diff options
Diffstat (limited to 'drivers/gpu/drm/amd')
534 files changed, 15442 insertions, 3809 deletions
diff --git a/drivers/gpu/drm/amd/amdgpu/Makefile b/drivers/gpu/drm/amd/amdgpu/Makefile index 4536c8ad0e11..078f588e99eb 100644 --- a/drivers/gpu/drm/amd/amdgpu/Makefile +++ b/drivers/gpu/drm/amd/amdgpu/Makefile @@ -23,7 +23,7 @@ # Makefile for the drm device driver. This driver provides support for the # Direct Rendering Infrastructure (DRI) in XFree86 4.1.0 and higher. -FULL_AMD_PATH=$(srctree)/$(src)/.. +FULL_AMD_PATH=$(src)/.. DISPLAY_FOLDER_NAME=display FULL_AMD_DISPLAY_PATH = $(FULL_AMD_PATH)/$(DISPLAY_FOLDER_NAME) @@ -70,7 +70,8 @@ amdgpu-y += amdgpu_device.o amdgpu_doorbell_mgr.o amdgpu_kms.o \ amdgpu_cs.o amdgpu_bios.o amdgpu_benchmark.o \ atombios_dp.o amdgpu_afmt.o amdgpu_trace_points.o \ atombios_encoders.o amdgpu_sa.o atombios_i2c.o \ - amdgpu_dma_buf.o amdgpu_vm.o amdgpu_vm_pt.o amdgpu_ib.o amdgpu_pll.o \ + amdgpu_dma_buf.o amdgpu_vm.o amdgpu_vm_pt.o amdgpu_vm_tlb_fence.o \ + amdgpu_ib.o amdgpu_pll.o \ amdgpu_ucode.o amdgpu_bo_list.o amdgpu_ctx.o amdgpu_sync.o \ amdgpu_gtt_mgr.o amdgpu_preempt_mgr.o amdgpu_vram_mgr.o amdgpu_virt.o \ amdgpu_atomfirmware.o amdgpu_vf_error.o amdgpu_sched.o \ @@ -80,7 +81,7 @@ amdgpu-y += amdgpu_device.o amdgpu_doorbell_mgr.o amdgpu_kms.o \ amdgpu_umc.o smu_v11_0_i2c.o amdgpu_fru_eeprom.o amdgpu_rap.o \ amdgpu_fw_attestation.o amdgpu_securedisplay.o \ amdgpu_eeprom.o amdgpu_mca.o amdgpu_psp_ta.o amdgpu_lsdma.o \ - amdgpu_ring_mux.o amdgpu_xcp.o amdgpu_seq64.o amdgpu_aca.o + amdgpu_ring_mux.o amdgpu_xcp.o amdgpu_seq64.o amdgpu_aca.o amdgpu_dev_coredump.o amdgpu-$(CONFIG_PROC_FS) += amdgpu_fdinfo.o @@ -247,7 +248,8 @@ amdgpu-y += \ smuio_v11_0_6.o \ smuio_v13_0.o \ smuio_v13_0_3.o \ - smuio_v13_0_6.o + smuio_v13_0_6.o \ + smuio_v14_0_2.o # add reset block amdgpu-y += \ diff --git a/drivers/gpu/drm/amd/amdgpu/aldebaran.c b/drivers/gpu/drm/amd/amdgpu/aldebaran.c index 576067d66bb9..d0a8da67dc2a 100644 --- a/drivers/gpu/drm/amd/amdgpu/aldebaran.c +++ b/drivers/gpu/drm/amd/amdgpu/aldebaran.c @@ -97,7 +97,7 @@ static int aldebaran_mode2_suspend_ip(struct amdgpu_device *adev) adev->ip_blocks[i].status.hw = false; } - return r; + return 0; } static int diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index 9c62552bec34..f87d53e183c3 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -139,6 +139,14 @@ enum amdgpu_ss { AMDGPU_SS_DRV_UNLOAD }; +struct amdgpu_hwip_reg_entry { + u32 hwip; + u32 inst; + u32 seg; + u32 reg_offset; + const char *reg_name; +}; + struct amdgpu_watchdog_timer { bool timeout_fatal_disable; uint32_t period; /* maxCycles = (1 << period), the number of cycles before a timeout */ @@ -210,6 +218,7 @@ extern int amdgpu_async_gfx_ring; extern int amdgpu_mcbp; extern int amdgpu_discovery; extern int amdgpu_mes; +extern int amdgpu_mes_log_enable; extern int amdgpu_mes_kiq; extern int amdgpu_noretry; extern int amdgpu_force_asic_type; @@ -493,6 +502,7 @@ struct amdgpu_wb { uint64_t gpu_addr; u32 num_wb; /* Number of wb slots actually reserved for amdgpu. */ unsigned long used[DIV_ROUND_UP(AMDGPU_MAX_WB, BITS_PER_LONG)]; + spinlock_t lock; }; int amdgpu_device_wb_get(struct amdgpu_device *adev, u32 *wb); @@ -605,7 +615,7 @@ struct amdgpu_asic_funcs { /* PCIe replay counter */ uint64_t (*get_pcie_replay_count)(struct amdgpu_device *adev); /* device supports BACO */ - bool (*supports_baco)(struct amdgpu_device *adev); + int (*supports_baco)(struct amdgpu_device *adev); /* pre asic_init quirks */ void (*pre_asic_init)(struct amdgpu_device *adev); /* enter/exit umd stable pstate */ @@ -1407,7 +1417,8 @@ bool amdgpu_device_supports_atpx(struct drm_device *dev); bool amdgpu_device_supports_px(struct drm_device *dev); bool amdgpu_device_supports_boco(struct drm_device *dev); bool amdgpu_device_supports_smart_shift(struct drm_device *dev); -bool amdgpu_device_supports_baco(struct drm_device *dev); +int amdgpu_device_supports_baco(struct drm_device *dev); +void amdgpu_device_detect_runtime_pm_mode(struct amdgpu_device *adev); bool amdgpu_device_is_peer_accessible(struct amdgpu_device *adev, struct amdgpu_device *peer_adev); int amdgpu_device_baco_enter(struct drm_device *dev); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_aca.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_aca.c index 493982f94649..c50202215f6b 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_aca.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_aca.c @@ -28,7 +28,7 @@ #define ACA_BANK_HWID(type, hwid, mcatype) [ACA_HWIP_TYPE_##type] = {hwid, mcatype} -typedef int bank_handler_t(struct aca_handle *handle, struct aca_bank *bank, enum aca_error_type type, void *data); +typedef int bank_handler_t(struct aca_handle *handle, struct aca_bank *bank, enum aca_smu_type type, void *data); struct aca_banks { int nr_banks; @@ -86,7 +86,7 @@ static void aca_banks_release(struct aca_banks *banks) } } -static int aca_smu_get_valid_aca_count(struct amdgpu_device *adev, enum aca_error_type type, u32 *count) +static int aca_smu_get_valid_aca_count(struct amdgpu_device *adev, enum aca_smu_type type, u32 *count) { struct amdgpu_aca *aca = &adev->aca; const struct aca_smu_funcs *smu_funcs = aca->smu_funcs; @@ -116,20 +116,22 @@ static struct aca_regs_dump { {"CONTROL_MASK", ACA_REG_IDX_CTL_MASK}, }; -static void aca_smu_bank_dump(struct amdgpu_device *adev, int idx, int total, struct aca_bank *bank) +static void aca_smu_bank_dump(struct amdgpu_device *adev, int idx, int total, struct aca_bank *bank, + struct ras_query_context *qctx) { + u64 event_id = qctx ? qctx->event_id : 0ULL; int i; - dev_info(adev->dev, HW_ERR "Accelerator Check Architecture events logged\n"); + RAS_EVENT_LOG(adev, event_id, HW_ERR "Accelerator Check Architecture events logged\n"); /* plus 1 for output format, e.g: ACA[08/08]: xxxx */ for (i = 0; i < ARRAY_SIZE(aca_regs); i++) - dev_info(adev->dev, HW_ERR "ACA[%02d/%02d].%s=0x%016llx\n", - idx + 1, total, aca_regs[i].name, bank->regs[aca_regs[i].reg_idx]); + RAS_EVENT_LOG(adev, event_id, HW_ERR "ACA[%02d/%02d].%s=0x%016llx\n", + idx + 1, total, aca_regs[i].name, bank->regs[aca_regs[i].reg_idx]); } -static int aca_smu_get_valid_aca_banks(struct amdgpu_device *adev, enum aca_error_type type, +static int aca_smu_get_valid_aca_banks(struct amdgpu_device *adev, enum aca_smu_type type, int start, int count, - struct aca_banks *banks) + struct aca_banks *banks, struct ras_query_context *qctx) { struct amdgpu_aca *aca = &adev->aca; const struct aca_smu_funcs *smu_funcs = aca->smu_funcs; @@ -143,13 +145,12 @@ static int aca_smu_get_valid_aca_banks(struct amdgpu_device *adev, enum aca_erro return -EOPNOTSUPP; switch (type) { - case ACA_ERROR_TYPE_UE: + case ACA_SMU_TYPE_UE: max_count = smu_funcs->max_ue_bank_count; break; - case ACA_ERROR_TYPE_CE: + case ACA_SMU_TYPE_CE: max_count = smu_funcs->max_ce_bank_count; break; - case ACA_ERROR_TYPE_DEFERRED: default: return -EINVAL; } @@ -164,7 +165,9 @@ static int aca_smu_get_valid_aca_banks(struct amdgpu_device *adev, enum aca_erro if (ret) return ret; - aca_smu_bank_dump(adev, i, count, &bank); + bank.type = type; + + aca_smu_bank_dump(adev, i, count, &bank, qctx); ret = aca_banks_add_bank(banks, &bank); if (ret) @@ -195,7 +198,7 @@ static bool aca_bank_hwip_is_matched(struct aca_bank *bank, enum aca_hwip_type t return hwip->hwid == hwid && hwip->mcatype == mcatype; } -static bool aca_bank_is_valid(struct aca_handle *handle, struct aca_bank *bank, enum aca_error_type type) +static bool aca_bank_is_valid(struct aca_handle *handle, struct aca_bank *bank, enum aca_smu_type type) { const struct aca_bank_ops *bank_ops = handle->bank_ops; @@ -273,59 +276,49 @@ static struct aca_bank_error *get_bank_error(struct aca_error *aerr, struct aca_ return new_bank_error(aerr, info); } -static int aca_log_errors(struct aca_handle *handle, enum aca_error_type type, - struct aca_bank_report *report) +int aca_error_cache_log_bank_error(struct aca_handle *handle, struct aca_bank_info *info, + enum aca_error_type type, u64 count) { struct aca_error_cache *error_cache = &handle->error_cache; struct aca_bank_error *bank_error; struct aca_error *aerr; - if (!handle || !report) + if (!handle || !info || type >= ACA_ERROR_TYPE_COUNT) return -EINVAL; - if (!report->count[type]) + if (!count) return 0; aerr = &error_cache->errors[type]; - bank_error = get_bank_error(aerr, &report->info); + bank_error = get_bank_error(aerr, info); if (!bank_error) return -ENOMEM; - bank_error->count[type] += report->count[type]; + bank_error->count += count; return 0; } -static int aca_generate_bank_report(struct aca_handle *handle, struct aca_bank *bank, - enum aca_error_type type, struct aca_bank_report *report) +static int aca_bank_parser(struct aca_handle *handle, struct aca_bank *bank, enum aca_smu_type type) { const struct aca_bank_ops *bank_ops = handle->bank_ops; - if (!bank || !report) + if (!bank) return -EINVAL; - if (!bank_ops->aca_bank_generate_report) + if (!bank_ops->aca_bank_parser) return -EOPNOTSUPP; - memset(report, 0, sizeof(*report)); - return bank_ops->aca_bank_generate_report(handle, bank, type, - report, handle->data); + return bank_ops->aca_bank_parser(handle, bank, type, + handle->data); } static int handler_aca_log_bank_error(struct aca_handle *handle, struct aca_bank *bank, - enum aca_error_type type, void *data) + enum aca_smu_type type, void *data) { - struct aca_bank_report report; int ret; - ret = aca_generate_bank_report(handle, bank, type, &report); - if (ret) - return ret; - - if (!report.count[type]) - return 0; - - ret = aca_log_errors(handle, type, &report); + ret = aca_bank_parser(handle, bank, type); if (ret) return ret; @@ -333,7 +326,7 @@ static int handler_aca_log_bank_error(struct aca_handle *handle, struct aca_bank } static int aca_dispatch_bank(struct aca_handle_manager *mgr, struct aca_bank *bank, - enum aca_error_type type, bank_handler_t handler, void *data) + enum aca_smu_type type, bank_handler_t handler, void *data) { struct aca_handle *handle; int ret; @@ -354,7 +347,7 @@ static int aca_dispatch_bank(struct aca_handle_manager *mgr, struct aca_bank *ba } static int aca_dispatch_banks(struct aca_handle_manager *mgr, struct aca_banks *banks, - enum aca_error_type type, bank_handler_t handler, void *data) + enum aca_smu_type type, bank_handler_t handler, void *data) { struct aca_bank_node *node; struct aca_bank *bank; @@ -378,8 +371,28 @@ static int aca_dispatch_banks(struct aca_handle_manager *mgr, struct aca_banks * return 0; } -static int aca_banks_update(struct amdgpu_device *adev, enum aca_error_type type, - bank_handler_t handler, void *data) +static bool aca_bank_should_update(struct amdgpu_device *adev, enum aca_smu_type type) +{ + struct amdgpu_aca *aca = &adev->aca; + bool ret = true; + + /* + * Because the UE Valid MCA count will only be cleared after reset, + * in order to avoid repeated counting of the error count, + * the aca bank is only updated once during the gpu recovery stage. + */ + if (type == ACA_SMU_TYPE_UE) { + if (amdgpu_ras_intr_triggered()) + ret = atomic_cmpxchg(&aca->ue_update_flag, 0, 1) == 0; + else + atomic_set(&aca->ue_update_flag, 0); + } + + return ret; +} + +static int aca_banks_update(struct amdgpu_device *adev, enum aca_smu_type type, + bank_handler_t handler, struct ras_query_context *qctx, void *data) { struct amdgpu_aca *aca = &adev->aca; struct aca_banks banks; @@ -389,9 +402,8 @@ static int aca_banks_update(struct amdgpu_device *adev, enum aca_error_type type if (list_empty(&aca->mgr.list)) return 0; - /* NOTE: pmfw is only support UE and CE */ - if (type == ACA_ERROR_TYPE_DEFERRED) - type = ACA_ERROR_TYPE_CE; + if (!aca_bank_should_update(adev, type)) + return 0; ret = aca_smu_get_valid_aca_count(adev, type, &count); if (ret) @@ -402,7 +414,7 @@ static int aca_banks_update(struct amdgpu_device *adev, enum aca_error_type type aca_banks_init(&banks); - ret = aca_smu_get_valid_aca_banks(adev, type, 0, count, &banks); + ret = aca_smu_get_valid_aca_banks(adev, type, 0, count, &banks, qctx); if (ret) goto err_release_banks; @@ -431,7 +443,7 @@ static int aca_log_aca_error_data(struct aca_bank_error *bank_error, enum aca_er if (type >= ACA_ERROR_TYPE_COUNT) return -EINVAL; - count = bank_error->count[type]; + count = bank_error->count; if (!count) return 0; @@ -447,6 +459,8 @@ static int aca_log_aca_error_data(struct aca_bank_error *bank_error, enum aca_er amdgpu_ras_error_statistic_ce_count(err_data, &mcm_info, NULL, count); break; case ACA_ERROR_TYPE_DEFERRED: + amdgpu_ras_error_statistic_de_count(err_data, &mcm_info, NULL, count); + break; default: break; } @@ -477,12 +491,25 @@ out_unlock: } static int __aca_get_error_data(struct amdgpu_device *adev, struct aca_handle *handle, enum aca_error_type type, - struct ras_err_data *err_data) + struct ras_err_data *err_data, struct ras_query_context *qctx) { + enum aca_smu_type smu_type; int ret; + switch (type) { + case ACA_ERROR_TYPE_UE: + smu_type = ACA_SMU_TYPE_UE; + break; + case ACA_ERROR_TYPE_CE: + case ACA_ERROR_TYPE_DEFERRED: + smu_type = ACA_SMU_TYPE_CE; + break; + default: + return -EINVAL; + } + /* udpate aca bank to aca source error_cache first */ - ret = aca_banks_update(adev, type, handler_aca_log_bank_error, NULL); + ret = aca_banks_update(adev, smu_type, handler_aca_log_bank_error, qctx, NULL); if (ret) return ret; @@ -498,10 +525,9 @@ static bool aca_handle_is_valid(struct aca_handle *handle) } int amdgpu_aca_get_error_data(struct amdgpu_device *adev, struct aca_handle *handle, - enum aca_error_type type, void *data) + enum aca_error_type type, struct ras_err_data *err_data, + struct ras_query_context *qctx) { - struct ras_err_data *err_data = (struct ras_err_data *)data; - if (!handle || !err_data) return -EINVAL; @@ -511,7 +537,7 @@ int amdgpu_aca_get_error_data(struct amdgpu_device *adev, struct aca_handle *han if (!(BIT(type) & handle->mask)) return 0; - return __aca_get_error_data(adev, handle, type, err_data); + return __aca_get_error_data(adev, handle, type, err_data, qctx); } static void aca_error_init(struct aca_error *aerr, enum aca_error_type type) @@ -668,6 +694,8 @@ int amdgpu_aca_init(struct amdgpu_device *adev) struct amdgpu_aca *aca = &adev->aca; int ret; + atomic_set(&aca->ue_update_flag, 0); + ret = aca_manager_init(&aca->mgr); if (ret) return ret; @@ -680,6 +708,8 @@ void amdgpu_aca_fini(struct amdgpu_device *adev) struct amdgpu_aca *aca = &adev->aca; aca_manager_fini(&aca->mgr); + + atomic_set(&aca->ue_update_flag, 0); } int amdgpu_aca_reset(struct amdgpu_device *adev) @@ -723,23 +753,13 @@ int aca_bank_info_decode(struct aca_bank *bank, struct aca_bank_info *info) static int aca_bank_get_error_code(struct amdgpu_device *adev, struct aca_bank *bank) { - int error_code; - - switch (amdgpu_ip_version(adev, MP1_HWIP, 0)) { - case IP_VERSION(13, 0, 6): - if (!(adev->flags & AMD_IS_APU) && adev->pm.fw_version >= 0x00555600) { - error_code = ACA_REG__SYND__ERRORINFORMATION(bank->regs[ACA_REG_IDX_SYND]); - return error_code & 0xff; - } - break; - default: - break; - } + struct amdgpu_aca *aca = &adev->aca; + const struct aca_smu_funcs *smu_funcs = aca->smu_funcs; - /* NOTE: the true error code is encoded in status.errorcode[0:7] */ - error_code = ACA_REG__STATUS__ERRORCODE(bank->regs[ACA_REG_IDX_STATUS]); + if (!smu_funcs || !smu_funcs->parse_error_code) + return -EOPNOTSUPP; - return error_code & 0xff; + return smu_funcs->parse_error_code(adev, bank); } int aca_bank_check_error_codes(struct amdgpu_device *adev, struct aca_bank *bank, int *err_codes, int size) @@ -750,6 +770,9 @@ int aca_bank_check_error_codes(struct amdgpu_device *adev, struct aca_bank *bank return -EINVAL; error_code = aca_bank_get_error_code(adev, bank); + if (error_code < 0) + return error_code; + for (i = 0; i < size; i++) { if (err_codes[i] == error_code) return 0; @@ -784,7 +807,7 @@ static int amdgpu_aca_smu_debug_mode_set(void *data, u64 val) return 0; } -static void aca_dump_entry(struct seq_file *m, struct aca_bank *bank, enum aca_error_type type, int idx) +static void aca_dump_entry(struct seq_file *m, struct aca_bank *bank, enum aca_smu_type type, int idx) { struct aca_bank_info info; int i, ret; @@ -793,7 +816,7 @@ static void aca_dump_entry(struct seq_file *m, struct aca_bank *bank, enum aca_e if (ret) return; - seq_printf(m, "aca entry[%d].type: %s\n", idx, type == ACA_ERROR_TYPE_UE ? "UE" : "CE"); + seq_printf(m, "aca entry[%d].type: %s\n", idx, type == ACA_SMU_TYPE_UE ? "UE" : "CE"); seq_printf(m, "aca entry[%d].info: socketid:%d aid:%d hwid:0x%03x mcatype:0x%04x\n", idx, info.socket_id, info.die_id, info.hwid, info.mcatype); @@ -807,7 +830,7 @@ struct aca_dump_context { }; static int handler_aca_bank_dump(struct aca_handle *handle, struct aca_bank *bank, - enum aca_error_type type, void *data) + enum aca_smu_type type, void *data) { struct aca_dump_context *ctx = (struct aca_dump_context *)data; @@ -816,7 +839,7 @@ static int handler_aca_bank_dump(struct aca_handle *handle, struct aca_bank *ban return handler_aca_log_bank_error(handle, bank, type, NULL); } -static int aca_dump_show(struct seq_file *m, enum aca_error_type type) +static int aca_dump_show(struct seq_file *m, enum aca_smu_type type) { struct amdgpu_device *adev = (struct amdgpu_device *)m->private; struct aca_dump_context context = { @@ -824,12 +847,12 @@ static int aca_dump_show(struct seq_file *m, enum aca_error_type type) .idx = 0, }; - return aca_banks_update(adev, type, handler_aca_bank_dump, (void *)&context); + return aca_banks_update(adev, type, handler_aca_bank_dump, NULL, (void *)&context); } static int aca_dump_ce_show(struct seq_file *m, void *unused) { - return aca_dump_show(m, ACA_ERROR_TYPE_CE); + return aca_dump_show(m, ACA_SMU_TYPE_CE); } static int aca_dump_ce_open(struct inode *inode, struct file *file) @@ -847,7 +870,7 @@ static const struct file_operations aca_ce_dump_debug_fops = { static int aca_dump_ue_show(struct seq_file *m, void *unused) { - return aca_dump_show(m, ACA_ERROR_TYPE_UE); + return aca_dump_show(m, ACA_SMU_TYPE_UE); } static int aca_dump_ue_open(struct inode *inode, struct file *file) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_aca.