diff options
98 files changed, 1549 insertions, 1981 deletions
diff --git a/MAINTAINERS b/MAINTAINERS index 225bc22c4106..41de15567b8a 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -14063,7 +14063,6 @@ F: drivers/net/wireless/quantenna RADEON and AMDGPU DRM DRIVERS M: Alex Deucher <alexander.deucher@amd.com> M: Christian König <christian.koenig@amd.com> -M: David (ChunMing) Zhou <David1.Zhou@amd.com> L: amd-gfx@lists.freedesktop.org S: Supported T: git git://people.freedesktop.org/~agd5f/linux diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index 99e5f474505d..4a03a24348f4 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -765,7 +765,6 @@ struct amdgpu_device { uint8_t *bios; uint32_t bios_size; struct amdgpu_bo *stolen_vga_memory; - struct amdgpu_bo *discovery_memory; uint32_t bios_scratch_reg_offset; uint32_t bios_scratch[AMDGPU_BIOS_NUM_SCRATCH]; @@ -918,7 +917,9 @@ struct amdgpu_device { struct amdgpu_display_manager dm; /* discovery */ - uint8_t *discovery; + uint8_t *discovery_bin; + uint32_t discovery_tmr_size; + struct amdgpu_bo *discovery_memory; /* mes */ bool enable_mes; @@ -957,6 +958,7 @@ struct amdgpu_device { /* s3/s4 mask */ bool in_suspend; + bool in_hibernate; bool in_gpu_reset; enum pp_mp1_state mp1_state; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c index 1e41367ef74e..956cbbda4793 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c @@ -444,7 +444,6 @@ static int amdgpu_atif_handler(struct amdgpu_device *adev, DRM_DEBUG_DRIVER("ATIF: %d pending SBIOS requests\n", count); - /* 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; @@ -463,6 +462,27 @@ static int amdgpu_atif_handler(struct amdgpu_device *adev, #endif } } +#if defined(CONFIG_DRM_AMD_DC) +#if defined(CONFIG_BACKLIGHT_CLASS_DEVICE) || defined(CONFIG_BACKLIGHT_CLASS_DEVICE_MODULE) + if ((req.pending & ATIF_PANEL_BRIGHTNESS_CHANGE_REQUEST) && + amdgpu_device_has_dc_support(adev)) { + struct amdgpu_display_manager *dm = &adev->dm; + struct backlight_device *bd = dm->backlight_dev; + + if (bd) { + DRM_DEBUG_DRIVER("Changing brightness to %d\n", + req.backlight_level); + + /* + * XXX backlight_device_set_brightness() is + * hardwired to post BACKLIGHT_UPDATE_SYSFS. + * It probably should accept 'reason' parameter. + */ + backlight_device_set_brightness(bd, req.backlight_level); + } + } +#endif +#endif if (req.pending & ATIF_DGPU_DISPLAY_EVENT) { if (adev->flags & AMD_IS_PX) { pm_runtime_get_sync(adev->ddev->dev); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h index d065c50582eb..3f2b695cf19e 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h @@ -65,6 +65,7 @@ struct kgd_mem { struct amdgpu_sync sync; bool aql_queue; + bool is_imported; }; /* KFD Memory Eviction */ @@ -148,6 +149,9 @@ int amdgpu_amdkfd_post_reset(struct amdgpu_device *adev); void amdgpu_amdkfd_gpu_reset(struct kgd_dev *kgd); +int amdgpu_queue_mask_bit_to_set_resource_bit(struct amdgpu_device *adev, + int queue_bit); + /* Shared API */ int amdgpu_amdkfd_alloc_gtt_mem(struct kgd_dev *kgd, size_t size, void **mem_obj, uint64_t *gpu_addr, @@ -219,7 +223,7 @@ int amdgpu_amdkfd_gpuvm_alloc_memory_of_gpu( void *vm, struct kgd_mem **mem, uint64_t *offset, uint32_t flags); int amdgpu_amdkfd_gpuvm_free_memory_of_gpu( - struct kgd_dev *kgd, struct kgd_mem *mem); + struct kgd_dev *kgd, struct kgd_mem *mem, uint64_t *size); int amdgpu_amdkfd_gpuvm_map_memory_to_gpu( struct kgd_dev *kgd, struct kgd_mem *mem, void *vm); int amdgpu_amdkfd_gpuvm_unmap_memory_from_gpu( diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c index 0768b7eb7683..da8b31a53291 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c @@ -1277,7 +1277,7 @@ err: } int amdgpu_amdkfd_gpuvm_free_memory_of_gpu( - struct kgd_dev *kgd, struct kgd_mem *mem) + struct kgd_dev *kgd, struct kgd_mem *mem, uint64_t *size) { struct amdkfd_process_info *process_info = mem->process_info; unsigned long bo_size = mem->bo->tbo.mem.size; @@ -1286,9 +1286,11 @@ int amdgpu_amdkfd_gpuvm_free_memory_of_gpu( struct ttm_validate_buffer *bo_list_entry; unsigned int mapped_to_gpu_memory; int ret; + bool is_imported = 0; mutex_lock(&mem->lock); mapped_to_gpu_memory = mem->mapped_to_gpu_memory; + is_imported = mem->is_imported; mutex_unlock(&mem->lock); /* lock is not needed after this, since mem is unused and will * be freed anyway @@ -1340,8 +1342,19 @@ int amdgpu_amdkfd_gpuvm_free_memory_of_gpu( kfree(mem->bo->tbo.sg); } + /* Update the size of the BO being freed if it was allocated from + * VRAM and is not imported. + */ + if (size) { + if ((mem->bo->preferred_domains == AMDGPU_GEM_DOMAIN_VRAM) && + (!is_imported)) + *size = bo_size; + else + *size = 0; + } + /* Free the BO*/ - amdgpu_bo_unref(&mem->bo); + drm_gem_object_put_unlocked(&mem->bo->tbo.base); mutex_destroy(&mem->lock); kfree(mem); @@ -1686,7 +1699,8 @@ int amdgpu_amdkfd_gpuvm_import_dmabuf(struct kgd_dev *kgd, | KFD_IOC_ALLOC_MEM_FLAGS_WRITABLE | KFD_IOC_ALLOC_MEM_FLAGS_EXECUTABLE; - (*mem)->bo = amdgpu_bo_ref(bo); + drm_gem_object_get(&bo->tbo.base); + (*mem)->bo = bo; (*mem)->va = va; (*mem)->domain = (bo->preferred_domains & AMDGPU_GEM_DOMAIN_VRAM) ? AMDGPU_GEM_DOMAIN_VRAM : AMDGPU_GEM_DOMAIN_GTT; @@ -1694,6 +1708,7 @@ int amdgpu_amdkfd_gpuvm_import_dmabuf(struct kgd_dev *kgd, (*mem)->process_info = avm->process_info; add_kgd_mem_to_kfd_bo_list(*mem, avm->process_info, false); amdgpu_sync_create(&(*mem)->sync); + (*mem)->is_imported = true; return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c index 7653f62b1b2d..19070226a945 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c @@ -1208,7 +1208,6 @@ static int amdgpu_cs_submit(struct amdgpu_cs_parser *p, { struct amdgpu_fpriv *fpriv = p->filp->driver_priv; struct drm_sched_entity *entity = p->entity; - enum drm_sched_priority priority; struct amdgpu_bo_list_entry *e; struct amdgpu_job *job; uint64_t seq; @@ -1258,7 +1257,6 @@ static int amdgpu_cs_submit(struct amdgpu_cs_parser *p, 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); amdgpu_vm_move_to_lru_tail(p->adev, &fpriv->vm); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index b038ddbb2ece..cc41e8f5ad14 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -68,6 +68,7 @@ #include <linux/suspend.h> #include <drm/task_barrier.h> +#include <linux/pm_runtime.h> MODULE_FIRMWARE("amdgpu/vega10_gpu_info.bin"); MODULE_FIRMWARE("amdgpu/vega12_gpu_info.bin"); @@ -254,6 +255,32 @@ void amdgpu_device_vram_access(struct amdgpu_device *adev, loff_t pos, uint32_t hi = ~0; uint64_t last; + +#ifdef CONFIG_64BIT + last = min(pos + size, adev->gmc.visible_vram_size); + if (last > pos) { + void __iomem *addr = adev->mman.aper_base_kaddr + pos; + size_t count = last - pos; + + if (write) { + memcpy_toio(addr, buf, count); + mb(); + amdgpu_asic_flush_hdp(adev, NULL); + } else { + amdgpu_asic_invalidate_hdp(adev, NULL); + mb(); + memcpy_fromio(buf, addr, count); + } + + if (count == size) + return; + + pos += count; + buf += count / 4; + size -= count; + } +#endif + spin_lock_irqsave(&adev->mmio_idx_lock, flags); for (last = pos + size; pos < last; pos += 4) { uint32_t tmp = pos >> 31; @@ -2891,6 +2918,14 @@ static int amdgpu_device_get_job_timeout_settings(struct amdgpu_device *adev) return ret; } +static const struct attribute *amdgpu_dev_attributes[] = { + &dev_attr_product_name.attr, + &dev_attr_product_number.attr, + &dev_attr_serial_number.attr, + &dev_attr_pcie_replay_count.attr, + NULL +}; + /** * amdgpu_device_init - initialize the driver * @@ -3240,27 +3275,9 @@ fence_driver_init: queue_delayed_work(system_wq, &adev->delayed_init_work, msecs_to_jiffies(AMDGPU_RESUME_MS)); - r = device_create_file(adev->dev, &dev_attr_pcie_replay_count); + r = sysfs_create_files(&adev->dev->kobj, amdgpu_dev_attributes); if (r) { - dev_err(adev->dev, "Could not create pcie_replay_count"); - return r; - } - - r = device_create_file(adev->dev, &dev_attr_product_name); - if (r) { - dev_err(adev->dev, "Could not create product_name"); - return r; - } - - r = device_create_file(adev->dev, &dev_attr_product_number); - if (r) { - dev_err(adev->dev, "Could not create product_number"); - return r; - } - - r = device_create_file(adev->dev, &dev_attr_serial_number); - if (r) { - dev_err(adev->dev, "Could not create serial_number"); + dev_err(adev->dev, "Could not create amdgpu device attr\n"); return r; } @@ -3343,12 +3360,10 @@ void amdgpu_device_fini(struct amdgpu_device *adev) adev->rmmio = NULL; amdgpu_device_doorbell_fini(adev); - device_remove_file(adev->dev, &dev_attr_pcie_replay_count); if (adev->ucode_sysfs_en) amdgpu_ucode_sysfs_fini(adev); - device_remove_file(adev->dev, &dev_attr_product_name); - device_remove_file(adev->dev, &dev_attr_product_number); - device_remove_file(adev->dev, &dev_attr_serial_number); + + sysfs_remove_files(&adev->dev->kobj, amdgpu_dev_attributes); if (IS_ENABLED(CONFIG_PERF_EVENTS)) amdgpu_pmu_fini(adev); if (amdgpu_discovery && adev->asic_type >= CHIP_NAVI10) @@ -4116,6 +4131,64 @@ static void amdgpu_device_unlock_adev(struct amdgpu_device *adev) mutex_unlock(&adev->lock_reset); } +static void amdgpu_device_resume_display_audio(struct amdgpu_device *adev) +{ + struct pci_dev *p = NULL; + + p = pci_get_domain_bus_and_slot(pci_domain_nr(adev->pdev->bus), + adev->pdev->bus->number, 1); + if (p) { + pm_runtime_enable(&(p->dev)); + pm_runtime_resume(&(p->dev)); + } +} + +static int amdgpu_device_suspend_display_audio(struct amdgpu_device *adev) +{ + enum amd_reset_method reset_method; + struct pci_dev *p = NULL; + u64 expires; + + /* + * For now, only BACO and mode1 reset are confirmed + * to suffer the audio issue without proper suspended. + */ + reset_method = amdgpu_asic_reset_method(adev); + if ((reset_method != AMD_RESET_METHOD_BACO) && + (reset_method != AMD_RESET_METHOD_MODE1)) + return -EINVAL; + + p = pci_get_domain_bus_and_slot(pci_domain_nr(adev->pdev->bus), + adev->pdev->bus->number, 1); + if (!p) + return -ENODEV; + + expires = pm_runtime_autosuspend_expiration(&(p->dev)); + if (!expires) + /* + * If we cannot get the audio device autosuspend delay, + * a fixed 4S interval will be used. Considering 3S is + * the audio controller default autosuspend delay setting. + * 4S used here is guaranteed to cover that. + */ + expires = ktime_get_mono_fast_ns() + NSEC_PER_SEC * 4ULL; + + while (!pm_runtime_status_suspended(&(p->dev))) { + if (!pm_runtime_suspend(&(p->dev))) + break; + + if (expires < ktime_get_mono_fast_ns()) { + dev_warn(adev->dev, "failed to suspend display audio\n"); + /* TODO: abort the succeeding gpu reset? */ + return -ETIMEDOUT; + } + } + + pm_runtime_disable(&(p->dev)); + + return 0; +} + /** * amdgpu_device_gpu_recover - reset the asic and recover scheduler * @@ -4140,6 +4213,7 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev, bool use_baco = (amdgpu_asic_reset_method(adev) == AMD_RESET_METHOD_BACO) ? true : false; + bool audio_suspended = false; /* * Flush RAM to disk so that after reboot @@ -4197,6 +4271,19 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev, return 0; } + /* + * Try to put the audio codec into suspend state + * before gpu reset started. + * + * Due to the power domain of the graphics device + * is shared with AZ power domain. Without this, + * we may change the audio hardware from behind + * the audio driver's back. That will trigger + * some audio codec errors. + */ + if (!amdgpu_device_suspend_display_audio(tmp_adev)) + audio_suspended = true; + amdgpu_ras_set_error_query_ready(tmp_adev, false); cancel_delayed_work_sync(&tmp_adev->delayed_init_work); @@ -4309,6 +4396,8 @@ skip_sched_resume: /*unlock kfd: SRIOV would do it separately */ if (!(in_ras_intr && !use_baco) && !amdgpu_sriov_vf(tmp_adev)) amdgpu_amdkfd_post_reset(tmp_adev); + if (audio_suspended) + amdgpu_device_resume_display_audio(tmp_adev); amdgpu_device_unlock_adev(tmp_adev); } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_df.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_df.h index 057f6ea645d7..61a26c15c8dd 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_df.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_df.h @@ -52,9 +52,6 @@ struct amdgpu_df_funcs { uint64_t (*get_fica)(struct amdgpu_device *adev, uint32_t ficaa_val); void (*set_fica)(struct amdgpu_device *adev, uint32_t ficaa_val, uint32_t ficadl_val, uint32_t ficadh_val); - uint64_t (*get_dram_base_addr)(struct amdgpu_device *adev, - uint32_t df_inst); - uint32_t (*get_df_inst_id)(struct amdgpu_device *adev); }; struct amdgpu_df { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c index 43bb22ad8add..b5d6274952a5 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c @@ -133,9 +133,10 @@ static int hw_id_map[MAX_HWIP] = { static int amdgpu_discovery_read_binary(struct amdgpu_device *adev, uint8_t *binary) { uint64_t vram_size = (uint64_t)RREG32(mmRCC_CONFIG_MEMSIZE) << 20; - uint64_t pos = vram_size - DISCOVERY_TMR_SIZE; + uint64_t pos = vram_size - adev->discovery_tmr_size; - amdgpu_device_vram_access(adev, pos, (uint32_t *)binary, DISCOVERY_TMR_SIZE, false); + amdgpu_device_vram_access(adev, pos, (uint32_t *)binary, + adev->discovery_tmr_size, false); return 0; } @@ -167,17 +168,18 @@ static int amdgpu_discovery_init(struct amdgpu_device *adev) uint16_t checksum; int r; - adev->discovery = kzalloc(DISCOVERY_TMR_SIZE, GFP_KERNEL); - if (!adev->discovery) + adev->discovery_tmr_size = DISCOVERY_TMR_SIZE; + adev->discovery_bin = kzalloc(adev->discovery_tmr_size, GFP_KERNEL); + if (!adev->discovery_bin) return -ENOMEM; - r = amdgpu_discovery_read_binary(adev, adev->discovery); + r = amdgpu_discovery_read_binary(adev, adev->discovery_bin); if (r) { DRM_ERROR("failed to read ip discovery binary\n"); goto out; } - bhdr = (struct binary_header *)adev->discovery; + bhdr = (struct binary_header *)adev->discovery_bin; if (le32_to_cpu(bhdr->binary_signature) != BINARY_SIGNATURE) { DRM_ERROR("invalid ip discovery binary signature\n"); @@ -190,7 +192,7 @@ static int amdgpu_discovery_init(struct amdgpu_device *adev) size = bhdr->binary_size - offset; checksum = bhdr->binary_checksum; - if (!amdgpu_discovery_verify_checksum(adev->discovery + offset, + if (!amdgpu_discovery_verify_checksum(adev->discovery_bin + offset, size, checksum)) { DRM_ERROR("invalid ip discovery binary checksum\n"); r = -EINVAL; @@ -200,7 +202,7 @@ static int amdgpu_discovery_init(struct amdgpu_device *adev) info = &bhdr->table_list[IP_DISCOVERY]; offset = le16_to_cpu(info->offset); checksum = le16_to_cpu(info->checksum); - ihdr = (struct ip_discovery_header *)(adev->discovery + offset); + ihdr = (struct ip_discovery_header *)(adev->discovery_bin + offset); if (le32_to_cpu(ihdr->signature) != DISCOVERY_TABLE_SIGNATURE) { DRM_ERROR("invalid ip discovery data table signature\n"); @@ -208,7 +210,7 @@ static int amdgpu_discovery_init(struct amdgpu_device *adev) goto out; } - if (!amdgpu_discovery_verify_checksum(adev->discovery + offset, + if (!amdgpu_discovery_verify_checksum(adev->discovery_bin + offset, ihdr->size, checksum)) { DRM_ERROR("invalid ip discovery data table checksum\n"); r = -EINVAL; @@ -218,9 +220,9 @@ static int amdgpu_discovery_init(struct amdgpu_device *adev) info = &bhdr->table_list[GC]; offset = le16_to_cpu(info->offset); checksum = le16_to_cpu(info->checksum); - ghdr = (struct gpu_info_header *)(adev->discovery + offset); + ghdr = (struct gpu_info_header *)(adev->discovery_bin + offset); - if (!amdgpu_discovery_verify_checksum(adev->discovery + offset, + if (!amdgpu_discovery_verify_checksum(adev->discovery_bin + offset, ghdr->size, checksum)) { DRM_ERROR("invalid gc data table checksum\n"); r = -EINVAL; @@ -230,16 +232,16 @@ static int amdgpu_discovery_init(struct amdgpu_device *adev) return 0; out: - kfree(adev->discovery); - adev->discovery = NULL; + kfree(adev->discovery_bin); + adev->discovery_bin = NULL; return r; } void amdgpu_discovery_fini(struct amdgpu_device *adev) { - kfree(adev->discovery); - adev->discovery = NULL; + kfree(adev->discovery_bin); + adev->discovery_bin = NULL; } int amdgpu_discovery_reg_base_init(struct amdgpu_device *adev) @@ -263,8 +265,8 @@ int amdgpu_discovery_reg_base_init(struct amdgpu_device *adev) return r; } - bhdr = (struct binary_header *)adev->discovery; - ihdr = (struct ip_discovery_header *)(adev->discovery + + bhdr = (struct binary_header *)adev->discovery_bin; + ihdr = (struct ip_discovery_header *)(adev->discovery_bin + le16_to_cpu(bhdr->table_list[IP_DISCOVERY].offset)); num_dies = le16_to_cpu(ihdr->num_dies); @@ -272,7 +274,7 @@ int amdgpu_discovery_reg_base_init(struct amdgpu_device *adev) for (i = 0; i < num_dies; i++) { die_offset = le16_to_cpu(ihdr->die_info[i].die_offset); - dhdr = (struct die_header *)(adev->discovery + die_offset); + dhdr = (struct die_header *)(adev->discovery_bin + die_offset); num_ips = le16_to_cpu(dhdr->num_ips); ip_offset = die_offset + sizeof(*dhdr); @@ -286,7 +288,7 @@ int amdgpu_discovery_reg_base_init(struct amdgpu_device *adev) le16_to_cpu(dhdr->die_id), num_ips); for (j = 0; j < num_ips; j++) { - ip = (struct ip *)(adev->discovery + ip_offset); + ip = (struct ip *)(adev->discovery_bin + ip_offset); num_base_address = ip->num_base_address; DRM_DEBUG("%s(%d) #%d v%d.%d.%d:\n", @@ -335,24 +337,24 @@ int amdgpu_discovery_get_ip_version(struct amdgpu_device *adev, int hw_id, uint16_t num_ips; int i, j; - if (!adev->discovery) { + if (!adev->discovery_bin) { DRM_ERROR("ip discovery uninitialized\n"); return -EINVAL; } - bhdr = (struct binary_header *)adev->discovery; - ihdr = (struct ip_discovery_header *)(adev->discovery + + bhdr = (struct binary_header *)adev->discovery_bin; + ihdr = (struct ip_discovery_header *)(adev->discovery_bin + le16_to_cpu(bhdr->table_list[IP_DISCOVERY].offset)); num_dies = le16_to_cpu(ihdr->num_dies); for (i = 0; i < num_dies; i++) { die_offset = le16_to_cpu(ihdr->die_info[i].die_offset); - dhdr = (struct die_header *)(adev->discovery + die_offset); + dhdr = (struct die_header *)(adev->discovery_bin + die_offset); num_ips = le16_to_cpu(dhdr->num_ips); ip_offset = die_offset + sizeof(*dhdr); for (j = 0; j < num_ips; j++) { - ip = (struct ip *)(adev->discovery + ip_offset); + ip = (struct ip *)(adev->discovery_bin + ip_offset); if (le16_to_cpu(ip->hw_id) == hw_id) { if (major) @@ -375,13 +377,13 @@ int amdgpu_discovery_get_gfx_info(struct amdgpu_device *adev) struct binary_header *bhdr; struct gc_info_v1_0 *gc_info; - if (!adev->discovery) { + if (!adev->discovery_bin) { DRM_ERROR("ip discovery uninitialized\n"); return -EINVAL; } - bhdr = (struct binary_header *)adev->discovery; - gc_info = (struct gc_info_v1_0 *)(adev->discovery + + bhdr = (struct binary_header *)adev->discovery_bin; + gc_info = (struct gc_info_v1_0 *)(adev->discovery_bin + le16_to_cpu(bhdr->table_list[GC].offset)); adev->gfx.config.max_shader_engines = le32_to_cpu(gc_info->gc_num_se); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c index 719a963c31a2..beb35dd12964 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c @@ -1174,14 +1174,6 @@ static int amdgpu_pmops_resume(struct device *dev) { struct drm_device *drm_dev = dev_get_drvdata(dev); - /* GPU comes up enabled by the bios on resume */ - if (amdgpu_device_supports_boco(drm_dev) || - amdgpu_device_supports_baco(drm_dev)) { - pm_runtime_disable(dev); - pm_runtime_set_active(dev); - pm_runtime_enable(dev); - } - return amdgpu_device_resume(drm_dev, true); } @@ -1191,7 +1183,9 @@ static int amdgpu_pmops_freeze(struct device *dev) struct amdgpu_device *adev = drm_dev->dev_private; int r; + adev->in_hibernate = true; r = amdgpu_device_suspend(drm_dev, true); + adev->in_hibernate = false; if (r) return r; return amdgpu_asic_reset(adev); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c index 9ae7b61f696a..25ddb482466a 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c @@ -133,8 +133,7 @@ static int amdgpufb_create_pinned_object(struct amdgpu_fbdev *rfbdev, u32 cpp; u64 flags = AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED | AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS | - AMDGPU_GEM_CREATE_VRAM_CLEARED | - AMDGPU_GEM_CREATE_CPU_GTT_USWC; + AMDGPU_GEM_CREATE_VRAM_CLEARED; info = drm_get_format_info(adev->ddev, mode_cmd); cpp = info->cpp[0]; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c index 0103acc57474..d612033a23ac 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c @@ -48,7 +48,7 @@ int amdgpu_gfx_mec_queue_to_bit(struct amdgpu_device *adev, int mec, return bit; } -void amdgpu_gfx_bit_to_mec_queue(struct amdgpu_device *adev, int bit, +void amdgpu_queue_mask_bit_to_mec_queue(struct amdgpu_device *adev, int bit, int *mec, int *pipe, int *queue) { *queue = bit % adev->gfx.mec.num_queue_per_pipe; @@ -274,7 +274,7 @@ static int amdgpu_gfx_kiq_acquire(struct amdgpu_device *adev, if (test_bit(queue_bit, adev->gfx.mec.queue_bitmap)) continue; - amdgpu_gfx_bit_to_mec_queue(adev, queue_bit, &mec, &pipe, &queue); + amdgpu_queue_mask_bit_to_mec_queue(adev, queue_bit, &mec, &pipe, &queue); /* * 1. Using pipes 2/3 from MEC 2 seems cause problems. @@ -485,6 +485,19 @@ int amdgpu_gfx_disable_kcq(struct amdgpu_device *adev) return amdgpu_ring_test_helper(kiq_ring); } +int amdgpu_queue_mask_bit_to_set_resource_bit(struct amdgpu_device *adev, + int queue_bit) +{ + int mec, pipe, queue; + int set_resource_bit = 0; + + amdgpu_queue_mask_bit_to_mec_queue(adev, queue_bit, &mec, &pipe, &queue); + + set_resource_bit = mec * 4 * 8 + pipe * 8 + queue; + + return set_resource_bit; +} + int amdgpu_gfx_enable_kcq(struct amdgpu_device *adev) { struct amdgpu_kiq *kiq = &adev->gfx.kiq; @@ -507,7 +520,7 @@ int amdgpu_gfx_enable_kcq(struct amdgpu_device *adev) break; } - queue_mask |= (1ull << i); + queue_mask |= (1ull << amdgpu_queue_mask_bit_to_set_resource_bit(adev, i)); } DRM_INFO("kiq ring mec %d pipe %d q %d\n", kiq_ring->me, kiq_ring->pipe, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h index ee698f0246d8..d43c11671a38 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h @@ -364,7 +364,7 @@ void amdgpu_gfx_graphics_queue_acquire(struct amdgpu_device *adev); int amdgpu_gfx_mec_queue_to_bit(struct amdgpu_device *adev, int mec, int pipe, int queue); -void amdgpu_gfx_bit_to_mec_queue(struct amdgpu_device *adev, int bit, +void amdgpu_queue_mask_bit_to_mec_queue(struct amdgpu_device *adev, int bit, int *mec, int *pipe, int *queue); bool amdgpu_gfx_is_mec_queue_enabled(struct amdgpu_device *adev, int mec, int pipe, int queue); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c index ea7e72ecaefa..682a514f1794 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c @@ -189,10 +189,12 @@ int amdgpu_driver_load_kms(struct drm_device *dev, unsigned long flags) dev_dbg(&dev->pdev->dev, "Error during ACPI methods call\n"); if (adev->runpm) { - dev_pm_set_driver_flags(dev->dev, DPM_FLAG_NEVER_SKIP); + /* only need to skip on ATPX */ + if (amdgpu_device_supports_boco(dev) && + !amdgpu_is_atpx_hybrid()) + dev_pm_set_driver_flags(dev->dev, DPM_FLAG_NEVER_SKIP); pm_runtime_use_autosuspend(dev->dev); pm_runtime_set_autosuspend_delay(dev->dev, 5000); - pm_runtime_set_active(dev->dev); pm_runtime_allow(dev->dev); pm_runtime_mark_last_busy(dev->dev); pm_runtime_put_autosuspend(dev->dev); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c index 730f98aab11b..ddb4af0cc702 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c @@ -664,6 +664,121 @@ int psp_xgmi_initialize(struct psp_context *psp) return ret; } +int psp_xgmi_get_hive_id(struct psp_context *psp, uint64_t *hive_id) +{ + struct ta_xgmi_shared_memory *xgmi_cmd; + int ret; + + xgmi_cmd = (struct ta_xgmi_shared_memory*)psp->xgmi_context.xgmi_shared_buf; + memset(xgmi_cmd, 0, sizeof(struct ta_xgmi_shared_memory)); + + xgmi_cmd->cmd_id = TA_COMMAND_XGMI__GET_HIVE_ID; + + /* Invoke xgmi ta to get hive id */ + ret = psp_xgmi_invoke(psp, xgmi_cmd->cmd_id); + if (ret) + return ret; + + *hive_id = xgmi_cmd->xgmi_out_message.get_hive_id.hive_id; + + return 0; +} + +int psp_xgmi_get_node_id(struct psp_context *psp, uint64_t *node_id) +{ + struct ta_xgmi_shared_memory *xgmi_cmd; + int ret; + + xgmi_cmd = (struct ta_xgmi_shared_memory*)psp->xgmi_context.xgmi_shared_buf; + memset(xgmi_cmd, 0, sizeof(struct ta_xgmi_shared_memory)); + + xgmi_cmd->cmd_id = TA_COMMAND_XGMI__GET_NODE_ID; + + /* Invoke xgmi ta to get the node id */ + ret = psp_xgmi_invoke(psp, xgmi_cmd->cmd_id); + if (ret) + return ret; + + *node_id = xgmi_cmd->xgmi_out_message.get_node_id.node_id; + + return 0; +} + +int psp_xgmi_get_topology_info(struct psp_context *psp, + int number_devices, + struct psp_xgmi_topology_info *topology) +{ + struct ta_xgmi_shared_memory *xgmi_cmd; + struct ta_xgmi_cmd_get_topology_info_input *topology_info_input; + struct ta_xgmi_cmd_get_topology_info_output *topology_info_output; + int i; + int ret; + + if (!topology || topology->num_nodes > TA_XGMI__MAX_CONNECTED_NODES) + return -EINVAL; + + xgmi_cmd = (struct ta_xgmi_shared_memory*)psp->xgmi_context.xgmi_shared_buf; + memset(xgmi_cmd, 0, sizeof(struct ta_xgmi_shared_memory)); + + /* Fill in the shared memory with topology information as input */ + topology_info_input = &xgmi_cmd->xgmi_in_message.get_topology_info; + xgmi_cmd->cmd_id = TA_COMMAND_XGMI__GET_GET_TOPOLOGY_INFO; + topology_info_input->num_nodes = number_devices; + + for (i = 0; i < topology_info_input->num_nodes; i++) { + topology_info_input->nodes[i].node_id = topology->nodes[i].node_id; + topology_info_input->nodes[i].num_hops = topology->nodes[i].num_hops; + topology_info_input->nodes[i].is_sharing_enabled = topology->nodes[i].is_sharing_enabled; + topology_info_input->nodes[i].sdma_engine = topology->nodes[i].sdma_engine; + } + + /* Invoke xgmi ta to get the topology information */ + ret = psp_xgmi_invoke(psp, TA_COMMAND_XGMI__GET_GET_TOPOLOGY_INFO); + if (ret) + return ret; + + /* Read the output topology information from the shared memory */ + topology_info_output = &xgmi_cmd->xgmi_out_message.get_topology_info; + topology->num_nodes = xgmi_cmd->xgmi_out_message.get_topology_info.num_nodes; + for (i = 0; i < topology->num_nodes; i++) { + topology->nodes[i].node_id = topology_info_output->nodes[i].node_id; + topology->nodes[i].num_hops = topology_info_output->nodes[i].num_hops; + topology->nodes[i].is_sharing_enabled = topology_info_output->nodes[i].is_sharing_enabled; + topology->nodes[i].sdma_engine = topology_info_output->nodes[i].sdma_engine; + } + + return 0; +} + +int psp_xgmi_set_topology_info(struct psp_context *psp, + int number_devices, + struct psp_xgmi_topology_info *topology) +{ + struct ta_xgmi_shared_memory *xgmi_cmd; + struct ta_xgmi_cmd_get_topology_info_input *topology_info_input; + int i; + + if (!topology || topology->num_nodes > TA_XGMI__MAX_CONNECTED_NODES) + return -EINVAL; + + xgmi_cmd = (struct ta_xgmi_shared_memory*)psp->xgmi_context.xgmi_shared_buf; + memset(xgmi_cmd, 0, sizeof(struct ta_xgmi_shared_memory)); + + topology_info_input = &xgmi_cmd->xgmi_in_message.get_topology_info; + xgmi_cmd->cmd_id = TA_COMMAND_XGMI__SET_TOPOLOGY_INFO; + topology_info_input->num_nodes = number_devices; + + for (i = 0; i < topology_info_input->num_nodes; i++) { + topology_info_input->nodes[i].node_id = topology->nodes[i].node_id; + topology_info_input->nodes[i].num_hops = topology->nodes[i].num_hops; + topology_info_input->nodes[i].is_sharing_enabled = 1; + topology_info_input->nodes[i].sdma_engine = topology->nodes[i].sdma_engine; + } + + /* Invoke xgmi ta to set topology information */ + return psp_xgmi_invoke(psp, TA_COMMAND_XGMI__SET_TOPOLOGY_INFO); +} + // ras begin static int psp_ras_init_shared_buf(struct psp_context *psp) { @@ -746,13 +861,40 @@ static int psp_ras_unload(struct psp_context *psp) int psp_ras_invoke(struct psp_context *psp, uint32_t ta_cmd_id) { + struct ta_ras_shared_memory *ras_cmd; + int ret; + + ras_cmd = (struct ta_ras_shared_memory *)psp->ras.ras_shared_buf; + /* * TODO: bypass the loading in sriov for now */ if (amdgpu_sriov_vf(psp->adev)) return 0; - return psp_ta_invoke(psp, ta_cmd_id, psp->ras.session_id); + ret = psp_ta_invoke(psp, ta_cmd_id, psp->ras.session_id); + + if (amdgpu_ras_intr_triggered()) + return ret; + + if (ras_cmd->if_version > RAS_TA_HOST_IF_VER) + { + DRM_WARN("RAS: Unsupported Interface"); + return -EINVAL; + } + + if (!ret) { + if (ras_cmd->ras_out_message.flags.err_inject_switch_disable_flag) { + dev_warn(psp->adev->dev, "ECC switch disabled\n"); + + ras_cmd->ras_status = TA_RAS_STATUS__ERROR_RAS_NOT_AVAILABLE; + } + else if (ras_cmd->ras_out_message.flags.reg_access_failure_flag) + dev_warn(psp->adev->dev, + "RAS internal register access blocked\n"); + } + + return ret; } int psp_ras_enable_features(struct psp_context *psp, @@ -836,6 +978,33 @@ static int psp_ras_initialize(struct psp_context *psp) return 0; } + +int psp_ras_trigger_error(struct psp_context *psp, + struct ta_ras_trigger_error_input *info) +{ + struct ta_ras_shared_memory *ras_cmd; + int ret; + + if (!psp->ras.ras_initialized) + return -EINVAL; + + ras_cmd = (struct ta_ras_shared_memory *)psp->ras.ras_shared_buf; + memset(ras_cmd, 0, sizeof(struct ta_ras_shared_memory)); + + ras_cmd->cmd_id = TA_RAS_COMMAND__TRIGGER_ERROR; + ras_cmd->ras_in_message.trigger_error = *info; + + ret = psp_ras_invoke(psp, ras_cmd->cmd_id); + if (ret) + return -EINVAL; + + /* If err_event_athub occurs error inject was successful, however + return status from TA is no long reliable */ + if (amdgpu_ras_intr_triggered()) + return 0; + + return ras_cmd->ras_status; +} // ras end // HDCP start @@ -1477,7 +1646,7 @@ static int psp_np_fw_load(struct psp_context *psp) /* Start rlc autoload after psp recieved all the gfx firmware */ if (psp->autoload_supported && ucode->ucode_id == (amdgpu_sriov_vf(adev) ? AMDGPU_UCODE_ID_CP_MEC2 : AMDGPU_UCODE_ID_RLC_G)) { - ret = psp_rlc_autoload(psp); + ret = psp_rlc_autoload_start(psp); if (ret) { DRM_ERROR("Failed to start rlc autoload\n"); return ret; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h index 7fcd63d5432c..2a56ad996d83 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h @@ -95,16 +95,6 @@ struct psp_funcs enum psp_ring_type ring_type); bool (*smu_reload_quirk)(struct psp_context *psp); int (*mode1_reset)(struct psp_context *psp); - int (*xgmi_get_node_id)(struct psp_context *psp, uint64_t *node_id); - int (*xgmi_get_hive_id)(struct psp_context *psp, uint64_t *hive_id); - int (*xgmi_get_topology_info)(struct psp_context *psp, int number_devices, - struct psp_xgmi_topology_info *topology); - int (*xgmi_set_topology_info)(struct psp_context *psp, int number_devices, - struct psp_xgmi_topology_info *topology); - int (*ras_trigger_error)(struct psp_context *psp, - struct ta_ras_trigger_error_input *info); - int (*ras_cure_posion)(struct psp_context *psp, uint64_t *mode_ptr); - int (*rlc_autoload_start)(struct psp_context *psp); int (*mem_training_init)(struct psp_context *psp); void (*mem_training_fini)(struct psp_context *psp); int (*mem_training)(struct psp_context *psp, uint32_t ops); @@ -316,18 +306,6 @@ struct amdgpu_psp_funcs { ((psp)->funcs->smu_reload_quirk ? (psp)->funcs->smu_reload_quirk((psp)) : false) #define psp_mode1_reset(psp) \ ((psp)->funcs->mode1_reset ? (psp)->funcs->mode1_reset((psp)) : false) -#define psp_xgmi_get_node_id(psp, node_id) \ - ((psp)->funcs->xgmi_get_node_id ? (psp)->funcs->xgmi_get_node_id((psp), (node_id)) : -EINVAL) -#define psp_xgmi_get_hive_id(psp, hive_id) \ - ((psp)->funcs->xgmi_get_hive_id ? (psp)->funcs->xgmi_get_hive_id((psp), (hive_id)) : -EINVAL) -#define psp_xgmi_get_topology_info(psp, num_device, topology) \ - ((psp)->funcs->xgmi_get_topology_info ? \ - (psp)->funcs->xgmi_get_topology_info((psp), (num_device), (topology)) : -EINVAL) -#define psp_xgmi_set_topology_info(psp, num_device, topology) \ - ((psp)->funcs->xgmi_set_topology_info ? \ - (psp)->funcs->xgmi_set_topology_info((psp), (num_device), (topology)) : -EINVAL) -#define psp_rlc_autoload(psp) \ - ((psp)->funcs->rlc_autoload_start ? (psp)->funcs->rlc_autoload_start((psp)) : 0) #define psp_mem_training_init(psp) \ ((psp)->funcs->mem_training_init ? (psp)->funcs->mem_training_init((psp)) : 0) #define psp_mem_training_fini(psp) \ @@ -335,13 +313,6 @@ struct amdgpu_psp_funcs { #define psp_mem_training(psp, ops) \ ((psp)->funcs->mem_training ? (psp)->funcs->mem_training((psp), (ops)) : 0) -#define psp_ras_trigger_error(psp, info) \ - ((psp)->funcs->ras_trigger_error ? \ - (psp)->funcs->ras_trigger_error((psp), (info)) : -EINVAL) -#define psp_ras_cure_posion(psp, addr) \ - ((psp)->funcs->ras_cure_posion ? \ - (psp)->funcs->ras_cure_posion(psp, (addr)) : -EINVAL) - #define psp_ring_get_wptr(psp) (psp)->funcs->ring_get_wptr((psp)) #define psp_ring_set_wptr(psp, value) (psp)->funcs->ring_set_wptr((psp), (value)) @@ -369,10 +340,21 @@ int psp_update_vcn_sram(struct amdgpu_device *adev, int inst_idx, int psp_xgmi_initialize(struct psp_context *psp); int psp_xgmi_terminate(struct psp_context *psp); int psp_xgmi_invoke(struct psp_context *psp, uint32_t ta_cmd_id); +int psp_xgmi_get_hive_id(struct psp_context *psp, uint64_t *hive_id); +int psp_xgmi_get_node_id(struct psp_context *psp, uint64_t *node_id); +int psp_xgmi_get_topology_info(struct psp_context *psp, + int number_devices, + struct psp_xgmi_topology_info *topology); +int psp_xgmi_set_topology_info(struct psp_context *psp, + int number_devices, + struct psp_xgmi_topology_info *topology); int psp_ras_invoke(struct psp_context *psp, uint32_t ta_cmd_id); int psp_ras_enable_features(struct psp_context *psp, union ta_ras_cmd_input *info, bool enable); +int psp_ras_trigger_error(struct psp_context *psp, + struct ta_ras_trigger_error_input *info); + int psp_hdcp_invoke(struct psp_context *psp, uint32_t ta_cmd_id); int psp_dtm_invoke(struct psp_context *psp, uint32_t ta_cmd_id); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c index 8b14aee370c8..7348619253c7 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c @@ -502,6 +502,29 @@ struct ras_manager *amdgpu_ras_find_obj(struct amdgpu_device *adev, } /* obj end */ +void amdgpu_ras_parse_status_code(struct amdgpu_device* adev, + const char* invoke_type, + const char* block_name, + enum ta_ras_status ret) +{ + switch (ret) { + case TA_RAS_STATUS__SUCCESS: + return; + case TA_RAS_STATUS__ERROR_RAS_NOT_AVAILABLE: + dev_warn(adev->dev, + "RAS WARN: %s %s currently unavailable\n", + invoke_type, + block_name); + break; + default: + dev_err(adev->dev, + "RAS ERROR: %s %s error failed ret 0x%X\n", + invoke_type, + block_name, + ret); + } +} + /* feature ctl begin */ static int amdgpu_ras_is_feature_allowed(struct amdgpu_device *adev, struct ras_common_if *head) @@ -565,19 +588,23 @@ int amdgpu_ras_feature_enable(struct amdgpu_device *adev, struct ras_common_if *head, bool enable) { struct amdgpu_ras *con = amdgpu_ras_get_context(adev); - union ta_ras_cmd_input info; + union ta_ras_cmd_input *info; int ret; if (!con) return -EINVAL; + info = kzalloc(sizeof(union ta_ras_cmd_input), GFP_KERNEL); + if (!info) + return -ENOMEM; + if (!enable) { - info.disable_features = (struct ta_ras_disable_features_input) { + info->disable_features = (struct ta_ras_disable_features_input) { .block_id = amdgpu_ras_block_to_ta(head->block), .error_type = amdgpu_ras_error_to_ta(head->type), }; } else { - info.enable_features = (struct ta_ras_enable_features_input) { + info->enable_features = (struct ta_ras_enable_features_input) { .block_id = amdgpu_ras_block_to_ta(head->block), .error_type = amdgpu_ras_error_to_ta(head->type), }; @@ -586,27 +613,33 @@ int amdgpu_ras_feature_enable(struct amdgpu_device *adev, /* Do not enable if it is not allowed. */ WARN_ON(enable && !amdgpu_ras_is_feature_allowed(adev, head)); /* Are we alerady in that state we are going to set? */ - if (!(!!enable ^ !!amdgpu_ras_is_feature_enabled(adev, head))) - return 0; + if (!(!!enable ^ !!amdgpu_ras_is_feature_enabled(adev, head))) { + ret = 0; + goto out; + } if (!amdgpu_ras_intr_triggered()) { - ret = psp_ras_enable_features(&adev->psp, &info, enable); + ret = psp_ras_enable_features(&adev->psp, info, enable); if (ret) { - dev_err(adev->dev, "RAS ERROR: %s %s feature " - "failed ret %d\n", - enable ? "enable":"disable", - ras_block_str(head->block), - ret); + amdgpu_ras_parse_status_code(adev, + enable ? "enable":"disable", + ras_block_str(head->block), + (enum ta_ras_status)ret); if (ret == TA_RAS_STATUS__RESET_NEEDED) - return -EAGAIN; - return -EINVAL; + ret = -EAGAIN; + else + ret = -EINVAL; + + goto out; } } /* setup the obj */ __amdgpu_ras_feature_enable(adev, head, enable); - - return 0; + ret = 0; +out: + kfree(info); + return ret; } /* Only used in device probe stage and called only once. */ @@ -821,10 +854,10 @@ int amdgpu_ras_error_inject(struct amdgpu_device *adev, ret = -EINVAL; } - if (ret) - dev_err(adev->dev, "RAS ERROR: inject %s error failed ret %d\n", - ras_block_str(info->head.block), - ret); + amdgpu_ras_parse_status_code(adev, + "inject", + ras_block_str(info->head.block), + (enum ta_ras_status)ret); return ret; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c index d5543c25f3c7..eff1f73302de 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c @@ -1957,17 +1957,19 @@ int amdgpu_ttm_init(struct amdgpu_device *adev) return r; /* - * reserve one TMR (64K) memory at the top of VRAM which holds + * reserve TMR memory at the top of VRAM which holds * IP Discovery data and is protected by PSP. */ - r = amdgpu_bo_create_kernel_at(adev, - adev->gmc.real_vram_size - DISCOVERY_TMR_SIZE, - DISCOVERY_TMR_SIZE, - AMDGPU_GEM_DOMAIN_VRAM, - &adev->discovery_memory, - NULL); - if (r) - return r; + if (adev->discovery_tmr_size > 0) { + r = amdgpu_bo_create_kernel_at(adev, + adev->gmc.real_vram_size - adev->discovery_tmr_size, + adev->discovery_tmr_size, + AMDGPU_GEM_DOMAIN_VRAM, + &adev->discovery_memory, + NULL); + if (r) + return r; + } DRM_INFO("amdgpu: %uM of VRAM memory ready\n", (unsigned) (adev->gmc.real_vram_size / (1024 * 1024))); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c index 128a667ed8fa..d399e5893170 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c @@ -149,6 +149,15 @@ static DEVICE_ATTR(mem_info_vis_vram_used, S_IRUGO, static DEVICE_ATTR(mem_info_vram_vendor, S_IRUGO, amdgpu_mem_info_vram_vendor, NULL); +static const struct attribute *amdgpu_vram_mgr_attributes[] = { + &dev_attr_mem_info_vram_total.attr, + &dev_attr_mem_info_vis_vram_total.attr, + &dev_attr_mem_info_vram_used.attr, + &dev_attr_mem_info_vis_vram_used.attr, + &dev_attr_mem_info_vram_vendor.attr, + NULL +}; + /** * amdgpu_vram_mgr_init - init VRAM manager and DRM MM * @@ -173,31 +182,9 @@ static int amdgpu_vram_mgr_init(struct ttm_mem_type_manager *man, man->priv = mgr; /* Add the two VRAM-related sysfs files */ - ret = device_create_file(adev->dev, &dev_attr_mem_info_vram_total); - if (ret) { - DRM_ERROR("Failed to create device file mem_info_vram_total\n"); - return ret; - } - ret = device_create_file(adev->dev, &dev_attr_mem_info_vis_vram_total); - if (ret) { - DRM_ERROR("Failed to create device file mem_info_vis_vram_total\n"); - return ret; - } - ret = device_create_file(adev->dev, &dev_attr_mem_info_vram_used); - if (ret) { - DRM_ERROR("Failed to create device file mem_info_vram_used\n"); - return ret; - } - ret = device_create_file(adev->dev, &dev_attr_mem_info_vis_vram_used); - if (ret) { - DRM_ERROR("Failed to create device file mem_info_vis_vram_used\n"); - return ret; - } - ret = device_create_file(adev->dev, &dev_attr_mem_info_vram_vendor); - if (ret) { - DRM_ERROR("Failed to create device file mem_info_vram_vendor\n"); - return ret; - } + ret = sysfs_create_files(&adev->dev->kobj, amdgpu_vram_mgr_attributes); + if (ret) + DRM_ERROR("Failed to register sysfs\n"); return 0; } @@ -220,11 +207,7 @@ static int amdgpu_vram_mgr_fini(struct ttm_mem_type_manager *man) spin_unlock(&mgr->lock); kfree(mgr); man->priv = NULL; - device_remove_file(adev->dev, &dev_attr_mem_info_vram_total); - device_remove_file(adev->dev, &dev_attr_mem_info_vis_vram_total); - device_remove_file(adev->dev, &dev_attr_mem_info_vram_used); - device_remove_file(adev->dev, &dev_attr_mem_info_vis_vram_used); - device_remove_file(adev->dev, &dev_attr_mem_info_vram_vendor); + sysfs_remove_files(&adev->dev->kobj, amdgpu_vram_mgr_attributes); return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c index 48c0ce13f68e..90610b4f2c75 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c @@ -649,31 +649,8 @@ void amdgpu_xgmi_ras_fini(struct amdgpu_device *adev) uint64_t amdgpu_xgmi_get_relative_phy_addr(struct amdgpu_device *adev, uint64_t addr) { - uint32_t df_inst_id; - uint64_t dram_base_addr = 0; - const struct amdgpu_df_funcs *df_funcs = adev->df.funcs; - - if ((!df_funcs) || - (!df_funcs->get_df_inst_id) || - (!df_funcs->get_dram_base_addr)) { - dev_warn(adev->dev, - "XGMI: relative phy_addr algorithm is not supported\n"); - return addr; - } - - if (amdgpu_dpm_set_df_cstate(adev, DF_CSTATE_DISALLOW)) { - dev_warn(adev->dev, - "failed to disable DF-Cstate, DF register may not be accessible\n"); - return addr; - } - - df_inst_id = df_funcs->get_df_inst_id(adev); - dram_base_addr = df_funcs->get_dram_base_addr(adev, df_inst_id); - - if (amdgpu_dpm_set_df_cstate(adev, DF_CSTATE_ALLOW)) - dev_warn(adev->dev, "failed to enable DF-Cstate\n"); - - return addr + dram_base_addr; + struct amdgpu_xgmi *xgmi = &adev->gmc.xgmi; + return (addr + xgmi->physical_node_id * xgmi->node_segment_size); } static void pcs_clear_status(struct amdgpu_device *adev, uint32_t pcs_status_reg) diff --git a/drivers/gpu/drm/amd/amdgpu/df_v3_6.c b/drivers/gpu/drm/amd/amdgpu/df_v3_6.c index 5a1bd8ed1a6c..a7b8292cefee 100644 --- a/drivers/gpu/drm/amd/amdgpu/df_v3_6.c +++ b/drivers/gpu/drm/amd/amdgpu/df_v3_6.c @@ -686,58 +686,6 @@ static void df_v3_6_pmc_get_count(struct amdgpu_device *adev, } } -static uint64_t df_v3_6_get_dram_base_addr(struct amdgpu_device *adev, - uint32_t df_inst) -{ - uint32_t base_addr_reg_val = 0; - uint64_t base_addr = 0; - - base_addr_reg_val = RREG32_PCIE(smnDF_CS_UMC_AON0_DramBaseAddress0 + - df_inst * DF_3_6_SMN_REG_INST_DIST); - - if (REG_GET_FIELD(base_addr_reg_val, - DF_CS_UMC_AON0_DramBaseAddress0, - AddrRngVal) == 0) { - DRM_WARN("address range not valid"); - return 0; - } - - base_addr = REG_GET_FIELD(base_addr_reg_val, - DF_CS_UMC_AON0_DramBaseAddress0, - DramBaseAddr); - - return base_addr << 28; -} - -static uint32_t df_v3_6_get_df_inst_id(struct amdgpu_device *adev) -{ - uint32_t xgmi_node_id = 0; - uint32_t df_inst_id = 0; - - /* Walk through DF dst nodes to find current XGMI node */ - for (df_inst_id = 0; df_inst_id < DF_3_6_INST_CNT; df_inst_id++) { - - xgmi_node_id = RREG32_PCIE(smnDF_CS_UMC_AON0_DramLimitAddress0 + - df_inst_id * DF_3_6_SMN_REG_INST_DIST); - xgmi_node_id = REG_GET_FIELD(xgmi_node_id, - DF_CS_UMC_AON0_DramLimitAddress0, - DstFabricID); - - /* TODO: establish reason dest fabric id is offset by 7 */ - xgmi_node_id = xgmi_node_id >> 7; - - if (adev->gmc.xgmi.physical_node_id == xgmi_node_id) - break; - } - - if (df_inst_id == DF_3_6_INST_CNT) { - DRM_WARN("cant match df dst id with gpu node"); - return 0; - } - - return df_inst_id; -} - const struct amdgpu_df_funcs df_v3_6_funcs = { .sw_init = df_v3_6_sw_init, .sw_fini = df_v3_6_sw_fini, @@ -752,6 +700,4 @@ const struct amdgpu_df_funcs df_v3_6_funcs = { .pmc_get_count = df_v3_6_pmc_get_count, .get_fica = df_v3_6_get_fica, .set_fica = df_v3_6_set_fica, - .get_dram_base_addr = df_v3_6_get_dram_base_addr, - .get_df_inst_id = df_v3_6_get_df_inst_id }; diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c index 9fe20b782e88..64080d209223 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c @@ -4577,11 +4577,13 @@ static int gfx_v10_0_init_csb(struct amdgpu_device *adev) adev->gfx.rlc.funcs->get_csb_buffer(adev, adev->gfx.rlc.cs_ptr); /* csib */ - WREG32_SOC15_RLC(GC, 0, mmRLC_CSIB_ADDR_HI, - adev->gfx.rlc.clear_state_gpu_addr >> 32); - WREG32_SOC15_RLC(GC, 0, mmRLC_CSIB_ADDR_LO, - adev->gfx.rlc.clear_state_gpu_addr & 0xfffffffc); - WREG32_SOC15_RLC(GC, 0, mmRLC_CSIB_LENGTH, adev->gfx.rlc.clear_state_size); + /* amdgpu_mm_wreg_mmio_rlc will fall back to mmio if doesn't support rlcg_write */ + amdgpu_mm_wreg_mmio_rlc(adev, SOC15_REG_OFFSET(GC, 0, mmRLC_CSIB_ADDR_HI), + adev->gfx.rlc.clear_state_gpu_addr >> 32, 0); + amdgpu_mm_wreg_mmio_rlc(adev, SOC15_REG_OFFSET(GC, 0, mmRLC_CSIB_ADDR_LO), + adev->gfx.rlc.clear_state_gpu_addr & 0xfffffffc, 0); + amdgpu_mm_wreg_mmio_rlc(adev, SOC15_REG_OFFSET(GC, 0, mmRLC_CSIB_LENGTH), + adev->gfx.rlc.clear_state_size, 0); return 0; } @@ -5190,7 +5192,7 @@ static int gfx_v10_0_cp_gfx_enable(struct amdgpu_device *adev, bool enable) tmp = REG_SET_FIELD(tmp, CP_ME_CNTL, ME_HALT, enable ? 0 : 1); tmp = REG_SET_FIELD(tmp, CP_ME_CNTL, PFP_HALT, enable ? 0 : 1); tmp = REG_SET_FIELD(tmp, CP_ME_CNTL, CE_HALT, enable ? 0 : 1); - WREG32_SOC15_RLC(GC, 0, mmCP_ME_CNTL, tmp); + amdgpu_mm_wreg_mmio_rlc(adev, SOC15_REG_OFFSET(GC, 0, mmCP_ME_CNTL), tmp, 0); for (i = 0; i < adev->usec_timeout; i++) { if (RREG32_SOC15(GC, 0, mmCP_STAT) == 0) @@ -6540,14 +6542,16 @@ static int gfx_v10_0_hw_init(void *handle) * loaded firstly, so in direct type, it has to load smc ucode * here before rlc. */ - r = smu_load_microcode(&adev->smu); - if (r) - return r; + if (adev->smu.ppt_funcs != NULL) { + r = smu_load_microcode(&adev->smu); + if (r) + return r; - r = smu_check_fw_status(&adev->smu); - if (r) { - pr_err("SMC firmware status is not correct\n"); - return r; + r = smu_check_fw_status(&adev->smu); + if (r) { + pr_err("SMC firmware status is not correct\n"); + return r; + } } } @@ -7011,7 +7015,7 @@ static int gfx_v10_0_update_gfx_clock_gating(struct amdgpu_device *adev, /* === CGCG /CGLS for GFX 3D Only === */ gfx_v10_0_update_3d_clock_gating(adev, enable); /* === MGCG + MGLS === */ - /* gfx_v10_0_update_medium_grain_clock_gating(adev, enable); */ + gfx_v10_0_update_medium_grain_clock_gating(adev, enable); } if (adev->cg_flags & @@ -7086,6 +7090,20 @@ static const struct amdgpu_rlc_funcs gfx_v10_0_rlc_funcs = { .reset = gfx_v10_0_rlc_reset, .start = gfx_v10_0_rlc_start, .update_spm_vmid = gfx_v10_0_update_spm_vmid, +}; + +static const struct amdgpu_rlc_funcs gfx_v10_0_rlc_funcs_sriov = { + .is_rlc_enabled = gfx_v10_0_is_rlc_enabled, + .set_safe_mode = gfx_v10_0_set_safe_mode, + .unset_safe_mode = gfx_v10_0_unset_safe_mode, + .init = gfx_v10_0_rlc_init, + .get_csb_size = gfx_v10_0_get_csb_size, + .get_csb_buffer = gfx_v10_0_get_csb_buffer, + .resume = gfx_v10_0_rlc_resume, + .stop = gfx_v10_0_rlc_stop, + .reset = gfx_v10_0_rlc_reset, + .start = gfx_v10_0_rlc_start, + .update_spm_vmid = gfx_v10_0_update_spm_vmid, .rlcg_wreg = gfx_v10_rlcg_wreg, .is_rlcg_access_range = gfx_v10_0_is_rlcg_access_range, }; @@ -7102,11 +7120,7 @@ static int gfx_v10_0_set_powergating_state(void *handle, switch (adev->asic_type) { case CHIP_NAVI10: case CHIP_NAVI14: - if (!enable) { - amdgpu_gfx_off_ctrl(adev, false); - cancel_delayed_work_sync(&adev->gfx.gfx_off_delay_work); - } else - amdgpu_gfx_off_ctrl(adev, true); + amdgpu_gfx_off_ctrl(adev, enable); break; default: break; @@ -7672,6 +7686,19 @@ static void gfx_v10_0_ring_emit_reg_write_reg_wait(struct amdgpu_ring *ring, ref, mask); } +static void gfx_v10_0_ring_soft_recovery(struct amdgpu_ring *ring, + unsigned vmid) +{ + struct amdgpu_device *adev = ring->adev; + uint32_t value = 0; + + value = REG_SET_FIELD(value, SQ_CMD, CMD, 0x03); + value = REG_SET_FIELD(value, SQ_CMD, MODE, 0x01); + value = REG_SET_FIELD(value, SQ_CMD, CHECK_VMID, 1); + value = REG_SET_FIELD(value, SQ_CMD, VM_ID, vmid); + WREG32_SOC15(GC, 0, mmSQ_CMD, value); +} + static void gfx_v10_0_set_gfx_eop_interrupt_state(struct amdgpu_device *adev, uint32_t me, uint32_t pipe, @@ -8063,6 +8090,7 @@ static const struct amdgpu_ring_funcs gfx_v10_0_ring_funcs_gfx = { .emit_wreg = gfx_v10_0_ring_emit_wreg, .emit_reg_wait = gfx_v10_0_ring_emit_reg_wait, .emit_reg_write_reg_wait = gfx_v10_0_ring_emit_reg_write_reg_wait, + .soft_recovery = gfx_v10_0_ring_soft_recovery, }; static const struct amdgpu_ring_funcs gfx_v10_0_ring_funcs_compute = { @@ -8183,9 +8211,11 @@ static void gfx_v10_0_set_rlc_funcs(struct amdgpu_device *adev) switch (adev->asic_type) { case CHIP_NAVI10: case CHIP_NAVI14: - case CHIP_NAVI12: adev->gfx.rlc.funcs = &gfx_v10_0_rlc_funcs; break; + case CHIP_NAVI12: + adev->gfx.rlc.funcs = &gfx_v10_0_rlc_funcs_sriov; + break; default: break; } diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c index eedb92218ba5..a0988634aeaf 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c @@ -507,8 +507,8 @@ static const struct soc15_reg_golden golden_settings_gc_9_0[] = SOC15_REG_GOLDEN_VALUE(GC, 0, mmSH_MEM_CONFIG, 0x00001000, 0x00001000), SOC15_REG_GOLDEN_VALUE(GC, 0, mmSPI_RESOURCE_RESERVE_CU_0, 0x0007ffff, 0x00000800), SOC15_REG_GOLDEN_VALUE(GC, 0, mmSPI_RESOURCE_RESERVE_CU_1, 0x0007ffff, 0x00000800), - SOC15_REG_GOLDEN_VALUE(GC, 0, mmSPI_RESOURCE_RESERVE_EN_CU_0, 0x01ffffff, 0x0000ff87), - SOC15_REG_GOLDEN_VALUE(GC, 0, mmSPI_RESOURCE_RESERVE_EN_CU_1, 0x01ffffff, 0x0000ff8f), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSPI_RESOURCE_RESERVE_EN_CU_0, 0x01ffffff, 0x00ffff87), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSPI_RESOURCE_RESERVE_EN_CU_1, 0x01ffffff, 0x00ffff8f), SOC15_REG_GOLDEN_VALUE(GC, 0, mmSQC_CONFIG, 0x03000000, 0x020a2000), SOC15_REG_GOLDEN_VALUE(GC, 0, mmTA_CNTL_AUX, 0xfffffeef, 0x010b0000), SOC15_REG_GOLDEN_VALUE(GC, 0, mmTCP_CHAN_STEER_HI, 0xffffffff, 0x4a2c0e68), @@ -1233,6 +1233,8 @@ static const struct amdgpu_gfxoff_quirk amdgpu_gfxoff_quirk_list[] = { { 0x1002, 0x15dd, 0x1002, 0x15dd, 0xc8 }, /* https://bugzilla.kernel.org/show_bug.cgi?id=207171 */ { 0x1002, 0x15dd, 0x103c, 0x83e7, 0xd3 }, + /* GFXOFF is unstable on C6 parts with a VBIOS 113-RAVEN-114 */ + { 0x1002, 0x15dd, 0x1002, 0x15dd, 0xc6 }, { 0, 0, 0, 0, 0 }, }; @@ -5039,10 +5041,9 @@ static int gfx_v9_0_set_powergating_state(void *handle, switch (adev->asic_type) { case CHIP_RAVEN: case CHIP_RENOIR: - if (!enable) { + if (!enable) amdgpu_gfx_off_ctrl(adev, false); - cancel_delayed_work_sync(&adev->gfx.gfx_off_delay_work); - } + if (adev->pg_flags & AMD_PG_SUPPORT_RLC_SMU_HS) { gfx_v9_0_enable_sck_slow_down_on_power_up(adev, true); gfx_v9_0_enable_sck_slow_down_on_power_down(adev, true); @@ -5066,12 +5067,7 @@ static int gfx_v9_0_set_powergating_state(void *handle, amdgpu_gfx_off_ctrl(adev, true); break; case CHIP_VEGA12: - if (!enable) { - amdgpu_gfx_off_ctrl(adev, false); - cancel_delayed_work_sync(&adev->gfx.gfx_off_delay_work); - } else { - amdgpu_gfx_off_ctrl(adev, true); - } + amdgpu_gfx_off_ctrl(adev, enable); 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 c1a530dbe162..a75e472b4a81 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c @@ -61,17 +61,6 @@ MODULE_FIRMWARE("amdgpu/si58_mc.bin"); #define MC_SEQ_MISC0__MT__HBM 0x60000000 #define MC_SEQ_MISC0__MT__DDR3 0xB0000000 - -static const u32 crtc_offsets[6] = -{ - SI_CRTC0_REGISTER_OFFSET, - SI_CRTC1_REGISTER_OFFSET, - SI_CRTC2_REGISTER_OFFSET, - SI_CRTC3_REGISTER_OFFSET, - SI_CRTC4_REGISTER_OFFSET, - SI_CRTC5_REGISTER_OFFSET -}; - static void gmc_v6_0_mc_stop(struct amdgpu_device *adev) { u32 blackout; diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c index e8529e244a2b..bcd4baecfe11 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c @@ -762,6 +762,7 @@ static void gmc_v7_0_gart_disable(struct amdgpu_device *adev) * @adev: amdgpu_device pointer * @status: VM_CONTEXT1_PROTECTION_FAULT_STATUS register value * @addr: VM_CONTEXT1_PROTECTION_FAULT_ADDR register value + * @mc_client: VM_CONTEXT1_PROTECTION_FAULT_MCCLIENT register value * * Print human readable fault information (CIK). */ diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c index 0aa5b82808d1..26976e50e2a2 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c @@ -1005,6 +1005,7 @@ static void gmc_v8_0_gart_disable(struct amdgpu_device *adev) * @adev: amdgpu_device pointer * @status: VM_CONTEXT1_PROTECTION_FAULT_STATUS register value * @addr: VM_CONTEXT1_PROTECTION_FAULT_ADDR register value + * @mc_client: VM_CONTEXT1_PROTECTION_FAULT_MCCLIENT register value * * Print human readable fault information (VI). */ diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v11_0.c b/drivers/gpu/drm/amd/amdgpu/psp_v11_0.c index 97c80f1d5731..1de89cc3c355 100644 --- a/drivers/gpu/drm/amd/amdgpu/psp_v11_0.c +++ b/drivers/gpu/drm/amd/amdgpu/psp_v11_0.c @@ -524,181 +524,6 @@ static int psp_v11_0_mode1_reset(struct psp_context *psp) return 0; } -/* TODO: Fill in follow functions once PSP firmware interface for XGMI is ready. - * For now, return success and hack the hive_id so high level code can - * start testing - */ -static int psp_v11_0_xgmi_get_topology_info(struct psp_context *psp, - int number_devices, struct psp_xgmi_topology_info *topology) -{ - struct ta_xgmi_shared_memory *xgmi_cmd; - struct ta_xgmi_cmd_get_topology_info_input *topology_info_input; - struct ta_xgmi_cmd_get_topology_info_output *topology_info_output; - int i; - int ret; - - if (!topology || topology->num_nodes > TA_XGMI__MAX_CONNECTED_NODES) - return -EINVAL; - - xgmi_cmd = (struct ta_xgmi_shared_memory*)psp->xgmi_context.xgmi_shared_buf; - memset(xgmi_cmd, 0, sizeof(struct ta_xgmi_shared_memory)); - - /* Fill in the shared memory with topology information as input */ - topology_info_input = &xgmi_cmd->xgmi_in_message.get_topology_info; - xgmi_cmd->cmd_id = TA_COMMAND_XGMI__GET_GET_TOPOLOGY_INFO; - topology_info_input->num_nodes = number_devices; - - for (i = 0; i < topology_info_input->num_nodes; i++) { - topology_info_input->nodes[i].node_id = topology->nodes[i].node_id; - topology_info_input->nodes[i].num_hops = topology->nodes[i].num_hops; - topology_info_input->nodes[i].is_sharing_enabled = topology->nodes[i].is_sharing_enabled; - topology_info_input->nodes[i].sdma_engine = topology->nodes[i].sdma_engine; - } - - /* Invoke xgmi ta to get the topology information */ - ret = psp_xgmi_invoke(psp, TA_COMMAND_XGMI__GET_GET_TOPOLOGY_INFO); - if (ret) - return ret; - - /* Read the output topology information from the shared memory */ - topology_info_output = &xgmi_cmd->xgmi_out_message.get_topology_info; - topology->num_nodes = xgmi_cmd->xgmi_out_message.get_topology_info.num_nodes; - for (i = 0; i < topology->num_nodes; i++) { - topology->nodes[i].node_id = topology_info_output->nodes[i].node_id; - topology->nodes[i].num_hops = topology_info_output->nodes[i].num_hops; - topology->nodes[i].is_sharing_enabled = topology_info_output->nodes[i].is_sharing_enabled; - topology->nodes[i].sdma_engine = topology_info_output->nodes[i].sdma_engine; - } - - return 0; -} - -static int psp_v11_0_xgmi_set_topology_info(struct psp_context *psp, - int number_devices, struct psp_xgmi_topology_info *topology) -{ - struct ta_xgmi_shared_memory *xgmi_cmd; - struct ta_xgmi_cmd_get_topology_info_input *topology_info_input; - int i; - - if (!topology || topology->num_nodes > TA_XGMI__MAX_CONNECTED_NODES) - return -EINVAL; - - xgmi_cmd = (struct ta_xgmi_shared_memory*)psp->xgmi_context.xgmi_shared_buf; - memset(xgmi_cmd, 0, sizeof(struct ta_xgmi_shared_memory)); - - topology_info_input = &xgmi_cmd->xgmi_in_message.get_topology_info; - xgmi_cmd->cmd_id = TA_COMMAND_XGMI__SET_TOPOLOGY_INFO; - topology_info_input->num_nodes = number_devices; - - for (i = 0; i < topology_info_input->num_nodes; i++) { - topology_info_input->nodes[i].node_id = topology->nodes[i].node_id; - topology_info_input->nodes[i].num_hops = topology->nodes[i].num_hops; - topology_info_input->nodes[i].is_sharing_enabled = 1; - topology_info_input->nodes[i].sdma_engine = topology->nodes[i].sdma_engine; - } - - /* Invoke xgmi ta to set topology information */ - return psp_xgmi_invoke(psp, TA_COMMAND_XGMI__SET_TOPOLOGY_INFO); -} - -static int psp_v11_0_xgmi_get_hive_id(struct psp_context *psp, uint64_t *hive_id) -{ - struct ta_xgmi_shared_memory *xgmi_cmd; - int ret; - - xgmi_cmd = (struct ta_xgmi_shared_memory*)psp->xgmi_context.xgmi_shared_buf; - memset(xgmi_cmd, 0, sizeof(struct ta_xgmi_shared_memory)); - - xgmi_cmd->cmd_id = TA_COMMAND_XGMI__GET_HIVE_ID; - - /* Invoke xgmi ta to get hive id */ - ret = psp_xgmi_invoke(psp, xgmi_cmd->cmd_id); - if (ret) - return ret; - - *hive_id = xgmi_cmd->xgmi_out_message.get_hive_id.hive_id; - - return 0; -} - -static int psp_v11_0_xgmi_get_node_id(struct psp_context *psp, uint64_t *node_id) -{ - struct ta_xgmi_shared_memory *xgmi_cmd; - int ret; - - xgmi_cmd = (struct ta_xgmi_shared_memory*)psp->xgmi_context.xgmi_shared_buf; - memset(xgmi_cmd, 0, sizeof(struct ta_xgmi_shared_memory)); - - xgmi_cmd->cmd_id = TA_COMMAND_XGMI__GET_NODE_ID; - - /* Invoke xgmi ta to get the node id */ - ret = psp_xgmi_invoke(psp, xgmi_cmd->cmd_id); - if (ret) - return ret; - - *node_id = xgmi_cmd->xgmi_out_message.get_node_id.node_id; - - return 0; -} - -static int psp_v11_0_ras_trigger_error(struct psp_context *psp, - struct ta_ras_trigger_error_input *info) -{ - struct ta_ras_shared_memory *ras_cmd; - int ret; - - if (!psp->ras.ras_initialized) - return -EINVAL; - - ras_cmd = (struct ta_ras_shared_memory *)psp->ras.ras_shared_buf; - memset(ras_cmd, 0, sizeof(struct ta_ras_shared_memory)); - - ras_cmd->cmd_id = TA_RAS_COMMAND__TRIGGER_ERROR; - ras_cmd->ras_in_message.trigger_error = *info; - - ret = psp_ras_invoke(psp, ras_cmd->cmd_id); - if (ret) - return -EINVAL; - - /* If err_event_athub occurs error inject was successful, however - return status from TA is no long reliable */ - if (amdgpu_ras_intr_triggered()) - return 0; - - return ras_cmd->ras_status; -} - -static int psp_v11_0_ras_cure_posion(struct psp_context *psp, uint64_t *mode_ptr) -{ -#if 0 - // not support yet. - struct ta_ras_shared_memory *ras_cmd; - int ret; - - if (!psp->ras.ras_initialized) - return -EINVAL; - - ras_cmd = (struct ta_ras_shared_memory *)psp->ras.ras_shared_buf; - memset(ras_cmd, 0, sizeof(struct