculation and programming + */ + if (pipe_ctx->plane_state->update_flags.bits.full_update) + dc->hwss.set_output_transfer_func(pipe_ctx, pipe_ctx->stream); +} + static void program_all_pipe_in_tree( struct dc *dc, struct pipe_ctx *pipe_ctx, @@ -2127,31 +2200,12 @@ static void program_all_pipe_in_tree( pipe_ctx->stream_res.tg->funcs->program_global_sync( pipe_ctx->stream_res.tg); - dc->hwss.blank_pixel_data(dc, &pipe_ctx->stream_res, - pipe_ctx->stream, blank); + dc->hwss.blank_pixel_data(dc, pipe_ctx, blank); + } if (pipe_ctx->plane_state != NULL) { - if (pipe_ctx->plane_state->update_flags.bits.full_update) - dcn10_enable_plane(dc, pipe_ctx, context); - - update_dchubp_dpp(dc, pipe_ctx, context); - - set_hdr_multiplier(pipe_ctx); - - if (pipe_ctx->plane_state->update_flags.bits.full_update || - pipe_ctx->plane_state->update_flags.bits.in_transfer_func_change || - pipe_ctx->plane_state->update_flags.bits.gamma_change) - dc->hwss.set_input_transfer_func(pipe_ctx, pipe_ctx->plane_state); - - /* dcn10_translate_regamma_to_hw_format takes 750us to finish - * only do gamma programming for full update. - * TODO: This can be further optimized/cleaned up - * Always call this for now since it does memcmp inside before - * doing heavy calculation and programming - */ - if (pipe_ctx->plane_state->update_flags.bits.full_update) - dc->hwss.set_output_transfer_func(pipe_ctx, pipe_ctx->stream); + dcn10_program_pipe(dc, pipe_ctx, context); } if (pipe_ctx->bottom_pipe != NULL && pipe_ctx->bottom_pipe != pipe_ctx) { @@ -2165,12 +2219,12 @@ static void dcn10_pplib_apply_display_requirements( { struct dm_pp_display_configuration *pp_display_cfg = &context->pp_display_cfg; - pp_display_cfg->min_engine_clock_khz = context->bw.dcn.cur_clk.dcfclk_khz; - pp_display_cfg->min_memory_clock_khz = context->bw.dcn.cur_clk.fclk_khz; - pp_display_cfg->min_engine_clock_deep_sleep_khz = context->bw.dcn.cur_clk.dcfclk_deep_sleep_khz; - pp_display_cfg->min_dcfc_deep_sleep_clock_khz = context->bw.dcn.cur_clk.dcfclk_deep_sleep_khz; - pp_display_cfg->min_dcfclock_khz = context->bw.dcn.cur_clk.dcfclk_khz; - pp_display_cfg->disp_clk_khz = context->bw.dcn.cur_clk.dispclk_khz; + pp_display_cfg->min_engine_clock_khz = dc->res_pool->dccg->clks.dcfclk_khz; + pp_display_cfg->min_memory_clock_khz = dc->res_pool->dccg->clks.fclk_khz; + pp_display_cfg->min_engine_clock_deep_sleep_khz = dc->res_pool->dccg->clks.dcfclk_deep_sleep_khz; + pp_display_cfg->min_dcfc_deep_sleep_clock_khz = dc->res_pool->dccg->clks.dcfclk_deep_sleep_khz; + pp_display_cfg->min_dcfclock_khz = dc->res_pool->dccg->clks.dcfclk_khz; + pp_display_cfg->disp_clk_khz = dc->res_pool->dccg->clks.dispclk_khz; dce110_fill_display_configs(context, pp_display_cfg); if (memcmp(&dc->prev_display_config, pp_display_cfg, sizeof( @@ -2232,8 +2286,6 @@ static void dcn10_apply_ctx_for_surface( int i; struct timing_generator *tg; bool removed_pipe[4] = { false }; - unsigned int ref_clk_mhz = dc->res_pool->ref_clock_inKhz/1000; - bool program_water_mark = false; struct pipe_ctx *top_pipe_to_program = find_top_pipe_for_stream(dc, context, stream); DC_LOGGER_INIT(dc->ctx->logger); @@ -2247,7 +2299,7 @@ static void dcn10_apply_ctx_for_surface( if (num_planes == 0) { /* OTG blank before remove all front end */ - dc->hwss.blank_pixel_data(dc, &top_pipe_to_program->stream_res, top_pipe_to_program->stream, true); + dc->hwss.blank_pixel_data(dc, top_pipe_to_program, true); } /* Disconnect unused mpcc */ @@ -2278,11 +2330,10 @@ static void dcn10_apply_ctx_for_surface( old_pipe_ctx->plane_state && old_pipe_ctx->stream_res.tg == tg) { - hwss1_plane_atomic_disconnect(dc, old_pipe_ctx); + dc->hwss.plane_atomic_disconnect(dc, old_pipe_ctx); removed_pipe[i] = true; - DC_LOG_DC( - "Reset mpcc for pipe %d\n", + DC_LOG_DC("Reset mpcc for pipe %d\n", old_pipe_ctx->pipe_idx); } } @@ -2295,248 +2346,41 @@ static void dcn10_apply_ctx_for_surface( if (num_planes == 0) false_optc_underflow_wa(dc, stream, tg); - for (i = 0; i < dc->res_pool->pipe_count; i++) { - struct pipe_ctx *old_pipe_ctx = - &dc->current_state->res_ctx.pipe_ctx[i]; - struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i]; - - if (pipe_ctx->stream == stream && - pipe_ctx->plane_state && - pipe_ctx->plane_state->update_flags.bits.full_update) - program_water_mark = true; - + for (i = 0; i < dc->res_pool->pipe_count; i++) if (removed_pipe[i]) - dcn10_disable_plane(dc, old_pipe_ctx); - } - - if (program_water_mark) { - if (dc->debug.sanity_checks) { - /* pstate stuck check after watermark update */ - dcn10_verify_allow_pstate_change_high(dc); - } - - /* watermark is for all pipes */ - hubbub1_program_watermarks(dc->res_pool->hubbub, - &context->bw.dcn.watermarks, ref_clk_mhz); + dcn10_disable_plane(dc, &dc->current_state->res_ctx.pipe_ctx[i]); - if (dc->debug.sanity_checks) { - /* pstate stuck check after watermark update */ - dcn10_verify_allow_pstate_change_high(dc); - } - } -/* DC_LOG_BANDWIDTH_CALCS(dc->ctx->logger, - "\n============== Watermark parameters ==============\n" - "a.urgent_ns: %d \n" - "a.cstate_enter_plus_exit: %d \n" - "a.cstate_exit: %d \n" - "a.pstate_change: %d \n" - "a.pte_meta_urgent: %d \n" - "b.urgent_ns: %d \n" - "b.cstate_enter_plus_exit: %d \n" - "b.cstate_exit: %d \n" - "b.pstate_change: %d \n" - "b.pte_meta_urgent: %d \n", - context->bw.dcn.watermarks.a.urgent_ns, - context->bw.dcn.watermarks.a.cstate_pstate.cstate_enter_plus_exit_ns, - context->bw.dcn.watermarks.a.cstate_pstate.cstate_exit_ns, - context->bw.dcn.watermarks.a.cstate_pstate.pstate_change_ns, - context->bw.dcn.watermarks.a.pte_meta_urgent_ns, - context->bw.dcn.watermarks.b.urgent_ns, - context->bw.dcn.watermarks.b.cstate_pstate.cstate_enter_plus_exit_ns, - context->bw.dcn.watermarks.b.cstate_pstate.cstate_exit_ns, - context->bw.dcn.watermarks.b.cstate_pstate.pstate_change_ns, - context->bw.dcn.watermarks.b.pte_meta_urgent_ns - ); - DC_LOG_BANDWIDTH_CALCS(dc->ctx->logger, - "\nc.urgent_ns: %d \n" - "c.cstate_enter_plus_exit: %d \n" - "c.cstate_exit: %d \n" - "c.pstate_change: %d \n" - "c.pte_meta_urgent: %d \n" - "d.urgent_ns: %d \n" - "d.cstate_enter_plus_exit: %d \n" - "d.cstate_exit: %d \n" - "d.pstate_change: %d \n" - "d.pte_meta_urgent: %d \n" - "========================================================\n", - context->bw.dcn.watermarks.c.urgent_ns, - context->bw.dcn.watermarks.c.cstate_pstate.cstate_enter_plus_exit_ns, - context->bw.dcn.watermarks.c.cstate_pstate.cstate_exit_ns, - context->bw.dcn.watermarks.c.cstate_pstate.pstate_change_ns, - context->bw.dcn.watermarks.c.pte_meta_urgent_ns, - context->bw.dcn.watermarks.d.urgent_ns, - context->bw.dcn.watermarks.d.cstate_pstate.cstate_enter_plus_exit_ns, - context->bw.dcn.watermarks.d.cstate_pstate.cstate_exit_ns, - context->bw.dcn.watermarks.d.cstate_pstate.pstate_change_ns, - context->bw.dcn.watermarks.d.pte_meta_urgent_ns - ); -*/ -} - -static inline bool should_set_clock(bool decrease_allowed, int calc_clk, int cur_clk) -{ - return ((decrease_allowed && calc_clk < cur_clk) || calc_clk > cur_clk); -} - -static int determine_dppclk_threshold(struct dc *dc, struct dc_state *context) -{ - bool request_dpp_div = context->bw.dcn.calc_clk.dispclk_khz > - context->bw.dcn.calc_clk.dppclk_khz; - bool dispclk_increase = context->bw.dcn.calc_clk.dispclk_khz > - context->bw.dcn.cur_clk.dispclk_khz; - int disp_clk_threshold = context->bw.dcn.calc_clk.max_supported_dppclk_khz; - bool cur_dpp_div = context->bw.dcn.cur_clk.dispclk_khz > - context->bw.dcn.cur_clk.dppclk_khz; - - /* increase clock, looking for div is 0 for current, request div is 1*/ - if (dispclk_increase) { - /* already divided by 2, no need to reach target clk with 2 steps*/ - if (cur_dpp_div) - return context->bw.dcn.calc_clk.dispclk_khz; - - /* request disp clk is lower than maximum supported dpp clk, - * no need to reach target clk with two steps. - */ - if (context->bw.dcn.calc_clk.dispclk_khz <= disp_clk_threshold) - return context->bw.dcn.calc_clk.dispclk_khz; - - /* target dpp clk not request divided by 2, still within threshold */ - if (!request_dpp_div) - return context->bw.dcn.calc_clk.dispclk_khz; - - } else { - /* decrease clock, looking for current dppclk divided by 2, - * request dppclk not divided by 2. - */ - - /* current dpp clk not divided by 2, no need to ramp*/ - if (!cur_dpp_div) - return context->bw.dcn.calc_clk.dispclk_khz; - - /* current disp clk is lower than current maximum dpp clk, - * no need to ramp - */ - if (context->bw.dcn.cur_clk.dispclk_khz <= disp_clk_threshold) - return context->bw.dcn.calc_clk.dispclk_khz; - - /* request dpp clk need to be divided by 2 */ - if (request_dpp_div) - return context->bw.dcn.calc_clk.dispclk_khz; - } - - return disp_clk_threshold; -} - -static void ramp_up_dispclk_with_dpp(struct dc *dc, struct dc_state *context) -{ - int i; - bool request_dpp_div = context->bw.dcn.calc_clk.dispclk_khz > - context->bw.dcn.calc_clk.dppclk_khz; - - int dispclk_to_dpp_threshold = determine_dppclk_threshold(dc, context); - - /* set disp clk to dpp clk threshold */ - dc->res_pool->display_clock->funcs->set_clock( - dc->res_pool->display_clock, - dispclk_to_dpp_threshold); - - /* update request dpp clk division option */ - for (i = 0; i < dc->res_pool->pipe_count; i++) { - struct pipe_ctx *pipe_ctx = &dc->current_state->res_ctx.pipe_ctx[i]; - - if (!pipe_ctx->plane_state) - continue; - - pipe_ctx->plane_res.dpp->funcs->dpp_dppclk_control( - pipe_ctx->plane_res.dpp, - request_dpp_div, - true); - } - - /* If target clk not same as dppclk threshold, set to target clock */ - if (dispclk_to_dpp_threshold != context->bw.dcn.calc_clk.dispclk_khz) { - dc->res_pool->display_clock->funcs->set_clock( - dc->res_pool->display_clock, - context->bw.dcn.calc_clk.dispclk_khz); - } - - context->bw.dcn.cur_clk.dispclk_khz = - context->bw.dcn.calc_clk.dispclk_khz; - context->bw.dcn.cur_clk.dppclk_khz = - context->bw.dcn.calc_clk.dppclk_khz; - context->bw.dcn.cur_clk.max_supported_dppclk_khz = - context->bw.dcn.calc_clk.max_supported_dppclk_khz; + if (dc->hwseq->wa.DEGVIDCN10_254) + hubbub1_wm_change_req_wa(dc->res_pool->hubbub); } static void dcn10_set_bandwidth( struct dc *dc, struct dc_state *context, - bool decrease_allowed) + bool safe_to_lower) { - struct pp_smu_display_requirement_rv *smu_req_cur = - &dc->res_pool->pp_smu_req; - struct pp_smu_display_requirement_rv smu_req = *smu_req_cur; - struct pp_smu_funcs_rv *pp_smu = dc->res_pool->pp_smu; - - if (dc->debug.sanity_checks) { + if (dc->debug.sanity_checks) dcn10_verify_allow_pstate_change_high(dc); - } - - if (IS_FPGA_MAXIMUS_DC(dc->ctx->dce_environment)) - return; - - if (should_set_clock( - decrease_allowed, - context->bw.dcn.calc_clk.dcfclk_khz, - dc->current_state->bw.dcn.cur_clk.dcfclk_khz)) { - context->bw.dcn.cur_clk.dcfclk_khz = - context->bw.dcn.calc_clk.dcfclk_khz; - smu_req.hard_min_dcefclk_khz = - context->bw.dcn.calc_clk.dcfclk_khz; - } - - if (should_set_clock( - decrease_allowed, - context->bw.dcn.calc_clk.dcfclk_deep_sleep_khz, - dc->current_state->bw.dcn.cur_clk.dcfclk_deep_sleep_khz)) { - context->bw.dcn.cur_clk.dcfclk_deep_sleep_khz = - context->bw.dcn.calc_clk.dcfclk_deep_sleep_khz; - } - - if (should_set_clock( - decrease_allowed, - context->bw.dcn.calc_clk.fclk_khz, - dc->current_state->bw.dcn.cur_clk.fclk_khz)) { - context->bw.dcn.cur_clk.fclk_khz = - context->bw.dcn.calc_clk.fclk_khz; - smu_req.hard_min_fclk_khz = context->bw.dcn.calc_clk.fclk_khz; - } - - smu_req.display_count = context->stream_count; - - if (pp_smu->set_display_requirement) - pp_smu->set_display_requirement(&pp_smu->pp_smu, &smu_req); - *smu_req_cur = smu_req; + if (!IS_FPGA_MAXIMUS_DC(dc->ctx->dce_environment)) { + if (context->stream_count == 0) + context->bw.dcn.clk.phyclk_khz = 0; - /* make sure dcf clk is before dpp clk to - * make sure we have enough voltage to run dpp clk - */ - if (should_set_clock( - decrease_allowed, - context->bw.dcn.calc_clk.dispclk_khz, - dc->current_state->bw.dcn.cur_clk.dispclk_khz)) { + dc->res_pool->dccg->funcs->update_clocks( + dc->res_pool->dccg, + &context->bw.dcn.clk, + safe_to_lower); - ramp_up_dispclk_with_dpp(dc, context); + dcn10_pplib_apply_display_requirements(dc, context); } - dcn10_pplib_apply_display_requirements(dc, context); + hubbub1_program_watermarks(dc->res_pool->hubbub, + &context->bw.dcn.watermarks, + dc->res_pool->ref_clock_inKhz / 1000, + true); - if (dc->debug.sanity_checks) { + if (dc->debug.sanity_checks) dcn10_verify_allow_pstate_change_high(dc); - } - - /* need to fix this function. not doing the right thing here */ } static void set_drr(struct pipe_ctx **pipe_ctx, @@ -2701,16 +2545,20 @@ static void dcn10_update_pending_status(struct pipe_ctx *pipe_ctx) { struct dc_plane_state *plane_state = pipe_ctx->plane_state; struct timing_generator *tg = pipe_ctx->stream_res.tg; + bool flip_pending; if (plane_state == NULL) return; - plane_state->status.is_flip_pending = - pipe_ctx->plane_res.hubp->funcs->hubp_is_flip_pending( + flip_pending = pipe_ctx->plane_res.hubp->funcs->hubp_is_flip_pending( pipe_ctx->plane_res.hubp); - plane_state->status.current_address = pipe_ctx->plane_res.hubp->current_address; - if (pipe_ctx->plane_res.hubp->current_address.type == PLN_ADDR_TYPE_GRPH_STEREO && + plane_state->status.is_flip_pending = flip_pending; + + if (!flip_pending) + plane_state->status.current_address = plane_state->status.requested_address; + + if (plane_state->status.current_address.type == PLN_ADDR_TYPE_GRPH_STEREO && tg->funcs->is_stereo_left_eye) { plane_state->status.is_right_eye = !tg->funcs->is_stereo_left_eye(pipe_ctx->stream_res.tg); @@ -2719,8 +2567,14 @@ static void dcn10_update_pending_status(struct pipe_ctx *pipe_ctx) static void dcn10_update_dchub(struct dce_hwseq *hws, struct dchub_init_data *dh_data) { - if (hws->ctx->dc->res_pool->hubbub != NULL) - hubbub1_update_dchub(hws->ctx->dc->res_pool->hubbub, dh_data); + if (hws->ctx->dc->res_pool->hubbub != NULL) { + struct hubp *hubp = hws->ctx->dc->res_pool->hubps[0]; + + if (hubp->funcs->hubp_update_dchub) + hubp->funcs->hubp_update_dchub(hubp, dh_data); + else + hubbub1_update_dchub(hws->ctx->dc->res_pool->hubbub, dh_data); + } } static void dcn10_set_cursor_position(struct pipe_ctx *pipe_ctx) @@ -2731,9 +2585,11 @@ static void dcn10_set_cursor_position(struct pipe_ctx *pipe_ctx) struct dc_cursor_mi_param param = { .pixel_clk_khz = pipe_ctx->stream->timing.pix_clk_khz, .ref_clk_khz = pipe_ctx->stream->ctx->dc->res_pool->ref_clock_inKhz, - .viewport_x_start = pipe_ctx->plane_res.scl_data.viewport.x, - .viewport_width = pipe_ctx->plane_res.scl_data.viewport.width, - .h_scale_ratio = pipe_ctx->plane_res.scl_data.ratios.horz + .viewport = pipe_ctx->plane_res.scl_data.viewport, + .h_scale_ratio = pipe_ctx->plane_res.scl_data.ratios.horz, + .v_scale_ratio = pipe_ctx->plane_res.scl_data.ratios.vert, + .rotation = pipe_ctx->plane_state->rotation, + .mirror = pipe_ctx->plane_state->horizontal_mirror }; if (pipe_ctx->plane_state->address.type @@ -2757,6 +2613,33 @@ static void dcn10_set_cursor_attribute(struct pipe_ctx *pipe_ctx) pipe_ctx->plane_res.dpp, attributes->color_format); } +static void dcn10_set_cursor_sdr_white_level(struct pipe_ctx *pipe_ctx) +{ + uint32_t sdr_white_level = pipe_ctx->stream->cursor_attributes.sdr_white_level; + struct fixed31_32 multiplier; + struct dpp_cursor_attributes opt_attr = { 0 }; + uint32_t hw_scale = 0x3c00; // 1.0 default multiplier + struct custom_float_format fmt; + + if (!pipe_ctx->plane_res.dpp->funcs->set_optional_cursor_attributes) + return; + + fmt.exponenta_bits = 5; + fmt.mantissa_bits = 10; + fmt.sign = true; + + if (sdr_white_level > 80) { + multiplier = dc_fixpt_from_fraction(sdr_white_level, 80); + convert_to_custom_float_format(multiplier, &fmt, &hw_scale); + } + + opt_attr.scale = hw_scale; + opt_attr.bias = 0; + + pipe_ctx->plane_res.dpp->funcs->set_optional_cursor_attributes( + pipe_ctx->plane_res.dpp, &opt_attr); +} + static const struct hw_sequencer_funcs dcn10_funcs = { .program_gamut_remap = program_gamut_remap, .program_csc_matrix = program_csc_matrix, @@ -2764,7 +2647,9 @@ static const struct hw_sequencer_funcs dcn10_funcs = { .apply_ctx_to_hw = dce110_apply_ctx_to_hw, .apply_ctx_for_surface = dcn10_apply_ctx_for_surface, .update_plane_addr = dcn10_update_plane_addr, + .plane_atomic_disconnect = hwss1_plane_atomic_disconnect, .update_dchub = dcn10_update_dchub, + .update_mpcc = dcn10_update_mpcc, .update_pending_status = dcn10_update_pending_status, .set_input_transfer_func = dcn10_set_input_transfer_func, .set_output_transfer_func = dcn10_set_output_transfer_func, @@ -2778,6 +2663,8 @@ static const struct hw_sequencer_funcs dcn10_funcs = { .disable_stream = dce110_disable_stream, .unblank_stream = dce110_unblank_stream, .blank_stream = dce110_blank_stream, + .enable_audio_stream = dce110_enable_audio_stream, + .disable_audio_stream = dce110_disable_audio_stream, .enable_display_power_gating = dcn10_dummy_display_power_gating, .disable_plane = dcn10_disable_plane, .blank_pixel_data = dcn10_blank_pixel_data, @@ -2800,7 +2687,8 @@ static const struct hw_sequencer_funcs dcn10_funcs = { .edp_power_control = hwss_edp_power_control, .edp_wait_for_hpd_ready = hwss_edp_wait_for_hpd_ready, .set_cursor_position = dcn10_set_cursor_position, - .set_cursor_attribute = dcn10_set_cursor_attribute + .set_cursor_attribute = dcn10_set_cursor_attribute, + .set_cursor_sdr_white_level = dcn10_set_cursor_sdr_white_level }; diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.h index 44f734b73f9e..7139fb73e966 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.h +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.h @@ -39,4 +39,11 @@ bool is_rgb_cspace(enum dc_color_space output_color_space); void hwss1_plane_atomic_disconnect(struct dc *dc, struct pipe_ctx *pipe_ctx); +void dcn10_verify_allow_pstate_change_high(struct dc *dc); + +void dcn10_program_pipe( + struct dc *dc, + struct pipe_ctx *pipe_ctx, + struct dc_state *context); + #endif /* __DC_HWSS_DCN10_H__ */ diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.c index 21fa40ac0786..6f675206a136 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.c @@ -65,11 +65,6 @@ enum { DP_MST_UPDATE_MAX_RETRY = 50 }; - - -static void aux_initialize(struct dcn10_link_encoder *enc10); - - static const struct link_encoder_funcs dcn10_lnk_enc_funcs = { .validate_output_with_stream = dcn10_link_encoder_validate_output_with_stream, @@ -445,12 +440,11 @@ static uint8_t get_frontend_source( } } -static void configure_encoder( +void configure_encoder( struct dcn10_link_encoder *enc10, const struct dc_link_settings *link_settings) { /* set number of lanes */ - REG_SET(DP_CONFIG, 0, DP_UDI_LANES, link_settings->lane_count - LANE_COUNT_ONE); @@ -602,6 +596,9 @@ static bool dcn10_link_encoder_validate_hdmi_output( if (!enc10->base.features.flags.bits.HDMI_6GB_EN && adjusted_pix_clk_khz >= 300000) return false; + if (enc10->base.ctx->dc->debug.hdmi20_disable && + crtc_timing->pixel_encoding == PIXEL_ENCODING_YCBCR420) + return false; return true; } @@ -734,6 +731,9 @@ void dcn10_link_encoder_construct( __func__, result); } + if (enc10->base.ctx->dc->debug.hdmi20_disable) { + enc10->base.features.flags.bits.HDMI_6GB_EN = 0; + } } bool dcn10_link_encoder_validate_output_with_stream( @@ -812,7 +812,7 @@ void dcn10_link_encoder_hw_init( ASSERT(result == BP_RESULT_OK); } - aux_initialize(enc10); + dcn10_aux_initialize(enc10); /* reinitialize HPD. * hpd_initialize() will pass DIG_FE id to HW context. @@ -995,6 +995,8 @@ void dcn10_link_encoder_disable_output( if (!dcn10_is_dig_enabled(enc)) { /* OF_SKIP_POWER_DOWN_INACTIVE_ENCODER */ + /*in DP_Alt_No_Connect case, we turn off the dig already, + after excuation the PHY w/a sequence, not allow touch PHY any more*/ return; } /* Power-down RX and disable GPU PHY should be paired. @@ -1347,8 +1349,7 @@ void dcn10_link_encoder_disable_hpd(struct link_encoder *enc) FN(reg, f1), v1,\ FN(reg, f2), v2) -static void aux_initialize( - struct dcn10_link_encoder *enc10) +void dcn10_aux_initialize(struct dcn10_link_encoder *enc10) { enum hpd_source_id hpd_source = enc10->base.hpd_source; diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.h index 2a97cdb2cfbb..49ead12b2532 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.h +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.h @@ -42,6 +42,7 @@ #define LE_DCN_COMMON_REG_LIST(id) \ SRI(DIG_BE_CNTL, DIG, id), \ SRI(DIG_BE_EN_CNTL, DIG, id), \ + SRI(TMDS_CTL_BITS, DIG, id), \ SRI(DP_CONFIG, DP, id), \ SRI(DP_DPHY_CNTL, DP, id), \ SRI(DP_DPHY_PRBS_CNTL, DP, id), \ @@ -64,6 +65,7 @@ SRI(DP_DPHY_INTERNAL_CTRL, DP, id), \ SRI(DP_DPHY_HBR2_PATTERN_CONTROL, DP, id) + #define LE_DCN10_REG_LIST(id)\ LE_DCN_COMMON_REG_LIST(id) @@ -100,6 +102,7 @@ struct dcn10_link_enc_registers { uint32_t DP_DPHY_BS_SR_SWAP_CNTL; uint32_t DP_DPHY_HBR2_PATTERN_CONTROL; uint32_t DP_SEC_CNTL1; + uint32_t TMDS_CTL_BITS; }; #define LE_SF(reg_name, field_name, post_fix)\ @@ -110,6 +113,7 @@ struct dcn10_link_enc_registers { LE_SF(DIG0_DIG_BE_CNTL, DIG_HPD_SELECT, mask_sh),\ LE_SF(DIG0_DIG_BE_CNTL, DIG_MODE, mask_sh),\ LE_SF(DIG0_DIG_BE_CNTL, DIG_FE_SOURCE_SELECT, mask_sh),\ + LE_SF(DIG0_TMDS_CTL_BITS, TMDS_CTL0, mask_sh), \ LE_SF(DP0_DP_DPHY_CNTL, DPHY_BYPASS, mask_sh),\ LE_SF(DP0_DP_DPHY_CNTL, DPHY_ATEST_SEL_LANE0, mask_sh),\ LE_SF(DP0_DP_DPHY_CNTL, DPHY_ATEST_SEL_LANE1, mask_sh),\ @@ -198,10 +202,11 @@ struct dcn10_link_enc_registers { type DP_MSE_SAT_SLOT_COUNT3;\ type DP_MSE_SAT_UPDATE;\ type DP_MSE_16_MTP_KEEPOUT;\ + type DC_HPD_EN;\ + type TMDS_CTL0;\ type AUX_HPD_SEL;\ type AUX_LS_READ_EN;\ - type AUX_RX_RECEIVE_WINDOW;\ - type DC_HPD_EN + type AUX_RX_RECEIVE_WINDOW struct dcn10_link_enc_shift { DCN_LINK_ENCODER_REG_FIELD_LIST(uint8_t); @@ -266,6 +271,10 @@ void dcn10_link_encoder_setup( struct link_encoder *enc, enum signal_type signal); +void configure_encoder( + struct dcn10_link_encoder *enc10, + const struct dc_link_settings *link_settings); + /* enables TMDS PHY output */ /* TODO: still need depth or just pass in adjusted pixel clock? */ void dcn10_link_encoder_enable_tmds_output( @@ -327,4 +336,6 @@ void dcn10_psr_program_secondary_packet(struct link_encoder *enc, bool dcn10_is_dig_enabled(struct link_encoder *enc); +void dcn10_aux_initialize(struct dcn10_link_encoder *enc10); + #endif /* __DC_LINK_ENCODER__DCN10_H__ */ diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_mpc.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_mpc.c index 9ca51ae46de7..958994edf2c4 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_mpc.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_mpc.c @@ -428,7 +428,7 @@ void mpc1_read_mpcc_state( MPCC_BUSY, &s->busy); } -const struct mpc_funcs dcn10_mpc_funcs = { +static const struct mpc_funcs dcn10_mpc_funcs = { .read_mpcc_state = mpc1_read_mpcc_state, .insert_plane = mpc1_insert_plane, .remove_mpcc = mpc1_remove_mpcc, diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_opp.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_opp.c index 77a1a9d541a4..ab958cff3b76 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_opp.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_opp.c @@ -385,7 +385,7 @@ void opp1_destroy(struct output_pixel_processor **opp) *opp = NULL; } -static struct opp_funcs dcn10_opp_funcs = { +static const struct opp_funcs dcn10_opp_funcs = { .opp_set_dyn_expansion = opp1_set_dyn_expansion, .opp_program_fmt = opp1_program_fmt, .opp_program_bit_depth_reduction = opp1_program_bit_depth_reduction, diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c index f2fbce0e3fc5..411f89218e01 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c @@ -1257,6 +1257,37 @@ void optc1_read_otg_state(struct optc *optc1, OPTC_UNDERFLOW_OCCURRED_STATUS, &s->underflow_occurred_status); } +bool optc1_get_otg_active_size(struct timing_generator *optc, + uint32_t *otg_active_width, + uint32_t *otg_active_height) +{ + uint32_t otg_enabled; + uint32_t v_blank_start; + uint32_t v_blank_end; + uint32_t h_blank_start; + uint32_t h_blank_end; + struct optc *optc1 = DCN10TG_FROM_TG(optc); + + + REG_GET(OTG_CONTROL, + OTG_MASTER_EN, &otg_enabled); + + if (otg_enabled == 0) + return false; + + REG_GET_2(OTG_V_BLANK_START_END, + OTG_V_BLANK_START, &v_blank_start, + OTG_V_BLANK_END, &v_blank_end); + + REG_GET_2(OTG_H_BLANK_START_END, + OTG_H_BLANK_START, &h_blank_start, + OTG_H_BLANK_END, &h_blank_end); + + *otg_active_width = v_blank_start - v_blank_end; + *otg_active_height = h_blank_start - h_blank_end; + return true; +} + void optc1_clear_optc_underflow(struct timing_generator *optc) { struct optc *optc1 = DCN10TG_FROM_TG(optc); @@ -1293,6 +1324,72 @@ bool optc1_is_optc_underflow_occurred(struct timing_generator *optc) return (underflow_occurred == 1); } +bool optc1_configure_crc(struct timing_generator *optc, + const struct crc_params *params) +{ + struct optc *optc1 = DCN10TG_FROM_TG(optc); + + /* Cannot configure crc on a CRTC that is disabled */ + if (!optc1_is_tg_enabled(optc)) + return false; + + REG_WRITE(OTG_CRC_CNTL, 0); + + if (!params->enable) + return true; + + /* Program frame boundaries */ + /* Window A x axis start and end. */ + REG_UPDATE_2(OTG_CRC0_WINDOWA_X_CONTROL, + OTG_CRC0_WINDOWA_X_START, params->windowa_x_start, + OTG_CRC0_WINDOWA_X_END, params->windowa_x_end); + + /* Window A y axis start and end. */ + REG_UPDATE_2(OTG_CRC0_WINDOWA_Y_CONTROL, + OTG_CRC0_WINDOWA_Y_START, params->windowa_y_start, + OTG_CRC0_WINDOWA_Y_END, params->windowa_y_end); + + /* Window B x axis start and end. */ + REG_UPDATE_2(OTG_CRC0_WINDOWB_X_CONTROL, + OTG_CRC0_WINDOWB_X_START, params->windowb_x_start, + OTG_CRC0_WINDOWB_X_END, params->windowb_x_end); + + /* Window B y axis start and end. */ + REG_UPDATE_2(OTG_CRC0_WINDOWB_Y_CONTROL, + OTG_CRC0_WINDOWB_Y_START, params->windowb_y_start, + OTG_CRC0_WINDOWB_Y_END, params->windowb_y_end); + + /* Set crc mode and selection, and enable. Only using CRC0*/ + REG_UPDATE_3(OTG_CRC_CNTL, + OTG_CRC_CONT_EN, params->continuous_mode ? 1 : 0, + OTG_CRC0_SELECT, params->selection, + OTG_CRC_EN, 1); + + return true; +} + +bool optc1_get_crc(struct timing_generator *optc, + uint32_t *r_cr, uint32_t *g_y, uint32_t *b_cb) +{ + uint32_t field = 0; + struct optc *optc1 = DCN10TG_FROM_TG(optc); + + REG_GET(OTG_CRC_CNTL, OTG_CRC_EN, &field); + + /* Early return if CRC is not enabled for this CRTC */ + if (!field) + return false; + + REG_GET_2(OTG_CRC0_DATA_RG, + CRC0_R_CR, r_cr, + CRC0_G_Y, g_y); + + REG_GET(OTG_CRC0_DATA_B, + CRC0_B_CB, b_cb); + + return true; +} + static const struct timing_generator_funcs dcn10_tg_funcs = { .validate_timing = optc1_validate_timing, .program_timing = optc1_program_timing, @@ -1305,6 +1402,7 @@ static const struct timing_generator_funcs dcn10_tg_funcs = { .get_position = optc1_get_position, .get_frame_count = optc1_get_vblank_counter, .get_scanoutpos = optc1_get_crtc_scanoutpos, + .get_otg_active_size = optc1_get_otg_active_size, .set_early_control = optc1_set_early_control, /* used by enable_timing_synchronization. Not need for FPGA */ .wait_for_state = optc1_wait_for_state, @@ -1328,6 +1426,8 @@ static const struct timing_generator_funcs dcn10_tg_funcs = { .is_tg_enabled = optc1_is_tg_enabled, .is_optc_underflow_occurred = optc1_is_optc_underflow_occurred, .clear_optc_underflow = optc1_clear_optc_underflow, + .get_crc = optc1_get_crc, + .configure_crc = optc1_configure_crc, }; void dcn10_timing_generator_init(struct optc *optc1) diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h index c62052f46460..c1b114209fe8 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h @@ -75,7 +75,14 @@ SRI(CONTROL, VTG, inst),\ SRI(OTG_VERT_SYNC_CONTROL, OTG, inst),\ SRI(OTG_MASTER_UPDATE_MODE, OTG, inst),\ - SRI(OTG_GSL_CONTROL, OTG, inst) + SRI(OTG_GSL_CONTROL, OTG, inst),\ + SRI(OTG_CRC_CNTL, OTG, inst),\ + SRI(OTG_CRC0_DATA_RG, OTG, inst),\ + SRI(OTG_CRC0_DATA_B, OTG, inst),\ + SRI(OTG_CRC0_WINDOWA_X_CONTROL, OTG, inst),\ + SRI(OTG_CRC0_WINDOWA_Y_CONTROL, OTG, inst),\ + SRI(OTG_CRC0_WINDOWB_X_CONTROL, OTG, inst),\ + SRI(OTG_CRC0_WINDOWB_Y_CONTROL, OTG, inst) #define TG_COMMON_REG_LIST_DCN1_0(inst) \ TG_COMMON_REG_LIST_DCN(inst),\ @@ -138,6 +145,13 @@ struct dcn_optc_registers { uint32_t OTG_GSL_WINDOW_X; uint32_t OTG_GSL_WINDOW_Y; uint32_t OTG_VUPDATE_KEEPOUT; + uint32_t OTG_CRC_CNTL; + uint32_t OTG_CRC0_DATA_RG; + uint32_t OTG_CRC0_DATA_B; + uint32_t OTG_CRC0_WINDOWA_X_CONTROL; + uint32_t OTG_CRC0_WINDOWA_Y_CONTROL; + uint32_t OTG_CRC0_WINDOWB_X_CONTROL; + uint32_t OTG_CRC0_WINDOWB_Y_CONTROL; }; #define TG_COMMON_MASK_SH_LIST_DCN(mask_sh)\ @@ -232,7 +246,21 @@ struct dcn_optc_registers { SF(OTG0_OTG_GSL_CONTROL, OTG_GSL2_EN, mask_sh),\ SF(OTG0_OTG_GSL_CONTROL, OTG_GSL_MASTER_EN, mask_sh),\ SF(OTG0_OTG_GSL_CONTROL, OTG_GSL_FORCE_DELAY, mask_sh),\ - SF(OTG0_OTG_GSL_CONTROL, OTG_GSL_CHECK_ALL_FIELDS, mask_sh) + SF(OTG0_OTG_GSL_CONTROL, OTG_GSL_CHECK_ALL_FIELDS, mask_sh),\ + SF(OTG0_OTG_CRC_CNTL, OTG_CRC_CONT_EN, mask_sh),\ + SF(OTG0_OTG_CRC_CNTL, OTG_CRC0_SELECT, mask_sh),\ + SF(OTG0_OTG_CRC_CNTL, OTG_CRC_EN, mask_sh),\ + SF(OTG0_OTG_CRC0_DATA_RG, CRC0_R_CR, mask_sh),\ + SF(OTG0_OTG_CRC0_DATA_RG, CRC0_G_Y, mask_sh),\ + SF(OTG0_OTG_CRC0_DATA_B, CRC0_B_CB, mask_sh),\ + SF(OTG0_OTG_CRC0_WINDOWA_X_CONTROL, OTG_CRC0_WINDOWA_X_START, mask_sh),\ + SF(OTG0_OTG_CRC0_WINDOWA_X_CONTROL, OTG_CRC0_WINDOWA_X_END, mask_sh),\ + SF(OTG0_OTG_CRC0_WINDOWA_Y_CONTROL, OTG_CRC0_WINDOWA_Y_START, mask_sh),\ + SF(OTG0_OTG_CRC0_WINDOWA_Y_CONTROL, OTG_CRC0_WINDOWA_Y_END, mask_sh),\ + SF(OTG0_OTG_CRC0_WINDOWB_X_CONTROL, OTG_CRC0_WINDOWB_X_START, mask_sh),\ + SF(OTG0_OTG_CRC0_WINDOWB_X_CONTROL, OTG_CRC0_WINDOWB_X_END, mask_sh),\ + SF(OTG0_OTG_CRC0_WINDOWB_Y_CONTROL, OTG_CRC0_WINDOWB_Y_START, mask_sh),\ + SF(OTG0_OTG_CRC0_WINDOWB_Y_CONTROL, OTG_CRC0_WINDOWB_Y_END, mask_sh) #define TG_COMMON_MASK_SH_LIST_DCN1_0(mask_sh)\ @@ -363,7 +391,22 @@ struct dcn_optc_registers { type OTG_MASTER_UPDATE_LOCK_GSL_EN;\ type MASTER_UPDATE_LOCK_VUPDATE_KEEPOUT_START_OFFSET;\ type MASTER_UPDATE_LOCK_VUPDATE_KEEPOUT_END_OFFSET;\ - type OTG_MASTER_UPDATE_LOCK_VUPDATE_KEEPOUT_EN; + type OTG_MASTER_UPDATE_LOCK_VUPDATE_KEEPOUT_EN;\ + type OTG_CRC_CONT_EN;\ + type OTG_CRC0_SELECT;\ + type OTG_CRC_EN;\ + type CRC0_R_CR;\ + type CRC0_G_Y;\ + type CRC0_B_CB;\ + type OTG_CRC0_WINDOWA_X_START;\ + type OTG_CRC0_WINDOWA_X_END;\ + type OTG_CRC0_WINDOWA_Y_START;\ + type OTG_CRC0_WINDOWA_Y_END;\ + type OTG_CRC0_WINDOWB_X_START;\ + type OTG_CRC0_WINDOWB_X_END;\ + type OTG_CRC0_WINDOWB_Y_START;\ + type OTG_CRC0_WINDOWB_Y_END; + #define TG_REG_FIELD_LIST(type) \ TG_REG_FIELD_LIST_DCN1_0(type) @@ -507,4 +550,19 @@ bool optc1_is_optc_underflow_occurred(struct timing_generator *optc); void optc1_set_blank_data_double_buffer(struct timing_generator *optc, bool enable); +bool optc1_get_otg_active_size(struct timing_generator *optc, + uint32_t *otg_active_width, + uint32_t *otg_active_height); + +void optc1_enable_crtc_reset( + struct timing_generator *optc, + int source_tg_inst, + struct crtc_trigger_info *crtc_tp); + +bool optc1_configure_crc(struct timing_generator *optc, + const struct crc_params *params); + +bool optc1_get_crc(struct timing_generator *optc, + uint32_t *r_cr, uint32_t *g_y, uint32_t *b_cb); + #endif /* __DC_TIMING_GENERATOR_DCN10_H__ */ diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c index 34dac84066a0..6b44ed3697a4 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c @@ -64,6 +64,69 @@ #include "reg_helper.h" #include "dce/dce_abm.h" #include "dce/dce_dmcu.h" +#include "dce/dce_aux.h" + +const struct _vcs_dpi_ip_params_st dcn1_0_ip = { + .rob_buffer_size_kbytes = 64, + .det_buffer_size_kbytes = 164, + .dpte_buffer_size_in_pte_reqs = 42, + .dpp_output_buffer_pixels = 2560, + .opp_output_buffer_lines = 1, + .pixel_chunk_size_kbytes = 8, + .pte_enable = 1, + .pte_chunk_size_kbytes = 2, + .meta_chunk_size_kbytes = 2, + .writeback_chunk_size_kbytes = 2, + .line_buffer_size_bits = 589824, + .max_line_buffer_lines = 12, + .IsLineBufferBppFixed = 0, + .LineBufferFixedBpp = -1, + .writeback_luma_buffer_size_kbytes = 12, + .writeback_chroma_buffer_size_kbytes = 8, + .max_num_dpp = 4, + .max_num_wb = 2, + .max_dchub_pscl_bw_pix_per_clk = 4, + .max_pscl_lb_bw_pix_per_clk = 2, + .max_lb_vscl_bw_pix_per_clk = 4, + .max_vscl_hscl_bw_pix_per_clk = 4, + .max_hscl_ratio = 4, + .max_vscl_ratio = 4, + .hscl_mults = 4, + .vscl_mults = 4, + .max_hscl_taps = 8, + .max_vscl_taps = 8, + .dispclk_ramp_margin_percent = 1, + .underscan_factor = 1.10, + .min_vblank_lines = 14, + .dppclk_delay_subtotal = 90, + .dispclk_delay_subtotal = 42, + .dcfclk_cstate_latency = 10, + .max_inter_dcn_tile_repeaters = 8, + .can_vstartup_lines_exceed_vsync_plus_back_porch_lines_minus_one = 0, + .bug_forcing_LC_req_same_size_fixed = 0, +}; + +const struct _vcs_dpi_soc_bounding_box_st dcn1_0_soc = { + .sr_exit_time_us = 9.0, + .sr_enter_plus_exit_time_us = 11.0, + .urgent_latency_us = 4.0, + .writeback_latency_us = 12.0, + .ideal_dram_bw_after_urgent_percent = 80.0, + .max_request_size_bytes = 256, + .downspread_percent = 0.5, + .dram_page_open_time_ns = 50.0, + .dram_rw_turnaround_time_ns = 17.5, + .dram_return_buffer_per_channel_bytes = 8192, + .round_trip_ping_latency_dcfclk_cycles = 128, + .urgent_out_of_order_return_per_channel_bytes = 256, + .channel_interleave_bytes = 256, + .num_banks = 8, + .num_chans = 2, + .vmm_page_size_bytes = 4096, + .dram_clock_change_latency_us = 17.0, + .writeback_dram_clock_change_latency_us = 23.0, + .return_bus_width_bytes = 64, +}; #ifndef mmDP0_DP_DPHY_INTERNAL_CTRL #define mmDP0_DP_DPHY_INTERNAL_CTRL 0x210f @@ -294,6 +357,21 @@ static const struct dcn10_opp_mask opp_mask = { OPP_MASK_SH_LIST_DCN10(_MASK), }; +#define aux_engine_regs(id)\ +[id] = {\ + AUX_COMMON_REG_LIST(id), \ + .AUX_RESET_MASK = 0 \ +} + +static const struct dce110_aux_registers aux_engine_regs[] = { + aux_engine_regs(0), + aux_engine_regs(1), + aux_engine_regs(2), + aux_engine_regs(3), + aux_engine_regs(4), + aux_engine_regs(5) +}; + #define tf_regs(id)\ [id] = {\ TF_REG_LIST_DCN10(id),\ @@ -417,13 +495,14 @@ static const struct dce110_clk_src_mask cs_mask = { static const struct resource_caps res_cap = { .num_timing_generator = 4, + .num_opp = 4, .num_video_plane = 4, .num_audio = 4, .num_stream_encoder = 4, .num_pll = 4, }; -static const struct dc_debug debug_defaults_drv = { +static const struct dc_debug_options debug_defaults_drv = { .sanity_checks = true, .disable_dmcu = true, .force_abm_enable = false, @@ -436,7 +515,7 @@ static const struct dc_debug debug_defaults_drv = { */ .min_disp_clk_khz = 100000, - .disable_pplib_clock_request = true, + .disable_pplib_clock_request = false, .disable_pplib_wm_range = false, .pplib_wm_report_mode = WM_REPORT_DEFAULT, .pipe_split_policy = MPC_SPLIT_AVOID_MULT_DISP, @@ -451,7 +530,7 @@ static const struct dc_debug debug_defaults_drv = { .max_downscale_src_width = 3840, }; -static const struct dc_debug debug_defaults_diags = { +static const struct dc_debug_options debug_defaults_diags = { .disable_dmcu = true, .force_abm_enable = false, .timing_trace = true, @@ -515,6 +594,23 @@ static struct output_pixel_processor *dcn10_opp_create( return &opp->base; } +struct aux_engine *dcn10_aux_engine_create( + struct dc_context *ctx, + uint32_t inst) +{ + struct aux_engine_dce110 *aux_engine = + kzalloc(sizeof(struct aux_engine_dce110), GFP_KERNEL); + + if (!aux_engine) + return NULL; + + dce110_aux_engine_construct(aux_engine, ctx, inst, + SW_AUX_TIMEOUT_PERIOD_MULTIPLIER * AUX_TIMEOUT_PERIOD, + &aux_engine_regs[inst]); + + return &aux_engine->base; +} + static struct mpc *dcn10_mpc_create(struct dc_context *ctx) { struct dcn10_mpc *mpc10 = kzalloc(sizeof(struct dcn10_mpc), @@ -680,6 +776,7 @@ static struct dce_hwseq *dcn10_hwseq_create( hws->masks = &hwseq_mask; hws->wa.DEGVIDCN10_253 = true; hws->wa.false_optc_underflow = true; + hws->wa.DEGVIDCN10_254 = true; } return hws; } @@ -762,6 +859,9 @@ static void destruct(struct dcn10_resource_pool *pool) kfree(DCN10TG_FROM_TG(pool->base.timing_generators[i])); pool->base.timing_generators[i] = NULL; } + + if (pool->base.engines[i] != NULL) + pool->base.engines[i]->funcs->destroy_engine(&pool->base.engines[i]); } for (i = 0; i < pool->base.stream_enc_count; i++) @@ -790,8 +890,8 @@ static void destruct(struct dcn10_resource_pool *pool) if (pool->base.dmcu != NULL) dce_dmcu_destroy(&pool->base.dmcu); - if (pool->base.display_clock != NULL) - dce_disp_clk_destroy(&pool->base.display_clock); + if (pool->base.dccg != NULL) + dce_dccg_destroy(&pool->base.dccg); kfree(pool->base.pp_smu); } @@ -971,11 +1071,11 @@ static enum dc_status dcn10_validate_plane(const struct dc_plane_state *plane_st return DC_OK; } -static struct dc_cap_funcs cap_funcs = { +static const struct dc_cap_funcs cap_funcs = { .get_dcc_compression_cap = dcn10_get_dcc_compression_cap }; -static struct resource_funcs dcn10_res_pool_funcs = { +static const struct resource_funcs dcn10_res_pool_funcs = { .destroy = dcn10_destroy_resource_pool, .link_enc_create = dcn10_link_encoder_create, .validate_bandwidth = dcn_validate_bandwidth, @@ -1072,8 +1172,8 @@ static bool construct( } } - pool->base.display_clock = dce120_disp_clk_create(ctx); - if (pool->base.display_clock == NULL) { + pool->base.dccg = dcn1_dccg_create(ctx); + if (pool->base.dccg == NULL) { dm_error("DC: failed to create display clock!\n"); BREAK_TO_DEBUGGER(); goto fail; @@ -1193,6 +1293,14 @@ static bool construct( goto fail; } + pool->base.engines[i] = dcn10_aux_engine_create(ctx, i); + if (pool->base.engines[i] == NULL) { + BREAK_TO_DEBUGGER(); + dm_error( + "DC:failed to create aux engine!!