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_aca.h index 2da50e095883..5ef6b745f222 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_aca.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_aca.h @@ -26,6 +26,9 @@ #include <linux/list.h> +struct ras_err_data; +struct ras_query_context; + #define ACA_MAX_REGS_COUNT (16) #define ACA_REG_FIELD(x, h, l) (((x) & GENMASK_ULL(h, l)) >> l) @@ -99,7 +102,14 @@ enum aca_error_type { ACA_ERROR_TYPE_COUNT }; +enum aca_smu_type { + ACA_SMU_TYPE_UE = 0, + ACA_SMU_TYPE_CE, + ACA_SMU_TYPE_COUNT, +}; + struct aca_bank { + enum aca_smu_type type; u64 regs[ACA_MAX_REGS_COUNT]; }; @@ -115,15 +125,10 @@ struct aca_bank_info { int mcatype; }; -struct aca_bank_report { - struct aca_bank_info info; - u64 count[ACA_ERROR_TYPE_COUNT]; -}; - struct aca_bank_error { struct list_head node; struct aca_bank_info info; - u64 count[ACA_ERROR_TYPE_COUNT]; + u64 count; }; struct aca_error { @@ -157,9 +162,8 @@ struct aca_handle { }; struct aca_bank_ops { - int (*aca_bank_generate_report)(struct aca_handle *handle, struct aca_bank *bank, enum aca_error_type type, - struct aca_bank_report *report, void *data); - bool (*aca_bank_is_valid)(struct aca_handle *handle, struct aca_bank *bank, enum aca_error_type type, + int (*aca_bank_parser)(struct aca_handle *handle, struct aca_bank *bank, enum aca_smu_type type, void *data); + bool (*aca_bank_is_valid)(struct aca_handle *handle, struct aca_bank *bank, enum aca_smu_type type, void *data); }; @@ -167,13 +171,15 @@ struct aca_smu_funcs { int max_ue_bank_count; int max_ce_bank_count; int (*set_debug_mode)(struct amdgpu_device *adev, bool enable); - int (*get_valid_aca_count)(struct amdgpu_device *adev, enum aca_error_type type, u32 *count); - int (*get_valid_aca_bank)(struct amdgpu_device *adev, enum aca_error_type type, int idx, struct aca_bank *bank); + int (*get_valid_aca_count)(struct amdgpu_device *adev, enum aca_smu_type type, u32 *count); + int (*get_valid_aca_bank)(struct amdgpu_device *adev, enum aca_smu_type type, int idx, struct aca_bank *bank); + int (*parse_error_code)(struct amdgpu_device *adev, struct aca_bank *bank); }; struct amdgpu_aca { struct aca_handle_manager mgr; const struct aca_smu_funcs *smu_funcs; + atomic_t ue_update_flag; bool is_enabled; }; @@ -196,7 +202,10 @@ int amdgpu_aca_add_handle(struct amdgpu_device *adev, struct aca_handle *handle, const char *name, const struct aca_info *aca_info, void *data); void amdgpu_aca_remove_handle(struct aca_handle *handle); int amdgpu_aca_get_error_data(struct amdgpu_device *adev, struct aca_handle *handle, - enum aca_error_type type, void *data); + enum aca_error_type type, struct ras_err_data *err_data, + struct ras_query_context *qctx); int amdgpu_aca_smu_set_debug_mode(struct amdgpu_device *adev, bool en); void amdgpu_aca_smu_debugfs_init(struct amdgpu_device *adev, struct dentry *root); +int aca_error_cache_log_bank_error(struct aca_handle *handle, struct aca_bank_info *info, + enum aca_error_type type, u64 count); #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_acp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_acp.c index 6d72355ac492..bf6c4a0d0525 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_acp.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_acp.c @@ -637,6 +637,8 @@ static const struct amd_ip_funcs acp_ip_funcs = { .soft_reset = acp_soft_reset, .set_clockgating_state = acp_set_clockgating_state, .set_powergating_state = acp_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; const struct amdgpu_ip_block_version acp_ip_block = { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c index 35dd6effa9a3..e3738d417245 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c @@ -455,6 +455,9 @@ void amdgpu_amdkfd_get_local_mem_info(struct amdgpu_device *adev, else mem_info->local_mem_size_private = KFD_XCP_MEMORY_SIZE(adev, xcp->id); + } else if (adev->flags & AMD_IS_APU) { + mem_info->local_mem_size_public = (ttm_tt_pages_limit() << PAGE_SHIFT); + mem_info->local_mem_size_private = 0; } else { mem_info->local_mem_size_public = adev->gmc.visible_vram_size; mem_info->local_mem_size_private = adev->gmc.real_vram_size - @@ -747,10 +750,17 @@ bool amdgpu_amdkfd_is_fed(struct amdgpu_device *adev) return amdgpu_ras_get_fed_status(adev); } +void amdgpu_amdkfd_ras_pasid_poison_consumption_handler(struct amdgpu_device *adev, + enum amdgpu_ras_block block, uint16_t pasid, + pasid_notify pasid_fn, void *data, uint32_t reset) +{ + amdgpu_umc_pasid_poison_handler(adev, block, pasid, pasid_fn, data, reset); +} + void amdgpu_amdkfd_ras_poison_consumption_handler(struct amdgpu_device *adev, - enum amdgpu_ras_block block, bool reset) + enum amdgpu_ras_block block, uint32_t reset) { - amdgpu_umc_poison_handler(adev, block, reset); + amdgpu_umc_pasid_poison_handler(adev, block, 0, NULL, NULL, reset); } int amdgpu_amdkfd_send_close_event_drain_irq(struct amdgpu_device *adev, @@ -769,12 +779,20 @@ int amdgpu_amdkfd_send_close_event_drain_irq(struct amdgpu_device *adev, return 0; } -bool amdgpu_amdkfd_ras_query_utcl2_poison_status(struct amdgpu_device *adev) +bool amdgpu_amdkfd_ras_query_utcl2_poison_status(struct amdgpu_device *adev, + int hub_inst, int hub_type) { - if (adev->gfx.ras && adev->gfx.ras->query_utcl2_poison_status) - return adev->gfx.ras->query_utcl2_poison_status(adev); - else - return false; + if (!hub_type) { + if (adev->gfxhub.funcs->query_utcl2_poison_status) + return adev->gfxhub.funcs->query_utcl2_poison_status(adev, hub_inst); + else + return false; + } else { + if (adev->mmhub.funcs->query_utcl2_poison_status) + return adev->mmhub.funcs->query_utcl2_poison_status(adev, hub_inst); + else + return false; + } } int amdgpu_amdkfd_check_and_lock_kfd(struct amdgpu_device *adev) @@ -809,6 +827,8 @@ u64 amdgpu_amdkfd_xcp_memory_size(struct amdgpu_device *adev, int xcp_id) } do_div(tmp, adev->xcp_mgr->num_xcp_per_mem_partition); return ALIGN_DOWN(tmp, PAGE_SIZE); + } else if (adev->flags & AMD_IS_APU) { + return (ttm_tt_pages_limit() << PAGE_SHIFT); } else { return adev->gmc.real_vram_size; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h index 0ef223c2affb..1de021ebdd46 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h @@ -336,12 +336,18 @@ void amdgpu_amdkfd_debug_mem_fence(struct amdgpu_device *adev); int amdgpu_amdkfd_get_tile_config(struct amdgpu_device *adev, struct tile_config *config); void amdgpu_amdkfd_ras_poison_consumption_handler(struct amdgpu_device *adev, - enum amdgpu_ras_block block, bool reset); + enum amdgpu_ras_block block, uint32_t reset); + +void amdgpu_amdkfd_ras_pasid_poison_consumption_handler(struct amdgpu_device *adev, + enum amdgpu_ras_block block, uint16_t pasid, + pasid_notify pasid_fn, void *data, uint32_t reset); + bool amdgpu_amdkfd_is_fed(struct amdgpu_device *adev); bool amdgpu_amdkfd_bo_mapped_to_dev(struct amdgpu_device *adev, struct kgd_mem *mem); void amdgpu_amdkfd_block_mmu_notifications(void *p); int amdgpu_amdkfd_criu_resume(void *p); -bool amdgpu_amdkfd_ras_query_utcl2_poison_status(struct amdgpu_device *adev); +bool amdgpu_amdkfd_ras_query_utcl2_poison_status(struct amdgpu_device *adev, + int hub_inst, int hub_type); int amdgpu_amdkfd_reserve_mem_limit(struct amdgpu_device *adev, uint64_t size, u32 alloc_flag, int8_t xcp_id); void amdgpu_amdkfd_unreserve_mem_limit(struct amdgpu_device *adev, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v10.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v10.c index 69810b3f1c63..3ab6c3aa0ad1 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v10.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v10.c @@ -881,6 +881,7 @@ uint32_t kgd_gfx_v10_set_wave_launch_mode(struct amdgpu_device *adev, } #define TCP_WATCH_STRIDE (mmTCP_WATCH1_ADDR_H - mmTCP_WATCH0_ADDR_H) +#define SQ_WATCH_STRIDE (mmSQ_WATCH1_ADDR_H - mmSQ_WATCH0_ADDR_H) uint32_t kgd_gfx_v10_set_address_watch(struct amdgpu_device *adev, uint64_t watch_address, uint32_t watch_address_mask, @@ -889,55 +890,93 @@ uint32_t kgd_gfx_v10_set_address_watch(struct amdgpu_device *adev, uint32_t debug_vmid, uint32_t inst) { + /* SQ_WATCH?_ADDR_* and TCP_WATCH?