ta_ras_shared_memory)); - - ras_cmd->cmd_id = TA_RAS_COMMAND__CURE_POISON; - ras_cmd->ras_in_message.cure_poison.mode_ptr = mode_ptr; - - ret = psp_ras_invoke(psp, ras_cmd->cmd_id); - if (ret) - return -EINVAL; - - return ras_cmd->ras_status; -#else - return -EINVAL; -#endif -} - -static int psp_v11_0_rlc_autoload_start(struct psp_context *psp) -{ - return psp_rlc_autoload_start(psp); -} - static int psp_v11_0_memory_training_send_msg(struct psp_context *psp, int msg) { int ret; @@ -995,13 +820,6 @@ static const struct psp_funcs psp_v11_0_funcs = { .ring_stop = psp_v11_0_ring_stop, .ring_destroy = psp_v11_0_ring_destroy, .mode1_reset = psp_v11_0_mode1_reset, - .xgmi_get_topology_info = psp_v11_0_xgmi_get_topology_info, - .xgmi_set_topology_info = psp_v11_0_xgmi_set_topology_info, - .xgmi_get_hive_id = psp_v11_0_xgmi_get_hive_id, - .xgmi_get_node_id = psp_v11_0_xgmi_get_node_id, - .ras_trigger_error = psp_v11_0_ras_trigger_error, - .ras_cure_posion = psp_v11_0_ras_cure_posion, - .rlc_autoload_start = psp_v11_0_rlc_autoload_start, .mem_training_init = psp_v11_0_memory_training_init, .mem_training_fini = psp_v11_0_memory_training_fini, .mem_training = psp_v11_0_memory_training, diff --git a/drivers/gpu/drm/amd/amdgpu/ta_ras_if.h b/drivers/gpu/drm/amd/amdgpu/ta_ras_if.h index ca7d05993ca2..745ed0fba1ed 100644 --- a/drivers/gpu/drm/amd/amdgpu/ta_ras_if.h +++ b/drivers/gpu/drm/amd/amdgpu/ta_ras_if.h @@ -24,6 +24,8 @@ #ifndef _TA_RAS_IF_H #define _TA_RAS_IF_H +#define RAS_TA_HOST_IF_VER 0 + /* Responses have bit 31 set */ #define RSP_ID_MASK (1U << 31) #define RSP_ID(cmdId) (((uint32_t)(cmdId)) | RSP_ID_MASK) @@ -36,18 +38,24 @@ enum ras_command { TA_RAS_COMMAND__TRIGGER_ERROR, }; -enum ta_ras_status { - TA_RAS_STATUS__SUCCESS = 0x00, - TA_RAS_STATUS__RESET_NEEDED = 0x01, - TA_RAS_STATUS__ERROR_INVALID_PARAMETER = 0x02, - TA_RAS_STATUS__ERROR_RAS_NOT_AVAILABLE = 0x03, - TA_RAS_STATUS__ERROR_RAS_DUPLICATE_CMD = 0x04, - TA_RAS_STATUS__ERROR_INJECTION_FAILED = 0x05, - TA_RAS_STATUS__ERROR_ASD_READ_WRITE = 0x06, - TA_RAS_STATUS__ERROR_TOGGLE_DF_CSTATE = 0x07, - TA_RAS_STATUS__ERROR_TIMEOUT = 0x08, - TA_RAS_STATUS__ERROR_BLOCK_DISABLED = 0x09, - TA_RAS_STATUS__ERROR_GENERIC = 0x10, +enum ta_ras_status +{ + TA_RAS_STATUS__SUCCESS = 0x00, + TA_RAS_STATUS__RESET_NEEDED = 0xA001, + TA_RAS_STATUS__ERROR_INVALID_PARAMETER = 0xA002, + TA_RAS_STATUS__ERROR_RAS_NOT_AVAILABLE = 0xA003, + TA_RAS_STATUS__ERROR_RAS_DUPLICATE_CMD = 0xA004, + TA_RAS_STATUS__ERROR_INJECTION_FAILED = 0xA005, + TA_RAS_STATUS__ERROR_ASD_READ_WRITE = 0xA006, + TA_RAS_STATUS__ERROR_TOGGLE_DF_CSTATE = 0xA007, + TA_RAS_STATUS__ERROR_TIMEOUT = 0xA008, + TA_RAS_STATUS__ERROR_BLOCK_DISABLED = 0XA009, + TA_RAS_STATUS__ERROR_GENERIC = 0xA00A, + TA_RAS_STATUS__ERROR_RAS_MMHUB_INIT = 0xA00B, + TA_RAS_STATUS__ERROR_GET_DEV_INFO = 0xA00C, + TA_RAS_STATUS__ERROR_UNSUPPORTED_DEV = 0xA00D, + TA_RAS_STATUS__ERROR_NOT_INITIALIZED = 0xA00E, + TA_RAS_STATUS__ERROR_TEE_INTERNAL = 0xA00F }; enum ta_ras_block { @@ -97,22 +105,39 @@ struct ta_ras_trigger_error_input { uint64_t value; // method if error injection. i.e persistent, coherent etc. }; +struct ta_ras_output_flags +{ + uint8_t ras_init_success_flag; + uint8_t err_inject_switch_disable_flag; + uint8_t reg_access_failure_flag; +}; + /* Common input structure for RAS callbacks */ /**********************************************************/ union ta_ras_cmd_input { struct ta_ras_enable_features_input enable_features; struct ta_ras_disable_features_input disable_features; struct ta_ras_trigger_error_input trigger_error; + + uint32_t reserve_pad[256]; +}; + +union ta_ras_cmd_output +{ + struct ta_ras_output_flags flags; + + uint32_t reserve_pad[256]; }; /* Shared Memory structures */ /**********************************************************/ struct ta_ras_shared_memory { - uint32_t cmd_id; - uint32_t resp_id; - enum ta_ras_status ras_status; - uint32_t reserved; - union ta_ras_cmd_input ras_in_message; + uint32_t cmd_id; + uint32_t resp_id; + uint32_t ras_status; + uint32_t if_version; + union ta_ras_cmd_input ras_in_message; + union ta_ras_cmd_output ras_out_message; }; #endif // TL_RAS_IF_H_ diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v2_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v2_0.c index 349da7bf7c68..90ed773695ea 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v2_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v2_0.c @@ -1845,7 +1845,6 @@ static int vcn_v2_0_start_sriov(struct amdgpu_device *adev) uint32_t table_size = 0; struct mmsch_v2_0_cmd_direct_write direct_wt = { {0} }; struct mmsch_v2_0_cmd_direct_read_modify_write direct_rd_mod_wt = { {0} }; - struct mmsch_v2_0_cmd_direct_polling direct_poll = { {0} }; struct mmsch_v2_0_cmd_end end = { {0} }; struct mmsch_v2_0_init_header *header; uint32_t *init_table = adev->virt.mm_table.cpu_addr; @@ -1855,8 +1854,6 @@ static int vcn_v2_0_start_sriov(struct amdgpu_device *adev) direct_wt.cmd_header.command_type = MMSCH_COMMAND__DIRECT_REG_WRITE; direct_rd_mod_wt.cmd_header.command_type = MMSCH_COMMAND__DIRECT_REG_READ_MODIFY_WRITE; - direct_poll.cmd_header.command_type = - MMSCH_COMMAND__DIRECT_REG_POLLING; end.cmd_header.command_type = MMSCH_COMMAND__END; if (header->vcn_table_offset == 0 && header->vcn_table_size == 0) { diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c b/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c index 38ca4a712f12..34ed906645c9 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c @@ -86,7 +86,7 @@ static int vcn_v2_5_early_init(void *handle) adev->vcn.num_vcn_inst = VCN25_MAX_HW_INSTANCES_ARCTURUS; for (i = 0; i < adev->vcn.num_vcn_inst; i++) { - harvest = RREG32_SOC15(UVD, i, mmCC_UVD_HARVESTING); + harvest = RREG32_SOC15(VCN, i, mmCC_UVD_HARVESTING); if (harvest & CC_UVD_HARVESTING__UVD_DISABLE_MASK) adev->vcn.harvest_config |= 1 << i; } @@ -177,15 +177,15 @@ static int vcn_v2_5_sw_init(void *handle) adev->vcn.internal.gp_scratch8 = mmUVD_GP_SCRATCH8_INTERNAL_OFFSET; adev->vcn.internal.scratch9 = mmUVD_SCRATCH9_INTERNAL_OFFSET; - adev->vcn.inst[j].external.scratch9 = SOC15_REG_OFFSET(UVD, j, mmUVD_SCRATCH9); + adev->vcn.inst[j].external.scratch9 = SOC15_REG_OFFSET(VCN, j, mmUVD_SCRATCH9); adev->vcn.internal.data0 = mmUVD_GPCOM_VCPU_DATA0_INTERNAL_OFFSET; - adev->vcn.inst[j].external.data0 = SOC15_REG_OFFSET(UVD, j, mmUVD_GPCOM_VCPU_DATA0); + adev->vcn.inst[j].external.data0 = SOC15_REG_OFFSET(VCN, j, mmUVD_GPCOM_VCPU_DATA0); adev->vcn.internal.data1 = mmUVD_GPCOM_VCPU_DATA1_INTERNAL_OFFSET; - adev->vcn.inst[j].external.data1 = SOC15_REG_OFFSET(UVD, j, mmUVD_GPCOM_VCPU_DATA1); + adev->vcn.inst[j].external.data1 = SOC15_REG_OFFSET(VCN, j, mmUVD_GPCOM_VCPU_DATA1); adev->vcn.internal.cmd = mmUVD_GPCOM_VCPU_CMD_INTERNAL_OFFSET; - adev->vcn.inst[j].external.cmd = SOC15_REG_OFFSET(UVD, j, mmUVD_GPCOM_VCPU_CMD); + adev->vcn.inst[j].external.cmd = SOC15_REG_OFFSET(VCN, j, mmUVD_GPCOM_VCPU_CMD); adev->vcn.internal.nop = mmUVD_NO_OP_INTERNAL_OFFSET; - adev->vcn.inst[j].external.nop = SOC15_REG_OFFSET(UVD, j, mmUVD_NO_OP); + adev->vcn.inst[j].external.nop = SOC15_REG_OFFSET(VCN, j, mmUVD_NO_OP); ring = &adev->vcn.inst[j].ring_dec; ring->use_doorbell = true; @@ -399,46 +399,46 @@ static void vcn_v2_5_mc_resume(struct amdgpu_device *adev) continue; /* cache window 0: fw */ if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) { - WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW, + WREG32_SOC15(VCN, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW, (adev->firmware.ucode[AMDGPU_UCODE_ID_VCN + i].tmr_mc_addr_lo)); - WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH, + WREG32_SOC15(VCN, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH, (adev->firmware.ucode[AMDGPU_UCODE_ID_VCN + i].tmr_mc_addr_hi)); - WREG32_SOC15(UVD, i, mmUVD_VCPU_CACHE_OFFSET0, 0); + WREG32_SOC15(VCN, i, mmUVD_VCPU_CACHE_OFFSET0, 0); offset = 0; } else { - WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW, + WREG32_SOC15(VCN, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW, lower_32_bits(adev->vcn.inst[i].gpu_addr)); - WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH, + WREG32_SOC15(VCN, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH, upper_32_bits(adev->vcn.inst[i].gpu_addr)); offset = size; - WREG32_SOC15(UVD, i, mmUVD_VCPU_CACHE_OFFSET0, + WREG32_SOC15(VCN, i, mmUVD_VCPU_CACHE_OFFSET0, AMDGPU_UVD_FIRMWARE_OFFSET >> 3); } - WREG32_SOC15(UVD, i, mmUVD_VCPU_CACHE_SIZE0, size); + WREG32_SOC15(VCN, i, mmUVD_VCPU_CACHE_SIZE0, size); /* cache window 1: stack */ - WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_LOW, + WREG32_SOC15(VCN, i, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_LOW, lower_32_bits(adev->vcn.inst[i].gpu_addr + offset)); - WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_HIGH, + WREG32_SOC15(VCN, i, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_HIGH, upper_32_bits(adev->vcn.inst[i].gpu_addr + offset)); - WREG32_SOC15(UVD, i, mmUVD_VCPU_CACHE_OFFSET1, 0); - WREG32_SOC15(UVD, i, mmUVD_VCPU_CACHE_SIZE1, AMDGPU_VCN_STACK_SIZE); + WREG32_SOC15(VCN, i, mmUVD_VCPU_CACHE_OFFSET1, 0); + WREG32_SOC15(VCN, i, mmUVD_VCPU_CACHE_SIZE1, AMDGPU_VCN_STACK_SIZE); /* cache window 2: context */ - WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_LOW, + WREG32_SOC15(VCN, i, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_LOW, lower_32_bits(adev->vcn.inst[i].gpu_addr + offset + AMDGPU_VCN_STACK_SIZE)); - WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_HIGH, + WREG32_SOC15(VCN, i, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_HIGH, upper_32_bits(adev->vcn.inst[i].gpu_addr + offset + AMDGPU_VCN_STACK_SIZE)); - WREG32_SOC15(UVD, i, mmUVD_VCPU_CACHE_OFFSET2, 0); - WREG32_SOC15(UVD, i, mmUVD_VCPU_CACHE_SIZE2, AMDGPU_VCN_CONTEXT_SIZE); + WREG32_SOC15(VCN, i, mmUVD_VCPU_CACHE_OFFSET2, 0); + WREG32_SOC15(VCN, i, mmUVD_VCPU_CACHE_SIZE2, AMDGPU_VCN_CONTEXT_SIZE); /* non-cache window */ - WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_NC0_64BIT_BAR_LOW, + WREG32_SOC15(VCN, i, mmUVD_LMI_VCPU_NC0_64BIT_BAR_LOW, lower_32_bits(adev->vcn.inst[i].fw_shared_gpu_addr)); - WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_NC0_64BIT_BAR_HIGH, + WREG32_SOC15(VCN, i, mmUVD_LMI_VCPU_NC0_64BIT_BAR_HIGH, upper_32_bits(adev->vcn.inst[i].fw_shared_gpu_addr)); - WREG32_SOC15(UVD, i, mmUVD_VCPU_NONCACHE_OFFSET0, 0); - WREG32_SOC15(UVD, i, mmUVD_VCPU_NONCACHE_SIZE0, + WREG32_SOC15(VCN, i, mmUVD_VCPU_NONCACHE_OFFSET0, 0); + WREG32_SOC15(VCN, i, mmUVD_VCPU_NONCACHE_SIZE0, AMDGPU_GPU_PAGE_ALIGN(sizeof(struct amdgpu_fw_shared))); } } @@ -452,91 +452,91 @@ static void vcn_v2_5_mc_resume_dpg_mode(struct amdgpu_device *adev, int inst_idx if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) { if (!indirect) { WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW), + VCN, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW), (adev->firmware.ucode[AMDGPU_UCODE_ID_VCN + inst_idx].tmr_mc_addr_lo), 0, indirect); WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH), + VCN, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH), (adev->firmware.ucode[AMDGPU_UCODE_ID_VCN + inst_idx].tmr_mc_addr_hi), 0, indirect); WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_VCPU_CACHE_OFFSET0), 0, 0, indirect); + VCN, 0, mmUVD_VCPU_CACHE_OFFSET0), 0, 0, indirect); } else { WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW), 0, 0, indirect); + VCN, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW), 0, 0, indirect); WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH), 0, 0, indirect); + VCN, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH), 0, 0, indirect); WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_VCPU_CACHE_OFFSET0), 0, 0, indirect); + VCN, 0, mmUVD_VCPU_CACHE_OFFSET0), 0, 0, indirect); } offset = 0; } else { WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW), + VCN, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW), lower_32_bits(adev->vcn.inst[inst_idx].gpu_addr), 0, indirect); WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH), + VCN, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH), upper_32_bits(adev->vcn.inst[inst_idx].gpu_addr), 0, indirect); offset = size; WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_VCPU_CACHE_OFFSET0), + VCN, 0, mmUVD_VCPU_CACHE_OFFSET0), AMDGPU_UVD_FIRMWARE_OFFSET >> 3, 0, indirect); } if (!indirect) WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_VCPU_CACHE_SIZE0), size, 0, indirect); + VCN, 0, mmUVD_VCPU_CACHE_SIZE0), size, 0, indirect); else WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_VCPU_CACHE_SIZE0), 0, 0, indirect); + VCN, 0, mmUVD_VCPU_CACHE_SIZE0), 0, 0, indirect); /* cache window 1: stack */ if (!indirect) { WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_LOW), + VCN, 0, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_LOW), lower_32_bits(adev->vcn.inst[inst_idx].gpu_addr + offset), 0, indirect); WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_HIGH), + VCN, 0, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_HIGH), upper_32_bits(adev->vcn.inst[inst_idx].gpu_addr + offset), 0, indirect); WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_VCPU_CACHE_OFFSET1), 0, 0, indirect); + VCN, 0, mmUVD_VCPU_CACHE_OFFSET1), 0, 0, indirect); } else { WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_LOW), 0, 0, indirect); + VCN, 0, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_LOW), 0, 0, indirect); WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_HIGH), 0, 0, indirect); + VCN, 0, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_HIGH), 0, 0, indirect); WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_VCPU_CACHE_OFFSET1), 0, 0, indirect); + VCN, 0, mmUVD_VCPU_CACHE_OFFSET1), 0, 0, indirect); } WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_VCPU_CACHE_SIZE1), AMDGPU_VCN_STACK_SIZE, 0, indirect); + VCN, 0, mmUVD_VCPU_CACHE_SIZE1), AMDGPU_VCN_STACK_SIZE, 0, indirect); /* cache window 2: context */ WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_LOW), + VCN, 0, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_LOW), lower_32_bits(adev->vcn.inst[inst_idx].gpu_addr + offset + AMDGPU_VCN_STACK_SIZE), 0, indirect); WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_HIGH), + VCN, 0, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_HIGH), upper_32_bits(adev->vcn.inst[inst_idx].gpu_addr + offset + AMDGPU_VCN_STACK_SIZE), 0, indirect); WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_VCPU_CACHE_OFFSET2), 0, 0, indirect); + VCN, 0, mmUVD_VCPU_CACHE_OFFSET2), 0, 0, indirect); WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_VCPU_CACHE_SIZE2), AMDGPU_VCN_CONTEXT_SIZE, 0, indirect); + VCN, 0, mmUVD_VCPU_CACHE_SIZE2), AMDGPU_VCN_CONTEXT_SIZE, 0, indirect); /* non-cache window */ WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_LMI_VCPU_NC0_64BIT_BAR_LOW), + VCN, 0, mmUVD_LMI_VCPU_NC0_64BIT_BAR_LOW), lower_32_bits(adev->vcn.inst[inst_idx].fw_shared_gpu_addr), 0, indirect); WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_LMI_VCPU_NC0_64BIT_BAR_HIGH), + VCN, 0, mmUVD_LMI_VCPU_NC0_64BIT_BAR_HIGH), upper_32_bits(adev->vcn.inst[inst_idx].fw_shared_gpu_addr), 0, indirect); WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_VCPU_NONCACHE_OFFSET0), 0, 0, indirect); + VCN, 0, mmUVD_VCPU_NONCACHE_OFFSET0), 0, 0, indirect); WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_VCPU_NONCACHE_SIZE0), + VCN, 0, mmUVD_VCPU_NONCACHE_SIZE0), AMDGPU_GPU_PAGE_ALIGN(sizeof(struct amdgpu_fw_shared)), 0, indirect); /* VCN global tiling registers */ WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_GFX8_ADDR_CONFIG), adev->gfx.config.gb_addr_config, 0, indirect); + VCN, 0, mmUVD_GFX8_ADDR_CONFIG), adev->gfx.config.gb_addr_config, 0, indirect); } /** @@ -690,19 +690,19 @@ static void vcn_v2_5_clock_gating_dpg_mode(struct amdgpu_device *adev, UVD_CGC_CTRL__VCPU_MODE_MASK | UVD_CGC_CTRL__MMSCH_MODE_MASK); WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_CGC_CTRL), reg_data, sram_sel, indirect); + VCN, 0, mmUVD_CGC_CTRL), reg_data, sram_sel, indirect); /* turn off clock gating */ WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_CGC_GATE), 0, sram_sel, indirect); + VCN, 0, mmUVD_CGC_GATE), 0, sram_sel, indirect); /* turn on SUVD clock gating */ WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_SUVD_CGC_GATE), 1, sram_sel, indirect); + VCN, 0, mmUVD_SUVD_CGC_GATE), 1, sram_sel, indirect); /* turn on sw mode in UVD_SUVD_CGC_CTRL */ WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_SUVD_CGC_CTRL), 0, sram_sel, indirect); + VCN, 0, mmUVD_SUVD_CGC_CTRL), 0, sram_sel, indirect); } /** @@ -774,13 +774,13 @@ static int vcn_v2_5_start_dpg_mode(struct amdgpu_device *adev, int inst_idx, boo uint32_t rb_bufsz, tmp; /* disable register anti-hang mechanism */ - WREG32_P(SOC15_REG_OFFSET(UVD, inst_idx, mmUVD_POWER_STATUS), 1, + WREG32_P(SOC15_REG_OFFSET(VCN, inst_idx, mmUVD_POWER_STATUS), 1, ~UVD_POWER_STATUS__UVD_POWER_STATUS_MASK); /* enable dynamic power gating mode */ - tmp = RREG32_SOC15(UVD, inst_idx, mmUVD_POWER_STATUS); + tmp = RREG32_SOC15(VCN, inst_idx, mmUVD_POWER_STATUS); tmp |= UVD_POWER_STATUS__UVD_PG_MODE_MASK; tmp |= UVD_POWER_STATUS__UVD_PG_EN_MASK; - WREG32_SOC15(UVD, inst_idx, mmUVD_POWER_STATUS, tmp); + WREG32_SOC15(VCN, inst_idx, mmUVD_POWER_STATUS, tmp); if (indirect) adev->vcn.inst[inst_idx].dpg_sram_curr_addr = (uint32_t*)adev->vcn.inst[inst_idx].dpg_sram_cpu_addr; @@ -793,11 +793,11 @@ static int vcn_v2_5_start_dpg_mode(struct amdgpu_device *adev, int inst_idx, boo tmp |= UVD_VCPU_CNTL__CLK_EN_MASK; tmp |= UVD_VCPU_CNTL__BLK_RST_MASK; WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_VCPU_CNTL), tmp, 0, indirect); + VCN, 0, mmUVD_VCPU_CNTL), tmp, 0, indirect); /* disable master interupt */ WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_MASTINT_EN), 0, 0, indirect); + VCN, 0, mmUVD_MASTINT_EN), 0, 0, indirect); /* setup mmUVD_LMI_CTRL */ tmp = (0x8 | UVD_LMI_CTRL__WRITE_CLEAN_TIMER_EN_MASK | @@ -809,28 +809,28 @@ static int vcn_v2_5_start_dpg_mode(struct amdgpu_device *adev, int inst_idx, boo (8 << UVD_LMI_CTRL__WRITE_CLEAN_TIMER__SHIFT) | 0x00100000L); WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_LMI_CTRL), tmp, 0, indirect); + VCN, 0, mmUVD_LMI_CTRL), tmp, 0, indirect); WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_MPC_CNTL), + VCN, 0, mmUVD_MPC_CNTL), 0x2 << UVD_MPC_CNTL__REPLACEMENT_MODE__SHIFT, 0, indirect); WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_MPC_SET_MUXA0), + VCN, 0, mmUVD_MPC_SET_MUXA0), ((0x1 << UVD_MPC_SET_MUXA0__VARA_1__SHIFT) | (0x2 << UVD_MPC_SET_MUXA0__VARA_2__SHIFT) | (0x3 << UVD_MPC_SET_MUXA0__VARA_3__SHIFT) | (0x4 << UVD_MPC_SET_MUXA0__VARA_4__SHIFT)), 0, indirect); WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_MPC_SET_MUXB0), + VCN, 0, mmUVD_MPC_SET_MUXB0), ((0x1 << UVD_MPC_SET_MUXB0__VARB_1__SHIFT) | (0x2 << UVD_MPC_SET_MUXB0__VARB_2__SHIFT) | (0x3 << UVD_MPC_SET_MUXB0__VARB_3__SHIFT) | (0x4 << UVD_MPC_SET_MUXB0__VARB_4__SHIFT)), 0, indirect); WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_MPC_SET_MUX), + VCN, 0, mmUVD_MPC_SET_MUX), ((0x0 << UVD_MPC_SET_MUX__SET_0__SHIFT) | (0x1 << UVD_MPC_SET_MUX__SET_1__SHIFT) | (0x2 << UVD_MPC_SET_MUX__SET_2__SHIFT)), 0, indirect); @@ -838,26 +838,26 @@ static int vcn_v2_5_start_dpg_mode(struct amdgpu_device *adev, int inst_idx, boo vcn_v2_5_mc_resume_dpg_mode(adev, inst_idx, indirect); WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_REG_XX_MASK), 0x10, 0, indirect); + VCN, 0, mmUVD_REG_XX_MASK), 0x10, 0, indirect); WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_RBC_XX_IB_REG_CHECK), 0x3, 0, indirect); + VCN, 0, mmUVD_RBC_XX_IB_REG_CHECK), 0x3, 0, indirect); /* enable LMI MC and UMC channels */ WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_LMI_CTRL2), 0, 0, indirect); + VCN, 0, mmUVD_LMI_CTRL2), 0, 0, indirect); /* unblock VCPU register access */ WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_RB_ARB_CTRL), 0, 0, indirect); + VCN, 0, mmUVD_RB_ARB_CTRL), 0, 0, indirect); tmp = (0xFF << UVD_VCPU_CNTL__PRB_TIMEOUT_VAL__SHIFT); tmp |= UVD_VCPU_CNTL__CLK_EN_MASK; WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_VCPU_CNTL), tmp, 0, indirect); + VCN, 0, mmUVD_VCPU_CNTL), tmp, 0, indirect); /* enable master interrupt */ WREG32_SOC15_DPG_MODE_2_0(inst_idx, SOC15_DPG_MODE_OFFSET_2_0( - UVD, 0, mmUVD_MASTINT_EN), + VCN, 0, mmUVD_MASTINT_EN), UVD_MASTINT_EN__VCPU_EN_MASK, 0, indirect); if (indirect) @@ -873,39 +873,39 @@ static int vcn_v2_5_start_dpg_mode(struct amdgpu_device *adev, int inst_idx, boo tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_NO_FETCH, 1); tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_NO_UPDATE, 1); tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_RPTR_WR_EN, 1); - WREG32_SOC15(UVD, inst_idx, mmUVD_RBC_RB_CNTL, tmp); + WREG32_SOC15(VCN, inst_idx, mmUVD_RBC_RB_CNTL, tmp); /* Stall DPG before WPTR/RPTR reset */ - WREG32_P(SOC15_REG_OFFSET(UVD, inst_idx, mmUVD_POWER_STATUS), + WREG32_P(SOC15_REG_OFFSET(VCN, inst_idx, mmUVD_POWER_STATUS), UVD_POWER_STATUS__STALL_DPG_POWER_UP_MASK, ~UVD_POWER_STATUS__STALL_DPG_POWER_UP_MASK); fw_shared->multi_queue.decode_queue_mode |= FW_QUEUE_RING_RESET; /* set the write pointer delay */ - WREG32_SOC15(UVD, inst_idx, mmUVD_RBC_RB_WPTR_CNTL, 0); + WREG32_SOC15(VCN, inst_idx, mmUVD_RBC_RB_WPTR_CNTL, 0); /* set the wb address */ - WREG32_SOC15(UVD, inst_idx, mmUVD_RBC_RB_RPTR_ADDR, + WREG32_SOC15(VCN, inst_idx, mmUVD_RBC_RB_RPTR_ADDR, (upper_32_bits(ring->gpu_addr) >> 2)); /* programm the RB_BASE for ring buffer */ - WREG32_SOC15(UVD, inst_idx, mmUVD_LMI_RBC_RB_64BIT_BAR_LOW, + WREG32_SOC15(VCN, inst_idx, mmUVD_LMI_RBC_RB_64BIT_BAR_LOW, lower_32_bits(ring->gpu_addr)); - WREG32_SOC15(UVD, inst_idx, mmUVD_LMI_RBC_RB_64BIT_BAR_HIGH, + WREG32_SOC15(VCN, inst_idx, mmUVD_LMI_RBC_RB_64BIT_BAR_HIGH, upper_32_bits(ring->gpu_addr)); /* Initialize the ring buffer's read and write pointers */ - WREG32_SOC15(UVD, inst_idx, mmUVD_RBC_RB_RPTR, 0); + WREG32_SOC15(VCN, inst_idx, mmUVD_RBC_RB_RPTR, 0); - WREG32_SOC15(UVD, inst_idx, mmUVD_SCRATCH2, 0); + WREG32_SOC15(VCN, inst_idx, mmUVD_SCRATCH2, 0); - ring->wptr = RREG32_SOC15(UVD, inst_idx, mmUVD_RBC_RB_RPTR); - WREG32_SOC15(UVD, inst_idx, mmUVD_RBC_RB_WPTR, + ring->wptr = RREG32_SOC15(VCN, inst_idx, mmUVD_RBC_RB_RPTR); + WREG32_SOC15(VCN, inst_idx, mmUVD_RBC_RB_WPTR, lower_32_bits(ring->wptr)); fw_shared->multi_queue.decode_queue_mode &= ~FW_QUEUE_RING_RESET; /* Unstall DPG */ - WREG32_P(SOC15_REG_OFFSET(UVD, inst_idx, mmUVD_POWER_STATUS), + WREG32_P(SOC15_REG_OFFSET(VCN, inst_idx, mmUVD_POWER_STATUS), 0, ~UVD_POWER_STATUS__STALL_DPG_POWER_UP_MASK); return 0; @@ -929,12 +929,12 @@ static int vcn_v2_5_start(struct amdgpu_device *adev) } /* disable register anti-hang mechanism */ - WREG32_P(SOC15_REG_OFFSET(UVD, i, mmUVD_POWER_STATUS), 0, + WREG32_P(SOC15_REG_OFFSET(VCN, i, mmUVD_POWER_STATUS), 0, ~UVD_POWER_STATUS__UVD_POWER_STATUS_MASK); /* set uvd status busy */ - tmp = RREG32_SOC15(UVD, i, mmUVD_STATUS) | UVD_STATUS__UVD_BUSY; - WREG32_SOC15(UVD, i, mmUVD_STATUS, tmp); + tmp = RREG32_SOC15(VCN, i, mmUVD_STATUS) | UVD_STATUS__UVD_BUSY; + WREG32_SOC15(VCN, i, mmUVD_STATUS, tmp); } if (adev->pg_flags & AMD_PG_SUPPORT_VCN_DPG) @@ -947,44 +947,44 @@ static int vcn_v2_5_start(struct amdgpu_device *adev) if (adev->vcn.harvest_config & (1 << i)) continue; /* enable VCPU clock */ - WREG32_P(SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CNTL), + WREG32_P(SOC15_REG_OFFSET(VCN, i, mmUVD_VCPU_CNTL), UVD_VCPU_CNTL__CLK_EN_MASK, ~UVD_VCPU_CNTL__CLK_EN_MASK); /* disable master interrupt */ - WREG32_P(SOC15_REG_OFFSET(UVD, i, mmUVD_MASTINT_EN), 0, + WREG32_P(SOC15_REG_OFFSET(VCN, i, mmUVD_MASTINT_EN), 0, ~UVD_MASTINT_EN__VCPU_EN_MASK); /* setup mmUVD_LMI_CTRL */ - tmp = RREG32_SOC15(UVD, i, mmUVD_LMI_CTRL); + tmp = RREG32_SOC15(VCN, i, mmUVD_LMI_CTRL); tmp &= ~0xff; - WREG32_SOC15(UVD, i, mmUVD_LMI_CTRL, tmp | 0x8| + WREG32_SOC15(VCN, i, mmUVD_LMI_CTRL, tmp | 0x8| UVD_LMI_CTRL__WRITE_CLEAN_TIMER_EN_MASK | UVD_LMI_CTRL__MASK_MC_URGENT_MASK | UVD_LMI_CTRL__DATA_COHERENCY_EN_MASK | UVD_LMI_CTRL__VCPU_DATA_COHERENCY_EN_MASK); /* setup mmUVD_MPC_CNTL */ - tmp = RREG32_SOC15(UVD, i, mmUVD_MPC_CNTL); + tmp = RREG32_SOC15(VCN, i, mmUVD_MPC_CNTL); tmp &= ~UVD_MPC_CNTL__REPLACEMENT_MODE_MASK; tmp |= 0x2 << UVD_MPC_CNTL__REPLACEMENT_MODE__SHIFT; WREG32_SOC15(VCN, i, mmUVD_MPC_CNTL, tmp); /* setup UVD_MPC_SET_MUXA0 */ - WREG32_SOC15(UVD, i, mmUVD_MPC_SET_MUXA0, + WREG32_SOC15(VCN, i, mmUVD_MPC_SET_MUXA0, ((0x1 << UVD_MPC_SET_MUXA0__VARA_1__SHIFT) | (0x2 << UVD_MPC_SET_MUXA0__VARA_2__SHIFT) | (0x3 << UVD_MPC_SET_MUXA0__VARA_3__SHIFT) | (0x4 << UVD_MPC_SET_MUXA0__VARA_4__SHIFT))); /* setup UVD_MPC_SET_MUXB0 */ - WREG32_SOC15(UVD, i, mmUVD_MPC_SET_MUXB0, + WREG32_SOC15(VCN, i, mmUVD_MPC_SET_MUXB0, ((0x1 << UVD_MPC_SET_MUXB0__VARB_1__SHIFT) | (0x2 << UVD_MPC_SET_MUXB0__VARB_2__SHIFT) | (0x3 << UVD_MPC_SET_MUXB0__VARB_3__SHIFT) | (0x4 << UVD_MPC_SET_MUXB0__VARB_4__SHIFT))); /* setup mmUVD_MPC_SET_MUX */ - WREG32_SOC15(UVD, i, mmUVD_MPC_SET_MUX, + WREG32_SOC15(VCN, i, mmUVD_MPC_SET_MUX, ((0x0 << UVD_MPC_SET_MUX__SET_0__SHIFT) | (0x1 << UVD_MPC_SET_MUX__SET_1__SHIFT) | (0x2 << UVD_MPC_SET_MUX__SET_2__SHIFT))); @@ -997,27 +997,27 @@ static int vcn_v2_5_start(struct amdgpu_device *adev) if (adev->vcn.harvest_config & (1 << i)) continue; /* VCN global tiling registers */ - WREG32_SOC15(UVD, i, mmUVD_GFX8_ADDR_CONFIG, + WREG32_SOC15(VCN, i, mmUVD_GFX8_ADDR_CONFIG, adev->gfx.config.gb_addr_config); - WREG32_SOC15(UVD, i, mmUVD_GFX8_ADDR_CONFIG, + WREG32_SOC15(VCN, i, mmUVD_GFX8_ADDR_CONFIG, adev->gfx.config.gb_addr_config); /* enable LMI MC and UMC channels */ - WREG32_P(SOC15_REG_OFFSET(UVD, i, mmUVD_LMI_CTRL2), 0, + WREG32_P(SOC15_REG_OFFSET(VCN, i, mmUVD_LMI_CTRL2), 0, ~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK); /* unblock VCPU register access */ - WREG32_P(SOC15_REG_OFFSET(UVD, i, mmUVD_RB_ARB_CTRL), 0, + WREG32_P(SOC15_REG_OFFSET(VCN, i, mmUVD_RB_ARB_CTRL), 0, ~UVD_RB_ARB_CTRL__VCPU_DIS_MASK); - WREG32_P(SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CNTL), 0, + WREG32_P(SOC15_REG_OFFSET(VCN, i, mmUVD_VCPU_CNTL), 0, ~UVD_VCPU_CNTL__BLK_RST_MASK); for (k = 0; k < 10; ++k) { uint32_t status; for (j = 0; j < 100; ++j) { - status = RREG32_SOC15(UVD, i, mmUVD_STATUS); + status = RREG32_SOC15(VCN, i, mmUVD_STATUS); if (status & 2) break; if (amdgpu_emu_mode == 1) @@ -1030,11 +1030,11 @@ static int vcn_v2_5_start(struct amdgpu_device *adev) break; DRM_ERROR("VCN decode not responding, trying to reset the VCPU!!!