\n"); + goto fail; + } + /* check next valid pipe */ j++; } diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.c index c928ee4cd382..6f9078f3c4d3 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.c @@ -257,20 +257,18 @@ void enc1_stream_encoder_dp_set_stream_attribute( uint8_t colorimetry_bpc; uint8_t dynamic_range_rgb = 0; /*full range*/ uint8_t dynamic_range_ycbcr = 1; /*bt709*/ + uint8_t dp_pixel_encoding = 0; + uint8_t dp_component_depth = 0; struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc); - REG_UPDATE(DP_DB_CNTL, DP_DB_DISABLE, 1); - /* set pixel encoding */ switch (crtc_timing->pixel_encoding) { case PIXEL_ENCODING_YCBCR422: - REG_UPDATE(DP_PIXEL_FORMAT, DP_PIXEL_ENCODING, - DP_PIXEL_ENCODING_TYPE_YCBCR422); + dp_pixel_encoding = DP_PIXEL_ENCODING_TYPE_YCBCR422; break; case PIXEL_ENCODING_YCBCR444: - REG_UPDATE(DP_PIXEL_FORMAT, DP_PIXEL_ENCODING, - DP_PIXEL_ENCODING_TYPE_YCBCR444); + dp_pixel_encoding = DP_PIXEL_ENCODING_TYPE_YCBCR444; if (crtc_timing->flags.Y_ONLY) if (crtc_timing->display_color_depth != COLOR_DEPTH_666) @@ -278,8 +276,8 @@ void enc1_stream_encoder_dp_set_stream_attribute( * Color depth of Y-only could be * 8, 10, 12, 16 bits */ - REG_UPDATE(DP_PIXEL_FORMAT, DP_PIXEL_ENCODING, - DP_PIXEL_ENCODING_TYPE_Y_ONLY); + dp_pixel_encoding = DP_PIXEL_ENCODING_TYPE_Y_ONLY; + /* Note: DP_MSA_MISC1 bit 7 is the indicator * of Y-only mode. * This bit is set in HW if register @@ -287,48 +285,55 @@ void enc1_stream_encoder_dp_set_stream_attribute( */ break; case PIXEL_ENCODING_YCBCR420: - REG_UPDATE(DP_PIXEL_FORMAT, DP_PIXEL_ENCODING, - DP_PIXEL_ENCODING_TYPE_YCBCR420); + dp_pixel_encoding = DP_PIXEL_ENCODING_TYPE_YCBCR420; REG_UPDATE(DP_VID_TIMING, DP_VID_N_MUL, 1); break; default: - REG_UPDATE(DP_PIXEL_FORMAT, DP_PIXEL_ENCODING, - DP_PIXEL_ENCODING_TYPE_RGB444); + dp_pixel_encoding = DP_PIXEL_ENCODING_TYPE_RGB444; break; } misc1 = REG_READ(DP_MSA_MISC); + /* For YCbCr420 and BT2020 Colorimetry Formats, VSC SDP shall be used. + * When MISC1, bit 6, is Set to 1, a Source device uses a VSC SDP to indicate the + * Pixel Encoding/Colorimetry Format and that a Sink device shall ignore MISC1, bit 7, + * and MISC0, bits 7:1 (MISC1, bit 7, and MISC0, bits 7:1, become "don't care"). + */ + if ((crtc_timing->pixel_encoding == PIXEL_ENCODING_YCBCR420) || + (output_color_space == COLOR_SPACE_2020_YCBCR) || + (output_color_space == COLOR_SPACE_2020_RGB_FULLRANGE) || + (output_color_space == COLOR_SPACE_2020_RGB_LIMITEDRANGE)) + misc1 = misc1 | 0x40; + else + misc1 = misc1 & ~0x40; /* set color depth */ - switch (crtc_timing->display_color_depth) { case COLOR_DEPTH_666: - REG_UPDATE(DP_PIXEL_FORMAT, DP_COMPONENT_DEPTH, - 0); + dp_component_depth = DP_COMPONENT_PIXEL_DEPTH_6BPC; break; case COLOR_DEPTH_888: - REG_UPDATE(DP_PIXEL_FORMAT, DP_COMPONENT_DEPTH, - DP_COMPONENT_PIXEL_DEPTH_8BPC); + dp_component_depth = DP_COMPONENT_PIXEL_DEPTH_8BPC; break; case COLOR_DEPTH_101010: - REG_UPDATE(DP_PIXEL_FORMAT, DP_COMPONENT_DEPTH, - DP_COMPONENT_PIXEL_DEPTH_10BPC); - + dp_component_depth = DP_COMPONENT_PIXEL_DEPTH_10BPC; break; case COLOR_DEPTH_121212: - REG_UPDATE(DP_PIXEL_FORMAT, DP_COMPONENT_DEPTH, - DP_COMPONENT_PIXEL_DEPTH_12BPC); + dp_component_depth = DP_COMPONENT_PIXEL_DEPTH_12BPC; break; case COLOR_DEPTH_161616: - REG_UPDATE(DP_PIXEL_FORMAT, DP_COMPONENT_DEPTH, - DP_COMPONENT_PIXEL_DEPTH_16BPC); + dp_component_depth = DP_COMPONENT_PIXEL_DEPTH_16BPC; break; default: - REG_UPDATE(DP_PIXEL_FORMAT, DP_COMPONENT_DEPTH, - DP_COMPONENT_PIXEL_DEPTH_6BPC); + dp_component_depth = DP_COMPONENT_PIXEL_DEPTH_6BPC; break; } + /* Set DP pixel encoding and component depth */ + REG_UPDATE_2(DP_PIXEL_FORMAT, + DP_PIXEL_ENCODING, dp_pixel_encoding, + DP_COMPONENT_DEPTH, dp_component_depth); + /* set dynamic range and YCbCr range */ switch (crtc_timing->display_color_depth) { @@ -354,7 +359,6 @@ void enc1_stream_encoder_dp_set_stream_attribute( switch (output_color_space) { case COLOR_SPACE_SRGB: - misc0 = misc0 | 0x0; misc1 = misc1 & ~0x80; /* bit7 = 0*/ dynamic_range_rgb = 0; /*full range*/ break; @@ -1087,27 +1091,6 @@ static union audio_cea_channels speakers_to_channels( return cea_channels; } -static uint32_t calc_max_audio_packets_per_line( - const struct audio_crtc_info *crtc_info) -{ - uint32_t max_packets_per_line; - - max_packets_per_line = - crtc_info->h_total - crtc_info->h_active; - - if (crtc_info->pixel_repetition) - max_packets_per_line *= crtc_info->pixel_repetition; - - /* for other hdmi features */ - max_packets_per_line -= 58; - /* for Control Period */ - max_packets_per_line -= 16; - /* Number of Audio Packets per Line */ - max_packets_per_line /= 32; - - return max_packets_per_line; -} - static void get_audio_clock_info( enum dc_color_depth color_depth, uint32_t crtc_pixel_clock_in_khz, @@ -1201,16 +1184,9 @@ static void enc1_se_setup_hdmi_audio( struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc); struct audio_clock_info audio_clock_info = {0}; - uint32_t max_packets_per_line; - - /* For now still do calculation, although this field is ignored when - * above HDMI_PACKET_GEN_VERSION set to 1 - */ - max_packets_per_line = calc_max_audio_packets_per_line(crtc_info); /* HDMI_AUDIO_PACKET_CONTROL */ - REG_UPDATE_2(HDMI_AUDIO_PACKET_CONTROL, - HDMI_AUDIO_PACKETS_PER_LINE, max_packets_per_line, + REG_UPDATE(HDMI_AUDIO_PACKET_CONTROL, HDMI_AUDIO_DELAY_EN, 1); /* AFMT_AUDIO_PACKET_CONTROL */ diff --git a/drivers/gpu/drm/amd/display/dc/dm_helpers.h b/drivers/gpu/drm/amd/display/dc/dm_helpers.h index 034369fbb9e2..5d4527d03045 100644 --- a/drivers/gpu/drm/amd/display/dc/dm_helpers.h +++ b/drivers/gpu/drm/amd/display/dc/dm_helpers.h @@ -40,6 +40,14 @@ enum dc_edid_status dm_helpers_parse_edid_caps( const struct dc_edid *edid, struct dc_edid_caps *edid_caps); + +/* + * Update DP branch info + */ +void dm_helpers_dp_update_branch_info( + struct dc_context *ctx, + const struct dc_link *link); + /* * Writes payload allocation table in immediate downstream device. */ @@ -103,6 +111,9 @@ bool dm_helpers_submit_i2c( const struct dc_link *link, struct i2c_command *cmd); +bool dm_helpers_is_dp_sink_present( + struct dc_link *link); + enum dc_edid_status dm_helpers_read_local_edid( struct dc_context *ctx, struct dc_link *link, diff --git a/drivers/gpu/drm/amd/display/dc/dm_pp_smu.h b/drivers/gpu/drm/amd/display/dc/dm_pp_smu.h index eac4bfe12257..58ed2055ef9f 100644 --- a/drivers/gpu/drm/amd/display/dc/dm_pp_smu.h +++ b/drivers/gpu/drm/amd/display/dc/dm_pp_smu.h @@ -40,7 +40,7 @@ enum wm_set_id { WM_B, WM_C, WM_D, - WM_COUNT, + WM_SET_COUNT, }; struct pp_smu_wm_set_range { @@ -53,10 +53,10 @@ struct pp_smu_wm_set_range { struct pp_smu_wm_range_sets { uint32_t num_reader_wm_sets; - struct pp_smu_wm_set_range reader_wm_sets[WM_COUNT]; + struct pp_smu_wm_set_range reader_wm_sets[WM_SET_COUNT]; uint32_t num_writer_wm_sets; - struct pp_smu_wm_set_range writer_wm_sets[WM_COUNT]; + struct pp_smu_wm_set_range writer_wm_sets[WM_SET_COUNT]; }; struct pp_smu_display_requirement_rv { diff --git a/drivers/gpu/drm/amd/display/dc/dm_services.h b/drivers/gpu/drm/amd/display/dc/dm_services.h index 4ff9b2bba178..eb5ab3978e84 100644 --- a/drivers/gpu/drm/amd/display/dc/dm_services.h +++ b/drivers/gpu/drm/amd/display/dc/dm_services.h @@ -339,7 +339,10 @@ bool dm_dmcu_set_pipe(struct dc_context *ctx, unsigned int controller_id); #define dm_log_to_buffer(buffer, size, fmt, args)\ vsnprintf(buffer, size, fmt, args) -unsigned long long dm_get_timestamp(struct dc_context *ctx); +static inline unsigned long long dm_get_timestamp(struct dc_context *ctx) +{ + return ktime_get_raw_ns(); +} unsigned long long dm_get_elapse_time_in_ns(struct dc_context *ctx, unsigned long long current_time_stamp, diff --git a/drivers/gpu/drm/amd/display/dc/dm_services_types.h b/drivers/gpu/drm/amd/display/dc/dm_services_types.h index ab8c77d4e6df..2b83f922ac02 100644 --- a/drivers/gpu/drm/amd/display/dc/dm_services_types.h +++ b/drivers/gpu/drm/amd/display/dc/dm_services_types.h @@ -137,7 +137,7 @@ struct dm_pp_clock_range_for_wm_set { enum dm_pp_wm_set_id wm_set_id; uint32_t wm_min_eng_clk_in_khz; uint32_t wm_max_eng_clk_in_khz; - uint32_t wm_min_memg_clk_in_khz; + uint32_t wm_min_mem_clk_in_khz; uint32_t wm_max_mem_clk_in_khz; }; @@ -150,7 +150,7 @@ struct dm_pp_clock_range_for_dmif_wm_set_soc15 { enum dm_pp_wm_set_id wm_set_id; uint32_t wm_min_dcfclk_clk_in_khz; uint32_t wm_max_dcfclk_clk_in_khz; - uint32_t wm_min_memg_clk_in_khz; + uint32_t wm_min_mem_clk_in_khz; uint32_t wm_max_mem_clk_in_khz; }; @@ -158,7 +158,7 @@ struct dm_pp_clock_range_for_mcif_wm_set_soc15 { enum dm_pp_wm_set_id wm_set_id; uint32_t wm_min_socclk_clk_in_khz; uint32_t wm_max_socclk_clk_in_khz; - uint32_t wm_min_memg_clk_in_khz; + uint32_t wm_min_mem_clk_in_khz; uint32_t wm_max_mem_clk_in_khz; }; diff --git a/drivers/gpu/drm/amd/display/dc/dml/Makefile b/drivers/gpu/drm/amd/display/dc/dml/Makefile index f83a608f93e9..d97ca6528f9d 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/Makefile +++ b/drivers/gpu/drm/amd/display/dc/dml/Makefile @@ -36,11 +36,10 @@ CFLAGS_display_mode_lib.o := $(dml_ccflags) CFLAGS_display_pipe_clocks.o := $(dml_ccflags) CFLAGS_dml1_display_rq_dlg_calc.o := $(dml_ccflags) CFLAGS_display_rq_dlg_helpers.o := $(dml_ccflags) -CFLAGS_soc_bounding_box.o := $(dml_ccflags) CFLAGS_dml_common_defs.o := $(dml_ccflags) DML = display_mode_lib.o display_rq_dlg_helpers.o dml1_display_rq_dlg_calc.o \ - soc_bounding_box.o dml_common_defs.o + dml_common_defs.o AMD_DAL_DML = $(addprefix $(AMDDALPATH)/dc/dml/,$(DML)) diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.c b/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.c index fd9d97aab071..dddeb0d4db8f 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.c +++ b/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.c @@ -26,67 +26,8 @@ #include "display_mode_lib.h" #include "dc_features.h" -static const struct _vcs_dpi_ip_params_st dcn1_0_ip = { - .rob_buffer_size_kbytes = 64, - .det_buffer_size_kbytes = 164, - .dpte_buffer_size_in_pte_reqs = 42, - .dpp_output_buffer_pixels = 2560, - .opp_output_buffer_lines = 1, - .pixel_chunk_size_kbytes = 8, - .pte_enable = 1, - .pte_chunk_size_kbytes = 2, - .meta_chunk_size_kbytes = 2, - .writeback_chunk_size_kbytes = 2, - .line_buffer_size_bits = 589824, - .max_line_buffer_lines = 12, - .IsLineBufferBppFixed = 0, - .LineBufferFixedBpp = -1, - .writeback_luma_buffer_size_kbytes = 12, - .writeback_chroma_buffer_size_kbytes = 8, - .max_num_dpp = 4, - .max_num_wb = 2, - .max_dchub_pscl_bw_pix_per_clk = 4, - .max_pscl_lb_bw_pix_per_clk = 2, - .max_lb_vscl_bw_pix_per_clk = 4, - .max_vscl_hscl_bw_pix_per_clk = 4, - .max_hscl_ratio = 4, - .max_vscl_ratio = 4, - .hscl_mults = 4, - .vscl_mults = 4, - .max_hscl_taps = 8, - .max_vscl_taps = 8, - .dispclk_ramp_margin_percent = 1, - .underscan_factor = 1.10, - .min_vblank_lines = 14, - .dppclk_delay_subtotal = 90, - .dispclk_delay_subtotal = 42, - .dcfclk_cstate_latency = 10, - .max_inter_dcn_tile_repeaters = 8, - .can_vstartup_lines_exceed_vsync_plus_back_porch_lines_minus_one = 0, - .bug_forcing_LC_req_same_size_fixed = 0, -}; - -static const struct _vcs_dpi_soc_bounding_box_st dcn1_0_soc = { - .sr_exit_time_us = 9.0, - .sr_enter_plus_exit_time_us = 11.0, - .urgent_latency_us = 4.0, - .writeback_latency_us = 12.0, - .ideal_dram_bw_after_urgent_percent = 80.0, - .max_request_size_bytes = 256, - .downspread_percent = 0.5, - .dram_page_open_time_ns = 50.0, - .dram_rw_turnaround_time_ns = 17.5, - .dram_return_buffer_per_channel_bytes = 8192, - .round_trip_ping_latency_dcfclk_cycles = 128, - .urgent_out_of_order_return_per_channel_bytes = 256, - .channel_interleave_bytes = 256, - .num_banks = 8, - .num_chans = 2, - .vmm_page_size_bytes = 4096, - .dram_clock_change_latency_us = 17.0, - .writeback_dram_clock_change_latency_us = 23.0, - .return_bus_width_bytes = 64, -}; +extern const struct _vcs_dpi_ip_params_st dcn1_0_ip; +extern const struct _vcs_dpi_soc_bounding_box_st dcn1_0_soc; static void set_soc_bounding_box(struct _vcs_dpi_soc_bounding_box_st *soc, enum dml_project project) { diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.h b/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.h index 3c2abcb8a1b0..635206248889 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.h +++ b/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.h @@ -27,7 +27,6 @@ #include "dml_common_defs.h" -#include "soc_bounding_box.h" #include "dml1_display_rq_dlg_calc.h" enum dml_project { diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h b/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h index 7fa0375939ae..cbafce649e33 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h +++ b/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h @@ -64,10 +64,9 @@ struct _vcs_dpi_voltage_scaling_st { double dscclk_mhz; double dcfclk_mhz; double socclk_mhz; - double dram_speed_mhz; + double dram_speed_mts; double fabricclk_mhz; double dispclk_mhz; - double dram_bw_per_chan_gbps; double phyclk_mhz; double dppclk_mhz; }; @@ -112,6 +111,8 @@ struct _vcs_dpi_soc_bounding_box_st { double xfc_bus_transport_time_us; double xfc_xbuf_latency_tolerance_us; int use_urgent_burst_bw; + double max_hscl_ratio; + double max_vscl_ratio; struct _vcs_dpi_voltage_scaling_st clock_limits[7]; }; @@ -304,6 +305,7 @@ struct _vcs_dpi_display_pipe_dest_params_st { unsigned char otg_inst; unsigned char odm_split_cnt; unsigned char odm_combine; + unsigned char use_maximum_vstartup; }; struct _vcs_dpi_display_pipe_params_st { diff --git a/drivers/gpu/drm/amd/display/dc/dml/soc_bounding_box.c b/drivers/gpu/drm/amd/display/dc/dml/soc_bounding_box.c deleted file mode 100644 index 324239c77958..000000000000 --- a/drivers/gpu/drm/amd/display/dc/dml/soc_bounding_box.c +++ /dev/null @@ -1,79 +0,0 @@ -/* - * 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. - * - * Authors: AMD - * - */ -#include "soc_bounding_box.h" -#include "display_mode_lib.h" -#include "dc_features.h" - -#include "dml_inline_defs.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. - */ - -void dml_socbb_set_latencies(soc_bounding_box_st *to_box, soc_bounding_box_st *from_box) -{ - to_box->dram_clock_change_latency_us = from_box->dram_clock_change_latency_us; - to_box->sr_exit_time_us = from_box->sr_exit_time_us; - to_box->sr_enter_plus_exit_time_us = from_box->sr_enter_plus_exit_time_us; - to_box->urgent_latency_us = from_box->urgent_latency_us; - to_box->writeback_latency_us = from_box->writeback_latency_us; -} - -voltage_scaling_st dml_socbb_voltage_scaling( - const soc_bounding_box_st *soc, - enum voltage_state voltage) -{ - const voltage_scaling_st *voltage_state; - const voltage_scaling_st * const voltage_end = soc->clock_limits + DC__VOLTAGE_STATES; - - for (voltage_state = soc->clock_limits; - voltage_state < voltage_end && voltage_state->state != voltage; - voltage_state++) { - } - - if (voltage_state < voltage_end) - return *voltage_state; - return soc->clock_limits[DC__VOLTAGE_STATES - 1]; -} - -double dml_socbb_return_bw_mhz(soc_bounding_box_st *box, enum voltage_state voltage) -{ - double return_bw; - - voltage_scaling_st state = dml_socbb_voltage_scaling(box, voltage); - - return_bw = dml_min((double) box->return_bus_width_bytes * state.dcfclk_mhz, - state.dram_bw_per_chan_gbps * 1000.0 * (double) box->num_chans - * box->ideal_dram_bw_after_urgent_percent / 100.0); - - return_bw = dml_min((double) box->return_bus_width_bytes * state.fabricclk_mhz, return_bw); - - return return_bw; -} diff --git a/drivers/gpu/drm/amd/display/dc/gpio/Makefile b/drivers/gpu/drm/amd/display/dc/gpio/Makefile index 562ee189d780..b9d9930a4974 100644 --- a/drivers/gpu/drm/amd/display/dc/gpio/Makefile +++ b/drivers/gpu/drm/amd/display/dc/gpio/Makefile @@ -61,7 +61,7 @@ AMD_DISPLAY_FILES += $(AMD_DAL_GPIO_DCE120) ############################################################################### # DCN 1x ############################################################################### -ifdef CONFIG_DRM_AMD_DC_DCN1_0 +ifdef CONFIG_X86 GPIO_DCN10 = hw_translate_dcn10.o hw_factory_dcn10.o AMD_DAL_GPIO_DCN10 = $(addprefix $(AMDDALPATH)/dc/gpio/dcn10/,$(GPIO_DCN10)) diff --git a/drivers/gpu/drm/amd/display/dc/gpio/ddc_regs.h b/drivers/gpu/drm/amd/display/dc/gpio/ddc_regs.h index 9c4a56c738c0..bf40725f982f 100644 --- a/drivers/gpu/drm/amd/display/dc/gpio/ddc_regs.h +++ b/drivers/gpu/drm/amd/display/dc/gpio/ddc_regs.h @@ -82,13 +82,16 @@ DDC_GPIO_I2C_REG_LIST(cd),\ .ddc_setup = 0 -#define DDC_MASK_SH_LIST(mask_sh) \ +#define DDC_MASK_SH_LIST_COMMON(mask_sh) \ SF_DDC(DC_I2C_DDC1_SETUP, DC_I2C_DDC1_ENABLE, mask_sh),\ SF_DDC(DC_I2C_DDC1_SETUP, DC_I2C_DDC1_EDID_DETECT_ENABLE, mask_sh),\ SF_DDC(DC_I2C_DDC1_SETUP, DC_I2C_DDC1_EDID_DETECT_MODE, mask_sh),\ SF_DDC(DC_GPIO_DDC1_MASK, DC_GPIO_DDC1DATA_PD_EN, mask_sh),\ SF_DDC(DC_GPIO_DDC1_MASK, DC_GPIO_DDC1CLK_PD_EN, mask_sh),\ - SF_DDC(DC_GPIO_DDC1_MASK, AUX_PAD1_MODE, mask_sh),\ + SF_DDC(DC_GPIO_DDC1_MASK, AUX_PAD1_MODE, mask_sh) + +#define DDC_MASK_SH_LIST(mask_sh) \ + DDC_MASK_SH_LIST_COMMON(mask_sh),\ SF_DDC(DC_GPIO_I2CPAD_MASK, DC_GPIO_SDA_PD_DIS, mask_sh),\ SF_DDC(DC_GPIO_I2CPAD_MASK, DC_GPIO_SCL_PD_DIS, mask_sh) diff --git a/drivers/gpu/drm/amd/display/dc/gpio/gpio_service.c b/drivers/gpu/drm/amd/display/dc/gpio/gpio_service.c index ab5483c0c502..f20161c5706d 100644 --- a/drivers/gpu/drm/amd/display/dc/gpio/gpio_service.c +++ b/drivers/gpu/drm/amd/display/dc/gpio/gpio_service.c @@ -375,6 +375,7 @@ struct gpio *dal_gpio_create_irq( case GPIO_ID_GPIO_PAD: break; default: + id = GPIO_ID_HPD; ASSERT_CRITICAL(false); return NULL; } diff --git a/drivers/gpu/drm/amd/display/dc/gpio/hw_factory.c b/drivers/gpu/drm/amd/display/dc/gpio/hw_factory.c index 0caee3523017..83df779984e5 100644 --- a/drivers/gpu/drm/amd/display/dc/gpio/hw_factory.c +++ b/drivers/gpu/drm/amd/display/dc/gpio/hw_factory.c @@ -43,7 +43,7 @@ #include "dce80/hw_factory_dce80.h" #include "dce110/hw_factory_dce110.h" #include "dce120/hw_factory_dce120.h" -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 #include "dcn10/hw_factory_dcn10.h" #endif @@ -81,7 +81,7 @@ bool dal_hw_factory_init( case DCE_VERSION_12_0: dal_hw_factory_dce120_init(factory); return true; -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 case DCN_VERSION_1_0: dal_hw_factory_dcn10_init(factory); return true; diff --git a/drivers/gpu/drm/amd/display/dc/gpio/hw_translate.c b/drivers/gpu/drm/amd/display/dc/gpio/hw_translate.c index 55c707488541..e7541310480b 100644 --- a/drivers/gpu/drm/amd/display/dc/gpio/hw_translate.c +++ b/drivers/gpu/drm/amd/display/dc/gpio/hw_translate.c @@ -43,7 +43,7 @@ #include "dce80/hw_translate_dce80.h" #include "dce110/hw_translate_dce110.h" #include "dce120/hw_translate_dce120.h" -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 #include "dcn10/hw_translate_dcn10.h" #endif @@ -78,7 +78,7 @@ bool dal_hw_translate_init( case DCE_VERSION_12_0: dal_hw_translate_dce120_init(translate); return true; -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 case DCN_VERSION_1_0: dal_hw_translate_dcn10_init(translate); return true; diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/Makefile b/drivers/gpu/drm/amd/display/dc/i2caux/Makefile index 352885cb4d07..a851d07f0190 100644 --- a/drivers/gpu/drm/amd/display/dc/i2caux/Makefile +++ b/drivers/gpu/drm/amd/display/dc/i2caux/Makefile @@ -71,7 +71,7 @@ AMD_DISPLAY_FILES += $(AMD_DAL_I2CAUX_DCE112) ############################################################################### # DCN 1.0 family ############################################################################### -ifdef CONFIG_DRM_AMD_DC_DCN1_0 +ifdef CONFIG_X86 I2CAUX_DCN1 = i2caux_dcn10.o AMD_DAL_I2CAUX_DCN1 = $(addprefix $(AMDDALPATH)/dc/i2caux/dcn10/,$(I2CAUX_DCN1)) diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/aux_engine.c b/drivers/gpu/drm/amd/display/dc/i2caux/aux_engine.c index bb526ad326e5..0afd2fa57bbe 100644 --- a/drivers/gpu/drm/amd/display/dc/i2caux/aux_engine.c +++ b/drivers/gpu/drm/amd/display/dc/i2caux/aux_engine.c @@ -128,8 +128,20 @@ static void process_read_reply( ctx->status = I2CAUX_TRANSACTION_STATUS_FAILED_PROTOCOL_ERROR; ctx->operation_succeeded = false; + } else if (ctx->returned_byte < ctx->current_read_length) { + ctx->current_read_length -= ctx->returned_byte; + + ctx->offset += ctx->returned_byte; + + ++ctx->invalid_reply_retry_aux_on_ack; + + if (ctx->invalid_reply_retry_aux_on_ack > + AUX_INVALID_REPLY_RETRY_COUNTER) { + ctx->status = + I2CAUX_TRANSACTION_STATUS_FAILED_PROTOCOL_ERROR; + ctx->operation_succeeded = false; + } } else { - ctx->current_read_length = ctx->returned_byte; ctx->status = I2CAUX_TRANSACTION_STATUS_SUCCEEDED; ctx->transaction_complete = true; ctx->operation_succeeded = true; @@ -157,6 +169,10 @@ static void process_read_reply( ctx->operation_succeeded = false; } break; + case AUX_TRANSACTION_REPLY_HPD_DISCON: + ctx->status = I2CAUX_TRANSACTION_STATUS_FAILED_HPD_DISCON; + ctx->operation_succeeded = false; + break; default: ctx->status = I2CAUX_TRANSACTION_STATUS_UNKNOWN; ctx->operation_succeeded = false; @@ -215,6 +231,10 @@ static void process_read_request( * so we should not wait here */ } break; + case AUX_CHANNEL_OPERATION_FAILED_HPD_DISCON: + ctx->status = I2CAUX_TRANSACTION_STATUS_FAILED_HPD_DISCON; + ctx->operation_succeeded = false; + break; default: ctx->status = I2CAUX_TRANSACTION_STATUS_UNKNOWN; ctx->operation_succeeded = false; @@ -282,7 +302,6 @@ static bool read_command( ctx.operation_succeeded); } - request->payload.length = ctx.reply.length; return ctx.operation_succeeded; } @@ -370,6 +389,10 @@ static void process_write_reply( ctx->operation_succeeded = false; } break; + case AUX_TRANSACTION_REPLY_HPD_DISCON: + ctx->status = I2CAUX_TRANSACTION_STATUS_FAILED_HPD_DISCON; + ctx->operation_succeeded = false; + break; default: ctx->status = I2CAUX_TRANSACTION_STATUS_UNKNOWN; ctx->operation_succeeded = false; @@ -422,6 +445,10 @@ static void process_write_request( * so we should not wait here */ } break; + case AUX_CHANNEL_OPERATION_FAILED_HPD_DISCON: + ctx->status = I2CAUX_TRANSACTION_STATUS_FAILED_HPD_DISCON; + ctx->operation_succeeded = false; + break; default: ctx->status = I2CAUX_TRANSACTION_STATUS_UNKNOWN; ctx->operation_succeeded = false; diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/aux_engine.h b/drivers/gpu/drm/amd/display/dc/i2caux/aux_engine.h index 8e71324ccb10..c33a2898d967 100644 --- a/drivers/gpu/drm/amd/display/dc/i2caux/aux_engine.h +++ b/drivers/gpu/drm/amd/display/dc/i2caux/aux_engine.h @@ -26,46 +26,7 @@ #ifndef __DAL_AUX_ENGINE_H__ #define __DAL_AUX_ENGINE_H__ -enum aux_transaction_type { - AUX_TRANSACTION_TYPE_DP, - AUX_TRANSACTION_TYPE_I2C -}; - -struct aux_request_transaction_data { - enum aux_transaction_type type; - enum i2caux_transaction_action action; - /* 20-bit AUX channel transaction address */ - uint32_t address; - /* delay, in 100-microsecond units */ - uint8_t delay; - uint32_t length; - uint8_t *data; -}; - -enum aux_transaction_reply { - AUX_TRANSACTION_REPLY_AUX_ACK = 0x00, - AUX_TRANSACTION_REPLY_AUX_NACK = 0x01, - AUX_TRANSACTION_REPLY_AUX_DEFER = 0x02, - - AUX_TRANSACTION_REPLY_I2C_ACK = 0x00, - AUX_TRANSACTION_REPLY_I2C_NACK = 0x10, - AUX_TRANSACTION_REPLY_I2C_DEFER = 0x20, - - AUX_TRANSACTION_REPLY_INVALID = 0xFF -}; - -struct aux_reply_transaction_data { - enum aux_transaction_reply status; - uint32_t length; - uint8_t *data; -}; - -enum aux_channel_operation_result { - AUX_CHANNEL_OPERATION_SUCCEEDED, - AUX_CHANNEL_OPERATION_FAILED_REASON_UNKNOWN, - AUX_CHANNEL_OPERATION_FAILED_INVALID_REPLY, - AUX_CHANNEL_OPERATION_FAILED_TIMEOUT -}; +#include "dc_ddc_types.h" struct aux_engine; @@ -83,6 +44,12 @@ struct aux_engine_funcs { void (*process_channel_reply)( struct aux_engine *engine, struct aux_reply_transaction_data *reply); + int (*read_channel_reply)( + struct aux_engine *engine, + uint32_t size, + uint8_t *buffer, + uint8_t *reply_result, + uint32_t *sw_status); enum aux_channel_operation_result (*get_channel_status)( struct aux_engine *engine, uint8_t *returned_bytes); diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/dce100/i2caux_dce100.c b/drivers/gpu/drm/amd/display/dc/i2caux/dce100/i2caux_dce100.c index e8d3781deaed..8b704ab0471c 100644 --- a/drivers/gpu/drm/amd/display/dc/i2caux/dce100/i2caux_dce100.c +++ b/drivers/gpu/drm/amd/display/dc/i2caux/dce100/i2caux_dce100.c @@ -97,6 +97,7 @@ struct i2caux *dal_i2caux_dce100_create( dal_i2caux_dce110_construct(i2caux_dce110, ctx, + ARRAY_SIZE(dce100_aux_regs), dce100_aux_regs, dce100_hw_engine_regs, &i2c_shift, diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/dce110/aux_engine_dce110.c b/drivers/gpu/drm/amd/display/dc/i2caux/dce110/aux_engine_dce110.c index 5f47f6c007ac..ae5caa97caca 100644 --- a/drivers/gpu/drm/amd/display/dc/i2caux/dce110/aux_engine_dce110.c +++ b/drivers/gpu/drm/amd/display/dc/i2caux/dce110/aux_engine_dce110.c @@ -198,27 +198,27 @@ static void submit_channel_request( ((request->type == AUX_TRANSACTION_TYPE_I2C) && ((request->action == I2CAUX_TRANSACTION_ACTION_I2C_WRITE) || (request->action == I2CAUX_TRANSACTION_ACTION_I2C_WRITE_MOT))); + if (REG(AUXN_IMPCAL)) { + /* clear_aux_error */ + REG_UPDATE_SEQ(AUXN_IMPCAL, AUXN_CALOUT_ERROR_AK, + 1, + 0); - /* clear_aux_error */ - REG_UPDATE_SEQ(AUXN_IMPCAL, AUXN_CALOUT_ERROR_AK, - 1, - 0); - - REG_UPDATE_SEQ(AUXP_IMPCAL, AUXP_CALOUT_ERROR_AK, - 1, - 0); - - /* force_default_calibrate */ - REG_UPDATE_1BY1_2(AUXN_IMPCAL, - AUXN_IMPCAL_ENABLE, 1, - AUXN_IMPCAL_OVERRIDE_ENABLE, 0); + REG_UPDATE_SEQ(AUXP_IMPCAL, AUXP_CALOUT_ERROR_AK, + 1, + 0); - /* bug? why AUXN update EN and OVERRIDE_EN 1 by 1 while AUX P toggles OVERRIDE? */ + /* force_default_calibrate */ + REG_UPDATE_1BY1_2(AUXN_IMPCAL, + AUXN_IMPCAL_ENABLE, 1, + AUXN_IMPCAL_OVERRIDE_ENABLE, 0); - REG_UPDATE_SEQ(AUXP_IMPCAL, AUXP_IMPCAL_OVERRIDE_ENABLE, - 1, - 0); + /* bug? why AUXN update EN and OVERRIDE_EN 1 by 1 while AUX P toggles OVERRIDE? */ + REG_UPDATE_SEQ(AUXP_IMPCAL, AUXP_IMPCAL_OVERRIDE_ENABLE, + 1, + 0); + } /* set the delay and the number of bytes to write */ /* The length include @@ -275,55 +275,92 @@ static void submit_channel_request( REG_UPDATE(AUX_SW_CONTROL, AUX_SW_GO, 1); } -static void process_channel_reply( - struct aux_engine *engine, - struct aux_reply_transaction_data *reply) +static int read_channel_reply(struct aux_engine *engine, uint32_t size, + uint8_t *buffer, uint8_t *reply_result, + uint32_t *sw_status) { struct aux_engine_dce110 *aux110 = FROM_AUX_ENGINE(engine); + uint32_t bytes_replied; + uint32_t reply_result_32; - /* Need to do a read to get the number of bytes to process - * Alternatively, this information can be passed - - * but that causes coupling which isn't good either. */ + *sw_status = REG_GET(AUX_SW_STATUS, AUX_SW_REPLY_BYTE_COUNT, + &bytes_replied); - uint32_t bytes_replied; - uint32_t value; + /* In case HPD is LOW, exit AUX transaction */ + if ((*sw_status & AUX_SW_STATUS__AUX_SW_HPD_DISCON_MASK)) + return -1; - value = REG_GET(AUX_SW_STATUS, - AUX_SW_REPLY_BYTE_COUNT, &bytes_replied); + /* Need at least the status byte */ + if (!bytes_replied) + return -1; - if (bytes_replied) { - uint32_t reply_result; + REG_UPDATE_1BY1_3(AUX_SW_DATA, + AUX_SW_INDEX, 0, + AUX_SW_AUTOINCREMENT_DISABLE, 1, + AUX_SW_DATA_RW, 1); - REG_UPDATE_1BY1_3(AUX_SW_DATA, - AUX_SW_INDEX, 0, - AUX_SW_AUTOINCREMENT_DISABLE, 1, - AUX_SW_DATA_RW, 1); + REG_GET(AUX_SW_DATA, AUX_SW_DATA, &reply_result_32); + reply_result_32 = reply_result_32 >> 4; + *reply_result = (uint8_t)reply_result_32; - REG_GET(AUX_SW_DATA, - AUX_SW_DATA, &reply_result); + if (reply_result_32 == 0) { /* ACK */ + uint32_t i = 0; - reply_result = reply_result >> 4; + /* First byte was already used to get the command status */ + --bytes_replied; - switch (reply_result) { - case 0: /* ACK */ { - uint32_t i = 0; + /* Do not overflow buffer */ + if (bytes_replied > size) + return -1; - /* first byte was already used - * to get the command status */ - --bytes_replied; + while (i < bytes_replied) { + uint32_t aux_sw_data_val; - while (i < bytes_replied) { - uint32_t aux_sw_data_val; + REG_GET(AUX_SW_DATA, AUX_SW_DATA, &aux_sw_data_val); + buffer[i] = aux_sw_data_val; + ++i; + } - REG_GET(AUX_SW_DATA, - AUX_SW_DATA, &aux_sw_data_val); + return i; + } - reply->data[i] = aux_sw_data_val; - ++i; - } + return 0; +} - reply->status = AUX_TRANSACTION_REPLY_AUX_ACK; +static void process_channel_reply( + struct aux_engine *engine, + struct aux_reply_transaction_data *reply) +{ + int bytes_replied; + uint8_t reply_result; + uint32_t sw_status; + + bytes_replied = read_channel_reply(engine, reply->length, reply->data, + &reply_result, &sw_status); + + /* in case HPD is LOW, exit AUX transaction */ + if ((sw_status & AUX_SW_STATUS__AUX_SW_HPD_DISCON_MASK)) { + reply->status = AUX_CHANNEL_OPERATION_FAILED_HPD_DISCON; + return; + } + + if (bytes_replied < 0) { + /* Need to handle an error case... + * Hopefully, upper layer function won't call this function if + * the number of bytes in the reply was 0, because there was + * surely an error that was asserted that should have been + * handled for hot plug case, this could happens + */ + if (!(sw_status & AUX_SW_STATUS__AUX_SW_HPD_DISCON_MASK)) { + reply->status = AUX_TRANSACTION_REPLY_INVALID; + ASSERT_CRITICAL(false); + return; } + } else { + + switch (reply_result) { + case 0: /* ACK */ + reply->status = AUX_TRANSACTION_REPLY_AUX_ACK; break; case 1: /* NACK */ reply->status = AUX_TRANSACTION_REPLY_AUX_NACK; @@ -340,15 +377,6 @@ static void process_channel_reply( default: reply->status = AUX_TRANSACTION_REPLY_INVALID; } - } else { - /* Need to handle an error case... - * hopefully, upper layer function won't call this function - * if the number of bytes in the reply was 0 - * because there was surely an error that was asserted - * that should have been handled - * for hot plug case, this could happens*/ - if (!(value & AUX_SW_STATUS__AUX_SW_HPD_DISCON_MASK)) - ASSERT_CRITICAL(false); } } @@ -371,6 +399,10 @@ static enum aux_channel_operation_result get_channel_status( value = REG_WAIT(AUX_SW_STATUS, AUX_SW_DONE, 1, 10, aux110->timeout_period/10); + /* in case HPD is LOW, exit AUX transaction */ + if ((value & AUX_SW_STATUS__AUX_SW_HPD_DISCON_MASK)) + return AUX_CHANNEL_OPERATION_FAILED_HPD_DISCON; + /* Note that the following bits are set in 'status.bits' * during CTS 4.2.1.2 (FW 3.3.1): * AUX_SW_RX_MIN_COUNT_VIOL, AUX_SW_RX_INVALID_STOP, @@ -402,10 +434,10 @@ static enum aux_channel_operation_result get_channel_status( return AUX_CHANNEL_OPERATION_SUCCEEDED; } } else { - /*time_elapsed >= aux_engine->timeout_period */ - if (!(value & AUX_SW_STATUS__AUX_SW_HPD_DISCON_MASK)) - ASSERT_CRITICAL(false); - + /*time_elapsed >= aux_engine->timeout_period + * AUX_SW_STATUS__AUX_SW_HPD_DISCON = at this point + */ + ASSERT_CRITICAL(false); return AUX_CHANNEL_OPERATION_FAILED_TIMEOUT; } } @@ -415,6 +447,7 @@ static const struct aux_engine_funcs aux_engine_funcs = { .acquire_engine = acquire_engine, .submit_channel_request = submit_channel_request, .process_channel_reply = process_channel_reply, + .read_channel_reply = read_channel_reply, .get_channel_status = get_channel_status, .is_engine_available = is_engine_available, }; diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2c_hw_engine_dce110.c b/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2c_hw_engine_dce110.c index b7256f595052..9cbe1a7a6bcb 100644 --- a/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2c_hw_engine_dce110.c +++ b/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2c_hw_engine_dce110.c @@ -62,12 +62,7 @@ enum dc_i2c_arbitration { DC_I2C_ARBITRATION__DC_I2C_SW_PRIORITY_HIGH }; -enum { - /* No timeout in HW - * (timeout implemented in SW by querying status) */ - I2C_SETUP_TIME_LIMIT = 255, - I2C_HW_BUFFER_SIZE = 538 -}; + /* * @brief @@ -152,6 +147,11 @@ static bool setup_engine( struct i2c_engine *i2c_engine) { struct i2c_hw_engine_dce110 *hw_engine = FROM_I2C_ENGINE(i2c_engine); + uint32_t i2c_setup_limit = I2C_SETUP_TIME_LIMIT_DCE; + uint32_t reset_length = 0; + + if (hw_engine->base.base.setup_limit != 0) + i2c_setup_limit = hw_engine->base.base.setup_limit; /* Program pin select */ REG_UPDATE_6( @@ -164,11 +164,15 @@ static bool setup_engine( DC_I2C_DDC_SELECT, hw_engine->engine_id); /* Program time limit */ - REG_UPDATE_N( - SETUP, 2, - FN(DC_I2C_DDC1_SETUP, DC_I2C_DDC1_TIME_LIMIT), I2C_SETUP_TIME_LIMIT, - FN(DC_I2C_DDC1_SETUP, DC_I2C_DDC1_ENABLE), 1); - + if (hw_engine->base.base.send_reset_length == 0) { + /*pre-dcn*/ + REG_UPDATE_N( + SETUP, 2, + FN(DC_I2C_DDC1_SETUP, DC_I2C_DDC1_TIME_LIMIT), i2c_setup_limit, + FN(DC_I2C_DDC1_SETUP, DC_I2C_DDC1_ENABLE), 1); + } else { + reset_length = hw_engine->base.base.send_reset_length; + } /* Program HW priority * set to High - interrupt software I2C at any time * Enable restart of SW I2C that was interrupted by HW diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2c_hw_engine_dce110.h b/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2c_hw_engine_dce110.h index 5bb04085f670..fea2946906ed 100644 --- a/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2c_hw_engine_dce110.h +++ b/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2c_hw_engine_dce110.h @@ -192,6 +192,7 @@ struct i2c_hw_engine_dce110 { /* number of pending transactions (before GO) */ uint32_t transaction_count; uint32_t engine_keep_power_up_count; + uint32_t i2_setup_time_limit; }; struct i2c_hw_engine_dce110_create_arg { @@ -207,4 +208,11 @@ struct i2c_hw_engine_dce110_create_arg { struct i2c_engine *dal_i2c_hw_engine_dce110_create( const struct i2c_hw_engine_dce110_create_arg *arg); +enum { + I2C_SETUP_TIME_LIMIT_DCE = 255, + I2C_SETUP_TIME_LIMIT_DCN = 3, + I2C_HW_BUFFER_SIZE = 538, + I2C_SEND_RESET_LENGTH_9 = 9, + I2C_SEND_RESET_LENGTH_10 = 10, +}; #endif diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2caux_dce110.c b/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2caux_dce110.c index 2a047f8ca0e9..1d748ac1d6d6 100644 --- a/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2caux_dce110.c +++ b/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2caux_dce110.c @@ -43,6 +43,9 @@ #include "i2c_sw_engine_dce110.h" #include "i2c_hw_engine_dce110.h" #include "aux_engine_dce110.h" +#include "../../dc.h" +#include "dc_types.h" + /* * Post-requisites: headers required by this unit @@ -199,6 +202,7 @@ static const struct dce110_i2c_hw_engine_mask i2c_mask = { void dal_i2caux_dce110_construct( struct i2caux_dce110 *i2caux_dce110, struct dc_context *ctx, + unsigned int num_i2caux_inst, const struct dce110_aux_registers aux_regs[], const struct dce110_i2c_hw_engine_registers i2c_hw_engine_regs[], const struct dce110_i2c_hw_engine_shift *i2c_shift, @@ -249,9 +253,22 @@ void dal_i2caux_dce110_construct( base->i2c_hw_engines[line_id] = dal_i2c_hw_engine_dce110_create(&hw_arg_dce110); - + if (base->i2c_hw_engines[line_id] != NULL) { + switch (ctx->dce_version) { + case DCN_VERSION_1_0: + base->i2c_hw_engines[line_id]->setup_limit = + I2C_SETUP_TIME_LIMIT_DCN; + base->i2c_hw_engines[line_id]->send_reset_length = 0; + break; + default: + base->i2c_hw_engines[line_id]->setup_limit = + I2C_SETUP_TIME_LIMIT_DCE; + base->i2c_hw_engines[line_id]->send_reset_length = 0; + break; + } + } ++i; - } while (i < ARRAY_SIZE(hw_ddc_lines)); + } while (i < num_i2caux_inst); /* Create AUX engines for all lines which has assisted HW AUX * 'i' (loop counter) used as DDC/AUX engine_id */ @@ -272,7 +289,7 @@ void dal_i2caux_dce110_construct( dal_aux_engine_dce110_create(&aux_init_data); ++i; - } while (i < ARRAY_SIZE(hw_aux_lines)); + } while (i < num_i2caux_inst); /*TODO Generic I2C SW and HW*/ } @@ -303,6 +320,7 @@ struct i2caux *dal_i2caux_dce110_create( dal_i2caux_dce110_construct(i2caux_dce110, ctx, + ARRAY_SIZE(dce110_aux_regs), dce110_aux_regs, i2c_hw_engine_regs, &i2c_shift, diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2caux_dce110.h b/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2caux_dce110.h index 1b1f71c60ac9..d3d8cc58666a 100644 --- a/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2caux_dce110.h +++ b/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2caux_dce110.h @@ -45,6 +45,7 @@ struct i2caux *dal_i2caux_dce110_create( void dal_i2caux_dce110_construct( struct i2caux_dce110 *i2caux_dce110, struct dc_context *ctx, + unsigned int num_i2caux_inst, const struct dce110_aux_registers *aux_regs, const struct dce110_i2c_hw_engine_registers *i2c_hw_engine_regs, const struct dce110_i2c_hw_engine_shift *i2c_shift, diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/dce112/i2caux_dce112.c b/drivers/gpu/drm/amd/display/dc/i2caux/dce112/i2caux_dce112.c index dafc1a727f7f..a9db04738724 100644 --- a/drivers/gpu/drm/amd/display/dc/i2caux/dce112/i2caux_dce112.c +++ b/drivers/gpu/drm/amd/display/dc/i2caux/dce112/i2caux_dce112.c @@ -93,6 +93,7 @@ static void construct( { dal_i2caux_dce110_construct(i2caux_dce110, ctx, + ARRAY_SIZE(dce112_aux_regs), dce112_aux_regs, dce112_hw_engine_regs, &i2c_shift, diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/dce120/i2caux_dce120.c b/drivers/gpu/drm/amd/display/dc/i2caux/dce120/i2caux_dce120.c index 0e7b18260027..6a4f344c1db4 100644 --- a/drivers/gpu/drm/amd/display/dc/i2caux/dce120/i2caux_dce120.c +++ b/drivers/gpu/drm/amd/display/dc/i2caux/dce120/i2caux_dce120.c @@ -111,6 +111,7 @@ struct i2caux *dal_i2caux_dce120_create( dal_i2caux_dce110_construct(i2caux_dce110, ctx, + ARRAY_SIZE(dce120_aux_regs), dce120_aux_regs, dce120_hw_engine_regs, &i2c_shift, diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/dcn10/i2caux_dcn10.c b/drivers/gpu/drm/amd/display/dc/i2caux/dcn10/i2caux_dcn10.c index e44a8901f38b..a59c1f50c1e8 100644 --- a/drivers/gpu/drm/amd/display/dc/i2caux/dcn10/i2caux_dcn10.c +++ b/drivers/gpu/drm/amd/display/dc/i2caux/dcn10/i2caux_dcn10.c @@ -111,6 +111,7 @@ struct i2caux *dal_i2caux_dcn10_create( dal_i2caux_dce110_construct(i2caux_dce110, ctx, + ARRAY_SIZE(dcn10_aux_regs), dcn10_aux_regs, dcn10_hw_engine_regs, &i2c_shift, diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/engine.h b/drivers/gpu/drm/amd/display/dc/i2caux/engine.h index 33de8a8834dc..b16fb1ff687d 100644 --- a/drivers/gpu/drm/amd/display/dc/i2caux/engine.h +++ b/drivers/gpu/drm/amd/display/dc/i2caux/engine.h @@ -26,6 +26,8 @@ #ifndef __DAL_ENGINE_H__ #define __DAL_ENGINE_H__ +#include "dc_ddc_types.h" + enum i2caux_transaction_operation { I2CAUX_TRANSACTION_READ, I2CAUX_TRANSACTION_WRITE @@ -53,7 +55,8 @@ enum i2caux_transaction_status { I2CAUX_TRANSACTION_STATUS_FAILED_INCOMPLETE, I2CAUX_TRANSACTION_STATUS_FAILED_OPERATION, I2CAUX_TRANSACTION_STATUS_FAILED_INVALID_OPERATION, - I2CAUX_TRANSACTION_STATUS_FAILED_BUFFER_OVERFLOW + I2CAUX_TRANSACTION_STATUS_FAILED_BUFFER_OVERFLOW, + I2CAUX_TRANSACTION_STATUS_FAILED_HPD_DISCON }; struct i2caux_transaction_request { @@ -75,19 +78,6 @@ enum i2c_default_speed { I2CAUX_DEFAULT_I2C_SW_SPEED = 50 }; -enum i2caux_transaction_action { - I2CAUX_TRANSACTION_ACTION_I2C_WRITE = 0x00, - I2CAUX_TRANSACTION_ACTION_I2C_READ = 0x10, - I2CAUX_TRANSACTION_ACTION_I2C_STATUS_REQUEST = 0x20, - - I2CAUX_TRANSACTION_ACTION_I2C_WRITE_MOT = 0x40, - I2CAUX_TRANSACTION_ACTION_I2C_READ_MOT = 0x50, - I2CAUX_TRANSACTION_ACTION_I2C_STATUS_REQUEST_MOT = 0x60, - - I2CAUX_TRANSACTION_ACTION_DP_WRITE = 0x80, - I2CAUX_TRANSACTION_ACTION_DP_READ = 0x90 -}; - struct engine; struct engine_funcs { @@ -106,6 +96,7 @@ struct engine_funcs { struct engine { const struct engine_funcs *funcs; + uint32_t inst; struct ddc *ddc; struct dc_context *ctx; }; diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/i2c_engine.h b/drivers/gpu/drm/amd/display/dc/i2caux/i2c_engine.h index 58fc0f25eceb..ded6ea34b714 100644 --- a/drivers/gpu/drm/amd/display/dc/i2caux/i2c_engine.h +++ b/drivers/gpu/drm/amd/display/dc/i2caux/i2c_engine.h @@ -86,6 +86,8 @@ struct i2c_engine { struct engine base; const struct i2c_engine_funcs *funcs; uint32_t timeout_delay; + uint32_t setup_limit; + uint32_t send_reset_length; }; void dal_i2c_engine_construct( diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/i2caux.c b/drivers/gpu/drm/amd/display/dc/i2caux/i2caux.c index 14dc8c94d862..f7ed355fc84f 100644 --- a/drivers/gpu/drm/amd/display/dc/i2caux/i2caux.c +++ b/drivers/gpu/drm/amd/display/dc/i2caux/i2caux.c @@ -59,7 +59,7 @@ #include "dce120/i2caux_dce120.h" -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 #include "dcn10/i2caux_dcn10.h" #endif @@ -91,7 +91,7 @@ struct i2caux *dal_i2caux_create( return dal_i2caux_dce100_create(ctx); case DCE_VERSION_12_0: return dal_i2caux_dce120_create(ctx); -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 case DCN_VERSION_1_0: return dal_i2caux_dcn10_create(ctx); #endif @@ -254,7 +254,6 @@ bool dal_i2caux_submit_aux_command( break; } - cmd->payloads->length = request.payload.length; ++index_of_payload; } diff --git a/drivers/gpu/drm/amd/display/dc/inc/core_types.h b/drivers/gpu/drm/amd/display/dc/inc/core_types.h index a94942d4e66b..9f33306f9014 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/core_types.h +++ b/drivers/gpu/drm/amd/display/dc/inc/core_types.h @@ -33,7 +33,7 @@ #include "dc_bios_types.h" #include "mem_input.h" #include "hubp.h" -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 #include "mpc.h" #endif @@ -138,7 +138,7 @@ struct resource_pool { struct output_pixel_processor *opps[MAX_PIPES]; struct timing_generator *timing_generators[MAX_PIPES]; struct stream_encoder *stream_enc[MAX_PIPES * 2]; - + struct aux_engine *engines[MAX_PIPES]; struct hubbub *hubbub; struct mpc *mpc; struct pp_smu_funcs_rv *pp_smu; @@ -162,7 +162,7 @@ struct resource_pool { unsigned int audio_count; struct audio_support audio_support; - struct display_clock *display_clock; + struct dccg *dccg; struct irq_service *irqs; struct abm *abm; @@ -221,7 +221,7 @@ struct pipe_ctx { struct pipe_ctx *top_pipe; struct pipe_ctx *bottom_pipe; -#ifdef CONFIG_DRM_AMD_DC_DCN1_0 +#ifdef CONFIG_X86 struct _vcs_dpi_display_dlg_regs_st dlg_regs; struct _vcs_dpi_display_ttu_regs_st ttu_regs; struct _vcs_dpi_display_rq_regs_st rq_regs; @@ -255,8 +255,7 @@ struct dce_bw_output { }; struct dcn_bw_output { - struct dc_clocks cur_clk; - struct dc_clocks calc_clk; + struct dc_clocks clk; struct dcn_watermark_set watermarks; }; @@ -277,11 +276,11 @@ struct dc_state { /* Note: these are big structures, do *not* put on stack! */ struct dm_pp_display_configuration pp_display_cfg; -#ifdef CONFIG_DRM_AMD_DC_DCN1_0 +#ifdef CONFIG_X86 struct dcn_bw_internal_vars dcn_bw_vars; #endif - struct display_clock *dis_clk; + struct dccg *dis_clk; struct kref refcount; }; diff --git a/drivers/gpu/drm/amd/display/dc/inc/dc_link_ddc.h b/drivers/gpu/drm/amd/display/dc/inc/dc_link_ddc.h index 090b7a8dd67b..538b83303b86 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/dc_link_ddc.h +++ b/drivers/gpu/drm/amd/display/dc/inc/dc_link_ddc.h @@ -102,21 +102,13 @@ bool dal_ddc_service_query_ddc_data( uint8_t *read_buf, uint32_t read_size); -ssize_t dal_ddc_service_read_dpcd_data( - struct ddc_service *ddc, - bool i2c, - enum i2c_mot_mode mot, - uint32_t address, - uint8_t *data, - uint32_t len); - -enum ddc_result dal_ddc_service_write_dpcd_data( - struct ddc_service *ddc, - bool i2c, - enum i2c_mot_mode mot, - uint32_t address, - const uint8_t *data, - uint32_t len); +int dc_link_aux_transfer(struct ddc_service *ddc, + unsigned int address, + uint8_t *reply, + void *buffer, + unsigned int size, + enum aux_transaction_type type, + enum i2caux_transaction_action action); void dal_ddc_service_write_scdc_data( struct ddc_service *ddc_service, diff --git a/drivers/gpu/drm/amd/display/dc/inc/dc_link_dp.h b/drivers/gpu/drm/amd/display/dc/inc/dc_link_dp.h index 2f783c650084..a37255c757e0 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/dc_link_dp.h +++ b/drivers/gpu/drm/amd/display/dc/inc/dc_link_dp.h @@ -33,9 +33,10 @@ struct dc_link; struct dc_stream_state; struct dc_link_settings; -bool dp_hbr_verify_link_cap( +bool dp_verify_link_cap( struct dc_link *link, - struct dc_link_settings *known_limit_link_setting); + struct dc_link_settings *known_limit_link_setting, + int *fail_count); bool dp_validate_mode_timing( struct dc_link *link, diff --git a/drivers/gpu/drm/amd/display/dc/inc/dcn_calcs.h b/drivers/gpu/drm/amd/display/dc/inc/dcn_calcs.h index 132d18d4b293..ddbb673caa08 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/dcn_calcs.h +++ b/drivers/gpu/drm/amd/display/dc/inc/dcn_calcs.h @@ -625,7 +625,7 @@ bool dcn_validate_bandwidth( unsigned int dcn_find_dcfclk_suits_all( const struct dc *dc, - struct clocks_value *clocks); + struct dc_clocks *clocks); void dcn_bw_update_from_pplib(struct dc *dc); void dcn_bw_notify_pplib_of_wm_ranges(struct dc *dc); diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/aux_engine.h b/drivers/gpu/drm/amd/display/dc/inc/hw/aux_engine.h new file mode 100644 index 000000000000..e79cd4e92919 --- /dev/null +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/aux_engine.h @@ -0,0 +1,180 @@ +/* + * Copyright 2012-15 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: AMD + * + */ + +#ifndef __DAL_AUX_ENGINE_H__ +#define __DAL_AUX_ENGINE_H__ + +#include "dc_ddc_types.h" +#include "include/i2caux_interface.h" + +enum i2caux_transaction_operation { + I2CAUX_TRANSACTION_READ, + I2CAUX_TRANSACTION_WRITE +}; + +enum i2caux_transaction_address_space { + I2CAUX_TRANSACTION_ADDRESS_SPACE_I2C = 1, + I2CAUX_TRANSACTION_ADDRESS_SPACE_DPCD +}; + +struct i2caux_transaction_payload { + enum i2caux_transaction_address_space address_space; + uint32_t address; + uint32_t length; + uint8_t *data; +}; + +enum i2caux_transaction_status { + I2CAUX_TRANSACTION_STATUS_UNKNOWN = (-1L), + I2CAUX_TRANSACTION_STATUS_SUCCEEDED, + I2CAUX_TRANSACTION_STATUS_FAILED_CHANNEL_BUSY, + I2CAUX_TRANSACTION_STATUS_FAILED_TIMEOUT, + I2CAUX_TRANSACTION_STATUS_FAILED_PROTOCOL_ERROR, + I2CAUX_TRANSACTION_STATUS_FAILED_NACK, + I2CAUX_TRANSACTION_STATUS_FAILED_INCOMPLETE, + I2CAUX_TRANSACTION_STATUS_FAILED_OPERATION, + I2CAUX_TRANSACTION_STATUS_FAILED_INVALID_OPERATION, + I2CAUX_TRANSACTION_STATUS_FAILED_BUFFER_OVERFLOW, + I2CAUX_TRANSACTION_STATUS_FAILED_HPD_DISCON +}; + +struct i2caux_transaction_request { + enum i2caux_transaction_operation operation; + struct i2caux_transaction_payload payload; + enum i2caux_transaction_status status; +}; + +enum i2caux_engine_type { + I2CAUX_ENGINE_TYPE_UNKNOWN = (-1L), + I2CAUX_ENGINE_TYPE_AUX, + I2CAUX_ENGINE_TYPE_I2C_DDC_HW, + I2CAUX_ENGINE_TYPE_I2C_GENERIC_HW, + I2CAUX_ENGINE_TYPE_I2C_SW +}; + +enum i2c_default_speed { + I2CAUX_DEFAULT_I2C_HW_SPEED = 50, + I2CAUX_DEFAULT_I2C_SW_SPEED = 50 +}; + +union aux_config; + +struct aux_engine { + uint32_t inst; + struct ddc *ddc; + struct dc_context *ctx; + const struct aux_engine_funcs *funcs; + /* following values are expressed in milliseconds */ + uint32_t delay; + uint32_t max_defer_write_retry; + bool acquire_reset; +}; + +struct read_command_context { + uint8_t *buffer; + uint32_t current_read_length; + uint32_t offset; + enum i2caux_transaction_status status; + + struct aux_request_transaction_data request; + struct aux_reply_transaction_data reply; + + uint8_t returned_byte; + + uint32_t timed_out_retry_aux; + uint32_t invalid_reply_retry_aux; + uint32_t defer_retry_aux; + uint32_t defer_retry_i2c; + uint32_t invalid_reply_retry_aux_on_ack; + + bool transaction_complete; + bool operation_succeeded; +}; + +struct write_command_context { + bool mot; + + uint8_t *buffer; + uint32_t current_write_length; + enum i2caux_transaction_status status; + + struct aux_request_transaction_data request; + struct aux_reply_transaction_data reply; + + uint8_t returned_byte; + + uint32_t timed_out_retry_aux; + uint32_t invalid_reply_retry_aux; + uint32_t defer_retry_aux; + uint32_t defer_retry_i2c; + uint32_t max_defer_retry; + uint32_t ack_m_retry; + + uint8_t reply_data[DEFAULT_AUX_MAX_DATA_SIZE]; + + bool transaction_complete; + bool operation_succeeded; +}; + + +struct aux_engine_funcs { + void (*destroy)( + struct aux_engine **ptr); + bool (*acquire_engine)( + struct aux_engine *engine); + void (*configure)( + struct aux_engine *engine, + union aux_config cfg); + void (*submit_channel_request)( + struct aux_engine *engine, + struct aux_request_transaction_data *request); + void (*process_channel_reply)( + struct aux_engine *engine, + struct aux_reply_transaction_data *reply); + int (*read_channel_reply)( + struct aux_engine *engine, + uint32_t size, + uint8_t *buffer, + uint8_t *reply_result, + uint32_t *sw_status); + enum aux_channel_operation_result (*get_channel_status)( + struct aux_engine *engine, + uint8_t *returned_bytes); + bool (*is_engine_available)(struct aux_engine *engine); + enum i2caux_engine_type (*get_engine_type)( + const struct aux_engine *engine); + bool (*acquire)( + struct aux_engine *engine, + struct ddc *ddc); + bool (*submit_request)( + struct aux_engine *engine, + struct i2caux_transaction_request *request, + bool middle_of_transaction); + void (*release_engine)( + struct aux_engine *engine); + void (*destroy_engine)( + struct aux_engine **engine); +}; +#endif diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/display_clock.h b/drivers/gpu/drm/amd/display/dc/inc/hw/display_clock.h index f5f69cd81f6f..3c7ccb68ecdb 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/display_clock.