_ADDR_* are programmed with the + * same values. + */ uint32_t watch_address_high; uint32_t watch_address_low; - uint32_t watch_address_cntl; - - watch_address_cntl = 0; + uint32_t tcp_watch_address_cntl; + uint32_t sq_watch_address_cntl; watch_address_low = lower_32_bits(watch_address); watch_address_high = upper_32_bits(watch_address) & 0xffff; - watch_address_cntl = REG_SET_FIELD(watch_address_cntl, + tcp_watch_address_cntl = 0; + tcp_watch_address_cntl = REG_SET_FIELD(tcp_watch_address_cntl, TCP_WATCH0_CNTL, VMID, debug_vmid); - watch_address_cntl = REG_SET_FIELD(watch_address_cntl, + tcp_watch_address_cntl = REG_SET_FIELD(tcp_watch_address_cntl, TCP_WATCH0_CNTL, MODE, watch_mode); - watch_address_cntl = REG_SET_FIELD(watch_address_cntl, + tcp_watch_address_cntl = REG_SET_FIELD(tcp_watch_address_cntl, TCP_WATCH0_CNTL, MASK, watch_address_mask >> 7); + sq_watch_address_cntl = 0; + sq_watch_address_cntl = REG_SET_FIELD(sq_watch_address_cntl, + SQ_WATCH0_CNTL, + VMID, + debug_vmid); + sq_watch_address_cntl = REG_SET_FIELD(sq_watch_address_cntl, + SQ_WATCH0_CNTL, + MODE, + watch_mode); + sq_watch_address_cntl = REG_SET_FIELD(sq_watch_address_cntl, + SQ_WATCH0_CNTL, + MASK, + watch_address_mask >> 6); + /* Turning off this watch point until we set all the registers */ - watch_address_cntl = REG_SET_FIELD(watch_address_cntl, + tcp_watch_address_cntl = REG_SET_FIELD(tcp_watch_address_cntl, TCP_WATCH0_CNTL, VALID, 0); - WREG32((SOC15_REG_OFFSET(GC, 0, mmTCP_WATCH0_CNTL) + (watch_id * TCP_WATCH_STRIDE)), - watch_address_cntl); + tcp_watch_address_cntl); + + sq_watch_address_cntl = REG_SET_FIELD(sq_watch_address_cntl, + SQ_WATCH0_CNTL, + VALID, + 0); + WREG32((SOC15_REG_OFFSET(GC, 0, mmSQ_WATCH0_CNTL) + + (watch_id * SQ_WATCH_STRIDE)), + sq_watch_address_cntl); + /* Program {TCP,SQ}_WATCH?_ADDR* */ WREG32((SOC15_REG_OFFSET(GC, 0, mmTCP_WATCH0_ADDR_H) + (watch_id * TCP_WATCH_STRIDE)), watch_address_high); - WREG32((SOC15_REG_OFFSET(GC, 0, mmTCP_WATCH0_ADDR_L) + (watch_id * TCP_WATCH_STRIDE)), watch_address_low); + WREG32((SOC15_REG_OFFSET(GC, 0, mmSQ_WATCH0_ADDR_H) + + (watch_id * SQ_WATCH_STRIDE)), + watch_address_high); + WREG32((SOC15_REG_OFFSET(GC, 0, mmSQ_WATCH0_ADDR_L) + + (watch_id * SQ_WATCH_STRIDE)), + watch_address_low); + /* Enable the watch point */ - watch_address_cntl = REG_SET_FIELD(watch_address_cntl, + tcp_watch_address_cntl = REG_SET_FIELD(tcp_watch_address_cntl, TCP_WATCH0_CNTL, VALID, 1); - WREG32((SOC15_REG_OFFSET(GC, 0, mmTCP_WATCH0_CNTL) + (watch_id * TCP_WATCH_STRIDE)), - watch_address_cntl); + tcp_watch_address_cntl); + + sq_watch_address_cntl = REG_SET_FIELD(sq_watch_address_cntl, + SQ_WATCH0_CNTL, + VALID, + 1); + WREG32((SOC15_REG_OFFSET(GC, 0, mmSQ_WATCH0_CNTL) + + (watch_id * SQ_WATCH_STRIDE)), + sq_watch_address_cntl); return 0; } @@ -953,8 +992,14 @@ uint32_t kgd_gfx_v10_clear_address_watch(struct amdgpu_device *adev, (watch_id * TCP_WATCH_STRIDE)), watch_address_cntl); + WREG32((SOC15_REG_OFFSET(GC, 0, mmSQ_WATCH0_CNTL) + + (watch_id * SQ_WATCH_STRIDE)), + watch_address_cntl); + return 0; } +#undef TCP_WATCH_STRIDE +#undef SQ_WATCH_STRIDE /* kgd_gfx_v10_get_iq_wait_times: Returns the mmCP_IQ_WAIT_TIME1/2 values diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c index df58a6a1a67e..48ad0c04aa72 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c @@ -196,7 +196,7 @@ int amdgpu_amdkfd_reserve_mem_limit(struct amdgpu_device *adev, return -EINVAL; vram_size = KFD_XCP_MEMORY_SIZE(adev, xcp_id); - if (adev->gmc.is_app_apu) { + if (adev->flags & AMD_IS_APU) { system_mem_needed = size; ttm_mem_needed = size; } @@ -220,7 +220,7 @@ int amdgpu_amdkfd_reserve_mem_limit(struct amdgpu_device *adev, (kfd_mem_limit.ttm_mem_used + ttm_mem_needed > kfd_mem_limit.max_ttm_mem_limit) || (adev && xcp_id >= 0 && adev->kfd.vram_used[xcp_id] + vram_needed > - vram_size - reserved_for_pt)) { + vram_size - reserved_for_pt - atomic64_read(&adev->vram_pin_size))) { ret = -ENOMEM; goto release; } @@ -232,7 +232,8 @@ int amdgpu_amdkfd_reserve_mem_limit(struct amdgpu_device *adev, "adev reference can't be null when vram is used"); if (adev && xcp_id >= 0) { adev->kfd.vram_used[xcp_id] += vram_needed; - adev->kfd.vram_used_aligned[xcp_id] += adev->gmc.is_app_apu ? + adev->kfd.vram_used_aligned[xcp_id] += + (adev->flags & AMD_IS_APU) ? vram_needed : ALIGN(vram_needed, VRAM_AVAILABLITY_ALIGN); } @@ -260,7 +261,7 @@ void amdgpu_amdkfd_unreserve_mem_limit(struct amdgpu_device *adev, if (adev) { adev->kfd.vram_used[xcp_id] -= size; - if (adev->gmc.is_app_apu) { + if (adev->flags & AMD_IS_APU) { adev->kfd.vram_used_aligned[xcp_id] -= size; kfd_mem_limit.system_mem_used -= size; kfd_mem_limit.ttm_mem_used -= size; @@ -889,7 +890,7 @@ static int kfd_mem_attach(struct amdgpu_device *adev, struct kgd_mem *mem, * if peer device has large BAR. In contrast, access over xGMI is * allowed for both small and large BAR configurations of peer device */ - if ((adev != bo_adev && !adev->gmc.is_app_apu) && + if ((adev != bo_adev && !(adev->flags & AMD_IS_APU)) && ((mem->domain == AMDGPU_GEM_DOMAIN_VRAM) || (mem->alloc_flags & KFD_IOC_ALLOC_MEM_FLAGS_DOORBELL) || (mem->alloc_flags & KFD_IOC_ALLOC_MEM_FLAGS_MMIO_REMAP))) { @@ -1188,7 +1189,8 @@ static int reserve_bo_and_cond_vms(struct kgd_mem *mem, int ret; ctx->sync = &mem->sync; - drm_exec_init(&ctx->exec, DRM_EXEC_INTERRUPTIBLE_WAIT, 0); + drm_exec_init(&ctx->exec, DRM_EXEC_INTERRUPTIBLE_WAIT | + DRM_EXEC_IGNORE_DUPLICATES, 0); drm_exec_until_all_locked(&ctx->exec) { ctx->n_vms = 0; list_for_each_entry(entry, &mem->attachments, list) { @@ -1656,7 +1658,7 @@ size_t amdgpu_amdkfd_get_available_memory(struct amdgpu_device *adev, - atomic64_read(&adev->vram_pin_size) - reserved_for_pt; - if (adev->gmc.is_app_apu) { + if (adev->flags & AMD_IS_APU) { system_mem_available = no_system_mem_limit ? kfd_mem_limit.max_system_mem_limit : kfd_mem_limit.max_system_mem_limit - @@ -1704,7 +1706,7 @@ int amdgpu_amdkfd_gpuvm_alloc_memory_of_gpu( if (flags & KFD_IOC_ALLOC_MEM_FLAGS_VRAM) { domain = alloc_domain = AMDGPU_GEM_DOMAIN_VRAM; - if (adev->gmc.is_app_apu) { + if (adev->flags & AMD_IS_APU) { domain = AMDGPU_GEM_DOMAIN_GTT; alloc_domain = AMDGPU_GEM_DOMAIN_GTT; alloc_flags = 0; @@ -1854,6 +1856,7 @@ err_node_allow: err_bo_create: amdgpu_amdkfd_unreserve_mem_limit(adev, aligned_size, flags, xcp_id); err_reserve_limit: + amdgpu_sync_free(&(*mem)->sync); mutex_destroy(&(*mem)->lock); if (gobj) drm_gem_object_put(gobj); @@ -1950,7 +1953,7 @@ int amdgpu_amdkfd_gpuvm_free_memory_of_gpu( if (size) { if (!is_imported && (mem->bo->preferred_domains == AMDGPU_GEM_DOMAIN_VRAM || - (adev->gmc.is_app_apu && + ((adev->flags & AMD_IS_APU) && mem->bo->preferred_domains == AMDGPU_GEM_DOMAIN_GTT))) *size = bo_size; else @@ -2372,8 +2375,9 @@ static int import_obj_create(struct amdgpu_device *adev, (*mem)->dmabuf = dma_buf; (*mem)->bo = bo; (*mem)->va = va; - (*mem)->domain = (bo->preferred_domains & AMDGPU_GEM_DOMAIN_VRAM) && !adev->gmc.is_app_apu ? - AMDGPU_GEM_DOMAIN_VRAM : AMDGPU_GEM_DOMAIN_GTT; + (*mem)->domain = (bo->preferred_domains & AMDGPU_GEM_DOMAIN_VRAM) && + !(adev->flags & AMD_IS_APU) ? + AMDGPU_GEM_DOMAIN_VRAM : AMDGPU_GEM_DOMAIN_GTT; (*mem)->mapped_to_gpu_memory = 0; (*mem)->process_info = avm->process_info; @@ -2900,13 +2904,12 @@ int amdgpu_amdkfd_gpuvm_restore_process_bos(void *info, struct dma_fence __rcu * amdgpu_sync_create(&sync_obj); - /* Validate BOs and map them to GPUVM (update VM page tables). */ + /* Validate BOs managed by KFD */ list_for_each_entry(mem, &process_info->kfd_bo_list, validate_list) { struct amdgpu_bo *bo = mem->bo; uint32_t domain = mem->domain; - struct kfd_mem_attachment *attachment; struct dma_resv_iter cursor; struct dma_fence *fence; @@ -2931,6 +2934,25 @@ int amdgpu_amdkfd_gpuvm_restore_process_bos(void *info, struct dma_fence __rcu * goto validate_map_fail; } } + } + + if (failed_size) + pr_debug("0x%lx/0x%lx in system\n", failed_size, total_size); + + /* Validate PDs, PTs and evicted DMABuf imports last. Otherwise BO + * validations above would invalidate DMABuf imports again. + */ + ret = process_validate_vms(process_info, &exec.ticket); + if (ret) { + pr_debug("Validating VMs failed, ret: %d\n", ret); + goto validate_map_fail; + } + + /* Update mappings managed by KFD. */ + list_for_each_entry(mem, &process_info->kfd_bo_list, + validate_list) { + struct kfd_mem_attachment *attachment; + list_for_each_entry(attachment, &mem->attachments, list) { if (!attachment->is_mapped) continue; @@ -2947,18 +2969,6 @@ int amdgpu_amdkfd_gpuvm_restore_process_bos(void *info, struct dma_fence __rcu * } } - if (failed_size) - pr_debug("0x%lx/0x%lx in system\n", failed_size, total_size); - - /* Validate PDs, PTs and evicted DMABuf imports last. Otherwise BO - * validations above would invalidate DMABuf imports again. - */ - ret = process_validate_vms(process_info, &exec.ticket); - if (ret) { - pr_debug("Validating VMs failed, ret: %d\n", ret); - goto validate_map_fail; - } - /* Update mappings not managed by KFD */ list_for_each_entry(peer_vm, &process_info->vm_list_head, vm_list_node) { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c index 6857c586ded7..108003bdf1e9 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c @@ -34,6 +34,7 @@ union firmware_info { struct atom_firmware_info_v3_2 v32; struct atom_firmware_info_v3_3 v33; struct atom_firmware_info_v3_4 v34; + struct atom_firmware_info_v3_5 v35; }; /* @@ -211,6 +212,7 @@ union igp_info { struct atom_integrated_system_info_v1_11 v11; struct atom_integrated_system_info_v1_12 v12; struct atom_integrated_system_info_v2_1 v21; + struct atom_integrated_system_info_v2_3 v23; }; union umc_info { @@ -359,6 +361,20 @@ amdgpu_atomfirmware_get_vram_info(struct amdgpu_device *adev, if (vram_type) *vram_type = convert_atom_mem_type_to_vram_type(adev, mem_type); break; + case 3: + mem_channel_number = igp_info->v23.umachannelnumber; + if (!mem_channel_number) + mem_channel_number = 1; + mem_type = igp_info->v23.memorytype; + if (mem_type == LpDdr5MemType) + mem_channel_width = 32; + else + mem_channel_width = 64; + if (vram_width) + *vram_width = mem_channel_number * mem_channel_width; + if (vram_type) + *vram_type = convert_atom_mem_type_to_vram_type(adev, mem_type); + break; default: return -EINVAL; } @@ -872,6 +888,10 @@ int amdgpu_atomfirmware_get_fw_reserved_fb_size(struct amdgpu_device *adev) fw_reserved_fb_size = (firmware_info->v34.fw_reserved_size_in_kb << 10); break; + case 5: + fw_reserved_fb_size = + (firmware_info->v35.fw_reserved_size_in_kb << 10); + break; default: fw_reserved_fb_size = 0; break; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_benchmark.