\n"); - WREG32_P(SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CNTL), + WREG32_P(SOC15_REG_OFFSET(VCN, i, mmUVD_VCPU_CNTL), UVD_VCPU_CNTL__BLK_RST_MASK, ~UVD_VCPU_CNTL__BLK_RST_MASK); mdelay(10); - WREG32_P(SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CNTL), 0, + WREG32_P(SOC15_REG_OFFSET(VCN, i, mmUVD_VCPU_CNTL), 0, ~UVD_VCPU_CNTL__BLK_RST_MASK); mdelay(10); @@ -1047,15 +1047,15 @@ static int vcn_v2_5_start(struct amdgpu_device *adev) } /* enable master interrupt */ - WREG32_P(SOC15_REG_OFFSET(UVD, i, mmUVD_MASTINT_EN), + WREG32_P(SOC15_REG_OFFSET(VCN, i, mmUVD_MASTINT_EN), UVD_MASTINT_EN__VCPU_EN_MASK, ~UVD_MASTINT_EN__VCPU_EN_MASK); /* clear the busy bit of VCN_STATUS */ - WREG32_P(SOC15_REG_OFFSET(UVD, i, mmUVD_STATUS), 0, + WREG32_P(SOC15_REG_OFFSET(VCN, i, mmUVD_STATUS), 0, ~(2 << UVD_STATUS__VCPU_REPORT__SHIFT)); - WREG32_SOC15(UVD, i, mmUVD_LMI_RBC_RB_VMID, 0); + WREG32_SOC15(VCN, i, mmUVD_LMI_RBC_RB_VMID, 0); ring = &adev->vcn.inst[i].ring_dec; /* force RBC into idle state */ @@ -1065,39 +1065,39 @@ static int vcn_v2_5_start(struct amdgpu_device *adev) tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_NO_FETCH, 1); tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_NO_UPDATE, 1); tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_RPTR_WR_EN, 1); - WREG32_SOC15(UVD, i, mmUVD_RBC_RB_CNTL, tmp); + WREG32_SOC15(VCN, i, mmUVD_RBC_RB_CNTL, tmp); fw_shared->multi_queue.decode_queue_mode |= FW_QUEUE_RING_RESET; /* programm the RB_BASE for ring buffer */ - WREG32_SOC15(UVD, i, mmUVD_LMI_RBC_RB_64BIT_BAR_LOW, + WREG32_SOC15(VCN, i, mmUVD_LMI_RBC_RB_64BIT_BAR_LOW, lower_32_bits(ring->gpu_addr)); - WREG32_SOC15(UVD, i, mmUVD_LMI_RBC_RB_64BIT_BAR_HIGH, + WREG32_SOC15(VCN, i, mmUVD_LMI_RBC_RB_64BIT_BAR_HIGH, upper_32_bits(ring->gpu_addr)); /* Initialize the ring buffer's read and write pointers */ - WREG32_SOC15(UVD, i, mmUVD_RBC_RB_RPTR, 0); + WREG32_SOC15(VCN, i, mmUVD_RBC_RB_RPTR, 0); - ring->wptr = RREG32_SOC15(UVD, i, mmUVD_RBC_RB_RPTR); - WREG32_SOC15(UVD, i, mmUVD_RBC_RB_WPTR, + ring->wptr = RREG32_SOC15(VCN, i, mmUVD_RBC_RB_RPTR); + WREG32_SOC15(VCN, i, mmUVD_RBC_RB_WPTR, lower_32_bits(ring->wptr)); fw_shared->multi_queue.decode_queue_mode &= ~FW_QUEUE_RING_RESET; fw_shared->multi_queue.encode_generalpurpose_queue_mode |= FW_QUEUE_RING_RESET; ring = &adev->vcn.inst[i].ring_enc[0]; - WREG32_SOC15(UVD, i, mmUVD_RB_RPTR, lower_32_bits(ring->wptr)); - WREG32_SOC15(UVD, i, mmUVD_RB_WPTR, lower_32_bits(ring->wptr)); - WREG32_SOC15(UVD, i, mmUVD_RB_BASE_LO, ring->gpu_addr); - WREG32_SOC15(UVD, i, mmUVD_RB_BASE_HI, upper_32_bits(ring->gpu_addr)); - WREG32_SOC15(UVD, i, mmUVD_RB_SIZE, ring->ring_size / 4); + WREG32_SOC15(VCN, i, mmUVD_RB_RPTR, lower_32_bits(ring->wptr)); + WREG32_SOC15(VCN, i, mmUVD_RB_WPTR, lower_32_bits(ring->wptr)); + WREG32_SOC15(VCN, i, mmUVD_RB_BASE_LO, ring->gpu_addr); + WREG32_SOC15(VCN, i, mmUVD_RB_BASE_HI, upper_32_bits(ring->gpu_addr)); + WREG32_SOC15(VCN, i, mmUVD_RB_SIZE, ring->ring_size / 4); fw_shared->multi_queue.encode_generalpurpose_queue_mode &= ~FW_QUEUE_RING_RESET; fw_shared->multi_queue.encode_lowlatency_queue_mode |= FW_QUEUE_RING_RESET; ring = &adev->vcn.inst[i].ring_enc[1]; - WREG32_SOC15(UVD, i, mmUVD_RB_RPTR2, lower_32_bits(ring->wptr)); - WREG32_SOC15(UVD, i, mmUVD_RB_WPTR2, lower_32_bits(ring->wptr)); - WREG32_SOC15(UVD, i, mmUVD_RB_BASE_LO2, ring->gpu_addr); - WREG32_SOC15(UVD, i, mmUVD_RB_BASE_HI2, upper_32_bits(ring->gpu_addr)); - WREG32_SOC15(UVD, i, mmUVD_RB_SIZE2, ring->ring_size / 4); + WREG32_SOC15(VCN, i, mmUVD_RB_RPTR2, lower_32_bits(ring->wptr)); + WREG32_SOC15(VCN, i, mmUVD_RB_WPTR2, lower_32_bits(ring->wptr)); + WREG32_SOC15(VCN, i, mmUVD_RB_BASE_LO2, ring->gpu_addr); + WREG32_SOC15(VCN, i, mmUVD_RB_BASE_HI2, upper_32_bits(ring->gpu_addr)); + WREG32_SOC15(VCN, i, mmUVD_RB_SIZE2, ring->ring_size / 4); fw_shared->multi_queue.encode_lowlatency_queue_mode &= ~FW_QUEUE_RING_RESET; } @@ -1118,33 +1118,33 @@ static int vcn_v2_5_mmsch_start(struct amdgpu_device *adev, * 1, write to vce_mmsch_vf_ctx_addr_lo/hi register with GPU mc addr of * memory descriptor location */ - WREG32_SOC15(UVD, 0, mmMMSCH_VF_CTX_ADDR_LO, lower_32_bits(addr)); - WREG32_SOC15(UVD, 0, mmMMSCH_VF_CTX_ADDR_HI, upper_32_bits(addr)); + WREG32_SOC15(VCN, 0, mmMMSCH_VF_CTX_ADDR_LO, lower_32_bits(addr)); + WREG32_SOC15(VCN, 0, mmMMSCH_VF_CTX_ADDR_HI, upper_32_bits(addr)); /* 2, update vmid of descriptor */ - data = RREG32_SOC15(UVD, 0, mmMMSCH_VF_VMID); + data = RREG32_SOC15(VCN, 0, mmMMSCH_VF_VMID); data &= ~MMSCH_VF_VMID__VF_CTX_VMID_MASK; /* use domain0 for MM scheduler */ data |= (0 << MMSCH_VF_VMID__VF_CTX_VMID__SHIFT); - WREG32_SOC15(UVD, 0, mmMMSCH_VF_VMID, data); + WREG32_SOC15(VCN, 0, mmMMSCH_VF_VMID, data); /* 3, notify mmsch about the size of this descriptor */ - WREG32_SOC15(UVD, 0, mmMMSCH_VF_CTX_SIZE, size); + WREG32_SOC15(VCN, 0, mmMMSCH_VF_CTX_SIZE, size); /* 4, set resp to zero */ - WREG32_SOC15(UVD, 0, mmMMSCH_VF_MAILBOX_RESP, 0); + WREG32_SOC15(VCN, 0, mmMMSCH_VF_MAILBOX_RESP, 0); /* * 5, kick off the initialization and wait until * VCE_MMSCH_VF_MAILBOX_RESP becomes non-zero */ - WREG32_SOC15(UVD, 0, mmMMSCH_VF_MAILBOX_HOST, 0x10000001); + WREG32_SOC15(VCN, 0, mmMMSCH_VF_MAILBOX_HOST, 0x10000001); - data = RREG32_SOC15(UVD, 0, mmMMSCH_VF_MAILBOX_RESP); + data = RREG32_SOC15(VCN, 0, mmMMSCH_VF_MAILBOX_RESP); loop = 10; while ((data & 0x10000002) != 0x10000002) { udelay(100); - data = RREG32_SOC15(UVD, 0, mmMMSCH_VF_MAILBOX_RESP); + data = RREG32_SOC15(VCN, 0, mmMMSCH_VF_MAILBOX_RESP); loop--; if (!loop) break; @@ -1167,14 +1167,12 @@ static int vcn_v2_5_sriov_start(struct amdgpu_device *adev) uint32_t table_size = 0; struct mmsch_v1_0_cmd_direct_write direct_wt = { { 0 } }; struct mmsch_v1_0_cmd_direct_read_modify_write direct_rd_mod_wt = { { 0 } }; - struct mmsch_v1_0_cmd_direct_polling direct_poll = { { 0 } }; struct mmsch_v1_0_cmd_end end = { { 0 } }; uint32_t *init_table = adev->virt.mm_table.cpu_addr; struct mmsch_v1_1_init_header *header = (struct mmsch_v1_1_init_header *)init_table; direct_wt.cmd_header.command_type = MMSCH_COMMAND__DIRECT_REG_WRITE; direct_rd_mod_wt.cmd_header.command_type = MMSCH_COMMAND__DIRECT_REG_READ_MODIFY_WRITE; - direct_poll.cmd_header.command_type = MMSCH_COMMAND__DIRECT_REG_POLLING; end.cmd_header.command_type = MMSCH_COMMAND__END; header->version = MMSCH_VERSION; @@ -1189,93 +1187,93 @@ static int vcn_v2_5_sriov_start(struct amdgpu_device *adev) table_size = 0; MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT( - SOC15_REG_OFFSET(UVD, i, mmUVD_STATUS), + SOC15_REG_OFFSET(VCN, i, mmUVD_STATUS), ~UVD_STATUS__UVD_BUSY, UVD_STATUS__UVD_BUSY); size = AMDGPU_GPU_PAGE_ALIGN(adev->vcn.fw->size + 4); /* mc resume*/ if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) { MMSCH_V1_0_INSERT_DIRECT_WT( - SOC15_REG_OFFSET(UVD, i, + SOC15_REG_OFFSET(VCN, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW), adev->firmware.ucode[AMDGPU_UCODE_ID_VCN + i].tmr_mc_addr_lo); MMSCH_V1_0_INSERT_DIRECT_WT( - SOC15_REG_OFFSET(UVD, i, + SOC15_REG_OFFSET(VCN, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH), adev->firmware.ucode[AMDGPU_UCODE_ID_VCN + i].tmr_mc_addr_hi); offset = 0; MMSCH_V1_0_INSERT_DIRECT_WT( - SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CACHE_OFFSET0), 0); + SOC15_REG_OFFSET(VCN, i, mmUVD_VCPU_CACHE_OFFSET0), 0); } else { MMSCH_V1_0_INSERT_DIRECT_WT( - SOC15_REG_OFFSET(UVD, i, + SOC15_REG_OFFSET(VCN, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW), lower_32_bits(adev->vcn.inst[i].gpu_addr)); MMSCH_V1_0_INSERT_DIRECT_WT( - SOC15_REG_OFFSET(UVD, i, + SOC15_REG_OFFSET(VCN, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH), upper_32_bits(adev->vcn.inst[i].gpu_addr)); offset = size; MMSCH_V1_0_INSERT_DIRECT_WT( - SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CACHE_OFFSET0), + SOC15_REG_OFFSET(VCN, i, mmUVD_VCPU_CACHE_OFFSET0), AMDGPU_UVD_FIRMWARE_OFFSET >> 3); } MMSCH_V1_0_INSERT_DIRECT_WT( - SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CACHE_SIZE0), + SOC15_REG_OFFSET(VCN, i, mmUVD_VCPU_CACHE_SIZE0), size); MMSCH_V1_0_INSERT_DIRECT_WT( - SOC15_REG_OFFSET(UVD, i, + SOC15_REG_OFFSET(VCN, i, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_LOW), lower_32_bits(adev->vcn.inst[i].gpu_addr + offset)); MMSCH_V1_0_INSERT_DIRECT_WT( - SOC15_REG_OFFSET(UVD, i, + SOC15_REG_OFFSET(VCN, i, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_HIGH), upper_32_bits(adev->vcn.inst[i].gpu_addr + offset)); MMSCH_V1_0_INSERT_DIRECT_WT( - SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CACHE_OFFSET1), + SOC15_REG_OFFSET(VCN, i, mmUVD_VCPU_CACHE_OFFSET1), 0); MMSCH_V1_0_INSERT_DIRECT_WT( - SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CACHE_SIZE1), + SOC15_REG_OFFSET(VCN, i, mmUVD_VCPU_CACHE_SIZE1), AMDGPU_VCN_STACK_SIZE); MMSCH_V1_0_INSERT_DIRECT_WT( - SOC15_REG_OFFSET(UVD, i, + SOC15_REG_OFFSET(VCN, i, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_LOW), lower_32_bits(adev->vcn.inst[i].gpu_addr + offset + AMDGPU_VCN_STACK_SIZE)); MMSCH_V1_0_INSERT_DIRECT_WT( - SOC15_REG_OFFSET(UVD, i, + SOC15_REG_OFFSET(VCN, i, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_HIGH), upper_32_bits(adev->vcn.inst[i].gpu_addr + offset + AMDGPU_VCN_STACK_SIZE)); MMSCH_V1_0_INSERT_DIRECT_WT( - SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CACHE_OFFSET2), + SOC15_REG_OFFSET(VCN, i, mmUVD_VCPU_CACHE_OFFSET2), 0); MMSCH_V1_0_INSERT_DIRECT_WT( - SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CACHE_SIZE2), + SOC15_REG_OFFSET(VCN, i, mmUVD_VCPU_CACHE_SIZE2), AMDGPU_VCN_CONTEXT_SIZE); ring = &adev->vcn.inst[i].ring_enc[0]; ring->wptr = 0; MMSCH_V1_0_INSERT_DIRECT_WT( - SOC15_REG_OFFSET(UVD, i, mmUVD_RB_BASE_LO), + SOC15_REG_OFFSET(VCN, i, mmUVD_RB_BASE_LO), lower_32_bits(ring->gpu_addr)); MMSCH_V1_0_INSERT_DIRECT_WT( - SOC15_REG_OFFSET(UVD, i, mmUVD_RB_BASE_HI), + SOC15_REG_OFFSET(VCN, i, mmUVD_RB_BASE_HI), upper_32_bits(ring->gpu_addr)); MMSCH_V1_0_INSERT_DIRECT_WT( - SOC15_REG_OFFSET(UVD, i, mmUVD_RB_SIZE), + SOC15_REG_OFFSET(VCN, i, mmUVD_RB_SIZE), ring->ring_size / 4); ring = &adev->vcn.inst[i].ring_dec; ring->wptr = 0; MMSCH_V1_0_INSERT_DIRECT_WT( - SOC15_REG_OFFSET(UVD, i, + SOC15_REG_OFFSET(VCN, i, mmUVD_LMI_RBC_RB_64BIT_BAR_LOW), lower_32_bits(ring->gpu_addr)); MMSCH_V1_0_INSERT_DIRECT_WT( - SOC15_REG_OFFSET(UVD, i, + SOC15_REG_OFFSET(VCN, i, mmUVD_LMI_RBC_RB_64BIT_BAR_HIGH), upper_32_bits(ring->gpu_addr)); @@ -1287,7 +1285,7 @@ static int vcn_v2_5_sriov_start(struct amdgpu_device *adev) tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_NO_UPDATE, 1); tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_RPTR_WR_EN, 1); MMSCH_V1_0_INSERT_DIRECT_WT( - SOC15_REG_OFFSET(UVD, i, mmUVD_RBC_RB_CNTL), tmp); + SOC15_REG_OFFSET(VCN, i, mmUVD_RBC_RB_CNTL), tmp); /* add end packet */ memcpy((void *)init_table, &end, sizeof(struct mmsch_v1_0_cmd_end)); @@ -1308,24 +1306,24 @@ static int vcn_v2_5_stop_dpg_mode(struct amdgpu_device *adev, int inst_idx) uint32_t tmp; /* Wait for power status to be 1 */ - SOC15_WAIT_ON_RREG(UVD, inst_idx, mmUVD_POWER_STATUS, 1, + SOC15_WAIT_ON_RREG(VCN, inst_idx, mmUVD_POWER_STATUS, 1, UVD_POWER_STATUS__UVD_POWER_STATUS_MASK, ret_code); /* wait for read ptr to be equal to write ptr */ - tmp = RREG32_SOC15(UVD, inst_idx, mmUVD_RB_WPTR); - SOC15_WAIT_ON_RREG(UVD, inst_idx, mmUVD_RB_RPTR, tmp, 0xFFFFFFFF, ret_code); + tmp = RREG32_SOC15(VCN, inst_idx, mmUVD_RB_WPTR); + SOC15_WAIT_ON_RREG(VCN, inst_idx, mmUVD_RB_RPTR, tmp, 0xFFFFFFFF, ret_code); - tmp = RREG32_SOC15(UVD, inst_idx, mmUVD_RB_WPTR2); - SOC15_WAIT_ON_RREG(UVD, inst_idx, mmUVD_RB_RPTR2, tmp, 0xFFFFFFFF, ret_code); + tmp = RREG32_SOC15(VCN, inst_idx, mmUVD_RB_WPTR2); + SOC15_WAIT_ON_RREG(VCN, inst_idx, mmUVD_RB_RPTR2, tmp, 0xFFFFFFFF, ret_code); - tmp = RREG32_SOC15(UVD, inst_idx, mmUVD_RBC_RB_WPTR) & 0x7FFFFFFF; - SOC15_WAIT_ON_RREG(UVD, inst_idx, mmUVD_RBC_RB_RPTR, tmp, 0xFFFFFFFF, ret_code); + tmp = RREG32_SOC15(VCN, inst_idx, mmUVD_RBC_RB_WPTR) & 0x7FFFFFFF; + SOC15_WAIT_ON_RREG(VCN, inst_idx, mmUVD_RBC_RB_RPTR, tmp, 0xFFFFFFFF, ret_code); - SOC15_WAIT_ON_RREG(UVD, inst_idx, mmUVD_POWER_STATUS, 1, + SOC15_WAIT_ON_RREG(VCN, inst_idx, mmUVD_POWER_STATUS, 1, UVD_POWER_STATUS__UVD_POWER_STATUS_MASK, ret_code); /* disable dynamic power gating mode */ - WREG32_P(SOC15_REG_OFFSET(UVD, inst_idx, mmUVD_POWER_STATUS), 0, + WREG32_P(SOC15_REG_OFFSET(VCN, inst_idx, mmUVD_POWER_STATUS), 0, ~UVD_POWER_STATUS__UVD_PG_MODE_MASK); return 0; @@ -1369,17 +1367,17 @@ static int vcn_v2_5_stop(struct amdgpu_device *adev) return r; /* block VCPU register access */ - WREG32_P(SOC15_REG_OFFSET(UVD, i, mmUVD_RB_ARB_CTRL), + WREG32_P(SOC15_REG_OFFSET(VCN, i, mmUVD_RB_ARB_CTRL), UVD_RB_ARB_CTRL__VCPU_DIS_MASK, ~UVD_RB_ARB_CTRL__VCPU_DIS_MASK); /* reset VCPU */ - WREG32_P(SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CNTL), + WREG32_P(SOC15_REG_OFFSET(VCN, i, mmUVD_VCPU_CNTL), UVD_VCPU_CNTL__BLK_RST_MASK, ~UVD_VCPU_CNTL__BLK_RST_MASK); /* disable VCPU clock */ - WREG32_P(SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CNTL), 0, + WREG32_P(SOC15_REG_OFFSET(VCN, i, mmUVD_VCPU_CNTL), 0, ~(UVD_VCPU_CNTL__CLK_EN_MASK)); /* clear status */ @@ -1388,7 +1386,7 @@ static int vcn_v2_5_stop(struct amdgpu_device *adev) vcn_v2_5_enable_clock_gating(adev); /* enable register anti-hang mechanism */ - WREG32_P(SOC15_REG_OFFSET(UVD, i, mmUVD_POWER_STATUS), + WREG32_P(SOC15_REG_OFFSET(VCN, i, mmUVD_POWER_STATUS), UVD_POWER_STATUS__UVD_POWER_STATUS_MASK, ~UVD_POWER_STATUS__UVD_POWER_STATUS_MASK); } @@ -1410,11 +1408,11 @@ static int vcn_v2_5_pause_dpg_mode(struct amdgpu_device *adev, if (adev->vcn.inst[inst_idx].pause_state.fw_based != new_state->fw_based) { DRM_DEBUG("dpg pause state changed %d -> %d", adev->vcn.inst[inst_idx].pause_state.fw_based, new_state->fw_based); - reg_data = RREG32_SOC15(UVD, inst_idx, mmUVD_DPG_PAUSE) & + reg_data = RREG32_SOC15(VCN, inst_idx, mmUVD_DPG_PAUSE) & (~UVD_DPG_PAUSE__NJ_PAUSE_DPG_ACK_MASK); if (new_state->fw_based == VCN_DPG_STATE__PAUSE) { - SOC15_WAIT_ON_RREG(UVD, inst_idx, mmUVD_POWER_STATUS, 0x1, + SOC15_WAIT_ON_RREG(VCN, inst_idx, mmUVD_POWER_STATUS, 0x1, UVD_POWER_STATUS__UVD_POWER_STATUS_MASK, ret_code); if (!ret_code) { @@ -1422,15 +1420,15 @@ static int vcn_v2_5_pause_dpg_mode(struct amdgpu_device *adev, /* pause DPG */ reg_data |= UVD_DPG_PAUSE__NJ_PAUSE_DPG_REQ_MASK; - WREG32_SOC15(UVD, inst_idx, mmUVD_DPG_PAUSE, reg_data); + WREG32_SOC15(VCN, inst_idx, mmUVD_DPG_PAUSE, reg_data); /* wait for ACK */ - SOC15_WAIT_ON_RREG(UVD, inst_idx, mmUVD_DPG_PAUSE, + SOC15_WAIT_ON_RREG(VCN, inst_idx, mmUVD_DPG_PAUSE, UVD_DPG_PAUSE__NJ_PAUSE_DPG_ACK_MASK, UVD_DPG_PAUSE__NJ_PAUSE_DPG_ACK_MASK, ret_code); /* Stall DPG before WPTR/RPTR reset */ - WREG32_P(SOC15_REG_OFFSET(UVD, inst_idx, mmUVD_POWER_STATUS), + WREG32_P(SOC15_REG_OFFSET(VCN, inst_idx, mmUVD_POWER_STATUS), UVD_POWER_STATUS__STALL_DPG_POWER_UP_MASK, ~UVD_POWER_STATUS__STALL_DPG_POWER_UP_MASK); @@ -1438,39 +1436,39 @@ static int vcn_v2_5_pause_dpg_mode(struct amdgpu_device *adev, fw_shared->multi_queue.encode_generalpurpose_queue_mode |= FW_QUEUE_RING_RESET; ring = &adev->vcn.inst[inst_idx].ring_enc[0]; ring->wptr = 0; - WREG32_SOC15(UVD, inst_idx, mmUVD_RB_BASE_LO, ring->gpu_addr); - WREG32_SOC15(UVD, inst_idx, mmUVD_RB_BASE_HI, upper_32_bits(ring->gpu_addr)); - WREG32_SOC15(UVD, inst_idx, mmUVD_RB_SIZE, ring->ring_size / 4); - WREG32_SOC15(UVD, inst_idx, mmUVD_RB_RPTR, lower_32_bits(ring->wptr)); - WREG32_SOC15(UVD, inst_idx, mmUVD_RB_WPTR, lower_32_bits(ring->wptr)); + WREG32_SOC15(VCN, inst_idx, mmUVD_RB_BASE_LO, ring->gpu_addr); + WREG32_SOC15(VCN, inst_idx, mmUVD_RB_BASE_HI, upper_32_bits(ring->gpu_addr)); + WREG32_SOC15(VCN, inst_idx, mmUVD_RB_SIZE, ring->ring_size / 4); + WREG32_SOC15(VCN, inst_idx, mmUVD_RB_RPTR, lower_32_bits(ring->wptr)); + WREG32_SOC15(VCN, inst_idx, mmUVD_RB_WPTR, lower_32_bits(ring->wptr)); fw_shared->multi_queue.encode_generalpurpose_queue_mode &= ~FW_QUEUE_RING_RESET; fw_shared->multi_queue.encode_lowlatency_queue_mode |= FW_QUEUE_RING_RESET; ring = &adev->vcn.inst[inst_idx].ring_enc[1]; ring->wptr = 0; - WREG32_SOC15(UVD, inst_idx, mmUVD_RB_BASE_LO2, ring->gpu_addr); - WREG32_SOC15(UVD, inst_idx, mmUVD_RB_BASE_HI2, upper_32_bits(ring->gpu_addr)); - WREG32_SOC15(UVD, inst_idx, mmUVD_RB_SIZE2, ring->ring_size / 4); - WREG32_SOC15(UVD, inst_idx, mmUVD_RB_RPTR2, lower_32_bits(ring->wptr)); - WREG32_SOC15(UVD, inst_idx, mmUVD_RB_WPTR2, lower_32_bits(ring->wptr)); + WREG32_SOC15(VCN, inst_idx, mmUVD_RB_BASE_LO2, ring->gpu_addr); + WREG32_SOC15(VCN, inst_idx, mmUVD_RB_BASE_HI2, upper_32_bits(ring->gpu_addr)); + WREG32_SOC15(VCN, inst_idx, mmUVD_RB_SIZE2, ring->ring_size / 4); + WREG32_SOC15(VCN, inst_idx, mmUVD_RB_RPTR2, lower_32_bits(ring->wptr)); + WREG32_SOC15(VCN, inst_idx, mmUVD_RB_WPTR2, lower_32_bits(ring->wptr)); fw_shared->multi_queue.encode_lowlatency_queue_mode &= ~FW_QUEUE_RING_RESET; fw_shared->multi_queue.decode_queue_mode |= FW_QUEUE_RING_RESET; - WREG32_SOC15(UVD, inst_idx, mmUVD_RBC_RB_WPTR, - RREG32_SOC15(UVD, inst_idx, mmUVD_SCRATCH2) & 0x7FFFFFFF); + WREG32_SOC15(VCN, inst_idx, mmUVD_RBC_RB_WPTR, + RREG32_SOC15(VCN, inst_idx, mmUVD_SCRATCH2) & 0x7FFFFFFF); fw_shared->multi_queue.decode_queue_mode &= ~FW_QUEUE_RING_RESET; /* Unstall DPG */ - WREG32_P(SOC15_REG_OFFSET(UVD, inst_idx, mmUVD_POWER_STATUS), + WREG32_P(SOC15_REG_OFFSET(VCN, inst_idx, mmUVD_POWER_STATUS), 0, ~UVD_POWER_STATUS__STALL_DPG_POWER_UP_MASK); - SOC15_WAIT_ON_RREG(UVD, inst_idx, mmUVD_POWER_STATUS, + SOC15_WAIT_ON_RREG(VCN, inst_idx, mmUVD_POWER_STATUS, UVD_PGFSM_CONFIG__UVDM_UVDU_PWR_ON, UVD_POWER_STATUS__UVD_POWER_STATUS_MASK, ret_code); } } else { reg_data &= ~UVD_DPG_PAUSE__NJ_PAUSE_DPG_REQ_MASK; - WREG32_SOC15(UVD, inst_idx, mmUVD_DPG_PAUSE, reg_data); - SOC15_WAIT_ON_RREG(UVD, inst_idx, mmUVD_POWER_STATUS, 0x1, + WREG32_SOC15(VCN, inst_idx, mmUVD_DPG_PAUSE, reg_data); + SOC15_WAIT_ON_RREG(VCN, inst_idx, mmUVD_POWER_STATUS, 0x1, UVD_POWER_STATUS__UVD_POWER_STATUS_MASK, ret_code); } adev->vcn.inst[inst_idx].pause_state.fw_based = new_state->fw_based; @@ -1490,7 +1488,7 @@ static uint64_t vcn_v2_5_dec_ring_get_rptr(struct amdgpu_ring *ring) { struct amdgpu_device *adev = ring->adev; - return RREG32_SOC15(UVD, ring->me, mmUVD_RBC_RB_RPTR); + return RREG32_SOC15(VCN, ring->me, mmUVD_RBC_RB_RPTR); } /** @@ -1507,7 +1505,7 @@ static uint64_t vcn_v2_5_dec_ring_get_wptr(struct amdgpu_ring *ring) if (ring->use_doorbell) return adev->wb.wb[ring->wptr_offs]; else - return RREG32_SOC15(UVD, ring->me, mmUVD_RBC_RB_WPTR); + return RREG32_SOC15(VCN, ring->me, mmUVD_RBC_RB_WPTR); } /** @@ -1522,14 +1520,14 @@ static void vcn_v2_5_dec_ring_set_wptr(struct amdgpu_ring *ring) struct amdgpu_device *adev = ring->adev; if (adev->pg_flags & AMD_PG_SUPPORT_VCN_DPG) - WREG32_SOC15(UVD, ring->me, mmUVD_SCRATCH2, + WREG32_SOC15(VCN, ring->me, mmUVD_SCRATCH2, lower_32_bits(ring->wptr) | 0x80000000); if (ring->use_doorbell) { adev->wb.wb[ring->wptr_offs] = lower_32_bits(ring->wptr); WDOORBELL32(ring->doorbell_index, lower_32_bits(ring->wptr)); } else { - WREG32_SOC15(UVD, ring->me, mmUVD_RBC_RB_WPTR, lower_32_bits(ring->wptr)); + WREG32_SOC15(VCN, ring->me, mmUVD_RBC_RB_WPTR, lower_32_bits(ring->wptr)); } } @@ -1575,9 +1573,9 @@ static uint64_t vcn_v2_5_enc_ring_get_rptr(struct amdgpu_ring *ring) struct amdgpu_device *adev = ring->adev; if (ring == &adev->vcn.inst[ring->me].ring_enc[0]) - return RREG32_SOC15(UVD, ring->me, mmUVD_RB_RPTR); + return RREG32_SOC15(VCN, ring->me, mmUVD_RB_RPTR); else - return RREG32_SOC15(UVD, ring->me, mmUVD_RB_RPTR2); + return RREG32_SOC15(VCN, ring->me, mmUVD_RB_RPTR2); } /** @@ -1595,12 +1593,12 @@ static uint64_t vcn_v2_5_enc_ring_get_wptr(struct amdgpu_ring *ring) if (ring->use_doorbell) return adev->wb.wb[ring->wptr_offs]; else - return RREG32_SOC15(UVD, ring->me, mmUVD_RB_WPTR); + return RREG32_SOC15(VCN, ring->me, mmUVD_RB_WPTR); } else { if (ring->use_doorbell) return adev->wb.wb[ring->wptr_offs]; else - return RREG32_SOC15(UVD, ring->me, mmUVD_RB_WPTR2); + return RREG32_SOC15(VCN, ring->me, mmUVD_RB_WPTR2); } } @@ -1620,14 +1618,14 @@ static void vcn_v2_5_enc_ring_set_wptr(struct amdgpu_ring *ring) adev->wb.wb[ring->wptr_offs] = lower_32_bits(ring->wptr); WDOORBELL32(ring->doorbell_index, lower_32_bits(ring->wptr)); } else { - WREG32_SOC15(UVD, ring->me, mmUVD_RB_WPTR, lower_32_bits(ring->wptr)); + WREG32_SOC15(VCN, ring->me, mmUVD_RB_WPTR, lower_32_bits(ring->wptr)); } } else { if (ring->use_doorbell) { adev->wb.wb[ring->wptr_offs] = lower_32_bits(ring->wptr); WDOORBELL32(ring->doorbell_index, lower_32_bits(ring->wptr)); } else { - WREG32_SOC15(UVD, ring->me, mmUVD_RB_WPTR2, lower_32_bits(ring->wptr)); + WREG32_SOC15(VCN, ring->me, mmUVD_RB_WPTR2, lower_32_bits(ring->wptr)); } } } diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c index ff47b1f69b68..cf0017f4d9d5 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c @@ -1323,6 +1323,10 @@ static int kfd_ioctl_alloc_memory_of_gpu(struct file *filep, goto err_free; } + /* Update the VRAM usage count */ + if (flags & KFD_IOC_ALLOC_MEM_FLAGS_VRAM) + WRITE_ONCE(pdd->vram_usage, pdd->vram_usage + args->size); + mutex_unlock(&p->mutex); args->handle = MAKE_HANDLE(args->gpu_id, idr_handle); @@ -1338,7 +1342,7 @@ static int kfd_ioctl_alloc_memory_of_gpu(struct file *filep, return 0; err_free: - amdgpu_amdkfd_gpuvm_free_memory_of_gpu(dev->kgd, (struct kgd_mem *)mem); + amdgpu_amdkfd_gpuvm_free_memory_of_gpu(dev->kgd, (struct kgd_mem *)mem, NULL); err_unlock: mutex_unlock(&p->mutex); return err; @@ -1352,6 +1356,7 @@ static int kfd_ioctl_free_memory_of_gpu(struct file *filep, void *mem; struct kfd_dev *dev; int ret; + uint64_t size = 0; dev = kfd_device_by_id(GET_GPU_ID(args->handle)); if (!dev) @@ -1374,7 +1379,7 @@ static int kfd_ioctl_free_memory_of_gpu(struct file *filep, } ret = amdgpu_amdkfd_gpuvm_free_memory_of_gpu(dev->kgd, - (struct kgd_mem *)mem); + (struct kgd_mem *)mem, &size); /* If freeing the buffer failed, leave the handle in place for * clean-up during process tear-down. @@ -1383,6 +1388,8 @@ static int kfd_ioctl_free_memory_of_gpu(struct file *filep, kfd_process_device_remove_obj_handle( pdd, GET_IDR_HANDLE(args->handle)); + WRITE_ONCE(pdd->vram_usage, pdd->vram_usage - size); + err_unlock: mutex_unlock(&p->mutex); return ret; @@ -1727,7 +1734,7 @@ static int kfd_ioctl_import_dmabuf(struct file *filep, return 0; err_free: - amdgpu_amdkfd_gpuvm_free_memory_of_gpu(dev->kgd, (struct kgd_mem *)mem); + amdgpu_amdkfd_gpuvm_free_memory_of_gpu(dev->kgd, (struct kgd_mem *)mem, NULL); err_unlock: mutex_unlock(&p->mutex); return r; 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 ae9547791813..e9c4867abeff 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c @@ -1089,7 +1089,9 @@ static int set_sched_resources(struct device_queue_manager *dqm) break; } - res.queue_mask |= (1ull << i); + res.queue_mask |= 1ull + << amdgpu_queue_mask_bit_to_set_resource_bit( + (struct amdgpu_device *)dqm->dev->kgd, i); } res.gws_mask = ~0ull; res.oac_mask = res.gds_heap_base = res.gds_heap_size = 0; diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h index d48b33449267..cde5e4c7caa1 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h @@ -629,6 +629,8 @@ enum kfd_pdd_bound { PDD_BOUND_SUSPENDED, }; +#define MAX_VRAM_FILENAME_LEN 11 + /* Data that is per-process-per device. */ struct kfd_process_device { /* @@ -671,6 +673,11 @@ struct kfd_process_device { /* Is this process/pasid bound to this device? (amd_iommu_bind_pasid) */ enum kfd_pdd_bound bound; + + /* VRAM usage */ + uint64_t vram_usage; + struct attribute attr_vram; + char vram_filename[MAX_VRAM_FILENAME_LEN]; }; #define qpd_to_pdd(x) container_of(x, struct kfd_process_device, qpd) diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process.c b/drivers/gpu/drm/amd/amdkfd/kfd_process.c index 82b4c5a9382a..d27221ddcdeb 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_process.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_process.c @@ -79,18 +79,22 @@ static struct kfd_procfs_tree procfs; static ssize_t kfd_procfs_show(struct kobject *kobj, struct attribute *attr, char *buffer) { - int val = 0; - if (strcmp(attr->name, "pasid") == 0) { struct kfd_process *p = container_of(attr, struct kfd_process, attr_pasid); - val = p->pasid; + + return snprintf(buffer, PAGE_SIZE, "%d\n", p->pasid); + } else if (strncmp(attr->name, "vram_", 5) == 0) { + struct kfd_process_device *pdd = container_of(attr, struct kfd_process_device, + attr_vram); + if (pdd) + return snprintf(buffer, PAGE_SIZE, "%llu\n", READ_ONCE(pdd->vram_usage)); } else { pr_err("Invalid attribute"); return -EINVAL; } - return snprintf(buffer, PAGE_SIZE, "%d\n", val); + return 0; } static void kfd_procfs_kobj_release(struct kobject *kobj) @@ -206,6 +210,34 @@ int kfd_procfs_add_queue(struct queue *q) return 0; } +int kfd_procfs_add_vram_usage(struct kfd_process *p) +{ + int ret = 0; + struct kfd_process_device *pdd; + + if (!p) + return -EINVAL; + + if (!p->kobj) + return -EFAULT; + + /* Create proc/<pid>/vram_<gpuid> file for each GPU */ + list_for_each_entry(pdd, &p->per_device_data, per_device_list) { + snprintf(pdd->vram_filename, MAX_VRAM_FILENAME_LEN, "vram_%u", + pdd->dev->id); + pdd->attr_vram.name = pdd->vram_filename; + pdd->attr_vram.mode = KFD_SYSFS_FILE_MODE; + sysfs_attr_init(&pdd->attr_vram); + ret = sysfs_create_file(p->kobj, &pdd->attr_vram); + if (ret) + pr_warn("Creating vram usage for gpu id %d failed", + (int)pdd->dev->id); + } + + return ret; +} + + void kfd_procfs_del_queue(struct queue *q) { if (!q) @@ -248,7 +280,7 @@ static void kfd_process_free_gpuvm(struct kgd_mem *mem, struct kfd_dev *dev = pdd->dev; amdgpu_amdkfd_gpuvm_unmap_memory_from_gpu(dev->kgd, mem, pdd->vm); - amdgpu_amdkfd_gpuvm_free_memory_of_gpu(dev->kgd, mem); + amdgpu_amdkfd_gpuvm_free_memory_of_gpu(dev->kgd, mem, NULL); } /* kfd_process_alloc_gpuvm - Allocate GPU VM for the KFD process @@ -312,7 +344,7 @@ sync_memory_failed: return err; err_map_mem: - amdgpu_amdkfd_gpuvm_free_memory_of_gpu(kdev->kgd, mem); + amdgpu_amdkfd_gpuvm_free_memory_of_gpu(kdev->kgd, mem, NULL); err_alloc_mem: *kptr = NULL; return err; @@ -411,6 +443,11 @@ struct kfd_process *kfd_create_process(struct file *filep) process->kobj); if (!process->kobj_queues) pr_warn("Creating KFD proc/queues folder failed"); + + ret = kfd_procfs_add_vram_usage(process); + if (ret) + pr_warn("Creating vram usage file for pid %d failed", + (int)process->lead_thread->pid); } out: if (!IS_ERR(process)) @@ -488,7 +525,7 @@ static void kfd_process_device_free_bos(struct kfd_process_device *pdd) peer_pdd->dev->kgd, mem, peer_pdd->vm); } - amdgpu_amdkfd_gpuvm_free_memory_of_gpu(pdd->dev->kgd, mem); + amdgpu_amdkfd_gpuvm_free_memory_of_gpu(pdd->dev->kgd, mem, NULL); kfd_process_device_remove_obj_handle(pdd, id); } } @@ -551,6 +588,7 @@ static void kfd_process_wq_release(struct work_struct *work) { struct kfd_process *p = container_of(work, struct kfd_process, release_work); + struct kfd_process_device *pdd; /* Remove the procfs files */ if (p->kobj) { @@ -558,6 +596,10 @@ static void kfd_process_wq_release(struct work_struct *work) kobject_del(p->kobj_queues); kobject_put(p->kobj_queues); p->kobj_queues = NULL; + + list_for_each_entry(pdd, &p->per_device_data, per_device_list) + sysfs_remove_file(p->kobj, &pdd->attr_vram); + kobject_del(p->kobj); kobject_put(p->kobj); p->kobj = NULL; @@ -863,6 +905,7 @@ struct kfd_process_device *kfd_create_process_device_data(struct kfd_dev *dev, pdd->bound = PDD_UNBOUND; pdd->already_dequeued = false; pdd->runtime_inuse = false; + pdd->vram_usage = 0; list_add(&pdd->per_device_list, &p->per_device_data); /* Init idr used for memory handle translation */ @@ -1079,7 +1122,7 @@ struct kfd_process *kfd_lookup_process_by_mm(const struct mm_struct *mm) return p; } -/* process_evict_queues - Evict all user queues of a process +/* kfd_process_evict_queues - Evict all user queues of a process * * Eviction is reference-counted per process-device. This means multiple * evictions from different sources can be nested safely. @@ -1119,7 +1162,7 @@ fail: return r; } -/* process_restore_queues - Restore all user queues of a process */ +/* kfd_process_restore_queues - Restore all user queues of a process */ int kfd_process_restore_queues(struct kfd_process *p) { struct kfd_process_device *pdd; diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c index bc4a22df12d7..bb77f7af2b6d 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c @@ -478,6 +478,8 @@ static ssize_t node_show(struct kobject *kobj, struct attribute *attr, dev->node_props.device_id); sysfs_show_32bit_prop(buffer, "location_id", dev->node_props.location_id); + sysfs_show_32bit_prop(buffer, "domain", + dev->node_props.domain); sysfs_show_32bit_prop(buffer, "drm_render_minor", dev->node_props.drm_render_minor); sysfs_show_64bit_prop(buffer, "hive_id", @@ -1306,6 +1308,7 @@ int kfd_topology_add_device(struct kfd_dev *gpu) HSA_CAP_ASIC_REVISION_SHIFT) & HSA_CAP_ASIC_REVISION_MASK); dev->node_props.location_id = pci_dev_id(gpu->pdev); + dev->node_props.domain = pci_domain_nr(gpu->pdev->bus); dev->node_props.max_engine_clk_fcompute = amdgpu_amdkfd_get_max_engine_clock_in_mhz(dev->gpu->kgd); dev->node_props.max_engine_clk_ccompute = diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_topology.h b/drivers/gpu/drm/amd/amdkfd/kfd_topology.h index 0c51bd3dcd59..326d9b26b7aa 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_topology.