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/display_clock.h @@ -27,23 +27,7 @@ #define __DISPLAY_CLOCK_H__ #include "dm_services_types.h" - - -struct clocks_value { - int dispclk_in_khz; - int max_pixelclk_in_khz; - int max_non_dp_phyclk_in_khz; - int max_dp_phyclk_in_khz; - bool dispclk_notify_pplib_done; - bool pixelclk_notify_pplib_done; - bool phyclk_notigy_pplib_done; - int dcfclock_in_khz; - int dppclk_in_khz; - int mclk_in_khz; - int phyclk_in_khz; - int common_vdd_level; -}; - +#include "dc.h" /* Structure containing all state-dependent clocks * (dependent on "enum clocks_state") */ @@ -52,34 +36,23 @@ struct state_dependent_clocks { int pixel_clk_khz; }; -struct display_clock { +struct dccg { struct dc_context *ctx; const struct display_clock_funcs *funcs; enum dm_pp_clocks_state max_clks_state; enum dm_pp_clocks_state cur_min_clks_state; - struct clocks_value cur_clocks_value; + struct dc_clocks clks; }; struct display_clock_funcs { - int (*set_clock)(struct display_clock *disp_clk, + void (*update_clocks)(struct dccg *dccg, + struct dc_clocks *new_clocks, + bool safe_to_lower); + int (*set_dispclk)(struct dccg *dccg, int requested_clock_khz); - enum dm_pp_clocks_state (*get_required_clocks_state)( - struct display_clock *disp_clk, - struct state_dependent_clocks *req_clocks); - - bool (*set_min_clocks_state)(struct display_clock *disp_clk, - enum dm_pp_clocks_state dm_pp_clocks_state); - - int (*get_dp_ref_clk_frequency)(struct display_clock *disp_clk); - - bool (*apply_clock_voltage_request)( - struct display_clock *disp_clk, - enum dm_pp_clock_type clocks_type, - int clocks_in_khz, - bool pre_mode_set, - bool update_dp_phyclk); + int (*get_dp_ref_clk_frequency)(struct dccg *dccg); }; #endif /* __DISPLAY_CLOCK_H__ */ diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/dmcu.h b/drivers/gpu/drm/amd/display/dc/inc/hw/dmcu.h index de60f940030d..4550747fb61c 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/dmcu.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/dmcu.h @@ -48,7 +48,7 @@ struct dmcu_funcs { const char *src, unsigned int bytes); void (*set_psr_enable)(struct dmcu *dmcu, bool enable, bool wait); - void (*setup_psr)(struct dmcu *dmcu, + bool (*setup_psr)(struct dmcu *dmcu, struct dc_link *link, struct psr_context *psr_context); void (*get_psr_state)(struct dmcu *dmcu, uint32_t *psr_state); diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h b/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h index 582458f028f8..74ad94b0e4f0 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h @@ -151,6 +151,9 @@ struct dpp_funcs { void (*dpp_set_hdr_multiplier)( struct dpp *dpp_base, uint32_t multiplier); + void (*set_optional_cursor_attributes)( + struct dpp *dpp_base, + struct dpp_cursor_attributes *attr); void (*dpp_dppclk_control)( struct dpp *dpp_base, diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h b/drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h index 97df82cddf82..4f3f9e68ccfa 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h @@ -43,10 +43,9 @@ enum cursor_lines_per_chunk { }; struct hubp { - struct hubp_funcs *funcs; + const struct hubp_funcs *funcs; struct dc_context *ctx; struct dc_plane_address request_address; - struct dc_plane_address current_address; int inst; /* run time states */ diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/mem_input.h b/drivers/gpu/drm/amd/display/dc/inc/hw/mem_input.h index 47f1dc5a43b7..da89c2edb07c 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/mem_input.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/mem_input.h @@ -64,7 +64,7 @@ struct stutter_modes { }; struct mem_input { - struct mem_input_funcs *funcs; + const struct mem_input_funcs *funcs; struct dc_context *ctx; struct dc_plane_address request_address; struct dc_plane_address current_address; diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h b/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h index 69cb0a105300..af700c7dac50 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h @@ -156,6 +156,9 @@ struct timing_generator_funcs { uint32_t *v_blank_end, uint32_t *h_position, uint32_t *v_position); + bool (*get_otg_active_size)(struct timing_generator *optc, + uint32_t *otg_active_width, + uint32_t *otg_active_height); void (*set_early_control)(struct timing_generator *tg, uint32_t early_cntl); void (*wait_for_state)(struct timing_generator *tg, diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h b/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h index 63fc6c499789..a14ce4de80b2 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h @@ -44,6 +44,7 @@ struct dce_hwseq_wa { bool blnd_crtc_trigger; bool DEGVIDCN10_253; bool false_optc_underflow; + bool DEGVIDCN10_254; }; struct hwseq_wa_state { @@ -101,10 +102,18 @@ struct hw_sequencer_funcs { const struct dc *dc, struct pipe_ctx *pipe_ctx); + void (*plane_atomic_disconnect)( + struct dc *dc, + struct pipe_ctx *pipe_ctx); + void (*update_dchub)( struct dce_hwseq *hws, struct dchub_init_data *dh_data); + void (*update_mpcc)( + struct dc *dc, + struct pipe_ctx *pipe_ctx); + void (*update_pending_status)( struct pipe_ctx *pipe_ctx); @@ -154,20 +163,24 @@ struct hw_sequencer_funcs { struct dc_link_settings *link_settings); void (*blank_stream)(struct pipe_ctx *pipe_ctx); + + void (*enable_audio_stream)(struct pipe_ctx *pipe_ctx); + + void (*disable_audio_stream)(struct pipe_ctx *pipe_ctx, int option); + void (*pipe_control_lock)( struct dc *dc, struct pipe_ctx *pipe, bool lock); void (*blank_pixel_data)( struct dc *dc, - struct stream_resource *stream_res, - struct dc_stream_state *stream, + struct pipe_ctx *pipe_ctx, bool blank); void (*set_bandwidth)( struct dc *dc, struct dc_state *context, - bool decrease_allowed); + bool safe_to_lower); void (*set_drr)(struct pipe_ctx **pipe_ctx, int num_pipes, int vmin, int vmax); @@ -210,6 +223,7 @@ struct hw_sequencer_funcs { void (*set_cursor_position)(struct pipe_ctx *pipe); void (*set_cursor_attribute)(struct pipe_ctx *pipe); + void (*set_cursor_sdr_white_level)(struct pipe_ctx *pipe); }; diff --git a/drivers/gpu/drm/amd/display/dc/inc/reg_helper.h b/drivers/gpu/drm/amd/display/dc/inc/reg_helper.h index 3306e7b0b3e3..cf5a84b9e27c 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/reg_helper.h +++ b/drivers/gpu/drm/amd/display/dc/inc/reg_helper.h @@ -445,4 +445,50 @@ uint32_t generic_reg_get8(const struct dc_context *ctx, uint32_t addr, uint8_t shift6, uint32_t mask6, uint32_t *field_value6, uint8_t shift7, uint32_t mask7, uint32_t *field_value7, uint8_t shift8, uint32_t mask8, uint32_t *field_value8); + + +/* indirect register access */ + +#define IX_REG_SET_N(index_reg_name, data_reg_name, index, n, initial_val, ...) \ + generic_indirect_reg_update_ex(CTX, \ + REG(index_reg_name), REG(data_reg_name), IND_REG(index), \ + initial_val, \ + n, __VA_ARGS__) + +#define IX_REG_SET_2(index_reg_name, data_reg_name, index, init_value, f1, v1, f2, v2) \ + IX_REG_SET_N(index_reg_name, data_reg_name, index, 2, init_value, \ + FN(reg, f1), v1,\ + FN(reg, f2), v2) + + +#define IX_REG_READ(index_reg_name, data_reg_name, index) \ + generic_read_indirect_reg(CTX, REG(index_reg_name), REG(data_reg_name), IND_REG(index)) + + + +#define IX_REG_UPDATE_N(index_reg_name, data_reg_name, index, n, ...) \ + generic_indirect_reg_update_ex(CTX, \ + REG(index_reg_name), REG(data_reg_name), IND_REG(index), \ + IX_REG_READ(index_reg_name, data_reg_name, index), \ + n, __VA_ARGS__) + +#define IX_REG_UPDATE_2(index_reg_name, data_reg_name, index, f1, v1, f2, v2) \ + IX_REG_UPDATE_N(index_reg_name, data_reg_name, index, 2,\ + FN(reg, f1), v1,\ + FN(reg, f2), v2) + +void generic_write_indirect_reg(const struct dc_context *ctx, + uint32_t addr_index, uint32_t addr_data, + uint32_t index, uint32_t data); + +uint32_t generic_read_indirect_reg(const struct dc_context *ctx, + uint32_t addr_index, uint32_t addr_data, + uint32_t index); + +uint32_t generic_indirect_reg_update_ex(const struct dc_context *ctx, + uint32_t addr_index, uint32_t addr_data, + uint32_t index, uint32_t reg_val, int n, + uint8_t shift1, uint32_t mask1, uint32_t field_value1, + ...); + #endif /* DRIVERS_GPU_DRM_AMD_DC_DEV_DC_INC_REG_HELPER_H_ */ diff --git a/drivers/gpu/drm/amd/display/dc/inc/resource.h b/drivers/gpu/drm/amd/display/dc/inc/resource.h index 640a647f4611..e92facbd038f 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/resource.h +++ b/drivers/gpu/drm/amd/display/dc/inc/resource.h @@ -38,6 +38,7 @@ enum dce_version resource_parse_asic_id( struct resource_caps { int num_timing_generator; + int num_opp; int num_video_plane; int num_audio; int num_stream_encoder; diff --git a/drivers/gpu/drm/amd/display/dc/irq/Makefile b/drivers/gpu/drm/amd/display/dc/irq/Makefile index 498515aad4a5..a76ee600ecee 100644 --- a/drivers/gpu/drm/amd/display/dc/irq/Makefile +++ b/drivers/gpu/drm/amd/display/dc/irq/Makefile @@ -60,7 +60,7 @@ AMD_DISPLAY_FILES += $(AMD_DAL_IRQ_DCE12) ############################################################################### # DCN 1x ############################################################################### -ifdef CONFIG_DRM_AMD_DC_DCN1_0 +ifdef CONFIG_X86 IRQ_DCN1 = irq_service_dcn10.o AMD_DAL_IRQ_DCN1 = $(addprefix $(AMDDALPATH)/dc/irq/dcn10/,$(IRQ_DCN1)) diff --git a/drivers/gpu/drm/amd/display/dc/irq/irq_service.c b/drivers/gpu/drm/amd/display/dc/irq/irq_service.c index dcdfa0f01551..ae3fd0a235ba 100644 --- a/drivers/gpu/drm/amd/display/dc/irq/irq_service.c +++ b/drivers/gpu/drm/amd/display/dc/irq/irq_service.c @@ -36,7 +36,7 @@ #include "dce120/irq_service_dce120.h" -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 #include "dcn10/irq_service_dcn10.h" #endif @@ -78,7 +78,7 @@ const struct irq_source_info *find_irq_source_info( struct irq_service *irq_service, enum dc_irq_source source) { - if (source > DAL_IRQ_SOURCES_NUMBER || source < DC_IRQ_SOURCE_INVALID) + if (source >= DAL_IRQ_SOURCES_NUMBER || source < DC_IRQ_SOURCE_INVALID) return NULL; return &irq_service->info[source]; diff --git a/drivers/gpu/drm/amd/display/dc/os_types.h b/drivers/gpu/drm/amd/display/dc/os_types.h index a407892905af..c9fce9066ad8 100644 --- a/drivers/gpu/drm/amd/display/dc/os_types.h +++ b/drivers/gpu/drm/amd/display/dc/os_types.h @@ -48,7 +48,7 @@ #define dm_error(fmt, ...) DRM_ERROR(fmt, ##__VA_ARGS__) -#if defined(CONFIG_DRM_AMD_DC_DCN1_0) +#ifdef CONFIG_X86 #include <asm/fpu/api.h> #endif diff --git a/drivers/gpu/drm/amd/display/include/ddc_service_types.h b/drivers/gpu/drm/amd/display/include/ddc_service_types.h index 019e7a095ea1..d968956a10cd 100644 --- a/drivers/gpu/drm/amd/display/include/ddc_service_types.h +++ b/drivers/gpu/drm/amd/display/include/ddc_service_types.h @@ -40,7 +40,8 @@ enum ddc_result { DDC_RESULT_FAILED_INCOMPLETE, DDC_RESULT_FAILED_OPERATION, DDC_RESULT_FAILED_INVALID_OPERATION, - DDC_RESULT_FAILED_BUFFER_OVERFLOW + DDC_RESULT_FAILED_BUFFER_OVERFLOW, + DDC_RESULT_FAILED_HPD_DISCON }; enum ddc_service_type { diff --git a/drivers/gpu/drm/amd/display/include/dpcd_defs.h b/drivers/gpu/drm/amd/display/include/dpcd_defs.h index d8e52e3b8e3c..1c66166d0a94 100644 --- a/drivers/gpu/drm/amd/display/include/dpcd_defs.h +++ b/drivers/gpu/drm/amd/display/include/dpcd_defs.h @@ -27,6 +27,9 @@ #define __DAL_DPCD_DEFS_H__ #include <drm/drm_dp_helper.h> +#ifndef DP_SINK_HW_REVISION_START // can remove this once the define gets into linux drm_dp_helper.h +#define DP_SINK_HW_REVISION_START 0x409 +#endif enum dpcd_revision { DPCD_REV_10 = 0x10, diff --git a/drivers/gpu/drm/amd/display/include/fixed31_32.h b/drivers/gpu/drm/amd/display/include/fixed31_32.h index a981b3e99ab3..52a73332befb 100644 --- a/drivers/gpu/drm/amd/display/include/fixed31_32.h +++ b/drivers/gpu/drm/amd/display/include/fixed31_32.h @@ -26,6 +26,13 @@ #ifndef __DAL_FIXED31_32_H__ #define __DAL_FIXED31_32_H__ +#ifndef LLONG_MAX +#define LLONG_MAX 9223372036854775807ll +#endif +#ifndef LLONG_MIN +#define LLONG_MIN (-LLONG_MAX - 1ll) +#endif + #define FIXED31_32_BITS_PER_FRACTIONAL_PART 32 #ifndef LLONG_MIN #define LLONG_MIN (1LL<<63) diff --git a/drivers/gpu/drm/amd/display/include/grph_object_ctrl_defs.h b/drivers/gpu/drm/amd/display/include/grph_object_ctrl_defs.h index 36bbad594267..f312834fef50 100644 --- a/drivers/gpu/drm/amd/display/include/grph_object_ctrl_defs.h +++ b/drivers/gpu/drm/amd/display/include/grph_object_ctrl_defs.h @@ -395,6 +395,8 @@ struct integrated_info { struct i2c_reg_info dp3_ext_hdmi_reg_settings[9]; unsigned char dp3_ext_hdmi_6g_reg_num; struct i2c_reg_info dp3_ext_hdmi_6g_reg_settings[3]; + /* V11 */ + uint32_t dp_ss_control; }; /** diff --git a/drivers/gpu/drm/amd/display/include/grph_object_defs.h b/drivers/gpu/drm/amd/display/include/grph_object_defs.h index 2941b882b0b6..58bb42ed85ca 100644 --- a/drivers/gpu/drm/amd/display/include/grph_object_defs.h +++ b/drivers/gpu/drm/amd/display/include/grph_object_defs.h @@ -37,6 +37,10 @@ * ******************************************************************** */ +#define MAX_CONNECTOR_NUMBER_PER_SLOT (16) +#define MAX_BOARD_SLOTS (4) +#define INVALID_CONNECTOR_INDEX ((unsigned int)(-1)) + /* HPD unit id - HW direct translation */ enum hpd_source_id { HPD_SOURCEID1 = 0, @@ -136,5 +140,47 @@ enum sync_source { SYNC_SOURCE_DUAL_GPU_PIN }; +/* connector sizes in millimeters - from BiosParserTypes.hpp */ +#define CONNECTOR_SIZE_DVI 40 +#define CONNECTOR_SIZE_VGA 32 +#define CONNECTOR_SIZE_HDMI 16 +#define CONNECTOR_SIZE_DP 16 +#define CONNECTOR_SIZE_MINI_DP 9 +#define CONNECTOR_SIZE_UNKNOWN 30 + +enum connector_layout_type { + CONNECTOR_LAYOUT_TYPE_UNKNOWN, + CONNECTOR_LAYOUT_TYPE_DVI_D, + CONNECTOR_LAYOUT_TYPE_DVI_I, + CONNECTOR_LAYOUT_TYPE_VGA, + CONNECTOR_LAYOUT_TYPE_HDMI, + CONNECTOR_LAYOUT_TYPE_DP, + CONNECTOR_LAYOUT_TYPE_MINI_DP, +}; +struct connector_layout_info { + struct graphics_object_id connector_id; + enum connector_layout_type connector_type; + unsigned int length; + unsigned int position; /* offset in mm from right side of the board */ +}; + +/* length and width in mm */ +struct slot_layout_info { + unsigned int length; + unsigned int width; + unsigned int num_of_connectors; + struct connector_layout_info connectors[MAX_CONNECTOR_NUMBER_PER_SLOT]; +}; + +struct board_layout_info { + unsigned int num_of_slots; + /* indicates valid information in bracket layout structure. */ + unsigned int is_number_of_slots_valid : 1; + unsigned int is_slots_size_valid : 1; + unsigned int is_connector_offsets_valid : 1; + unsigned int is_connector_lengths_valid : 1; + + struct slot_layout_info slots[MAX_BOARD_SLOTS]; +}; #endif diff --git a/drivers/gpu/drm/amd/display/include/grph_object_id.h b/drivers/gpu/drm/amd/display/include/grph_object_id.h index c4197432eb7c..33b3d755fe65 100644 --- a/drivers/gpu/drm/amd/display/include/grph_object_id.h +++ b/drivers/gpu/drm/amd/display/include/grph_object_id.h @@ -197,6 +197,11 @@ enum transmitter_color_depth { TRANSMITTER_COLOR_DEPTH_48 /* 16 bits */ }; +enum dp_alt_mode { + DP_Alt_mode__Unknown = 0, + DP_Alt_mode__Connect, + DP_Alt_mode__NoConnect, +}; /* ***************************************************************************** * graphics_object_id struct @@ -287,4 +292,15 @@ static inline enum engine_id dal_graphics_object_id_get_engine_id( return (enum engine_id) id.id; return ENGINE_ID_UNKNOWN; } + +static inline bool dal_graphics_object_id_equal( + struct graphics_object_id id_1, + struct graphics_object_id id_2) +{ + if ((id_1.id == id_2.id) && (id_1.enum_id == id_2.enum_id) && + (id_1.type == id_2.type)) { + return true; + } + return false; +} #endif diff --git a/drivers/gpu/drm/amd/display/include/logger_interface.h b/drivers/gpu/drm/amd/display/include/logger_interface.h index dc98d6d4b2bd..e3c79616682d 100644 --- a/drivers/gpu/drm/amd/display/include/logger_interface.h +++ b/drivers/gpu/drm/amd/display/include/logger_interface.h @@ -40,47 +40,7 @@ struct dc_state; * */ -struct dal_logger *dal_logger_create(struct dc_context *ctx, uint32_t log_mask); - -uint32_t dal_logger_destroy(struct dal_logger **logger); - -void dm_logger_flush_buffer(struct dal_logger *logger, bool should_warn); - -void dm_logger_write( - struct dal_logger *logger, - enum dc_log_type log_type, - const char *msg, - ...); - -void dm_logger_append( - struct log_entry *entry, - const char *msg, - ...); - -void dm_logger_append_va( - struct log_entry *entry, - const char *msg, - va_list args); - -void dm_logger_open( - struct dal_logger *logger, - struct log_entry *entry, - enum dc_log_type log_type); - -void dm_logger_close(struct log_entry *entry); - -void dc_conn_log(struct dc_context *ctx, - const struct dc_link *link, - uint8_t *hex_data, - int hex_data_count, - enum dc_log_type event, - const char *msg, - ...); - -void logger_write(struct dal_logger *logger, - enum dc_log_type log_type, - const char *msg, - void *paralist); +void dc_conn_log_hex_linux(const uint8_t *hex_data, int hex_data_count); void pre_surface_trace( struct dc *dc, @@ -106,28 +66,31 @@ void context_clock_trace( * marked by this macro. * Note that the message will be printed exactly once for every function * it is used in order to avoid repeating of the same message. */ + #define DAL_LOGGER_NOT_IMPL(fmt, ...) \ -{ \ - static bool print_not_impl = true; \ -\ - if (print_not_impl == true) { \ - print_not_impl = false; \ - dm_logger_write(ctx->logger, LOG_WARNING, \ - "DAL_NOT_IMPL: " fmt, ##__VA_ARGS__); \ - } \ -} + do { \ + static bool print_not_impl = true; \ + if (print_not_impl == true) { \ + print_not_impl = false; \ + DRM_WARN("DAL_NOT_IMPL: " fmt, ##__VA_ARGS__); \ + } \ + } while (0) /****************************************************************************** * Convenience macros to save on typing. *****************************************************************************/ #define DC_ERROR(...) \ - dm_logger_write(dc_ctx->logger, LOG_ERROR, \ - __VA_ARGS__) + do { \ + (void)(dc_ctx); \ + DC_LOG_ERROR(__VA_ARGS__); \ + } while (0) #define DC_SYNC_INFO(...) \ - dm_logger_write(dc_ctx->logger, LOG_SYNC, \ - __VA_ARGS__) + do { \ + (void)(dc_ctx); \ + DC_LOG_SYNC(__VA_ARGS__); \ + } while (0) /* Connectivity log format: * [time stamp] [drm] [Major_minor] [connector name] message..... @@ -137,20 +100,30 @@ void context_clock_trace( */ #define CONN_DATA_DETECT(link, hex_data, hex_len, ...) \ - dc_conn_log(link->ctx, link, hex_data, hex_len, \ - LOG_EVENT_DETECTION, ##__VA_ARGS__) + do { \ + (void)(link); \ + dc_conn_log_hex_linux(hex_data, hex_len); \ + DC_LOG_EVENT_DETECTION(__VA_ARGS__); \ + } while (0) #define CONN_DATA_LINK_LOSS(link, hex_data, hex_len, ...) \ - dc_conn_log(link->ctx, link, hex_data, hex_len, \ - LOG_EVENT_LINK_LOSS, ##__VA_ARGS__) + do { \ + (void)(link); \ + dc_conn_log_hex_linux(hex_data, hex_len); \ + DC_LOG_EVENT_LINK_LOSS(__VA_ARGS__); \ + } while (0) #define CONN_MSG_LT(link, ...) \ - dc_conn_log(link->ctx, link, NULL, 0, \ - LOG_EVENT_LINK_TRAINING, ##__VA_ARGS__) + do { \ + (void)(link); \ + DC_LOG_EVENT_LINK_TRAINING(__VA_ARGS__); \ + } while (0) #define CONN_MSG_MODE(link, ...) \ - dc_conn_log(link->ctx, link, NULL, 0, \ - LOG_EVENT_MODE_SET, ##__VA_ARGS__) + do { \ + (void)(link); \ + DC_LOG_EVENT_MODE_SET(__VA_ARGS__); \ + } while (0) /* * Display Test Next logging @@ -165,38 +138,21 @@ void context_clock_trace( dm_dtn_log_end(dc_ctx) #define PERFORMANCE_TRACE_START() \ - unsigned long long perf_trc_start_stmp = dm_get_timestamp(dc->ctx); \ - unsigned long long perf_trc_start_log_msk = dc->ctx->logger->mask; \ - unsigned int perf_trc_start_log_flags = dc->ctx->logger->flags.value; \ - if (dc->debug.performance_trace) {\ - dm_logger_flush_buffer(dc->ctx->logger, false);\ - dc->ctx->logger->mask = 1<<LOG_PERF_TRACE;\ - dc->ctx->logger->flags.bits.ENABLE_CONSOLE = 0;\ - dc->ctx->logger->flags.bits.ENABLE_BUFFER = 1;\ - } - -#define PERFORMANCE_TRACE_END() do {\ - unsigned long long perf_trc_end_stmp = dm_get_timestamp(dc->ctx);\ - if (dc->debug.performance_trace) {\ - dm_logger_write(dc->ctx->logger, \ - LOG_PERF_TRACE, \ - "%s duration: %d ticks\n", __func__,\ + unsigned long long perf_trc_start_stmp = dm_get_timestamp(dc->ctx) + +#define PERFORMANCE_TRACE_END() \ + do { \ + unsigned long long perf_trc_end_stmp = dm_get_timestamp(dc->ctx); \ + if (dc->debug.performance_trace) { \ + DC_LOG_PERF_TRACE("%s duration: %lld ticks\n", __func__, \ perf_trc_end_stmp - perf_trc_start_stmp); \ - if (perf_trc_start_log_msk != 1<<LOG_PERF_TRACE) {\ - dc->ctx->logger->mask = perf_trc_start_log_msk;\ - dc->ctx->logger->flags.value = perf_trc_start_log_flags;\ - dm_logger_flush_buffer(dc->ctx->logger, false);\ } \ - } \ -} while (0) + } while (0) -#define DISPLAY_STATS_BEGIN(entry) \ - dm_logger_open(dc->ctx->logger, &entry, LOG_DISPLAYSTATS) +#define DISPLAY_STATS_BEGIN(entry) (void)(entry) -#define DISPLAY_STATS(msg, ...) \ - dm_logger_append(&log_entry, msg, ##__VA_ARGS__) +#define DISPLAY_STATS(msg, ...) DC_LOG_PERF_TRACE(msg, __VA_ARGS__) -#define DISPLAY_STATS_END(entry) \ - dm_logger_close(&entry) +#define DISPLAY_STATS_END(entry) (void)(entry) #endif /* __DAL_LOGGER_INTERFACE_H__ */ diff --git a/drivers/gpu/drm/amd/display/include/logger_types.h b/drivers/gpu/drm/amd/display/include/logger_types.h index 0a540b9897a6..ad3695e67b76 100644 --- a/drivers/gpu/drm/amd/display/include/logger_types.h +++ b/drivers/gpu/drm/amd/display/include/logger_types.h @@ -138,63 +138,4 @@ enum dc_log_type { (1 << LOG_HW_AUDIO)| \ (1 << LOG_BANDWIDTH_CALCS)*/ -union logger_flags { - struct { - uint32_t ENABLE_CONSOLE:1; /* Print to console */ - uint32_t ENABLE_BUFFER:1; /* Print to buffer */ - uint32_t RESERVED:30; - } bits; - uint32_t value; -}; - -struct log_entry { - struct dal_logger *logger; - enum dc_log_type type; - - char *buf; - uint32_t buf_offset; - uint32_t max_buf_bytes; -}; - -/** -* Structure for enumerating log types -*/ -struct dc_log_type_info { - enum dc_log_type type; - char name[MAX_NAME_LEN]; -}; - -/* Structure for keeping track of offsets, buffer, etc */ - -#define DAL_LOGGER_BUFFER_MAX_SIZE 2048 - -/*Connectivity log needs to output EDID, which needs at lease 256x3 bytes, - * change log line size to 896 to meet the request. - */ -#define LOG_MAX_LINE_SIZE 896 - -struct dal_logger { - - /* How far into the circular buffer has been read by dsat - * Read offset should never cross write offset. Write \0's to - * read data just to be sure? - */ - uint32_t buffer_read_offset; - - /* How far into the circular buffer we have written - * Write offset should never cross read offset - */ - uint32_t buffer_write_offset; - - uint32_t open_count; - - char *log_buffer; /* Pointer to malloc'ed buffer */ - uint32_t log_buffer_size; /* Size of circular buffer */ - - uint32_t mask; /*array of masks for major elements*/ - - union logger_flags flags; - struct dc_context *ctx; -}; - #endif /* __DAL_LOGGER_TYPES_H__ */ diff --git a/drivers/gpu/drm/amd/display/modules/color/color_gamma.c b/drivers/gpu/drm/amd/display/modules/color/color_gamma.c index eee0dfad6962..bf29733958c3 100644 --- a/drivers/gpu/drm/amd/display/modules/color/color_gamma.c +++ b/drivers/gpu/drm/amd/display/modules/color/color_gamma.c @@ -131,6 +131,63 @@ static void compute_de_pq(struct fixed31_32 in_x, struct fixed31_32 *out_y) dc_fixpt_div(dc_fixpt_one, m1)); } + +/*de gamma, none linear to linear*/ +static void compute_hlg_oetf(struct fixed31_32 in_x, bool is_light0_12, struct fixed31_32 *out_y) +{ + struct fixed31_32 a; + struct fixed31_32 b; + struct fixed31_32 c; + struct fixed31_32 threshold; + struct fixed31_32 reference_white_level; + + a = dc_fixpt_from_fraction(17883277, 100000000); + if (is_light0_12) { + /*light 0-12*/ + b = dc_fixpt_from_fraction(28466892, 100000000); + c = dc_fixpt_from_fraction(55991073, 100000000); + threshold = dc_fixpt_one; + reference_white_level = dc_fixpt_half; + } else { + /*light 0-1*/ + b = dc_fixpt_from_fraction(2372241, 100000000); + c = dc_fixpt_add(dc_fixpt_one, dc_fixpt_from_fraction(429347, 100000000)); + threshold = dc_fixpt_from_fraction(1, 12); + reference_white_level = dc_fixpt_pow(dc_fixpt_from_fraction(3, 1), dc_fixpt_half); + } + if (dc_fixpt_lt(threshold, in_x)) + *out_y = dc_fixpt_add(c, dc_fixpt_mul(a, dc_fixpt_log(dc_fixpt_sub(in_x, b)))); + else + *out_y = dc_fixpt_mul(dc_fixpt_pow(in_x, dc_fixpt_half), reference_white_level); +} + +/*re gamma, linear to none linear*/ +static void compute_hlg_eotf(struct fixed31_32 in_x, bool is_light0_12, struct fixed31_32 *out_y) +{ + struct fixed31_32 a; + struct fixed31_32 b; + struct fixed31_32 c; + struct fixed31_32 reference_white_level; + + a = dc_fixpt_from_fraction(17883277, 100000000); + if (is_light0_12) { + /*light 0-12*/ + b = dc_fixpt_from_fraction(28466892, 100000000); + c = dc_fixpt_from_fraction(55991073, 100000000); + reference_white_level = dc_fixpt_from_fraction(4, 1); + } else { + /*light 0-1*/ + b = dc_fixpt_from_fraction(2372241, 100000000); + c = dc_fixpt_add(dc_fixpt_one, dc_fixpt_from_fraction(429347, 100000000)); + reference_white_level = dc_fixpt_from_fraction(1, 3); + } + if (dc_fixpt_lt(dc_fixpt_half, in_x)) + *out_y = dc_fixpt_add(dc_fixpt_exp(dc_fixpt_div(dc_fixpt_sub(in_x, c), a)), b); + else + *out_y = dc_fixpt_mul(dc_fixpt_pow(in_x, dc_fixpt_from_fraction(2, 1)), reference_white_level); +} + + /* one-time pre-compute PQ values - only for sdr_white_level 80 */ void precompute_pq(void) { @@ -691,6 +748,48 @@ static void build_degamma(struct pwl_float_data_ex *curve, } } +static void build_hlg_degamma(struct pwl_float_data_ex *degamma, + uint32_t hw_points_num, + const struct hw_x_point *coordinate_x, bool is_light0_12) +{ + uint32_t i; + + struct pwl_float_data_ex *rgb = degamma; + const struct hw_x_point *coord_x = coordinate_x; + + i = 0; + + while (i != hw_points_num + 1) { + compute_hlg_oetf(coord_x->x, is_light0_12, &rgb->r); + rgb->g = rgb->r; + rgb->b = rgb->r; + ++coord_x; + ++rgb; + ++i; + } +} + +static void build_hlg_regamma(struct pwl_float_data_ex *regamma, + uint32_t hw_points_num, + const struct hw_x_point *coordinate_x, bool is_light0_12) +{ + uint32_t i; + + struct pwl_float_data_ex *rgb = regamma; + const struct hw_x_point *coord_x = coordinate_x; + + i = 0; + + while (i != hw_points_num + 1) { + compute_hlg_eotf(coord_x->x, is_light0_12, &rgb->r); + rgb->g = rgb->r; + rgb->b = rgb->r; + ++coord_x; + ++rgb; + ++i; + } +} + static void scale_gamma(struct pwl_float_data *pwl_rgb, const struct dc_gamma *ramp, struct dividers dividers) @@ -898,7 +997,9 @@ static void scale_user_regamma_ramp(struct pwl_float_data *pwl_rgb, * norm_y = 4095*regamma_y, and index is just truncating to nearest integer * lut1 = lut1D[index], lut2 = lut1D[index+1] * - *adjustedY is then linearly interpolating regamma Y between lut1 and lut2 + * adjustedY is then linearly interpolating regamma Y between lut1 and lut2 + * + * Custom degamma on Linux uses the same interpolation math, so is handled here */ static void apply_lut_1d( const struct dc_gamma *ramp, @@ -919,7 +1020,7 @@ static void apply_lut_1d( struct fixed31_32 delta_lut; struct fixed31_32 delta_index; - if (ramp->type != GAMMA_CS_TFM_1D) + if (ramp->type != GAMMA_CS_TFM_1D && ramp->type != GAMMA_CUSTOM) return; // this is not expected for (i = 0; i < num_hw_points; i++) { @@ -1537,7 +1638,9 @@ bool mod_color_calculate_degamma_params(struct dc_transfer_func *input_tf, map_regamma_hw_to_x_user(ramp, coeff, rgb_user, coordinates_x, axix_x, curve, MAX_HW_POINTS, tf_pts, - mapUserRamp); + mapUserRamp && ramp->type != GAMMA_CUSTOM); + if (ramp->type == GAMMA_CUSTOM) + apply_lut_1d(ramp, MAX_HW_POINTS, tf_pts); ret = true; @@ -1622,6 +1725,25 @@ bool mod_color_calculate_curve(enum dc_transfer_func_predefined trans, ret = true; kvfree(rgb_regamma); + } else if (trans == TRANSFER_FUNCTION_HLG || + trans == TRANSFER_FUNCTION_HLG12) { + rgb_regamma = kvcalloc(MAX_HW_POINTS + _EXTRA_POINTS, + sizeof(*rgb_regamma), + GFP_KERNEL); + if (!rgb_regamma) + goto rgb_regamma_alloc_fail; + + build_hlg_regamma(rgb_regamma, + MAX_HW_POINTS, + coordinates_x, + trans == TRANSFER_FUNCTION_HLG12 ? true:false); + for (i = 0; i <= MAX_HW_POINTS ; i++) { + points->red[i] = rgb_regamma[i].r; + points->green[i] = rgb_regamma[i].g; + points->blue[i] = rgb_regamma[i].b; + } + ret = true; + kvfree(rgb_regamma); } rgb_regamma_alloc_fail: return ret; @@ -1682,6 +1804,25 @@ bool mod_color_calculate_degamma_curve(enum dc_transfer_func_predefined trans, ret = true; kvfree(rgb_degamma); + } else if (trans == TRANSFER_FUNCTION_HLG || + trans == TRANSFER_FUNCTION_HLG12) { + rgb_degamma = kvcalloc(MAX_HW_POINTS + _EXTRA_POINTS, + sizeof(*rgb_degamma), + GFP_KERNEL); + if (!rgb_degamma) + goto rgb_degamma_alloc_fail; + + build_hlg_degamma(rgb_degamma, + MAX_HW_POINTS, + coordinates_x, + trans == TRANSFER_FUNCTION_HLG12 ? true:false); + for (i = 0; i <= MAX_HW_POINTS ; i++) { + points->red[i] = rgb_degamma[i].r; + points->green[i] = rgb_degamma[i].g; + points->blue[i] = rgb_degamma[i].b; + } + ret = true; + kvfree(rgb_degamma); } points->end_exponent = 0; points->x_point_at_y1_red = 1; diff --git a/drivers/gpu/drm/amd/display/modules/color/luts_1d.h b/drivers/gpu/drm/amd/display/modules/color/luts_1d.h new file mode 100644 index 000000000000..66b1fad572ac --- /dev/null +++ b/drivers/gpu/drm/amd/display/modules/color/luts_1d.h @@ -0,0 +1,51 @@ +/* + * Copyright 2016 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: AMD + * + */ +#ifndef LUTS_1D_H +#define LUTS_1D_H + +#include "hw_shared.h" + +struct point_config { + uint32_t custom_float_x; + uint32_t custom_float_y; + uint32_t custom_float_slope; +}; + +struct lut_point { + uint32_t red; + uint32_t green; + uint32_t blue; + uint32_t delta_red; + uint32_t delta_green; + uint32_t delta_blue; +}; + +struct pwl_1dlut_parameter { + struct gamma_curve arr_curve_points[34]; + struct point_config arr_points[2]; + struct lut_point rgb_resulted[256]; + uint32_t hw_points_num; +}; +#endif // LUTS_1D_H diff --git a/drivers/gpu/drm/amd/display/modules/stats/stats.c b/drivers/gpu/drm/amd/display/modules/stats/stats.c index 710852ad03f3..3d4c1b1ab8c4 100644 --- a/drivers/gpu/drm/amd/display/modules/stats/stats.c +++ b/drivers/gpu/drm/amd/display/modules/stats/stats.c @@ -29,7 +29,7 @@ #include "core_types.h" #define DAL_STATS_ENABLE_REGKEY "DalStatsEnable" -#define DAL_STATS_ENABLE_REGKEY_DEFAULT 0x00000001 +#define DAL_STATS_ENABLE_REGKEY_DEFAULT 0x00000000 #define DAL_STATS_ENABLE_REGKEY_ENABLED 0x00000001 #define DAL_STATS_ENTRIES_REGKEY "DalStatsEntries" @@ -238,7 +238,7 @@ void mod_stats_dump(struct mod_stats *mod_stats) for (int i = 0; i < core_stats->entry_id; i++) { if (event_index < core_stats->event_index && i == events[event_index].entry_id) { - DISPLAY_STATS("%s\n", events[event_index].event_string); + DISPLAY_STATS("==Event==%s\n", events[event_index].event_string); event_index++; } else if (time_index < core_stats->index && i == time[time_index].entry_id) { diff --git a/drivers/gpu/drm/amd/include/amd_pcie.h b/drivers/gpu/drm/amd/include/amd_pcie.h index 5eb895fd98bf..9cb9ceb4d74d 100644 --- a/drivers/gpu/drm/amd/include/amd_pcie.h +++ b/drivers/gpu/drm/amd/include/amd_pcie.h @@ -27,6 +27,7 @@ #define CAIL_PCIE_LINK_SPEED_SUPPORT_GEN1 0x00010000 #define CAIL_PCIE_LINK_SPEED_SUPPORT_GEN2 0x00020000 #define CAIL_PCIE_LINK_SPEED_SUPPORT_GEN3 0x00040000 +#define CAIL_PCIE_LINK_SPEED_SUPPORT_GEN4 0x00080000 #define CAIL_PCIE_LINK_SPEED_SUPPORT_MASK 0xFFFF0000 #define CAIL_PCIE_LINK_SPEED_SUPPORT_SHIFT 16 @@ -34,6 +35,7 @@ #define CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN1 0x00000001 #define CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN2 0x00000002 #define CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN3 0x00000004 +#define CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN4 0x00000008 #define CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_MASK 0x0000FFFF #define CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_SHIFT 0 diff --git a/drivers/gpu/drm/amd/include/amd_shared.h b/drivers/gpu/drm/amd/include/amd_shared.h index b178176b72ac..265621d8945c 100644 --- a/drivers/gpu/drm/amd/include/amd_shared.h +++ b/drivers/gpu/drm/amd/include/amd_shared.h @@ -128,47 +128,57 @@ enum PP_FEATURE_MASK { PP_OVERDRIVE_MASK = 0x4000, PP_GFXOFF_MASK = 0x8000, PP_ACG_MASK = 0x10000, + PP_STUTTER_MODE = 0x20000, }; +/** + * struct amd_ip_funcs - general hooks for managing amdgpu IP Blocks + */ struct amd_ip_funcs { - /* Name of IP block */ + /** @name: Name of IP block */ char *name; - /* sets up early driver state (pre sw_init), does not configure hw - Optional */ + /** + * @early_init: + * + * sets up early driver state (pre sw_init), + * does not configure hw - Optional + */ int (*early_init)(void *handle); - /* sets up late driver/hw state (post hw_init) - Optional */ + /** @late_init: sets up late driver/hw state (post hw_init) - Optional */ int (*late_init)(void *handle); - /* sets up driver state, does not configure hw */ + /** @sw_init: sets up driver state, does not configure hw */ int (*sw_init)(void *handle); - /* tears down driver state, does not configure hw */ + /** @sw_fini: tears down driver state, does not configure hw */ int (*sw_fini)(void *handle); - /* sets up the hw state */ + /** @hw_init: sets up the hw state */ int (*hw_init)(void *handle); - /* tears down the hw state */ + /** @hw_fini: tears down the hw state */ int (*hw_fini)(void *handle); + /** @late_fini: final cleanup */ void (*late_fini)(void *handle); - /* handles IP specific hw/sw changes for suspend */ + /** @suspend: handles IP specific hw/sw changes for suspend */ int (*suspend)(void *handle); - /* handles IP specific hw/sw changes for resume */ + /** @resume: handles IP specific hw/sw changes for resume */ int (*resume)(void *handle); - /* returns current IP block idle status */ + /** @is_idle: returns current IP block idle status */ bool (*is_idle)(void *handle); - /* poll for idle */ + /** @wait_for_idle: poll for idle */ int (*wait_for_idle)(void *handle); - /* check soft reset the IP block */ + /** @check_soft_reset: check soft reset the IP block */ bool (*check_soft_reset)(void *handle); - /* pre soft reset the IP block */ + /** @pre_soft_reset: pre soft reset the IP block */ int (*pre_soft_reset)(void *handle); - /* soft reset the IP block */ + /** @soft_reset: soft reset the IP block */ int (*soft_reset)(void *handle); - /* post soft reset the IP block */ + /** @post_soft_reset: post soft reset the IP block */ int (*post_soft_reset)(void *handle); - /* enable/disable cg for the IP block */ + /** @set_clockgating_state: enable/disable cg for the IP block */ int (*set_clockgating_state)(void *handle, enum amd_clockgating_state state); - /* enable/disable pg for the IP block */ + /** @set_powergating_state: enable/disable pg for the IP block */ int (*set_powergating_state)(void *handle, enum amd_powergating_state state); - /* get current clockgating status */ + /** @get_clockgating_state: get current clockgating