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_benchmark.c index edc6377ec5ff..199693369c7c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_benchmark.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_benchmark.c @@ -39,7 +39,7 @@ static int amdgpu_benchmark_do_move(struct amdgpu_device *adev, unsigned size, for (i = 0; i < n; i++) { struct amdgpu_ring *ring = adev->mman.buffer_funcs_ring; r = amdgpu_copy_buffer(ring, saddr, daddr, size, NULL, &fence, - false, false, false); + false, false, 0); if (r) goto exit_do_move; r = dma_fence_wait(fence, false); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c index 0a4b09709cfb..ec888fc6ead8 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c @@ -819,7 +819,7 @@ retry: p->bytes_moved += ctx.bytes_moved; if (!amdgpu_gmc_vram_full_visible(&adev->gmc) && - amdgpu_bo_in_cpu_visible_vram(bo)) + amdgpu_res_cpu_visible(adev, bo->tbo.resource)) p->bytes_moved_vis += ctx.bytes_moved; if (unlikely(r == -ENOMEM) && domain != bo->allowed_domains) { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c index f5d0fa207a88..b62ae3c91a9d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c @@ -2065,12 +2065,13 @@ static ssize_t amdgpu_reset_dump_register_list_write(struct file *f, struct amdgpu_device *adev = (struct amdgpu_device *)file_inode(f)->i_private; char reg_offset[11]; uint32_t *new = NULL, *tmp = NULL; - int ret, i = 0, len = 0; + unsigned int len = 0; + int ret, i = 0; do { memset(reg_offset, 0, 11); if (copy_from_user(reg_offset, buf + len, - min(10, ((int)size-len)))) { + min(10, (size-len)))) { ret = -EFAULT; goto error_free; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_dev_coredump.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_dev_coredump.c new file mode 100644 index 000000000000..c1cb62683695 --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_dev_coredump.c @@ -0,0 +1,360 @@ +// SPDX-License-Identifier: MIT +/* + * Copyright 2024 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ + +#include <generated/utsrelease.h> +#include <linux/devcoredump.h> +#include "amdgpu_dev_coredump.h" +#include "atom.h" + +#ifndef CONFIG_DEV_COREDUMP +void amdgpu_coredump(struct amdgpu_device *adev, bool vram_lost, + struct amdgpu_reset_context *reset_context) +{ +} +#else + +const char *hw_ip_names[MAX_HWIP] = { + [GC_HWIP] = "GC", + [HDP_HWIP] = "HDP", + [SDMA0_HWIP] = "SDMA0", + [SDMA1_HWIP] = "SDMA1", + [SDMA2_HWIP] = "SDMA2", + [SDMA3_HWIP] = "SDMA3", + [SDMA4_HWIP] = "SDMA4", + [SDMA5_HWIP] = "SDMA5", + [SDMA6_HWIP] = "SDMA6", + [SDMA7_HWIP] = "SDMA7", + [LSDMA_HWIP] = "LSDMA", + [MMHUB_HWIP] = "MMHUB", + [ATHUB_HWIP] = "ATHUB", + [NBIO_HWIP] = "NBIO", + [MP0_HWIP] = "MP0", + [MP1_HWIP] = "MP1", + [UVD_HWIP] = "UVD/JPEG/VCN", + [VCN1_HWIP] = "VCN1", + [VCE_HWIP] = "VCE", + [VPE_HWIP] = "VPE", + [DF_HWIP] = "DF", + [DCE_HWIP] = "DCE", + [OSSSYS_HWIP] = "OSSSYS", + [SMUIO_HWIP] = "SMUIO", + [PWR_HWIP] = "PWR", + [NBIF_HWIP] = "NBIF", + [THM_HWIP] = "THM", + [CLK_HWIP] = "CLK", + [UMC_HWIP] = "UMC", + [RSMU_HWIP] = "RSMU", + [XGMI_HWIP] = "XGMI", + [DCI_HWIP] = "DCI", + [PCIE_HWIP] = "PCIE", +}; + +static void amdgpu_devcoredump_fw_info(struct amdgpu_device *adev, + struct drm_printer *p) +{ + uint32_t version; + uint32_t feature; + uint8_t smu_program, smu_major, smu_minor, smu_debug; + struct atom_context *ctx = adev->mode_info.atom_context; + + drm_printf(p, "VCE feature version: %u, fw version: 0x%08x\n", + adev->vce.fb_version, adev->vce.fw_version); + drm_printf(p, "UVD feature version: %u, fw version: 0x%08x\n", 0, + adev->uvd.fw_version); + drm_printf(p, "GMC feature version: %u, fw version: 0x%08x\n", 0, + adev->gmc.fw_version); + drm_printf(p, "ME feature version: %u, fw version: 0x%08x\n", + adev->gfx.me_feature_version, adev->gfx.me_fw_version); + drm_printf(p, "PFP feature version: %u, fw version: 0x%08x\n", + adev->gfx.pfp_feature_version, adev->gfx.pfp_fw_version); + drm_printf(p, "CE feature version: %u, fw version: 0x%08x\n", + adev->gfx.ce_feature_version, adev->gfx.ce_fw_version); + drm_printf(p, "RLC feature version: %u, fw version: 0x%08x\n", + adev->gfx.rlc_feature_version, adev->gfx.rlc_fw_version); + + drm_printf(p, "RLC SRLC feature version: %u, fw version: 0x%08x\n", + adev->gfx.rlc_srlc_feature_version, + adev->gfx.rlc_srlc_fw_version); + drm_printf(p, "RLC SRLG feature version: %u, fw version: 0x%08x\n", + adev->gfx.rlc_srlg_feature_version, + adev->gfx.rlc_srlg_fw_version); + drm_printf(p, "RLC SRLS feature version: %u, fw version: 0x%08x\n", + adev->gfx.rlc_srls_feature_version, + adev->gfx.rlc_srls_fw_version); + drm_printf(p, "RLCP feature version: %u, fw version: 0x%08x\n", + adev->gfx.rlcp_ucode_feature_version, + adev->gfx.rlcp_ucode_version); + drm_printf(p, "RLCV feature version: %u, fw version: 0x%08x\n", + adev->gfx.rlcv_ucode_feature_version, + adev->gfx.rlcv_ucode_version); + drm_printf(p, "MEC feature version: %u, fw version: 0x%08x\n", + adev->gfx.mec_feature_version, adev->gfx.mec_fw_version); + + if (adev->gfx.mec2_fw) + drm_printf(p, "MEC2 feature version: %u, fw version: 0x%08x\n", + adev->gfx.mec2_feature_version, + adev->gfx.mec2_fw_version); + + drm_printf(p, "IMU feature version: %u, fw version: 0x%08x\n", 0, + adev->gfx.imu_fw_version); + drm_printf(p, "PSP SOS feature version: %u, fw version: 0x%08x\n", + adev->psp.sos.feature_version, adev->psp.sos.fw_version); + drm_printf(p, "PSP ASD feature version: %u, fw version: 0x%08x\n", + adev->psp.asd_context.bin_desc.feature_version, + adev->psp.asd_context.bin_desc.fw_version); + + drm_printf(p, "TA XGMI feature version: 0x%08x, fw version: 0x%08x\n", + adev->psp.xgmi_context.context.bin_desc.feature_version, + adev->psp.xgmi_context.context.bin_desc.fw_version); + drm_printf(p, "TA RAS feature version: 0x%08x, fw version: 0x%08x\n", + adev->psp.ras_context.context.bin_desc.feature_version, + adev->psp.ras_context.context.bin_desc.fw_version); + drm_printf(p, "TA HDCP feature version: 0x%08x, fw version: 0x%08x\n", + adev->psp.hdcp_context.context.bin_desc.feature_version, + adev->psp.hdcp_context.context.bin_desc.fw_version); + drm_printf(p, "TA DTM feature version: 0x%08x, fw version: 0x%08x\n", + adev->psp.dtm_context.context.bin_desc.feature_version, + adev->psp.dtm_context.context.bin_desc.fw_version); + drm_printf(p, "TA RAP feature version: 0x%08x, fw version: 0x%08x\n", + adev->psp.rap_context.context.bin_desc.feature_version, + adev->psp.rap_context.context.bin_desc.fw_version); + drm_printf(p, + "TA SECURE DISPLAY feature version: 0x%08x, fw version: 0x%08x\n", + adev->psp.securedisplay_context.context.bin_desc.feature_version, + adev->psp.securedisplay_context.context.bin_desc.fw_version); + + /* SMC firmware */ + version = adev->pm.fw_version; + + smu_program = (version >> 24) & 0xff; + smu_major = (version >> 16) & 0xff; + smu_minor = (version >> 8) & 0xff; + smu_debug = (version >> 0) & 0xff; + drm_printf(p, + "SMC feature version: %u, program: %d, fw version: 0x%08x (%d.%d.