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_topology.h @@ -80,6 +80,7 @@ struct kfd_node_properties { uint32_t vendor_id; uint32_t device_id; uint32_t location_id; + uint32_t domain; uint32_t max_engine_clk_fcompute; uint32_t max_engine_clk_ccompute; int32_t drm_render_minor; 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 71309ee3aca3..48f2b3710e7c 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -30,7 +30,7 @@ #include "dc.h" #include "dc/inc/core_types.h" #include "dal_asic_id.h" -#include "dmub/inc/dmub_srv.h" +#include "dmub/dmub_srv.h" #include "dc/inc/hw/dmcu.h" #include "dc/inc/hw/abm.h" #include "dc/dc_dmub_srv.h" @@ -441,7 +441,7 @@ static void dm_vupdate_high_irq(void *interrupt_params) /** * dm_crtc_high_irq() - Handles CRTC interrupt - * @interrupt_params: ignored + * @interrupt_params: used for determining the CRTC instance * * Handles the CRTC/VSYNC interrupt by notfying DRM's VBLANK * event handler. @@ -455,70 +455,6 @@ static void dm_crtc_high_irq(void *interrupt_params) unsigned long flags; acrtc = get_crtc_by_otg_inst(adev, irq_params->irq_src - IRQ_TYPE_VBLANK); - - if (acrtc) { - acrtc_state = to_dm_crtc_state(acrtc->base.state); - - DRM_DEBUG_VBL("crtc:%d, vupdate-vrr:%d\n", - acrtc->crtc_id, - amdgpu_dm_vrr_active(acrtc_state)); - - /* Core vblank handling at start of front-porch is only possible - * in non-vrr mode, as only there vblank timestamping will give - * valid results while done in front-porch. Otherwise defer it - * to dm_vupdate_high_irq after end of front-porch. - */ - if (!amdgpu_dm_vrr_active(acrtc_state)) - drm_crtc_handle_vblank(&acrtc->base); - - /* Following stuff must happen at start of vblank, for crc - * computation and below-the-range btr support in vrr mode. - */ - amdgpu_dm_crtc_handle_crc_irq(&acrtc->base); - - if (acrtc_state->stream && adev->family >= AMDGPU_FAMILY_AI && - acrtc_state->vrr_params.supported && - acrtc_state->freesync_config.state == VRR_STATE_ACTIVE_VARIABLE) { - spin_lock_irqsave(&adev->ddev->event_lock, flags); - mod_freesync_handle_v_update( - adev->dm.freesync_module, - acrtc_state->stream, - &acrtc_state->vrr_params); - - dc_stream_adjust_vmin_vmax( - adev->dm.dc, - acrtc_state->stream, - &acrtc_state->vrr_params.adjust); - spin_unlock_irqrestore(&adev->ddev->event_lock, flags); - } - } -} - -#if defined(CONFIG_DRM_AMD_DC_DCN) -/** - * dm_dcn_crtc_high_irq() - Handles VStartup interrupt for DCN generation ASICs - * @interrupt params - interrupt parameters - * - * Notify DRM's vblank event handler at VSTARTUP - * - * Unlike DCE hardware, we trigger the handler at VSTARTUP. at which: - * * We are close enough to VUPDATE - the point of no return for hw - * * We are in the fixed portion of variable front porch when vrr is enabled - * * We are before VUPDATE, where double-buffered vrr registers are swapped - * - * It is therefore the correct place to signal vblank, send user flip events, - * and update VRR. - */ -static void dm_dcn_crtc_high_irq(void *interrupt_params) -{ - struct common_irq_params *irq_params = interrupt_params; - struct amdgpu_device *adev = irq_params->adev; - struct amdgpu_crtc *acrtc; - struct dm_crtc_state *acrtc_state; - unsigned long flags; - - acrtc = get_crtc_by_otg_inst(adev, irq_params->irq_src - IRQ_TYPE_VBLANK); - if (!acrtc) return; @@ -528,22 +464,35 @@ static void dm_dcn_crtc_high_irq(void *interrupt_params) amdgpu_dm_vrr_active(acrtc_state), acrtc_state->active_planes); + /** + * Core vblank handling at start of front-porch is only possible + * in non-vrr mode, as only there vblank timestamping will give + * valid results while done in front-porch. Otherwise defer it + * to dm_vupdate_high_irq after end of front-porch. + */ + if (!amdgpu_dm_vrr_active(acrtc_state)) + drm_crtc_handle_vblank(&acrtc->base); + + /** + * Following stuff must happen at start of vblank, for crc + * computation and below-the-range btr support in vrr mode. + */ amdgpu_dm_crtc_handle_crc_irq(&acrtc->base); - drm_crtc_handle_vblank(&acrtc->base); + + /* BTR updates need to happen before VUPDATE on Vega and above. */ + if (adev->family < AMDGPU_FAMILY_AI) + return; spin_lock_irqsave(&adev->ddev->event_lock, flags); - if (acrtc_state->vrr_params.supported && + if (acrtc_state->stream && acrtc_state->vrr_params.supported && acrtc_state->freesync_config.state == VRR_STATE_ACTIVE_VARIABLE) { - mod_freesync_handle_v_update( - adev->dm.freesync_module, - acrtc_state->stream, - &acrtc_state->vrr_params); + mod_freesync_handle_v_update(adev->dm.freesync_module, + acrtc_state->stream, + &acrtc_state->vrr_params); - dc_stream_adjust_vmin_vmax( - adev->dm.dc, - acrtc_state->stream, - &acrtc_state->vrr_params.adjust); + dc_stream_adjust_vmin_vmax(adev->dm.dc, acrtc_state->stream, + &acrtc_state->vrr_params.adjust); } /* @@ -556,7 +505,8 @@ static void dm_dcn_crtc_high_irq(void *interrupt_params) * avoid race conditions between flip programming and completion, * which could cause too early flip completion events. */ - if (acrtc->pflip_status == AMDGPU_FLIP_SUBMITTED && + if (adev->family >= AMDGPU_FAMILY_RV && + acrtc->pflip_status == AMDGPU_FLIP_SUBMITTED && acrtc_state->active_planes == 0) { if (acrtc->event) { drm_crtc_send_vblank_event(&acrtc->base, acrtc->event); @@ -568,7 +518,6 @@ static void dm_dcn_crtc_high_irq(void *interrupt_params) spin_unlock_irqrestore(&adev->ddev->event_lock, flags); } -#endif static int dm_set_clockgating_state(void *handle, enum amd_clockgating_state state) @@ -1389,9 +1338,14 @@ static int dm_late_init(void *handle) struct dmcu_iram_parameters params; unsigned int linear_lut[16]; int i; - struct dmcu *dmcu = adev->dm.dc->res_pool->dmcu; + struct dmcu *dmcu = NULL; bool ret = false; + if (!adev->dm.fw_dmcu) + return detect_mst_link_for_all_connectors(adev->ddev); + + dmcu = adev->dm.dc->res_pool->dmcu; + for (i = 0; i < 16; i++) linear_lut[i] = 0xFFFF * i / 15; @@ -1571,7 +1525,6 @@ static int dm_suspend(void *handle) { struct amdgpu_device *adev = handle; struct amdgpu_display_manager *dm = &adev->dm; - int ret = 0; WARN_ON(adev->dm.cached_state); adev->dm.cached_state = drm_atomic_helper_suspend(adev->ddev); @@ -1583,7 +1536,7 @@ static int dm_suspend(void *handle) dc_set_power_state(dm->dc, DC_ACPI_CM_POWER_STATE_D3); - return ret; + return 0; } static struct amdgpu_dm_connector * @@ -2013,17 +1966,22 @@ void amdgpu_dm_update_connector_after_detect( dc_sink_retain(aconnector->dc_sink); if (sink->dc_edid.length == 0) { aconnector->edid = NULL; - drm_dp_cec_unset_edid(&aconnector->dm_dp_aux.aux); + if (aconnector->dc_link->aux_mode) { + drm_dp_cec_unset_edid( + &aconnector->dm_dp_aux.aux); + } } else { aconnector->edid = - (struct edid *) sink->dc_edid.raw_edid; - + (struct edid *)sink->dc_edid.raw_edid; drm_connector_update_edid_property(connector, - aconnector->edid); - drm_dp_cec_set_edid(&aconnector->dm_dp_aux.aux, - aconnector->edid); + aconnector->edid); + + if (aconnector->dc_link->aux_mode) + drm_dp_cec_set_edid(&aconnector->dm_dp_aux.aux, + aconnector->edid); } + amdgpu_dm_update_freesync_caps(connector, aconnector->edid); update_connector_ext_caps(aconnector); } else { @@ -2445,8 +2403,36 @@ static int dcn10_register_irq_handlers(struct amdgpu_device *adev) c_irq_params->adev = adev; c_irq_params->irq_src = int_params.irq_source; + amdgpu_dm_irq_register_interrupt( + adev, &int_params, dm_crtc_high_irq, c_irq_params); + } + + /* Use VUPDATE_NO_LOCK interrupt on DCN, which seems to correspond to + * the regular VUPDATE interrupt on DCE. We want DC_IRQ_SOURCE_VUPDATEx + * to trigger at end of each vblank, regardless of state of the lock, + * matching DCE behaviour. + */ + for (i = DCN_1_0__SRCID__OTG0_IHC_V_UPDATE_NO_LOCK_INTERRUPT; + i <= DCN_1_0__SRCID__OTG0_IHC_V_UPDATE_NO_LOCK_INTERRUPT + adev->mode_info.num_crtc - 1; + i++) { + r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_DCE, i, &adev->vupdate_irq); + + if (r) { + DRM_ERROR("Failed to add vupdate irq id!\n"); + return r; + } + + int_params.int_context = INTERRUPT_HIGH_IRQ_CONTEXT; + int_params.irq_source = + dc_interrupt_to_irq_source(dc, i, 0); + + c_irq_params = &adev->dm.vupdate_params[int_params.irq_source - DC_IRQ_SOURCE_VUPDATE1]; + + c_irq_params->adev = adev; + c_irq_params->irq_src = int_params.irq_source; + amdgpu_dm_irq_register_interrupt(adev, &int_params, - dm_dcn_crtc_high_irq, c_irq_params); + dm_vupdate_high_irq, c_irq_params); } /* Use GRPH_PFLIP interrupt */ @@ -3661,6 +3647,10 @@ fill_dc_plane_info_and_addr(struct amdgpu_device *adev, case DRM_FORMAT_P010: plane_info->format = SURFACE_PIXEL_FORMAT_VIDEO_420_10bpc_YCrCb; break; + case DRM_FORMAT_XRGB16161616F: + case DRM_FORMAT_ARGB16161616F: + plane_info->format = SURFACE_PIXEL_FORMAT_GRPH_ARGB16161616F; + break; default: DRM_ERROR( "Unsupported screen format %s\n", @@ -4458,10 +4448,6 @@ static inline int dm_set_vupdate_irq(struct drm_crtc *crtc, bool enable) struct amdgpu_device *adev = crtc->dev->dev_private; int rc; - /* Do not set vupdate for DCN hardware */ - if (adev->family > AMDGPU_FAMILY_AI) - return 0; - irq_source = IRQ_TYPE_VUPDATE + acrtc->otg_inst; rc = dc_interrupt_set(adev->dm.dc, irq_source, enable) ? 0 : -EBUSY; @@ -5577,6 +5563,10 @@ static int get_plane_formats(const struct drm_plane *plane, formats[num_formats++] = DRM_FORMAT_NV12; if (plane_cap && plane_cap->pixel_format_support.p010) formats[num_formats++] = DRM_FORMAT_P010; + if (plane_cap && plane_cap->pixel_format_support.fp16) { + formats[num_formats++] = DRM_FORMAT_XRGB16161616F; + formats[num_formats++] = DRM_FORMAT_ARGB16161616F; + } break; case DRM_PLANE_TYPE_OVERLAY: @@ -6865,7 +6855,7 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state, dc_state); if ((acrtc_state->update_type > UPDATE_TYPE_FAST) && - acrtc_state->stream->link->psr_settings.psr_version != PSR_VERSION_UNSUPPORTED && + acrtc_state->stream->link->psr_settings.psr_version != DC_PSR_VERSION_UNSUPPORTED && !acrtc_state->stream->link->psr_settings.psr_feature_enabled) amdgpu_dm_link_setup_psr(acrtc_state->stream); else if ((acrtc_state->update_type == UPDATE_TYPE_FAST) && @@ -7895,6 +7885,7 @@ static int dm_update_plane_state(struct dc *dc, struct drm_crtc_state *old_crtc_state, *new_crtc_state; struct dm_crtc_state *dm_new_crtc_state, *dm_old_crtc_state; struct dm_plane_state *dm_new_plane_state, *dm_old_plane_state; + struct amdgpu_crtc *new_acrtc; bool needs_reset; int ret = 0; @@ -7904,9 +7895,30 @@ static int dm_update_plane_state(struct dc *dc, dm_new_plane_state = to_dm_plane_state(new_plane_state); dm_old_plane_state = to_dm_plane_state(old_plane_state); - /*TODO Implement atomic check for cursor plane */ - if (plane->type == DRM_PLANE_TYPE_CURSOR) + /*TODO Implement better atomic check for cursor plane */ + if (plane->type == DRM_PLANE_TYPE_CURSOR) { + if (!enable || !new_plane_crtc || + drm_atomic_plane_disabling(plane->state, new_plane_state)) + return 0; + + new_acrtc = to_amdgpu_crtc(new_plane_crtc); + + if ((new_plane_state->crtc_w > new_acrtc->max_cursor_width) || + (new_plane_state->crtc_h > new_acrtc->max_cursor_height)) { + DRM_DEBUG_ATOMIC("Bad cursor size %d x %d\n", + new_plane_state->crtc_w, new_plane_state->crtc_h); + return -EINVAL; + } + + if (new_plane_state->crtc_x <= -new_acrtc->max_cursor_width || + new_plane_state->crtc_y <= -new_acrtc->max_cursor_height) { + DRM_DEBUG_ATOMIC("Bad cursor position %d, %d\n", + new_plane_state->crtc_x, new_plane_state->crtc_y); + return -EINVAL; + } + return 0; + } needs_reset = should_reset_plane(state, plane, old_plane_state, new_plane_state); @@ -8640,10 +8652,10 @@ static void amdgpu_dm_set_psr_caps(struct dc_link *link) link->dpcd_caps.psr_caps.psr_version = dpcd_data[0]; if (dpcd_data[0] == 0) { - link->psr_settings.psr_version = PSR_VERSION_UNSUPPORTED; + link->psr_settings.psr_version = DC_PSR_VERSION_UNSUPPORTED; link->psr_settings.psr_feature_enabled = false; } else { - link->psr_settings.psr_version = PSR_VERSION_1; + link->psr_settings.psr_version = DC_PSR_VERSION_1; link->psr_settings.psr_feature_enabled = true; } @@ -8662,14 +8674,12 @@ static bool amdgpu_dm_link_setup_psr(struct dc_stream_state *stream) struct dc_link *link = NULL; struct psr_config psr_config = {0}; struct psr_context psr_context = {0}; - struct dc *dc = NULL; bool ret = false; if (stream == NULL) return false; link = stream->link; - dc = link->ctx->dc; psr_config.psr_version = link->dpcd_caps.psr_caps.psr_version; 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 index b3b7efd973ca..076af267b488 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c @@ -32,7 +32,7 @@ #include "amdgpu_dm.h" #include "amdgpu_dm_debugfs.h" #include "dm_helpers.h" -#include "dmub/inc/dmub_srv.h" +#include "dmub/dmub_srv.h" struct dmub_debugfs_trace_header { uint32_t entry_count; diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_hdcp.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_hdcp.c index 78e1c11d4ae5..dcf84a61de37 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_hdcp.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_hdcp.c @@ -398,15 +398,15 @@ static void update_config(void *handle, struct cp_psp_stream_config *config) struct mod_hdcp_display *display = &hdcp_work[link_index].display; struct mod_hdcp_link *link = &hdcp_work[link_index].link; - memset(display, 0, sizeof(*display)); - memset(link, 0, sizeof(*link)); - - display->index = aconnector->base.index; - if (config->dpms_off) { hdcp_remove_display(hdcp_work, link_index, aconnector); return; } + + memset(display, 0, sizeof(*display)); + memset(link, 0, sizeof(*link)); + + display->index = aconnector->base.index; display->state = MOD_HDCP_DISPLAY_ACTIVE; if (aconnector->dc_sink != NULL) 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 c407f06cd1f5..b086d5c906e0 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 @@ -554,6 +554,7 @@ enum dc_edid_status dm_helpers_read_local_edid( struct dc_sink *sink) { struct amdgpu_dm_connector *aconnector = link->priv; + struct drm_connector *connector = &aconnector->base; struct i2c_adapter *ddc; int retry = 3; enum dc_edid_status edid_status; @@ -571,6 +572,15 @@ enum dc_edid_status dm_helpers_read_local_edid( edid = drm_get_edid(&aconnector->base, ddc); + /* DP Compliance Test 4.2.2.6 */ + if (link->aux_mode && connector->edid_corrupt) + drm_dp_send_real_edid_checksum(&aconnector->dm_dp_aux.aux, connector->real_edid_checksum); + + if (!edid && connector->edid_corrupt) { + connector->edid_corrupt = false; + return EDID_BAD_CHECKSUM; + } + if (!edid) return EDID_NO_RESPONSE; @@ -605,34 +615,10 @@ 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)); - - } + /* DP Compliance Test 4.2.2.3 */ + if (link->aux_mode) + drm_dp_send_real_edid_checksum(&aconnector->dm_dp_aux.aux, sink->dc_edid.raw_edid[sink->dc_edid.length-1]); return edid_status; } diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c index ad817bd74586..45cfb7c45566 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc.c @@ -66,7 +66,7 @@ #include "dce/dce_i2c.h" -#include "dmub/inc/dmub_cmd_dal.h" +#include "dmub/dmub_srv.h" #define CTX \ dc->ctx @@ -839,11 +839,10 @@ static void disable_dangling_plane(struct dc *dc, struct dc_state *context) static void wait_for_no_pipes_pending(struct dc *dc, struct dc_state *context) { int i; - int count = 0; - struct pipe_ctx *pipe; PERF_TRACE(); for (i = 0; i < MAX_PIPES; i++) { - pipe = &context->res_ctx.pipe_ctx[i]; + int count = 0; + struct pipe_ctx *pipe = &context->res_ctx.pipe_ctx[i]; if (!pipe->plane_state) continue; @@ -2210,9 +2209,7 @@ static void commit_planes_do_stream_update(struct dc *dc, if (should_program_abm) { if (*stream_update->abm_level == ABM_LEVEL_IMMEDIATE_DISABLE) { - pipe_ctx->stream_res.abm->funcs->set_abm_immediate_disable( - pipe_ctx->stream_res.abm, - pipe_ctx->stream->link->panel_cntl->inst); + dc->hwss.set_abm_immediate_disable(pipe_ctx); } else { pipe_ctx->stream_res.abm->funcs->set_abm_level( pipe_ctx->stream_res.abm, stream->abm_level); 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 67c5342cf89a..c08de6823db4 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_link.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_link.c @@ -26,7 +26,7 @@ #include <linux/slab.h> #include "dm_services.h" -#include "atom.h" +#include "atomfirmware.h" #include "dm_helpers.h" #include "dc.h" #include "grph_object_id.h" @@ -46,7 +46,7 @@ #include "dmcu.h" #include "hw/clk_mgr.h" #include "dce/dmub_psr.h" -#include "dmub/inc/dmub_cmd_dal.h" +#include "dmub/dmub_srv.h" #include "inc/hw/panel_cntl.h" #define DC_LOGGER_INIT(logger) @@ -1552,7 +1552,7 @@ static bool dc_link_construct(struct dc_link *link, */ program_hpd_filter(link); - link->psr_settings.psr_version = PSR_VERSION_UNSUPPORTED; + link->psr_settings.psr_version = DC_PSR_VERSION_UNSUPPORTED; return true; device_tag_fail: @@ -2504,59 +2504,56 @@ int dc_link_get_target_backlight_pwm(const struct dc_link *link) return (int) abm->funcs->get_target_backlight(abm); } +static struct pipe_ctx *get_pipe_from_link(const struct dc_link *link) +{ + int i; + struct dc *dc = link->ctx->dc; + struct pipe_ctx *pipe_ctx = NULL; + + for (i = 0; i < MAX_PIPES; i++) { + if (dc->current_state->res_ctx.pipe_ctx[i].stream) { + if (dc->current_state->res_ctx.pipe_ctx[i].stream->link == link) { + pipe_ctx = &dc->current_state->res_ctx.pipe_ctx[i]; + break; + } + } + } + + return pipe_ctx; +} + bool dc_link_set_backlight_level(const struct dc_link *link, uint32_t backlight_pwm_u16_16, uint32_t frame_ramp) { struct dc *dc = link->ctx->dc; - int i; DC_LOGGER_INIT(link->ctx->logger); DC_LOG_BACKLIGHT("New Backlight level: %d (0x%X)\n", backlight_pwm_u16_16, backlight_pwm_u16_16); if (dc_is_embedded_signal(link->connector_signal)) { - struct pipe_ctx *pipe_ctx = NULL; - - for (i = 0; i < MAX_PIPES; i++) { - if (dc->current_state->res_ctx.pipe_ctx[i].stream) { - if (dc->current_state->res_ctx. - pipe_ctx[i].stream->link - == link) { - pipe_ctx = &dc->current_state->res_ctx.pipe_ctx[i]; - - /* Disable brightness ramping when the display is blanked - * as it can hang the DMCU - */ - if (dc->current_state->res_ctx.pipe_ctx[i].plane_state == NULL) - frame_ramp = 0; - } - } - } + struct pipe_ctx *pipe_ctx = get_pipe_from_link(link); - if (pipe_ctx == NULL) + if (pipe_ctx) { + /* Disable brightness ramping when the display is blanked + * as it can hang the DMCU + */ + if (pipe_ctx->plane_state == NULL) + frame_ramp = 0; + } else { ASSERT(false); + return false; + } dc->hwss.set_backlight_level( pipe_ctx, backlight_pwm_u16_16, frame_ramp); } - return true; } -bool dc_link_set_abm_disable(const struct dc_link *link) -{ - struct abm *abm = get_abm_from_stream_res(link); - bool success = false; - - if (abm) - success = abm->funcs->set_abm_immediate_disable(abm, link->panel_cntl->inst); - - return success; -} - bool dc_link_set_psr_allow_active(struct dc_link *link, bool allow_active, bool wait) { struct dc *dc = link->ctx->dc; 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 9ef9e50a34fa..1db592372435 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 @@ -13,7 +13,6 @@ #include "core_status.h" #include "dpcd_defs.h" -#include "resource.h" #define DC_LOGGER \ link->ctx->logger 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 1a01c038632b..cb5d11f11cad 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c @@ -1547,35 +1547,6 @@ bool dc_add_all_planes_for_stream( return add_all_planes_for_stream(dc, stream, &set, 1, context); } - -static bool is_hdr_static_meta_changed(struct dc_stream_state *cur_stream, - struct dc_stream_state *new_stream) -{ - if (cur_stream == NULL) - return true; - - if (memcmp(&cur_stream->hdr_static_metadata, - &new_stream->hdr_static_metadata, - sizeof(struct dc_info_packet)) != 0) - return true; - - return false; -} - -static bool is_vsc_info_packet_changed(struct dc_stream_state *cur_stream, - struct dc_stream_state *new_stream) -{ - if (cur_stream == NULL) - return true; - - if (memcmp(&cur_stream->vsc_infopacket, - &new_stream->vsc_infopacket, - sizeof(struct dc_info_packet)) != 0) - return true; - - return false; -} - static bool is_timing_changed(struct dc_stream_state *cur_stream, struct dc_stream_state *new_stream) { @@ -1610,15 +1581,9 @@ static bool are_stream_backends_same( if (is_timing_changed(stream_a, stream_b)) return false; - if (is_hdr_static_meta_changed(stream_a, stream_b)) - return false; - if (stream_a->dpms_off != stream_b->dpms_off) return false; - if (is_vsc_info_packet_changed(stream_a, stream_b)) - return false; - return true; } @@ -1758,21 +1723,6 @@ static struct audio *find_first_free_audio( return 0; } -bool resource_is_stream_unchanged( - struct dc_state *old_context, struct dc_stream_state *stream) -{ - int i; - - for (i = 0; i < old_context->stream_count; i++) { - struct dc_stream_state *old_stream = old_context->streams[i]; - - if (are_stream_backends_same(old_stream, stream)) - return true; - } - - return false; -} - /** * dc_add_stream_to_ctx() - Add a new dc_stream_state to a dc_state. */ @@ -2027,17 +1977,6 @@ enum dc_status resource_map_pool_resources( int pipe_idx = -1; struct dc_bios *dcb = dc->ctx->dc_bios; - /* TODO Check if this is needed */ - /*if (!resource_is_stream_unchanged(old_context, stream)) { - if (stream != NULL && old_context->streams[i] != NULL) { - stream->bit_depth_params = - old_context->streams[i]->bit_depth_params; - stream->clamping = old_context->streams[i]->clamping; - continue; - } - } - */ - calculate_phy_pix_clks(stream); /* TODO: Check Linux */ @@ -2720,15 +2659,9 @@ bool pipe_need_reprogram( if (is_timing_changed(pipe_ctx_old->stream, pipe_ctx->stream)) return true; - if (is_hdr_static_meta_changed(pipe_ctx_old->stream, pipe_ctx->stream)) - return true; - if (pipe_ctx_old->stream->dpms_off != pipe_ctx->stream->dpms_off) return true; - if (is_vsc_info_packet_changed(pipe_ctx_old->stream, pipe_ctx->stream)) - return true; - if (false == pipe_ctx_old->stream->link->link_state_valid && false == pipe_ctx_old->stream->dpms_off) return true; diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h index 17075f99bc54..85908561c741 100644 --- a/drivers/gpu/drm/amd/display/dc/dc.h +++ b/drivers/gpu/drm/amd/display/dc/dc.h @@ -42,7 +42,7 @@ #include "inc/hw/dmcu.h" #include "dml/display_mode_lib.h" -#define DC_VER "3.2.83.1" +#define DC_VER "3.2.84" #define MAX_SURFACES 3 #define MAX_PLANES 6 @@ -277,6 +277,7 @@ struct dc_config { bool disable_extended_timeout_support; // Used to disable extended timeout and lttpr feature as well bool multi_mon_pp_mclk_switch; bool disable_dmcu; + bool enable_4to1MPC; }; enum visual_confirm { @@ -476,6 +477,7 @@ struct dc_debug_options { bool enable_dmcub_surface_flip; bool usbc_combo_phy_reset_wa; bool disable_dsc; + bool enable_dram_clock_change_one_display_vactive; }; struct dc_debug_data { diff --git a/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c b/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c index 907e0c5374bb..eea2429ac67d 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c +++ b/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c @@ -25,7 +25,7 @@ #include "dc.h" #include "dc_dmub_srv.h" -#include "../dmub/inc/dmub_srv.h" +#include "../dmub/dmub_srv.h" static void dc_dmub_srv_construct(struct dc_dmub_srv *dc_srv, struct dc *dc, struct dmub_srv *dmub) diff --git a/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.h b/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.h index 6689ae33dee8..a3a09ccb6d26 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.h +++ b/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.h @@ -27,7 +27,7 @@ #define _DMUB_DC_SRV_H_ #include "os_types.h" -#include "dmub/inc/dmub_cmd.h" +#include "dmub/dmub_srv.h" struct dmub_srv; diff --git a/drivers/gpu/drm/amd/display/dc/dc_link.h b/drivers/gpu/drm/amd/display/dc/dc_link.h index 80fb4149f36a..f63fc25aa6c5 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_link.h +++ b/drivers/gpu/drm/amd/display/dc/dc_link.h @@ -29,7 +29,6 @@ #include "dc.h" #include "dc_types.h" #include "grph_object_defs.h" -#include "dmub/inc/dmub_cmd_dal.h" enum dc_link_fec_state { dc_link_fec_not_ready, @@ -72,7 +71,7 @@ struct link_trace { struct psr_settings { bool psr_feature_enabled; // PSR is supported by sink bool psr_allow_active; // PSR is currently active - enum psr_version psr_version; // Internal PSR version, determined based on DPCD + enum dc_psr_version psr_version; // Internal PSR version, determined based on DPCD /* These parameters are calculated in Driver, * based on display timing and Sink capabilities. @@ -220,8 +219,6 @@ int dc_link_get_backlight_level(const struct dc_link *dc_link); int dc_link_get_target_backlight_pwm(const struct dc_link *link); -bool dc_link_set_abm_disable(const struct dc_link *dc_link); - bool dc_link_set_psr_allow_active(struct dc_link *dc_link, bool enable, bool wait); bool dc_link_get_psr_state(const struct dc_link *dc_link, uint32_t *psr_state); diff --git a/drivers/gpu/drm/amd/display/dc/dc_types.h b/drivers/gpu/drm/amd/display/dc/dc_types.h index 0d210104ba0a..f236da1c1859 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_types.h +++ b/drivers/gpu/drm/amd/display/dc/dc_types.h @@ -862,4 +862,9 @@ struct dsc_dec_dpcd_caps { uint32_t branch_max_line_width; }; +enum dc_psr_version { + DC_PSR_VERSION_1 = 0, + DC_PSR_VERSION_UNSUPPORTED = 0xFFFFFFFF, +}; + #endif /* DC_TYPES_H_ */ diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_abm.c b/drivers/gpu/drm/amd/display/dc/dce/dce_abm.c index c15e60fb5ebc..4e87e70237e3 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_abm.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_abm.c @@ -83,120 +83,6 @@ static bool dce_abm_set_pipe(struct abm *abm, uint32_t controller_id, uint32_t p return true; } -static unsigned int calculate_16_bit_backlight_from_pwm(struct dce_abm *abm_dce) -{ - uint64_t current_backlight; - uint32_t round_result; - uint32_t pwm_period_cntl, bl_period, bl_int_count; - uint32_t bl_pwm_cntl, bl_pwm, fractional_duty_cycle_en; - uint32_t bl_period_mask, bl_pwm_mask; - - pwm_period_cntl = REG_READ(BL_PWM_PERIOD_CNTL); - REG_GET(BL_PWM_PERIOD_CNTL, BL_PWM_PERIOD, &bl_period); - REG_GET(BL_PWM_PERIOD_CNTL, BL_PWM_PERIOD_BITCNT, &bl_int_count); - - bl_pwm_cntl = REG_READ(BL_PWM_CNTL); - REG_GET(BL_PWM_CNTL, BL_ACTIVE_INT_FRAC_CNT, (uint32_t *)(&bl_pwm)); - REG_GET(BL_PWM_CNTL, BL_PWM_FRACTIONAL_EN, &fractional_duty_cycle_en); - - if (bl_int_count == 0) - bl_int_count = 16; - - bl_period_mask = (1 << bl_int_count) - 1; - bl_period &= bl_period_mask; - - bl_pwm_mask = bl_period_mask << (16 - bl_int_count); - - if (fractional_duty_cycle_en == 0) - bl_pwm &= bl_pwm_mask; - else - bl_pwm &= 0xFFFF; - - current_backlight = bl_pwm << (1 + bl_int_count); - - if (bl_period == 0) - bl_period = 0xFFFF; - - current_backlight = div_u64(current_backlight, bl_period); - current_backlight = (current_backlight + 1) >> 1; - - current_backlight = (uint64_t)(current_backlight) * bl_period; - - round_result = (uint32_t)(current_backlight & 0xFFFFFFFF); - - round_result = (round_result >> (bl_int_count-1)) & 1; - - current_backlight >>= bl_int_count; - current_backlight += round_result; - - return (uint32_t)(current_backlight); -} - -static void driver_set_backlight_level(struct dce_abm *abm_dce, - uint32_t backlight_pwm_u16_16) -{ - uint32_t backlight_16bit; - uint32_t masked_pwm_period; - uint8_t bit_count; - uint64_t active_duty_cycle; - uint32_t pwm_period_bitcnt; - - /* - * 1. Find 16 bit backlight active duty cycle, where 0 <= backlight - * active duty cycle <= backlight period - */ - - /* 1.1 Apply bitmask for backlight period value based on value of BITCNT - */ - REG_GET_2(BL_PWM_PERIOD_CNTL, - BL_PWM_PERIOD_BITCNT, &pwm_period_bitcnt, - BL_PWM_PERIOD, &masked_pwm_period); - - if (pwm_period_bitcnt == 0) - bit_count = 16; - else - bit_count = pwm_period_bitcnt; - - /* e.g. maskedPwmPeriod = 0x24 when bitCount is 6 */ - masked_pwm_period = masked_pwm_period & ((1 << bit_count) - 1); - - /* 1.2 Calculate integer active duty cycle required upper 16 bits - * contain integer component, lower 16 bits contain fractional component - * of active duty cycle e.g. 0x21BDC0 = 0xEFF0 * 0x24 - */ - active_duty_cycle = backlight_pwm_u16_16 * masked_pwm_period; - - /* 1.3 Calculate 16 bit active duty cycle from integer and fractional - * components shift by bitCount then mask 16 bits and add rounding bit - * from MSB of fraction e.g. 0x86F7 = ((0x21BDC0 >> 6) & 0xFFF) + 0 - */ - backlight_16bit = active_duty_cycle >> bit_count; - backlight_16bit &= 0xFFFF; - backlight_16bit += (active_duty_cycle >> (bit_count - 1)) & 0x1; - - /* - * 2. Program register with updated value - */ - - /* 2.1 Lock group 2 backlight registers */ - - REG_UPDATE_2(BL_PWM_GRP1_REG_LOCK, - BL_PWM_GRP1_IGNORE_MASTER_LOCK_EN, 1, - BL_PWM_GRP1_REG_LOCK, 1); - - // 2.2 Write new active duty cycle - REG_UPDATE(BL_PWM_CNTL, BL_ACTIVE_INT_FRAC_CNT, backlight_16bit); - - /* 2.3 Unlock group 2 backlight registers */ - REG_UPDATE(BL_PWM_GRP1_REG_LOCK, - BL_PWM_GRP1_REG_LOCK, 0); - - /* 3 Wait for pending bit to be cleared */ - REG_WAIT(BL_PWM_GRP1_REG_LOCK, - BL_PWM_GRP1_REG_UPDATE_PENDING, 0, - 1, 10000); -} - static void dmcu_set_backlight_level( struct dce_abm *abm_dce, uint32_t backlight_pwm_u16_16, @@ -249,10 +135,9 @@ static void dmcu_set_backlight_level( 0, 1, 80000); } -static void dce_abm_init(struct abm *abm) +static void dce_abm_init(struct abm *abm, uint32_t backlight) { struct dce_abm *abm_dce = TO_DCE_ABM(abm); - unsigned int backlight = calculate_16_bit_backlight_from_pwm(abm_dce); REG_WRITE(DC_ABM1_HG_SAMPLE_RATE, 0x103); REG_WRITE(DC_ABM1_HG_SAMPLE_RATE, 0x101); @@ -334,85 +219,11 @@ static bool dce_abm_set_level(struct abm *abm, uint32_t level) static bool dce_abm_immediate_disable(struct abm *abm, uint32_t panel_inst) { - struct dce_abm *abm_dce = TO_DCE_ABM(abm); - if (abm->dmcu_is_running == false) return true; dce_abm_set_pipe(abm, MCP_DISABLE_ABM_IMMEDIATELY, panel_inst); - abm->stored_backlight_registers.BL_PWM_CNTL = - REG_READ(BL_PWM_CNTL); - abm->stored_backlight_registers.BL_PWM_CNTL2 = - REG_READ(BL_PWM_CNTL2); - abm->stored_backlight_registers.BL_PWM_PERIOD_CNTL = - REG_READ(BL_PWM_PERIOD_CNTL); - - REG_GET(LVTMA_PWRSEQ_REF_DIV, BL_PWM_REF_DIV, - &abm->stored_backlight_registers.LVTMA_PWRSEQ_REF_DIV_BL_PWM_REF_DIV); - return true; -} - -static bool dce_abm_init_backlight(struct abm *abm) -{ - struct dce_abm *abm_dce = TO_DCE_ABM(abm); - uint32_t value; - - /* It must not be 0, so we have to restore them - * Bios bug w/a - period resets to zero, - * restoring to cache values which is always correct - */ - REG_GET(BL_PWM_CNTL, BL_ACTIVE_INT_FRAC_CNT, &value); - if (value == 0 || value == 1) { - if (abm->stored_backlight_registers.BL_PWM_CNTL != 0) { - REG_WRITE(BL_PWM_CNTL, - abm->stored_backlight_registers.BL_PWM_CNTL); - REG_WRITE(BL_PWM_CNTL2, - abm->stored_backlight_registers.BL_PWM_CNTL2); - REG_WRITE(BL_PWM_PERIOD_CNTL, - abm->stored_backlight_registers.BL_PWM_PERIOD_CNTL); - REG_UPDATE(LVTMA_PWRSEQ_REF_DIV, - BL_PWM_REF_DIV, - abm->stored_backlight_registers. - LVTMA_PWRSEQ_REF_DIV_BL_PWM_REF_DIV); - } else { - /* TODO: Note: This should not really happen since VBIOS - * should have initialized PWM registers on boot. - */ - REG_WRITE(BL_PWM_CNTL, 0xC000FA00); - REG_WRITE(BL_PWM_PERIOD_CNTL, 0x000C0FA0); - } - } else { - abm->stored_backlight_registers.BL_PWM_CNTL = - REG_READ(BL_PWM_CNTL); - abm->stored_backlight_registers.BL_PWM_CNTL2 = - REG_READ(BL_PWM_CNTL2); - abm->stored_backlight_registers.BL_PWM_PERIOD_CNTL = - REG_READ(BL_PWM_PERIOD_CNTL); - - REG_GET(LVTMA_PWRSEQ_REF_DIV, BL_PWM_REF_DIV, - &abm->stored_backlight_registers. - LVTMA_PWRSEQ_REF_DIV_BL_PWM_REF_DIV); - } - - /* Have driver take backlight control - * TakeBacklightControl(true) - */ - value = REG_READ(BIOS_SCRATCH_2); - value |= ATOM_S2_VRI_BRIGHT_ENABLE; - REG_WRITE(BIOS_SCRATCH_2, value); - - /* Enable the backlight output */ - REG_UPDATE(BL_PWM_CNTL, BL_PWM_EN, 1); - - /* Disable fractional pwm if configured */ - REG_UPDATE(BL_PWM_CNTL, BL_PWM_FRACTIONAL_EN, - abm->ctx->dc->config.disable_fractional_pwm ? 