status */ void (*get_clockgating_state)(void *handle, u32 *flags); }; diff --git a/drivers/gpu/drm/amd/include/asic_reg/vcn/vcn_1_0_offset.h b/drivers/gpu/drm/amd/include/asic_reg/vcn/vcn_1_0_offset.h index 18a32477ed1d..fe0cbaade3c3 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/vcn/vcn_1_0_offset.h +++ b/drivers/gpu/drm/amd/include/asic_reg/vcn/vcn_1_0_offset.h @@ -89,6 +89,8 @@ #define mmUVD_JPEG_RB_SIZE_BASE_IDX 1 #define mmUVD_JPEG_ADDR_CONFIG 0x021f #define mmUVD_JPEG_ADDR_CONFIG_BASE_IDX 1 +#define mmUVD_JPEG_PITCH 0x0222 +#define mmUVD_JPEG_PITCH_BASE_IDX 1 #define mmUVD_JPEG_GPCOM_CMD 0x022c #define mmUVD_JPEG_GPCOM_CMD_BASE_IDX 1 #define mmUVD_JPEG_GPCOM_DATA0 0x022d @@ -203,6 +205,8 @@ #define mmUVD_RB_WPTR4_BASE_IDX 1 #define mmUVD_JRBC_RB_RPTR 0x0457 #define mmUVD_JRBC_RB_RPTR_BASE_IDX 1 +#define mmUVD_LMI_JPEG_VMID 0x045d +#define mmUVD_LMI_JPEG_VMID_BASE_IDX 1 #define mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH 0x045e #define mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH_BASE_IDX 1 #define mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW 0x045f @@ -231,6 +235,8 @@ #define mmUVD_LMI_JRBC_IB_64BIT_BAR_HIGH_BASE_IDX 1 #define mmUVD_LMI_JRBC_IB_VMID 0x0507 #define mmUVD_LMI_JRBC_IB_VMID_BASE_IDX 1 +#define mmUVD_LMI_JRBC_RB_VMID 0x0508 +#define mmUVD_LMI_JRBC_RB_VMID_BASE_IDX 1 #define mmUVD_JRBC_RB_WPTR 0x0509 #define mmUVD_JRBC_RB_WPTR_BASE_IDX 1 #define mmUVD_JRBC_RB_CNTL 0x050a @@ -239,6 +245,20 @@ #define mmUVD_JRBC_IB_SIZE_BASE_IDX 1 #define mmUVD_JRBC_LMI_SWAP_CNTL 0x050d #define mmUVD_JRBC_LMI_SWAP_CNTL_BASE_IDX 1 +#define mmUVD_LMI_JRBC_RB_MEM_WR_64BIT_BAR_LOW 0x050e +#define mmUVD_LMI_JRBC_RB_MEM_WR_64BIT_BAR_LOW_BASE_IDX 1 +#define mmUVD_LMI_JRBC_RB_MEM_WR_64BIT_BAR_HIGH 0x050f +#define mmUVD_LMI_JRBC_RB_MEM_WR_64BIT_BAR_HIGH_BASE_IDX 1 +#define mmUVD_LMI_JRBC_RB_MEM_RD_64BIT_BAR_LOW 0x0510 +#define mmUVD_LMI_JRBC_RB_MEM_RD_64BIT_BAR_LOW_BASE_IDX 1 +#define mmUVD_LMI_JRBC_RB_MEM_RD_64BIT_BAR_HIGH 0x0511 +#define mmUVD_LMI_JRBC_RB_MEM_RD_64BIT_BAR_HIGH_BASE_IDX 1 +#define mmUVD_JRBC_RB_REF_DATA 0x0512 +#define mmUVD_JRBC_RB_REF_DATA_BASE_IDX 1 +#define mmUVD_JRBC_RB_COND_RD_TIMER 0x0513 +#define mmUVD_JRBC_RB_COND_RD_TIMER_BASE_IDX 1 +#define mmUVD_JRBC_EXTERNAL_REG_BASE 0x0517 +#define mmUVD_JRBC_EXTERNAL_REG_BASE_BASE_IDX 1 #define mmUVD_JRBC_SOFT_RESET 0x0519 #define mmUVD_JRBC_SOFT_RESET_BASE_IDX 1 #define mmUVD_JRBC_STATUS 0x051a diff --git a/drivers/gpu/drm/amd/include/atomfirmware.h b/drivers/gpu/drm/amd/include/atomfirmware.h index 33b4de4ad66e..4bc118df3bc4 100644 --- a/drivers/gpu/drm/amd/include/atomfirmware.h +++ b/drivers/gpu/drm/amd/include/atomfirmware.h @@ -1074,7 +1074,7 @@ struct atom_integrated_system_info_v1_11 uint16_t dpphy_override; // bit vector, enum of atom_sysinfo_dpphy_override_def uint16_t lvds_misc; // enum of atom_sys_info_lvds_misc_def uint16_t backlight_pwm_hz; // pwm frequency in hz - uint8_t memorytype; // enum of atom_sys_mem_type + uint8_t memorytype; // enum of atom_dmi_t17_mem_type_def, APU memory type indication. uint8_t umachannelnumber; // number of memory channels uint8_t pwr_on_digon_to_de; /* all pwr sequence numbers below are in uint of 4ms */ uint8_t pwr_on_de_to_vary_bl; @@ -1084,18 +1084,25 @@ struct atom_integrated_system_info_v1_11 uint8_t pwr_on_vary_bl_to_blon; uint8_t pwr_down_bloff_to_vary_bloff; uint8_t min_allowed_bl_level; + uint8_t htc_hyst_limit; + uint8_t htc_tmp_limit; + uint8_t reserved1; + uint8_t reserved2; struct atom_external_display_connection_info extdispconninfo; struct atom_14nm_dpphy_dvihdmi_tuningset dvi_tuningset; struct atom_14nm_dpphy_dvihdmi_tuningset hdmi_tuningset; struct atom_14nm_dpphy_dvihdmi_tuningset hdmi6g_tuningset; - struct atom_14nm_dpphy_dp_tuningset dp_tuningset; - struct atom_14nm_dpphy_dp_tuningset dp_hbr3_tuningset; + struct atom_14nm_dpphy_dp_tuningset dp_tuningset; // rbr 1.62G dp tuning set + struct atom_14nm_dpphy_dp_tuningset dp_hbr3_tuningset; // HBR3 dp tuning set struct atom_camera_data camera_info; struct atom_hdmi_retimer_redriver_set dp0_retimer_set; //for DP0 struct atom_hdmi_retimer_redriver_set dp1_retimer_set; //for DP1 struct atom_hdmi_retimer_redriver_set dp2_retimer_set; //for DP2 struct atom_hdmi_retimer_redriver_set dp3_retimer_set; //for DP3 - uint32_t reserved[108]; + struct atom_14nm_dpphy_dp_tuningset dp_hbr_tuningset; //hbr 2.7G dp tuning set + struct atom_14nm_dpphy_dp_tuningset dp_hbr2_tuningset; //hbr2 5.4G dp turnig set + struct atom_14nm_dpphy_dp_tuningset edp_tuningset; //edp tuning set + uint32_t reserved[66]; }; diff --git a/drivers/gpu/drm/amd/include/dm_pp_interface.h b/drivers/gpu/drm/amd/include/dm_pp_interface.h index 7852952d1fde..1d93a0c574c9 100644 --- a/drivers/gpu/drm/amd/include/dm_pp_interface.h +++ b/drivers/gpu/drm/amd/include/dm_pp_interface.h @@ -23,6 +23,8 @@ #ifndef _DM_PP_INTERFACE_ #define _DM_PP_INTERFACE_ +#include "dm_services_types.h" + #define PP_MAX_CLOCK_LEVELS 16 enum amd_pp_display_config_type{ @@ -189,39 +191,4 @@ struct pp_display_clock_request { uint32_t clock_freq_in_khz; }; -#define PP_MAX_WM_SETS 4 - -enum pp_wm_set_id { - DC_WM_SET_A = 0, - DC_WM_SET_B, - DC_WM_SET_C, - DC_WM_SET_D, - DC_WM_SET_INVALID = 0xffff, -}; - -struct pp_wm_set_with_dmif_clock_range_soc15 { - enum pp_wm_set_id wm_set_id; - uint32_t wm_min_dcefclk_in_khz; - uint32_t wm_max_dcefclk_in_khz; - uint32_t wm_min_memclk_in_khz; - uint32_t wm_max_memclk_in_khz; -}; - -struct pp_wm_set_with_mcif_clock_range_soc15 { - enum pp_wm_set_id wm_set_id; - uint32_t wm_min_socclk_in_khz; - uint32_t wm_max_socclk_in_khz; - uint32_t wm_min_memclk_in_khz; - uint32_t wm_max_memclk_in_khz; -}; - -struct pp_wm_sets_with_clock_ranges_soc15 { - uint32_t num_wm_sets_dmif; - uint32_t num_wm_sets_mcif; - struct pp_wm_set_with_dmif_clock_range_soc15 - wm_sets_dmif[PP_MAX_WM_SETS]; - struct pp_wm_set_with_mcif_clock_range_soc15 - wm_sets_mcif[PP_MAX_WM_SETS]; -}; - #endif /* _DM_PP_INTERFACE_ */ diff --git a/drivers/gpu/drm/amd/include/ivsrcid/gfx/irqsrcs_gfx_9_0.h b/drivers/gpu/drm/amd/include/ivsrcid/gfx/irqsrcs_gfx_9_0.h new file mode 100644 index 000000000000..36306c57a2b4 --- /dev/null +++ b/drivers/gpu/drm/amd/include/ivsrcid/gfx/irqsrcs_gfx_9_0.h @@ -0,0 +1,55 @@ +/* + * 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. + * + * Authors: AMD + * + */ + +#ifndef __IRQSRCS_GFX_9_0_H__ +#define __IRQSRCS_GFX_9_0_H__ + + +#define GFX_9_0__SRCID__CP_RB_INTERRUPT_PKT 176 /* B0 CP_INTERRUPT pkt in RB */ +#define GFX_9_0__SRCID__CP_IB1_INTERRUPT_PKT 177 /* B1 CP_INTERRUPT pkt in IB1 */ +#define GFX_9_0__SRCID__CP_IB2_INTERRUPT_PKT 178 /* B2 CP_INTERRUPT pkt in IB2 */ +#define GFX_9_0__SRCID__CP_PM4_PKT_RSVD_BIT_ERROR 180 /* B4 PM4 Pkt Rsvd Bits Error */ +#define GFX_9_0__SRCID__CP_EOP_INTERRUPT 181 /* B5 End-of-Pipe Interrupt */ +#define GFX_9_0__SRCID__CP_BAD_OPCODE_ERROR 183 /* B7 Bad Opcode Error */ +#define GFX_9_0__SRCID__CP_PRIV_REG_FAULT 184 /* B8 Privileged Register Fault */ +#define GFX_9_0__SRCID__CP_PRIV_INSTR_FAULT 185 /* B9 Privileged Instr Fault */ +#define GFX_9_0__SRCID__CP_WAIT_MEM_SEM_FAULT 186 /* BA Wait Memory Semaphore Fault (Synchronization Object Fault) */ +#define GFX_9_0__SRCID__CP_CTX_EMPTY_INTERRUPT 187 /* BB Context Empty Interrupt */ +#define GFX_9_0__SRCID__CP_CTX_BUSY_INTERRUPT 188 /* BC Context Busy Interrupt */ +#define GFX_9_0__SRCID__CP_ME_WAIT_REG_MEM_POLL_TIMEOUT 192 /* C0 CP.ME Wait_Reg_Mem Poll Timeout */ +#define GFX_9_0__SRCID__CP_SIG_INCOMPLETE 193 /* C1 "Surface Probe Fault Signal Incomplete" */ +#define GFX_9_0__SRCID__CP_PREEMPT_ACK 194 /* C2 Preemption Ack-wledge */ +#define GFX_9_0__SRCID__CP_GPF 195 /* C3 General Protection Fault (GPF) */ +#define GFX_9_0__SRCID__CP_GDS_ALLOC_ERROR 196 /* C4 GDS Alloc Error */ +#define GFX_9_0__SRCID__CP_ECC_ERROR 197 /* C5 ECC Error */ +#define GFX_9_0__SRCID__CP_COMPUTE_QUERY_STATUS 199 /* C7 Compute query status */ +#define GFX_9_0__SRCID__CP_VM_DOORBELL 200 /* C8 Unattached VM Doorbell Received */ +#define GFX_9_0__SRCID__CP_FUE_ERROR 201 /* C9 ECC FUE Error */ +#define GFX_9_0__SRCID__RLC_STRM_PERF_MONITOR_INTERRUPT 202 /* CA Streaming Perf Monitor Interrupt */ +#define GFX_9_0__SRCID__GRBM_RD_TIMEOUT_ERROR 232 /* E8 CRead timeout error */ +#define GFX_9_0__SRCID__GRBM_REG_GUI_IDLE 233 /* E9 Register GUI Idle */ +#define GFX_9_0__SRCID__SQ_INTERRUPT_ID 239 /* EF SQ Interrupt (ttrace wrap, errors) */ + +#endif /* __IRQSRCS_GFX_9_0_H__ */ diff --git a/drivers/gpu/drm/amd/include/ivsrcid/ivsrcid_vislands30.h b/drivers/gpu/drm/amd/include/ivsrcid/ivsrcid_vislands30.h index c6b6f97de9de..aaed7f59e0e2 100644 --- a/drivers/gpu/drm/amd/include/ivsrcid/ivsrcid_vislands30.h +++ b/drivers/gpu/drm/amd/include/ivsrcid/ivsrcid_vislands30.h @@ -198,4 +198,102 @@ #define VISLANDS30_IV_SRCID_HPD_RX_F 42 // 0x2a #define VISLANDS30_IV_EXTID_HPD_RX_F 11 +#define VISLANDS30_IV_SRCID_GPIO_19 0x00000053 /* 83 */ + +#define VISLANDS30_IV_SRCID_SRBM_READ_TIMEOUT_ERR 0x00000060 /* 96 */ +#define VISLANDS30_IV_SRCID_SRBM_CTX_SWITCH 0x00000061 /* 97 */ + +#define VISLANDS30_IV_SRBM_REG_ACCESS_ERROR 0x00000062 /* 98 */ + + +#define VISLANDS30_IV_SRCID_UVD_ENC_GEN_PURP 0x00000077 /* 119 */ +#define VISLANDS30_IV_SRCID_UVD_SYSTEM_MESSAGE 0x0000007c /* 124 */ + +#define VISLANDS30_IV_SRCID_BIF_PF_VF_MSGBUF_VALID 0x00000087 /* 135 */ + +#define VISLANDS30_IV_SRCID_BIF_VF_PF_MSGBUF_ACK 0x0000008a /* 138 */ + +#define VISLANDS30_IV_SRCID_SYS_PAGE_INV_FAULT 0x0000008c /* 140 */ +#define VISLANDS30_IV_SRCID_SYS_MEM_PROT_FAULT 0x0000008d /* 141 */ + +#define VISLANDS30_IV_SRCID_SEM_PAGE_INV_FAULT 0x00000090 /* 144 */ +#define VISLANDS30_IV_SRCID_SEM_MEM_PROT_FAULT 0x00000091 /* 145 */ + +#define VISLANDS30_IV_SRCID_GFX_PAGE_INV_FAULT 0x00000092 /* 146 */ +#define VISLANDS30_IV_SRCID_GFX_MEM_PROT_FAULT 0x00000093 /* 147 */ + +#define VISLANDS30_IV_SRCID_ACP 0x000000a2 /* 162 */ + +#define VISLANDS30_IV_SRCID_VCE_TRAP 0x000000a7 /* 167 */ +#define VISLANDS30_IV_EXTID_VCE_TRAP_GENERAL_PURPOSE 0 +#define VISLANDS30_IV_EXTID_VCE_TRAP_LOW_LATENCY 1 +#define VISLANDS30_IV_EXTID_VCE_TRAP_REAL_TIME 2 + +#define VISLANDS30_IV_SRCID_CP_INT_RB 0x000000b0 /* 176 */ +#define VISLANDS30_IV_SRCID_CP_INT_IB1 0x000000b1 /* 177 */ +#define VISLANDS30_IV_SRCID_CP_INT_IB2 0x000000b2 /* 178 */ +#define VISLANDS30_IV_SRCID_CP_PM4_RES_BITS_ERR 0x000000b4 /* 180 */ +#define VISLANDS30_IV_SRCID_CP_END_OF_PIPE 0x000000b5 /* 181 */ +#define VISLANDS30_IV_SRCID_CP_BAD_OPCODE 0x000000b7 /* 183 */ +#define VISLANDS30_IV_SRCID_CP_PRIV_REG_FAULT 0x000000b8 /* 184 */ +#define VISLANDS30_IV_SRCID_CP_PRIV_INSTR_FAULT 0x000000b9 /* 185 */ +#define VISLANDS30_IV_SRCID_CP_WAIT_MEM_SEM_FAULT 0x000000ba /* 186 */ +#define VISLANDS30_IV_SRCID_CP_GUI_IDLE 0x000000bb /* 187 */ +#define VISLANDS30_IV_SRCID_CP_GUI_BUSY 0x000000bc /* 188 */ + +#define VISLANDS30_IV_SRCID_CP_COMPUTE_QUERY_STATUS 0x000000bf /* 191 */ +#define VISLANDS30_IV_SRCID_CP_ECC_ERROR 0x000000c5 /* 197 */ + +#define CARRIZO_IV_SRCID_CP_COMPUTE_QUERY_STATUS 0x000000c7 /* 199 */ + +#define VISLANDS30_IV_SRCID_CP_WAIT_REG_MEM_POLL_TIMEOUT 0x000000c0 /* 192 */ +#define VISLANDS30_IV_SRCID_CP_SEM_SIG_INCOMPL 0x000000c1 /* 193 */ +#define VISLANDS30_IV_SRCID_CP_PREEMPT_ACK 0x000000c2 /* 194 */ +#define VISLANDS30_IV_SRCID_CP_GENERAL_PROT_FAULT 0x000000c3 /* 195 */ +#define VISLANDS30_IV_SRCID_CP_GDS_ALLOC_ERROR 0x000000c4 /* 196 */ +#define VISLANDS30_IV_SRCID_CP_ECC_ERROR 0x000000c5 /* 197 */ + +#define VISLANDS30_IV_SRCID_RLC_STRM_PERF_MONITOR 0x000000ca /* 202 */ + +#define VISLANDS30_IV_SDMA_ATOMIC_SRC_ID 0x000000da /* 218 */ + +#define VISLANDS30_IV_SRCID_SDMA_ECC_ERROR 0x000000dc /* 220 */ + +#define VISLANDS30_IV_SRCID_SDMA_TRAP 0x000000e0 /* 224 */ +#define VISLANDS30_IV_SRCID_SDMA_SEM_INCOMPLETE 0x000000e1 /* 225 */ +#define VISLANDS30_IV_SRCID_SDMA_SEM_WAIT 0x000000e2 /* 226 */ + + +#define VISLANDS30_IV_SRCID_SMU_DISP_TIMER2_TRIGGER 0x000000e5 /* 229 */ + +#define VISLANDS30_IV_SRCID_CG_TSS_THERMAL_LOW_TO_HIGH 0x000000e6 /* 230 */ +#define VISLANDS30_IV_SRCID_CG_TSS_THERMAL_HIGH_TO_LOW 0x000000e7 /* 231 */ + +#define VISLANDS30_IV_SRCID_GRBM_READ_TIMEOUT_ERR 0x000000e8 /* 232 */ +#define VISLANDS30_IV_SRCID_GRBM_REG_GUI_IDLE 0x000000e9 /* 233 */ + +#define VISLANDS30_IV_SRCID_SQ_INTERRUPT_MSG 0x000000ef /* 239 */ + +#define VISLANDS30_IV_SRCID_SDMA_PREEMPT 0x000000f0 /* 240 */ +#define VISLANDS30_IV_SRCID_SDMA_VM_HOLE 0x000000f2 /* 242 */ +#define VISLANDS30_IV_SRCID_SDMA_CTXEMPTY 0x000000f3 /* 243 */ +#define VISLANDS30_IV_SRCID_SDMA_DOORBELL_INVALID 0x000000f4 /* 244 */ +#define VISLANDS30_IV_SRCID_SDMA_FROZEN 0x000000f5 /* 245 */ +#define VISLANDS30_IV_SRCID_SDMA_POLL_TIMEOUT 0x000000f6 /* 246 */ +#define VISLANDS30_IV_SRCID_SDMA_SRBM_WRITE 0x000000f7 /* 247 */ + +#define VISLANDS30_IV_SRCID_CG_THERMAL_TRIG 0x000000f8 /* 248 */ + +#define VISLANDS30_IV_SRCID_SMU_DISP_TIMER_TRIGGER 0x000000fd /* 253 */ + +/* These are not "real" source ids defined by HW */ +#define VISLANDS30_IV_SRCID_VM_CONTEXT_ALL 0x00000100 /* 256 */ +#define VISLANDS30_IV_EXTID_VM_CONTEXT0_ALL 0 +#define VISLANDS30_IV_EXTID_VM_CONTEXT1_ALL 1 + + +/* IV Extended IDs */ +#define VISLANDS30_IV_EXTID_NONE 0x00000000 +#define VISLANDS30_IV_EXTID_INVALID 0xffffffff + #endif // _IVSRCID_VISLANDS30_H_ diff --git a/drivers/gpu/drm/amd/include/ivsrcid/sdma0/irqsrcs_sdma0_4_0.h b/drivers/gpu/drm/amd/include/ivsrcid/sdma0/irqsrcs_sdma0_4_0.h new file mode 100644 index 000000000000..802413832fe8 --- /dev/null +++ b/drivers/gpu/drm/amd/include/ivsrcid/sdma0/irqsrcs_sdma0_4_0.h @@ -0,0 +1,50 @@ +/* + * 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. + * + * Authors: AMD + * + */ + +#ifndef __IRQSRCS_SDMA0_4_0_H__ +#define __IRQSRCS_SDMA0_4_0_H__ + +#define SDMA0_4_0__SRCID__SDMA_ATOMIC_RTN_DONE 217 /* 0xD9 SDMA atomic*_rtn ops complete */ +#define SDMA0_4_0__SRCID__SDMA_ATOMIC_TIMEOUT 218 /* 0xDA SDMA atomic CMPSWAP loop timeout */ +#define SDMA0_4_0__SRCID__SDMA_IB_PREEMPT 219 /* 0xDB sdma mid-command buffer preempt interrupt */ +#define SDMA0_4_0__SRCID__SDMA_ECC 220 /* 0xDC ECC Error */ +#define SDMA0_4_0__SRCID__SDMA_PAGE_FAULT 221 /* 0xDD Page Fault Error from UTCL2 when nack=3 */ +#define SDMA0_4_0__SRCID__SDMA_PAGE_NULL 222 /* 0xDE Page Null from UTCL2 when nack=2 */ +#define SDMA0_4_0__SRCID__SDMA_XNACK 223 /* 0xDF Page retry timeout after UTCL2 return nack=1 */ +#define SDMA0_4_0__SRCID__SDMA_TRAP 224 /* 0xE0 Trap */ +#define SDMA0_4_0__SRCID__SDMA_SEM_INCOMPLETE_TIMEOUT 225 /* 0xE1 0xDAGPF (Sem incomplete timeout) */ +#define SDMA0_4_0__SRCID__SDMA_SEM_WAIT_FAIL_TIMEOUT 226 /* 0xE2 Semaphore wait fail timeout */ +#define SDMA0_4_0__SRCID__SDMA_SRAM_ECC 228 /* 0xE4 SRAM ECC Error */ +#define SDMA0_4_0__SRCID__SDMA_PREEMPT 240 /* 0xF0 SDMA New Run List */ +#define SDMA0_4_0__SRCID__SDMA_VM_HOLE 242 /* 0xF2 MC or SEM address in VM hole */ +#define SDMA0_4_0__SRCID__SDMA_CTXEMPTY 243 /* 0xF3 Context Empty */ +#define SDMA0_4_0__SRCID__SDMA_DOORBELL_INVALID 244 /* 0xF4 Doorbell BE invalid */ +#define SDMA0_4_0__SRCID__SDMA_FROZEN 245 /* 0xF5 SDMA Frozen */ +#define SDMA0_4_0__SRCID__SDMA_POLL_TIMEOUT 246 /* 0xF6 SRBM read poll timeout */ +#define SDMA0_4_0__SRCID__SDMA_SRBMWRITE 247 /* 0xF7 SRBM write Protection */ + +#endif /* __IRQSRCS_SDMA_4_0_H__ */ + + diff --git a/drivers/gpu/drm/amd/include/ivsrcid/sdma1/irqsrcs_sdma1_4_0.h b/drivers/gpu/drm/amd/include/ivsrcid/sdma1/irqsrcs_sdma1_4_0.h new file mode 100644 index 000000000000..d12a35619f9a --- /dev/null +++ b/drivers/gpu/drm/amd/include/ivsrcid/sdma1/irqsrcs_sdma1_4_0.h @@ -0,0 +1,50 @@ +/* + * 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. + * + * Authors: AMD + * + */ + +#ifndef __IRQSRCS_SDMA1_4_0_H__ +#define __IRQSRCS_SDMA1_4_0_H__ + +#define SDMA1_4_0__SRCID__SDMA_ATOMIC_RTN_DONE 217 /* 0xD9 SDMA atomic*_rtn ops complete */ +#define SDMA1_4_0__SRCID__SDMA_ATOMIC_TIMEOUT 218 /* 0xDA SDMA atomic CMPSWAP loop timeout */ +#define SDMA1_4_0__SRCID__SDMA_IB_PREEMPT 219 /* 0xDB sdma mid-command buffer preempt interrupt */ +#define SDMA1_4_0__SRCID__SDMA_ECC 220 /* 0xDC ECC Error */ +#define SDMA1_4_0__SRCID__SDMA_PAGE_FAULT 221 /* 0xDD Page Fault Error from UTCL2 when nack=3 */ +#define SDMA1_4_0__SRCID__SDMA_PAGE_NULL 222 /* 0xDE Page Null from UTCL2 when nack=2 */ +#define SDMA1_4_0__SRCID__SDMA_XNACK 223 /* 0xDF Page retry timeout after UTCL2 return nack=1 */ +#define SDMA1_4_0__SRCID__SDMA_TRAP 224 /* 0xE0 Trap */ +#define SDMA1_4_0__SRCID__SDMA_SEM_INCOMPLETE_TIMEOUT 225 /* 0xE1 0xDAGPF (Sem incomplete timeout) */ +#define SDMA1_4_0__SRCID__SDMA_SEM_WAIT_FAIL_TIMEOUT 226 /* 0xE2 Semaphore wait fail timeout */ +#define SDMA1_4_0__SRCID__SDMA_SRAM_ECC 228 /* 0xE4 SRAM ECC Error */ +#define SDMA1_4_0__SRCID__SDMA_PREEMPT 240 /* 0xF0 SDMA New Run List */ +#define SDMA1_4_0__SRCID__SDMA_VM_HOLE 242 /* 0xF2 MC or SEM address in VM hole */ +#define SDMA1_4_0__SRCID__SDMA_CTXEMPTY 243 /* 0xF3 Context Empty */ +#define SDMA1_4_0__SRCID__SDMA_DOORBELL_INVALID 244 /* 0xF4 Doorbell BE invalid */ +#define SDMA1_4_0__SRCID__SDMA_FROZEN 245 /* 0xF5 SDMA Frozen */ +#define SDMA1_4_0__SRCID__SDMA_POLL_TIMEOUT 246 /* 0xF6 SRBM read poll timeout */ +#define SDMA1_4_0__SRCID__SDMA_SRBMWRITE 247 /* 0xF7 SRBM write Protection */ + +#endif /* __IRQSRCS_SDMA1_4_0_H__ */ + + diff --git a/drivers/gpu/drm/amd/display/dc/dml/soc_bounding_box.h b/drivers/gpu/drm/amd/include/ivsrcid/smuio/irqsrcs_smuio_9_0.h index 7a65206a6d21..02bab4673cd4 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/soc_bounding_box.h +++ b/drivers/gpu/drm/amd/include/ivsrcid/smuio/irqsrcs_smuio_9_0.h @@ -23,13 +23,10 @@ * */ -#ifndef __SOC_BOUNDING_BOX_H__ -#define __SOC_BOUNDING_BOX_H__ +#ifndef __IRQSRCS_SMUIO_9_0_H__ +#define __IRQSRCS_SMUIO_9_0_H__ -#include "dml_common_defs.h" +#define SMUIO_9_0__SRCID__SMUIO_GPIO19 83 /* GPIO19 interrupt */ -void dml_socbb_set_latencies(soc_bounding_box_st *to_box, soc_bounding_box_st *from_box); -voltage_scaling_st dml_socbb_voltage_scaling(const soc_bounding_box_st *box, enum voltage_state voltage); -double dml_socbb_return_bw_mhz(soc_bounding_box_st *box, enum voltage_state voltage); +#endif /* __IRQSRCS_SMUIO_9_0_H__ */ -#endif diff --git a/drivers/gpu/drm/amd/include/ivsrcid/thm/irqsrcs_thm_9_0.h b/drivers/gpu/drm/amd/include/ivsrcid/thm/irqsrcs_thm_9_0.h new file mode 100644 index 000000000000..5218bc53fb2d --- /dev/null +++ b/drivers/gpu/drm/amd/include/ivsrcid/thm/irqsrcs_thm_9_0.h @@ -0,0 +1,33 @@ +/* + * 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. + * + * Authors: AMD + * + */ + +#ifndef __IRQSRCS_THM_9_0_H__ +#define __IRQSRCS_THM_9_0_H__ + +#define THM_9_0__SRCID__THM_DIG_THERM_L2H 0 /* ASIC_TEMP > CG_THERMAL_INT.DIG_THERM_INTH */ +#define THM_9_0__SRCID__THM_DIG_THERM_H2L 1 /* ASIC_TEMP < CG_THERMAL_INT.DIG_THERM_INTL */ + +#endif /* __IRQSRCS_THM_9_0_H__ */ + diff --git a/drivers/gpu/drm/amd/include/ivsrcid/uvd/irqsrcs_uvd_7_0.h b/drivers/gpu/drm/amd/include/ivsrcid/uvd/irqsrcs_uvd_7_0.h new file mode 100644 index 000000000000..fb041aee6c66 --- /dev/null +++ b/drivers/gpu/drm/amd/include/ivsrcid/uvd/irqsrcs_uvd_7_0.h @@ -0,0 +1,34 @@ +/* + * 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. + * + * Authors: AMD + * + */ + +#ifndef __IRQSRCS_UVD_7_0_H__ +#define __IRQSRCS_UVD_7_0_H__ + +#define UVD_7_0__SRCID__UVD_ENC_GEN_PURP 119 +#define UVD_7_0__SRCID__UVD_ENC_LOW_LATENCY 120 +#define UVD_7_0__SRCID__UVD_SYSTEM_MESSAGE_INTERRUPT 124 /* UVD system message interrupt */ + +#endif /* __IRQSRCS_UVD_7_0_H__ */ + diff --git a/drivers/gpu/drm/amd/include/ivsrcid/vce/irqsrcs_vce_4_0.h b/drivers/gpu/drm/amd/include/ivsrcid/vce/irqsrcs_vce_4_0.h new file mode 100644 index 000000000000..3440bab565af --- /dev/null +++ b/drivers/gpu/drm/amd/include/ivsrcid/vce/irqsrcs_vce_4_0.h @@ -0,0 +1,34 @@ +/* + * 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. + * + * Authors: AMD + * + */ + +#ifndef __IRQSRCS_VCE_4_0_H__ +#define __IRQSRCS_VCE_4_0_H__ + +#define VCE_4_0__CTXID__VCE_TRAP_GENERAL_PURPOSE 0 +#define VCE_4_0__CTXID__VCE_TRAP_LOW_LATENCY 1 +#define VCE_4_0__CTXID__VCE_TRAP_REAL_TIME 2 + +#endif /* __IRQSRCS_VCE_4_0_H__ */ + diff --git a/drivers/gpu/drm/amd/include/ivsrcid/vcn/irqsrcs_vcn_1_0.h b/drivers/gpu/drm/amd/include/ivsrcid/vcn/irqsrcs_vcn_1_0.h new file mode 100644 index 000000000000..e5951709bfc3 --- /dev/null +++ b/drivers/gpu/drm/amd/include/ivsrcid/vcn/irqsrcs_vcn_1_0.h @@ -0,0 +1,34 @@ +/* + * 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. + * + * Authors: AMD + * + */ + +#ifndef __IRQSRCS_VCN_1_0_H__ +#define __IRQSRCS_VCN_1_0_H__ + +#define VCN_1_0__SRCID__UVD_ENC_GENERAL_PURPOSE 119 /* 0x77 Encoder General Purpose */ +#define VCN_1_0__SRCID__UVD_ENC_LOW_LATENCY 120 /* 0x78 Encoder Low Latency */ +#define VCN_1_0__SRCID__UVD_SYSTEM_MESSAGE_INTERRUPT 124 /* 0x7c UVD system message interrupt */ + +#endif /* __IRQSRCS_VCN_1_0_H__ */ + diff --git a/drivers/gpu/drm/amd/include/ivsrcid/vmc/irqsrcs_vmc_1_0.h b/drivers/gpu/drm/amd/include/ivsrcid/vmc/irqsrcs_vmc_1_0.h new file mode 100644 index 000000000000..d130936c9989 --- /dev/null +++ b/drivers/gpu/drm/amd/include/ivsrcid/vmc/irqsrcs_vmc_1_0.h @@ -0,0 +1,37 @@ +/* + * 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. + * + * Authors: AMD + * + */ + +#ifndef __IRQSRCS_VMC_1_0_H__ +#define __IRQSRCS_VMC_1_0_H__ + + +#define VMC_1_0__SRCID__VM_FAULT 0 +#define VMC_1_0__SRCID__VM_CONTEXT0_ALL 256 +#define VMC_1_0__SRCID__VM_CONTEXT1_ALL 257 + +#define UTCL2_1_0__SRCID__FAULT 0 /* UTC L2 has encountered a fault or retry scenario */ + + +#endif /* __IRQSRCS_VMC_1_0_H__ */ diff --git a/drivers/gpu/drm/amd/include/kgd_kfd_interface.h b/drivers/gpu/drm/amd/include/kgd_kfd_interface.h index 5733fbee07f7..14391b06080c 100644 --- a/drivers/gpu/drm/amd/include/kgd_kfd_interface.h +++ b/drivers/gpu/drm/amd/include/kgd_kfd_interface.h @@ -47,6 +47,17 @@ enum kfd_preempt_type { KFD_PREEMPT_TYPE_WAVEFRONT_RESET, }; +struct kfd_vm_fault_info { + uint64_t page_addr; + uint32_t vmid; + uint32_t mc_id; + uint32_t status; + bool prot_valid; + bool prot_read; + bool prot_write; + bool prot_exec; +}; + struct kfd_cu_info { uint32_t num_shader_engines; uint32_t num_shader_arrays_per_engine; @@ -259,6 +270,21 @@ struct tile_config { * IB to the corresponding ring (ring type). The IB is executed with the * specified VMID in a user mode context. * + * @get_vm_fault_info: Return information about a recent VM fault on + * GFXv7 and v8. If multiple VM faults occurred since the last call of + * this function, it will return information about the first of those + * faults. On GFXv9 VM fault information is fully contained in the IH + * packet and this function is not needed. + * + * @read_vmid_from_vmfault_reg: On Hawaii the VMID is not set in the + * IH ring entry. This function allows the KFD ISR to get the VMID + * from the fault status register as early as possible. + * + * @gpu_recover: let kgd reset gpu after kfd detect CPC hang + * + * @set_compute_idle: Indicates that compute is idle on a device. This + * can be used to change power profiles depending on compute activity. + * * This structure contains function pointers to services that the kgd driver * provides to amdkfd driver. * @@ -374,6 +400,14 @@ struct kfd2kgd_calls { int (*submit_ib)(struct kgd_dev *kgd, enum kgd_engine_type engine, uint32_t vmid, uint64_t gpu_addr, uint32_t *ib_cmd, uint32_t ib_len); + + int (*get_vm_fault_info)(struct kgd_dev *kgd, + struct kfd_vm_fault_info *info); + uint32_t (*read_vmid_from_vmfault_reg)(struct kgd_dev *kgd); + + void (*gpu_recover)(struct kgd_dev *kgd); + + void (*set_compute_idle)(struct kgd_dev *kgd, bool idle); }; /** @@ -399,6 +433,10 @@ struct kfd2kgd_calls { * @schedule_evict_and_restore_process: Schedules work queue that will prepare * for safe eviction of KFD BOs that belong to the specified process. * + * @pre_reset: Notifies amdkfd that amdgpu about to reset the gpu + * + * @post_reset: Notify amdkfd that amgpu successfully reseted the gpu + * * This structure contains function callback pointers so the kgd driver * will notify to the amdkfd about certain status changes. * @@ -417,6 +455,8 @@ struct kgd2kfd_calls { int (*resume_mm)(struct mm_struct *mm); int (*schedule_evict_and_restore_process)(struct mm_struct *mm, struct dma_fence *fence); + int (*pre_reset)(struct kfd_dev *kfd); + int (*post_reset)(struct kfd_dev *kfd); }; int kgd2kfd_init(unsigned interface_version, diff --git a/drivers/gpu/drm/amd/include/kgd_pp_interface.h b/drivers/gpu/drm/amd/include/kgd_pp_interface.h index 06f08f34a110..6a41b81c7325 100644 --- a/drivers/gpu/drm/amd/include/kgd_pp_interface.h +++ b/drivers/gpu/drm/amd/include/kgd_pp_interface.h @@ -192,7 +192,6 @@ struct amd_pp_simple_clock_info; struct amd_pp_display_configuration; struct amd_pp_clock_info; struct pp_display_clock_request; -struct pp_wm_sets_with_clock_ranges_soc15; struct pp_clock_levels_with_voltage; struct pp_clock_levels_with_latency; struct amd_pp_clocks; @@ -232,16 +231,19 @@ struct amd_pm_funcs { void (*debugfs_print_current_performance_level)(void *handle, struct seq_file *m); int (*switch_power_profile)(void *handle, enum PP_SMC_POWER_PROFILE type, bool en); /* export to amdgpu */ - void (*powergate_uvd)(void *handle, bool gate); - void (*powergate_vce)(void *handle, bool gate); struct amd_vce_state *(*get_vce_clock_state)(void *handle, u32 idx); int (*dispatch_tasks)(void *handle, enum amd_pp_task task_id, enum amd_pm_state_type *user_state); int (*load_firmware)(void *handle); int (*wait_for_fw_loading_complete)(void *handle); + int (*set_powergating_by_smu)(void *handle, + uint32_t block_type, bool gate); int (*set_clockgating_by_smu)(void *handle, uint32_t msg_id); int (*set_power_limit)(void *handle, uint32_t n); int (*get_power_limit)(void *handle, uint32_t *limit, bool default_limit); + int (*get_power_profile_mode)(void *handle, char *buf); + int (*set_power_profile_mode)(void *handle, long *input, uint32_t size); + int (*odn_edit_dpm_table)(void *handle, uint32_t type, long *input, uint32_t size); /* export to DC */ u32 (*get_sclk)(void *handle, bool low); u32 (*get_mclk)(void *handle, bool low); @@ -261,15 +263,12 @@ struct amd_pm_funcs { enum amd_pp_clock_type type, struct pp_clock_levels_with_voltage *clocks); int (*set_watermarks_for_clocks_ranges)(void *handle, - struct pp_wm_sets_with_clock_ranges_soc15 *wm_with_clock_ranges); + void *clock_ranges); int (*display_clock_voltage_request)(void *handle, struct pp_display_clock_request *clock); int (*get_display_mode_validation_clocks)(void *handle, struct amd_pp_simple_clock_info *clocks); - int (*get_power_profile_mode)(void *handle, char *buf); - int (*set_power_profile_mode)(void *handle, long *input, uint32_t size); - int (*odn_edit_dpm_table)(void *handle, uint32_t type, long *input, uint32_t size); - int (*set_mmhub_powergating_by_smu)(void *handle); + int (*notify_smu_enable_pwe)(void *handle); }; #endif diff --git a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c index d567be49c31b..7a646f94b478 100644 --- a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c +++ b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c @@ -221,29 +221,7 @@ static int pp_sw_reset(void *handle) static int pp_set_powergating_state(void *handle, enum amd_powergating_state state) { - struct amdgpu_device *adev = handle; - struct pp_hwmgr *hwmgr = adev->powerplay.pp_handle; - int ret; - - if (!hwmgr || !hwmgr->pm_en) - return 0; - - if (hwmgr->hwmgr_func->gfx_off_control) { - /* Enable/disable GFX off through SMU */ - ret = hwmgr->hwmgr_func->gfx_off_control(hwmgr, - state == AMD_PG_STATE_GATE); - if (ret) - pr_err("gfx off control failed!\n"); - } - - if (hwmgr->hwmgr_func->enable_per_cu_power_gating == NULL) { - pr_debug("%s was not implemented.\n", __func__); - return 0; - } - - /* Enable/disable GFX per cu powergating through SMU */ - return hwmgr->hwmgr_func->enable_per_cu_power_gating(hwmgr, - state == AMD_PG_STATE_GATE); + return 0; } static int pp_suspend(void *handle) @@ -1020,7 +998,7 @@ static int pp_get_display_power_level(void *handle, static int pp_get_current_clocks(void *handle, struct amd_pp_clock_info *clocks) { - struct amd_pp_simple_clock_info simple_clocks; + struct amd_pp_simple_clock_info simple_clocks = { 0 }; struct pp_clock_info hw_clocks; struct pp_hwmgr *hwmgr = handle; int ret = 0; @@ -1056,7 +1034,10 @@ static int pp_get_current_clocks(void *handle, clocks->max_engine_clock_in_sr = hw_clocks.max_eng_clk; clocks->min_engine_clock_in_sr = hw_clocks.min_eng_clk; - clocks->max_clocks_state = simple_clocks.level; + if (simple_clocks.level == 0) + clocks->max_clocks_state = PP_DAL_POWERLEVEL_7; + else + clocks->max_clocks_state = simple_clocks.level; if (0 == phm_get_current_shallow_sleep_clocks(hwmgr, &hwmgr->current_ps->hardware, &hw_clocks)) { clocks->max_engine_clock_in_sr = hw_clocks.max_eng_clk; @@ -1118,17 +1099,17 @@ static int pp_get_clock_by_type_with_voltage(void *handle, } static int pp_set_watermarks_for_clocks_ranges(void *handle, - struct pp_wm_sets_with_clock_ranges_soc15 *wm_with_clock_ranges) + void *clock_ranges) { struct pp_hwmgr *hwmgr = handle; int ret = 0; - if (!hwmgr || !hwmgr->pm_en ||!wm_with_clock_ranges) + if (!hwmgr || !hwmgr->pm_en || !clock_ranges) return -EINVAL; mutex_lock(&hwmgr->smu_lock); ret = phm_set_watermarks_for_clocks_ranges(hwmgr, - wm_with_clock_ranges); + clock_ranges); mutex_unlock(&hwmgr->smu_lock); return ret; @@ -1159,6 +1140,8 @@ static int pp_get_display_mode_validation_clocks(void *handle, if (!hwmgr || !hwmgr->pm_en ||!clocks) return -EINVAL; + clocks->level = PP_DAL_POWERLEVEL_7; + mutex_lock(&hwmgr->smu_lock); if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DynamicPatchPowerState)) @@ -1168,19 +1151,78 @@ static int pp_get_display_mode_validation_clocks(void *handle, return ret; } -static int pp_set_mmhub_powergating_by_smu(void *handle) +static int pp_dpm_powergate_mmhub(void *handle) { struct pp_hwmgr *hwmgr = handle; if (!hwmgr || !hwmgr->pm_en) return -EINVAL; - if (hwmgr->hwmgr_func->set_mmhub_powergating_by_smu == NULL) { + if (hwmgr->hwmgr_func->powergate_mmhub == NULL) { + pr_info("%s was not implemented.\n", __func__); + return 0; + } + + return hwmgr->hwmgr_func->powergate_mmhub(hwmgr); +} + +static int pp_dpm_powergate_gfx(void *handle, bool gate) +{ + struct pp_hwmgr *hwmgr = handle; + + if (!hwmgr || !hwmgr->pm_en) + return 0; + + if (hwmgr->hwmgr_func->powergate_gfx == NULL) { pr_info("%s was not implemented.\n", __func__); return 0; } - return hwmgr->hwmgr_func->set_mmhub_powergating_by_smu(hwmgr); + return hwmgr->hwmgr_func->powergate_gfx(hwmgr, gate); +} + +static int pp_set_powergating_by_smu(void *handle, + uint32_t block_type, bool gate) +{ + int ret = 0; + + switch (block_type) { + case AMD_IP_BLOCK_TYPE_UVD: + case AMD_IP_BLOCK_TYPE_VCN: + pp_dpm_powergate_uvd(handle, gate); + break; + case AMD_IP_BLOCK_TYPE_VCE: + pp_dpm_powergate_vce(handle, gate); + break; + case AMD_IP_BLOCK_TYPE_GMC: + pp_dpm_powergate_mmhub(handle); + break; + case AMD_IP_BLOCK_TYPE_GFX: + ret = pp_dpm_powergate_gfx(handle, gate); + break; + default: + break; + } + return ret; +} + +static int pp_notify_smu_enable_pwe(void *handle) +{ + struct pp_hwmgr *hwmgr = handle; + + if (!hwmgr || !hwmgr->pm_en) + return -EINVAL; + + if (hwmgr->hwmgr_func->smus_notify_pwe == NULL) { + pr_info("%s was not implemented.\n", __func__); + return -EINVAL;; + } + + mutex_lock(&hwmgr->smu_lock); + hwmgr->hwmgr_func->smus_notify_pwe(hwmgr); + mutex_unlock(&hwmgr->smu_lock); + + return 0; } static const struct amd_pm_funcs pp_dpm_funcs = { @@ -1189,8 +1231,6 @@ static const struct amd_pm_funcs pp_dpm_funcs = { .force_performance_level = pp_dpm_force_performance_level, .get_performance_level = pp_dpm_get_performance_level, .get_current_power_state = pp_dpm_get_current_power_state, - .powergate_vce = pp_dpm_powergate_vce, - .powergate_uvd = pp_dpm_powergate_uvd, .dispatch_tasks = pp_dpm_dispatch_tasks, .set_fan_control_mode = pp_dpm_set_fan_control_mode, .get_fan_control_mode = pp_dpm_get_fan_control_mode, @@ -1210,6 +1250,7 @@ static const struct amd_pm_funcs pp_dpm_funcs = { .get_vce_clock_state = pp_dpm_get_vce_clock_state, .switch_power_profile = pp_dpm_switch_power_profile, .set_clockgating_by_smu = pp_set_clockgating_by_smu, + .set_powergating_by_smu = pp_set_powergating_by_smu, .get_power_profile_mode = pp_get_power_profile_mode, .set_power_profile_mode = pp_set_power_profile_mode, .odn_edit_dpm_table = pp_odn_edit_dpm_table, @@ -1227,5 +1268,5 @@ static const struct amd_pm_funcs pp_dpm_funcs = { .set_watermarks_for_clocks_ranges = pp_set_watermarks_for_clocks_ranges, .display_clock_voltage_request = pp_display_clock_voltage_request, .get_display_mode_validation_clocks = pp_get_display_mode_validation_clocks, - .set_mmhub_powergating_by_smu = pp_set_mmhub_powergating_by_smu, + .notify_smu_enable_pwe = pp_notify_smu_enable_pwe, }; diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c b/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c index a0bb921fac22..53207e76b0f3 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c @@ -435,7 +435,7 @@ int phm_get_clock_by_type_with_voltage(struct pp_hwmgr *hwmgr, } int phm_set_watermarks_for_clocks_ranges(struct pp_hwmgr *hwmgr, - struct pp_wm_sets_with_clock_ranges_soc15 *wm_with_clock_ranges) + void *clock_ranges) { PHM_FUNC_CHECK(hwmgr); @@ -443,7 +443,7 @@ int phm_set_watermarks_for_clocks_ranges(struct pp_hwmgr *hwmgr, return -EINVAL; return hwmgr->hwmgr_func->set_watermarks_for_clocks_ranges(hwmgr, - wm_with_clock_ranges); + clock_ranges); } int phm_display_clock_voltage_request(struct pp_hwmgr *hwmgr, diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c index e63bc47dc715..8994aa5c8cf8 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c @@ -81,7 +81,6 @@ int hwmgr_early_init(struct pp_hwmgr *hwmgr) return -EINVAL; hwmgr->usec_timeout = AMD_MAX_USEC_TIMEOUT; - hwmgr->power_source = PP_PowerSource_AC; hwmgr->pp_table_version = PP_TABLE_V1; hwmgr->dpm_level = AMD_DPM_FORCED_LEVEL_AUTO; hwmgr->request_dpm_level = AMD_DPM_FORCED_LEVEL_AUTO; @@ -148,10 +147,10 @@ int hwmgr_early_init(struct pp_hwmgr *hwmgr) smu7_init_function_pointers(hwmgr); break; case AMDGPU_FAMILY_AI: - hwmgr->feature_mask &= ~PP_GFXOFF_MASK; switch (hwmgr->chip_id) { case CHIP_VEGA10: case CHIP_VEGA20: + hwmgr->feature_mask &= ~PP_GFXOFF_MASK; hwmgr->smumgr_funcs = &vega10_smu_funcs; vega10_hwmgr_init(hwmgr); break; @@ -236,6 +235,11 @@ int hwmgr_hw_init(struct pp_hwmgr *hwmgr) ret = hwmgr->hwmgr_func->backend_init(hwmgr); if (ret) goto err1; + /* make sure dc limits are valid */ + if ((hwmgr->dyn_state.max_clock_voltage_on_dc.sclk == 0) || + (hwmgr->dyn_state.max_clock_voltage_on_dc.mclk == 0)) + hwmgr->dyn_state.max_clock_voltage_on_dc = + hwmgr->dyn_state.max_clock_voltage_on_ac; ret = psm_init_power_state_table(hwmgr); if (ret) diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c index 7047e29755c3..01dc46dc9c8a 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c @@ -1544,14 +1544,14 @@ void atomctrl_get_voltage_range(struct pp_hwmgr *hwmgr, uint32_t *max_vddc, switch (hwmgr->chip_id) { case CHIP_TONGA: case CHIP_FIJI: - *max_vddc = le32_to_cpu(((ATOM_ASIC_PROFILING_INFO_V3_3 *)profile)->ulMaxVddc/4); - *min_vddc = le32_to_cpu(((ATOM_ASIC_PROFILING_INFO_V3_3 *)profile)->ulMinVddc/4); + *max_vddc = le32_to_cpu(((ATOM_ASIC_PROFILING_INFO_V3_3 *)profile)->ulMaxVddc) / 4; + *min_vddc = le32_to_cpu(((ATOM_ASIC_PROFILING_INFO_V3_3 *)profile)->ulMinVddc) / 4; return; case CHIP_POLARIS11: case CHIP_POLARIS10: case CHIP_POLARIS12: - *max_vddc = le32_to_cpu(((ATOM_ASIC_PROFILING_INFO_V3_6 *)profile)->ulMaxVddc/100); - *min_vddc = le32_to_cpu(((ATOM_ASIC_PROFILING_INFO_V3_6 *)profile)->ulMinVddc/100); + *max_vddc = le32_to_cpu(((ATOM_ASIC_PROFILING_INFO_V3_6 *)profile)->ulMaxVddc) / 100; + *min_vddc = le32_to_cpu(((ATOM_ASIC_PROFILING_INFO_V3_6 *)profile)->ulMinVddc) / 100; return; default: break; diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/process_pptables_v1_0.c b/drivers/gpu/drm/amd/powerplay/hwmgr/process_pptables_v1_0.c index 35bd9870ab10..4e1fd5393845 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/process_pptables_v1_0.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/process_pptables_v1_0.c @@ -183,10 +183,10 @@ static int get_vddc_lookup_table( ATOM_Tonga_Voltage_Lookup_Record, entries, vddc_lookup_pp_tables, i); record->us_calculated = 0; - record->us_vdd = atom_record->usVdd; - record->us_cac_low = atom_record->usCACLow; - record->us_cac_mid = atom_record->usCACMid; - record->us_cac_high = atom_record->usCACHigh; + record->us_vdd = le16_to_cpu(atom_record->usVdd); + record->us_cac_low = le16_to_cpu(atom_record->usCACLow); + record->us_cac_mid = le16_to_cpu(atom_record->usCACMid); + record->us_cac_high = le16_to_cpu(atom_record->usCACHigh); } *lookup_table = table; diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c index d4bc83e81389..a63e00653324 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c @@ -993,7 +993,7 @@ static int smu10_get_clock_by_type_with_latency(struct pp_hwmgr *hwmgr, clocks->num_levels = 0; for (i = 0; i < pclk_vol_table->count; i++) { - clocks->data[i].clocks_in_khz = pclk_vol_table->entries[i].clk; + clocks->data[i].clocks_in_khz = pclk_vol_table->entries[i].clk * 10; clocks->data[i].latency_in_us = latency_required ? smu10_get_mem_latency(hwmgr, pclk_vol_table->entries[i].clk) : @@ -1044,7 +1044,7 @@ static int smu10_get_clock_by_type_with_voltage(struct pp_hwmgr *hwmgr, clocks->num_levels = 0; for (i = 0; i < pclk_vol_table->count; i++) { - clocks->data[i].clocks_in_khz = pclk_vol_table->entries[i].clk; + clocks->data[i].clocks_in_khz = pclk_vol_table->entries[i].clk * 10; clocks->data[i].voltage_in_mv = pclk_vol_table->entries[i].vol; clocks->num_levels++; } @@ -1108,9 +1108,10 @@ static int smu10_read_sensor(struct pp_hwmgr *hwmgr, int idx, } static int smu10_set_watermarks_for_clocks_ranges(struct pp_hwmgr *hwmgr, - struct pp_wm_sets_with_clock_ranges_soc15 *wm_with_clock_ranges) + void *clock_ranges) { struct smu10_hwmgr *data = hwmgr->backend; + struct dm_pp_wm_sets_with_clock_ranges_soc15 *wm_with_clock_ranges = clock_ranges; Watermarks_t *table = &(data->water_marks_table); int result = 0; @@ -1126,7 +1127,7 @@ static int smu10_smus_notify_pwe(struct pp_hwmgr *hwmgr) return smum_send_msg_to_smc(hwmgr, PPSMC_MSG_SetRccPfcPmeRestoreRegister); } -static int smu10_set_mmhub_powergating_by_smu(struct pp_hwmgr *hwmgr) +static int smu10_powergate_mmhub(struct pp_hwmgr *hwmgr) { return smum_send_msg_to_smc(hwmgr, PPSMC_MSG_PowerGateMmHub); } @@ -1182,10 +1183,11 @@ static const struct pp_hwmgr_func smu10_hwmgr_funcs = { .asic_setup = smu10_setup_asic_task, .power_state_set = smu10_set_power_state_tasks, .dynamic_state_management_disable = smu10_disable_dpm_tasks, - .set_mmhub_powergating_by_smu = smu10_set_mmhub_powergating_by_smu, + .powergate_mmhub = smu10_powergate_mmhub, .smus_notify_pwe = smu10_smus_notify_pwe, .gfx_off_control = smu10_gfx_off_control, .display_clock_voltage_request = smu10_display_clock_voltage_request, + .powergate_gfx = smu10_gfx_off_control, }; int smu10_init_function_pointers(struct pp_hwmgr *hwmgr) diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_clockpowergating.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_clockpowergating.c index 6d72a5600917..683b29a99366 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_clockpowergating.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_clockpowergating.c @@ -39,13 +39,6 @@ static int smu7_enable_disable_vce_dpm(struct pp_hwmgr *hwmgr, bool enable) PPSMC_MSG_VCEDPM_Disable); } -static int smu7_enable_disable_samu_dpm(struct pp_hwmgr *hwmgr, bool enable) -{ - return smum_send_msg_to_smc(hwmgr, enable ? - PPSMC_MSG_SAMUDPM_Enable : - PPSMC_MSG_SAMUDPM_Disable); -} - static int smu7_update_uvd_dpm(struct pp_hwmgr *hwmgr, bool bgate) { if (!bgate) @@ -60,13 +53,6 @@ static int smu7_update_vce_dpm(struct pp_hwmgr *hwmgr, bool bgate) return smu7_enable_disable_vce_dpm(hwmgr, !bgate); } -static int smu7_update_samu_dpm(struct pp_hwmgr *hwmgr, bool bgate) -{ - if (!bgate) - smum_update_smc_table(hwmgr, SMU_SAMU_TABLE); - return smu7_enable_disable_samu_dpm(hwmgr, !bgate); -} - int smu7_powerdown_uvd(struct pp_hwmgr *hwmgr) { if (phm_cf_want_uvd_power_gating(hwmgr)) @@ -107,35 +93,15 @@ static int smu7_powerup_vce(struct pp_hwmgr *hwmgr) return 0; } -static int smu7_powerdown_samu(struct pp_hwmgr *hwmgr) -{ - if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, - PHM_PlatformCaps_SamuPowerGating)) - return smum_send_msg_to_smc(hwmgr, - PPSMC_MSG_SAMPowerOFF); - return 0; -} - -static int smu7_powerup_samu(struct pp_hwmgr *hwmgr) -{ - if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, - PHM_PlatformCaps_SamuPowerGating)) - return smum_send_msg_to_smc(hwmgr, - PPSMC_MSG_SAMPowerON); - return 0; -} - int smu7_disable_clock_power_gating(struct pp_hwmgr *hwmgr) { struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend); data->uvd_power_gated = false; data->vce_power_gated = false; - data->samu_power_gated = false; smu7_powerup_uvd(hwmgr); smu7_powerup_vce(hwmgr); - smu7_powerup_samu(hwmgr); return 0; } @@ -195,26 +161,6 @@ void smu7_powergate_vce(struct pp_hwmgr *hwmgr, bool bgate) } } -int smu7_powergate_samu(struct pp_hwmgr *hwmgr, bool bgate) -{ - struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend); - - if (data->samu_power_gated == bgate) - return 0; - - data->samu_power_gated = bgate; - - if (bgate) { - smu7_update_samu_dpm(hwmgr, true); - smu7_powerdown_samu(hwmgr); - } else { - smu7_powerup_samu(hwmgr); - smu7_update_samu_dpm(hwmgr, false); - } - - return 0; -} - int smu7_update_clock_gatings(struct pp_hwmgr *hwmgr, const uint32_t *msg_id) { @@ -470,7 +416,7 @@ int smu7_update_clock_gatings(struct pp_hwmgr *hwmgr, * Powerplay will only control the static per CU Power Gating. * Dynamic per CU Power Gating will be done in gfx. */ -int smu7_enable_per_cu_power_gating(struct pp_hwmgr *hwmgr, bool enable) +int smu7_powergate_gfx(struct pp_hwmgr *hwmgr, bool enable) { struct amdgpu_device *adev = hwmgr->adev; diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_clockpowergating.h b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_clockpowergating.h index 1ddce023218a..fc8f8a6acc72 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_clockpowergating.h +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_clockpowergating.h @@ -29,11 +29,10 @@ void smu7_powergate_vce(struct pp_hwmgr *hwmgr, bool bgate); void smu7_powergate_uvd(struct pp_hwmgr *hwmgr, bool bgate); int smu7_powerdown_uvd(struct pp_hwmgr *hwmgr); -int smu7_powergate_samu(struct pp_hwmgr *hwmgr, bool bgate); int smu7_powergate_acp(struct pp_hwmgr *hwmgr, bool bgate); int smu7_disable_clock_power_gating(struct pp_hwmgr *hwmgr); int smu7_update_clock_gatings(struct pp_hwmgr *hwmgr, const uint32_t *msg_id); -int smu7_enable_per_cu_power_gating(struct pp_hwmgr *hwmgr, bool enable); +int smu7_powergate_gfx(struct pp_hwmgr *hwmgr, bool enable); #endif diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c index f8e866ceda02..052e60dfaf9f 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c @@ -48,6 +48,8 @@ #include "processpptables.h" #include "pp_thermal.h" +#include "ivsrcid/ivsrcid_vislands30.h" + #define MC_CG_ARB_FREQ_F0 0x0a #define MC_CG_ARB_FREQ_F1 0x0b #define MC_CG_ARB_FREQ_F2 0x0c @@ -885,6 +887,60 @@ static void smu7_setup_voltage_range_from_vbios(struct pp_hwmgr *hwmgr) data->odn_dpm_table.max_vddc = max_vddc; } +static void smu7_check_dpm_table_updated(struct pp_hwmgr *hwmgr) +{ + struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend); + struct smu7_odn_dpm_table *odn_table = &(data->odn_dpm_table); + struct phm_ppt_v1_information *table_info = + (struct phm_ppt_v1_information *)(hwmgr->pptable); + uint32_t i; + + struct phm_ppt_v1_clock_voltage_dependency_table *dep_table; + struct phm_ppt_v1_clock_voltage_dependency_table *odn_dep_table; + + if (table_info == NULL) + return; + + for (i = 0; i < data->dpm_table.sclk_table.count; i++) { + if (odn_table->odn_core_clock_dpm_levels.entries[i].clock != + data->dpm_table.sclk_table.dpm_levels[i].value) { + data->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_SCLK; + break; + } + } + + for (i = 0; i < data->dpm_table.mclk_table.count; i++) { + if (odn_table->odn_memory_clock_dpm_levels.entries[i].clock != + data->dpm_table.mclk_table.dpm_levels[i].value) { + data->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_MCLK; + break; + } + } + + dep_table = table_info->vdd_dep_on_mclk; + odn_dep_table = (struct phm_ppt_v1_clock_voltage_dependency_table *)&(odn_table->vdd_dependency_on_mclk); + + for (i = 0; i < dep_table->count; i++) { + if (dep_table->entries[i].vddc != odn_dep_table->entries[i].vddc) { + data->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_VDDC | DPMTABLE_OD_UPDATE_MCLK; + return; + } + } + + dep_table = table_info->vdd_dep_on_sclk; + odn_dep_table = (struct phm_ppt_v1_clock_voltage_dependency_table *)&(odn_table->vdd_dependency_on_sclk); + for (i = 0; i < dep_table->count; i++) { + if (dep_table->entries[i].vddc != odn_dep_table->entries[i].vddc) { + data->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_VDDC | DPMTABLE_OD_UPDATE_SCLK; + return; + } + } + if (data->need_update_smu7_dpm_table & DPMTABLE_OD_UPDATE_VDDC) { + data->need_update_smu7_dpm_table &= ~DPMTABLE_OD_UPDATE_VDDC; + data->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_SCLK | DPMTABLE_OD_UPDATE_MCLK; + } +} + static int smu7_setup_default_dpm_tables(struct pp_hwmgr *hwmgr) { struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend); @@ -904,10 +960,13 @@ static int smu7_setup_default_dpm_tables(struct pp_hwmgr *hwmgr) /* initialize ODN table */ if (hwmgr->od_enabled) { - smu7_setup_voltage_range_from_vbios(hwmgr); - smu7_odn_initial_default_setting(hwmgr); + if (data->odn_dpm_table.max_vddc) { + smu7_check_dpm_table_updated(hwmgr); + } else { + smu7_setup_voltage_range_from_vbios(hwmgr); + smu7_odn_initial_default_setting(hwmgr); + } } - return 0; } @@ -1521,7 +1580,7 @@ static void smu7_init_dpm_defaults(struct pp_hwmgr *hwmgr) data->current_profile_setting.sclk_up_hyst = 0; data->current_profile_setting.sclk_down_hyst = 100; data->current_profile_setting.sclk_activity = SMU7_SCLK_TARGETACTIVITY_DFLT; - data->current_profile_setting.bupdate_sclk = 1; + data->current_profile_setting.bupdate_mclk = 1; data->current_profile_setting.mclk_up_hyst = 0; data->current_profile_setting.mclk_down_hyst = 100; data->current_profile_setting.mclk_activity = SMU7_MCLK_TARGETACTIVITY_DFLT; @@ -2820,7 +2879,7 @@ static int smu7_apply_state_adjust_rules(struct pp_hwmgr *hwmgr, struct pp_power_state *request_ps, const struct pp_power_state *current_ps) { - + struct amdgpu_device *adev = hwmgr->adev; struct smu7_power_state *smu7_ps = cast_phw_smu7_power_state(&request_ps->hardware); uint32_t sclk; @@ -2843,12 +2902,12 @@ static int smu7_apply_state_adjust_rules(struct pp_hwmgr *hwmgr, "VI should always have 2 performance levels", ); - max_limits = (PP_PowerSource_AC == hwmgr->power_source) ? + max_limits = adev->pm.ac_power ? &(hwmgr->dyn_state.max_clock_voltage_on_ac) : &(hwmgr->dyn_state.max_clock_voltage_on_dc); /* Cap clock DPM tables at DC MAX if it is in DC. */ - if (PP_PowerSource_DC == hwmgr->power_source) { + if (!adev->pm.ac_power) { for (i = 0; i < smu7_ps->performance_level_count; i++) { if (smu7_ps->performance_levels[i].memory_clock > max_limits->mclk) smu7_ps->performance_levels[i].memory_clock = max_limits->mclk; @@ -3126,7 +3185,7 @@ static int smu7_get_pp_table_entry_callback_func_v1(struct pp_hwmgr *hwmgr, performance_level->pcie_gen = get_pcie_gen_support(data->pcie_gen_cap, state_entry->ucPCIEGenLow); performance_level->pcie_lane = get_pcie_lane_support(data->pcie_lane_cap, - state_entry->ucPCIELaneHigh); + state_entry->ucPCIELaneLow); performance_level = &(smu7_power_state->performance_levels [smu7_power_state->performance_level_count++]); @@ -3717,8 +3776,9 @@ static int smu7_trim_single_dpm_states(struct pp_hwmgr *hwmgr, uint32_t i; for (i = 0; i < dpm_table->count; i++) { - if ((dpm_table->dpm_levels[i].value < low_limit) - || (dpm_table->dpm_levels[i].value > high_limit)) + /*skip the trim if od is enabled*/ + if (!hwmgr->od_enabled && (dpm_table->dpm_levels[i].value < low_limit + || dpm_table->dpm_levels[i].value > high_limit)) dpm_table->dpm_levels[i].enabled = false; else dpm_table->dpm_levels[i].enabled = true; @@ -3762,10 +3822,8 @@ static int smu7_generate_dpm_level_enable_mask( const struct smu7_power_state *smu7_ps = cast_const_phw_smu7_power_state(states->pnew_state); - /*skip the trim if od is enabled*/ - if (!hwmgr->od_enabled) - result = smu7_trim_dpm_states(hwmgr, smu7_ps); + result = smu7_trim_dpm_states(hwmgr, smu7_ps); if (result) return result; @@ -4049,17 +4107,17 @@ static int smu7_register_irq_handlers(struct pp_hwmgr *hwmgr) amdgpu_irq_add_id((struct amdgpu_device *)(hwmgr->adev), AMDGPU_IH_CLIENTID_LEGACY, - 230, + VISLANDS30_IV_SRCID_CG_TSS_THERMAL_LOW_TO_HIGH, source); amdgpu_irq_add_id((struct amdgpu_device *)(hwmgr->adev), AMDGPU_IH_CLIENTID_LEGACY, - 231, + VISLANDS30_IV_SRCID_CG_TSS_THERMAL_HIGH_TO_LOW, source); /* Register CTF(GPIO_19) interrupt */ amdgpu_irq_add_id((struct amdgpu_device *)(hwmgr->adev), AMDGPU_IH_CLIENTID_LEGACY, - 83, + VISLANDS30_IV_SRCID_GPIO_19, source); return 0; @@ -4244,7 +4302,6 @@ static int smu7_init_power_gate_state(struct pp_hwmgr *hwmgr) data->uvd_power_gated = false; data->vce_power_gated = false; - data->samu_power_gated = false; return 0; } @@ -4555,12 +4612,12 @@ static int smu7_get_sclks(struct pp_hwmgr *hwmgr, struct amd_pp_clocks *clocks) return -EINVAL; dep_sclk_table = table_info->vdd_dep_on_sclk; for (i = 0; i < dep_sclk_table->count; i++) - clocks->clock[i] = dep_sclk_table->entries[i].clk; + clocks->clock[i] = dep_sclk_table->entries[i].clk * 10; clocks->count = dep_sclk_table->count; } else if (hwmgr->pp_table_version == PP_TABLE_V0) { sclk_table = hwmgr->dyn_state.vddc_dependency_on_sclk; for (i = 0; i < sclk_table->count; i++) - clocks->clock[i] = sclk_table->entries[i].clk; + clocks->clock[i] = sclk_table->entries[i].clk * 10; clocks->count = sclk_table->count; } @@ -4592,7 +4649,7 @@ static int smu7_get_mclks(struct pp_hwmgr *hwmgr, struct amd_pp_clocks *clocks) return -EINVAL; dep_mclk_table = table_info->vdd_dep_on_mclk; for (i = 0; i < dep_mclk_table->count; i++) { - clocks->clock[i] = dep_mclk_table->entries[i].clk; + clocks->clock[i] = dep_mclk_table->entries[i].clk * 10; clocks->latency[i] = smu7_get_mem_latency(hwmgr, dep_mclk_table->entries[i].clk); } @@ -4600,7 +4657,7 @@ static int smu7_get_mclks(struct pp_hwmgr *hwmgr, struct amd_pp_clocks *clocks) } else if (hwmgr->pp_table_version == PP_TABLE_V0) { mclk_table = hwmgr->dyn_state.vddc_dependency_on_mclk; for (i = 0; i < mclk_table->count; i++) - clocks->clock[i] = mclk_table->entries[i].clk; + clocks->clock[i] = mclk_table->entries[i].clk * 10; clocks->count = mclk_table->count; } return 0; @@ -4739,60 +4796,6 @@ static bool smu7_check_clk_voltage_valid(struct pp_hwmgr *hwmgr, return true; } -static void smu7_check_dpm_table_updated(struct pp_hwmgr *hwmgr) -{ - struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend); - struct smu7_odn_dpm_table *odn_table = &(data->odn_dpm_table); - struct phm_ppt_v1_information *table_info = - (struct phm_ppt_v1_information *)(hwmgr->pptable); - uint32_t i; - - struct phm_ppt_v1_clock_voltage_dependency_table *dep_table; - struct phm_ppt_v1_clock_voltage_dependency_table *odn_dep_table; - - if (table_info == NULL) - return; - - for (i=0; i<data->dpm_table.sclk_table.count; i++) { - if (odn_table->odn_core_clock_dpm_levels.entries[i].clock != - data->dpm_table.sclk_table.dpm_levels[i].value) { - data->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_SCLK; - break; - } - } - - for (i=0; i<data->dpm_table.mclk_table.count; i++) { - if (odn_table->odn_memory_clock_dpm_levels.entries[i].clock != - data->dpm_table.mclk_table.dpm_levels[i].value) { - data->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_MCLK; - break; - } - } - - dep_table = table_info->vdd_dep_on_mclk; - odn_dep_table = (struct phm_ppt_v1_clock_voltage_dependency_table *)&(odn_table->vdd_dependency_on_mclk); - - for (i=0; i < dep_table->count; i++) { - if (dep_table->entries[i].vddc != odn_dep_table->entries[i].vddc) { - data->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_VDDC | DPMTABLE_OD_UPDATE_MCLK; - return; - } - } - - dep_table = table_info->vdd_dep_on_sclk; - odn_dep_table = (struct phm_ppt_v1_clock_voltage_dependency_table *)&(odn_table->vdd_dependency_on_sclk); - for (i=0; i < dep_table->count; i++) { - if (dep_table->entries[i].vddc != odn_dep_table->entries[i].vddc) { - data->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_VDDC | DPMTABLE_OD_UPDATE_SCLK; - return; - } - } - if (data->need_update_smu7_dpm_table & DPMTABLE_OD_UPDATE_VDDC) { - data->need_update_smu7_dpm_table &= ~DPMTABLE_OD_UPDATE_VDDC; - data->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_SCLK | DPMTABLE_OD_UPDATE_MCLK; - } -} - static int smu7_odn_edit_dpm_table(struct pp_hwmgr *hwmgr, enum PP_OD_DPM_TABLE_COMMAND type, long *input, uint32_t size) @@ -5043,7 +5046,7 @@ static const struct pp_hwmgr_func smu7_hwmgr_funcs = { .get_fan_control_mode = smu7_get_fan_control_mode, .force_clock_level = smu7_force_clock_level, .print_clock_levels = smu7_print_clock_levels, - .enable_per_cu_power_gating = smu7_enable_per_cu_power_gating, + .powergate_gfx = smu7_powergate_gfx, .get_sclk_od = smu7_get_sclk_od, .set_sclk_od = smu7_set_sclk_od, .get_mclk_od = smu7_get_mclk_od, diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.h b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.h index c91e75db6a8e..3784ce6e50ab 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.h +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.h @@ -310,7 +310,6 @@ struct smu7_hwmgr { /* ---- Power Gating States ---- */ bool uvd_power_gated; bool vce_power_gated; - bool samu_power_gated; bool need_long_memory_training; /* Application power optimization parameters */ diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c index c952845833d7..5e19f5977eb1 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c @@ -403,6 +403,49 @@ static const struct gpu_pt_config_reg DIDTConfig_Polaris12[] = { { ixDIDT_SQ_CTRL1, DIDT_SQ_CTRL1__MAX_POWER_MASK, DIDT_SQ_CTRL1__MAX_POWER__SHIFT, 0xffff, GPU_CONFIGREG_DIDT_IND }, { ixDIDT_SQ_CTRL_OCP, DIDT_SQ_CTRL_OCP__UNUSED_0_MASK, DIDT_SQ_CTRL_OCP__UNUSED_0__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_CTRL_OCP, DIDT_SQ_CTRL_OCP__OCP_MAX_POWER_MASK, DIDT_SQ_CTRL_OCP__OCP_MAX_POWER__SHIFT, 0xffff, GPU_CONFIGREG_DIDT_IND }, + + { ixDIDT_SQ_CTRL2, DIDT_SQ_CTRL2__MAX_POWER_DELTA_MASK, DIDT_SQ_CTRL2__MAX_POWER_DELTA__SHIFT, 0x3853, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_CTRL2, DIDT_SQ_CTRL2__UNUSED_0_MASK, DIDT_SQ_CTRL2__UNUSED_0__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_CTRL2, DIDT_SQ_CTRL2__SHORT_TERM_INTERVAL_SIZE_MASK, DIDT_SQ_CTRL2__SHORT_TERM_INTERVAL_SIZE__SHIFT, 0x005a, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_CTRL2, DIDT_SQ_CTRL2__UNUSED_1_MASK, DIDT_SQ_CTRL2__UNUSED_1__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_CTRL2, DIDT_SQ_CTRL2__LONG_TERM_INTERVAL_RATIO_MASK, DIDT_SQ_CTRL2__LONG_TERM_INTERVAL_RATIO__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_CTRL2, DIDT_SQ_CTRL2__UNUSED_2_MASK, DIDT_SQ_CTRL2__UNUSED_2__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + + { ixDIDT_SQ_STALL_CTRL, DIDT_SQ_STALL_CTRL__DIDT_STALL_CTRL_ENABLE_MASK, DIDT_SQ_STALL_CTRL__DIDT_STALL_CTRL_ENABLE__SHIFT, 0x0001, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_STALL_CTRL, DIDT_SQ_STALL_CTRL__DIDT_STALL_DELAY_HI_MASK, DIDT_SQ_STALL_CTRL__DIDT_STALL_DELAY_HI__SHIFT, 0x0001, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_STALL_CTRL, DIDT_SQ_STALL_CTRL__DIDT_STALL_DELAY_LO_MASK, DIDT_SQ_STALL_CTRL__DIDT_STALL_DELAY_LO__SHIFT, 0x0001, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_STALL_CTRL, DIDT_SQ_STALL_CTRL__DIDT_HI_POWER_THRESHOLD_MASK, DIDT_SQ_STALL_CTRL__DIDT_HI_POWER_THRESHOLD__SHIFT, 0x0ebb, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_STALL_CTRL, DIDT_SQ_STALL_CTRL__UNUSED_0_MASK, DIDT_SQ_STALL_CTRL__UNUSED_0__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + + { ixDIDT_SQ_TUNING_CTRL, DIDT_SQ_TUNING_CTRL__DIDT_TUNING_ENABLE_MASK, DIDT_SQ_TUNING_CTRL__DIDT_TUNING_ENABLE__SHIFT, 0x0001, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_TUNING_CTRL, DIDT_SQ_TUNING_CTRL__MAX_POWER_DELTA_HI_MASK, DIDT_SQ_TUNING_CTRL__MAX_POWER_DELTA_HI__SHIFT, 0x3853, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_TUNING_CTRL, DIDT_SQ_TUNING_CTRL__MAX_POWER_DELTA_LO_MASK, DIDT_SQ_TUNING_CTRL__MAX_POWER_DELTA_LO__SHIFT, 0x3153, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_TUNING_CTRL, DIDT_SQ_TUNING_CTRL__UNUSED_0_MASK, DIDT_SQ_TUNING_CTRL__UNUSED_0__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + + { ixDIDT_SQ_CTRL0, DIDT_SQ_CTRL0__DIDT_CTRL_EN_MASK, DIDT_SQ_CTRL0__DIDT_CTRL_EN__SHIFT, 0x0001, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_CTRL0, DIDT_SQ_CTRL0__USE_REF_CLOCK_MASK, DIDT_SQ_CTRL0__USE_REF_CLOCK__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_CTRL0, DIDT_SQ_CTRL0__PHASE_OFFSET_MASK, DIDT_SQ_CTRL0__PHASE_OFFSET__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_CTRL0, DIDT_SQ_CTRL0__DIDT_CTRL_RST_MASK, DIDT_SQ_CTRL0__DIDT_CTRL_RST__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_CTRL0, DIDT_SQ_CTRL0__DIDT_CLK_EN_OVERRIDE_MASK, DIDT_SQ_CTRL0__DIDT_CLK_EN_OVERRIDE__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_CTRL0, DIDT_SQ_CTRL0__DIDT_MAX_STALLS_ALLOWED_HI_MASK, DIDT_SQ_CTRL0__DIDT_MAX_STALLS_ALLOWED_HI__SHIFT, 0x0010, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_CTRL0, DIDT_SQ_CTRL0__DIDT_MAX_STALLS_ALLOWED_LO_MASK, DIDT_SQ_CTRL0__DIDT_MAX_STALLS_ALLOWED_LO__SHIFT, 0x0010, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_SQ_CTRL0, DIDT_SQ_CTRL0__UNUSED_0_MASK, DIDT_SQ_CTRL0__UNUSED_0__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + + { ixDIDT_TD_WEIGHT0_3, DIDT_TD_WEIGHT0_3__WEIGHT0_MASK, DIDT_TD_WEIGHT0_3__WEIGHT0__SHIFT, 0x000a, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TD_WEIGHT0_3, DIDT_TD_WEIGHT0_3__WEIGHT1_MASK, DIDT_TD_WEIGHT0_3__WEIGHT1__SHIFT, 0x0010, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TD_WEIGHT0_3, DIDT_TD_WEIGHT0_3__WEIGHT2_MASK, DIDT_TD_WEIGHT0_3__WEIGHT2__SHIFT, 0x0017, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TD_WEIGHT0_3, DIDT_TD_WEIGHT0_3__WEIGHT3_MASK, DIDT_TD_WEIGHT0_3__WEIGHT3__SHIFT, 0x002f, GPU_CONFIGREG_DIDT_IND }, + + { ixDIDT_TD_WEIGHT4_7, DIDT_TD_WEIGHT4_7__WEIGHT4_MASK, DIDT_TD_WEIGHT4_7__WEIGHT4__SHIFT, 0x0046, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TD_WEIGHT4_7, DIDT_TD_WEIGHT4_7__WEIGHT5_MASK, DIDT_TD_WEIGHT4_7__WEIGHT5__SHIFT, 0x005d, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TD_WEIGHT4_7, DIDT_TD_WEIGHT4_7__WEIGHT6_MASK, DIDT_TD_WEIGHT4_7__WEIGHT6__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TD_WEIGHT4_7, DIDT_TD_WEIGHT4_7__WEIGHT7_MASK, DIDT_TD_WEIGHT4_7__WEIGHT7__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + + { ixDIDT_TD_CTRL1, DIDT_TD_CTRL1__MIN_POWER_MASK, DIDT_TD_CTRL1__MIN_POWER__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, + { ixDIDT_TD_CTRL1, DIDT_TD_CTRL1__MAX_POWER_MASK, DIDT_TD_CTRL1__MAX_POWER__SHIFT, 0xffff, GPU_CONFIGREG_DIDT_IND }, + + { ixDIDT_TD_CTRL_OCP, DIDT_TD_CTRL_OCP__UNUSED_0_MASK, DIDT_TD_CTRL_OCP__UNUSED_0__SHIFT, 0x0000, GPU_CONFIGREG_DIDT_IND }, { ixDIDT_TD_CTRL_OCP, DIDT_TD_CTRL_OCP__OCP_MAX_POWER_MASK, DIDT_TD_CTRL_OCP__OCP_MAX_POWER__SHIFT, 0x00ff, GPU_CONFIGREG_DIDT_IND }, { ixDIDT_TD_CTRL2, DIDT_TD_CTRL2__MAX_POWER_DELTA_MASK, DIDT_TD_CTRL2__MAX_POWER_DELTA__SHIFT, 0x3fff, GPU_CONFIGREG_DIDT_IND }, diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu8_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu8_hwmgr.c index 50690c72b2ea..0adfc5392cd3 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu8_hwmgr.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu8_hwmgr.c @@ -244,6 +244,7 @@ static int smu8_initialize_dpm_defaults(struct pp_hwmgr *hwmgr) return 0; } +/* convert form 8bit vid to real voltage in mV*4 */ static uint32_t smu8_convert_8Bit_index_to_voltage( struct pp_hwmgr *hwmgr, uint16_t voltage) { @@ -1604,17 +1605,17 @@ static int smu8_get_clock_by_type(struct pp_hwmgr *hwmgr, enum amd_pp_clock_type switch (type) { case amd_pp_disp_clock: for (i = 0; i < clocks->count; i++) - clocks->clock[i] = data->sys_info.display_clock[i]; + clocks->clock[i] = data->sys_info.display_clock[i] * 10; break; case amd_pp_sys_clock: table = hwmgr->dyn_state.vddc_dependency_on_sclk; for (i = 0; i < clocks->count; i++) - clocks->clock[i] = table->entries[i].clk; + clocks->clock[i] = table->entries[i].clk * 10; break; case amd_pp_mem_clock: clocks->count = SMU8_NUM_NBPMEMORYCLOCK; for (i = 0; i < clocks->count; i++) - clocks->clock[i] = data->sys_info.nbp_memory_clock[clocks->count - 1 - i]; + clocks->clock[i] = data->sys_info.nbp_memory_clock[clocks->count - 1 - i] * 10; break; default: return -1; @@ -1702,13 +1703,13 @@ static int smu8_read_sensor(struct pp_hwmgr *hwmgr, int idx, case AMDGPU_PP_SENSOR_VDDNB: tmp = (cgs_read_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixSMUSVI_NB_CURRENTVID) & CURRENT_NB_VID_MASK) >> CURRENT_NB_VID__SHIFT; - vddnb = smu8_convert_8Bit_index_to_voltage(hwmgr, tmp); + vddnb = smu8_convert_8Bit_index_to_voltage(hwmgr, tmp) / 4; *((uint32_t *)value) = vddnb; return 0; case AMDGPU_PP_SENSOR_VDDGFX: tmp = (cgs_read_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixSMUSVI_GFX_CURRENTVID) & CURRENT_GFX_VID_MASK) >> CURRENT_GFX_VID__SHIFT; - vddgfx = smu8_convert_8Bit_index_to_voltage(hwmgr, (u16)tmp); + vddgfx = smu8_convert_8Bit_index_to_voltage(hwmgr, (u16)tmp) / 4; *((uint32_t *)value) = vddgfx; return 0; case AMDGPU_PP_SENSOR_UVD_VCLK: diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.c index 93a3d022ba47..2aab1b475945 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.c @@ -25,6 +25,9 @@ #include "ppatomctrl.h" #include "ppsmc.h" #include "atom.h" +#include "ivsrcid/thm/irqsrcs_thm_9_0.h" +#include "ivsrcid/smuio/irqsrcs_smuio_9_0.h" +#include "ivsrcid/ivsrcid_vislands30.h" uint8_t convert_to_vid(uint16_t vddc) { @@ -543,17 +546,17 @@ int phm_irq_process(struct amdgpu_device *adev, uint32_t src_id = entry->src_id; if (client_id == AMDGPU_IH_CLIENTID_LEGACY) { - if (src_id == 230) + if (src_id == VISLANDS30_IV_SRCID_CG_TSS_THERMAL_LOW_TO_HIGH) pr_warn("GPU over temperature range detected on PCIe %d:%d.%d!\n", PCI_BUS_NUM(adev->pdev->devfn), PCI_SLOT(adev->pdev->devfn), PCI_FUNC(adev->pdev->devfn)); - else if (src_id == 231) + else if (src_id == VISLANDS30_IV_SRCID_CG_TSS_THERMAL_HIGH_TO_LOW) pr_warn("GPU under temperature range detected on PCIe %d:%d.%d!\n", PCI_BUS_NUM(adev->pdev->devfn), PCI_SLOT(adev->pdev->devfn), PCI_FUNC(adev->pdev->devfn)); - else if (src_id == 83) + else if (src_id == VISLANDS30_IV_SRCID_GPIO_19) pr_warn("GPU Critical Temperature Fault detected on PCIe %d:%d.%d!\n", PCI_BUS_NUM(adev->pdev->devfn), PCI_SLOT(adev->pdev->devfn), @@ -594,17 +597,17 @@ int smu9_register_irq_handlers(struct pp_hwmgr *hwmgr) amdgpu_irq_add_id((struct amdgpu_device *)(hwmgr->adev), SOC15_IH_CLIENTID_THM, - 0, + THM_9_0__SRCID__THM_DIG_THERM_L2H, source); amdgpu_irq_add_id((struct amdgpu_device *)(hwmgr->adev), SOC15_IH_CLIENTID_THM, - 1, + THM_9_0__SRCID__THM_DIG_THERM_H2L, source); /* Register CTF(GPIO_19) interrupt */ amdgpu_irq_add_id((struct amdgpu_device *)(hwmgr->adev), SOC15_IH_CLIENTID_ROM_SMUIO, - 83, + SMUIO_9_0__SRCID__SMUIO_GPIO19, source); return 0; @@ -652,7 +655,7 @@ int smu_get_voltage_dependency_table_ppt_v1( } int smu_set_watermarks_for_clocks_ranges(void *wt_table, - struct pp_wm_sets_with_clock_ranges_soc15 *wm_with_clock_ranges) + struct dm_pp_wm_sets_with_clock_ranges_soc15 *wm_with_clock_ranges) { uint32_t i; struct watermarks *table = wt_table; @@ -660,49 +663,49 @@ int smu_set_watermarks_for_clocks_ranges(void *wt_table, if (!table || !wm_with_clock_ranges) return -EINVAL; - if (wm_with_clock_ranges->num_wm_sets_dmif > 4 || wm_with_clock_ranges->num_wm_sets_mcif > 4) + if (wm_with_clock_ranges->num_wm_dmif_sets > 4 || wm_with_clock_ranges->num_wm_mcif_sets > 4) return -EINVAL; - for (i = 0; i < wm_with_clock_ranges->num_wm_sets_dmif; i++) { + for (i = 0; i < wm_with_clock_ranges->num_wm_dmif_sets; i++) { table->WatermarkRow[1][i].MinClock = cpu_to_le16((uint16_t) - (wm_with_clock_ranges->wm_sets_dmif[i].wm_min_dcefclk_in_khz) / - 100); + (wm_with_clock_ranges->wm_dmif_clocks_ranges[i].wm_min_dcfclk_clk_in_khz) / + 1000); table->WatermarkRow[1][i].MaxClock = cpu_to_le16((uint16_t) - (wm_with_clock_ranges->wm_sets_dmif[i].wm_max_dcefclk_in_khz) / + (wm_with_clock_ranges->wm_dmif_clocks_ranges[i].wm_max_dcfclk_clk_in_khz) / 100); table->WatermarkRow[1][i].MinUclk = cpu_to_le16((uint16_t) - (wm_with_clock_ranges->wm_sets_dmif[i].wm_min_memclk_in_khz) / - 100); + (wm_with_clock_ranges->wm_dmif_clocks_ranges[i].wm_min_mem_clk_in_khz) / + 1000); table->WatermarkRow[1][i].MaxUclk = cpu_to_le16((uint16_t) - (wm_with_clock_ranges->wm_sets_dmif[i].wm_max_memclk_in_khz) / - 100); + (wm_with_clock_ranges->wm_dmif_clocks_ranges[i].wm_max_mem_clk_in_khz) / + 1000); table->WatermarkRow[1][i].WmSetting = (uint8_t) - wm_with_clock_ranges->wm_sets_dmif[i].wm_set_id; + wm_with_clock_ranges->wm_dmif_clocks_ranges[i].wm_set_id; } - for (i = 0; i < wm_with_clock_ranges->num_wm_sets_mcif; i++) { + for (i = 0; i < wm_with_clock_ranges->num_wm_mcif_sets; i++) { table->WatermarkRow[0][i].MinClock = cpu_to_le16((uint16_t) - (wm_with_clock_ranges->wm_sets_mcif[i].wm_min_socclk_in_khz) / - 100); + (wm_with_clock_ranges->wm_mcif_clocks_ranges[i].wm_min_socclk_clk_in_khz) / + 1000); table->WatermarkRow[0][i].MaxClock = cpu_to_le16((uint16_t) - (wm_with_clock_ranges->wm_sets_mcif[i].wm_max_socclk_in_khz) / - 100); + (wm_with_clock_ranges->wm_mcif_clocks_ranges[i].wm_max_socclk_clk_in_khz) / + 1000); table->WatermarkRow[0][i].MinUclk = cpu_to_le16((uint16_t) - (wm_with_clock_ranges->wm_sets_mcif[i].wm_min_memclk_in_khz) / - 100); + (wm_with_clock_ranges->wm_mcif_clocks_ranges[i].wm_min_mem_clk_in_khz) / + 1000); table->WatermarkRow[0][i].MaxUclk = cpu_to_le16((uint16_t) - (wm_with_clock_ranges->wm_sets_mcif[i].wm_max_memclk_in_khz) / - 100); + (wm_with_clock_ranges->wm_mcif_clocks_ranges[i].wm_max_mem_clk_in_khz) / + 1000); table->WatermarkRow[0][i].WmSetting = (uint8_t) - wm_with_clock_ranges->wm_sets_mcif[i].wm_set_id; + wm_with_clock_ranges->wm_mcif_clocks_ranges[i].wm_set_id; } return 0; } diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.h b/drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.h index 916cc01e7652..5454289d5226 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.h +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.h @@ -107,7 +107,7 @@ int smu_get_voltage_dependency_table_ppt_v1( struct phm_ppt_v1_clock_voltage_dependency_table *dep_table); int smu_set_watermarks_for_clocks_ranges(void *wt_table, - struct pp_wm_sets_with_clock_ranges_soc15 *wm_with_clock_ranges); + struct dm_pp_wm_sets_with_clock_ranges_soc15 *wm_with_clock_ranges); #define PHM_FIELD_SHIFT(reg, field) reg##__##field##__SHIFT #define PHM_FIELD_MASK(reg, field) reg##__##field##_MASK diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c index 05e680d55dbb..fb86c24394ff 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c @@ -55,12 +55,6 @@ static const uint32_t channel_number[] = {1, 2, 0, 4, 0, 8, 0, 16, 2}; -#define MEM_FREQ_LOW_LATENCY 25000 -#define MEM_FREQ_HIGH_LATENCY 80000 -#define MEM_LATENCY_HIGH 245 -#define MEM_LATENCY_LOW 35 -#define MEM_LATENCY_ERR 0xFFFF - #define mmDF_CS_AON0_DramBaseAddress0 0x0044 #define mmDF_CS_AON0_DramBaseAddress0_BASE_IDX 0 @@ -295,7 +289,15 @@ static int vega10_odn_initial_default_setting(struct pp_hwmgr *hwmgr) struct phm_ppt_v1_voltage_lookup_table *vddc_lookup_table; struct phm_ppt_v1_clock_voltage_dependency_table *dep_table[3]; struct phm_ppt_v1_clock_voltage_dependency_table *od_table[3]; + struct pp_atomfwctrl_avfs_parameters avfs_params = {0}; uint32_t i; + int result; + + result = pp_atomfwctrl_get_avfs_information(hwmgr, &avfs_params); + if (!result) { + data->odn_dpm_table.max_vddc = avfs_params.ulMaxVddc; + data->odn_dpm_table.min_vddc = avfs_params.ulMinVddc; + } od_lookup_table = &odn_table->vddc_lookup_table; vddc_lookup_table = table_info->vddc_lookup_table; @@ -2078,9 +2080,6 @@ static int vega10_populate_avfs_parameters(struct pp_hwmgr *hwmgr) if (data->smu_features[GNLD_AVFS].supported) { result = pp_atomfwctrl_get_avfs_information(hwmgr, &avfs_params); if (!result) { - data->odn_dpm_table.max_vddc = avfs_params.ulMaxVddc; - data->odn_dpm_table.min_vddc = avfs_params.ulMinVddc; - pp_table->MinVoltageVid = (uint8_t) convert_to_vid((uint16_t)(avfs_params.ulMinVddc)); pp_table->MaxVoltageVid = (uint8_t) @@ -2414,6 +2413,40 @@ static int vega10_populate_and_upload_avfs_fuse_override(struct pp_hwmgr *hwmgr) return result; } +static void vega10_check_dpm_table_updated(struct pp_hwmgr *hwmgr) +{ + struct vega10_hwmgr *data = hwmgr->backend; + struct vega10_odn_dpm_table *odn_table = &(data->odn_dpm_table); + struct phm_ppt_v2_information *table_info = hwmgr->pptable; + struct phm_ppt_v1_clock_voltage_dependency_table *dep_table; + struct phm_ppt_v1_clock_voltage_dependency_table *odn_dep_table; + uint32_t i; + + dep_table = table_info->vdd_dep_on_mclk; + odn_dep_table = (struct phm_ppt_v1_clock_voltage_dependency_table *)&(odn_table->vdd_dep_on_mclk); + + for (i = 0; i < dep_table->count; i++) { + if (dep_table->entries[i].vddc != odn_dep_table->entries[i].vddc) { + data->need_update_dpm_table |= DPMTABLE_OD_UPDATE_VDDC | DPMTABLE_OD_UPDATE_MCLK; + return; + } + } + + dep_table = table_info->vdd_dep_on_sclk; + odn_dep_table = (struct phm_ppt_v1_clock_voltage_dependency_table *)&(odn_table->vdd_dep_on_sclk); + for (i = 0; i < dep_table->count; i++) { + if (dep_table->entries[i].vddc != odn_dep_table->entries[i].vddc) { + data->need_update_dpm_table |= DPMTABLE_OD_UPDATE_VDDC | DPMTABLE_OD_UPDATE_SCLK; + return; + } + } + + if (data->need_update_dpm_table & DPMTABLE_OD_UPDATE_VDDC) { + data->need_update_dpm_table &= ~DPMTABLE_OD_UPDATE_VDDC; + data->need_update_dpm_table |= DPMTABLE_OD_UPDATE_SCLK | DPMTABLE_OD_UPDATE_MCLK; + } +} + /** * Initializes the SMC table and uploads it * @@ -2430,6 +2463,7 @@ static int vega10_init_smc_table(struct pp_hwmgr *hwmgr) PPTable_t *pp_table = &(data->smc_state_table.pp_table); struct pp_atomfwctrl_voltage_table voltage_table; struct pp_atomfwctrl_bios_boot_up_values boot_up_values; + struct vega10_odn_dpm_table *odn_table = &(data->odn_dpm_table); result = vega10_setup_default_dpm_tables(hwmgr); PP_ASSERT_WITH_CODE(!result, @@ -2437,8 +2471,14 @@ static int vega10_init_smc_table(struct pp_hwmgr *hwmgr) return result); /* initialize ODN table */ - if (hwmgr->od_enabled) - vega10_odn_initial_default_setting(hwmgr); + if (hwmgr->od_enabled) { + if (odn_table->max_vddc) { + data->need_update_dpm_table |= DPMTABLE_OD_UPDATE_SCLK | DPMTABLE_OD_UPDATE_MCLK; + vega10_check_dpm_table_updated(hwmgr); + } else { + vega10_odn_initial_default_setting(hwmgr); + } + } pp_atomfwctrl_get_voltage_table_v4(hwmgr, VOLTAGE_TYPE_VDDC, VOLTAGE_OBJ_SVID2, &voltage_table); @@ -2861,11 +2901,6 @@ static int vega10_enable_dpm_tasks(struct pp_hwmgr *hwmgr) vega10_enable_disable_PCC_limit_feature(hwmgr, true); - if ((hwmgr->smu_version == 0x001c2c00) || - (hwmgr->smu_version == 0x001c2d00)) - smum_send_msg_to_smc_with_parameter(hwmgr, - PPSMC_MSG_UpdatePkgPwrPidAlpha, 1); - smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_ConfigureTelemetry, data->config_telemetry); @@ -3061,6 +3096,7 @@ static int vega10_apply_state_adjust_rules(struct pp_hwmgr *hwmgr, struct pp_power_state *request_ps, const struct pp_power_state *current_ps) { + struct amdgpu_device *adev = hwmgr->adev; struct vega10_power_state *vega10_ps = cast_phw_vega10_power_state(&request_ps->hardware); uint32_t sclk; @@ -3086,12 +3122,12 @@ static int vega10_apply_state_adjust_rules(struct pp_hwmgr *hwmgr, if (vega10_ps->performance_level_count != 2) pr_info("VI should always have 2 performance levels"); - max_limits = (PP_PowerSource_AC == hwmgr->power_source) ? + max_limits = adev->pm.ac_power ? &(hwmgr->dyn_state.max_clock_voltage_on_ac) : &(hwmgr->dyn_state.max_clock_voltage_on_dc); /* Cap clock DPM tables at DC MAX if it is in DC. */ - if (PP_PowerSource_DC == hwmgr->power_source) { + if (!adev->pm.ac_power) { for (i = 0; i < vega10_ps->performance_level_count; i++) { if (vega10_ps->performance_levels[i].mem_clock > max_limits->mclk) @@ -3181,7 +3217,7 @@ static int vega10_apply_state_adjust_rules(struct pp_hwmgr *hwmgr, /* Find the lowest MCLK frequency that is within * the tolerable latency defined in DAL */ - latency = 0; + latency = hwmgr->display_config->dce_tolerable_mclk_in_active_latency; for (i = 0; i < data->mclk_latency_table.count; i++) { if ((data->mclk_latency_table.entries[i].latency <= latency) && (data->mclk_latency_table.entries[i].frequency >= @@ -3223,10 +3259,25 @@ static int vega10_populate_and_upload_sclk_mclk_dpm_levels( { int result = 0; struct vega10_hwmgr *data = hwmgr->backend; + struct vega10_dpm_table *dpm_table = &data->dpm_table; + struct vega10_odn_dpm_table *odn_table = &data->odn_dpm_table; + struct vega10_odn_clock_voltage_dependency_table *odn_clk_table = &odn_table->vdd_dep_on_sclk; + int count; if (!data->need_update_dpm_table) return 0; + if (hwmgr->od_enabled && data->need_update_dpm_table & DPMTABLE_OD_UPDATE_SCLK) { + for (count = 0; count < dpm_table->gfx_table.count; count++) + dpm_table->gfx_table.dpm_levels[count].value = odn_clk_table->entries[count].clk; + } + + odn_clk_table = &odn_table->vdd_dep_on_mclk; + if (hwmgr->od_enabled && data->need_update_dpm_table & DPMTABLE_OD_UPDATE_MCLK) { + for (count = 0; count < dpm_table->mem_table.count; count++) + dpm_table->mem_table.dpm_levels[count].value = odn_clk_table->entries[count].clk; + } + if (data->need_update_dpm_table & (DPMTABLE_OD_UPDATE_SCLK + DPMTABLE_UPDATE_SCLK + DPMTABLE_UPDATE_SOCCLK)) { result = vega10_populate_all_graphic_levels(hwmgr); @@ -3674,7 +3725,7 @@ static void vega10_notify_smc_display_change(struct pp_hwmgr *hwmgr, { smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_SetUclkFastSwitch, - has_disp ? 0 : 1); + has_disp ? 1 : 0); } int vega10_display_clock_voltage_request(struct pp_hwmgr *hwmgr, @@ -3749,7 +3800,9 @@ static int vega10_notify_smc_display_config_after_ps_adjustment( uint32_t i; struct pp_display_clock_request clock_req; - if (hwmgr->display_config->num_display > 1) + if ((hwmgr->display_config->num_display > 1) && + !hwmgr->display_config->multi_monitor_in_sync && + !hwmgr->display_config->nb_pstate_switch_disable) vega10_notify_smc_display_change(hwmgr, false); else vega10_notify_smc_display_change(hwmgr, true); @@ -3765,7 +3818,7 @@ static int vega10_notify_smc_display_config_after_ps_adjustment( if (i < dpm_table->count) { clock_req.clock_type = amd_pp_dcef_clock; - clock_req.clock_freq_in_khz = dpm_table->dpm_levels[i].value; + clock_req.clock_freq_in_khz = dpm_table->dpm_levels[i].value * 10; if (!vega10_display_clock_voltage_request(hwmgr, &clock_req)) { smum_send_msg_to_smc_with_parameter( hwmgr, PPSMC_MSG_SetMinDeepSleepDcefclk, @@ -4022,28 +4075,17 @@ static void vega10_get_sclks(struct pp_hwmgr *hwmgr, table_info->vdd_dep_on_sclk; uint32_t i; + clocks->num_levels = 0; for (i = 0; i < dep_table->count; i++) { if (dep_table->entries[i].clk) { clocks->data[clocks->num_levels].clocks_in_khz = - dep_table->entries[i].clk; + dep_table->entries[i].clk * 10; clocks->num_levels++; } } } -static uint32_t vega10_get_mem_latency(struct pp_hwmgr *hwmgr, - uint32_t clock) -{ - if (clock >= MEM_FREQ_LOW_LATENCY && - clock < MEM_FREQ_HIGH_LATENCY) - return MEM_LATENCY_HIGH; - else if (clock >= MEM_FREQ_HIGH_LATENCY) - return MEM_LATENCY_LOW; - else - return MEM_LATENCY_ERR; -} - static void vega10_get_memclocks(struct pp_hwmgr *hwmgr, struct pp_clock_levels_with_latency *clocks) { @@ -4052,26 +4094,22 @@ static void vega10_get_memclocks(struct pp_hwmgr *hwmgr, struct phm_ppt_v1_clock_voltage_dependency_table *dep_table = table_info->vdd_dep_on_mclk; struct vega10_hwmgr *data = hwmgr->backend; + uint32_t j = 0; uint32_t i; - clocks->num_levels = 0; - data->mclk_latency_table.count = 0; - for (i = 0; i < dep_table->count; i++) { if (dep_table->entries[i].clk) { - clocks->data[clocks->num_levels].clocks_in_khz = - data->mclk_latency_table.entries - [data->mclk_latency_table.count].frequency = - dep_table->entries[i].clk; - clocks->data[clocks->num_levels].latency_in_us = - data->mclk_latency_table.entries - [data->mclk_latency_table.count].latency = - vega10_get_mem_latency(hwmgr, - dep_table->entries[i].clk); - clocks->num_levels++; - data->mclk_latency_table.count++; + + clocks->data[j].clocks_in_khz = + dep_table->entries[i].clk * 10; + data->mclk_latency_table.entries[j].frequency = + dep_table->entries[i].clk; + clocks->data[j].latency_in_us = + data->mclk_latency_table.entries[j].latency = 25; + j++; } } + clocks->num_levels = data->mclk_latency_table.count = j; } static void vega10_get_dcefclocks(struct pp_hwmgr *hwmgr, @@ -4084,7 +4122,7 @@ static void vega10_get_dcefclocks(struct pp_hwmgr *hwmgr, uint32_t i; for (i = 0; i < dep_table->count; i++) { - clocks->data[i].clocks_in_khz = dep_table->entries[i].clk; + clocks->data[i].clocks_in_khz = dep_table->entries[i].clk * 10; clocks->data[i].latency_in_us = 0; clocks->num_levels++; } @@ -4100,7 +4138,7 @@ static void vega10_get_socclocks(struct pp_hwmgr *hwmgr, uint32_t i; for (i = 0; i < dep_table->count; i++) { - clocks->data[i].clocks_in_khz = dep_table->entries[i].clk; + clocks->data[i].clocks_in_khz = dep_table->entries[i].clk * 10; clocks->data[i].latency_in_us = 0; clocks->num_levels++; } @@ -4160,7 +4198,7 @@ static int vega10_get_clock_by_type_with_voltage(struct pp_hwmgr *hwmgr, } for (i = 0; i < dep_table->count; i++) { - clocks->data[i].clocks_in_khz = dep_table->entries[i].clk; + clocks->data[i].clocks_in_khz = dep_table->entries[i].clk * 10; clocks->data[i].voltage_in_mv = (uint32_t)(table_info->vddc_lookup_table-> entries[dep_table->entries[i].vddInd].us_vdd); clocks->num_levels++; @@ -4173,9 +4211,10 @@ static int vega10_get_clock_by_type_with_voltage(struct pp_hwmgr *hwmgr, } static int vega10_set_watermarks_for_clocks_ranges(struct pp_hwmgr *hwmgr, - struct pp_wm_sets_with_clock_ranges_soc15 *wm_with_clock_ranges) + void *clock_range) { struct vega10_hwmgr *data = hwmgr->backend; + struct dm_pp_wm_sets_with_clock_ranges_soc15 *wm_with_clock_ranges = clock_range; Watermarks_t *table = &(data->smc_state_table.water_marks_table); int result = 0; @@ -4695,40 +4734,6 @@ static bool vega10_check_clk_voltage_valid(struct pp_hwmgr *hwmgr, return true; } -static void vega10_check_dpm_table_updated(struct pp_hwmgr *hwmgr) -{ - struct vega10_hwmgr *data = hwmgr->backend; - struct vega10_odn_dpm_table *odn_table = &(data->odn_dpm_table); - struct phm_ppt_v2_information *table_info = hwmgr->pptable; - struct phm_ppt_v1_clock_voltage_dependency_table *dep_table; - struct phm_ppt_v1_clock_voltage_dependency_table *odn_dep_table; - uint32_t i; - - dep_table = table_info->vdd_dep_on_mclk; - odn_dep_table = (struct phm_ppt_v1_clock_voltage_dependency_table *)&(odn_table->vdd_dep_on_mclk); - - for (i = 0; i < dep_table->count; i++) { - if (dep_table->entries[i].vddc != odn_dep_table->entries[i].vddc) { - data->need_update_dpm_table |= DPMTABLE_OD_UPDATE_VDDC | DPMTABLE_OD_UPDATE_MCLK; - return; - } - } - - dep_table = table_info->vdd_dep_on_sclk; - odn_dep_table = (struct phm_ppt_v1_clock_voltage_dependency_table *)&(odn_table->vdd_dep_on_sclk); - for (i = 0; i < dep_table->count; i++) { - if (dep_table->entries[i].vddc != odn_dep_table->entries[i].vddc) { - data->need_update_dpm_table |= DPMTABLE_OD_UPDATE_VDDC | DPMTABLE_OD_UPDATE_SCLK; - return; - } - } - - if (data->need_update_dpm_table & DPMTABLE_OD_UPDATE_VDDC) { - data->need_update_dpm_table &= ~DPMTABLE_OD_UPDATE_VDDC; - data->need_update_dpm_table |= DPMTABLE_OD_UPDATE_SCLK | DPMTABLE_OD_UPDATE_MCLK; - } -} - static void vega10_odn_update_soc_table(struct pp_hwmgr *hwmgr, enum PP_OD_DPM_TABLE_COMMAND type) { diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.h b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.h index aadd6cbc7e85..339820da9e6a 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.h +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.h @@ -370,7 +370,6 @@ struct vega10_hwmgr { /* ---- Power Gating States ---- */ bool uvd_power_gated; bool vce_power_gated; - bool samu_power_gated; bool need_long_memory_training; /* Internal settings to apply the application power optimization parameters */ diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c index c98e5de777cd..0789d64246ca 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c @@ -423,6 +423,11 @@ static int vega12_hwmgr_backend_init(struct pp_hwmgr *hwmgr) hwmgr->thermal_controller.advanceFanControlParameters.usFanPWMMinLimit * hwmgr->thermal_controller.fanInfo.ulMaxRPM / 100; + if (hwmgr->feature_mask & PP_GFXOFF_MASK) + data->gfxoff_controlled_by_driver = true; + else + data->gfxoff_controlled_by_driver = false; + return result; } @@ -454,43 +459,36 @@ static int vega12_setup_asic_task(struct pp_hwmgr *hwmgr) */ static void vega12_init_dpm_state(struct vega12_dpm_state *dpm_state) { - dpm_state->soft_min_level = 0xff; - dpm_state->soft_max_level = 0xff; - dpm_state->hard_min_level = 0xff; - dpm_state->hard_max_level = 0xff; + dpm_state->soft_min_level = 0x0; + dpm_state->soft_max_level = 0xffff; + dpm_state->hard_min_level = 0x0; + dpm_state->hard_max_level = 0xffff; } -static int vega12_get_number_dpm_level(struct pp_hwmgr *hwmgr, - PPCLK_e clkID, uint32_t *num_dpm_level) +static int vega12_get_number_of_dpm_level(struct pp_hwmgr *hwmgr, + PPCLK_e clk_id, uint32_t *num_of_levels) { - int result; - /* - * SMU expects the Clock ID to be in the top 16 bits. - * Lower 16 bits specify the level however 0xFF is a - * special argument the returns the total number of levels - */ - PP_ASSERT_WITH_CODE(smum_send_msg_to_smc_with_parameter(hwmgr, - PPSMC_MSG_GetDpmFreqByIndex, (clkID << 16 | 0xFF)) == 0, - "[GetNumberDpmLevel] Failed to get DPM levels from SMU for CLKID!", - return -EINVAL); - - result = vega12_read_arg_from_smc(hwmgr, num_dpm_level); + int ret = 0; - PP_ASSERT_WITH_CODE(*num_dpm_level < MAX_REGULAR_DPM_NUMBER, - "[GetNumberDPMLevel] Number of DPM levels is greater than limit", - return -EINVAL); + ret = smum_send_msg_to_smc_with_parameter(hwmgr, + PPSMC_MSG_GetDpmFreqByIndex, + (clk_id << 16 | 0xFF)); + PP_ASSERT_WITH_CODE(!ret, + "[GetNumOfDpmLevel] failed to get dpm levels!", + return ret); - PP_ASSERT_WITH_CODE(*num_dpm_level != 0, - "[GetNumberDPMLevel] Number of CLK Levels is zero!", - return -EINVAL); + *num_of_levels = smum_get_argument(hwmgr); + PP_ASSERT_WITH_CODE(*num_of_levels > 0, + "[GetNumOfDpmLevel] number of clk levels is invalid!", + return -EINVAL); - return result; + return ret; } static int vega12_get_dpm_frequency_by_index(struct pp_hwmgr *hwmgr, PPCLK_e clkID, uint32_t index, uint32_t *clock) { - int result; + int result = 0; /* *SMU expects the Clock ID to be in the top 16 bits. @@ -501,15 +499,36 @@ static int vega12_get_dpm_frequency_by_index(struct pp_hwmgr *hwmgr, "[GetDpmFrequencyByIndex] Failed to get dpm frequency from SMU!", return -EINVAL); - result = vega12_read_arg_from_smc(hwmgr, clock); - - PP_ASSERT_WITH_CODE(*clock != 0, - "[GetDPMFrequencyByIndex] Failed to get dpm frequency by index.!", - return -EINVAL); + *clock = smum_get_argument(hwmgr); return result; } +static int vega12_setup_single_dpm_table(struct pp_hwmgr *hwmgr, + struct vega12_single_dpm_table *dpm_table, PPCLK_e clk_id) +{ + int ret = 0; + uint32_t i, num_of_levels, clk; + + ret = vega12_get_number_of_dpm_level(hwmgr, clk_id, &num_of_levels); + PP_ASSERT_WITH_CODE(!ret, + "[SetupSingleDpmTable] failed to get clk levels!", + return ret); + + dpm_table->count = num_of_levels; + + for (i = 0; i < num_of_levels; i++) { + ret = vega12_get_dpm_frequency_by_index(hwmgr, clk_id, i, &clk); + PP_ASSERT_WITH_CODE(!ret, + "[SetupSingleDpmTable] failed to get clk of specific level!", + return ret); + dpm_table->dpm_levels[i].value = clk; + dpm_table->dpm_levels[i].enabled = true; + } + + return ret; +} + /* * This function is to initialize all DPM state tables * for SMU based on the dependency table. @@ -520,224 +539,136 @@ static int vega12_get_dpm_frequency_by_index(struct pp_hwmgr *hwmgr, */ static int vega12_setup_default_dpm_tables(struct pp_hwmgr *hwmgr) { - uint32_t num_levels, i, clock; struct vega12_hwmgr *data = (struct vega12_hwmgr *)(hwmgr->backend); - struct vega12_single_dpm_table *dpm_table; + int ret = 0; memset(&data->dpm_table, 0, sizeof(data->dpm_table)); - /* Initialize Sclk DPM and SOC DPM table based on allow Sclk values */ + /* socclk */ dpm_table = &(data->dpm_table.soc_table); - - PP_ASSERT_WITH_CODE(vega12_get_number_dpm_level(hwmgr, PPCLK_SOCCLK, - &num_levels) == 0, - "[SetupDefaultDPMTables] Failed to get DPM levels from SMU for SOCCLK!", - return -EINVAL); - - dpm_table->count = num_levels; - - for (i = 0; i < num_levels; i++) { - PP_ASSERT_WITH_CODE(vega12_get_dpm_frequency_by_index(hwmgr, - PPCLK_SOCCLK, i, &clock) == 0, - "[SetupDefaultDPMTables] Failed to get DPM levels from SMU for SOCCLK!", - return -EINVAL); - - dpm_table->dpm_levels[i].value = clock; - dpm_table->dpm_levels[i].enabled = true; + if (data->smu_features[GNLD_DPM_SOCCLK].enabled) { + ret = vega12_setup_single_dpm_table(hwmgr, dpm_table, PPCLK_SOCCLK); + PP_ASSERT_WITH_CODE(!ret, + "[SetupDefaultDpmTable] failed to get socclk dpm levels!", + return ret); + } else { + dpm_table->count = 1; + dpm_table->dpm_levels[0].value = data->vbios_boot_state.soc_clock / 100; } - vega12_init_dpm_state(&(dpm_table->dpm_state)); + /* gfxclk */ dpm_table = &(data->dpm_table.gfx_table); - - PP_ASSERT_WITH_CODE(vega12_get_number_dpm_level(hwmgr, PPCLK_GFXCLK, - &num_levels) == 0, - "[SetupDefaultDPMTables] Failed to get DPM levels from SMU for GFXCLK!", - return -EINVAL); - - dpm_table->count = num_levels; - for (i = 0; i < num_levels; i++) { - PP_ASSERT_WITH_CODE(vega12_get_dpm_frequency_by_index(hwmgr, - PPCLK_GFXCLK, i, &clock) == 0, - "[SetupDefaultDPMTables] Failed to get DPM levels from SMU for GFXCLK!", - return -EINVAL); - - dpm_table->dpm_levels[i].value = clock; - dpm_table->dpm_levels[i].enabled = true; + if (data->smu_features[GNLD_DPM_GFXCLK].enabled) { + ret = vega12_setup_single_dpm_table(hwmgr, dpm_table, PPCLK_GFXCLK); + PP_ASSERT_WITH_CODE(!ret, + "[SetupDefaultDpmTable] failed to get gfxclk dpm levels!", + return ret); + } else { + dpm_table->count = 1; + dpm_table->dpm_levels[0].value = data->vbios_boot_state.gfx_clock / 100; } - vega12_init_dpm_state(&(dpm_table->dpm_state)); - /* Initialize Mclk DPM table based on allow Mclk values */ - dpm_table = &(data->dpm_table.mem_table); - PP_ASSERT_WITH_CODE(vega12_get_number_dpm_level(hwmgr, PPCLK_UCLK, - &num_levels) == 0, - "[SetupDefaultDPMTables] Failed to get DPM levels from SMU for UCLK!", - return -EINVAL); - - dpm_table->count = num_levels; - - for (i = 0; i < num_levels; i++) { - PP_ASSERT_WITH_CODE(vega12_get_dpm_frequency_by_index(hwmgr, - PPCLK_UCLK, i, &clock) == 0, - "[SetupDefaultDPMTables] Failed to get DPM levels from SMU for UCLK!", - return -EINVAL); - - dpm_table->dpm_levels[i].value = clock; - dpm_table->dpm_levels[i].enabled = true; + /* memclk */ + dpm_table = &(data->dpm_table.mem_table); + if (data->smu_features[GNLD_DPM_UCLK].enabled) { + ret = vega12_setup_single_dpm_table(hwmgr, dpm_table, PPCLK_UCLK); + PP_ASSERT_WITH_CODE(!ret, + "[SetupDefaultDpmTable] failed to get memclk dpm levels!", + return ret); + } else { + dpm_table->count = 1; + dpm_table->dpm_levels[0].value = data->vbios_boot_state.mem_clock / 100; } - vega12_init_dpm_state(&(dpm_table->dpm_state)); + /* eclk */ dpm_table = &(data->dpm_table.eclk_table); - - PP_ASSERT_WITH_CODE(vega12_get_number_dpm_level(hwmgr, PPCLK_ECLK, - &num_levels) == 0, - "[SetupDefaultDPMTables] Failed to get DPM levels from SMU for ECLK!", - return -EINVAL); - - dpm_table->count = num_levels; - - for (i = 0; i < num_levels; i++) { - PP_ASSERT_WITH_CODE(vega12_get_dpm_frequency_by_index(hwmgr, - PPCLK_ECLK, i, &clock) == 0, - "[SetupDefaultDPMTables] Failed to get DPM levels from SMU for ECLK!", - return -EINVAL); - - dpm_table->dpm_levels[i].value = clock; - dpm_table->dpm_levels[i].enabled = true; + if (data->smu_features[GNLD_DPM_VCE].enabled) { + ret = vega12_setup_single_dpm_table(hwmgr, dpm_table, PPCLK_ECLK); + PP_ASSERT_WITH_CODE(!ret, + "[SetupDefaultDpmTable] failed to get eclk dpm levels!", + return ret); + } else { + dpm_table->count = 1; + dpm_table->dpm_levels[0].value = data->vbios_boot_state.eclock / 100; } - vega12_init_dpm_state(&(dpm_table->dpm_state)); + /* vclk */ dpm_table = &(data->dpm_table.vclk_table); - - PP_ASSERT_WITH_CODE(vega12_get_number_dpm_level(hwmgr, PPCLK_VCLK, - &num_levels) == 0, - "[SetupDefaultDPMTables] Failed to get DPM levels from SMU for VCLK!", - return -EINVAL); - - dpm_table->count = num_levels; - - for (i = 0; i < num_levels; i++) { - PP_ASSERT_WITH_CODE(vega12_get_dpm_frequency_by_index(hwmgr, - PPCLK_VCLK, i, &clock) == 0, - "[SetupDefaultDPMTables] Failed to get DPM levels from SMU for VCLK!", - return -EINVAL); - - dpm_table->dpm_levels[i].value = clock; - dpm_table->dpm_levels[i].enabled = true; + if (data->smu_features[GNLD_DPM_UVD].enabled) { + ret = vega12_setup_single_dpm_table(hwmgr, dpm_table, PPCLK_VCLK); + PP_ASSERT_WITH_CODE(!ret, + "[SetupDefaultDpmTable] failed to get vclk dpm levels!", + return ret); + } else { + dpm_table->count = 1; + dpm_table->dpm_levels[0].value = data->vbios_boot_state.vclock / 100; } - vega12_init_dpm_state(&(dpm_table->dpm_state)); + /* dclk */ dpm_table = &(data->dpm_table.dclk_table); - - PP_ASSERT_WITH_CODE(vega12_get_number_dpm_level(hwmgr, PPCLK_DCLK, - &num_levels) == 0, - "[SetupDefaultDPMTables] Failed to get DPM levels from SMU for DCLK!", - return -EINVAL); - - dpm_table->count = num_levels; - - for (i = 0; i < num_levels; i++) { - PP_ASSERT_WITH_CODE(vega12_get_dpm_frequency_by_index(hwmgr, - PPCLK_DCLK, i, &clock) == 0, - "[SetupDefaultDPMTables] Failed to get DPM levels from SMU for DCLK!", - return -EINVAL); - - dpm_table->dpm_levels[i].value = clock; - dpm_table->dpm_levels[i].enabled = true; + if (data->smu_features[GNLD_DPM_UVD].enabled) { + ret = vega12_setup_single_dpm_table(hwmgr, dpm_table, PPCLK_DCLK); + PP_ASSERT_WITH_CODE(!ret, + "[SetupDefaultDpmTable] failed to get dclk dpm levels!", + return ret); + } else { + dpm_table->count = 1; + dpm_table->dpm_levels[0].value = data->vbios_boot_state.dclock / 100; } - vega12_init_dpm_state(&(dpm_table->dpm_state)); - /* Assume there is no headless Vega12 for now */ + /* dcefclk */ dpm_table = &(data->dpm_table.dcef_table); - - PP_ASSERT_WITH_CODE(vega12_get_number_dpm_level(hwmgr, - PPCLK_DCEFCLK, &num_levels) == 0, - "[SetupDefaultDPMTables] Failed to get DPM levels from SMU for DCEFCLK!", - return -EINVAL); - - dpm_table->count = num_levels; - - for (i = 0; i < num_levels; i++) { - PP_ASSERT_WITH_CODE(vega12_get_dpm_frequency_by_index(hwmgr, - PPCLK_DCEFCLK, i, &clock) == 0, - "[SetupDefaultDPMTables] Failed to get DPM levels from SMU for DCEFCLK!", - return -EINVAL); - - dpm_table->dpm_levels[i].value = clock; - dpm_table->dpm_levels[i].enabled = true; + if (data->smu_features[GNLD_DPM_DCEFCLK].enabled) { + ret = vega12_setup_single_dpm_table(hwmgr, dpm_table, PPCLK_DCEFCLK); + PP_ASSERT_WITH_CODE(!ret, + "[SetupDefaultDpmTable] failed to get dcefclk dpm levels!", + return ret); + } else { + dpm_table->count = 1; + dpm_table->dpm_levels[0].value = data->vbios_boot_state.dcef_clock / 100; } - vega12_init_dpm_state(&(dpm_table->dpm_state)); + /* pixclk */ dpm_table = &(data->dpm_table.pixel_table); - - PP_ASSERT_WITH_CODE(vega12_get_number_dpm_level(hwmgr, - PPCLK_PIXCLK, &num_levels) == 0, - "[SetupDefaultDPMTables] Failed to get DPM levels from SMU for PIXCLK!", - return -EINVAL); - - dpm_table->count = num_levels; - - for (i = 0; i < num_levels; i++) { - PP_ASSERT_WITH_CODE(vega12_get_dpm_frequency_by_index(hwmgr, - PPCLK_PIXCLK, i, &clock) == 0, - "[SetupDefaultDPMTables] Failed to get DPM levels from SMU for PIXCLK!", - return -EINVAL); - - dpm_table->dpm_levels[i].value = clock; - dpm_table->dpm_levels[i].enabled = true; - } - + if (data->smu_features[GNLD_DPM_DCEFCLK].enabled) { + ret = vega12_setup_single_dpm_table(hwmgr, dpm_table, PPCLK_PIXCLK); + PP_ASSERT_WITH_CODE(!ret, + "[SetupDefaultDpmTable] failed to get pixclk dpm levels!", + return ret); + } else + dpm_table->count = 0; vega12_init_dpm_state(&(dpm_table->dpm_state)); + /* dispclk */ dpm_table = &(data->dpm_table.display_table); - - PP_ASSERT_WITH_CODE(vega12_get_number_dpm_level(hwmgr, - PPCLK_DISPCLK, &num_levels) == 0, - "[SetupDefaultDPMTables] Failed to get DPM levels from SMU for DISPCLK!", - return -EINVAL); - - dpm_table->count = num_levels; - - for (i = 0; i < num_levels; i++) { - PP_ASSERT_WITH_CODE(vega12_get_dpm_frequency_by_index(hwmgr, - PPCLK_DISPCLK, i, &clock) == 0, - "[SetupDefaultDPMTables] Failed to get DPM levels from SMU for DISPCLK!", - return -EINVAL); - - dpm_table->dpm_levels[i].value = clock; - dpm_table->dpm_levels[i].enabled = true; - } - + if (data->smu_features[GNLD_DPM_DCEFCLK].enabled) { + ret = vega12_setup_single_dpm_table(hwmgr, dpm_table, PPCLK_DISPCLK); + PP_ASSERT_WITH_CODE(!ret, + "[SetupDefaultDpmTable] failed to get dispclk dpm levels!", + return ret); + } else + dpm_table->count = 0; vega12_init_dpm_state(&(dpm_table->dpm_state)); + /* phyclk */ dpm_table = &(data->dpm_table.phy_table); - - PP_ASSERT_WITH_CODE(vega12_get_number_dpm_level(hwmgr, - PPCLK_PHYCLK, &num_levels) == 0, - "[SetupDefaultDPMTables] Failed to get DPM levels from SMU for PHYCLK!", - return -EINVAL); - - dpm_table->count = num_levels; - - for (i = 0; i < num_levels; i++) { - PP_ASSERT_WITH_CODE(vega12_get_dpm_frequency_by_index(hwmgr, - PPCLK_PHYCLK, i, &clock) == 0, - "[SetupDefaultDPMTables] Failed to get DPM levels from SMU for PHYCLK!", - return -EINVAL); - - dpm_table->dpm_levels[i].value = clock; - dpm_table->dpm_levels[i].enabled = true; - } - + if (data->smu_features[GNLD_DPM_DCEFCLK].enabled) { + ret = vega12_setup_single_dpm_table(hwmgr, dpm_table, PPCLK_PHYCLK); + PP_ASSERT_WITH_CODE(!ret, + "[SetupDefaultDpmTable] failed to get phyclk dpm levels!", + return ret); + } else + dpm_table->count = 0; vega12_init_dpm_state(&(dpm_table->dpm_state)); /* save a copy of the default DPM table */ @@ -848,6 +779,21 @@ static int vega12_set_allowed_featuresmask(struct pp_hwmgr *hwmgr) return 0; } +static void vega12_init_powergate_state(struct pp_hwmgr *hwmgr) +{ + struct vega12_hwmgr *data = + (struct vega12_hwmgr *)(hwmgr->backend); + + data->uvd_power_gated = true; + data->vce_power_gated = true; + + if (data->smu_features[GNLD_DPM_UVD].enabled) + data->uvd_power_gated = false; + + if (data->smu_features[GNLD_DPM_VCE].enabled) + data->vce_power_gated = false; +} + static int vega12_enable_all_smu_features(struct pp_hwmgr *hwmgr) { struct vega12_hwmgr *data = @@ -866,12 +812,11 @@ static int vega12_enable_all_smu_features(struct pp_hwmgr *hwmgr) enabled = (features_enabled & data->smu_features[i].smu_feature_bitmap) ? true : false; data->smu_features[i].enabled = enabled; data->smu_features[i].supported = enabled; - PP_ASSERT( - !data->smu_features[i].allowed || enabled, - "[EnableAllSMUFeatures] Enabled feature is different from allowed, expected disabled!"); } } + vega12_init_powergate_state(hwmgr); + return 0; } @@ -927,6 +872,48 @@ static int vega12_power_control_set_level(struct pp_hwmgr *hwmgr) return result; } +static int vega12_get_all_clock_ranges_helper(struct pp_hwmgr *hwmgr, + PPCLK_e clkid, struct vega12_clock_range *clock) +{ + /* AC Max */ + PP_ASSERT_WITH_CODE( + smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_GetMaxDpmFreq, (clkid << 16)) == 0, + "[GetClockRanges] Failed to get max ac clock from SMC!", + return -EINVAL); + clock->ACMax = smum_get_argument(hwmgr); + + /* AC Min */ + PP_ASSERT_WITH_CODE( + smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_GetMinDpmFreq, (clkid << 16)) == 0, + "[GetClockRanges] Failed to get min ac clock from SMC!", + return -EINVAL); + clock->ACMin = smum_get_argument(hwmgr); + + /* DC Max */ + PP_ASSERT_WITH_CODE( + smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_GetDcModeMaxDpmFreq, (clkid << 16)) == 0, + "[GetClockRanges] Failed to get max dc clock from SMC!", + return -EINVAL); + clock->DCMax = smum_get_argument(hwmgr); + + return 0; +} + +static int vega12_get_all_clock_ranges(struct pp_hwmgr *hwmgr) +{ + struct vega12_hwmgr *data = + (struct vega12_hwmgr *)(hwmgr->backend); + uint32_t i; + + for (i = 0; i < PPCLK_COUNT; i++) + PP_ASSERT_WITH_CODE(!vega12_get_all_clock_ranges_helper(hwmgr, + i, &(data->clk_range[i])), + "Failed to get clk range from SMC!", + return -EINVAL); + + return 0; +} + static int vega12_enable_dpm_tasks(struct pp_hwmgr *hwmgr) { int tmp_result, result = 0; @@ -954,6 +941,11 @@ static int vega12_enable_dpm_tasks(struct pp_hwmgr *hwmgr) "Failed to power control set level!", result = tmp_result); + result = vega12_get_all_clock_ranges(hwmgr); + PP_ASSERT_WITH_CODE(!result, + "Failed to get all clock ranges!", + return result); + result = vega12_odn_initialize_default_settings(hwmgr); PP_ASSERT_WITH_CODE(!result, "Failed to power control set level!", @@ -982,76 +974,172 @@ static uint32_t vega12_find_lowest_dpm_level( break; } + if (i >= table->count) { + i = 0; + table->dpm_levels[i].enabled = true; + } + return i; } static uint32_t vega12_find_highest_dpm_level( struct vega12_single_dpm_table *table) { - uint32_t i = 0; + int32_t i = 0; + PP_ASSERT_WITH_CODE(table->count <= MAX_REGULAR_DPM_NUMBER, + "[FindHighestDPMLevel] DPM Table has too many entries!", + return MAX_REGULAR_DPM_NUMBER - 1); - if (table->count <= MAX_REGULAR_DPM_NUMBER) { - for (i = table->count; i > 0; i--) { - if (table->dpm_levels[i - 1].enabled) - return i - 1; - } - } else { - pr_info("DPM Table Has Too Many Entries!"); - return MAX_REGULAR_DPM_NUMBER - 1; + for (i = table->count - 1; i >= 0; i--) { + if (table->dpm_levels[i].enabled) + break; } - return i; + if (i < 0) { + i = 0; + table->dpm_levels[i].enabled = true; + } + + return (uint32_t)i; } static int vega12_upload_dpm_min_level(struct pp_hwmgr *hwmgr) { struct vega12_hwmgr *data = hwmgr->backend; - if (data->smc_state_table.gfx_boot_level != - data->dpm_table.gfx_table.dpm_state.soft_min_level) { - smum_send_msg_to_smc_with_parameter(hwmgr, - PPSMC_MSG_SetSoftMinByFreq, - PPCLK_GFXCLK<<16 | data->dpm_table.gfx_table.dpm_levels[data->smc_state_table.gfx_boot_level].value); - data->dpm_table.gfx_table.dpm_state.soft_min_level = - data->smc_state_table.gfx_boot_level; + uint32_t min_freq; + int ret = 0; + + if (data->smu_features[GNLD_DPM_GFXCLK].enabled) { + min_freq = data->dpm_table.gfx_table.dpm_state.soft_min_level; + PP_ASSERT_WITH_CODE(!(ret = smum_send_msg_to_smc_with_parameter( + hwmgr, PPSMC_MSG_SetSoftMinByFreq, + (PPCLK_GFXCLK << 16) | (min_freq & 0xffff))), + "Failed to set soft min gfxclk !", + return ret); } - if (data->smc_state_table.mem_boot_level != - data->dpm_table.mem_table.dpm_state.soft_min_level) { - smum_send_msg_to_smc_with_parameter(hwmgr, - PPSMC_MSG_SetSoftMinByFreq, - PPCLK_UCLK<<16 | data->dpm_table.mem_table.dpm_levels[data->smc_state_table.mem_boot_level].value); - data->dpm_table.mem_table.dpm_state.soft_min_level = - data->smc_state_table.mem_boot_level; + if (data->smu_features[GNLD_DPM_UCLK].enabled) { + min_freq = data->dpm_table.mem_table.dpm_state.soft_min_level; + PP_ASSERT_WITH_CODE(!(ret = smum_send_msg_to_smc_with_parameter( + hwmgr, PPSMC_MSG_SetSoftMinByFreq, + (PPCLK_UCLK << 16) | (min_freq & 0xffff))), + "Failed to set soft min memclk !", + return ret); + + min_freq = data->dpm_table.mem_table.dpm_state.hard_min_level; + PP_ASSERT_WITH_CODE(!(ret = smum_send_msg_to_smc_with_parameter( + hwmgr, PPSMC_MSG_SetHardMinByFreq, + (PPCLK_UCLK << 16) | (min_freq & 0xffff))), + "Failed to set hard min memclk !", + return ret); } - return 0; + if (data->smu_features[GNLD_DPM_UVD].enabled) { + min_freq = data->dpm_table.vclk_table.dpm_state.soft_min_level; + + PP_ASSERT_WITH_CODE(!(ret = smum_send_msg_to_smc_with_parameter( + hwmgr, PPSMC_MSG_SetSoftMinByFreq, + (PPCLK_VCLK << 16) | (min_freq & 0xffff))), + "Failed to set soft min vclk!", + return ret); + + min_freq = data->dpm_table.dclk_table.dpm_state.soft_min_level; + + PP_ASSERT_WITH_CODE(!(ret = smum_send_msg_to_smc_with_parameter( + hwmgr, PPSMC_MSG_SetSoftMinByFreq, + (PPCLK_DCLK << 16) | (min_freq & 0xffff))), + "Failed to set soft min dclk!", + return ret); + } + + if (data->smu_features[GNLD_DPM_VCE].enabled) { + min_freq = data->dpm_table.eclk_table.dpm_state.soft_min_level; + + PP_ASSERT_WITH_CODE(!(ret = smum_send_msg_to_smc_with_parameter( + hwmgr, PPSMC_MSG_SetSoftMinByFreq, + (PPCLK_ECLK << 16) | (min_freq & 0xffff))), + "Failed to set soft min eclk!", + return ret); + } + + if (data->smu_features[GNLD_DPM_SOCCLK].enabled) { + min_freq = data->dpm_table.soc_table.dpm_state.soft_min_level; + + PP_ASSERT_WITH_CODE(!(ret = smum_send_msg_to_smc_with_parameter( + hwmgr, PPSMC_MSG_SetSoftMinByFreq, + (PPCLK_SOCCLK << 16) | (min_freq & 0xffff))), + "Failed to set soft min socclk!", + return ret); + } + + return ret; } static int vega12_upload_dpm_max_level(struct pp_hwmgr *hwmgr) { struct vega12_hwmgr *data = hwmgr->backend; - if (data->smc_state_table.gfx_max_level != - data->dpm_table.gfx_table.dpm_state.soft_max_level) { - smum_send_msg_to_smc_with_parameter(hwmgr, - PPSMC_MSG_SetSoftMaxByFreq, - /* plus the vale by 1 to align the resolution */ - PPCLK_GFXCLK<<16 | (data->dpm_table.gfx_table.dpm_levels[data->smc_state_table.gfx_max_level].value + 1)); - data->dpm_table.gfx_table.dpm_state.soft_max_level = - data->smc_state_table.gfx_max_level; + uint32_t max_freq; + int ret = 0; + + if (data->smu_features[GNLD_DPM_GFXCLK].enabled) { + max_freq = data->dpm_table.gfx_table.dpm_state.soft_max_level; + + PP_ASSERT_WITH_CODE(!(ret = smum_send_msg_to_smc_with_parameter( + hwmgr, PPSMC_MSG_SetSoftMaxByFreq, + (PPCLK_GFXCLK << 16) | (max_freq & 0xffff))), + "Failed to set soft max gfxclk!", + return ret); } - if (data->smc_state_table.mem_max_level != - data->dpm_table.mem_table.dpm_state.soft_max_level) { - smum_send_msg_to_smc_with_parameter(hwmgr, - PPSMC_MSG_SetSoftMaxByFreq, - /* plus the vale by 1 to align the resolution */ - PPCLK_UCLK<<16 | (data->dpm_table.mem_table.dpm_levels[data->smc_state_table.mem_max_level].value + 1)); - data->dpm_table.mem_table.dpm_state.soft_max_level = - data->smc_state_table.mem_max_level; + if (data->smu_features[GNLD_DPM_UCLK].enabled) { + max_freq = data->dpm_table.mem_table.dpm_state.soft_max_level; + + PP_ASSERT_WITH_CODE(!(ret = smum_send_msg_to_smc_with_parameter( + hwmgr, PPSMC_MSG_SetSoftMaxByFreq, + (PPCLK_UCLK << 16) | (max_freq & 0xffff))), + "Failed to set soft max memclk!", + return ret); } - return 0; + if (data->smu_features[GNLD_DPM_UVD].enabled) { + max_freq = data->dpm_table.vclk_table.dpm_state.soft_max_level; + + PP_ASSERT_WITH_CODE(!(ret = smum_send_msg_to_smc_with_parameter( + hwmgr, PPSMC_MSG_SetSoftMaxByFreq, + (PPCLK_VCLK << 16) | (max_freq & 0xffff))), + "Failed to set soft max vclk!", + return ret); + + max_freq = data->dpm_table.dclk_table.dpm_state.soft_max_level; + PP_ASSERT_WITH_CODE(!(ret = smum_send_msg_to_smc_with_parameter( + hwmgr, PPSMC_MSG_SetSoftMaxByFreq, + (PPCLK_DCLK << 16) | (max_freq & 0xffff))), + "Failed to set soft max dclk!", + return ret); + } + + if (data->smu_features[GNLD_DPM_VCE].enabled) { + max_freq = data->dpm_table.eclk_table.dpm_state.soft_max_level; + + PP_ASSERT_WITH_CODE(!(ret = smum_send_msg_to_smc_with_parameter( + hwmgr, PPSMC_MSG_SetSoftMaxByFreq, + (PPCLK_ECLK << 16) | (max_freq & 0xffff))), + "Failed to set soft max eclk!", + return ret); + } + + if (data->smu_features[GNLD_DPM_SOCCLK].enabled) { + max_freq = data->dpm_table.soc_table.dpm_state.soft_max_level; + + PP_ASSERT_WITH_CODE(!(ret = smum_send_msg_to_smc_with_parameter( + hwmgr, PPSMC_MSG_SetSoftMaxByFreq, + (PPCLK_SOCCLK << 16) | (max_freq & 0xffff))), + "Failed to set soft max socclk!", + return ret); + } + + return ret; } int vega12_enable_disable_vce_dpm(struct pp_hwmgr *hwmgr, bool enable) @@ -1127,7 +1215,7 @@ static int vega12_get_gpu_power(struct pp_hwmgr *hwmgr, uint32_t *query) "Failed to get current package power!", return -EINVAL); - vega12_read_arg_from_smc(hwmgr, &value); + value = smum_get_argument(hwmgr); /* power value is an integer */ *query = value << 8; #endif @@ -1140,14 +1228,11 @@ static int vega12_get_current_gfx_clk_freq(struct pp_hwmgr *hwmgr, uint32_t *gfx *gfx_freq = 0; - PP_ASSERT_WITH_CODE( - smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_GetDpmClockFreq, (PPCLK_GFXCLK << 16)) == 0, + PP_ASSERT_WITH_CODE(smum_send_msg_to_smc_with_parameter(hwmgr, + PPSMC_MSG_GetDpmClockFreq, (PPCLK_GFXCLK << 16)) == 0, "[GetCurrentGfxClkFreq] Attempt to get Current GFXCLK Frequency Failed!", - return -1); - PP_ASSERT_WITH_CODE( - vega12_read_arg_from_smc(hwmgr, &gfx_clk) == 0, - "[GetCurrentGfxClkFreq] Attempt to read arg from SMC Failed", - return -1); + return -EINVAL); + gfx_clk = smum_get_argument(hwmgr); *gfx_freq = gfx_clk * 100; @@ -1163,11 +1248,8 @@ static int vega12_get_current_mclk_freq(struct pp_hwmgr *hwmgr, uint32_t *mclk_f PP_ASSERT_WITH_CODE( smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_GetDpmClockFreq, (PPCLK_UCLK << 16)) == 0, "[GetCurrentMClkFreq] Attempt to get Current MCLK Frequency Failed!", - return -1); - PP_ASSERT_WITH_CODE( - vega12_read_arg_from_smc(hwmgr, &mem_clk) == 0, - "[GetCurrentMClkFreq] Attempt to read arg from SMC Failed", - return -1); + return -EINVAL); + mem_clk = smum_get_argument(hwmgr); *mclk_freq = mem_clk * 100; @@ -1184,16 +1266,12 @@ static int vega12_get_current_activity_percent( #if 0 ret = smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_GetAverageGfxActivity, 0); if (!ret) { - ret = vega12_read_arg_from_smc(hwmgr, ¤t_activity); - if (!ret) { - if (current_activity > 100) { - PP_ASSERT(false, - "[GetCurrentActivityPercent] Activity Percentage Exceeds 100!"); - current_activity = 100; - } - } else + current_activity = smum_get_argument(hwmgr); + if (current_activity > 100) { PP_ASSERT(false, - "[GetCurrentActivityPercent] Attempt To Read Average Graphics Activity from SMU Failed!"); + "[GetCurrentActivityPercent] Activity Percentage Exceeds 100!"); + current_activity = 100; + } } else PP_ASSERT(false, "[GetCurrentActivityPercent] Attempt To Send Get Average Graphics Activity to SMU Failed!"); @@ -1256,7 +1334,7 @@ static int vega12_notify_smc_display_change(struct pp_hwmgr *hwmgr, if (data->smu_features[GNLD_DPM_UCLK].enabled) return smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_SetUclkFastSwitch, - has_disp ? 0 : 1); + has_disp ? 1 : 0); return 0; } @@ -1274,7 +1352,6 @@ int vega12_display_clock_voltage_request(struct pp_hwmgr *hwmgr, if (data->smu_features[GNLD_DPM_DCEFCLK].enabled) { switch (clk_type) { case amd_pp_dcef_clock: - clk_freq = clock_req->clock_freq_in_khz / 100; clk_select = PPCLK_DCEFCLK; break; case amd_pp_disp_clock: @@ -1310,9 +1387,10 @@ static int vega12_notify_smc_display_config_after_ps_adjustment( (struct vega12_hwmgr *)(hwmgr->backend); struct PP_Clocks min_clocks = {0}; struct pp_display_clock_request clock_req; - uint32_t clk_request; - if (hwmgr->display_config->num_display > 1) + if ((hwmgr->display_config->num_display > 1) && + !hwmgr->display_config->multi_monitor_in_sync && + !hwmgr->display_config->nb_pstate_switch_disable) vega12_notify_smc_display_change(hwmgr, false); else vega12_notify_smc_display_change(hwmgr, true); @@ -1323,7 +1401,7 @@ static int vega12_notify_smc_display_config_after_ps_adjustment( if (data->smu_features[GNLD_DPM_DCEFCLK].supported) { clock_req.clock_type = amd_pp_dcef_clock; - clock_req.clock_freq_in_khz = min_clocks.dcefClock; + clock_req.clock_freq_in_khz = min_clocks.dcefClock/10; if (!vega12_display_clock_voltage_request(hwmgr, &clock_req)) { if (data->smu_features[GNLD_DS_DCEFCLK].supported) PP_ASSERT_WITH_CODE( @@ -1337,15 +1415,6 @@ static int vega12_notify_smc_display_config_after_ps_adjustment( } } - if (data->smu_features[GNLD_DPM_UCLK].enabled) { - clk_request = (PPCLK_UCLK << 16) | (min_clocks.memoryClock) / 100; - PP_ASSERT_WITH_CODE( - smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_SetHardMinByFreq, clk_request) == 0, - "[PhwVega12_NotifySMCDisplayConfigAfterPowerStateAdjustment] Attempt to set UCLK HardMin Failed!", - return -1); - data->dpm_table.mem_table.dpm_state.hard_min_level = min_clocks.memoryClock; - } - return 0; } @@ -1354,12 +1423,19 @@ static int vega12_force_dpm_highest(struct pp_hwmgr *hwmgr) struct vega12_hwmgr *data = (struct vega12_hwmgr *)(hwmgr->backend); - data->smc_state_table.gfx_boot_level = - data->smc_state_table.gfx_max_level = - vega12_find_highest_dpm_level(&(data->dpm_table.gfx_table)); - data->smc_state_table.mem_boot_level = - data->smc_state_table.mem_max_level = - vega12_find_highest_dpm_level(&(data->dpm_table.mem_table)); + uint32_t soft_level; + + soft_level = vega12_find_highest_dpm_level(&(data->dpm_table.gfx_table)); + + data->dpm_table.gfx_table.dpm_state.soft_min_level = + data->dpm_table.gfx_table.dpm_state.soft_max_level = + data->dpm_table.gfx_table.dpm_levels[soft_level].value; + + soft_level = vega12_find_highest_dpm_level(&(data->dpm_table.mem_table)); + + data->dpm_table.mem_table.dpm_state.soft_min_level = + data->dpm_table.mem_table.dpm_state.soft_max_level = + data->dpm_table.mem_table.dpm_levels[soft_level].value; PP_ASSERT_WITH_CODE(!vega12_upload_dpm_min_level(hwmgr), "Failed to upload boot level to highest!", @@ -1376,13 +1452,19 @@ static int vega12_force_dpm_lowest(struct pp_hwmgr *hwmgr) { struct vega12_hwmgr *data = (struct vega12_hwmgr *)(hwmgr->backend); + uint32_t soft_level; + + soft_level = vega12_find_lowest_dpm_level(&(data->dpm_table.gfx_table)); + + data->dpm_table.gfx_table.dpm_state.soft_min_level = + data->dpm_table.gfx_table.dpm_state.soft_max_level = + data->dpm_table.gfx_table.dpm_levels[soft_level].value; - data->smc_state_table.gfx_boot_level = - data->smc_state_table.gfx_max_level = - vega12_find_lowest_dpm_level(&(data->dpm_table.gfx_table)); - data->smc_state_table.mem_boot_level = - data->smc_state_table.mem_max_level = - vega12_find_lowest_dpm_level(&(data->dpm_table.mem_table)); + soft_level = vega12_find_lowest_dpm_level(&(data->dpm_table.mem_table)); + + data->dpm_table.mem_table.dpm_state.soft_min_level = + data->dpm_table.mem_table.dpm_state.soft_max_level = + data->dpm_table.mem_table.dpm_levels[soft_level].value; PP_ASSERT_WITH_CODE(!vega12_upload_dpm_min_level(hwmgr), "Failed to upload boot level to highest!", @@ -1398,17 +1480,6 @@ static int vega12_force_dpm_lowest(struct pp_hwmgr *hwmgr) static int vega12_unforce_dpm_levels(struct pp_hwmgr *hwmgr) { - struct vega12_hwmgr *data = (struct vega12_hwmgr *)(hwmgr->backend); - - data->smc_state_table.gfx_boot_level = - vega12_find_lowest_dpm_level(&(data->dpm_table.gfx_table)); - data->smc_state_table.gfx_max_level = - vega12_find_highest_dpm_level(&(data->dpm_table.gfx_table)); - data->smc_state_table.mem_boot_level = - vega12_find_lowest_dpm_level(&(data->dpm_table.mem_table)); - data->smc_state_table.mem_max_level = - vega12_find_highest_dpm_level(&(data->dpm_table.mem_table)); - PP_ASSERT_WITH_CODE(!vega12_upload_dpm_min_level(hwmgr), "Failed to upload DPM Bootup Levels!", return -1); @@ -1416,22 +1487,28 @@ static int vega12_unforce_dpm_levels(struct pp_hwmgr *hwmgr) PP_ASSERT_WITH_CODE(!vega12_upload_dpm_max_level(hwmgr), "Failed to upload DPM Max Levels!", return -1); + return 0; } -#if 0 static int vega12_get_profiling_clk_mask(struct pp_hwmgr *hwmgr, enum amd_dpm_forced_level level, uint32_t *sclk_mask, uint32_t *mclk_mask, uint32_t *soc_mask) { - struct phm_ppt_v2_information *table_info = - (struct phm_ppt_v2_information *)(hwmgr->pptable); + struct vega12_hwmgr *data = (struct vega12_hwmgr *)(hwmgr->backend); + struct vega12_single_dpm_table *gfx_dpm_table = &(data->dpm_table.gfx_table); + struct vega12_single_dpm_table *mem_dpm_table = &(data->dpm_table.mem_table); + struct vega12_single_dpm_table *soc_dpm_table = &(data->dpm_table.soc_table); - if (table_info->vdd_dep_on_sclk->count > VEGA12_UMD_PSTATE_GFXCLK_LEVEL && - table_info->vdd_dep_on_socclk->count > VEGA12_UMD_PSTATE_SOCCLK_LEVEL && - table_info->vdd_dep_on_mclk->count > VEGA12_UMD_PSTATE_MCLK_LEVEL) { + *sclk_mask = 0; + *mclk_mask = 0; + *soc_mask = 0; + + if (gfx_dpm_table->count > VEGA12_UMD_PSTATE_GFXCLK_LEVEL && + mem_dpm_table->count > VEGA12_UMD_PSTATE_MCLK_LEVEL && + soc_dpm_table->count > VEGA12_UMD_PSTATE_SOCCLK_LEVEL) { *sclk_mask = VEGA12_UMD_PSTATE_GFXCLK_LEVEL; - *soc_mask = VEGA12_UMD_PSTATE_SOCCLK_LEVEL; *mclk_mask = VEGA12_UMD_PSTATE_MCLK_LEVEL; + *soc_mask = VEGA12_UMD_PSTATE_SOCCLK_LEVEL; } if (level == AMD_DPM_FORCED_LEVEL_PROFILE_MIN_SCLK) { @@ -1439,13 +1516,13 @@ static int vega12_get_profiling_clk_mask(struct pp_hwmgr *hwmgr, enum amd_dpm_fo } else if (level == AMD_DPM_FORCED_LEVEL_PROFILE_MIN_MCLK) { *mclk_mask = 0; } else if (level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK) { - *sclk_mask = table_info->vdd_dep_on_sclk->count - 1; - *soc_mask = table_info->vdd_dep_on_socclk->count - 1; - *mclk_mask = table_info->vdd_dep_on_mclk->count - 1; + *sclk_mask = gfx_dpm_table->count - 1; + *mclk_mask = mem_dpm_table->count - 1; + *soc_mask = soc_dpm_table->count - 1; } + return 0; } -#endif static void vega12_set_fan_control_mode(struct pp_hwmgr *hwmgr, uint32_t mode) { @@ -1469,11 +1546,9 @@ static int vega12_dpm_force_dpm_level(struct pp_hwmgr *hwmgr, enum amd_dpm_forced_level level) { int ret = 0; -#if 0 uint32_t sclk_mask = 0; uint32_t mclk_mask = 0; uint32_t soc_mask = 0; -#endif switch (level) { case AMD_DPM_FORCED_LEVEL_HIGH: @@ -1489,27 +1564,18 @@ static int vega12_dpm_force_dpm_level(struct pp_hwmgr *hwmgr, case AMD_DPM_FORCED_LEVEL_PROFILE_MIN_SCLK: case AMD_DPM_FORCED_LEVEL_PROFILE_MIN_MCLK: case AMD_DPM_FORCED_LEVEL_PROFILE_PEAK: -#if 0 ret = vega12_get_profiling_clk_mask(hwmgr, level, &sclk_mask, &mclk_mask, &soc_mask); if (ret) return ret; - vega12_force_clock_level(hwmgr, PP_SCLK, 1<<sclk_mask); - vega12_force_clock_level(hwmgr, PP_MCLK, 1<<mclk_mask); -#endif + vega12_force_clock_level(hwmgr, PP_SCLK, 1 << sclk_mask); + vega12_force_clock_level(hwmgr, PP_MCLK, 1 << mclk_mask); break; case AMD_DPM_FORCED_LEVEL_MANUAL: case AMD_DPM_FORCED_LEVEL_PROFILE_EXIT: default: break; } -#if 0 - if (!ret) { - if (level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK && hwmgr->dpm_level != AMD_DPM_FORCED_LEVEL_PROFILE_PEAK) - vega12_set_fan_control_mode(hwmgr, AMD_FAN_CTRL_NONE); - else if (level != AMD_DPM_FORCED_LEVEL_PROFILE_PEAK && hwmgr->dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK) - vega12_set_fan_control_mode(hwmgr, AMD_FAN_CTRL_AUTO); - } -#endif + return ret; } @@ -1543,24 +1609,14 @@ static int vega12_get_clock_ranges(struct pp_hwmgr *hwmgr, PPCLK_e clock_select, bool max) { - int result; - *clock = 0; + struct vega12_hwmgr *data = (struct vega12_hwmgr *)(hwmgr->backend); - if (max) { - PP_ASSERT_WITH_CODE( - smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_GetMaxDpmFreq, (clock_select << 16)) == 0, - "[GetClockRanges] Failed to get max clock from SMC!", - return -1); - result = vega12_read_arg_from_smc(hwmgr, clock); - } else { - PP_ASSERT_WITH_CODE( - smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_GetMinDpmFreq, (clock_select << 16)) == 0, - "[GetClockRanges] Failed to get min clock from SMC!", - return -1); - result = vega12_read_arg_from_smc(hwmgr, clock); - } + if (max) + *clock = data->clk_range[clock_select].ACMax; + else + *clock = data->clk_range[clock_select].ACMin; - return result; + return 0; } static int vega12_get_sclks(struct pp_hwmgr *hwmgr, @@ -1575,12 +1631,12 @@ static int vega12_get_sclks(struct pp_hwmgr *hwmgr, return -1; dpm_table = &(data->dpm_table.gfx_table); - ucount = (dpm_table->count > VG12_PSUEDO_NUM_GFXCLK_DPM_LEVELS) ? - VG12_PSUEDO_NUM_GFXCLK_DPM_LEVELS : dpm_table->count; + ucount = (dpm_table->count > MAX_NUM_CLOCKS) ? + MAX_NUM_CLOCKS : dpm_table->count; for (i = 0; i < ucount; i++) { clocks->data[i].clocks_in_khz = - dpm_table->dpm_levels[i].value * 100; + dpm_table->dpm_levels[i].value * 1000; clocks->data[i].latency_in_us = 0; } @@ -1607,13 +1663,12 @@ static int vega12_get_memclocks(struct pp_hwmgr *hwmgr, return -1; dpm_table = &(data->dpm_table.mem_table); - ucount = (dpm_table->count > VG12_PSUEDO_NUM_UCLK_DPM_LEVELS) ? - VG12_PSUEDO_NUM_UCLK_DPM_LEVELS : dpm_table->count; + ucount = (dpm_table->count > MAX_NUM_CLOCKS) ? + MAX_NUM_CLOCKS : dpm_table->count; for (i = 0; i < ucount; i++) { - clocks->data[i].clocks_in_khz = - dpm_table->dpm_levels[i].value * 100; - + clocks->data[i].clocks_in_khz = dpm_table->dpm_levels[i].value * 1000; + data->mclk_latency_table.entries[i].frequency = dpm_table->dpm_levels[i].value * 100; clocks->data[i].latency_in_us = data->mclk_latency_table.entries[i].latency = vega12_get_mem_latency(hwmgr, dpm_table->dpm_levels[i].value); @@ -1637,12 +1692,12 @@ static int vega12_get_dcefclocks(struct pp_hwmgr *hwmgr, dpm_table = &(data->dpm_table.dcef_table); - ucount = (dpm_table->count > VG12_PSUEDO_NUM_DCEFCLK_DPM_LEVELS) ? - VG12_PSUEDO_NUM_DCEFCLK_DPM_LEVELS : dpm_table->count; + ucount = (dpm_table->count > MAX_NUM_CLOCKS) ? + MAX_NUM_CLOCKS : dpm_table->count; for (i = 0; i < ucount; i++) { clocks->data[i].clocks_in_khz = - dpm_table->dpm_levels[i].value * 100; + dpm_table->dpm_levels[i].value * 1000; clocks->data[i].latency_in_us = 0; } @@ -1665,12 +1720,12 @@ static int vega12_get_socclocks(struct pp_hwmgr *hwmgr, dpm_table = &(data->dpm_table.soc_table); - ucount = (dpm_table->count > VG12_PSUEDO_NUM_SOCCLK_DPM_LEVELS) ? - VG12_PSUEDO_NUM_SOCCLK_DPM_LEVELS : dpm_table->count; + ucount = (dpm_table->count > MAX_NUM_CLOCKS) ? + MAX_NUM_CLOCKS : dpm_table->count; for (i = 0; i < ucount; i++) { clocks->data[i].clocks_in_khz = - dpm_table->dpm_levels[i].value * 100; + dpm_table->dpm_levels[i].value * 1000; clocks->data[i].latency_in_us = 0; } @@ -1717,99 +1772,69 @@ static int vega12_get_clock_by_type_with_voltage(struct pp_hwmgr *hwmgr, } static int vega12_set_watermarks_for_clocks_ranges(struct pp_hwmgr *hwmgr, - struct pp_wm_sets_with_clock_ranges_soc15 *wm_with_clock_ranges) + void *clock_ranges) { struct vega12_hwmgr *data = (struct vega12_hwmgr *)(hwmgr->backend); Watermarks_t *table = &(data->smc_state_table.water_marks_table); - int result = 0; - uint32_t i; + struct dm_pp_wm_sets_with_clock_ranges_soc15 *wm_with_clock_ranges = clock_ranges; if (!data->registry_data.disable_water_mark && data->smu_features[GNLD_DPM_DCEFCLK].supported && data->smu_features[GNLD_DPM_SOCCLK].supported) { - for (i = 0; i < wm_with_clock_ranges->num_wm_sets_dmif; i++) { - table->WatermarkRow[WM_DCEFCLK][i].MinClock = - cpu_to_le16((uint16_t) - (wm_with_clock_ranges->wm_sets_dmif[i].wm_min_dcefclk_in_khz) / - 100); - table->WatermarkRow[WM_DCEFCLK][i].MaxClock = - cpu_to_le16((uint16_t) - (wm_with_clock_ranges->wm_sets_dmif[i].wm_max_dcefclk_in_khz) / - 100); - table->WatermarkRow[WM_DCEFCLK][i].MinUclk = - cpu_to_le16((uint16_t) - (wm_with_clock_ranges->wm_sets_dmif[i].wm_min_memclk_in_khz) / - 100); - table->WatermarkRow[WM_DCEFCLK][i].MaxUclk = - cpu_to_le16((uint16_t) - (wm_with_clock_ranges->wm_sets_dmif[i].wm_max_memclk_in_khz) / - 100); - table->WatermarkRow[WM_DCEFCLK][i].WmSetting = (uint8_t) - wm_with_clock_ranges->wm_sets_dmif[i].wm_set_id; - } - - for (i = 0; i < wm_with_clock_ranges->num_wm_sets_mcif; i++) { - table->WatermarkRow[WM_SOCCLK][i].MinClock = - cpu_to_le16((uint16_t) - (wm_with_clock_ranges->wm_sets_mcif[i].wm_min_socclk_in_khz) / - 100); - table->WatermarkRow[WM_SOCCLK][i].MaxClock = - cpu_to_le16((uint16_t) - (wm_with_clock_ranges->wm_sets_mcif[i].wm_max_socclk_in_khz) / - 100); - table->WatermarkRow[WM_SOCCLK][i].MinUclk = - cpu_to_le16((uint16_t) - (wm_with_clock_ranges->wm_sets_mcif[i].wm_min_memclk_in_khz) / - 100); - table->WatermarkRow[WM_SOCCLK][i].MaxUclk = - cpu_to_le16((uint16_t) - (wm_with_clock_ranges->wm_sets_mcif[i].wm_max_memclk_in_khz) / - 100); - table->WatermarkRow[WM_SOCCLK][i].WmSetting = (uint8_t) - wm_with_clock_ranges->wm_sets_mcif[i].wm_set_id; - } + smu_set_watermarks_for_clocks_ranges(table, wm_with_clock_ranges); data->water_marks_bitmap |= WaterMarksExist; data->water_marks_bitmap &= ~WaterMarksLoaded; } - return result; + return 0; } static int vega12_force_clock_level(struct pp_hwmgr *hwmgr, enum pp_clock_type type, uint32_t mask) { struct vega12_hwmgr *data = (struct vega12_hwmgr *)(hwmgr->backend); - - if (hwmgr->request_dpm_level & (AMD_DPM_FORCED_LEVEL_AUTO | - AMD_DPM_FORCED_LEVEL_LOW | - AMD_DPM_FORCED_LEVEL_HIGH)) - return -EINVAL; + uint32_t soft_min_level, soft_max_level; + int ret = 0; switch (type) { case PP_SCLK: - data->smc_state_table.gfx_boot_level = mask ? (ffs(mask) - 1) : 0; - data->smc_state_table.gfx_max_level = mask ? (fls(mask) - 1) : 0; + soft_min_level = mask ? (ffs(mask) - 1) : 0; + soft_max_level = mask ? (fls(mask) - 1) : 0; + + data->dpm_table.gfx_table.dpm_state.soft_min_level = + data->dpm_table.gfx_table.dpm_levels[soft_min_level].value; + data->dpm_table.gfx_table.dpm_state.soft_max_level = + data->dpm_table.gfx_table.dpm_levels[soft_max_level].value; - PP_ASSERT_WITH_CODE(!vega12_upload_dpm_min_level(hwmgr), + ret = vega12_upload_dpm_min_level(hwmgr); + PP_ASSERT_WITH_CODE(!ret, "Failed to upload boot level to lowest!", - return -EINVAL); + return ret); - PP_ASSERT_WITH_CODE(!vega12_upload_dpm_max_level(hwmgr), + ret = vega12_upload_dpm_max_level(hwmgr); + PP_ASSERT_WITH_CODE(!ret, "Failed to upload dpm max level to highest!", - return -EINVAL); + return ret); break; case PP_MCLK: - data->smc_state_table.mem_boot_level = mask ? (ffs(mask) - 1) : 0; - data->smc_state_table.mem_max_level = mask ? (fls(mask) - 1) : 0; + soft_min_level = mask ? (ffs(mask) - 1) : 0; + soft_max_level = mask ? (fls(mask) - 1) : 0; - PP_ASSERT_WITH_CODE(!vega12_upload_dpm_min_level(hwmgr), + data->dpm_table.mem_table.dpm_state.soft_min_level = + data->dpm_table.mem_table.dpm_levels[soft_min_level].value; + data->dpm_table.mem_table.dpm_state.soft_max_level = + data->dpm_table.mem_table.dpm_levels[soft_max_level].value; + + ret = vega12_upload_dpm_min_level(hwmgr); + PP_ASSERT_WITH_CODE(!ret, "Failed to upload boot level to lowest!", - return -EINVAL); + return ret); - PP_ASSERT_WITH_CODE(!vega12_upload_dpm_max_level(hwmgr), + ret = vega12_upload_dpm_max_level(hwmgr); + PP_ASSERT_WITH_CODE(!ret, "Failed to upload dpm max level to highest!", - return -EINVAL); + return ret); break; @@ -1842,8 +1867,8 @@ static int vega12_print_clock_levels(struct pp_hwmgr *hwmgr, return -1); for (i = 0; i < clocks.num_levels; i++) size += sprintf(buf + size, "%d: %uMhz %s\n", - i, clocks.data[i].clocks_in_khz / 100, - (clocks.data[i].clocks_in_khz == now) ? "*" : ""); + i, clocks.data[i].clocks_in_khz / 1000, + (clocks.data[i].clocks_in_khz / 1000 == now / 100) ? "*" : ""); break; case PP_MCLK: @@ -1858,8 +1883,8 @@ static int vega12_print_clock_levels(struct pp_hwmgr *hwmgr, return -1); for (i = 0; i < clocks.num_levels; i++) size += sprintf(buf + size, "%d: %uMhz %s\n", - i, clocks.data[i].clocks_in_khz / 100, - (clocks.data[i].clocks_in_khz == now) ? "*" : ""); + i, clocks.data[i].clocks_in_khz / 1000, + (clocks.data[i].clocks_in_khz / 1000 == now / 100) ? "*" : ""); break; case PP_PCIE: @@ -1871,6 +1896,205 @@ static int vega12_print_clock_levels(struct pp_hwmgr *hwmgr, return size; } +static int vega12_apply_clocks_adjust_rules(struct pp_hwmgr *hwmgr) +{ + struct vega12_hwmgr *data = (struct vega12_hwmgr *)(hwmgr->backend); + struct vega12_single_dpm_table *dpm_table; + bool vblank_too_short = false; + bool disable_mclk_switching; + uint32_t i, latency; + + disable_mclk_switching = ((1 < hwmgr->display_config->num_display) && + !hwmgr->display_config->multi_monitor_in_sync) || + vblank_too_short; + latency = hwmgr->display_config->dce_tolerable_mclk_in_active_latency; + + /* gfxclk */ + dpm_table = &(data->dpm_table.gfx_table); + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[0].