0 : 1); - - /* Unlock group 2 backlight registers */ - REG_UPDATE(BL_PWM_GRP1_REG_LOCK, - BL_PWM_GRP1_REG_LOCK, 0); - return true; } @@ -421,23 +232,18 @@ static bool dce_abm_set_backlight_level_pwm( unsigned int backlight_pwm_u16_16, unsigned int frame_ramp, unsigned int controller_id, - unsigned int panel_inst, - bool fw_set_brightness) + unsigned int panel_inst) { struct dce_abm *abm_dce = TO_DCE_ABM(abm); DC_LOG_BACKLIGHT("New Backlight level: %d (0x%X)\n", backlight_pwm_u16_16, backlight_pwm_u16_16); - /* If DMCU is in reset state, DMCU is uninitialized */ - if (fw_set_brightness) - dmcu_set_backlight_level(abm_dce, - backlight_pwm_u16_16, - frame_ramp, - controller_id, - panel_inst); - else - driver_set_backlight_level(abm_dce, backlight_pwm_u16_16); + dmcu_set_backlight_level(abm_dce, + backlight_pwm_u16_16, + frame_ramp, + controller_id, + panel_inst); return true; } @@ -445,13 +251,12 @@ static bool dce_abm_set_backlight_level_pwm( static const struct abm_funcs dce_funcs = { .abm_init = dce_abm_init, .set_abm_level = dce_abm_set_level, - .init_backlight = dce_abm_init_backlight, .set_pipe = dce_abm_set_pipe, .set_backlight_level_pwm = dce_abm_set_backlight_level_pwm, .get_current_backlight = dce_abm_get_current_backlight, .get_target_backlight = dce_abm_get_target_backlight, .init_abm_config = NULL, - .set_abm_immediate_disable = dce_abm_immediate_disable + .set_abm_immediate_disable = dce_abm_immediate_disable, }; static void dce_abm_construct( @@ -465,10 +270,6 @@ static void dce_abm_construct( base->ctx = ctx; base->funcs = &dce_funcs; - base->stored_backlight_registers.BL_PWM_CNTL = 0; - base->stored_backlight_registers.BL_PWM_CNTL2 = 0; - base->stored_backlight_registers.BL_PWM_PERIOD_CNTL = 0; - base->stored_backlight_registers.LVTMA_PWRSEQ_REF_DIV_BL_PWM_REF_DIV = 0; base->dmcu_is_running = false; abm_dce->regs = regs; diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_abm.h b/drivers/gpu/drm/amd/display/dc/dce/dce_abm.h index ba0caaffa24b..9718a4823372 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_abm.h +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_abm.h @@ -30,11 +30,6 @@ #include "abm.h" #define ABM_COMMON_REG_LIST_DCE_BASE() \ - SR(BL_PWM_PERIOD_CNTL), \ - SR(BL_PWM_CNTL), \ - SR(BL_PWM_CNTL2), \ - SR(BL_PWM_GRP1_REG_LOCK), \ - SR(LVTMA_PWRSEQ_REF_DIV), \ SR(MASTER_COMM_CNTL_REG), \ SR(MASTER_COMM_CMD_REG), \ SR(MASTER_COMM_DATA_REG1) @@ -85,15 +80,6 @@ .field_name = reg_name ## __ ## field_name ## post_fix #define ABM_COMMON_MASK_SH_LIST_DCE_COMMON_BASE(mask_sh) \ - ABM_SF(BL_PWM_PERIOD_CNTL, BL_PWM_PERIOD, mask_sh), \ - ABM_SF(BL_PWM_PERIOD_CNTL, BL_PWM_PERIOD_BITCNT, mask_sh), \ - ABM_SF(BL_PWM_CNTL, BL_ACTIVE_INT_FRAC_CNT, mask_sh), \ - ABM_SF(BL_PWM_CNTL, BL_PWM_FRACTIONAL_EN, mask_sh), \ - ABM_SF(BL_PWM_CNTL, BL_PWM_EN, mask_sh), \ - ABM_SF(BL_PWM_GRP1_REG_LOCK, BL_PWM_GRP1_IGNORE_MASTER_LOCK_EN, mask_sh), \ - ABM_SF(BL_PWM_GRP1_REG_LOCK, BL_PWM_GRP1_REG_LOCK, mask_sh), \ - ABM_SF(BL_PWM_GRP1_REG_LOCK, BL_PWM_GRP1_REG_UPDATE_PENDING, mask_sh), \ - ABM_SF(LVTMA_PWRSEQ_REF_DIV, BL_PWM_REF_DIV, mask_sh), \ ABM_SF(MASTER_COMM_CNTL_REG, MASTER_COMM_INTERRUPT, mask_sh), \ ABM_SF(MASTER_COMM_CMD_REG, MASTER_COMM_CMD_REG_BYTE0, mask_sh), \ ABM_SF(MASTER_COMM_CMD_REG, MASTER_COMM_CMD_REG_BYTE1, mask_sh), \ @@ -178,19 +164,10 @@ type ABM1_HG_REG_READ_MISSED_FRAME_CLEAR; \ type ABM1_LS_REG_READ_MISSED_FRAME_CLEAR; \ type ABM1_BL_REG_READ_MISSED_FRAME_CLEAR; \ - type BL_PWM_PERIOD; \ - type BL_PWM_PERIOD_BITCNT; \ - type BL_ACTIVE_INT_FRAC_CNT; \ - type BL_PWM_FRACTIONAL_EN; \ type MASTER_COMM_INTERRUPT; \ type MASTER_COMM_CMD_REG_BYTE0; \ type MASTER_COMM_CMD_REG_BYTE1; \ - type MASTER_COMM_CMD_REG_BYTE2; \ - type BL_PWM_REF_DIV; \ - type BL_PWM_EN; \ - type BL_PWM_GRP1_IGNORE_MASTER_LOCK_EN; \ - type BL_PWM_GRP1_REG_LOCK; \ - type BL_PWM_GRP1_REG_UPDATE_PENDING + type MASTER_COMM_CMD_REG_BYTE2 struct dce_abm_shift { ABM_REG_FIELD_LIST(uint8_t); @@ -201,10 +178,6 @@ struct dce_abm_mask { }; struct dce_abm_registers { - uint32_t BL_PWM_PERIOD_CNTL; - uint32_t BL_PWM_CNTL; - uint32_t BL_PWM_CNTL2; - uint32_t LVTMA_PWRSEQ_REF_DIV; uint32_t DC_ABM1_HG_SAMPLE_RATE; uint32_t DC_ABM1_LS_SAMPLE_RATE; uint32_t BL1_PWM_BL_UPDATE_SAMPLE_RATE; @@ -219,7 +192,6 @@ struct dce_abm_registers { uint32_t MASTER_COMM_CMD_REG; uint32_t MASTER_COMM_DATA_REG1; uint32_t BIOS_SCRATCH_2; - uint32_t BL_PWM_GRP1_REG_LOCK; }; struct dce_abm { diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_panel_cntl.c b/drivers/gpu/drm/amd/display/dc/dce/dce_panel_cntl.c index d9b0ff7eb2a4..ebff9b1e312e 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_panel_cntl.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_panel_cntl.c @@ -28,6 +28,7 @@ #include "dc_dmub_srv.h" #include "panel_cntl.h" #include "dce_panel_cntl.h" +#include "atom.h" #define TO_DCE_PANEL_CNTL(panel_cntl)\ container_of(panel_cntl, struct dce_panel_cntl, base) @@ -45,9 +46,113 @@ #define FN(reg_name, field_name) \ dce_panel_cntl->shift->field_name, dce_panel_cntl->mask->field_name -void dce_panel_cntl_hw_init(struct panel_cntl *panel_cntl) +static unsigned int calculate_16_bit_backlight_from_pwm(struct dce_panel_cntl *dce_panel_cntl) { + uint64_t current_backlight; + uint32_t round_result; + uint32_t pwm_period_cntl, bl_period, bl_int_count; + uint32_t bl_pwm_cntl, bl_pwm, fractional_duty_cycle_en; + uint32_t bl_period_mask, bl_pwm_mask; + pwm_period_cntl = REG_READ(BL_PWM_PERIOD_CNTL); + REG_GET(BL_PWM_PERIOD_CNTL, BL_PWM_PERIOD, &bl_period); + REG_GET(BL_PWM_PERIOD_CNTL, BL_PWM_PERIOD_BITCNT, &bl_int_count); + + bl_pwm_cntl = REG_READ(BL_PWM_CNTL); + REG_GET(BL_PWM_CNTL, BL_ACTIVE_INT_FRAC_CNT, (uint32_t *)(&bl_pwm)); + REG_GET(BL_PWM_CNTL, BL_PWM_FRACTIONAL_EN, &fractional_duty_cycle_en); + + if (bl_int_count == 0) + bl_int_count = 16; + + bl_period_mask = (1 << bl_int_count) - 1; + bl_period &= bl_period_mask; + + bl_pwm_mask = bl_period_mask << (16 - bl_int_count); + + if (fractional_duty_cycle_en == 0) + bl_pwm &= bl_pwm_mask; + else + bl_pwm &= 0xFFFF; + + current_backlight = bl_pwm << (1 + bl_int_count); + + if (bl_period == 0) + bl_period = 0xFFFF; + + current_backlight = div_u64(current_backlight, bl_period); + current_backlight = (current_backlight + 1) >> 1; + + current_backlight = (uint64_t)(current_backlight) * bl_period; + + round_result = (uint32_t)(current_backlight & 0xFFFFFFFF); + + round_result = (round_result >> (bl_int_count-1)) & 1; + + current_backlight >>= bl_int_count; + current_backlight += round_result; + + return (uint32_t)(current_backlight); +} + +uint32_t dce_panel_cntl_hw_init(struct panel_cntl *panel_cntl) +{ + struct dce_panel_cntl *dce_panel_cntl = TO_DCE_PANEL_CNTL(panel_cntl); + uint32_t value; + uint32_t current_backlight; + + /* It must not be 0, so we have to restore them + * Bios bug w/a - period resets to zero, + * restoring to cache values which is always correct + */ + REG_GET(BL_PWM_CNTL, BL_ACTIVE_INT_FRAC_CNT, &value); + + if (value == 0 || value == 1) { + if (panel_cntl->stored_backlight_registers.BL_PWM_CNTL != 0) { + REG_WRITE(BL_PWM_CNTL, + panel_cntl->stored_backlight_registers.BL_PWM_CNTL); + REG_WRITE(BL_PWM_CNTL2, + panel_cntl->stored_backlight_registers.BL_PWM_CNTL2); + REG_WRITE(BL_PWM_PERIOD_CNTL, + panel_cntl->stored_backlight_registers.BL_PWM_PERIOD_CNTL); + REG_UPDATE(PWRSEQ_REF_DIV, + BL_PWM_REF_DIV, + panel_cntl->stored_backlight_registers.LVTMA_PWRSEQ_REF_DIV_BL_PWM_REF_DIV); + } else { + /* TODO: Note: This should not really happen since VBIOS + * should have initialized PWM registers on boot. + */ + REG_WRITE(BL_PWM_CNTL, 0xC000FA00); + REG_WRITE(BL_PWM_PERIOD_CNTL, 0x000C0FA0); + } + } else { + panel_cntl->stored_backlight_registers.BL_PWM_CNTL = + REG_READ(BL_PWM_CNTL); + panel_cntl->stored_backlight_registers.BL_PWM_CNTL2 = + REG_READ(BL_PWM_CNTL2); + panel_cntl->stored_backlight_registers.BL_PWM_PERIOD_CNTL = + REG_READ(BL_PWM_PERIOD_CNTL); + + REG_GET(PWRSEQ_REF_DIV, BL_PWM_REF_DIV, + &panel_cntl->stored_backlight_registers.LVTMA_PWRSEQ_REF_DIV_BL_PWM_REF_DIV); + } + + // Have driver take backlight control + // TakeBacklightControl(true) + value = REG_READ(BIOS_SCRATCH_2); + value |= ATOM_S2_VRI_BRIGHT_ENABLE; + REG_WRITE(BIOS_SCRATCH_2, value); + + // Enable the backlight output + REG_UPDATE(BL_PWM_CNTL, BL_PWM_EN, 1); + + // Unlock group 2 backlight registers + REG_UPDATE(BL_PWM_GRP1_REG_LOCK, + BL_PWM_GRP1_REG_LOCK, 0); + + current_backlight = calculate_16_bit_backlight_from_pwm(dce_panel_cntl); + + return current_backlight; } bool dce_is_panel_backlight_on(struct panel_cntl *panel_cntl) @@ -55,7 +160,7 @@ bool dce_is_panel_backlight_on(struct panel_cntl *panel_cntl) struct dce_panel_cntl *dce_panel_cntl = TO_DCE_PANEL_CNTL(panel_cntl); uint32_t value; - REG_GET(PWRSEQ_CNTL, BLON, &value); + REG_GET(PWRSEQ_CNTL, LVTMA_BLON, &value); return value; } @@ -65,13 +170,94 @@ bool dce_is_panel_powered_on(struct panel_cntl *panel_cntl) struct dce_panel_cntl *dce_panel_cntl = TO_DCE_PANEL_CNTL(panel_cntl); uint32_t pwr_seq_state, dig_on, dig_on_ovrd; - REG_GET(PWRSEQ_STATE, PWRSEQ_TARGET_STATE_R, &pwr_seq_state); + REG_GET(PWRSEQ_STATE, LVTMA_PWRSEQ_TARGET_STATE_R, &pwr_seq_state); - REG_GET_2(PWRSEQ_CNTL, DIGON, &dig_on, DIGON_OVRD, &dig_on_ovrd); + REG_GET_2(PWRSEQ_CNTL, LVTMA_DIGON, &dig_on, LVTMA_DIGON_OVRD, &dig_on_ovrd); return (pwr_seq_state == 1) || (dig_on == 1 && dig_on_ovrd == 1); } +void dce_store_backlight_level(struct panel_cntl *panel_cntl) +{ + struct dce_panel_cntl *dce_panel_cntl = TO_DCE_PANEL_CNTL(panel_cntl); + + panel_cntl->stored_backlight_registers.BL_PWM_CNTL = + REG_READ(BL_PWM_CNTL); + panel_cntl->stored_backlight_registers.BL_PWM_CNTL2 = + REG_READ(BL_PWM_CNTL2); + panel_cntl->stored_backlight_registers.BL_PWM_PERIOD_CNTL = + REG_READ(BL_PWM_PERIOD_CNTL); + + REG_GET(PWRSEQ_REF_DIV, BL_PWM_REF_DIV, + &panel_cntl->stored_backlight_registers.LVTMA_PWRSEQ_REF_DIV_BL_PWM_REF_DIV); +} + +void dce_driver_set_backlight(struct panel_cntl *panel_cntl, + uint32_t backlight_pwm_u16_16) +{ + uint32_t backlight_16bit; + uint32_t masked_pwm_period; + uint8_t bit_count; + uint64_t active_duty_cycle; + uint32_t pwm_period_bitcnt; + struct dce_panel_cntl *dce_panel_cntl = TO_DCE_PANEL_CNTL(panel_cntl); + + /* + * 1. Find 16 bit backlight active duty cycle, where 0 <= backlight + * active duty cycle <= backlight period + */ + + /* 1.1 Apply bitmask for backlight period value based on value of BITCNT + */ + REG_GET_2(BL_PWM_PERIOD_CNTL, + BL_PWM_PERIOD_BITCNT, &pwm_period_bitcnt, + BL_PWM_PERIOD, &masked_pwm_period); + + if (pwm_period_bitcnt == 0) + bit_count = 16; + else + bit_count = pwm_period_bitcnt; + + /* e.g. maskedPwmPeriod = 0x24 when bitCount is 6 */ + masked_pwm_period = masked_pwm_period & ((1 << bit_count) - 1); + + /* 1.2 Calculate integer active duty cycle required upper 16 bits + * contain integer component, lower 16 bits contain fractional component + * of active duty cycle e.g. 0x21BDC0 = 0xEFF0 * 0x24 + */ + active_duty_cycle = backlight_pwm_u16_16 * masked_pwm_period; + + /* 1.3 Calculate 16 bit active duty cycle from integer and fractional + * components shift by bitCount then mask 16 bits and add rounding bit + * from MSB of fraction e.g. 0x86F7 = ((0x21BDC0 >> 6) & 0xFFF) + 0 + */ + backlight_16bit = active_duty_cycle >> bit_count; + backlight_16bit &= 0xFFFF; + backlight_16bit += (active_duty_cycle >> (bit_count - 1)) & 0x1; + + /* + * 2. Program register with updated value + */ + + /* 2.1 Lock group 2 backlight registers */ + + REG_UPDATE_2(BL_PWM_GRP1_REG_LOCK, + BL_PWM_GRP1_IGNORE_MASTER_LOCK_EN, 1, + BL_PWM_GRP1_REG_LOCK, 1); + + // 2.2 Write new active duty cycle + REG_UPDATE(BL_PWM_CNTL, BL_ACTIVE_INT_FRAC_CNT, backlight_16bit); + + /* 2.3 Unlock group 2 backlight registers */ + REG_UPDATE(BL_PWM_GRP1_REG_LOCK, + BL_PWM_GRP1_REG_LOCK, 0); + + /* 3 Wait for pending bit to be cleared */ + REG_WAIT(BL_PWM_GRP1_REG_LOCK, + BL_PWM_GRP1_REG_UPDATE_PENDING, 0, + 1, 10000); +} + static void dce_panel_cntl_destroy(struct panel_cntl **panel_cntl) { struct dce_panel_cntl *dce_panel_cntl = TO_DCE_PANEL_CNTL(*panel_cntl); @@ -85,7 +271,8 @@ static const struct panel_cntl_funcs dce_link_panel_cntl_funcs = { .hw_init = dce_panel_cntl_hw_init, .is_panel_backlight_on = dce_is_panel_backlight_on, .is_panel_powered_on = dce_is_panel_powered_on, - + .store_backlight_level = dce_store_backlight_level, + .driver_set_backlight = dce_driver_set_backlight, }; void dce_panel_cntl_construct( @@ -95,6 +282,13 @@ void dce_panel_cntl_construct( const struct dce_panel_cntl_shift *shift, const struct dce_panel_cntl_mask *mask) { + struct panel_cntl *base = &dce_panel_cntl->base; + + base->stored_backlight_registers.BL_PWM_CNTL = 0; + base->stored_backlight_registers.BL_PWM_CNTL2 = 0; + base->stored_backlight_registers.BL_PWM_PERIOD_CNTL = 0; + base->stored_backlight_registers.LVTMA_PWRSEQ_REF_DIV_BL_PWM_REF_DIV = 0; + dce_panel_cntl->regs = regs; dce_panel_cntl->shift = shift; dce_panel_cntl->mask = mask; diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_panel_cntl.h b/drivers/gpu/drm/amd/display/dc/dce/dce_panel_cntl.h index 6dc6984f9248..70ec691e14d2 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_panel_cntl.h +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_panel_cntl.h @@ -35,10 +35,12 @@ #define DCE_PANEL_CNTL_REG_LIST()\ DCE_PANEL_CNTL_SR(PWRSEQ_CNTL, LVTMA), \ DCE_PANEL_CNTL_SR(PWRSEQ_STATE, LVTMA), \ + DCE_PANEL_CNTL_SR(PWRSEQ_REF_DIV, LVTMA), \ SR(BL_PWM_CNTL), \ SR(BL_PWM_CNTL2), \ SR(BL_PWM_PERIOD_CNTL), \ - SR(BL_PWM_GRP1_REG_LOCK) + SR(BL_PWM_GRP1_REG_LOCK), \ + SR(BIOS_SCRATCH_2) #define DCN_PANEL_CNTL_SR(reg_name, block)\ .reg_name = BASE(mm ## block ## _ ## reg_name ## _BASE_IDX) + \ @@ -47,33 +49,37 @@ #define DCN_PANEL_CNTL_REG_LIST()\ DCN_PANEL_CNTL_SR(PWRSEQ_CNTL, LVTMA), \ DCN_PANEL_CNTL_SR(PWRSEQ_STATE, LVTMA), \ + DCE_PANEL_CNTL_SR(PWRSEQ_REF_DIV, LVTMA), \ SR(BL_PWM_CNTL), \ SR(BL_PWM_CNTL2), \ SR(BL_PWM_PERIOD_CNTL), \ - SR(BL_PWM_GRP1_REG_LOCK) + SR(BL_PWM_GRP1_REG_LOCK), \ + SR(BIOS_SCRATCH_2) -#define DCE_PANEL_CNTL_SF(block, reg_name, field_name, post_fix)\ - .field_name = block ## reg_name ## __ ## block ## field_name ## post_fix +#define DCE_PANEL_CNTL_SF(reg_name, field_name, post_fix)\ + .field_name = reg_name ## __ ## field_name ## post_fix #define DCE_PANEL_CNTL_MASK_SH_LIST(mask_sh) \ - DCE_PANEL_CNTL_SF(LVTMA_, PWRSEQ_CNTL, BLON, mask_sh),\ - DCE_PANEL_CNTL_SF(LVTMA_, PWRSEQ_CNTL, DIGON, mask_sh),\ - DCE_PANEL_CNTL_SF(LVTMA_, PWRSEQ_CNTL, DIGON_OVRD, mask_sh),\ - DCE_PANEL_CNTL_SF(LVTMA_, PWRSEQ_STATE, PWRSEQ_TARGET_STATE_R, mask_sh), \ - DCE_PANEL_CNTL_SF(, BL_PWM_PERIOD_CNTL, BL_PWM_PERIOD, mask_sh), \ - DCE_PANEL_CNTL_SF(, BL_PWM_PERIOD_CNTL, BL_PWM_PERIOD_BITCNT, mask_sh), \ - DCE_PANEL_CNTL_SF(, BL_PWM_CNTL, BL_ACTIVE_INT_FRAC_CNT, mask_sh), \ - DCE_PANEL_CNTL_SF(, BL_PWM_CNTL, BL_PWM_FRACTIONAL_EN, mask_sh), \ - DCE_PANEL_CNTL_SF(, BL_PWM_CNTL, BL_PWM_EN, mask_sh), \ - DCE_PANEL_CNTL_SF(, BL_PWM_GRP1_REG_LOCK, BL_PWM_GRP1_IGNORE_MASTER_LOCK_EN, mask_sh), \ - DCE_PANEL_CNTL_SF(, BL_PWM_GRP1_REG_LOCK, BL_PWM_GRP1_REG_LOCK, mask_sh), \ - DCE_PANEL_CNTL_SF(, BL_PWM_GRP1_REG_LOCK, BL_PWM_GRP1_REG_UPDATE_PENDING, mask_sh) + DCE_PANEL_CNTL_SF(LVTMA_PWRSEQ_CNTL, LVTMA_BLON, mask_sh),\ + DCE_PANEL_CNTL_SF(LVTMA_PWRSEQ_CNTL, LVTMA_DIGON, mask_sh),\ + DCE_PANEL_CNTL_SF(LVTMA_PWRSEQ_CNTL, LVTMA_DIGON_OVRD, mask_sh),\ + DCE_PANEL_CNTL_SF(LVTMA_PWRSEQ_STATE, LVTMA_PWRSEQ_TARGET_STATE_R, mask_sh), \ + DCE_PANEL_CNTL_SF(LVTMA_PWRSEQ_REF_DIV, BL_PWM_REF_DIV, mask_sh), \ + DCE_PANEL_CNTL_SF(BL_PWM_PERIOD_CNTL, BL_PWM_PERIOD, mask_sh), \ + DCE_PANEL_CNTL_SF(BL_PWM_PERIOD_CNTL, BL_PWM_PERIOD_BITCNT, mask_sh), \ + DCE_PANEL_CNTL_SF(BL_PWM_CNTL, BL_ACTIVE_INT_FRAC_CNT, mask_sh), \ + DCE_PANEL_CNTL_SF(BL_PWM_CNTL, BL_PWM_FRACTIONAL_EN, mask_sh), \ + DCE_PANEL_CNTL_SF(BL_PWM_CNTL, BL_PWM_EN, mask_sh), \ + DCE_PANEL_CNTL_SF(BL_PWM_GRP1_REG_LOCK, BL_PWM_GRP1_IGNORE_MASTER_LOCK_EN, mask_sh), \ + DCE_PANEL_CNTL_SF(BL_PWM_GRP1_REG_LOCK, BL_PWM_GRP1_REG_LOCK, mask_sh), \ + DCE_PANEL_CNTL_SF(BL_PWM_GRP1_REG_LOCK, BL_PWM_GRP1_REG_UPDATE_PENDING, mask_sh) #define DCE_PANEL_CNTL_REG_FIELD_LIST(type) \ - type BLON;\ - type DIGON;\ - type DIGON_OVRD;\ - type PWRSEQ_TARGET_STATE_R; \ + type LVTMA_BLON;\ + type LVTMA_DIGON;\ + type LVTMA_DIGON_OVRD;\ + type LVTMA_PWRSEQ_TARGET_STATE_R; \ + type BL_PWM_REF_DIV; \ type BL_PWM_EN; \ type BL_ACTIVE_INT_FRAC_CNT; \ type BL_PWM_FRACTIONAL_EN; \ @@ -98,6 +104,8 @@ struct dce_panel_cntl_registers { uint32_t BL_PWM_CNTL2; uint32_t BL_PWM_PERIOD_CNTL; uint32_t BL_PWM_GRP1_REG_LOCK; + uint32_t PWRSEQ_REF_DIV; + uint32_t BIOS_SCRATCH_2; }; struct dce_panel_cntl { 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 451574971b96..4cdaaf4d881c 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 @@ -1336,7 +1336,6 @@ static void dce110_se_audio_setup( { struct dce110_stream_encoder *enc110 = DCE110STRENC_FROM_STRENC(enc); - uint32_t speakers = 0; uint32_t channels = 0; ASSERT(audio_info); @@ -1344,7 +1343,6 @@ static void dce110_se_audio_setup( /* This should not happen.it does so we don't get BSOD*/ return; - speakers = audio_info->flags.info.ALLSPEAKERS; channels = speakers_to_channels(audio_info->flags.speaker_flags).all; /* setup the audio stream source select (audio -> dig mapping) */ diff --git a/drivers/gpu/drm/amd/display/dc/dce/dmub_abm.c b/drivers/gpu/drm/amd/display/dc/dce/dmub_abm.c index 06d39d529c09..da0b29abfbda 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dmub_abm.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dmub_abm.c @@ -27,7 +27,7 @@ #include "dce_abm.h" #include "dc.h" #include "dc_dmub_srv.h" -#include "dmub/inc/dmub_srv.h" +#include "dmub/dmub_srv.h" #include "core_types.h" #include "dm_services.h" #include "reg_helper.h" @@ -70,53 +70,6 @@ static bool dmub_abm_set_pipe(struct abm *abm, uint32_t otg_inst, uint32_t panel return true; } -static unsigned int calculate_16_bit_backlight_from_pwm(struct dce_abm *dce_abm) -{ - uint64_t current_backlight; - uint32_t round_result; - uint32_t bl_period, bl_int_count; - uint32_t bl_pwm, fractional_duty_cycle_en; - uint32_t bl_period_mask, bl_pwm_mask; - - REG_GET(BL_PWM_PERIOD_CNTL, BL_PWM_PERIOD, &bl_period); - REG_GET(BL_PWM_PERIOD_CNTL, BL_PWM_PERIOD_BITCNT, &bl_int_count); - - REG_GET(BL_PWM_CNTL, BL_ACTIVE_INT_FRAC_CNT, &bl_pwm); - REG_GET(BL_PWM_CNTL, BL_PWM_FRACTIONAL_EN, &fractional_duty_cycle_en); - - if (bl_int_count == 0) - bl_int_count = 16; - - bl_period_mask = (1 << bl_int_count) - 1; - bl_period &= bl_period_mask; - - bl_pwm_mask = bl_period_mask << (16 - bl_int_count); - - if (fractional_duty_cycle_en == 0) - bl_pwm &= bl_pwm_mask; - else - bl_pwm &= 0xFFFF; - - current_backlight = (uint64_t)bl_pwm << (1 + bl_int_count); - - if (bl_period == 0) - bl_period = 0xFFFF; - - current_backlight = div_u64(current_backlight, bl_period); - current_backlight = (current_backlight + 1) >> 1; - - current_backlight = (uint64_t)(current_backlight) * bl_period; - - round_result = (uint32_t)(current_backlight & 0xFFFFFFFF); - - round_result = (round_result >> (bl_int_count-1)) & 1; - - current_backlight >>= bl_int_count; - current_backlight += round_result; - - return (uint32_t)(current_backlight); -} - static void dmcub_set_backlight_level( struct dce_abm *dce_abm, uint32_t backlight_pwm_u16_16, @@ -178,10 +131,9 @@ static void dmub_abm_enable_fractional_pwm(struct dc_context *dc) dc_dmub_srv_wait_idle(dc->dmub_srv); } -static void dmub_abm_init(struct abm *abm) +static void dmub_abm_init(struct abm *abm, uint32_t backlight) { struct dce_abm *dce_abm = TO_DMUB_ABM(abm); - unsigned int backlight = calculate_16_bit_backlight_from_pwm(dce_abm); REG_WRITE(DC_ABM1_HG_SAMPLE_RATE, 0x103); REG_WRITE(DC_ABM1_HG_SAMPLE_RATE, 0x101); @@ -261,77 +213,8 @@ static bool dmub_abm_set_level(struct abm *abm, uint32_t level) static bool dmub_abm_immediate_disable(struct abm *abm, uint32_t panel_inst) { - struct dce_abm *dce_abm = TO_DMUB_ABM(abm); - dmub_abm_set_pipe(abm, DISABLE_ABM_IMMEDIATELY, panel_inst); - abm->stored_backlight_registers.BL_PWM_CNTL = - REG_READ(BL_PWM_CNTL); - abm->stored_backlight_registers.BL_PWM_CNTL2 = - REG_READ(BL_PWM_CNTL2); - abm->stored_backlight_registers.BL_PWM_PERIOD_CNTL = - REG_READ(BL_PWM_PERIOD_CNTL); - - REG_GET(LVTMA_PWRSEQ_REF_DIV, BL_PWM_REF_DIV, - &abm->stored_backlight_registers.LVTMA_PWRSEQ_REF_DIV_BL_PWM_REF_DIV); - - return true; -} - -static bool dmub_abm_init_backlight(struct abm *abm) -{ - struct dce_abm *dce_abm = TO_DMUB_ABM(abm); - uint32_t value; - - /* It must not be 0, so we have to restore them - * Bios bug w/a - period resets to zero, - * restoring to cache values which is always correct - */ - REG_GET(BL_PWM_CNTL, BL_ACTIVE_INT_FRAC_CNT, &value); - - if (value == 0 || value == 1) { - if (abm->stored_backlight_registers.BL_PWM_CNTL != 0) { - REG_WRITE(BL_PWM_CNTL, - abm->stored_backlight_registers.BL_PWM_CNTL); - REG_WRITE(BL_PWM_CNTL2, - abm->stored_backlight_registers.BL_PWM_CNTL2); - REG_WRITE(BL_PWM_PERIOD_CNTL, - abm->stored_backlight_registers.BL_PWM_PERIOD_CNTL); - REG_UPDATE(LVTMA_PWRSEQ_REF_DIV, - BL_PWM_REF_DIV, - abm->stored_backlight_registers.LVTMA_PWRSEQ_REF_DIV_BL_PWM_REF_DIV); - } else { - /* TODO: Note: This should not really happen since VBIOS - * should have initialized PWM registers on boot. - */ - REG_WRITE(BL_PWM_CNTL, 0xC000FA00); - REG_WRITE(BL_PWM_PERIOD_CNTL, 0x000C0FA0); - } - } else { - abm->stored_backlight_registers.BL_PWM_CNTL = - REG_READ(BL_PWM_CNTL); - abm->stored_backlight_registers.BL_PWM_CNTL2 = - REG_READ(BL_PWM_CNTL2); - abm->stored_backlight_registers.BL_PWM_PERIOD_CNTL = - REG_READ(BL_PWM_PERIOD_CNTL); - - REG_GET(LVTMA_PWRSEQ_REF_DIV, BL_PWM_REF_DIV, - &abm->stored_backlight_registers.LVTMA_PWRSEQ_REF_DIV_BL_PWM_REF_DIV); - } - - // Have driver take backlight control - // TakeBacklightControl(true) - value = REG_READ(BIOS_SCRATCH_2); - value |= ATOM_S2_VRI_BRIGHT_ENABLE; - REG_WRITE(BIOS_SCRATCH_2, value); - - // Enable the backlight output - REG_UPDATE(BL_PWM_CNTL, BL_PWM_EN, 1); - - // Unlock group 2 backlight registers - REG_UPDATE(BL_PWM_GRP1_REG_LOCK, - BL_PWM_GRP1_REG_LOCK, 0); - return true; } @@ -340,8 +223,7 @@ static bool dmub_abm_set_backlight_level_pwm( unsigned int backlight_pwm_u16_16, unsigned int frame_ramp, unsigned int otg_inst, - uint32_t panel_inst, - bool fw_set_brightness) + uint32_t panel_inst) { struct dce_abm *dce_abm = TO_DMUB_ABM(abm); @@ -384,7 +266,6 @@ static bool dmub_abm_init_config(struct abm *abm, static const struct abm_funcs abm_funcs = { .abm_init = dmub_abm_init, .set_abm_level = dmub_abm_set_level, - .init_backlight = dmub_abm_init_backlight, .set_pipe = dmub_abm_set_pipe, .set_backlight_level_pwm = dmub_abm_set_backlight_level_pwm, .get_current_backlight = dmub_abm_get_current_backlight, @@ -404,10 +285,6 @@ static void dmub_abm_construct( base->ctx = ctx; base->funcs = &abm_funcs; - base->stored_backlight_registers.BL_PWM_CNTL = 0; - base->stored_backlight_registers.BL_PWM_CNTL2 = 0; - base->stored_backlight_registers.BL_PWM_PERIOD_CNTL = 0; - base->stored_backlight_registers.LVTMA_PWRSEQ_REF_DIV_BL_PWM_REF_DIV = 0; base->dmcu_is_running = false; abm_dce->regs = regs; diff --git a/drivers/gpu/drm/amd/display/dc/dce/dmub_psr.c b/drivers/gpu/drm/amd/display/dc/dce/dmub_psr.c index 9f12c76f21ab..044a0133ebb1 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dmub_psr.