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[0].value; + dpm_table->dpm_state.hard_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + + if (PP_CAP(PHM_PlatformCaps_UMDPState)) { + if (VEGA12_UMD_PSTATE_GFXCLK_LEVEL < dpm_table->count) { + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[VEGA12_UMD_PSTATE_GFXCLK_LEVEL].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[VEGA12_UMD_PSTATE_GFXCLK_LEVEL].value; + } + + if (hwmgr->dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_MIN_SCLK) { + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[0].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[0].value; + } + + if (hwmgr->dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK) { + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + } + } + + /* memclk */ + dpm_table = &(data->dpm_table.mem_table); + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[0].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[0].value; + dpm_table->dpm_state.hard_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + + if (PP_CAP(PHM_PlatformCaps_UMDPState)) { + if (VEGA12_UMD_PSTATE_MCLK_LEVEL < dpm_table->count) { + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[VEGA12_UMD_PSTATE_MCLK_LEVEL].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[VEGA12_UMD_PSTATE_MCLK_LEVEL].value; + } + + if (hwmgr->dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_MIN_MCLK) { + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[0].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[0].value; + } + + if (hwmgr->dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK) { + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + } + } + + /* honour DAL's UCLK Hardmin */ + if (dpm_table->dpm_state.hard_min_level < (hwmgr->display_config->min_mem_set_clock / 100)) + dpm_table->dpm_state.hard_min_level = hwmgr->display_config->min_mem_set_clock / 100; + + /* Hardmin is dependent on displayconfig */ + if (disable_mclk_switching) { + dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + for (i = 0; i < data->mclk_latency_table.count - 1; i++) { + if (data->mclk_latency_table.entries[i].latency <= latency) { + if (dpm_table->dpm_levels[i].value >= (hwmgr->display_config->min_mem_set_clock / 100)) { + dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[i].value; + break; + } + } + } + } + + if (hwmgr->display_config->nb_pstate_switch_disable) + dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + + /* vclk */ + dpm_table = &(data->dpm_table.vclk_table); + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[0].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[0].value; + dpm_table->dpm_state.hard_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + + if (PP_CAP(PHM_PlatformCaps_UMDPState)) { + if (VEGA12_UMD_PSTATE_UVDCLK_LEVEL < dpm_table->count) { + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[VEGA12_UMD_PSTATE_UVDCLK_LEVEL].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[VEGA12_UMD_PSTATE_UVDCLK_LEVEL].value; + } + + if (hwmgr->dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK) { + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + } + } + + /* dclk */ + dpm_table = &(data->dpm_table.dclk_table); + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[0].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[0].value; + dpm_table->dpm_state.hard_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + + if (PP_CAP(PHM_PlatformCaps_UMDPState)) { + if (VEGA12_UMD_PSTATE_UVDCLK_LEVEL < dpm_table->count) { + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[VEGA12_UMD_PSTATE_UVDCLK_LEVEL].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[VEGA12_UMD_PSTATE_UVDCLK_LEVEL].value; + } + + if (hwmgr->dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK) { + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + } + } + + /* socclk */ + dpm_table = &(data->dpm_table.soc_table); + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[0].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[0].value; + dpm_table->dpm_state.hard_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + + if (PP_CAP(PHM_PlatformCaps_UMDPState)) { + if (VEGA12_UMD_PSTATE_SOCCLK_LEVEL < dpm_table->count) { + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[VEGA12_UMD_PSTATE_SOCCLK_LEVEL].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[VEGA12_UMD_PSTATE_SOCCLK_LEVEL].value; + } + + if (hwmgr->dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK) { + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + } + } + + /* eclk */ + dpm_table = &(data->dpm_table.eclk_table); + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[0].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[0].value; + dpm_table->dpm_state.hard_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + + if (PP_CAP(PHM_PlatformCaps_UMDPState)) { + if (VEGA12_UMD_PSTATE_VCEMCLK_LEVEL < dpm_table->count) { + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[VEGA12_UMD_PSTATE_VCEMCLK_LEVEL].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[VEGA12_UMD_PSTATE_VCEMCLK_LEVEL].value; + } + + if (hwmgr->dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK) { + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + } + } + + return 0; +} + +static int vega12_set_uclk_to_highest_dpm_level(struct pp_hwmgr *hwmgr, + struct vega12_single_dpm_table *dpm_table) +{ + struct vega12_hwmgr *data = (struct vega12_hwmgr *)(hwmgr->backend); + int ret = 0; + + if (data->smu_features[GNLD_DPM_UCLK].enabled) { + PP_ASSERT_WITH_CODE(dpm_table->count > 0, + "[SetUclkToHightestDpmLevel] Dpm table has no entry!", + return -EINVAL); + PP_ASSERT_WITH_CODE(dpm_table->count <= NUM_UCLK_DPM_LEVELS, + "[SetUclkToHightestDpmLevel] Dpm table has too many entries!", + return -EINVAL); + + dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value; + PP_ASSERT_WITH_CODE(!(ret = smum_send_msg_to_smc_with_parameter(hwmgr, + PPSMC_MSG_SetHardMinByFreq, + (PPCLK_UCLK << 16 ) | dpm_table->dpm_state.hard_min_level)), + "[SetUclkToHightestDpmLevel] Set hard min uclk failed!", + return ret); + } + + return ret; +} + +static int vega12_pre_display_configuration_changed_task(struct pp_hwmgr *hwmgr) +{ + struct vega12_hwmgr *data = (struct vega12_hwmgr *)(hwmgr->backend); + int ret = 0; + + smum_send_msg_to_smc_with_parameter(hwmgr, + PPSMC_MSG_NumOfDisplays, 0); + + ret = vega12_set_uclk_to_highest_dpm_level(hwmgr, + &data->dpm_table.mem_table); + + return ret; +} + static int vega12_display_configuration_changed_task(struct pp_hwmgr *hwmgr) { struct vega12_hwmgr *data = (struct vega12_hwmgr *)(hwmgr->backend); @@ -1915,6 +2139,9 @@ static void vega12_power_gate_vce(struct pp_hwmgr *hwmgr, bool bgate) { struct vega12_hwmgr *data = (struct vega12_hwmgr *)(hwmgr->backend); + if (data->vce_power_gated == bgate) + return; + data->vce_power_gated = bgate; vega12_enable_disable_vce_dpm(hwmgr, !bgate); } @@ -1923,6 +2150,9 @@ static void vega12_power_gate_uvd(struct pp_hwmgr *hwmgr, bool bgate) { struct vega12_hwmgr *data = (struct vega12_hwmgr *)(hwmgr->backend); + if (data->uvd_power_gated == bgate) + return; + data->uvd_power_gated = bgate; vega12_enable_disable_uvd_dpm(hwmgr, !bgate); } @@ -2090,6 +2320,38 @@ static int vega12_get_thermal_temperature_range(struct pp_hwmgr *hwmgr, return 0; } +static int vega12_enable_gfx_off(struct pp_hwmgr *hwmgr) +{ + struct vega12_hwmgr *data = + (struct vega12_hwmgr *)(hwmgr->backend); + int ret = 0; + + if (data->gfxoff_controlled_by_driver) + ret = smum_send_msg_to_smc(hwmgr, PPSMC_MSG_AllowGfxOff); + + return ret; +} + +static int vega12_disable_gfx_off(struct pp_hwmgr *hwmgr) +{ + struct vega12_hwmgr *data = + (struct vega12_hwmgr *)(hwmgr->backend); + int ret = 0; + + if (data->gfxoff_controlled_by_driver) + ret = smum_send_msg_to_smc(hwmgr, PPSMC_MSG_DisallowGfxOff); + + return ret; +} + +static int vega12_gfx_off_control(struct pp_hwmgr *hwmgr, bool enable) +{ + if (enable) + return vega12_enable_gfx_off(hwmgr); + else + return vega12_disable_gfx_off(hwmgr); +} + static const struct pp_hwmgr_func vega12_hwmgr_funcs = { .backend_init = vega12_hwmgr_backend_init, .backend_fini = vega12_hwmgr_backend_fini, @@ -2117,6 +2379,10 @@ static const struct pp_hwmgr_func vega12_hwmgr_funcs = { .display_clock_voltage_request = vega12_display_clock_voltage_request, .force_clock_level = vega12_force_clock_level, .print_clock_levels = vega12_print_clock_levels, + .apply_clocks_adjust_rules = + vega12_apply_clocks_adjust_rules, + .pre_display_config_changed = + vega12_pre_display_configuration_changed_task, .display_config_changed = vega12_display_configuration_changed_task, .powergate_uvd = vega12_power_gate_uvd, .powergate_vce = vega12_power_gate_vce, @@ -2135,6 +2401,7 @@ static const struct pp_hwmgr_func vega12_hwmgr_funcs = { .get_thermal_temperature_range = vega12_get_thermal_temperature_range, .register_irq_handlers = smu9_register_irq_handlers, .start_thermal_controller = vega12_start_thermal_controller, + .powergate_gfx = vega12_gfx_off_control, }; int vega12_hwmgr_init(struct pp_hwmgr *hwmgr) diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.h b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.h index 49b38df8c7f2..b3e424d28994 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.h +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.h @@ -304,6 +304,12 @@ struct vega12_odn_fan_table { bool force_fan_pwm; }; +struct vega12_clock_range { + uint32_t ACMax; + uint32_t ACMin; + uint32_t DCMax; +}; + struct vega12_hwmgr { struct vega12_dpm_table dpm_table; struct vega12_dpm_table golden_dpm_table; @@ -385,6 +391,11 @@ struct vega12_hwmgr { uint32_t smu_version; struct smu_features smu_features[GNLD_FEATURES_MAX]; struct vega12_smc_state_table smc_state_table; + + struct vega12_clock_range clk_range[PPCLK_COUNT]; + + /* ---- Gfxoff ---- */ + bool gfxoff_controlled_by_driver; }; #define VEGA12_DPM2_NEAR_TDP_DEC 10 @@ -435,6 +446,8 @@ struct vega12_hwmgr { #define VEGA12_UMD_PSTATE_GFXCLK_LEVEL 0x3 #define VEGA12_UMD_PSTATE_SOCCLK_LEVEL 0x3 #define VEGA12_UMD_PSTATE_MCLK_LEVEL 0x2 +#define VEGA12_UMD_PSTATE_UVDCLK_LEVEL 0x3 +#define VEGA12_UMD_PSTATE_VCEMCLK_LEVEL 0x3 int vega12_enable_disable_vce_dpm(struct pp_hwmgr *hwmgr, bool enable); diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_processpptables.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_processpptables.c index 29914700ee82..cb3a5b1737c8 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_processpptables.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_processpptables.c @@ -224,11 +224,7 @@ static int append_vbios_pptable(struct pp_hwmgr *hwmgr, PPTable_t *ppsmc_pptable ppsmc_pptable->AcgGfxclkSpreadPercent = smc_dpm_table.acggfxclkspreadpercent; ppsmc_pptable->AcgGfxclkSpreadFreq = smc_dpm_table.acggfxclkspreadfreq; - /* 0xFFFF will disable the ACG feature */ - if (!(hwmgr->feature_mask & PP_ACG_MASK)) { - ppsmc_pptable->AcgThresholdFreqHigh = 0xFFFF; - ppsmc_pptable->AcgThresholdFreqLow = 0xFFFF; - } + ppsmc_pptable->Vr2_I2C_address = smc_dpm_table.Vr2_I2C_address; ppsmc_pptable->Vr2_I2C_address = smc_dpm_table.Vr2_I2C_address; diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_thermal.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_thermal.c index cfd9e6ccb790..904eb2c9155b 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_thermal.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_thermal.c @@ -34,11 +34,9 @@ static int vega12_get_current_rpm(struct pp_hwmgr *hwmgr, uint32_t *current_rpm) PP_ASSERT_WITH_CODE(!smum_send_msg_to_smc(hwmgr, PPSMC_MSG_GetCurrentRpm), "Attempt to get current RPM from SMC Failed!", - return -1); - PP_ASSERT_WITH_CODE(!vega12_read_arg_from_smc(hwmgr, - current_rpm), - "Attempt to read current RPM from SMC Failed!", - return -1); + return -EINVAL); + *current_rpm = smum_get_argument(hwmgr); + return 0; } diff --git a/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h b/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h index a202247c9894..429c9c4322da 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h +++ b/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h @@ -455,7 +455,7 @@ extern int phm_get_clock_by_type_with_voltage(struct pp_hwmgr *hwmgr, enum amd_pp_clock_type type, struct pp_clock_levels_with_voltage *clocks); extern int phm_set_watermarks_for_clocks_ranges(struct pp_hwmgr *hwmgr, - struct pp_wm_sets_with_clock_ranges_soc15 *wm_with_clock_ranges); + void *clock_ranges); extern int phm_display_clock_voltage_request(struct pp_hwmgr *hwmgr, struct pp_display_clock_request *clock); diff --git a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h b/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h index b99fb8ac822c..d3d96260f440 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h +++ b/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h @@ -26,7 +26,6 @@ #include <linux/seq_file.h> #include "amd_powerplay.h" #include "hardwaremanager.h" -#include "pp_power_source.h" #include "hwmgr_ppt.h" #include "ppatomctrl.h" #include "hwmgr_ppt.h" @@ -195,7 +194,7 @@ struct pp_smumgr_func { int (*request_smu_load_fw)(struct pp_hwmgr *hwmgr); int (*request_smu_load_specific_fw)(struct pp_hwmgr *hwmgr, uint32_t firmware); - int (*get_argument)(struct pp_hwmgr *hwmgr); + uint32_t (*get_argument)(struct pp_hwmgr *hwmgr); int (*send_msg_to_smc)(struct pp_hwmgr *hwmgr, uint16_t msg); int (*send_msg_to_smc_with_parameter)(struct pp_hwmgr *hwmgr, uint16_t msg, uint32_t parameter); @@ -294,8 +293,7 @@ struct pp_hwmgr_func { int (*get_clock_by_type_with_voltage)(struct pp_hwmgr *hwmgr, enum amd_pp_clock_type type, struct pp_clock_levels_with_voltage *clocks); - int (*set_watermarks_for_clocks_ranges)(struct pp_hwmgr *hwmgr, - struct pp_wm_sets_with_clock_ranges_soc15 *wm_with_clock_ranges); + int (*set_watermarks_for_clocks_ranges)(struct pp_hwmgr *hwmgr, void *clock_ranges); int (*display_clock_voltage_request)(struct pp_hwmgr *hwmgr, struct pp_display_clock_request *clock); int (*get_max_high_clocks)(struct pp_hwmgr *hwmgr, struct amd_pp_simple_clock_info *clocks); @@ -303,7 +301,7 @@ struct pp_hwmgr_func { int (*power_off_asic)(struct pp_hwmgr *hwmgr); int (*force_clock_level)(struct pp_hwmgr *hwmgr, enum pp_clock_type type, uint32_t mask); int (*print_clock_levels)(struct pp_hwmgr *hwmgr, enum pp_clock_type type, char *buf); - int (*enable_per_cu_power_gating)(struct pp_hwmgr *hwmgr, bool enable); + int (*powergate_gfx)(struct pp_hwmgr *hwmgr, bool enable); int (*get_sclk_od)(struct pp_hwmgr *hwmgr); int (*set_sclk_od)(struct pp_hwmgr *hwmgr, uint32_t value); int (*get_mclk_od)(struct pp_hwmgr *hwmgr); @@ -328,7 +326,7 @@ struct pp_hwmgr_func { enum PP_OD_DPM_TABLE_COMMAND type, long *input, uint32_t size); int (*set_power_limit)(struct pp_hwmgr *hwmgr, uint32_t n); - int (*set_mmhub_powergating_by_smu)(struct pp_hwmgr *hwmgr); + int (*powergate_mmhub)(struct pp_hwmgr *hwmgr); int (*smus_notify_pwe)(struct pp_hwmgr *hwmgr); }; @@ -741,7 +739,6 @@ struct pp_hwmgr { const struct pp_table_func *pptable_func; struct pp_power_state *ps; - enum pp_power_source power_source; uint32_t num_ps; struct pp_thermal_controller_info thermal_controller; bool fan_ctrl_is_in_default_mode; diff --git a/drivers/gpu/drm/amd/powerplay/inc/smumgr.h b/drivers/gpu/drm/amd/powerplay/inc/smumgr.h index 6c22ed9249bf..82550a8a3a3f 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/smumgr.h +++ b/drivers/gpu/drm/amd/powerplay/inc/smumgr.h @@ -29,7 +29,6 @@ enum SMU_TABLE { SMU_UVD_TABLE = 0, SMU_VCE_TABLE, - SMU_SAMU_TABLE, SMU_BIF_TABLE, }; @@ -47,7 +46,6 @@ enum SMU_MEMBER { UcodeLoadStatus, UvdBootLevel, VceBootLevel, - SamuBootLevel, LowSclkInterruptThreshold, DRAM_LOG_ADDR_H, DRAM_LOG_ADDR_L, @@ -82,7 +80,7 @@ enum SMU10_TABLE_ID { SMU10_CLOCKTABLE, }; -extern int smum_get_argument(struct pp_hwmgr *hwmgr); +extern uint32_t smum_get_argument(struct pp_hwmgr *hwmgr); extern int smum_download_powerplay_table(struct pp_hwmgr *hwmgr, void **table); diff --git a/drivers/gpu/drm/amd/powerplay/inc/vega12/smu9_driver_if.h b/drivers/gpu/drm/amd/powerplay/inc/vega12/smu9_driver_if.h index b08526fd1619..b6ffd08784e7 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/vega12/smu9_driver_if.h +++ b/drivers/gpu/drm/amd/powerplay/inc/vega12/smu9_driver_if.h @@ -412,10 +412,10 @@ typedef struct { QuadraticInt_t ReservedEquation2; QuadraticInt_t ReservedEquation3; - uint16_t MinVoltageUlvGfx; - uint16_t MinVoltageUlvSoc; + uint16_t MinVoltageUlvGfx; + uint16_t MinVoltageUlvSoc; - uint32_t Reserved[14]; + uint32_t Reserved[14]; @@ -483,9 +483,9 @@ typedef struct { uint8_t padding8_4; - uint8_t PllGfxclkSpreadEnabled; - uint8_t PllGfxclkSpreadPercent; - uint16_t PllGfxclkSpreadFreq; + uint8_t PllGfxclkSpreadEnabled; + uint8_t PllGfxclkSpreadPercent; + uint16_t PllGfxclkSpreadFreq; uint8_t UclkSpreadEnabled; uint8_t UclkSpreadPercent; @@ -495,9 +495,9 @@ typedef struct { uint8_t SocclkSpreadPercent; uint16_t SocclkSpreadFreq; - uint8_t AcgGfxclkSpreadEnabled; - uint8_t AcgGfxclkSpreadPercent; - uint16_t AcgGfxclkSpreadFreq; + uint8_t AcgGfxclkSpreadEnabled; + uint8_t AcgGfxclkSpreadPercent; + uint16_t AcgGfxclkSpreadFreq; uint8_t Vr2_I2C_address; uint8_t padding_vr2[3]; diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/Makefile b/drivers/gpu/drm/amd/powerplay/smumgr/Makefile index 0a200406a1ec..8d557accaef2 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/Makefile +++ b/drivers/gpu/drm/amd/powerplay/smumgr/Makefile @@ -26,7 +26,7 @@ SMU_MGR = smumgr.o smu8_smumgr.o tonga_smumgr.o fiji_smumgr.o \ polaris10_smumgr.o iceland_smumgr.o \ smu7_smumgr.o vega10_smumgr.o smu10_smumgr.o ci_smumgr.o \ - vega12_smumgr.o vegam_smumgr.o + vega12_smumgr.o vegam_smumgr.o smu9_smumgr.o AMD_PP_SMUMGR = $(addprefix $(AMD_PP_PATH)/smumgr/,$(SMU_MGR)) diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/ci_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/ci_smumgr.c index 2d4ec8ac3a08..fbe3ef4ee45c 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/ci_smumgr.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/ci_smumgr.c @@ -1614,37 +1614,6 @@ static int ci_populate_smc_acp_level(struct pp_hwmgr *hwmgr, return result; } -static int ci_populate_smc_samu_level(struct pp_hwmgr *hwmgr, - SMU7_Discrete_DpmTable *table) -{ - int result = -EINVAL; - uint8_t count; - struct pp_atomctrl_clock_dividers_vi dividers; - struct phm_samu_clock_voltage_dependency_table *samu_table = - hwmgr->dyn_state.samu_clock_voltage_dependency_table; - - table->SamuBootLevel = 0; - table->SamuLevelCount = (uint8_t)(samu_table->count); - - for (count = 0; count < table->SamuLevelCount; count++) { - table->SamuLevel[count].Frequency = samu_table->entries[count].samclk; - table->SamuLevel[count].MinVoltage = samu_table->entries[count].v * VOLTAGE_SCALE; - table->SamuLevel[count].MinPhases = 1; - - /* retrieve divider value for VBIOS */ - result = atomctrl_get_dfs_pll_dividers_vi(hwmgr, - table->SamuLevel[count].Frequency, ÷rs); - PP_ASSERT_WITH_CODE((0 == result), - "can not find divide id for samu clock", return result); - - table->SamuLevel[count].Divider = (uint8_t)dividers.pll_post_divider; - - CONVERT_FROM_HOST_TO_SMC_UL(table->SamuLevel[count].Frequency); - CONVERT_FROM_HOST_TO_SMC_US(table->SamuLevel[count].MinVoltage); - } - return result; -} - static int ci_populate_memory_timing_parameters( struct pp_hwmgr *hwmgr, uint32_t engine_clock, @@ -2026,10 +1995,6 @@ static int ci_init_smc_table(struct pp_hwmgr *hwmgr) PP_ASSERT_WITH_CODE(0 == result, "Failed to initialize ACP Level!", return result); - result = ci_populate_smc_samu_level(hwmgr, table); - PP_ASSERT_WITH_CODE(0 == result, - "Failed to initialize SAMU Level!", return result); - /* Since only the initial state is completely set up at this point (the other states are just copies of the boot state) we only */ /* need to populate the ARB settings for the initial state. */ result = ci_program_memory_timing_parameters(hwmgr); @@ -2881,6 +2846,89 @@ static int ci_update_dpm_settings(struct pp_hwmgr *hwmgr, return 0; } +static int ci_update_uvd_smc_table(struct pp_hwmgr *hwmgr) +{ + struct amdgpu_device *adev = hwmgr->adev; + struct smu7_hwmgr *data = hwmgr->backend; + struct ci_smumgr *smu_data = hwmgr->smu_backend; + struct phm_uvd_clock_voltage_dependency_table *uvd_table = + hwmgr->dyn_state.uvd_clock_voltage_dependency_table; + uint32_t profile_mode_mask = AMD_DPM_FORCED_LEVEL_PROFILE_STANDARD | + AMD_DPM_FORCED_LEVEL_PROFILE_MIN_SCLK | + AMD_DPM_FORCED_LEVEL_PROFILE_MIN_MCLK | + AMD_DPM_FORCED_LEVEL_PROFILE_PEAK; + uint32_t max_vddc = adev->pm.ac_power ? hwmgr->dyn_state.max_clock_voltage_on_ac.vddc : + hwmgr->dyn_state.max_clock_voltage_on_dc.vddc; + int32_t i; + + if (PP_CAP(PHM_PlatformCaps_UVDDPM) || uvd_table->count <= 0) + smu_data->smc_state_table.UvdBootLevel = 0; + else + smu_data->smc_state_table.UvdBootLevel = uvd_table->count - 1; + + PHM_WRITE_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, DPM_TABLE_475, + UvdBootLevel, smu_data->smc_state_table.UvdBootLevel); + + data->dpm_level_enable_mask.uvd_dpm_enable_mask = 0; + + for (i = uvd_table->count - 1; i >= 0; i--) { + if (uvd_table->entries[i].v <= max_vddc) + data->dpm_level_enable_mask.uvd_dpm_enable_mask |= 1 << i; + if (hwmgr->dpm_level & profile_mode_mask || !PP_CAP(PHM_PlatformCaps_UVDDPM)) + break; + } + ci_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_UVDDPM_SetEnabledMask, + data->dpm_level_enable_mask.uvd_dpm_enable_mask); + + return 0; +} + +static int ci_update_vce_smc_table(struct pp_hwmgr *hwmgr) +{ + struct amdgpu_device *adev = hwmgr->adev; + struct smu7_hwmgr *data = hwmgr->backend; + struct phm_vce_clock_voltage_dependency_table *vce_table = + hwmgr->dyn_state.vce_clock_voltage_dependency_table; + uint32_t profile_mode_mask = AMD_DPM_FORCED_LEVEL_PROFILE_STANDARD | + AMD_DPM_FORCED_LEVEL_PROFILE_MIN_SCLK | + AMD_DPM_FORCED_LEVEL_PROFILE_MIN_MCLK | + AMD_DPM_FORCED_LEVEL_PROFILE_PEAK; + uint32_t max_vddc = adev->pm.ac_power ? hwmgr->dyn_state.max_clock_voltage_on_ac.vddc : + hwmgr->dyn_state.max_clock_voltage_on_dc.vddc; + int32_t i; + + PHM_WRITE_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, DPM_TABLE_475, + VceBootLevel, 0); /* temp hard code to level 0, vce can set min evclk*/ + + data->dpm_level_enable_mask.vce_dpm_enable_mask = 0; + + for (i = vce_table->count - 1; i >= 0; i--) { + if (vce_table->entries[i].v <= max_vddc) + data->dpm_level_enable_mask.vce_dpm_enable_mask |= 1 << i; + if (hwmgr->dpm_level & profile_mode_mask || !PP_CAP(PHM_PlatformCaps_VCEDPM)) + break; + } + ci_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_VCEDPM_SetEnabledMask, + data->dpm_level_enable_mask.vce_dpm_enable_mask); + + return 0; +} + +static int ci_update_smc_table(struct pp_hwmgr *hwmgr, uint32_t type) +{ + switch (type) { + case SMU_UVD_TABLE: + ci_update_uvd_smc_table(hwmgr); + break; + case SMU_VCE_TABLE: + ci_update_vce_smc_table(hwmgr); + break; + default: + break; + } + return 0; +} + const struct pp_smumgr_func ci_smu_funcs = { .smu_init = ci_smu_init, .smu_fini = ci_smu_fini, @@ -2903,4 +2951,5 @@ const struct pp_smumgr_func ci_smu_funcs = { .initialize_mc_reg_table = ci_initialize_mc_reg_table, .is_dpm_running = ci_is_dpm_running, .update_dpm_settings = ci_update_dpm_settings, + .update_smc_table = ci_update_smc_table, }; diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c index 53df9405f43a..18048f8e2f13 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c @@ -1503,44 +1503,6 @@ static int fiji_populate_smc_acp_level(struct pp_hwmgr *hwmgr, return result; } -static int fiji_populate_smc_samu_level(struct pp_hwmgr *hwmgr, - SMU73_Discrete_DpmTable *table) -{ - int result = -EINVAL; - uint8_t count; - struct pp_atomctrl_clock_dividers_vi dividers; - struct phm_ppt_v1_information *table_info = - (struct phm_ppt_v1_information *)(hwmgr->pptable); - struct phm_ppt_v1_mm_clock_voltage_dependency_table *mm_table = - table_info->mm_dep_table; - - table->SamuBootLevel = 0; - table->SamuLevelCount = (uint8_t)(mm_table->count); - - for (count = 0; count < table->SamuLevelCount; count++) { - /* not sure whether we need evclk or not */ - table->SamuLevel[count].MinVoltage = 0; - table->SamuLevel[count].Frequency = mm_table->entries[count].samclock; - table->SamuLevel[count].MinVoltage |= (mm_table->entries[count].vddc * - VOLTAGE_SCALE) << VDDC_SHIFT; - table->SamuLevel[count].MinVoltage |= ((mm_table->entries[count].vddc - - VDDC_VDDCI_DELTA) * VOLTAGE_SCALE) << VDDCI_SHIFT; - table->SamuLevel[count].MinVoltage |= 1 << PHASES_SHIFT; - - /* retrieve divider value for VBIOS */ - result = atomctrl_get_dfs_pll_dividers_vi(hwmgr, - table->SamuLevel[count].Frequency, ÷rs); - PP_ASSERT_WITH_CODE((0 == result), - "can not find divide id for samu clock", return result); - - table->SamuLevel[count].Divider = (uint8_t)dividers.pll_post_divider; - - CONVERT_FROM_HOST_TO_SMC_UL(table->SamuLevel[count].Frequency); - CONVERT_FROM_HOST_TO_SMC_UL(table->SamuLevel[count].MinVoltage); - } - return result; -} - static int fiji_populate_memory_timing_parameters(struct pp_hwmgr *hwmgr, int32_t eng_clock, int32_t mem_clock, struct SMU73_Discrete_MCArbDramTimingTableEntry *arb_regs) @@ -2028,10 +1990,6 @@ static int fiji_init_smc_table(struct pp_hwmgr *hwmgr) PP_ASSERT_WITH_CODE(0 == result, "Failed to initialize ACP Level!", return result); - result = fiji_populate_smc_samu_level(hwmgr, table); - PP_ASSERT_WITH_CODE(0 == result, - "Failed to initialize SAMU Level!", return result); - /* Since only the initial state is completely set up at this point * (the other states are just copies of the boot state) we only * need to populate the ARB settings for the initial state. @@ -2378,8 +2336,6 @@ static uint32_t fiji_get_offsetof(uint32_t type, uint32_t member) return offsetof(SMU73_Discrete_DpmTable, UvdBootLevel); case VceBootLevel: return offsetof(SMU73_Discrete_DpmTable, VceBootLevel); - case SamuBootLevel: - return offsetof(SMU73_Discrete_DpmTable, SamuBootLevel); case LowSclkInterruptThreshold: return offsetof(SMU73_Discrete_DpmTable, LowSclkInterruptThreshold); } @@ -2478,33 +2434,6 @@ static int fiji_update_vce_smc_table(struct pp_hwmgr *hwmgr) return 0; } -static int fiji_update_samu_smc_table(struct pp_hwmgr *hwmgr) -{ - struct fiji_smumgr *smu_data = (struct fiji_smumgr *)(hwmgr->smu_backend); - uint32_t mm_boot_level_offset, mm_boot_level_value; - - - smu_data->smc_state_table.SamuBootLevel = 0; - mm_boot_level_offset = smu_data->smu7_data.dpm_table_start + - offsetof(SMU73_Discrete_DpmTable, SamuBootLevel); - - mm_boot_level_offset /= 4; - mm_boot_level_offset *= 4; - mm_boot_level_value = cgs_read_ind_register(hwmgr->device, - CGS_IND_REG__SMC, mm_boot_level_offset); - mm_boot_level_value &= 0xFFFFFF00; - mm_boot_level_value |= smu_data->smc_state_table.SamuBootLevel << 0; - cgs_write_ind_register(hwmgr->device, - CGS_IND_REG__SMC, mm_boot_level_offset, mm_boot_level_value); - - if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, - PHM_PlatformCaps_StablePState)) - smum_send_msg_to_smc_with_parameter(hwmgr, - PPSMC_MSG_SAMUDPM_SetEnabledMask, - (uint32_t)(1 << smu_data->smc_state_table.SamuBootLevel)); - return 0; -} - static int fiji_update_smc_table(struct pp_hwmgr *hwmgr, uint32_t type) { switch (type) { @@ -2514,9 +2443,6 @@ static int fiji_update_smc_table(struct pp_hwmgr *hwmgr, uint32_t type) case SMU_VCE_TABLE: fiji_update_vce_smc_table(hwmgr); break; - case SMU_SAMU_TABLE: - fiji_update_samu_smc_table(hwmgr); - break; default: break; } diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/iceland_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/iceland_smumgr.c index 415f691c3fa9..9299b93aa09a 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/iceland_smumgr.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/iceland_smumgr.c @@ -1578,12 +1578,6 @@ static int iceland_populate_smc_acp_level(struct pp_hwmgr *hwmgr, return 0; } -static int iceland_populate_smc_samu_level(struct pp_hwmgr *hwmgr, - SMU71_Discrete_DpmTable *table) -{ - return 0; -} - static int iceland_populate_memory_timing_parameters( struct pp_hwmgr *hwmgr, uint32_t engine_clock, @@ -1992,10 +1986,6 @@ static int iceland_init_smc_table(struct pp_hwmgr *hwmgr) PP_ASSERT_WITH_CODE(0 == result, "Failed to initialize ACP Level!", return result;); - result = iceland_populate_smc_samu_level(hwmgr, table); - PP_ASSERT_WITH_CODE(0 == result, - "Failed to initialize SAMU Level!", return result;); - /* Since only the initial state is completely set up at this point (the other states are just copies of the boot state) we only */ /* need to populate the ARB settings for the initial state. */ result = iceland_program_memory_timing_parameters(hwmgr); diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c index a8c6524f07e4..1276f168ff68 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c @@ -1204,7 +1204,6 @@ static int polaris10_populate_smc_acpi_level(struct pp_hwmgr *hwmgr, (struct phm_ppt_v1_information *)(hwmgr->pptable); SMIO_Pattern vol_level; uint32_t mvdd; - uint16_t us_mvdd; table->ACPILevel.Flags &= ~PPSMC_SWSTATE_FLAG_DC; @@ -1255,16 +1254,11 @@ static int polaris10_populate_smc_acpi_level(struct pp_hwmgr *hwmgr, "in Clock Dependency Table", ); - us_mvdd = 0; - if ((SMU7_VOLTAGE_CONTROL_NONE == data->mvdd_control) || - (data->mclk_dpm_key_disabled)) - us_mvdd = data->vbios_boot_state.mvdd_bootup_value; - else { - if (!polaris10_populate_mvdd_value(hwmgr, + if (!((SMU7_VOLTAGE_CONTROL_NONE == data->mvdd_control) || + (data->mclk_dpm_key_disabled))) + polaris10_populate_mvdd_value(hwmgr, data->dpm_table.mclk_table.dpm_levels[0].value, - &vol_level)) - us_mvdd = vol_level.Voltage; - } + &vol_level); if (0 == polaris10_populate_mvdd_value(hwmgr, 0, &vol_level)) table->MemoryACPILevel.MinMvdd = PP_HOST_TO_SMC_UL(vol_level.Voltage); @@ -1337,55 +1331,6 @@ static int polaris10_populate_smc_vce_level(struct pp_hwmgr *hwmgr, return result; } - -static int polaris10_populate_smc_samu_level(struct pp_hwmgr *hwmgr, - SMU74_Discrete_DpmTable *table) -{ - int result = -EINVAL; - uint8_t count; - struct pp_atomctrl_clock_dividers_vi dividers; - struct phm_ppt_v1_information *table_info = - (struct phm_ppt_v1_information *)(hwmgr->pptable); - struct phm_ppt_v1_mm_clock_voltage_dependency_table *mm_table = - table_info->mm_dep_table; - struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend); - uint32_t vddci; - - table->SamuBootLevel = 0; - table->SamuLevelCount = (uint8_t)(mm_table->count); - - for (count = 0; count < table->SamuLevelCount; count++) { - /* not sure whether we need evclk or not */ - table->SamuLevel[count].MinVoltage = 0; - table->SamuLevel[count].Frequency = mm_table->entries[count].samclock; - table->SamuLevel[count].MinVoltage |= (mm_table->entries[count].vddc * - VOLTAGE_SCALE) << VDDC_SHIFT; - - if (SMU7_VOLTAGE_CONTROL_BY_GPIO == data->vddci_control) - vddci = (uint32_t)phm_find_closest_vddci(&(data->vddci_voltage_table), - mm_table->entries[count].vddc - VDDC_VDDCI_DELTA); - else if (SMU7_VOLTAGE_CONTROL_BY_SVID2 == data->vddci_control) - vddci = mm_table->entries[count].vddc - VDDC_VDDCI_DELTA; - else - vddci = (data->vbios_boot_state.vddci_bootup_value * VOLTAGE_SCALE) << VDDCI_SHIFT; - - table->SamuLevel[count].MinVoltage |= (vddci * VOLTAGE_SCALE) << VDDCI_SHIFT; - table->SamuLevel[count].MinVoltage |= 1 << PHASES_SHIFT; - - /* retrieve divider value for VBIOS */ - result = atomctrl_get_dfs_pll_dividers_vi(hwmgr, - table->SamuLevel[count].Frequency, ÷rs); - PP_ASSERT_WITH_CODE((0 == result), - "can not find divide id for samu clock", return result); - - table->SamuLevel[count].Divider = (uint8_t)dividers.pll_post_divider; - - CONVERT_FROM_HOST_TO_SMC_UL(table->SamuLevel[count].Frequency); - CONVERT_FROM_HOST_TO_SMC_UL(table->SamuLevel[count].MinVoltage); - } - return result; -} - static int polaris10_populate_memory_timing_parameters(struct pp_hwmgr *hwmgr, int32_t eng_clock, int32_t mem_clock, SMU74_Discrete_MCArbDramTimingTableEntry *arb_regs) @@ -1566,7 +1511,7 @@ static int polaris10_populate_clock_stretcher_data_table(struct pp_hwmgr *hwmgr) uint32_t ro, efuse, volt_without_cks, volt_with_cks, value, max, min; struct polaris10_smumgr *smu_data = (struct polaris10_smumgr *)(hwmgr->smu_backend); - uint8_t i, stretch_amount, stretch_amount2, volt_offset = 0; + uint8_t i, stretch_amount, volt_offset = 0; struct phm_ppt_v1_information *table_info = (struct phm_ppt_v1_information *)(hwmgr->pptable); struct phm_ppt_v1_clock_voltage_dependency_table *sclk_table = @@ -1617,11 +1562,7 @@ static int polaris10_populate_clock_stretcher_data_table(struct pp_hwmgr *hwmgr) smu_data->smc_state_table.LdoRefSel = (table_info->cac_dtp_table->ucCKS_LDO_REFSEL != 0) ? table_info->cac_dtp_table->ucCKS_LDO_REFSEL : 6; /* Populate CKS Lookup Table */ - if (stretch_amount == 1 || stretch_amount == 2 || stretch_amount == 5) - stretch_amount2 = 0; - else if (stretch_amount == 3 || stretch_amount == 4) - stretch_amount2 = 1; - else { + if (stretch_amount == 0 || stretch_amount > 5) { phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_ClockStretcher); PP_ASSERT_WITH_CODE(false, @@ -1865,10 +1806,6 @@ static int polaris10_init_smc_table(struct pp_hwmgr *hwmgr) PP_ASSERT_WITH_CODE(0 == result, "Failed to initialize VCE Level!", return result); - result = polaris10_populate_smc_samu_level(hwmgr, table); - PP_ASSERT_WITH_CODE(0 == result, - "Failed to initialize SAMU Level!", return result); - /* Since only the initial state is completely set up at this point * (the other states are just copies of the boot state) we only * need to populate the ARB settings for the initial state. @@ -2222,34 +2159,6 @@ static int polaris10_update_vce_smc_table(struct pp_hwmgr *hwmgr) return 0; } -static int polaris10_update_samu_smc_table(struct pp_hwmgr *hwmgr) -{ - struct polaris10_smumgr *smu_data = (struct polaris10_smumgr *)(hwmgr->smu_backend); - uint32_t mm_boot_level_offset, mm_boot_level_value; - - - smu_data->smc_state_table.SamuBootLevel = 0; - mm_boot_level_offset = smu_data->smu7_data.dpm_table_start + - offsetof(SMU74_Discrete_DpmTable, SamuBootLevel); - - mm_boot_level_offset /= 4; - mm_boot_level_offset *= 4; - mm_boot_level_value = cgs_read_ind_register(hwmgr->device, - CGS_IND_REG__SMC, mm_boot_level_offset); - mm_boot_level_value &= 0xFFFFFF00; - mm_boot_level_value |= smu_data->smc_state_table.SamuBootLevel << 0; - cgs_write_ind_register(hwmgr->device, - CGS_IND_REG__SMC, mm_boot_level_offset, mm_boot_level_value); - - if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, - PHM_PlatformCaps_StablePState)) - smum_send_msg_to_smc_with_parameter(hwmgr, - PPSMC_MSG_SAMUDPM_SetEnabledMask, - (uint32_t)(1 << smu_data->smc_state_table.SamuBootLevel)); - return 0; -} - - static int polaris10_update_bif_smc_table(struct pp_hwmgr *hwmgr) { struct polaris10_smumgr *smu_data = (struct polaris10_smumgr *)(hwmgr->smu_backend); @@ -2276,9 +2185,6 @@ static int polaris10_update_smc_table(struct pp_hwmgr *hwmgr, uint32_t type) case SMU_VCE_TABLE: polaris10_update_vce_smc_table(hwmgr); break; - case SMU_SAMU_TABLE: - polaris10_update_samu_smc_table(hwmgr); - break; case SMU_BIF_TABLE: polaris10_update_bif_smc_table(hwmgr); default: @@ -2357,8 +2263,6 @@ static uint32_t polaris10_get_offsetof(uint32_t type, uint32_t member) return offsetof(SMU74_Discrete_DpmTable, UvdBootLevel); case VceBootLevel: return offsetof(SMU74_Discrete_DpmTable, VceBootLevel); - case SamuBootLevel: - return offsetof(SMU74_Discrete_DpmTable, SamuBootLevel); case LowSclkInterruptThreshold: return offsetof(SMU74_Discrete_DpmTable, LowSclkInterruptThreshold); } diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smu10_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/smu10_smumgr.c index 0a563f6fe9ea..bb07d43f3874 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/smu10_smumgr.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/smu10_smumgr.c @@ -68,7 +68,7 @@ static int smu10_send_msg_to_smc_without_waiting(struct pp_hwmgr *hwmgr, return 0; } -static int smu10_read_arg_from_smc(struct pp_hwmgr *hwmgr) +static uint32_t smu10_read_arg_from_smc(struct pp_hwmgr *hwmgr) { struct amdgpu_device *adev = hwmgr->adev; diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c index 9f407c48d4f0..a029e47c2319 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c @@ -379,9 +379,7 @@ int smu7_request_smu_load_fw(struct pp_hwmgr *hwmgr) { struct smu7_smumgr *smu_data = (struct smu7_smumgr *)(hwmgr->smu_backend); uint32_t fw_to_load; - int result = 0; - struct SMU_DRAMData_TOC *toc; - uint32_t num_entries = 0; + int r = 0; if (!hwmgr->reload_fw) { pr_info("skip reloading...\n"); @@ -422,49 +420,62 @@ int smu7_request_smu_load_fw(struct pp_hwmgr *hwmgr) + UCODE_ID_CP_MEC_JT2_MASK; } - toc = (struct SMU_DRAMData_TOC *)smu_data->header; - toc->structure_version = 1; - - PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr, - UCODE_ID_RLC_G, &toc->entry[num_entries++]), - "Failed to Get Firmware Entry.", return -EINVAL); - PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr, - UCODE_ID_CP_CE, &toc->entry[num_entries++]), - "Failed to Get Firmware Entry.", return -EINVAL); - PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr, - UCODE_ID_CP_PFP, &toc->entry[num_entries++]), - "Failed to Get Firmware Entry.", return -EINVAL); - PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr, - UCODE_ID_CP_ME, &toc->entry[num_entries++]), - "Failed to Get Firmware Entry.", return -EINVAL); - PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr, - UCODE_ID_CP_MEC, &toc->entry[num_entries++]), - "Failed to Get Firmware Entry.", return -EINVAL); - PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr, - UCODE_ID_CP_MEC_JT1, &toc->entry[num_entries++]), - "Failed to Get Firmware Entry.", return -EINVAL); - PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr, - UCODE_ID_CP_MEC_JT2, &toc->entry[num_entries++]), - "Failed to Get Firmware Entry.", return -EINVAL); - PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr, - UCODE_ID_SDMA0, &toc->entry[num_entries++]), - "Failed to Get Firmware Entry.", return -EINVAL); - PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr, - UCODE_ID_SDMA1, &toc->entry[num_entries++]), - "Failed to Get Firmware Entry.", return -EINVAL); - if (!hwmgr->not_vf) - PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr, - UCODE_ID_MEC_STORAGE, &toc->entry[num_entries++]), - "Failed to Get Firmware Entry.", return -EINVAL); + if (!smu_data->toc) { + struct SMU_DRAMData_TOC *toc; - toc->num_entries = num_entries; + smu_data->toc = kzalloc(sizeof(struct SMU_DRAMData_TOC), GFP_KERNEL); + if (!smu_data->toc) + return -ENOMEM; + toc = smu_data->toc; + toc->num_entries = 0; + toc->structure_version = 1; + + PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr, + UCODE_ID_RLC_G, &toc->entry[toc->num_entries++]), + "Failed to Get Firmware Entry.", r = -EINVAL; goto failed); + PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr, + UCODE_ID_CP_CE, &toc->entry[toc->num_entries++]), + "Failed to Get Firmware Entry.", r = -EINVAL; goto failed); + PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr, + UCODE_ID_CP_PFP, &toc->entry[toc->num_entries++]), + "Failed to Get Firmware Entry.", r = -EINVAL; goto failed); + PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr, + UCODE_ID_CP_ME, &toc->entry[toc->num_entries++]), + "Failed to Get Firmware Entry.", r = -EINVAL; goto failed); + PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr, + UCODE_ID_CP_MEC, &toc->entry[toc->num_entries++]), + "Failed to Get Firmware Entry.", r = -EINVAL; goto failed); + PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr, + UCODE_ID_CP_MEC_JT1, &toc->entry[toc->num_entries++]), + "Failed to Get Firmware Entry.", r = -EINVAL; goto failed); + PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr, + UCODE_ID_CP_MEC_JT2, &toc->entry[toc->num_entries++]), + "Failed to Get Firmware Entry.", r = -EINVAL; goto failed); + PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr, + UCODE_ID_SDMA0, &toc->entry[toc->num_entries++]), + "Failed to Get Firmware Entry.", r = -EINVAL; goto failed); + PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr, + UCODE_ID_SDMA1, &toc->entry[toc->num_entries++]), + "Failed to Get Firmware Entry.", r = -EINVAL; goto failed); + if (!hwmgr->not_vf) + PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr, + UCODE_ID_MEC_STORAGE, &toc->entry[toc->num_entries++]), + "Failed to Get Firmware Entry.", r = -EINVAL; goto failed); + } + memcpy_toio(smu_data->header_buffer.kaddr, smu_data->toc, + sizeof(struct SMU_DRAMData_TOC)); smu7_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_DRV_DRAM_ADDR_HI, upper_32_bits(smu_data->header_buffer.mc_addr)); smu7_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_DRV_DRAM_ADDR_LO, lower_32_bits(smu_data->header_buffer.mc_addr)); if (smu7_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_LoadUcodes, fw_to_load)) pr_err("Fail to Request SMU Load uCode"); - return result; + return r; + +failed: + kfree(smu_data->toc); + smu_data->toc = NULL; + return r; } /* Check if the FW has been loaded, SMU will not return if loading has not finished. */ @@ -571,7 +582,6 @@ int smu7_setup_pwr_virus(struct pp_hwmgr *hwmgr) int smu7_init(struct pp_hwmgr *hwmgr) { struct smu7_smumgr *smu_data; - uint64_t mc_addr = 0; int r; /* Allocate memory for backend private data */ smu_data = (struct smu7_smumgr *)(hwmgr->smu_backend); @@ -585,15 +595,12 @@ int smu7_init(struct pp_hwmgr *hwmgr) PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM, &smu_data->header_buffer.handle, - &mc_addr, + &smu_data->header_buffer.mc_addr, &smu_data->header_buffer.kaddr); if (r) return -EINVAL; - smu_data->header = smu_data->header_buffer.kaddr; - smu_data->header_buffer.mc_addr = mc_addr; - if (!hwmgr->not_vf) return 0; @@ -603,7 +610,7 @@ int smu7_init(struct pp_hwmgr *hwmgr) PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM, &smu_data->smu_buffer.handle, - &mc_addr, + &smu_data->smu_buffer.mc_addr, &smu_data->smu_buffer.kaddr); if (r) { @@ -612,7 +619,6 @@ int smu7_init(struct pp_hwmgr *hwmgr) &smu_data->header_buffer.kaddr); return -EINVAL; } - smu_data->smu_buffer.mc_addr = mc_addr; if (smum_is_hw_avfs_present(hwmgr)) hwmgr->avfs_supported = true; @@ -634,6 +640,9 @@ int smu7_smu_fini(struct pp_hwmgr *hwmgr) &smu_data->smu_buffer.mc_addr, &smu_data->smu_buffer.kaddr); + + kfree(smu_data->toc); + smu_data->toc = NULL; kfree(hwmgr->smu_backend); hwmgr->smu_backend = NULL; return 0; diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.h b/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.h index 39c9bfda0ab4..01f0538fba6b 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.h +++ b/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.h @@ -37,10 +37,9 @@ struct smu7_buffer_entry { }; struct smu7_smumgr { - uint8_t *header; - uint8_t *mec_image; struct smu7_buffer_entry smu_buffer; struct smu7_buffer_entry header_buffer; + struct SMU_DRAMData_TOC *toc; uint32_t soft_regs_start; uint32_t dpm_table_start; diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smu8_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/smu8_smumgr.c index c861d3023474..f7e3bc22bb93 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/smu8_smumgr.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/smu8_smumgr.c @@ -52,10 +52,10 @@ static const enum smu8_scratch_entry firmware_list[] = { SMU8_SCRATCH_ENTRY_UCODE_ID_RLC_G, }; -static int smu8_get_argument(struct pp_hwmgr *hwmgr) +static uint32_t smu8_get_argument(struct pp_hwmgr *hwmgr) { if (hwmgr == NULL || hwmgr->device == NULL) - return -EINVAL; + return 0; return cgs_read_register(hwmgr->device, mmSMU_MP1_SRBM2P_ARG_0); diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smu9_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/smu9_smumgr.c new file mode 100644 index 000000000000..079fc8e8f709 --- /dev/null +++ b/drivers/gpu/drm/amd/powerplay/smumgr/smu9_smumgr.c @@ -0,0 +1,150 @@ +/* + * Copyright 2018 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ + +#include "smumgr.h" +#include "vega10_inc.h" +#include "soc15_common.h" +#include "pp_debug.h" + + +/* MP Apertures */ +#define MP0_Public 0x03800000 +#define MP0_SRAM 0x03900000 +#define MP1_Public 0x03b00000 +#define MP1_SRAM 0x03c00004 + +#define smnMP1_FIRMWARE_FLAGS 0x3010028 + +bool smu9_is_smc_ram_running(struct pp_hwmgr *hwmgr) +{ + struct amdgpu_device *adev = hwmgr->adev; + uint32_t mp1_fw_flags; + + WREG32_SOC15(NBIF, 0, mmPCIE_INDEX2, + (MP1_Public | (smnMP1_FIRMWARE_FLAGS & 0xffffffff))); + + mp1_fw_flags = RREG32_SOC15(NBIF, 0, mmPCIE_DATA2); + + if (mp1_fw_flags & MP1_FIRMWARE_FLAGS__INTERRUPTS_ENABLED_MASK) + return true; + + return false; +} + +/* + * Check if SMC has responded to previous message. + * + * @param smumgr the address of the powerplay hardware manager. + * @return TRUE SMC has responded, FALSE otherwise. + */ +static uint32_t smu9_wait_for_response(struct pp_hwmgr *hwmgr) +{ + struct amdgpu_device *adev = hwmgr->adev; + uint32_t reg; + uint32_t ret; + + reg = SOC15_REG_OFFSET(MP1, 0, mmMP1_SMN_C2PMSG_90); + + ret = phm_wait_for_register_unequal(hwmgr, reg, + 0, MP1_C2PMSG_90__CONTENT_MASK); + + if (ret) + pr_err("No response from smu\n"); + + return RREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_90); +} + +/* + * Send a message to the SMC, and do not wait for its response. + * @param smumgr the address of the powerplay hardware manager. + * @param msg the message to send. + * @return Always return 0. + */ +static int smu9_send_msg_to_smc_without_waiting(struct pp_hwmgr *hwmgr, + uint16_t msg) +{ + struct amdgpu_device *adev = hwmgr->adev; + + WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_66, msg); + + return 0; +} + +/* + * Send a message to the SMC, and wait for its response. + * @param hwmgr the address of the powerplay hardware manager. + * @param msg the message to send. + * @return Always return 0. + */ +int smu9_send_msg_to_smc(struct pp_hwmgr *hwmgr, uint16_t msg) +{ + struct amdgpu_device *adev = hwmgr->adev; + uint32_t ret; + + smu9_wait_for_response(hwmgr); + + WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_90, 0); + + smu9_send_msg_to_smc_without_waiting(hwmgr, msg); + + ret = smu9_wait_for_response(hwmgr); + if (ret != 1) + pr_err("Failed to send message: 0x%x, ret value: 0x%x\n", msg, ret); + + return 0; +} + +/* + * Send a message to the SMC with parameter + * @param hwmgr: the address of the powerplay hardware manager. + * @param msg: the message to send. + * @param parameter: the parameter to send + * @return Always return 0. + */ +int smu9_send_msg_to_smc_with_parameter(struct pp_hwmgr *hwmgr, + uint16_t msg, uint32_t parameter) +{ + struct amdgpu_device *adev = hwmgr->adev; + uint32_t ret; + + smu9_wait_for_response(hwmgr); + + WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_90, 0); + + WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_82, parameter); + + smu9_send_msg_to_smc_without_waiting(hwmgr, msg); + + ret = smu9_wait_for_response(hwmgr); + if (ret != 1) + pr_err("Failed message: 0x%x, input parameter: 0x%x, error code: 0x%x\n", msg, parameter, ret); + + return 0; +} + +uint32_t smu9_get_argument(struct pp_hwmgr *hwmgr) +{ + struct amdgpu_device *adev = hwmgr->adev; + + return RREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_82); +} diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smu9_smumgr.h b/drivers/gpu/drm/amd/powerplay/smumgr/smu9_smumgr.h new file mode 100644 index 000000000000..1462279ca128 --- /dev/null +++ b/drivers/gpu/drm/amd/powerplay/smumgr/smu9_smumgr.h @@ -0,0 +1,32 @@ +/* + * Copyright 2018 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ +#ifndef _SMU9_SMUMANAGER_H_ +#define _SMU9_SMUMANAGER_H_ + +bool smu9_is_smc_ram_running(struct pp_hwmgr *hwmgr); +int smu9_send_msg_to_smc(struct pp_hwmgr *hwmgr, uint16_t msg); +int smu9_send_msg_to_smc_with_parameter(struct pp_hwmgr *hwmgr, + uint16_t msg, uint32_t parameter); +uint32_t smu9_get_argument(struct pp_hwmgr *hwmgr); + +#endif diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c index c9837935f0f5..99d5e4f98f49 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c @@ -96,7 +96,7 @@ int smum_process_firmware_header(struct pp_hwmgr *hwmgr) return 0; } -int smum_get_argument(struct pp_hwmgr *hwmgr) +uint32_t smum_get_argument(struct pp_hwmgr *hwmgr) { if (NULL != hwmgr->smumgr_funcs->get_argument) return hwmgr->smumgr_funcs->get_argument(hwmgr); diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/tonga_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/tonga_smumgr.c index 782b19fc2e70..7dabc6c456e1 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/tonga_smumgr.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/tonga_smumgr.c @@ -1443,51 +1443,6 @@ static int tonga_populate_smc_acp_level(struct pp_hwmgr *hwmgr, return result; } -static int tonga_populate_smc_samu_level(struct pp_hwmgr *hwmgr, - SMU72_Discrete_DpmTable *table) -{ - int result = 0; - uint8_t count; - pp_atomctrl_clock_dividers_vi dividers; - struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend); - struct phm_ppt_v1_information *pptable_info = - (struct phm_ppt_v1_information *)(hwmgr->pptable); - phm_ppt_v1_mm_clock_voltage_dependency_table *mm_table = - pptable_info->mm_dep_table; - - table->SamuBootLevel = 0; - table->SamuLevelCount = (uint8_t) (mm_table->count); - - for (count = 0; count < table->SamuLevelCount; count++) { - /* not sure whether we need evclk or not */ - table->SamuLevel[count].Frequency = - pptable_info->mm_dep_table->entries[count].samclock; - table->SamuLevel[count].MinVoltage.Vddc = - phm_get_voltage_index(pptable_info->vddc_lookup_table, - mm_table->entries[count].vddc); - table->SamuLevel[count].MinVoltage.VddGfx = - (data->vdd_gfx_control == SMU7_VOLTAGE_CONTROL_BY_SVID2) ? - phm_get_voltage_index(pptable_info->vddgfx_lookup_table, - mm_table->entries[count].vddgfx) : 0; - table->SamuLevel[count].MinVoltage.Vddci = - phm_get_voltage_id(&data->vddci_voltage_table, - mm_table->entries[count].vddc - VDDC_VDDCI_DELTA); - table->SamuLevel[count].MinVoltage.Phases = 1; - - /* retrieve divider value for VBIOS */ - result = atomctrl_get_dfs_pll_dividers_vi(hwmgr, - table->SamuLevel[count].Frequency, ÷rs); - PP_ASSERT_WITH_CODE((!result), - "can not find divide id for samu clock", return result); - - table->SamuLevel[count].Divider = (uint8_t)dividers.pll_post_divider; - - CONVERT_FROM_HOST_TO_SMC_UL(table->SamuLevel[count].Frequency); - } - - return result; -} - static int tonga_populate_memory_timing_parameters( struct pp_hwmgr *hwmgr, uint32_t engine_clock, @@ -2323,10 +2278,6 @@ static int tonga_init_smc_table(struct pp_hwmgr *hwmgr) PP_ASSERT_WITH_CODE(!result, "Failed to initialize ACP Level !", return result); - result = tonga_populate_smc_samu_level(hwmgr, table); - PP_ASSERT_WITH_CODE(!result, - "Failed to initialize SAMU Level !", return result); - /* Since only the initial state is completely set up at this * point (the other states are just copies of the boot state) we only * need to populate the ARB settings for the initial state. @@ -2673,8 +2624,6 @@ static uint32_t tonga_get_offsetof(uint32_t type, uint32_t member) return offsetof(SMU72_Discrete_DpmTable, UvdBootLevel); case VceBootLevel: return offsetof(SMU72_Discrete_DpmTable, VceBootLevel); - case SamuBootLevel: - return offsetof(SMU72_Discrete_DpmTable, SamuBootLevel); case LowSclkInterruptThreshold: return offsetof(SMU72_Discrete_DpmTable, LowSclkInterruptThreshold); } @@ -2773,32 +2722,6 @@ static int tonga_update_vce_smc_table(struct pp_hwmgr *hwmgr) return 0; } -static int tonga_update_samu_smc_table(struct pp_hwmgr *hwmgr) -{ - struct tonga_smumgr *smu_data = (struct tonga_smumgr *)(hwmgr->smu_backend); - uint32_t mm_boot_level_offset, mm_boot_level_value; - - smu_data->smc_state_table.SamuBootLevel = 0; - mm_boot_level_offset = smu_data->smu7_data.dpm_table_start + - offsetof(SMU72_Discrete_DpmTable, SamuBootLevel); - - mm_boot_level_offset /= 4; - mm_boot_level_offset *= 4; - mm_boot_level_value = cgs_read_ind_register(hwmgr->device, - CGS_IND_REG__SMC, mm_boot_level_offset); - mm_boot_level_value &= 0xFFFFFF00; - mm_boot_level_value |= smu_data->smc_state_table.SamuBootLevel << 0; - cgs_write_ind_register(hwmgr->device, - CGS_IND_REG__SMC, mm_boot_level_offset, mm_boot_level_value); - - if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, - PHM_PlatformCaps_StablePState)) - smum_send_msg_to_smc_with_parameter(hwmgr, - PPSMC_MSG_SAMUDPM_SetEnabledMask, - (uint32_t)(1 << smu_data->smc_state_table.SamuBootLevel)); - return 0; -} - static int tonga_update_smc_table(struct pp_hwmgr *hwmgr, uint32_t type) { switch (type) { @@ -2808,9 +2731,6 @@ static int tonga_update_smc_table(struct pp_hwmgr *hwmgr, uint32_t type) case SMU_VCE_TABLE: tonga_update_vce_smc_table(hwmgr); break; - case SMU_SAMU_TABLE: - tonga_update_samu_smc_table(hwmgr); - break; default: break; } diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.c index e84669c448a3..5d19115f410c 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.c @@ -28,142 +28,11 @@ #include "vega10_hwmgr.h" #include "vega10_ppsmc.h" #include "smu9_driver_if.h" +#include "smu9_smumgr.h" #include "ppatomctrl.h" #include "pp_debug.h" -#define AVFS_EN_MSB 1568 -#define AVFS_EN_LSB 1568 - -/* Microcode file is stored in this buffer */ -#define BUFFER_SIZE 80000 -#define MAX_STRING_SIZE 15 -#define BUFFER_SIZETWO 131072 /* 128 *1024 */ - -/* MP Apertures */ -#define MP0_Public 0x03800000 -#define MP0_SRAM 0x03900000 -#define MP1_Public 0x03b00000 -#define MP1_SRAM 0x03c00004 - -#define smnMP1_FIRMWARE_FLAGS 0x3010028 -#define smnMP0_FW_INTF 0x3010104 -#define smnMP1_PUB_CTRL 0x3010b14 - -static bool vega10_is_smc_ram_running(struct pp_hwmgr *hwmgr) -{ - struct amdgpu_device *adev = hwmgr->adev; - uint32_t mp1_fw_flags; - - WREG32_SOC15(NBIF, 0, mmPCIE_INDEX2, - (MP1_Public | (smnMP1_FIRMWARE_FLAGS & 0xffffffff))); - - mp1_fw_flags = RREG32_SOC15(NBIF, 0, mmPCIE_DATA2); - - if (mp1_fw_flags & MP1_FIRMWARE_FLAGS__INTERRUPTS_ENABLED_MASK) - return true; - - return false; -} - -/* - * Check if SMC has responded to previous message. - * - * @param smumgr the address of the powerplay hardware manager. - * @return TRUE SMC has responded, FALSE otherwise. - */ -static uint32_t vega10_wait_for_response(struct pp_hwmgr *hwmgr) -{ - struct amdgpu_device *adev = hwmgr->adev; - uint32_t reg; - uint32_t ret; - - reg = SOC15_REG_OFFSET(MP1, 0, mmMP1_SMN_C2PMSG_90); - - ret = phm_wait_for_register_unequal(hwmgr, reg, - 0, MP1_C2PMSG_90__CONTENT_MASK); - - if (ret) - pr_err("No response from smu\n"); - - return RREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_90); -} - -/* - * Send a message to the SMC, and do not wait for its response. - * @param smumgr the address of the powerplay hardware manager. - * @param msg the message to send. - * @return Always return 0. - */ -static int vega10_send_msg_to_smc_without_waiting(struct pp_hwmgr *hwmgr, - uint16_t msg) -{ - struct amdgpu_device *adev = hwmgr->adev; - - WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_66, msg); - - return 0; -} - -/* - * Send a message to the SMC, and wait for its response. - * @param hwmgr the address of the powerplay hardware manager. - * @param msg the message to send. - * @return Always return 0. - */ -static int vega10_send_msg_to_smc(struct pp_hwmgr *hwmgr, uint16_t msg) -{ - struct amdgpu_device *adev = hwmgr->adev; - uint32_t ret; - - vega10_wait_for_response(hwmgr); - - WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_90, 0); - - vega10_send_msg_to_smc_without_waiting(hwmgr, msg); - - ret = vega10_wait_for_response(hwmgr); - if (ret != 1) - pr_err("Failed to send message: 0x%x, ret value: 0x%x\n", msg, ret); - - return 0; -} - -/* - * Send a message to the SMC with parameter - * @param hwmgr: the address of the powerplay hardware manager. - * @param msg: the message to send. - * @param parameter: the parameter to send - * @return Always return 0. - */ -static int vega10_send_msg_to_smc_with_parameter(struct pp_hwmgr *hwmgr, - uint16_t msg, uint32_t parameter) -{ - struct amdgpu_device *adev = hwmgr->adev; - uint32_t ret; - - vega10_wait_for_response(hwmgr); - - WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_90, 0); - - WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_82, parameter); - - vega10_send_msg_to_smc_without_waiting(hwmgr, msg); - - ret = vega10_wait_for_response(hwmgr); - if (ret != 1) - pr_err("Failed message: 0x%x, input parameter: 0x%x, error code: 0x%x\n", msg, parameter, ret); - - return 0; -} - -static int vega10_get_argument(struct pp_hwmgr *hwmgr) -{ - struct amdgpu_device *adev = hwmgr->adev; - - return RREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_82); -} - static int vega10_copy_table_from_smc(struct pp_hwmgr *hwmgr, uint8_t *table, int16_t table_id) { @@ -175,13 +44,13 @@ static int vega10_copy_table_from_smc(struct pp_hwmgr *hwmgr, "Invalid SMU Table version!", return -EINVAL); PP_ASSERT_WITH_CODE(priv->smu_tables.entry[table_id].size != 0, "Invalid SMU Table Length!", return -EINVAL); - vega10_send_msg_to_smc_with_parameter(hwmgr, + smu9_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_SetDriverDramAddrHigh, upper_32_bits(priv->smu_tables.entry[table_id].mc_addr)); - vega10_send_msg_to_smc_with_parameter(hwmgr, + smu9_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_SetDriverDramAddrLow, lower_32_bits(priv->smu_tables.entry[table_id].mc_addr)); - vega10_send_msg_to_smc_with_parameter(hwmgr, + smu9_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_TransferTableSmu2Dram, priv->smu_tables.entry[table_id].table_id); @@ -206,13 +75,13 @@ static int vega10_copy_table_to_smc(struct pp_hwmgr *hwmgr, memcpy(priv->smu_tables.entry[table_id].table, table, priv->smu_tables.entry[table_id].size); - vega10_send_msg_to_smc_with_parameter(hwmgr, + smu9_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_SetDriverDramAddrHigh, upper_32_bits(priv->smu_tables.entry[table_id].mc_addr)); - vega10_send_msg_to_smc_with_parameter(hwmgr, + smu9_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_SetDriverDramAddrLow, lower_32_bits(priv->smu_tables.entry[table_id].mc_addr)); - vega10_send_msg_to_smc_with_parameter(hwmgr, + smu9_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_TransferTableDram2Smu, priv->smu_tables.entry[table_id].table_id); @@ -225,8 +94,8 @@ static int vega10_get_smc_features(struct pp_hwmgr *hwmgr, if (features_enabled == NULL) return -EINVAL; - vega10_send_msg_to_smc(hwmgr, PPSMC_MSG_GetEnabledSmuFeatures); - *features_enabled = vega10_get_argument(hwmgr); + smu9_send_msg_to_smc(hwmgr, PPSMC_MSG_GetEnabledSmuFeatures); + *features_enabled = smu9_get_argument(hwmgr); return 0; } @@ -248,10 +117,10 @@ static int vega10_set_tools_address(struct pp_hwmgr *hwmgr) struct vega10_smumgr *priv = hwmgr->smu_backend; if (priv->smu_tables.entry[TOOLSTABLE].mc_addr) { - vega10_send_msg_to_smc_with_parameter(hwmgr, + smu9_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_SetToolsDramAddrHigh, upper_32_bits(priv->smu_tables.entry[TOOLSTABLE].mc_addr)); - vega10_send_msg_to_smc_with_parameter(hwmgr, + smu9_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_SetToolsDramAddrLow, lower_32_bits(priv->smu_tables.entry[TOOLSTABLE].mc_addr)); } @@ -265,11 +134,11 @@ static int vega10_verify_smc_interface(struct pp_hwmgr *hwmgr) uint32_t dev_id; uint32_t rev_id; - PP_ASSERT_WITH_CODE(!vega10_send_msg_to_smc(hwmgr, + PP_ASSERT_WITH_CODE(!smu9_send_msg_to_smc(hwmgr, PPSMC_MSG_GetDriverIfVersion), "Attempt to get SMC IF Version Number Failed!", return -EINVAL); - smc_driver_if_version = vega10_get_argument(hwmgr); + smc_driver_if_version = smu9_get_argument(hwmgr); dev_id = adev->pdev->device; rev_id = adev->pdev->revision; @@ -441,7 +310,7 @@ static int vega10_smu_fini(struct pp_hwmgr *hwmgr) static int vega10_start_smu(struct pp_hwmgr *hwmgr) { - if (!vega10_is_smc_ram_running(hwmgr)) + if (!smu9_is_smc_ram_running(hwmgr)) return -EINVAL; PP_ASSERT_WITH_CODE(!vega10_verify_smc_interface(hwmgr), @@ -453,7 +322,8 @@ static int vega10_start_smu(struct pp_hwmgr *hwmgr) return 0; } -static int vega10_smc_table_manager(struct pp_hwmgr *hwmgr, uint8_t *table, uint16_t table_id, bool rw) +static int vega10_smc_table_manager(struct pp_hwmgr *hwmgr, uint8_t *table, + uint16_t table_id, bool rw) { int ret; @@ -470,11 +340,11 @@ const struct pp_smumgr_func vega10_smu_funcs = { .smu_fini = &vega10_smu_fini, .start_smu = &vega10_start_smu, .request_smu_load_specific_fw = NULL, - .send_msg_to_smc = &vega10_send_msg_to_smc, - .send_msg_to_smc_with_parameter = &vega10_send_msg_to_smc_with_parameter, + .send_msg_to_smc = &smu9_send_msg_to_smc, + .send_msg_to_smc_with_parameter = &smu9_send_msg_to_smc_with_parameter, .download_pptable_settings = NULL, .upload_pptable_settings = NULL, .is_dpm_running = vega10_is_dpm_running, - .get_argument = vega10_get_argument, + .get_argument = smu9_get_argument, .smc_table_manager = vega10_smc_table_manager, }; diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/vega12_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/vega12_smumgr.c index 7d9b40e8b1bf..7f0e2109f40d 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/vega12_smumgr.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/vega12_smumgr.c @@ -24,157 +24,14 @@ #include "smumgr.h" #include "vega12_inc.h" #include "soc15_common.h" +#include "smu9_smumgr.h" #include "vega12_smumgr.h" #include "vega12_ppsmc.h" #include "vega12/smu9_driver_if.h" - #include "ppatomctrl.h" #include "pp_debug.h" -/* MP Apertures */ -#define MP0_Public 0x03800000 -#define MP0_SRAM 0x03900000 -#define MP1_Public 0x03b00000 -#define MP1_SRAM 0x03c00004 - -#define smnMP1_FIRMWARE_FLAGS 0x3010028 -#define smnMP0_FW_INTF 0x3010104 -#define smnMP1_PUB_CTRL 0x3010b14 - -static bool vega12_is_smc_ram_running(struct pp_hwmgr *hwmgr) -{ - struct amdgpu_device *adev = hwmgr->adev; - uint32_t mp1_fw_flags; - - WREG32_SOC15(NBIF, 0, mmPCIE_INDEX2, - (MP1_Public | (smnMP1_FIRMWARE_FLAGS & 0xffffffff))); - - mp1_fw_flags = RREG32_SOC15(NBIF, 0, mmPCIE_DATA2); - - if ((mp1_fw_flags & MP1_FIRMWARE_FLAGS__INTERRUPTS_ENABLED_MASK) >> - MP1_FIRMWARE_FLAGS__INTERRUPTS_ENABLED__SHIFT) - return true; - - return false; -} - -/* - * Check if SMC has responded to previous message. - * - * @param smumgr the address of the powerplay hardware manager. - * @return TRUE SMC has responded, FALSE otherwise. - */ -static uint32_t vega12_wait_for_response(struct pp_hwmgr *hwmgr) -{ - struct amdgpu_device *adev = hwmgr->adev; - uint32_t reg; - - reg = SOC15_REG_OFFSET(MP1, 0, mmMP1_SMN_C2PMSG_90); - - phm_wait_for_register_unequal(hwmgr, reg, - 0, MP1_C2PMSG_90__CONTENT_MASK); - - return RREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_90); -} - -/* - * Send a message to the SMC, and do not wait for its response. - * @param smumgr the address of the powerplay hardware manager. - * @param msg the message to send. - * @return Always return 0. - */ -int vega12_send_msg_to_smc_without_waiting(struct pp_hwmgr *hwmgr, - uint16_t msg) -{ - struct amdgpu_device *adev = hwmgr->adev; - - WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_66, msg); - - return 0; -} - -/* - * Send a message to the SMC, and wait for its response. - * @param hwmgr the address of the powerplay hardware manager. - * @param msg the message to send. - * @return Always return 0. - */ -int vega12_send_msg_to_smc(struct pp_hwmgr *hwmgr, uint16_t msg) -{ - struct amdgpu_device *adev = hwmgr->adev; - - vega12_wait_for_response(hwmgr); - - WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_90, 0); - - vega12_send_msg_to_smc_without_waiting(hwmgr, msg); - - if (vega12_wait_for_response(hwmgr) != 1) - pr_err("Failed to send message: 0x%x\n", msg); - - return 0; -} - -/* - * Send a message to the SMC with parameter - * @param hwmgr: the address of the powerplay hardware manager. - * @param msg: the message to send. - * @param parameter: the parameter to send - * @return Always return 0. - */ -int vega12_send_msg_to_smc_with_parameter(struct pp_hwmgr *hwmgr, - uint16_t msg, uint32_t parameter) -{ - struct amdgpu_device *adev = hwmgr->adev; - - vega12_wait_for_response(hwmgr); - - WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_90, 0); - - WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_82, parameter); - - vega12_send_msg_to_smc_without_waiting(hwmgr, msg); - - if (vega12_wait_for_response(hwmgr) != 1) - pr_err("Failed to send message: 0x%x\n", msg); - - return 0; -} - - -/* - * Send a message to the SMC with parameter, do not wait for response - * @param hwmgr: the address of the powerplay hardware manager. - * @param msg: the message to send. - * @param parameter: the parameter to send - * @return The response that came from the SMC. - */ -int vega12_send_msg_to_smc_with_parameter_without_waiting( - struct pp_hwmgr *hwmgr, uint16_t msg, uint32_t parameter) -{ - struct amdgpu_device *adev = hwmgr->adev; - - WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_66, parameter); - - return vega12_send_msg_to_smc_without_waiting(hwmgr, msg); -} - -/* - * Retrieve an argument from SMC. - * @param hwmgr the address of the powerplay hardware manager. - * @param arg pointer to store the argument from SMC. - * @return Always return 0. - */ -int vega12_read_arg_from_smc(struct pp_hwmgr *hwmgr, uint32_t *arg) -{ - struct amdgpu_device *adev = hwmgr->adev; - - *arg = RREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_82); - - return 0; -} - /* * Copy table from SMC into driver FB * @param hwmgr the address of the HW manager @@ -192,16 +49,16 @@ int vega12_copy_table_from_smc(struct pp_hwmgr *hwmgr, "Invalid SMU Table version!", return -EINVAL); PP_ASSERT_WITH_CODE(priv->smu_tables.entry[table_id].size != 0, "Invalid SMU Table Length!", return -EINVAL); - PP_ASSERT_WITH_CODE(vega12_send_msg_to_smc_with_parameter(hwmgr, + PP_ASSERT_WITH_CODE(smu9_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_SetDriverDramAddrHigh, upper_32_bits(priv->smu_tables.entry[table_id].mc_addr)) == 0, "[CopyTableFromSMC] Attempt to Set Dram Addr High Failed!", return -EINVAL); - PP_ASSERT_WITH_CODE(vega12_send_msg_to_smc_with_parameter(hwmgr, + PP_ASSERT_WITH_CODE(smu9_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_SetDriverDramAddrLow, lower_32_bits(priv->smu_tables.entry[table_id].mc_addr)) == 0, "[CopyTableFromSMC] Attempt to Set Dram Addr Low Failed!", return -EINVAL); - PP_ASSERT_WITH_CODE(vega12_send_msg_to_smc_with_parameter(hwmgr, + PP_ASSERT_WITH_CODE(smu9_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_TransferTableSmu2Dram, table_id) == 0, "[CopyTableFromSMC] Attempt to Transfer Table From SMU Failed!", @@ -234,17 +91,17 @@ int vega12_copy_table_to_smc(struct pp_hwmgr *hwmgr, memcpy(priv->smu_tables.entry[table_id].table, table, priv->smu_tables.entry[table_id].size); - PP_ASSERT_WITH_CODE(vega12_send_msg_to_smc_with_parameter(hwmgr, + PP_ASSERT_WITH_CODE(smu9_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_SetDriverDramAddrHigh, upper_32_bits(priv->smu_tables.entry[table_id].mc_addr)) == 0, "[CopyTableToSMC] Attempt to Set Dram Addr High Failed!", return -EINVAL;); - PP_ASSERT_WITH_CODE(vega12_send_msg_to_smc_with_parameter(hwmgr, + PP_ASSERT_WITH_CODE(smu9_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_SetDriverDramAddrLow, lower_32_bits(priv->smu_tables.entry[table_id].mc_addr)) == 0, "[CopyTableToSMC] Attempt to Set Dram Addr Low Failed!", return -EINVAL); - PP_ASSERT_WITH_CODE(vega12_send_msg_to_smc_with_parameter(hwmgr, + PP_ASSERT_WITH_CODE(smu9_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_TransferTableDram2Smu, table_id) == 0, "[CopyTableToSMC] Attempt to Transfer Table To SMU Failed!", @@ -262,20 +119,20 @@ int vega12_enable_smc_features(struct pp_hwmgr *hwmgr, smu_features_high = (uint32_t)((feature_mask & SMU_FEATURES_HIGH_MASK) >> SMU_FEATURES_HIGH_SHIFT); if (enable) { - PP_ASSERT_WITH_CODE(vega12_send_msg_to_smc_with_parameter(hwmgr, + PP_ASSERT_WITH_CODE(smu9_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_EnableSmuFeaturesLow, smu_features_low) == 0, "[EnableDisableSMCFeatures] Attemp to enable SMU features Low failed!", return -EINVAL); - PP_ASSERT_WITH_CODE(vega12_send_msg_to_smc_with_parameter(hwmgr, + PP_ASSERT_WITH_CODE(smu9_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_EnableSmuFeaturesHigh, smu_features_high) == 0, "[EnableDisableSMCFeatures] Attemp to enable SMU features High failed!", return -EINVAL); } else { - PP_ASSERT_WITH_CODE(vega12_send_msg_to_smc_with_parameter(hwmgr, + PP_ASSERT_WITH_CODE(smu9_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_DisableSmuFeaturesLow, smu_features_low) == 0, "[EnableDisableSMCFeatures] Attemp to disable SMU features Low failed!", return -EINVAL); - PP_ASSERT_WITH_CODE(vega12_send_msg_to_smc_with_parameter(hwmgr, + PP_ASSERT_WITH_CODE(smu9_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_DisableSmuFeaturesHigh, smu_features_high) == 0, "[EnableDisableSMCFeatures] Attemp to disable SMU features High failed!", return -EINVAL); @@ -292,22 +149,17 @@ int vega12_get_enabled_smc_features(struct pp_hwmgr *hwmgr, if (features_enabled == NULL) return -EINVAL; - PP_ASSERT_WITH_CODE(vega12_send_msg_to_smc(hwmgr, + PP_ASSERT_WITH_CODE(smu9_send_msg_to_smc(hwmgr, PPSMC_MSG_GetEnabledSmuFeaturesLow) == 0, "[GetEnabledSMCFeatures] Attemp to get SMU features Low failed!", return -EINVAL); - PP_ASSERT_WITH_CODE(vega12_read_arg_from_smc(hwmgr, - &smc_features_low) == 0, - "[GetEnabledSMCFeatures] Attemp to read SMU features Low argument failed!", - return -EINVAL); - PP_ASSERT_WITH_CODE(vega12_send_msg_to_smc(hwmgr, + smc_features_low = smu9_get_argument(hwmgr); + + PP_ASSERT_WITH_CODE(smu9_send_msg_to_smc(hwmgr, PPSMC_MSG_GetEnabledSmuFeaturesHigh) == 0, "[GetEnabledSMCFeatures] Attemp to get SMU features High failed!", return -EINVAL); - PP_ASSERT_WITH_CODE(vega12_read_arg_from_smc(hwmgr, - &smc_features_high) == 0, - "[GetEnabledSMCFeatures] Attemp to read SMU features High argument failed!", - return -EINVAL); + smc_features_high = smu9_get_argument(hwmgr); *features_enabled = ((((uint64_t)smc_features_low << SMU_FEATURES_LOW_SHIFT) & SMU_FEATURES_LOW_MASK) | (((uint64_t)smc_features_high << SMU_FEATURES_HIGH_SHIFT) & SMU_FEATURES_HIGH_MASK)); @@ -333,39 +185,16 @@ static int vega12_set_tools_address(struct pp_hwmgr *hwmgr) (struct vega12_smumgr *)(hwmgr->smu_backend); if (priv->smu_tables.entry[TABLE_PMSTATUSLOG].mc_addr) { - if (!vega12_send_msg_to_smc_with_parameter(hwmgr, + if (!smu9_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_SetToolsDramAddrHigh, upper_32_bits(priv->smu_tables.entry[TABLE_PMSTATUSLOG].mc_addr))) - vega12_send_msg_to_smc_with_parameter(hwmgr, + smu9_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_SetToolsDramAddrLow, lower_32_bits(priv->smu_tables.entry[TABLE_PMSTATUSLOG].mc_addr)); } return 0; } -#if 0 /* tentatively remove */ -static int vega12_verify_smc_interface(struct pp_hwmgr *hwmgr) -{ - uint32_t smc_driver_if_version; - - PP_ASSERT_WITH_CODE(!vega12_send_msg_to_smc(hwmgr, - PPSMC_MSG_GetDriverIfVersion), - "Attempt to get SMC IF Version Number Failed!", - return -EINVAL); - vega12_read_arg_from_smc(hwmgr, &smc_driver_if_version); - - if (smc_driver_if_version != SMU9_DRIVER_IF_VERSION) { - pr_err("Your firmware(0x%x) doesn't match \ - SMU9_DRIVER_IF_VERSION(0x%x). \ - Please update your firmware!\n", - smc_driver_if_version, SMU9_DRIVER_IF_VERSION); - return -EINVAL; - } - - return 0; -} -#endif - static int vega12_smu_init(struct pp_hwmgr *hwmgr) { struct vega12_smumgr *priv; @@ -513,16 +342,10 @@ static int vega12_smu_fini(struct pp_hwmgr *hwmgr) static int vega12_start_smu(struct pp_hwmgr *hwmgr) { - PP_ASSERT_WITH_CODE(vega12_is_smc_ram_running(hwmgr), + PP_ASSERT_WITH_CODE(smu9_is_smc_ram_running(hwmgr), "SMC is not running!", return -EINVAL); -#if 0 /* tentatively remove */ - PP_ASSERT_WITH_CODE(!vega12_verify_smc_interface(hwmgr), - "Failed to verify SMC interface!", - return -EINVAL); -#endif - vega12_set_tools_address(hwmgr); return 0; @@ -533,9 +356,10 @@ const struct pp_smumgr_func vega12_smu_funcs = { .smu_fini = &vega12_smu_fini, .start_smu = &vega12_start_smu, .request_smu_load_specific_fw = NULL, - .send_msg_to_smc = &vega12_send_msg_to_smc, - .send_msg_to_smc_with_parameter = &vega12_send_msg_to_smc_with_parameter, + .send_msg_to_smc = &smu9_send_msg_to_smc, + .send_msg_to_smc_with_parameter = &smu9_send_msg_to_smc_with_parameter, .download_pptable_settings = NULL, .upload_pptable_settings = NULL, .is_dpm_running = vega12_is_dpm_running, + .get_argument = smu9_get_argument, }; diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/vega12_smumgr.h b/drivers/gpu/drm/amd/powerplay/smumgr/vega12_smumgr.h index 2810d387b611..b285cbc04019 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/vega12_smumgr.h +++ b/drivers/gpu/drm/amd/powerplay/smumgr/vega12_smumgr.h @@ -48,7 +48,6 @@ struct vega12_smumgr { #define SMU_FEATURES_HIGH_MASK 0xFFFFFFFF00000000 #define SMU_FEATURES_HIGH_SHIFT 32 -int vega12_read_arg_from_smc(struct pp_hwmgr *hwmgr, uint32_t *arg); int vega12_copy_table_from_smc(struct pp_hwmgr *hwmgr, uint8_t *table, int16_t table_id); int vega12_copy_table_to_smc(struct pp_hwmgr *hwmgr, diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/vegam_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/vegam_smumgr.c index 2de48959ac93..57420d7caa4e 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/vegam_smumgr.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/vegam_smumgr.c @@ -393,34 +393,6 @@ static int vegam_update_vce_smc_table(struct pp_hwmgr *hwmgr) return 0; } -static int vegam_update_samu_smc_table(struct pp_hwmgr *hwmgr) -{ - struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend); - uint32_t mm_boot_level_offset, mm_boot_level_value; - - - smu_data->smc_state_table.SamuBootLevel = 0; - mm_boot_level_offset = smu_data->smu7_data.dpm_table_start + - offsetof(SMU75_Discrete_DpmTable, SamuBootLevel); - - mm_boot_level_offset /= 4; - mm_boot_level_offset *= 4; - mm_boot_level_value = cgs_read_ind_register(hwmgr->device, - CGS_IND_REG__SMC, mm_boot_level_offset); - mm_boot_level_value &= 0xFFFFFF00; - mm_boot_level_value |= smu_data->smc_state_table.SamuBootLevel << 0; - cgs_write_ind_register(hwmgr->device, - CGS_IND_REG__SMC, mm_boot_level_offset, mm_boot_level_value); - - if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, - PHM_PlatformCaps_StablePState)) - smum_send_msg_to_smc_with_parameter(hwmgr, - PPSMC_MSG_SAMUDPM_SetEnabledMask, - (uint32_t)(1 << smu_data->smc_state_table.SamuBootLevel)); - return 0; -} - - static int vegam_update_bif_smc_table(struct pp_hwmgr *hwmgr) { struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend); @@ -447,9 +419,6 @@ static int vegam_update_smc_table(struct pp_hwmgr *hwmgr, uint32_t type) case SMU_VCE_TABLE: vegam_update_vce_smc_table(hwmgr); break; - case SMU_SAMU_TABLE: - vegam_update_samu_smc_table(hwmgr); - break; case SMU_BIF_TABLE: vegam_update_bif_smc_table(hwmgr); break; @@ -1281,54 +1250,6 @@ static int vegam_populate_smc_vce_level(struct pp_hwmgr *hwmgr, return result; } -static int vegam_populate_smc_samu_level(struct pp_hwmgr *hwmgr, - SMU75_Discrete_DpmTable *table) -{ - int result = -EINVAL; - uint8_t count; - struct pp_atomctrl_clock_dividers_vi dividers; - struct phm_ppt_v1_information *table_info = - (struct phm_ppt_v1_information *)(hwmgr->pptable); - struct phm_ppt_v1_mm_clock_voltage_dependency_table *mm_table = - table_info->mm_dep_table; - struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend); - uint32_t vddci; - - table->SamuBootLevel = 0; - table->SamuLevelCount = (uint8_t)(mm_table->count); - - for (count = 0; count < table->SamuLevelCount; count++) { - /* not sure whether we need evclk or not */ - table->SamuLevel[count].MinVoltage = 0; - table->SamuLevel[count].Frequency = mm_table->entries[count].samclock; - table->SamuLevel[count].MinVoltage |= (mm_table->entries[count].vddc * - VOLTAGE_SCALE) << VDDC_SHIFT; - - if (SMU7_VOLTAGE_CONTROL_BY_GPIO == data->vddci_control) - vddci = (uint32_t)phm_find_closest_vddci(&(data->vddci_voltage_table), - mm_table->entries[count].vddc - VDDC_VDDCI_DELTA); - else if (SMU7_VOLTAGE_CONTROL_BY_SVID2 == data->vddci_control) - vddci = mm_table->entries[count].vddc - VDDC_VDDCI_DELTA; - else - vddci = (data->vbios_boot_state.vddci_bootup_value * VOLTAGE_SCALE) << VDDCI_SHIFT; - - table->SamuLevel[count].MinVoltage |= (vddci * VOLTAGE_SCALE) << VDDCI_SHIFT; - table->SamuLevel[count].MinVoltage |= 1 << PHASES_SHIFT; - - /* retrieve divider value for VBIOS */ - result = atomctrl_get_dfs_pll_dividers_vi(hwmgr, - table->SamuLevel[count].Frequency, ÷rs); - PP_ASSERT_WITH_CODE((0 == result), - "can not find divide id for samu clock", return result); - - table->SamuLevel[count].Divider = (uint8_t)dividers.pll_post_divider; - - CONVERT_FROM_HOST_TO_SMC_UL(table->SamuLevel[count].Frequency); - CONVERT_FROM_HOST_TO_SMC_UL(table->SamuLevel[count].MinVoltage); - } - return result; -} - static int vegam_populate_memory_timing_parameters(struct pp_hwmgr *hwmgr, int32_t eng_clock, int32_t mem_clock, SMU75_Discrete_MCArbDramTimingTableEntry *arb_regs) @@ -2062,10 +1983,6 @@ static int vegam_init_smc_table(struct pp_hwmgr *hwmgr) PP_ASSERT_WITH_CODE(!result, "Failed to initialize VCE Level!", return result); - result = vegam_populate_smc_samu_level(hwmgr, table); - PP_ASSERT_WITH_CODE(!result, - "Failed to initialize SAMU Level!", return result); - /* Since only the initial state is completely set up at this point * (the other states are just copies of the boot state) we only * need to populate the ARB settings for the initial state. @@ -2273,8 +2190,6 @@ static uint32_t vegam_get_offsetof(uint32_t type, uint32_t member) return offsetof(SMU75_Discrete_DpmTable, UvdBootLevel); case VceBootLevel: return offsetof(SMU75_Discrete_DpmTable, VceBootLevel); - case SamuBootLevel: - return offsetof(SMU75_Discrete_DpmTable, SamuBootLevel); case LowSclkInterruptThreshold: return offsetof(SMU75_Discrete_DpmTable, LowSclkInterruptThreshold); } |