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dmub_psr.c @@ -26,8 +26,7 @@ #include "dmub_psr.h" #include "dc.h" #include "dc_dmub_srv.h" -#include "dmub/inc/dmub_srv.h" -#include "dmub/inc/dmub_gpint_cmd.h" +#include "dmub/dmub_srv.h" #include "core_types.h" #define MAX_PIPES 6 @@ -94,12 +93,20 @@ static bool dmub_psr_set_version(struct dmub_psr *dmub, struct dc_stream_state * union dmub_rb_cmd cmd; struct dc_context *dc = dmub->ctx; - if (stream->link->psr_settings.psr_version == PSR_VERSION_UNSUPPORTED) + if (stream->link->psr_settings.psr_version == DC_PSR_VERSION_UNSUPPORTED) return false; cmd.psr_set_version.header.type = DMUB_CMD__PSR; cmd.psr_set_version.header.sub_type = DMUB_CMD__PSR_SET_VERSION; - cmd.psr_set_version.psr_set_version_data.version = stream->link->psr_settings.psr_version; + switch (stream->link->psr_settings.psr_version) { + case DC_PSR_VERSION_1: + cmd.psr_set_version.psr_set_version_data.version = PSR_VERSION_1; + break; + case DC_PSR_VERSION_UNSUPPORTED: + default: + cmd.psr_set_version.psr_set_version_data.version = PSR_VERSION_UNSUPPORTED; + break; + } cmd.psr_set_version.header.payload_bytes = sizeof(struct dmub_cmd_psr_set_version_data); dc_dmub_srv_cmd_queue(dc->dmub_srv, &cmd); 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 6bd8d4e1c294..b77e9dc16086 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 @@ -1066,7 +1066,7 @@ void dce110_blank_stream(struct pipe_ctx *pipe_ctx) if (link->local_sink && link->local_sink->sink_signal == SIGNAL_TYPE_EDP) { hws->funcs.edp_backlight_control(link, false); - dc_link_set_abm_disable(link); + link->dc->hwss.set_abm_immediate_disable(pipe_ctx); } if (dc_is_dp_signal(pipe_ctx->stream->signal)) @@ -2355,6 +2355,7 @@ static void init_hw(struct dc *dc) struct abm *abm; struct dmcu *dmcu; struct dce_hwseq *hws = dc->hwseq; + uint32_t backlight = MAX_BACKLIGHT_LEVEL; bp = dc->ctx->dc_bios; for (i = 0; i < dc->res_pool->pipe_count; i++) { @@ -2401,12 +2402,17 @@ static void init_hw(struct dc *dc) audio->funcs->hw_init(audio); } - abm = dc->res_pool->abm; - if (abm != NULL) { - abm->funcs->init_backlight(abm); - abm->funcs->abm_init(abm); + for (i = 0; i < dc->link_count; i++) { + struct dc_link *link = dc->links[i]; + + if (link->panel_cntl) + backlight = link->panel_cntl->funcs->hw_init(link->panel_cntl); } + abm = dc->res_pool->abm; + if (abm != NULL) + abm->funcs->abm_init(abm, backlight); + dmcu = dc->res_pool->dmcu; if (dmcu != NULL && abm != NULL) abm->dmcu_is_running = dmcu->funcs->is_dmcu_initialized(dmcu); @@ -2721,6 +2727,7 @@ bool dce110_set_backlight_level(struct pipe_ctx *pipe_ctx, struct dc_link *link = pipe_ctx->stream->link; struct dc *dc = link->ctx->dc; struct abm *abm = pipe_ctx->stream_res.abm; + struct panel_cntl *panel_cntl = link->panel_cntl; struct dmcu *dmcu = dc->res_pool->dmcu; bool fw_set_brightness = true; /* DMCU -1 for all controller id values, @@ -2728,23 +2735,38 @@ bool dce110_set_backlight_level(struct pipe_ctx *pipe_ctx, */ uint32_t controller_id = pipe_ctx->stream_res.tg->inst + 1; - if (abm == NULL || (abm->funcs->set_backlight_level_pwm == NULL)) + if (abm == NULL || panel_cntl == NULL || (abm->funcs->set_backlight_level_pwm == NULL)) return false; if (dmcu) fw_set_brightness = dmcu->funcs->is_dmcu_initialized(dmcu); - abm->funcs->set_backlight_level_pwm( - abm, - backlight_pwm_u16_16, - frame_ramp, - controller_id, - link->panel_cntl->inst, - fw_set_brightness); + if (!fw_set_brightness && panel_cntl->funcs->driver_set_backlight) + panel_cntl->funcs->driver_set_backlight(panel_cntl, backlight_pwm_u16_16); + else + abm->funcs->set_backlight_level_pwm( + abm, + backlight_pwm_u16_16, + frame_ramp, + controller_id, + link->panel_cntl->inst); return true; } +void dce110_set_abm_immediate_disable(struct pipe_ctx *pipe_ctx) +{ + struct abm *abm = pipe_ctx->stream_res.abm; + struct panel_cntl *panel_cntl = pipe_ctx->stream->link->panel_cntl; + + if (abm) + abm->funcs->set_abm_immediate_disable(abm, + pipe_ctx->stream->link->panel_cntl->inst); + + if (panel_cntl) + panel_cntl->funcs->store_backlight_level(panel_cntl); +} + static const struct hw_sequencer_funcs dce110_funcs = { .program_gamut_remap = program_gamut_remap, .program_output_csc = program_output_csc, @@ -2781,6 +2803,7 @@ static const struct hw_sequencer_funcs dce110_funcs = { .set_cursor_position = dce110_set_cursor_position, .set_cursor_attribute = dce110_set_cursor_attribute, .set_backlight_level = dce110_set_backlight_level, + .set_abm_immediate_disable = dce110_set_abm_immediate_disable, }; static const struct hwseq_private_funcs dce110_private_funcs = { 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 e609a72b4634..fe5326df00f7 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 @@ -88,6 +88,7 @@ void dce110_edp_wait_for_hpd_ready( bool dce110_set_backlight_level(struct pipe_ctx *pipe_ctx, uint32_t backlight_pwm_u16_16, uint32_t frame_ramp); +void dce110_set_abm_immediate_disable(struct pipe_ctx *pipe_ctx); #endif /* __DC_HWSS_DCE110_H__ */ 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 0e682b5aa3eb..7f8456b9988b 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c @@ -134,13 +134,6 @@ bool dpp1_get_optimal_number_of_taps( struct scaler_data *scl_data, const struct scaling_taps *in_taps) { - uint32_t pixel_width; - - if (scl_data->viewport.width > scl_data->recout.width) - pixel_width = scl_data->recout.width; - else - pixel_width = scl_data->viewport.width; - /* Some ASICs does not support FP16 scaling, so we reject modes require this*/ if (scl_data->format == PIXEL_FORMAT_FP16 && dpp->caps->dscl_data_proc_format == DSCL_DATA_PRCESSING_FIXED_FORMAT && 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 2eb5d0e3e425..f36d1f57b846 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 @@ -826,6 +826,14 @@ enum dc_status dcn10_enable_stream_timing( color_space = stream->output_color_space; color_space_to_black_color(dc, color_space, &black_color); + /* + * The way 420 is packed, 2 channels carry Y component, 1 channel + * alternate between Cb and Cr, so both channels need the pixel + * value for Y + */ + if (stream->timing.pixel_encoding == PIXEL_ENCODING_YCBCR420) + black_color.color_r_cr = black_color.color_g_y; + if (pipe_ctx->stream_res.tg->funcs->set_blank_color) pipe_ctx->stream_res.tg->funcs->set_blank_color( pipe_ctx->stream_res.tg, @@ -903,8 +911,7 @@ static void dcn10_reset_back_end_for_pipe( if (pipe_ctx->top_pipe == NULL) { if (pipe_ctx->stream_res.abm) - pipe_ctx->stream_res.abm->funcs->set_abm_immediate_disable(pipe_ctx->stream_res.abm, - pipe_ctx->stream->link->panel_cntl->inst); + dc->hwss.set_abm_immediate_disable(pipe_ctx); pipe_ctx->stream_res.tg->funcs->disable_crtc(pipe_ctx->stream_res.tg); @@ -1245,6 +1252,7 @@ void dcn10_init_hw(struct dc *dc) struct dce_hwseq *hws = dc->hwseq; struct dc_bios *dcb = dc->ctx->dc_bios; struct resource_pool *res_pool = dc->res_pool; + uint32_t backlight = MAX_BACKLIGHT_LEVEL; if (dc->clk_mgr && dc->clk_mgr->funcs->init_clocks) dc->clk_mgr->funcs->init_clocks(dc->clk_mgr); @@ -1411,11 +1419,16 @@ void dcn10_init_hw(struct dc *dc) audio->funcs->hw_init(audio); } - if (abm != NULL) { - abm->funcs->init_backlight(abm); - abm->funcs->abm_init(abm); + for (i = 0; i < dc->link_count; i++) { + struct dc_link *link = dc->links[i]; + + if (link->panel_cntl) + backlight = link->panel_cntl->funcs->hw_init(link->panel_cntl); } + if (abm != NULL) + abm->funcs->abm_init(abm, backlight); + if (dmcu != NULL && !dmcu->auto_load_dmcu) dmcu->funcs->dmcu_init(dmcu); @@ -2249,6 +2262,14 @@ void dcn10_update_mpcc(struct dc *dc, struct pipe_ctx *pipe_ctx) &blnd_cfg.black_color); } + /* + * The way 420 is packed, 2 channels carry Y component, 1 channel + * alternate between Cb and Cr, so both channels need the pixel + * value for Y + */ + if (pipe_ctx->stream->timing.pixel_encoding == PIXEL_ENCODING_YCBCR420) + blnd_cfg.black_color.color_r_cr = blnd_cfg.black_color.color_g_y; + if (per_pixel_alpha) blnd_cfg.alpha_mode = MPCC_ALPHA_BLEND_MODE_PER_PIXEL_ALPHA; else @@ -2490,9 +2511,7 @@ void dcn10_blank_pixel_data( stream_res->abm->funcs->set_abm_level(stream_res->abm, stream->abm_level); } } else if (blank) { - if (stream_res->abm) - stream_res->abm->funcs->set_abm_immediate_disable(stream_res->abm, - stream->link->panel_cntl->inst); + dc->hwss.set_abm_immediate_disable(pipe_ctx); if (stream_res->tg->funcs->set_blank) stream_res->tg->funcs->set_blank(stream_res->tg, blank); } diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_init.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_init.c index 6ff7e2bd1d4a..897a3d25685a 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_init.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_init.c @@ -73,6 +73,7 @@ static const struct hw_sequencer_funcs dcn10_funcs = { .get_clock = dcn10_get_clock, .get_vupdate_offset_from_vsync = dcn10_get_vupdate_offset_from_vsync, .set_backlight_level = dce110_set_backlight_level, + .set_abm_immediate_disable = dce110_set_abm_immediate_disable, }; static const struct hwseq_private_funcs dcn10_private_funcs = { 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 bacf865f55c0..17d5cb422025 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c @@ -1121,24 +1121,6 @@ static enum dc_status build_mapped_resource( { struct pipe_ctx *pipe_ctx = resource_get_head_pipe_for_stream(&context->res_ctx, stream); - /*TODO Seems unneeded anymore */ - /* if (old_context && resource_is_stream_unchanged(old_context, stream)) { - if (stream != NULL && old_context->streams[i] != NULL) { - todo: shouldn't have to copy missing parameter here - resource_build_bit_depth_reduction_params(stream, - &stream->bit_depth_params); - stream->clamping.pixel_encoding = - stream->timing.pixel_encoding; - - resource_build_bit_depth_reduction_params(stream, - &stream->bit_depth_params); - build_clamping_params(stream); - - continue; - } - } - */ - if (!pipe_ctx) return DC_ERROR_UNEXPECTED; 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 7eba9333c328..07b2f9399671 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 @@ -1274,7 +1274,6 @@ static void enc1_se_audio_setup( { struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc); - uint32_t speakers = 0; uint32_t channels = 0; ASSERT(audio_info); @@ -1282,7 +1281,6 @@ static void enc1_se_audio_setup( /* This should not happen.it does so we don't get BSOD*/ return; - speakers = audio_info->flags.info.ALLSPEAKERS; channels = speakers_to_channels(audio_info->flags.speaker_flags).all; /* setup the audio stream source select (audio -> dig mapping) */ diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c index 3a290ccaeb82..da5333d165ac 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c @@ -961,9 +961,7 @@ void dcn20_blank_pixel_data( width = width / odm_cnt; if (blank) { - if (stream_res->abm) - stream_res->abm->funcs->set_abm_immediate_disable(stream_res->abm, - stream->link->panel_cntl->inst); + dc->hwss.set_abm_immediate_disable(pipe_ctx); if (dc->debug.visual_confirm != VISUAL_CONFIRM_DISABLE) { test_pattern = CONTROLLER_DP_TEST_PATTERN_COLORSQUARES; @@ -2042,9 +2040,7 @@ static void dcn20_reset_back_end_for_pipe( */ if (pipe_ctx->top_pipe == NULL) { - if (pipe_ctx->stream_res.abm) - pipe_ctx->stream_res.abm->funcs->set_abm_immediate_disable(pipe_ctx->stream_res.abm, - pipe_ctx->stream->link->panel_cntl->inst); + dc->hwss.set_abm_immediate_disable(pipe_ctx); pipe_ctx->stream_res.tg->funcs->disable_crtc(pipe_ctx->stream_res.tg); diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_init.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_init.c index c83d98e0d211..a8bcd747d7ba 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_init.c +++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_init.c @@ -84,6 +84,7 @@ static const struct hw_sequencer_funcs dcn20_funcs = { .set_flip_control_gsl = dcn20_set_flip_control_gsl, .get_vupdate_offset_from_vsync = dcn10_get_vupdate_offset_from_vsync, .set_backlight_level = dce110_set_backlight_level, + .set_abm_immediate_disable = dce110_set_abm_immediate_disable, }; static const struct hwseq_private_funcs dcn20_private_funcs = { diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c index 60ea499c1ca8..778e2e8fd2c6 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c @@ -1653,24 +1653,6 @@ enum dc_status dcn20_build_mapped_resource(const struct dc *dc, struct dc_state enum dc_status status = DC_OK; struct pipe_ctx *pipe_ctx = resource_get_head_pipe_for_stream(&context->res_ctx, stream); - /*TODO Seems unneeded anymore */ - /* if (old_context && resource_is_stream_unchanged(old_context, stream)) { - if (stream != NULL && old_context->streams[i] != NULL) { - todo: shouldn't have to copy missing parameter here - resource_build_bit_depth_reduction_params(stream, - &stream->bit_depth_params); - stream->clamping.pixel_encoding = - stream->timing.pixel_encoding; - - resource_build_bit_depth_reduction_params(stream, - &stream->bit_depth_params); - build_clamping_params(stream); - - continue; - } - } - */ - if (!pipe_ctx) return DC_ERROR_UNEXPECTED; @@ -1940,7 +1922,7 @@ bool dcn20_split_stream_for_odm( return true; } -bool dcn20_split_stream_for_mpc( +void dcn20_split_stream_for_mpc( struct resource_context *res_ctx, const struct resource_pool *pool, struct pipe_ctx *primary_pipe, @@ -1969,11 +1951,6 @@ bool dcn20_split_stream_for_mpc( secondary_pipe->top_pipe = primary_pipe; ASSERT(primary_pipe->plane_state); - if (!resource_build_scaling_params(primary_pipe) || - !resource_build_scaling_params(secondary_pipe)) - return false; - - return true; } void dcn20_populate_dml_writeback_from_context( @@ -2599,11 +2576,32 @@ static void dcn20_merge_pipes_for_validate( } } +int dcn20_find_previous_split_count(struct pipe_ctx *pipe) +{ + int previous_split = 1; + struct pipe_ctx *current_pipe = pipe; + + while (current_pipe->bottom_pipe) { + if (current_pipe->plane_state != current_pipe->bottom_pipe->plane_state) + break; + previous_split++; + current_pipe = current_pipe->bottom_pipe; + } + current_pipe = pipe; + while (current_pipe->top_pipe) { + if (current_pipe->plane_state != current_pipe->top_pipe->plane_state) + break; + previous_split++; + current_pipe = current_pipe->top_pipe; + } + return previous_split; +} + int dcn20_validate_apply_pipe_split_flags( struct dc *dc, struct dc_state *context, int vlevel, - bool *split, + int *split, bool *merge) { int i, pipe_idx, vlevel_split; @@ -2658,8 +2656,14 @@ int dcn20_validate_apply_pipe_split_flags( if (!context->res_ctx.pipe_ctx[i].stream) continue; - if (force_split || context->bw_ctx.dml.vba.NoOfDPP[vlevel][context->bw_ctx.dml.vba.maxMpcComb][pipe_plane] > 1) - split[i] = true; + if (force_split + || context->bw_ctx.dml.vba.NoOfDPP[vlevel][context->bw_ctx.dml.vba.maxMpcComb][pipe_plane] > 1) { + if (context->stream_count == 1 && plane_count == 1 + && dc->config.enable_4to1MPC && dc->res_pool->pipe_count >= 4) + split[i] = 4; + else + split[i] = 2; + } if ((pipe->stream->view_format == VIEW_3D_FORMAT_SIDE_BY_SIDE || pipe->stream->view_format == @@ -2668,9 +2672,9 @@ int dcn20_validate_apply_pipe_split_flags( TIMING_3D_FORMAT_TOP_AND_BOTTOM || pipe->stream->timing.timing_3d_format == TIMING_3D_FORMAT_SIDE_BY_SIDE)) - split[i] = true; + split[i] = 2; if (dc->debug.force_odm_combine & (1 << pipe->stream_res.tg->inst)) { - split[i] = true; + split[i] = 2; context->bw_ctx.dml.vba.ODMCombineEnablePerState[vlevel][pipe_plane] = dm_odm_combine_mode_2to1; } context->bw_ctx.dml.vba.ODMCombineEnabled[pipe_plane] = @@ -2678,39 +2682,58 @@ int dcn20_validate_apply_pipe_split_flags( if (pipe->prev_odm_pipe && context->bw_ctx.dml.vba.ODMCombineEnabled[pipe_plane] != dm_odm_combine_mode_disabled) { /*Already split odm pipe tree, don't try to split again*/ - split[i] = false; - split[pipe->prev_odm_pipe->pipe_idx] = false; + split[i] = 0; + split[pipe->prev_odm_pipe->pipe_idx] = 0; } else if (pipe->top_pipe && pipe->plane_state == pipe->top_pipe->plane_state && context->bw_ctx.dml.vba.ODMCombineEnabled[pipe_plane] == dm_odm_combine_mode_disabled) { - /*Already split mpc tree, don't try to split again, assumes only 2x mpc combine*/ - split[i] = false; - split[pipe->top_pipe->pipe_idx] = false; - } else if (pipe->prev_odm_pipe || (pipe->top_pipe && pipe->plane_state == pipe->top_pipe->plane_state)) { - if (split[i] == false) { + /*If 2 way split but can support 4 way split, then split each pipe again*/ + if (context->stream_count == 1 && plane_count == 1 + && dc->config.enable_4to1MPC && dc->res_pool->pipe_count >= 4) { + split[i] = 2; + } else { + split[i] = 0; + split[pipe->top_pipe->pipe_idx] = 0; + } + } else if (pipe->prev_odm_pipe || (dcn20_find_previous_split_count(pipe) == 2 && pipe->top_pipe)) { + if (split[i] == 0) { /*Exiting mpc/odm combine*/ merge[i] = true; - if (pipe->prev_odm_pipe) { - ASSERT(0); /*should not actually happen yet*/ - merge[pipe->prev_odm_pipe->pipe_idx] = true; - } else - merge[pipe->top_pipe->pipe_idx] = true; } else { /*Transition from mpc combine to odm combine or vice versa*/ ASSERT(0); /*should not actually happen yet*/ - split[i] = true; + split[i] = 2; merge[i] = true; if (pipe->prev_odm_pipe) { - split[pipe->prev_odm_pipe->pipe_idx] = true; + split[pipe->prev_odm_pipe->pipe_idx] = 2; merge[pipe->prev_odm_pipe->pipe_idx] = true; } else { - split[pipe->top_pipe->pipe_idx] = true; + split[pipe->top_pipe->pipe_idx] = 2; merge[pipe->top_pipe->pipe_idx] = true; } } + } else if (dcn20_find_previous_split_count(pipe) == 3) { + if (split[i] == 0 && !pipe->top_pipe) { + merge[pipe->bottom_pipe->pipe_idx] = true; + merge[pipe->bottom_pipe->bottom_pipe->pipe_idx] = true; + } else if (split[i] == 2 && !pipe->top_pipe) { + merge[pipe->bottom_pipe->bottom_pipe->pipe_idx] = true; + split[i] = 0; + } + } else if (dcn20_find_previous_split_count(pipe) == 4) { + if (split[i] == 0 && !pipe->top_pipe) { + merge[pipe->bottom_pipe->pipe_idx] = true; + merge[pipe->bottom_pipe->bottom_pipe->pipe_idx] = true; + merge[pipe->bottom_pipe->bottom_pipe->bottom_pipe->pipe_idx] = true; + } else if (split[i] == 2 && !pipe->top_pipe) { + merge[pipe->bottom_pipe->bottom_pipe->pipe_idx] = true; + merge[pipe->bottom_pipe->bottom_pipe->bottom_pipe->pipe_idx] = true; + split[i] = 0; + } } /* Adjust dppclk when split is forced, do not bother with dispclk */ - if (split[i] && context->bw_ctx.dml.vba.NoOfDPP[vlevel][context->bw_ctx.dml.vba.maxMpcComb][pipe_idx] == 1) + if (split[i] != 0 + && context->bw_ctx.dml.vba.NoOfDPP[vlevel][context->bw_ctx.dml.vba.maxMpcComb][pipe_idx] == 1) context->bw_ctx.dml.vba.RequiredDPPCLK[vlevel][context->bw_ctx.dml.vba.maxMpcComb][pipe_idx] /= 2; pipe_idx++; } @@ -2727,7 +2750,7 @@ bool dcn20_fast_validate_bw( int *vlevel_out) { bool out = false; - bool split[MAX_PIPES] = { false }; + int split[MAX_PIPES] = { 0 }; int pipe_cnt, i, pipe_idx, vlevel; ASSERT(pipes); @@ -2787,7 +2810,7 @@ bool dcn20_fast_validate_bw( && context->bw_ctx.dml.vba.ODMCombineEnabled[pipe_idx]) goto validate_fail; - if (split[i]) { + if (split[i] == 2) { if (!hsplit_pipe || hsplit_pipe->plane_state != pipe->plane_state) { /* pipe not split previously needs split */ hsplit_pipe = dcn20_find_secondary_pipe(dc, &context->res_ctx, dc->res_pool, pipe); @@ -2802,11 +2825,13 @@ bool dcn20_fast_validate_bw( pipe, hsplit_pipe)) goto validate_fail; dcn20_build_mapped_resource(dc, context, pipe->stream); - } else - if (!dcn20_split_stream_for_mpc( + } else { + dcn20_split_stream_for_mpc( &context->res_ctx, dc->res_pool, - pipe, hsplit_pipe)) + pipe, hsplit_pipe); + if (!resource_build_scaling_params(pipe) || !resource_build_scaling_params(hsplit_pipe)) goto validate_fail; + } pipe_split_from[hsplit_pipe->pipe_idx] = pipe_idx; } } else if (hsplit_pipe && hsplit_pipe->plane_state == pipe->plane_state) { @@ -3107,25 +3132,34 @@ validate_out: return out; } - -bool dcn20_validate_bandwidth(struct dc *dc, struct dc_state *context, - bool fast_validate) +/* + * This must be noinline to ensure anything that deals with FP registers + * is contained within this call; previously our compiling with hard-float + * would result in fp instructions being emitted outside of the boundaries + * of the DC_FP_START/END macros, which makes sense as the compiler has no + * idea about what is wrapped and what is not + * + * This is largely just a workaround to avoid breakage introduced with 5.6, + * ideally all fp-using code should be moved into its own file, only that + * should be compiled with hard-float, and all code exported from there + * should be strictly wrapped with DC_FP_START/END + */ +static noinline bool dcn20_validate_bandwidth_fp(struct dc *dc, + struct dc_state *context, bool fast_validate) { bool voltage_supported = false; bool full_pstate_supported = false; bool dummy_pstate_supported = false; double p_state_latency_us; - DC_FP_START(); p_state_latency_us = context->bw_ctx.dml.soc.dram_clock_change_latency_us; context->bw_ctx.dml.soc.disable_dram_clock_change_vactive_support = dc->debug.disable_dram_clock_change_vactive_support; + context->bw_ctx.dml.soc.allow_dram_clock_one_display_vactive = + dc->debug.enable_dram_clock_change_one_display_vactive; if (fast_validate) { - voltage_supported = dcn20_validate_bandwidth_internal(dc, context, true); - - DC_FP_END(); - return voltage_supported; + return dcn20_validate_bandwidth_internal(dc, context, true); } // Best case, we support full UCLK switch latency @@ -3154,7 +3188,15 @@ bool dcn20_validate_bandwidth(struct dc *dc, struct dc_state *context, restore_dml_state: context->bw_ctx.dml.soc.dram_clock_change_latency_us = p_state_latency_us; + return voltage_supported; +} +bool dcn20_validate_bandwidth(struct dc *dc, struct dc_state *context, + bool fast_validate) +{ + bool voltage_supported = false; + DC_FP_START(); + voltage_supported = dcn20_validate_bandwidth_fp(dc, context, fast_validate); DC_FP_END(); return voltage_supported; } @@ -3213,8 +3255,6 @@ static struct dc_cap_funcs cap_funcs = { enum dc_status dcn20_patch_unknown_plane_state(struct dc_plane_state *plane_state) { - enum dc_status result = DC_OK; - enum surface_pixel_format surf_pix_format = plane_state->format; unsigned int bpp = resource_pixel_format_to_bpp(surf_pix_format); @@ -3226,7 +3266,7 @@ enum dc_status dcn20_patch_unknown_plane_state(struct dc_plane_state *plane_stat swizzle = DC_SW_64KB_S; plane_state->tiling_info.gfx9.swizzle = swizzle; - return result; + return DC_OK; } static struct resource_funcs dcn20_res_pool_funcs = { diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.h b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.h index 578265ccbf5b..d5448c9b0e15 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.h +++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.h @@ -119,17 +119,18 @@ void dcn20_set_mcif_arb_params( display_e2e_pipe_params_st *pipes, int pipe_cnt); bool dcn20_validate_bandwidth(struct dc *dc, struct dc_state *context, bool fast_validate); +int dcn20_find_previous_split_count(struct pipe_ctx *pipe); int dcn20_validate_apply_pipe_split_flags( struct dc *dc, struct dc_state *context, int vlevel, - bool *split, + int *split, bool *merge); void dcn20_release_dsc(struct resource_context *res_ctx, const struct resource_pool *pool, struct display_stream_compressor **dsc); bool dcn20_validate_dsc(struct dc *dc, struct dc_state *new_ctx); -bool dcn20_split_stream_for_mpc( +void dcn20_split_stream_for_mpc( struct resource_context *res_ctx, const struct resource_pool *pool, struct pipe_ctx *primary_pipe, diff --git a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_init.c b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_init.c index ae05a8bfdae1..e97dfaa656e9 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_init.c +++ b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_init.c @@ -86,13 +86,9 @@ static const struct hw_sequencer_funcs dcn21_funcs = { .optimize_pwr_state = dcn21_optimize_pwr_state, .exit_optimized_pwr_state = dcn21_exit_optimized_pwr_state, .get_vupdate_offset_from_vsync = dcn10_get_vupdate_offset_from_vsync, - .set_cursor_position = dcn10_set_cursor_position, - .set_cursor_attribute = dcn10_set_cursor_attribute, - .set_cursor_sdr_white_level = dcn10_set_cursor_sdr_white_level, - .optimize_pwr_state = dcn21_optimize_pwr_state, - .exit_optimized_pwr_state = dcn21_exit_optimized_pwr_state, .power_down = dce110_power_down, .set_backlight_level = dce110_set_backlight_level, + .set_abm_immediate_disable = dce110_set_abm_immediate_disable, }; static const struct hwseq_private_funcs dcn21_private_funcs = { diff --git a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_resource.c b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_resource.c index ceaf70a934c3..419cdde624f5 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_resource.c @@ -1384,7 +1384,8 @@ static void update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_param struct dcn21_resource_pool *pool = TO_DCN21_RES_POOL(dc->res_pool); struct clk_limit_table *clk_table = &bw_params->clk_table; struct _vcs_dpi_voltage_scaling_st clock_limits[DC__VOLTAGE_STATES]; - unsigned int i, j, closest_clk_lvl; + unsigned int i, closest_clk_lvl; + int j; // Default clock levels are used for diags, which may lead to overclocking. if (!IS_DIAG_DC(dc->ctx->dce_environment)) { diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_mode_vba_20v2.c b/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_mode_vba_20v2.c index 922ab7169e52..80170f9721ce 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_mode_vba_20v2.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_mode_vba_20v2.c @@ -2599,18 +2599,40 @@ static void dml20v2_DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndP } } + { + float SecondMinActiveDRAMClockChangeMarginOneDisplayInVBLank = 999999; + int PlaneWithMinActiveDRAMClockChangeMargin = -1; + mode_lib->vba.MinActiveDRAMClockChangeMargin = 999999; for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { if (mode_lib->vba.ActiveDRAMClockChangeLatencyMargin[k] < mode_lib->vba.MinActiveDRAMClockChangeMargin) { mode_lib->vba.MinActiveDRAMClockChangeMargin = mode_lib->vba.ActiveDRAMClockChangeLatencyMargin[k]; + if (mode_lib->vba.BlendingAndTiming[k] == k) { + PlaneWithMinActiveDRAMClockChangeMargin = k; + } else { + for (j = 0; j < mode_lib->vba.NumberOfActivePlanes; ++j) { + if (mode_lib->vba.BlendingAndTiming[k] == j) { + PlaneWithMinActiveDRAMClockChangeMargin = j; + } + } + } } } mode_lib->vba.MinActiveDRAMClockChangeLatencySupported = mode_lib->vba.MinActiveDRAMClockChangeMargin + mode_lib->vba.DRAMClockChangeLatency; + for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { + if (!((k == PlaneWithMinActiveDRAMClockChangeMargin) && (mode_lib->vba.BlendingAndTiming[k] == k)) + && !(mode_lib->vba.BlendingAndTiming[k] == PlaneWithMinActiveDRAMClockChangeMargin) + && mode_lib->vba.ActiveDRAMClockChangeLatencyMargin[k] + < SecondMinActiveDRAMClockChangeMarginOneDisplayInVBLank) { + SecondMinActiveDRAMClockChangeMarginOneDisplayInVBLank = + mode_lib->vba.ActiveDRAMClockChangeLatencyMargin[k]; + } + } if (mode_lib->vba.DRAMClockChangeSupportsVActive && mode_lib->vba.MinActiveDRAMClockChangeMargin > 60) { @@ -2629,7 +2651,11 @@ static void dml20v2_DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndP mode_lib->vba.MinActiveDRAMClockChangeMargin > 0) { mode_lib->vba.DRAMClockChangeSupport[0][0] = dm_dram_clock_change_vactive; } else { - if (mode_lib->vba.SynchronizedVBlank || mode_lib->vba.NumberOfActivePlanes == 1) { + if ((mode_lib->vba.SynchronizedVBlank + || mode_lib->vba.NumberOfActivePlanes == 1 + || (SecondMinActiveDRAMClockChangeMarginOneDisplayInVBLank > 0 && + mode_lib->vba.AllowDramClockChangeOneDisplayVactive)) + && mode_lib->vba.PrefetchMode[mode_lib->vba.VoltageLevel][mode_lib->vba.maxMpcComb] == 0) { mode_lib->vba.DRAMClockChangeSupport[0][0] = dm_dram_clock_change_vblank; for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { if (!mode_lib->vba.AllowDRAMClockChangeDuringVBlank[k]) { @@ -2641,6 +2667,7 @@ static void dml20v2_DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndP mode_lib->vba.DRAMClockChangeSupport[0][0] = dm_dram_clock_change_unsupported; } } + } for (k = 0; k <= mode_lib->vba.soc.num_states; k++) for (j = 0; j < 2; j++) mode_lib->vba.DRAMClockChangeSupport[k][j] = mode_lib->vba.DRAMClockChangeSupport[0][0]; diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn21/display_mode_vba_21.c b/drivers/gpu/drm/amd/display/dc/dml/dcn21/display_mode_vba_21.c index 5bc80b6084da..a576eed94d9b 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn21/display_mode_vba_21.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn21/display_mode_vba_21.c @@ -3190,6 +3190,7 @@ static void CalculateFlipSchedule( double TimeForFetchingRowInVBlankImmediateFlip; double ImmediateFlipBW; double HostVMInefficiencyFactor; + double VRatioClamped; if (GPUVMEnable == true && HostVMEnable == true) { HostVMInefficiencyFactor = @@ -3222,31 +3223,32 @@ static void CalculateFlipSchedule( *DestinationLinesToRequestRowInImmediateFlip = dml_ceil(4.0 * (TimeForFetchingRowInVBlankImmediateFlip / LineTime), 1) / 4.0; *final_flip_bw = dml_max(PDEAndMetaPTEBytesPerFrame * HostVMInefficiencyFactor / (*DestinationLinesToRequestVMInImmediateFlip * LineTime), (MetaRowBytes + DPTEBytesPerRow) * HostVMInefficiencyFactor / (*DestinationLinesToRequestRowInImmediateFlip * LineTime)); + VRatioClamped = (VRatio < 1.0) ? 1.0 : VRatio; if (SourcePixelFormat == dm_420_8 || SourcePixelFormat == dm_420_10) { if (GPUVMEnable == true && DCCEnable != true) { min_row_time = dml_min( - dpte_row_height * LineTime / VRatio, - dpte_row_height_chroma * LineTime / (VRatio / 2)); + dpte_row_height * LineTime / VRatioClamped, + dpte_row_height_chroma * LineTime / (VRatioClamped / 2)); } else if (GPUVMEnable != true && DCCEnable == true) { min_row_time = dml_min( - meta_row_height * LineTime / VRatio, - meta_row_height_chroma * LineTime / (VRatio / 2)); + meta_row_height * LineTime / VRatioClamped, + meta_row_height_chroma * LineTime / (VRatioClamped / 2)); } else { min_row_time = dml_min4( - dpte_row_height * LineTime / VRatio, - meta_row_height * LineTime / VRatio, - dpte_row_height_chroma * LineTime / (VRatio / 2), - meta_row_height_chroma * LineTime / (VRatio / 2)); + dpte_row_height * LineTime / VRatioClamped, + meta_row_height * LineTime / VRatioClamped, + dpte_row_height_chroma * LineTime / (VRatioClamped / 2), + meta_row_height_chroma * LineTime / (VRatioClamped / 2)); } } else { if (GPUVMEnable == true && DCCEnable != true) { - min_row_time = dpte_row_height * LineTime / VRatio; + min_row_time = dpte_row_height * LineTime / VRatioClamped; } else if (GPUVMEnable != true && DCCEnable == true) { - min_row_time = meta_row_height * LineTime / VRatio; + min_row_time = meta_row_height * LineTime / VRatioClamped; } else { min_row_time = dml_min( - dpte_row_height * LineTime / VRatio, - meta_row_height * LineTime / VRatio); + dpte_row_height * LineTime / VRatioClamped, + meta_row_height * LineTime / VRatioClamped); } } diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn21/display_rq_dlg_calc_21.c b/drivers/gpu/drm/amd/display/dc/dml/dcn21/display_rq_dlg_calc_21.c index 193f31b8ac4a..90a5fefef05b 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn21/display_rq_dlg_calc_21.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn21/display_rq_dlg_calc_21.c @@ -1200,7 +1200,7 @@ static void dml_rq_dlg_get_dlg_params( min_hratio_fact_l = 1.0; min_hratio_fact_c = 1.0; - if (htaps_l <= 1) + if (hratio_l <= 1) min_hratio_fact_l = 2.0; else if (htaps_l <= 6) { if ((hratio_l * 2.0) > 4.0) @@ -1216,7 +1216,7 @@ static void dml_rq_dlg_get_dlg_params( hscale_pixel_rate_l = min_hratio_fact_l * dppclk_freq_in_mhz; - if (htaps_c <= 1) + if (hratio_c <= 1) min_hratio_fact_c = 2.0; else if (htaps_c <= 6) { if ((hratio_c * 2.0) > 4.0) @@ -1533,8 +1533,8 @@ static void dml_rq_dlg_get_dlg_params( disp_dlg_regs->refcyc_per_vm_group_vblank = get_refcyc_per_vm_group_vblank(mode_lib, e2e_pipe_param, num_pipes, pipe_idx) * refclk_freq_in_mhz; disp_dlg_regs->refcyc_per_vm_group_flip = get_refcyc_per_vm_group_flip(mode_lib, e2e_pipe_param, num_pipes, pipe_idx) * refclk_freq_in_mhz; - disp_dlg_regs->refcyc_per_vm_req_vblank = get_refcyc_per_vm_req_vblank(mode_lib, e2e_pipe_param, num_pipes, pipe_idx) * refclk_freq_in_mhz; - disp_dlg_regs->refcyc_per_vm_req_flip = get_refcyc_per_vm_req_flip(mode_lib, e2e_pipe_param, num_pipes, pipe_idx) * refclk_freq_in_mhz; + disp_dlg_regs->refcyc_per_vm_req_vblank = get_refcyc_per_vm_req_vblank(mode_lib, e2e_pipe_param, num_pipes, pipe_idx) * refclk_freq_in_mhz * dml_pow(2, 10); + disp_dlg_regs->refcyc_per_vm_req_flip = get_refcyc_per_vm_req_flip(mode_lib, e2e_pipe_param, num_pipes, pipe_idx) * refclk_freq_in_mhz * dml_pow(2, 10); // Clamp to max for now if (disp_dlg_regs->refcyc_per_vm_group_vblank >= (unsigned int)dml_pow(2, 23)) 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 b2ecb174a93f..439ffd04be34 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 @@ -118,6 +118,7 @@ struct _vcs_dpi_soc_bounding_box_st { double urgent_latency_adjustment_fabric_clock_component_us; double urgent_latency_adjustment_fabric_clock_reference_mhz; bool disable_dram_clock_change_vactive_support; + bool allow_dram_clock_one_display_vactive; }; struct _vcs_dpi_ip_params_st { diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.c b/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.c index 6e4e8a452e66..b19988f54721 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.c +++ b/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.c @@ -224,6 +224,7 @@ static void fetch_socbb_params(struct display_mode_lib *mode_lib) mode_lib->vba.DummyPStateCheck = soc->dram_clock_change_latency_us == soc->dummy_pstate_latency_us; mode_lib->vba.DRAMClockChangeSupportsVActive = !soc->disable_dram_clock_change_vactive_support || mode_lib->vba.DummyPStateCheck; + mode_lib->vba.AllowDramClockChangeOneDisplayVactive = soc->allow_dram_clock_one_display_vactive; mode_lib->vba.Downspreading = soc->downspread_percent; mode_lib->vba.DRAMChannelWidth = soc->dram_channel_width_bytes; // new! diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.h b/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.h index a1884ffe63ae..6a7b20927a6b 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.h +++ b/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.h @@ -899,6 +899,7 @@ struct vba_vars_st { double BPP; enum odm_combine_policy ODMCombinePolicy; bool UseMinimumRequiredDCFCLK; + bool AllowDramClockChangeOneDisplayVactive; }; bool CalculateMinAndMaxPrefetchMode( diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/abm.h b/drivers/gpu/drm/amd/display/dc/inc/hw/abm.h index 91a42f0f1fc4..e8ce8c85adf1 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/abm.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/abm.h @@ -27,27 +27,17 @@ #include "dm_services_types.h" -struct abm_backlight_registers { - unsigned int BL_PWM_CNTL; - unsigned int BL_PWM_CNTL2; - unsigned int BL_PWM_PERIOD_CNTL; - unsigned int LVTMA_PWRSEQ_REF_DIV_BL_PWM_REF_DIV; -}; - struct abm { struct dc_context *ctx; const struct abm_funcs *funcs; bool dmcu_is_running; - /* registers setting needs to be saved and restored at InitBacklight */ - struct abm_backlight_registers stored_backlight_registers; }; struct abm_funcs { - void (*abm_init)(struct abm *abm); + void (*abm_init)(struct abm *abm, uint32_t back_light); bool (*set_abm_level)(struct abm *abm, unsigned int abm_level); bool (*set_abm_immediate_disable)(struct abm *abm, unsigned int panel_inst); bool (*set_pipe)(struct abm *abm, unsigned int controller_id, unsigned int panel_inst); - bool (*init_backlight)(struct abm *abm); /* backlight_pwm_u16_16 is unsigned 32 bit, * 16 bit integer + 16 fractional, where 1.0 is max backlight value. @@ -56,8 +46,7 @@ struct abm_funcs { unsigned int backlight_pwm_u16_16, unsigned int frame_ramp, unsigned int controller_id, - unsigned int panel_inst, - bool fw_set_brightness); + unsigned int panel_inst); unsigned int (*get_current_backlight)(struct abm *abm); unsigned int (*get_target_backlight)(struct abm *abm); diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/panel_cntl.h b/drivers/gpu/drm/amd/display/dc/inc/hw/panel_cntl.h index 5a75d035f1fa..f9ab5abb6462 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/panel_cntl.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/panel_cntl.h @@ -32,11 +32,23 @@ #include "dc_types.h" +#define MAX_BACKLIGHT_LEVEL 0xFFFF + +struct panel_cntl_backlight_registers { + unsigned int BL_PWM_CNTL; + unsigned int BL_PWM_CNTL2; + unsigned int BL_PWM_PERIOD_CNTL; + unsigned int LVTMA_PWRSEQ_REF_DIV_BL_PWM_REF_DIV; +}; + struct panel_cntl_funcs { void (*destroy)(struct panel_cntl **panel_cntl); - void (*hw_init)(struct panel_cntl *panel_cntl); + uint32_t (*hw_init)(struct panel_cntl *panel_cntl); bool (*is_panel_backlight_on)(struct panel_cntl *panel_cntl); bool (*is_panel_powered_on)(struct panel_cntl *panel_cntl); + void (*store_backlight_level)(struct panel_cntl *panel_cntl); + void (*driver_set_backlight)(struct panel_cntl *panel_cntl, + uint32_t backlight_pwm_u16_16); }; struct panel_cntl_init_data { @@ -48,6 +60,8 @@ struct panel_cntl { const struct panel_cntl_funcs *funcs; struct dc_context *ctx; uint32_t inst; + /* registers setting needs to be saved and restored at InitBacklight */ + struct panel_cntl_backlight_registers stored_backlight_registers; }; #endif /* DC_PANEL_CNTL_H_ */ 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 1570fed20de0..3b2ea9bdb62c 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h @@ -196,6 +196,8 @@ struct hw_sequencer_funcs { uint32_t backlight_pwm_u16_16, uint32_t frame_ramp); + void (*set_abm_immediate_disable)(struct pipe_ctx *pipe_ctx); + }; diff --git a/drivers/gpu/drm/amd/display/dc/inc/resource.h b/drivers/gpu/drm/amd/display/dc/inc/resource.h index ca4c36c0c9bc..109c589eb97c 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/resource.h +++ b/drivers/gpu/drm/amd/display/dc/inc/resource.h @@ -138,9 +138,6 @@ struct pipe_ctx *find_idle_secondary_pipe( const struct resource_pool *pool, const struct pipe_ctx *primary_pipe); -bool resource_is_stream_unchanged( - struct dc_state *old_context, struct dc_stream_state *stream); - bool resource_validate_attach_surfaces( const struct dc_validation_set set[], int set_count, diff --git a/drivers/gpu/drm/amd/display/dc/os_types.h b/drivers/gpu/drm/amd/display/dc/os_types.h index c34eba19860a..6d7bca562eec 100644 --- a/drivers/gpu/drm/amd/display/dc/os_types.h +++ b/drivers/gpu/drm/amd/display/dc/os_types.h @@ -108,7 +108,7 @@ #define ASSERT(expr) ASSERT_CRITICAL(expr) #else -#define ASSERT(expr) WARN_ON(!(expr)) +#define ASSERT(expr) WARN_ON_ONCE(!(expr)) #endif #define BREAK_TO_DEBUGGER() ASSERT(0) diff --git a/drivers/gpu/drm/amd/display/dmub/inc/dmub_srv.h b/drivers/gpu/drm/amd/display/dmub/dmub_srv.h index 9f10bfe3d2df..26d94eb5ab58 100644 --- a/drivers/gpu/drm/amd/display/dmub/inc/dmub_srv.h +++ b/drivers/gpu/drm/amd/display/dmub/dmub_srv.h @@ -64,10 +64,11 @@ * other component within DAL. */ -#include "dmub_types.h" -#include "dmub_cmd.h" -#include "dmub_gpint_cmd.h" -#include "dmub_rb.h" +#include "inc/dmub_types.h" +#include "inc/dmub_cmd.h" +#include "inc/dmub_gpint_cmd.h" +#include "inc/dmub_cmd_dal.h" +#include "inc/dmub_rb.h" #if defined(__cplusplus) extern "C" { diff --git a/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h b/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h index eb10518dc058..599bf2055bcb 100644 --- a/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h +++ b/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h @@ -262,6 +262,7 @@ struct dmub_cmd_abm_set_pipe_data { uint32_t ramping_boundary; uint32_t otg_inst; uint32_t panel_inst; + uint32_t set_pipe_option; }; struct dmub_rb_cmd_abm_set_pipe { diff --git a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn20.c b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn20.c index 45638d61b73d..edc73d6d7ba2 100644 --- a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn20.c +++ b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn20.c @@ -23,7 +23,7 @@ * */ -#include "../inc/dmub_srv.h" +#include "../dmub_srv.h" #include "dmub_reg.h" #include "dmub_dcn20.h" diff --git a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn21.c b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn21.c index 5bed9fcd6b5c..e8f488232e34 100644 --- a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn21.c +++ b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn21.c @@ -23,7 +23,7 @@ * */ -#include "../inc/dmub_srv.h" +#include "../dmub_srv.h" #include "dmub_reg.h" #include "dmub_dcn21.h" diff --git a/drivers/gpu/drm/amd/display/dmub/src/dmub_reg.c b/drivers/gpu/drm/amd/display/dmub/src/dmub_reg.c index 4094eca212f0..ca0c8a54b635 100644 --- a/drivers/gpu/drm/amd/display/dmub/src/dmub_reg.c +++ b/drivers/gpu/drm/amd/display/dmub/src/dmub_reg.c @@ -24,7 +24,7 @@ */ #include "dmub_reg.h" -#include "../inc/dmub_srv.h" +#include "../dmub_srv.h" struct dmub_reg_value_masks { uint32_t value; diff --git a/drivers/gpu/drm/amd/display/dmub/src/dmub_srv.c b/drivers/gpu/drm/amd/display/dmub/src/dmub_srv.c index 66ba0b2d80d2..0e3751d94cb0 100644 --- a/drivers/gpu/drm/amd/display/dmub/src/dmub_srv.c +++ b/drivers/gpu/drm/amd/display/dmub/src/dmub_srv.c @@ -23,7 +23,7 @@ * */ -#include "../inc/dmub_srv.h" +#include "../dmub_srv.h" #include "dmub_dcn20.h" #include "dmub_dcn21.h" #include "dmub_fw_meta.h" diff --git a/drivers/gpu/drm/amd/display/modules/info_packet/info_packet.c b/drivers/gpu/drm/amd/display/modules/info_packet/info_packet.c index 22a5484d9f28..7cd8a43d1889 100644 --- a/drivers/gpu/drm/amd/display/modules/info_packet/info_packet.c +++ b/drivers/gpu/drm/amd/display/modules/info_packet/info_packet.c @@ -29,7 +29,6 @@ #include "mod_shared.h" #include "mod_freesync.h" #include "dc.h" -#include "dmub/inc/dmub_cmd_dal.h" enum vsc_packet_revision { vsc_packet_undefined = 0, @@ -145,7 +144,7 @@ void mod_build_vsc_infopacket(const struct dc_stream_state *stream, } /*VSC packet set to 2 when DP revision >= 1.2*/ - if (stream->link->psr_settings.psr_version != PSR_VERSION_UNSUPPORTED) + if (stream->link->psr_settings.psr_version != DC_PSR_VERSION_UNSUPPORTED) vsc_packet_revision = vsc_packet_rev2; /* Update to revision 5 for extended colorimetry support */ diff --git a/drivers/gpu/drm/amd/display/modules/stats/stats.c b/drivers/gpu/drm/amd/display/modules/stats/stats.c deleted file mode 100644 index 03121ca64fe4..000000000000 --- a/drivers/gpu/drm/amd/display/modules/stats/stats.c +++ /dev/null @@ -1,448 +0,0 @@ -/* - * 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 - * - */ - -#include "mod_stats.h" -#include "dm_services.h" -#include "dc.h" -#include "core_types.h" - -#define DAL_STATS_ENABLE_REGKEY "DalStatsEnable" -#define DAL_STATS_ENABLE_REGKEY_DEFAULT 0x00000000 -#define DAL_STATS_ENABLE_REGKEY_ENABLED 0x00000001 - -#define DAL_STATS_ENTRIES_REGKEY "DalStatsEntries" -#define DAL_STATS_ENTRIES_REGKEY_DEFAULT 0x00350000 -#define DAL_STATS_ENTRIES_REGKEY_MAX 0x01000000 - -#define DAL_STATS_EVENT_ENTRIES_DEFAULT 0x00000100 - -#define MOD_STATS_NUM_VSYNCS 5 -#define MOD_STATS_EVENT_STRING_MAX 512 - -struct stats_time_cache { - unsigned int entry_id; - - unsigned long flip_timestamp_in_ns; - unsigned long vupdate_timestamp_in_ns; - - unsigned int render_time_in_us; - unsigned int avg_render_time_in_us_last_ten; - unsigned int v_sync_time_in_us[MOD_STATS_NUM_VSYNCS]; - unsigned int num_vsync_between_flips; - - unsigned int flip_to_vsync_time_in_us; - unsigned int vsync_to_flip_time_in_us; - - unsigned int min_window; - unsigned int max_window; - unsigned int v_total_min; - unsigned int v_total_max; - unsigned int event_triggers; - - unsigned int lfc_mid_point_in_us; - unsigned int num_frames_inserted; - unsigned int inserted_duration_in_us; - - unsigned int flags; -}; - -struct stats_event_cache { - unsigned int entry_id; - char event_string[MOD_STATS_EVENT_STRING_MAX]; -}; - -struct core_stats { - struct mod_stats public; - struct dc *dc; - - bool enabled; - unsigned int entries; - unsigned int event_entries; - unsigned int entry_id; - - struct stats_time_cache *time; - unsigned int index; - - struct stats_event_cache *events; - unsigned int event_index; - -}; - -#define MOD_STATS_TO_CORE(mod_stats)\ - container_of(mod_stats, struct core_stats, public) - -bool mod_stats_init(struct mod_stats *mod_stats) -{ - bool result = false; - struct core_stats *core_stats = NULL; - struct dc *dc = NULL; - - if (mod_stats == NULL) - return false; - - core_stats = MOD_STATS_TO_CORE(mod_stats); - dc = core_stats->dc; - - return result; -} - -struct mod_stats *mod_stats_create(struct dc *dc) -{ - struct core_stats *core_stats = NULL; - struct persistent_data_flag flag; - unsigned int reg_data; - int i = 0; - - if (dc == NULL) - goto fail_construct; - - core_stats = kzalloc(sizeof(struct core_stats), GFP_KERNEL); - - if (core_stats == NULL) - goto fail_construct; - - core_stats->dc = dc; - - core_stats->enabled = DAL_STATS_ENABLE_REGKEY_DEFAULT; - if (dm_read_persistent_data(dc->ctx, NULL, NULL, - DAL_STATS_ENABLE_REGKEY, - ®_data, sizeof(unsigned int), &flag)) - core_stats->enabled = reg_data; - - if (core_stats->enabled) { - core_stats->entries = DAL_STATS_ENTRIES_REGKEY_DEFAULT; - if (dm_read_persistent_data(dc->ctx, NULL, NULL, - DAL_STATS_ENTRIES_REGKEY, - ®_data, sizeof(unsigned int), &flag)) { - if (reg_data > DAL_STATS_ENTRIES_REGKEY_MAX) - core_stats->entries = DAL_STATS_ENTRIES_REGKEY_MAX; - else - core_stats->entries = reg_data; - } - core_stats->time = kcalloc(core_stats->entries, - sizeof(struct stats_time_cache), - GFP_KERNEL); - - if (core_stats->time == NULL) - goto fail_construct_time; - - core_stats->event_entries = DAL_STATS_EVENT_ENTRIES_DEFAULT; - core_stats->events = kcalloc(core_stats->event_entries, - sizeof(struct stats_event_cache), - GFP_KERNEL); - - if (core_stats->events == NULL) - goto fail_construct_events; - - } else { - core_stats->entries = 0; - } - - /* Purposely leave index 0 unused so we don't need special logic to - * handle calculation cases that depend on previous flip data. - */ - core_stats->index = 1; - core_stats->event_index = 0; - - // Keeps track of ordering within the different stats structures - core_stats->entry_id = 0; - - return &core_stats->public; - -fail_construct_events: - kfree(core_stats->time); - -fail_construct_time: - kfree(core_stats); - -fail_construct: - return NULL; -} - -void mod_stats_destroy(struct mod_stats *mod_stats) -{ - if (mod_stats != NULL) { - struct core_stats *core_stats = MOD_STATS_TO_CORE(mod_stats); - - kfree(core_stats->time); - kfree(core_stats->events); - kfree(core_stats); - } -} - -void mod_stats_dump(struct mod_stats *mod_stats) -{ - struct dc *dc = NULL; - struct dal_logger *logger = NULL; - struct core_stats *core_stats = NULL; - struct stats_time_cache *time = NULL; - struct stats_event_cache *events = NULL; - unsigned int time_index = 1; - unsigned int event_index = 0; - unsigned int index = 0; - struct log_entry log_entry; - - if (mod_stats == NULL) - return; - - core_stats = MOD_STATS_TO_CORE(mod_stats); - dc = core_stats->dc; - logger = dc->ctx->logger; - time = core_stats->time; - events = core_stats->events; - - DISPLAY_STATS_BEGIN(log_entry); - - DISPLAY_STATS("==Display Caps==\n"); - - DISPLAY_STATS("==Display Stats==\n"); - - DISPLAY_STATS("%10s %10s %10s %10s %10s" - " %11s %11s %17s %10s %14s" - " %10s %10s %10s %10s %10s" - " %10s %10s %10s %10s\n", - "render", "avgRender", - "minWindow", "midPoint", "maxWindow", - "vsyncToFlip", "flipToVsync", "vsyncsBetweenFlip", - "numFrame", "insertDuration", - "vTotalMin", "vTotalMax", "eventTrigs", - "vSyncTime1", "vSyncTime2", "vSyncTime3", - "vSyncTime4", "vSyncTime5", "flags"); - - 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("==Event==%s\n", events[event_index].event_string); - event_index++; - } else if (time_index < core_stats->index && - i == time[time_index].entry_id) { - DISPLAY_STATS("%10u %10u %10u %10u %10u" - " %11u %11u %17u %10u %14u" - " %10u %10u %10u %10u %10u" - " %10u %10u %10u %10u\n", - time[time_index].render_time_in_us, - time[time_index].avg_render_time_in_us_last_ten, - time[time_index].min_window, - time[time_index].lfc_mid_point_in_us, - time[time_index].max_window, - time[time_index].vsync_to_flip_time_in_us, - time[time_index].flip_to_vsync_time_in_us, - time[time_index].num_vsync_between_flips, - time[time_index].num_frames_inserted, - time[time_index].inserted_duration_in_us, - time[time_index].v_total_min, - time[time_index].v_total_max, - time[time_index].event_triggers, - time[time_index].v_sync_time_in_us[0], - time[time_index].v_sync_time_in_us[1], - time[time_index].v_sync_time_in_us[2], - time[time_index].v_sync_time_in_us[3], - time[time_index].v_sync_time_in_us[4], - time[time_index].flags); - - time_index++; - } - } - - DISPLAY_STATS_END(log_entry); -} - -void mod_stats_reset_data(struct mod_stats *mod_stats) -{ - struct core_stats *core_stats = NULL; - struct stats_time_cache *time = NULL; - unsigned int index = 0; - - if (mod_stats == NULL) - return; - - core_stats = MOD_STATS_TO_CORE(mod_stats); - - memset(core_stats->time, 0, - sizeof(struct stats_time_cache) * core_stats->entries); - - memset(core_stats->events, 0, - sizeof(struct stats_event_cache) * core_stats->event_entries); - - core_stats->index = 1; - core_stats->event_index = 0; - - // Keeps track of ordering within the different stats structures - core_stats->entry_id = 0; -} - -void mod_stats_update_event(struct mod_stats *mod_stats, - char *event_string, - unsigned int length) -{ - struct core_stats *core_stats = NULL; - struct stats_event_cache *events = NULL; - unsigned int index = 0; - unsigned int copy_length = 0; - - if (mod_stats == NULL) - return; - - core_stats = MOD_STATS_TO_CORE(mod_stats); - - if (core_stats->event_index >= core_stats->event_entries) - return; - - events = core_stats->events; - index = core_stats->event_index; - - copy_length = length; - if (length > MOD_STATS_EVENT_STRING_MAX) - copy_length = MOD_STATS_EVENT_STRING_MAX; - - memcpy(&events[index].event_string, event_string, copy_length); - events[index].event_string[copy_length - 1] = '\0'; - - events[index].entry_id = core_stats->entry_id; - core_stats->event_index++; - core_stats->entry_id++; -} - -void mod_stats_update_flip(struct mod_stats *mod_stats, - unsigned long timestamp_in_ns) -{ - struct core_stats *core_stats = NULL; - struct stats_time_cache *time = NULL; - unsigned int index = 0; - - if (mod_stats == NULL) - return; - - core_stats = MOD_STATS_TO_CORE(mod_stats); - - if (core_stats->index >= core_stats->entries) - return; - - time = core_stats->time; - index = core_stats->index; - - time[index].flip_timestamp_in_ns = timestamp_in_ns; - time[index].render_time_in_us = - (timestamp_in_ns - time[index - 1].flip_timestamp_in_ns) / 1000; - - if (index >= 10) { - for (unsigned int i = 0; i < 10; i++) - time[index].avg_render_time_in_us_last_ten += - time[index - i].render_time_in_us; - time[index].avg_render_time_in_us_last_ten /= 10; - } - - if (time[index].num_vsync_between_flips > 0) - time[index].vsync_to_flip_time_in_us = - (timestamp_in_ns - - time[index].vupdate_timestamp_in_ns) / 1000; - else - time[index].vsync_to_flip_time_in_us = - (timestamp_in_ns - - time[index - 1].vupdate_timestamp_in_ns) / 1000; - - time[index].entry_id = core_stats->entry_id; - core_stats->index++; - core_stats->entry_id++; -} - -void mod_stats_update_vupdate(struct mod_stats *mod_stats, - unsigned long timestamp_in_ns) -{ - struct core_stats *core_stats = NULL; - struct stats_time_cache *time = NULL; - unsigned int index = 0; - unsigned int num_vsyncs = 0; - unsigned int prev_vsync_in_ns = 0; - - if (mod_stats == NULL) - return; - - core_stats = MOD_STATS_TO_CORE(mod_stats); - - if (core_stats->index >= core_stats->entries) - return; - - time = core_stats->time; - index = core_stats->index; - num_vsyncs = time[index].num_vsync_between_flips; - - if (num_vsyncs < MOD_STATS_NUM_VSYNCS) { - if (num_vsyncs == 0) { - prev_vsync_in_ns = - time[index - 1].vupdate_timestamp_in_ns; - - time[index].flip_to_vsync_time_in_us = - (timestamp_in_ns - - time[index - 1].flip_timestamp_in_ns) / - 1000; - } else { - prev_vsync_in_ns = - time[index].vupdate_timestamp_in_ns; - } - - time[index].v_sync_time_in_us[num_vsyncs] = - (timestamp_in_ns - prev_vsync_in_ns) / 1000; - } - - time[index].vupdate_timestamp_in_ns = timestamp_in_ns; - time[index].num_vsync_between_flips++; -} - -void mod_stats_update_freesync(struct mod_stats *mod_stats, - unsigned int v_total_min, - unsigned int v_total_max, - unsigned int event_triggers, - unsigned int window_min, - unsigned int window_max, - unsigned int lfc_mid_point_in_us, - unsigned int inserted_frames, - unsigned int inserted_duration_in_us) -{ - struct core_stats *core_stats = NULL; - struct stats_time_cache *time = NULL; - unsigned int index = 0; - - if (mod_stats == NULL) - return; - - core_stats = MOD_STATS_TO_CORE(mod_stats); - - if (core_stats->index >= core_stats->entries) - return; - - time = core_stats->time; - index = core_stats->index; - - time[index].v_total_min = v_total_min; - time[index].v_total_max = v_total_max; - time[index].event_triggers = event_triggers; - time[index].min_window = window_min; - time[index].max_window = window_max; - time[index].lfc_mid_point_in_us = lfc_mid_point_in_us; - time[index].num_frames_inserted = inserted_frames; - time[index].inserted_duration_in_us = inserted_duration_in_us; -} - diff --git a/drivers/gpu/drm/amd/include/asic_reg/gc/gc_10_1_0_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/gc/gc_10_1_0_sh_mask.h index e7db6f9f9c86..8b0b9a2a8fed 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/gc/gc_10_1_0_sh_mask.h +++ b/drivers/gpu/drm/amd/include/asic_reg/gc/gc_10_1_0_sh_mask.h @@ -5599,6 +5599,7 @@ #define GRBM_PWR_CNTL__ALL_REQ_EN_MASK 0x00008000L //GRBM_STATUS #define GRBM_STATUS__ME0PIPE0_CMDFIFO_AVAIL__SHIFT 0x0 +#define GRBM_STATUS__RSMU_RQ_PENDING__SHIFT 0x5 #define GRBM_STATUS__ME0PIPE0_CF_RQ_PENDING__SHIFT 0x7 #define GRBM_STATUS__ME0PIPE0_PF_RQ_PENDING__SHIFT 0x8 #define GRBM_STATUS__GDS_DMA_RQ_PENDING__SHIFT 0x9 @@ -5619,6 +5620,7 @@ #define GRBM_STATUS__CB_BUSY__SHIFT 0x1e #define GRBM_STATUS__GUI_ACTIVE__SHIFT 0x1f #define GRBM_STATUS__ME0PIPE0_CMDFIFO_AVAIL_MASK 0x0000000FL +#define GRBM_STATUS__RSMU_RQ_PENDING_MASK 0x00000020L #define GRBM_STATUS__ME0PIPE0_CF_RQ_PENDING_MASK 0x00000080L #define GRBM_STATUS__ME0PIPE0_PF_RQ_PENDING_MASK 0x00000100L #define GRBM_STATUS__GDS_DMA_RQ_PENDING_MASK 0x00000200L @@ -5832,6 +5834,7 @@ #define GRBM_READ_ERROR__READ_ERROR_MASK 0x80000000L //GRBM_READ_ERROR2 #define GRBM_READ_ERROR2__READ_REQUESTER_CPF__SHIFT 0x10 +#define GRBM_READ_ERROR2__READ_REQUESTER_RSMU__SHIFT 0x11 #define GRBM_READ_ERROR2__READ_REQUESTER_RLC__SHIFT 0x12 #define GRBM_READ_ERROR2__READ_REQUESTER_GDS_DMA__SHIFT 0x13 #define GRBM_READ_ERROR2__READ_REQUESTER_ME0PIPE0_CF__SHIFT 0x14 @@ -5847,6 +5850,7 @@ #define GRBM_READ_ERROR2__READ_REQUESTER_ME2PIPE2__SHIFT 0x1e #define GRBM_READ_ERROR2__READ_REQUESTER_ME2PIPE3__SHIFT 0x1f #define GRBM_READ_ERROR2__READ_REQUESTER_CPF_MASK 0x00010000L +#define GRBM_READ_ERROR2__READ_REQUESTER_RSMU_MASK 0x00020000L #define GRBM_READ_ERROR2__READ_REQUESTER_RLC_MASK 0x00040000L #define GRBM_READ_ERROR2__READ_REQUESTER_GDS_DMA_MASK 0x00080000L #define GRBM_READ_ERROR2__READ_REQUESTER_ME0PIPE0_CF_MASK 0x00100000L diff --git a/drivers/gpu/drm/amd/include/atomfirmware.h b/drivers/gpu/drm/amd/include/atomfirmware.h index a6f6e6bf5992..b36ea8340afa 100644 --- a/drivers/gpu/drm/amd/include/atomfirmware.h +++ b/drivers/gpu/drm/amd/include/atomfirmware.h @@ -972,11 +972,13 @@ struct atom_ext_display_path }; //usCaps -enum ext_display_path_cap_def -{ - EXT_DISPLAY_PATH_CAPS__HBR2_DISABLE =0x0001, - EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN =0x0002, - EXT_DISPLAY_PATH_CAPS__EXT_CHIP_MASK =0x007C, +enum ext_display_path_cap_def { + EXT_DISPLAY_PATH_CAPS__HBR2_DISABLE = 0x0001, + EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN = 0x0002, + EXT_DISPLAY_PATH_CAPS__EXT_CHIP_MASK = 0x007C, + EXT_DISPLAY_PATH_CAPS__HDMI20_PI3EQX1204 = (0x01 << 2), //PI redriver chip + EXT_DISPLAY_PATH_CAPS__HDMI20_TISN65DP159RSBT = (0x02 << 2), //TI retimer chip + EXT_DISPLAY_PATH_CAPS__HDMI20_PARADE_PS175 = (0x03 << 2) //Parade DP->HDMI recoverter chip }; struct atom_external_display_connection_info diff --git a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c index fc31499c2e5c..7e6dcdf7df73 100644 --- a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c +++ b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c @@ -322,12 +322,12 @@ static void pp_dpm_en_umd_pstate(struct pp_hwmgr *hwmgr, if (*level & profile_mode_mask) { hwmgr->saved_dpm_level = hwmgr->dpm_level; hwmgr->en_umd_pstate = true; - amdgpu_device_ip_set_clockgating_state(hwmgr->adev, - AMD_IP_BLOCK_TYPE_GFX, - AMD_CG_STATE_UNGATE); amdgpu_device_ip_set_powergating_state(hwmgr->adev, AMD_IP_BLOCK_TYPE_GFX, AMD_PG_STATE_UNGATE); + amdgpu_device_ip_set_clockgating_state(hwmgr->adev, + AMD_IP_BLOCK_TYPE_GFX, + AMD_CG_STATE_UNGATE); } } else { /* exit umd pstate, restore level, enable gfx cg*/ diff --git a/drivers/gpu/drm/amd/powerplay/amdgpu_smu.c b/drivers/gpu/drm/amd/powerplay/amdgpu_smu.c index 5964d6323a13..de14542de775 100644 --- a/drivers/gpu/drm/amd/powerplay/amdgpu_smu.c +++ b/drivers/gpu/drm/amd/powerplay/amdgpu_smu.c @@ -1498,7 +1498,7 @@ static int smu_disable_dpm(struct smu_context *smu) bool use_baco = !smu->is_apu && ((adev->in_gpu_reset && (amdgpu_asic_reset_method(adev) == AMD_RESET_METHOD_BACO)) || - (adev->in_runpm && amdgpu_asic_supports_baco(adev))); + ((adev->in_runpm || adev->in_hibernate) && amdgpu_asic_supports_baco(adev))); ret = smu_get_smc_version(smu, NULL, &smu_version); if (ret) { @@ -1784,12 +1784,12 @@ static int smu_enable_umd_pstate(void *handle, if (*level & profile_mode_mask) { smu_dpm_ctx->saved_dpm_level = smu_dpm_ctx->dpm_level; smu_dpm_ctx->enable_umd_pstate = true; - amdgpu_device_ip_set_clockgating_state(smu->adev, - AMD_IP_BLOCK_TYPE_GFX, - AMD_CG_STATE_UNGATE); amdgpu_device_ip_set_powergating_state(smu->adev, AMD_IP_BLOCK_TYPE_GFX, AMD_PG_STATE_UNGATE); + amdgpu_device_ip_set_clockgating_state(smu->adev, + AMD_IP_BLOCK_TYPE_GFX, + AMD_CG_STATE_UNGATE); } } else { /* exit umd pstate, restore level, enable gfx cg*/ diff --git a/drivers/gpu/drm/amd/powerplay/arcturus_ppt.c b/drivers/gpu/drm/amd/powerplay/arcturus_ppt.c index 1c66b7d7139c..cfae4bcaf32e 100644 --- a/drivers/gpu/drm/amd/powerplay/arcturus_ppt.c +++ b/drivers/gpu/drm/amd/powerplay/arcturus_ppt.c @@ -2251,7 +2251,7 @@ static bool arcturus_is_baco_supported(struct smu_context *smu) struct amdgpu_device *adev = smu->adev; uint32_t val; - if (!smu_v11_0_baco_is_support(smu)) + if (!smu_v11_0_baco_is_support(smu) || amdgpu_sriov_vf(adev)) return false; val = RREG32_SOC15(NBIO, 0, mmRCC_BIF_STRAP0); diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c index 1cc30f750c26..4f8c1b85e688 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c @@ -1357,7 +1357,6 @@ static int smu10_asic_reset(struct pp_hwmgr *hwmgr, enum SMU_ASIC_RESET_MODE mod static const struct pp_hwmgr_func smu10_hwmgr_funcs = { .backend_init = smu10_hwmgr_backend_init, .backend_fini = smu10_hwmgr_backend_fini, - .asic_setup = NULL, .apply_state_adjust_rules = smu10_apply_state_adjust_rules, .force_dpm_level = smu10_dpm_force_dpm_level, .get_power_state_size = smu10_get_power_state_size, diff --git a/drivers/gpu/drm/amd/powerplay/navi10_ppt.c b/drivers/gpu/drm/amd/powerplay/navi10_ppt.c index 2184d247a9f7..0c9be864d072 100644 --- a/drivers/gpu/drm/amd/powerplay/navi10_ppt.c +++ b/drivers/gpu/drm/amd/powerplay/navi10_ppt.c @@ -1293,8 +1293,6 @@ static int navi10_set_power_profile_mode(struct smu_context *smu, long *input, u } if (smu->power_profile_mode == PP_SMC_POWER_PROFILE_CUSTOM) { - if (size < 0) - return -EINVAL; ret = smu_update_table(smu, SMU_TABLE_ACTIVITY_MONITOR_COEFF, WORKLOAD_PPLIB_CUSTOM_BIT, diff --git a/drivers/gpu/drm/radeon/atom.c b/drivers/gpu/drm/radeon/atom.c index 2c27627b6659..f15b20da5315 100644 --- a/drivers/gpu/drm/radeon/atom.c +++ b/drivers/gpu/drm/radeon/atom.c @@ -1211,8 +1211,7 @@ static int atom_execute_table_locked(struct atom_context *ctx, int index, uint32 SDEBUG("<<\n"); free: - if (ws) - kfree(ectx.ws); + kfree(ectx.ws); return ret; } diff --git a/drivers/gpu/drm/radeon/radeon_kms.c b/drivers/gpu/drm/radeon/radeon_kms.c index 58176db85952..95006cbf42c3 100644 --- a/drivers/gpu/drm/radeon/radeon_kms.c +++ b/drivers/gpu/drm/radeon/radeon_kms.c @@ -828,7 +828,7 @@ int radeon_enable_vblank_kms(struct drm_crtc *crtc) unsigned long irqflags; int r; - if (pipe < 0 || pipe >= rdev->num_crtc) { + if (pipe >= rdev->num_crtc) { DRM_ERROR("Invalid crtc %d\n", pipe); return -EINVAL; } @@ -854,7 +854,7 @@ void radeon_disable_vblank_kms(struct drm_crtc *crtc) struct radeon_device *rdev = dev->dev_private; unsigned long irqflags; - if (pipe < 0 || pipe >= rdev->num_crtc) { + if (pipe >= rdev->num_crtc) { DRM_ERROR("Invalid crtc %d\n", pipe); return; } diff --git a/include/uapi/linux/kfd_ioctl.h b/include/uapi/linux/kfd_ioctl.h index 4f6676428c5c..b6be62356d34 100644 --- a/include/uapi/linux/kfd_ioctl.h +++ b/include/uapi/linux/kfd_ioctl.h @@ -251,7 +251,7 @@ struct kfd_memory_exception_failure { __u32 imprecise; /* Can't determine the exact fault address */ }; -/* memory exception data*/ +/* memory exception data */ struct kfd_hsa_memory_exception_data { struct kfd_memory_exception_failure failure; __u64 va; |