%d)\n", + 0, smu_program, version, smu_major, smu_minor, smu_debug); + + /* SDMA firmware */ + for (int i = 0; i < adev->sdma.num_instances; i++) { + drm_printf(p, + "SDMA%d feature version: %u, firmware version: 0x%08x\n", + i, adev->sdma.instance[i].feature_version, + adev->sdma.instance[i].fw_version); + } + + drm_printf(p, "VCN feature version: %u, fw version: 0x%08x\n", 0, + adev->vcn.fw_version); + drm_printf(p, "DMCU feature version: %u, fw version: 0x%08x\n", 0, + adev->dm.dmcu_fw_version); + drm_printf(p, "DMCUB feature version: %u, fw version: 0x%08x\n", 0, + adev->dm.dmcub_fw_version); + drm_printf(p, "PSP TOC feature version: %u, fw version: 0x%08x\n", + adev->psp.toc.feature_version, adev->psp.toc.fw_version); + + version = adev->mes.kiq_version & AMDGPU_MES_VERSION_MASK; + feature = (adev->mes.kiq_version & AMDGPU_MES_FEAT_VERSION_MASK) >> + AMDGPU_MES_FEAT_VERSION_SHIFT; + drm_printf(p, "MES_KIQ feature version: %u, fw version: 0x%08x\n", + feature, version); + + version = adev->mes.sched_version & AMDGPU_MES_VERSION_MASK; + feature = (adev->mes.sched_version & AMDGPU_MES_FEAT_VERSION_MASK) >> + AMDGPU_MES_FEAT_VERSION_SHIFT; + drm_printf(p, "MES feature version: %u, fw version: 0x%08x\n", feature, + version); + + drm_printf(p, "VPE feature version: %u, fw version: 0x%08x\n", + adev->vpe.feature_version, adev->vpe.fw_version); + + drm_printf(p, "\nVBIOS Information\n"); + drm_printf(p, "vbios name : %s\n", ctx->name); + drm_printf(p, "vbios pn : %s\n", ctx->vbios_pn); + drm_printf(p, "vbios version : %d\n", ctx->version); + drm_printf(p, "vbios ver_str : %s\n", ctx->vbios_ver_str); + drm_printf(p, "vbios date : %s\n", ctx->date); +} + +static ssize_t +amdgpu_devcoredump_read(char *buffer, loff_t offset, size_t count, + void *data, size_t datalen) +{ + struct drm_printer p; + struct amdgpu_coredump_info *coredump = data; + struct drm_print_iterator iter; + struct amdgpu_vm_fault_info *fault_info; + int i, ver; + + iter.data = buffer; + iter.offset = 0; + iter.start = offset; + iter.remain = count; + + p = drm_coredump_printer(&iter); + + drm_printf(&p, "**** AMDGPU Device Coredump ****\n"); + drm_printf(&p, "version: " AMDGPU_COREDUMP_VERSION "\n"); + drm_printf(&p, "kernel: " UTS_RELEASE "\n"); + drm_printf(&p, "module: " KBUILD_MODNAME "\n"); + drm_printf(&p, "time: %lld.%09ld\n", coredump->reset_time.tv_sec, + coredump->reset_time.tv_nsec); + + if (coredump->reset_task_info.pid) + drm_printf(&p, "process_name: %s PID: %d\n", + coredump->reset_task_info.process_name, + coredump->reset_task_info.pid); + + /* GPU IP's information of the SOC */ + drm_printf(&p, "\nIP Information\n"); + drm_printf(&p, "SOC Family: %d\n", coredump->adev->family); + drm_printf(&p, "SOC Revision id: %d\n", coredump->adev->rev_id); + drm_printf(&p, "SOC External Revision id: %d\n", coredump->adev->external_rev_id); + + for (int i = 1; i < MAX_HWIP; i++) { + for (int j = 0; j < HWIP_MAX_INSTANCE; j++) { + ver = coredump->adev->ip_versions[i][j]; + if (ver) + drm_printf(&p, "HWIP: %s[%d][%d]: v%d.%d.%d.%d.%d\n", + hw_ip_names[i], i, j, + IP_VERSION_MAJ(ver), + IP_VERSION_MIN(ver), + IP_VERSION_REV(ver), + IP_VERSION_VARIANT(ver), + IP_VERSION_SUBREV(ver)); + } + } + + /* IP firmware information */ + drm_printf(&p, "\nIP Firmwares\n"); + amdgpu_devcoredump_fw_info(coredump->adev, &p); + + if (coredump->ring) { + drm_printf(&p, "\nRing timed out details\n"); + drm_printf(&p, "IP Type: %d Ring Name: %s\n", + coredump->ring->funcs->type, + coredump->ring->name); + } + + /* Add page fault information */ + fault_info = &coredump->adev->vm_manager.fault_info; + drm_printf(&p, "\n[%s] Page fault observed\n", + fault_info->vmhub ? "mmhub" : "gfxhub"); + drm_printf(&p, "Faulty page starting at address: 0x%016llx\n", fault_info->addr); + drm_printf(&p, "Protection fault status register: 0x%x\n\n", fault_info->status); + + /* dump the ip state for each ip */ + drm_printf(&p, "IP Dump\n"); + for (int i = 0; i < coredump->adev->num_ip_blocks; i++) { + if (coredump->adev->ip_blocks[i].version->funcs->print_ip_state) { + drm_printf(&p, "IP: %s\n", + coredump->adev->ip_blocks[i] + .version->funcs->name); + coredump->adev->ip_blocks[i] + .version->funcs->print_ip_state( + (void *)coredump->adev, &p); + drm_printf(&p, "\n"); + } + } + + /* Add ring buffer information */ + drm_printf(&p, "Ring buffer information\n"); + for (int i = 0; i < coredump->adev->num_rings; i++) { + int j = 0; + struct amdgpu_ring *ring = coredump->adev->rings[i]; + + drm_printf(&p, "ring name: %s\n", ring->name); + drm_printf(&p, "Rptr: 0x%llx Wptr: 0x%llx RB mask: %x\n", + amdgpu_ring_get_rptr(ring), + amdgpu_ring_get_wptr(ring), + ring->buf_mask); + drm_printf(&p, "Ring size in dwords: %d\n", + ring->ring_size / 4); + drm_printf(&p, "Ring contents\n"); + drm_printf(&p, "Offset \t Value\n"); + + while (j < ring->ring_size) { + drm_printf(&p, "0x%x \t 0x%x\n", j, ring->ring[j / 4]); + j += 4; + } + } + + if (coredump->reset_vram_lost) + drm_printf(&p, "VRAM is lost due to GPU reset!\n"); + if (coredump->adev->reset_info.num_regs) { + drm_printf(&p, "AMDGPU register dumps:\nOffset: Value:\n"); + + for (i = 0; i < coredump->adev->reset_info.num_regs; i++) + drm_printf(&p, "0x%08x: 0x%08x\n", + coredump->adev->reset_info.reset_dump_reg_list[i], + coredump->adev->reset_info.reset_dump_reg_value[i]); + } + + return count - iter.remain; +} + +static void amdgpu_devcoredump_free(void *data) +{ + kfree(data); +} + +void amdgpu_coredump(struct amdgpu_device *adev, bool vram_lost, + struct amdgpu_reset_context *reset_context) +{ + struct amdgpu_coredump_info *coredump; + struct drm_device *dev = adev_to_drm(adev); + struct amdgpu_job *job = reset_context->job; + struct drm_sched_job *s_job; + + coredump = kzalloc(sizeof(*coredump), GFP_NOWAIT); + + if (!coredump) { + DRM_ERROR("%s: failed to allocate memory for coredump\n", __func__); + return; + } + + coredump->reset_vram_lost = vram_lost; + + if (reset_context->job && reset_context->job->vm) { + struct amdgpu_task_info *ti; + struct amdgpu_vm *vm = reset_context->job->vm; + + ti = amdgpu_vm_get_task_info_vm(vm); + if (ti) { + coredump->reset_task_info = *ti; + amdgpu_vm_put_task_info(ti); + } + } + + if (job) { + s_job = &job->base; + coredump->ring = to_amdgpu_ring(s_job->sched); + } + + coredump->adev = adev; + + ktime_get_ts64(&coredump->reset_time); + + dev_coredumpm(dev->dev, THIS_MODULE, coredump, 0, GFP_NOWAIT, + amdgpu_devcoredump_read, amdgpu_devcoredump_free); +} +#endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_dev_coredump.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_dev_coredump.h new file mode 100644 index 000000000000..52459512cb2b --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_dev_coredump.h @@ -0,0 +1,47 @@ +/* SPDX-License-Identifier: MIT */ +/* + * Copyright 2024 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ + +#ifndef __AMDGPU_DEV_COREDUMP_H__ +#define __AMDGPU_DEV_COREDUMP_H__ + +#include "amdgpu.h" +#include "amdgpu_reset.h" + +#ifdef CONFIG_DEV_COREDUMP + +#define AMDGPU_COREDUMP_VERSION "1" + +struct amdgpu_coredump_info { + struct amdgpu_device *adev; + struct amdgpu_task_info reset_task_info; + struct timespec64 reset_time; + bool reset_vram_lost; + struct amdgpu_ring *ring; +}; +#endif + +void amdgpu_coredump(struct amdgpu_device *adev, bool vram_lost, + struct amdgpu_reset_context *reset_context); + +#endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index 5dc24c971b41..932dc93b2e63 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -74,6 +74,7 @@ #include "amdgpu_fru_eeprom.h" #include "amdgpu_reset.h" #include "amdgpu_virt.h" +#include "amdgpu_dev_coredump.h" #include <linux/suspend.h> #include <drm/task_barrier.h> @@ -143,6 +144,8 @@ const char *amdgpu_asic_name[] = { "LAST", }; +static inline void amdgpu_device_stop_pending_resets(struct amdgpu_device *adev); + /** * DOC: pcie_replay_count * @@ -335,16 +338,93 @@ bool amdgpu_device_supports_boco(struct drm_device *dev) * * @dev: drm_device pointer * - * Returns true if the device supporte BACO, - * otherwise return false. + * Return: + * 1 if the device supporte BACO; + * 3 if the device support MACO (only works if BACO is supported) + * otherwise return 0. */ -bool amdgpu_device_supports_baco(struct drm_device *dev) +int amdgpu_device_supports_baco(struct drm_device *dev) { struct amdgpu_device *adev = drm_to_adev(dev); return amdgpu_asic_supports_baco(adev); } +void amdgpu_device_detect_runtime_pm_mode(struct amdgpu_device *adev) +{ + struct drm_device *dev; + int bamaco_support; + + dev = adev_to_drm(adev); + + adev->pm.rpm_mode = AMDGPU_RUNPM_NONE; + bamaco_support = amdgpu_device_supports_baco(dev); + + switch (amdgpu_runtime_pm) { + case 2: + if (bamaco_support & MACO_SUPPORT) { + adev->pm.rpm_mode = AMDGPU_RUNPM_BAMACO; + dev_info(adev->dev, "Forcing BAMACO for runtime pm\n"); + } else if (bamaco_support == BACO_SUPPORT) { + adev->pm.rpm_mode = AMDGPU_RUNPM_BACO; + dev_info(adev->dev, "Requested mode BAMACO not available,fallback to use BACO\n"); + } + break; + case 1: + if (bamaco_support & BACO_SUPPORT) { + adev->pm.rpm_mode = AMDGPU_RUNPM_BACO; + dev_info(adev->dev, "Forcing BACO for runtime pm\n"); + } + break; + case -1: + case -2: + if (amdgpu_device_supports_px(dev)) { /* enable PX as runtime mode */ + adev->pm.rpm_mode = AMDGPU_RUNPM_PX; + dev_info(adev->dev, "Using ATPX for runtime pm\n"); + } else if (amdgpu_device_supports_boco(dev)) { /* enable boco as runtime mode */ + adev->pm.rpm_mode = AMDGPU_RUNPM_BOCO; + dev_info(adev->dev, "Using BOCO for runtime pm\n"); + } else { + if (!bamaco_support) + goto no_runtime_pm; + + switch (adev->asic_type) { + case CHIP_VEGA20: + case CHIP_ARCTURUS: + /* BACO are not supported on vega20 and arctrus */ + break; + case CHIP_VEGA10: + /* enable BACO as runpm mode if noretry=0 */ + if (!adev->gmc.noretry) + adev->pm.rpm_mode = AMDGPU_RUNPM_BACO; + break; + default: + /* enable BACO as runpm mode on CI+ */ + adev->pm.rpm_mode = AMDGPU_RUNPM_BACO; + break; + } + + if (adev->pm.rpm_mode == AMDGPU_RUNPM_BACO) { + if (bamaco_support & MACO_SUPPORT) { + adev->pm.rpm_mode = AMDGPU_RUNPM_BAMACO; + dev_info(adev->dev, "Using BAMACO for runtime pm\n"); + } else { + dev_info(adev->dev, "Using BACO for runtime pm\n"); + } + } + } + break; + case 0: + dev_info(adev->dev, "runtime pm is manually disabled\n"); + break; + default: + break; + } + +no_runtime_pm: + if (adev->pm.rpm_mode == AMDGPU_RUNPM_NONE) + dev_info(adev->dev, "Runtime PM not available\n"); +} /** * amdgpu_device_supports_smart_shift - Is the device dGPU with * smart shift support @@ -1402,13 +1482,17 @@ static int amdgpu_device_wb_init(struct amdgpu_device *adev) */ int amdgpu_device_wb_get(struct amdgpu_device *adev, u32 *wb) { - unsigned long offset = find_first_zero_bit(adev->wb.used, adev->wb.num_wb); + unsigned long flags, offset; + spin_lock_irqsave(&adev->wb.lock, flags); + offset = find_first_zero_bit(adev->wb.used, adev->wb.num_wb); if (offset < adev->wb.num_wb) { __set_bit(offset, adev->wb.used); + spin_unlock_irqrestore(&adev->wb.lock, flags); *wb = offset << 3; /* convert to dw offset */ return 0; } else { + spin_unlock_irqrestore(&adev->wb.lock, flags); return -EINVAL; } } @@ -1423,9 +1507,13 @@ int amdgpu_device_wb_get(struct amdgpu_device *adev, u32 *wb) */ void amdgpu_device_wb_free(struct amdgpu_device *adev, u32 wb) { + unsigned long flags; + wb >>= 3; + spin_lock_irqsave(&adev->wb.lock, flags); if (wb < adev->wb.num_wb) __clear_bit(wb, adev->wb.used); + spin_unlock_irqrestore(&adev->wb.lock, flags); } /** @@ -1455,7 +1543,7 @@ int amdgpu_device_resize_fb_bar(struct amdgpu_device *adev) /* PCI_EXT_CAP_ID_VNDR extended capability is located at 0x100 */ if (!pci_find_ext_capability(adev->pdev, PCI_EXT_CAP_ID_VNDR)) - DRM_WARN("System can't access extended configuration space,please check!!\n"); + DRM_WARN("System can't access extended configuration space, please check!!\n"); /* skip if the bios has already enabled large BAR */ if (adev->gmc.real_vram_size && @@ -3981,6 +4069,7 @@ int amdgpu_device_init(struct amdgpu_device *adev, spin_lock_init(&adev->se_cac_idx_lock); spin_lock_init(&adev->audio_endpt_idx_lock); spin_lock_init(&adev->mm_stats.lock); + spin_lock_init(&adev->wb.lock); INIT_LIST_HEAD(&adev->shadow_list); mutex_init(&adev->shadow_list_lock); @@ -4069,6 +4158,13 @@ int amdgpu_device_init(struct amdgpu_device *adev, /* Enable TMZ based on IP_VERSION */ amdgpu_gmc_tmz_set(adev); + if (amdgpu_sriov_vf(adev) && + amdgpu_ip_version(adev, GC_HWIP, 0) >= IP_VERSION(10, 3, 0)) + /* VF MMIO access (except mailbox range) from CPU + * will be blocked during sriov runtime + */ + adev->virt.caps |= AMDGPU_VF_MMIO_ACCESS_PROTECT; + amdgpu_gmc_noretry_set(adev); /* Need to get xgmi info early to decide the reset behavior*/ if (adev->gmc.xgmi.supported) { @@ -4135,18 +4231,22 @@ int amdgpu_device_init(struct amdgpu_device *adev, adev->ip_blocks[i].status.hw = true; } } + } else if (amdgpu_ip_version(adev, MP1_HWIP, 0) == IP_VERSION(13, 0, 10) && + !amdgpu_device_has_display_hardware(adev)) { + r = psp_gpu_reset(adev); } else { - tmp = amdgpu_reset_method; - /* It should do a default reset when loading or reloading the driver, - * regardless of the module parameter reset_method. - */ - amdgpu_reset_method = AMD_RESET_METHOD_NONE; - r = amdgpu_asic_reset(adev); - amdgpu_reset_method = tmp; - if (r) { - dev_err(adev->dev, "asic reset on init failed\n"); - goto failed; - } + tmp = amdgpu_reset_method; + /* It should do a default reset when loading or reloading the driver, + * regardless of the module parameter reset_method. + */ + amdgpu_reset_method = AMD_RESET_METHOD_NONE; + r = amdgpu_asic_reset(adev); + amdgpu_reset_method = tmp; + } + + if (r) { + dev_err(adev->dev, "asic reset on init failed\n"); + goto failed; } } @@ -4539,6 +4639,8 @@ int amdgpu_device_prepare(struct drm_device *dev) if (r) goto unprepare; + flush_delayed_work(&adev->gfx.gfx_off_delay_work); + for (i = 0; i < adev->num_ip_blocks; i++) { if (!adev->ip_blocks[i].status.valid) continue; @@ -4968,12 +5070,15 @@ static int amdgpu_device_reset_sriov(struct amdgpu_device *adev, retry: amdgpu_amdkfd_pre_reset(adev); + amdgpu_device_stop_pending_resets(adev); + if (from_hypervisor) r = amdgpu_virt_request_full_gpu(adev, true); else r = amdgpu_virt_reset_gpu(adev); if (r) return r; + amdgpu_ras_set_fed(adev, false); amdgpu_irq_gpu_reset_resume_helper(adev); /* some sw clean up VF needs to do before recover */ @@ -5257,11 +5362,21 @@ int amdgpu_do_asic_reset(struct list_head *device_list_handle, struct amdgpu_device *tmp_adev = NULL; bool need_full_reset, skip_hw_reset, vram_lost = false; int r = 0; + uint32_t i; /* Try reset handler method first */ tmp_adev = list_first_entry(device_list_handle, struct amdgpu_device, reset_list); - amdgpu_reset_reg_dumps(tmp_adev); + + if (!test_bit(AMDGPU_SKIP_COREDUMP, &reset_context->flags)) { + amdgpu_reset_reg_dumps(tmp_adev); + + /* Trigger ip dump before we reset the asic */ + for (i = 0; i < tmp_adev->num_ip_blocks; i++) + if (tmp_adev->ip_blocks[i].version->funcs->dump_ip_state) + tmp_adev->ip_blocks[i].version->funcs + ->dump_ip_state((void *)tmp_adev); + } reset_context->reset_device_list = device_list_handle; r = amdgpu_reset_perform_reset(tmp_adev, reset_context); @@ -5334,7 +5449,8 @@ int amdgpu_do_asic_reset(struct list_head *device_list_handle, vram_lost = amdgpu_device_check_vram_lost(tmp_adev); - amdgpu_coredump(tmp_adev, vram_lost, reset_context); + if (!test_bit(AMDGPU_SKIP_COREDUMP, &reset_context->flags)) + amdgpu_coredump(tmp_adev, vram_lost, reset_context); if (vram_lost) { DRM_INFO("VRAM is lost due to GPU reset!\n"); @@ -5532,6 +5648,23 @@ static inline void amdgpu_device_stop_pending_resets(struct amdgpu_device *adev) } +static int amdgpu_device_health_check(struct list_head *device_list_handle) +{ + struct amdgpu_device *tmp_adev; + int ret = 0; + u32 status; + + list_for_each_entry(tmp_adev, device_list_handle, reset_list) { + pci_read_config_dword(tmp_adev->pdev, PCI_COMMAND, &status); + if (PCI_POSSIBLE_ERROR(status)) { + dev_err(tmp_adev->dev, "device lost from bus!"); + ret = -ENODEV; + } + } + + return ret; +} + /** * amdgpu_device_gpu_recover - reset the asic and recover scheduler * @@ -5603,6 +5736,12 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev, device_list_handle = &device_list; } + if (!amdgpu_sriov_vf(adev)) { + r = amdgpu_device_health_check(device_list_handle); + if (r) + goto end_reset; + } + /* We need to lock reset domain only once both for XGMI and single device */ tmp_adev = list_first_entry(device_list_handle, struct amdgpu_device, reset_list); @@ -5685,11 +5824,12 @@ retry: /* Rest of adevs pre asic reset from XGMI hive. */ tmp_adev->asic_reset_res = r; } - /* - * Drop all pending non scheduler resets. Scheduler resets - * were already dropped during drm_sched_stop - */ - amdgpu_device_stop_pending_resets(tmp_adev); + if (!amdgpu_sriov_vf(tmp_adev)) + /* + * Drop all pending non scheduler resets. Scheduler resets + * were already dropped during drm_sched_stop + */ + amdgpu_device_stop_pending_resets(tmp_adev); } /* Actual ASIC resets if needed.*/ @@ -5768,6 +5908,7 @@ skip_sched_resume: reset_list); amdgpu_device_unlock_reset_domain(tmp_adev->reset_domain); +end_reset: if (hive) { mutex_unlock(&hive->hive_lock); amdgpu_put_xgmi_hive(hive); @@ -5803,13 +5944,18 @@ static void amdgpu_device_partner_bandwidth(struct amdgpu_device *adev, *speed = PCI_SPEED_UNKNOWN; *width = PCIE_LNK_WIDTH_UNKNOWN; - while ((parent = pci_upstream_bridge(parent))) { - /* skip upstream/downstream switches internal to dGPU*/ - if (parent->vendor == PCI_VENDOR_ID_ATI) - continue; - *speed = pcie_get_speed_cap(parent); - *width = pcie_get_width_cap(parent); - break; + if (amdgpu_device_pcie_dynamic_switching_supported(adev)) { + while ((parent = pci_upstream_bridge(parent))) { + /* skip upstream/downstream switches internal to dGPU*/ + if (parent->vendor == PCI_VENDOR_ID_ATI) + continue; + *speed = pcie_get_speed_cap(parent); + *width = pcie_get_width_cap(parent); + break; + } + } else { + /* use the current speeds rather than max if switching is not supported */ + pcie_bandwidth_available(adev->pdev, NULL, speed, width); } } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c index a07e4b87d4ca..0e31bdb4b7cb 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c @@ -97,6 +97,7 @@ #include "smuio_v13_0.h" #include "smuio_v13_0_3.h" #include "smuio_v13_0_6.h" +#include "smuio_v14_0_2.h" #include "vcn_v5_0_0.h" #include "jpeg_v5_0_0.h" @@ -245,6 +246,9 @@ static int amdgpu_discovery_read_binary_from_sysmem(struct amdgpu_device *adev, return -ENOENT; } +#define IP_DISCOVERY_V2 2 +#define IP_DISCOVERY_V4 4 + static int amdgpu_discovery_read_binary_from_mem(struct amdgpu_device *adev, uint8_t *binary) { @@ -259,14 +263,14 @@ static int amdgpu_discovery_read_binary_from_mem(struct amdgpu_device *adev, * wait for this to complete. Once the C2PMSG is updated, we can * continue. */ - if (dev_is_removable(&adev->pdev->dev)) { - for (i = 0; i < 1000; i++) { - msg = RREG32(mmMP0_SMN_C2PMSG_33); - if (msg & 0x80000000) - break; - msleep(1); - } + + for (i = 0; i < 1000; i++) { + msg = RREG32(mmMP0_SMN_C2PMSG_33); + if (msg & 0x80000000) + break; + usleep_range(1000, 1100); } + vram_size = (uint64_t)RREG32(mmRCC_CONFIG_MEMSIZE) << 20; if (vram_size) { @@ -1896,6 +1900,9 @@ static int amdgpu_discovery_set_smu_ip_blocks(struct amdgpu_device *adev) amdgpu_device_ip_block_add(adev, &smu_v13_0_ip_block); break; case IP_VERSION(14, 0, 0): + case IP_VERSION(14, 0, 1): + case IP_VERSION(14, 0, 2): + case IP_VERSION(14, 0, 3): amdgpu_device_ip_block_add(adev, &smu_v14_0_ip_block); break; default: @@ -2237,6 +2244,7 @@ static int amdgpu_discovery_set_umsch_mm_ip_blocks(struct amdgpu_device *adev) { switch (amdgpu_ip_version(adev, VCN_HWIP, 0)) { case IP_VERSION(4, 0, 5): + case IP_VERSION(4, 0, 6): if (amdgpu_umsch_mm & 0x1) { amdgpu_device_ip_block_add(adev, &umsch_mm_v4_0_ip_block); adev->enable_umsch_mm = true; @@ -2676,6 +2684,9 @@ int amdgpu_discovery_set_ip_blocks(struct amdgpu_device *adev) case IP_VERSION(14, 0, 1): adev->smuio.funcs = &smuio_v13_0_6_funcs; break; + case IP_VERSION(14, 0, 2): + adev->smuio.funcs = &smuio_v14_0_2_funcs; + break; default: break; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c index 80b9642f2bc4..ea14f1c8f430 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c @@ -195,6 +195,7 @@ int amdgpu_async_gfx_ring = 1; int amdgpu_mcbp = -1; int amdgpu_discovery = -1; int amdgpu_mes; +int amdgpu_mes_log_enable = 0; int amdgpu_mes_kiq; int amdgpu_noretry = -1; int amdgpu_force_asic_type = -1; @@ -668,6 +669,15 @@ MODULE_PARM_DESC(mes, module_param_named(mes, amdgpu_mes, int, 0444); /** + * DOC: mes_log_enable (int) + * Enable Micro Engine Scheduler log. This is used to enable/disable MES internal log. + * (0 = disabled (default), 1 = enabled) + */ +MODULE_PARM_DESC(mes_log_enable, + "Enable Micro Engine Scheduler log (0 = disabled (default), 1 = enabled)"); +module_param_named(mes_log_enable, amdgpu_mes_log_enable, int, 0444); + +/** * DOC: mes_kiq (int) * Enable Micro Engine Scheduler KIQ. This is a new engine pipe for kiq. * (0 = disabled (default), 1 = enabled) @@ -915,7 +925,7 @@ module_param_named(freesync_video, amdgpu_freesync_vid_mode, uint, 0444); * GPU reset method (-1 = auto (default), 0 = legacy, 1 = mode0, 2 = mode1, 3 = mode2, 4 = baco) */ MODULE_PARM_DESC(reset_method, "GPU reset method (-1 = auto (default), 0 = legacy, 1 = mode0, 2 = mode1, 3 = mode2, 4 = baco/bamaco)"); -module_param_named(reset_method, amdgpu_reset_method, int, 0444); +module_param_named(reset_method, amdgpu_reset_method, int, 0644); /** * DOC: bad_page_threshold (int) Bad page threshold is specifies the @@ -2471,6 +2481,7 @@ static void amdgpu_drv_delayed_reset_work_handler(struct work_struct *work) /* Use a common context, just need to make sure full reset is done */ set_bit(AMDGPU_SKIP_HW_RESET, &reset_context.flags); + set_bit(AMDGPU_SKIP_COREDUMP, &reset_context.flags); r = amdgpu_do_asic_reset(&device_list, &reset_context); if (r) { @@ -2734,7 +2745,8 @@ static int amdgpu_pmops_runtime_suspend(struct device *dev) drm_dev->switch_power_state = DRM_SWITCH_POWER_DYNAMIC_OFF; } else if (adev->pm.rpm_mode == AMDGPU_RUNPM_BOCO) { /* nothing to do */ - } else if (adev->pm.rpm_mode == AMDGPU_RUNPM_BACO) { + } else if ((adev->pm.rpm_mode == AMDGPU_RUNPM_BACO) || + (adev->pm.rpm_mode == AMDGPU_RUNPM_BAMACO)) { amdgpu_device_baco_enter(drm_dev); } @@ -2774,7 +2786,8 @@ static int amdgpu_pmops_runtime_resume(struct device *dev) * PCI core handles it for _PR3. */ pci_set_master(pdev); - } else if (adev->pm.rpm_mode == AMDGPU_RUNPM_BACO) { + } else if ((adev->pm.rpm_mode == AMDGPU_RUNPM_BACO) || + (adev->pm.rpm_mode == AMDGPU_RUNPM_BAMACO)) { amdgpu_device_baco_exit(drm_dev); } ret = amdgpu_device_resume(drm_dev, false); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c index 55d5508987ff..1d955652f3ba 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c @@ -1206,7 +1206,8 @@ void amdgpu_gfx_cp_init_microcode(struct amdgpu_device *adev, fw_size = le32_to_cpu(cp_hdr_v2_0->data_size_bytes); break; default: - break; + dev_err(adev->dev, "Invalid ucode id %u\n", ucode_id); + return; } if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h index 8fcf889ddce9..64f197bbc866 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h @@ -259,7 +259,6 @@ struct amdgpu_cu_info { struct amdgpu_gfx_ras { struct amdgpu_ras_block_object ras_block; void (*enable_watchdog_timer)(struct amdgpu_device *adev); - bool (*query_utcl2_poison_status)(struct amdgpu_device *adev); int (*rlc_gc_fed_irq)(struct amdgpu_device *adev, struct amdgpu_irq_src *source, struct amdgpu_iv_entry *entry); @@ -434,6 +433,10 @@ struct amdgpu_gfx { uint32_t num_xcc_per_xcp; struct mutex partition_mutex; bool mcbp; /* mid command buffer preemption */ + + /* IP reg dump */ + uint32_t *ip_dump; + uint32_t reg_count; }; struct amdgpu_gfx_ras_reg_entry { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfxhub.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfxhub.h index c7b44aeb671b..103a837ccc71 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfxhub.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfxhub.h @@ -38,6 +38,8 @@ struct amdgpu_gfxhub_funcs { void (*mode2_save_regs)(struct amdgpu_device *adev); void (*mode2_restore_regs)(struct amdgpu_device *adev); void (*halt)(struct amdgpu_device *adev); + bool (*query_utcl2_poison_status)(struct amdgpu_device *adev, + int xcc_id); }; struct amdgpu_gfxhub { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_i2c.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_i2c.c index d79cb13e1aa8..00d6211e0fbf 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_i2c.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_i2c.c @@ -279,7 +279,7 @@ amdgpu_i2c_lookup(struct amdgpu_device *adev, return NULL; } -static void amdgpu_i2c_get_byte(struct amdgpu_i2c_chan *i2c_bus, +static int amdgpu_i2c_get_byte(struct amdgpu_i2c_chan *i2c_bus, u8 slave_addr, u8 addr, u8 *val) @@ -304,16 +304,18 @@ static void amdgpu_i2c_get_byte(struct amdgpu_i2c_chan *i2c_bus, out_buf[0] = addr; out_buf[1] = 0; - if (i2c_transfer(&i2c_bus->adapter, msgs, 2) == 2) { - *val = in_buf[0]; - DRM_DEBUG("val = 0x%02x\n", *val); - } else { - DRM_DEBUG("i2c 0x%02x 0x%02x read failed\n", - addr, *val); + if (i2c_transfer(&i2c_bus->adapter, msgs, 2) != 2) { + DRM_DEBUG("i2c 0x%02x read failed\n", addr); + return -EIO; } + + *val = in_buf[0]; + DRM_DEBUG("val = 0x%02x\n", *val); + + return 0; } -static void amdgpu_i2c_put_byte(struct amdgpu_i2c_chan *i2c_bus, +static int amdgpu_i2c_put_byte(struct amdgpu_i2c_chan *i2c_bus, u8 slave_addr, u8 addr, u8 val) @@ -329,9 +331,12 @@ static void amdgpu_i2c_put_byte(struct amdgpu_i2c_chan *i2c_bus, out_buf[0] = addr; out_buf[1] = val; - if (i2c_transfer(&i2c_bus->adapter, &msg, 1) != 1) - DRM_DEBUG("i2c 0x%02x 0x%02x write failed\n", - addr, val); + if (i2c_transfer(&i2c_bus->adapter, &msg, 1) != 1) { + DRM_DEBUG("i2c 0x%02x 0x%02x write failed\n", addr, val); + return -EIO; + } + + return 0; } /* ddc router switching */ @@ -346,16 +351,18 @@ amdgpu_i2c_router_select_ddc_port(const struct amdgpu_connector *amdgpu_connecto if (!amdgpu_connector->router_bus) return; - amdgpu_i2c_get_byte(amdgpu_connector->router_bus, + if (amdgpu_i2c_get_byte(amdgpu_connector->router_bus, amdgpu_connector->router.i2c_addr, - 0x3, &val); + 0x3, &val)) + return; val &= ~amdgpu_connector->router.ddc_mux_control_pin; amdgpu_i2c_put_byte(amdgpu_connector->router_bus, amdgpu_connector->router.i2c_addr, 0x3, val); - amdgpu_i2c_get_byte(amdgpu_connector->router_bus, + if (amdgpu_i2c_get_byte(amdgpu_connector->router_bus, amdgpu_connector->router.i2c_addr, - 0x1, &val); + 0x1, &val)) + return; val &= ~amdgpu_connector->router.ddc_mux_control_pin; val |= amdgpu_connector->router.ddc_mux_state; amdgpu_i2c_put_byte(amdgpu_connector->router_bus, @@ -375,16 +382,18 @@ amdgpu_i2c_router_select_cd_port(const struct amdgpu_connector *amdgpu_connector if (!amdgpu_connector->router_bus) return; - amdgpu_i2c_get_byte(amdgpu_connector->router_bus, + if (amdgpu_i2c_get_byte(amdgpu_connector->router_bus, amdgpu_connector->router.i2c_addr, - 0x3, &val); + 0x3, &val)) + return; val &= ~amdgpu_connector->router.cd_mux_control_pin; amdgpu_i2c_put_byte(amdgpu_connector->router_bus, amdgpu_connector->router.i2c_addr, 0x3, val); - amdgpu_i2c_get_byte(amdgpu_connector->router_bus, + if (amdgpu_i2c_get_byte(amdgpu_connector->router_bus, amdgpu_connector->router.i2c_addr, - 0x1, &val); + 0x1, &val)) + return; val &= ~amdgpu_connector->router.cd_mux_control_pin; val |= amdgpu_connector->router.cd_mux_state; amdgpu_i2c_put_byte(amdgpu_connector->router_bus, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c index 7e6d09730e6d..013ff373e067 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c @@ -279,7 +279,7 @@ int amdgpu_irq_init(struct amdgpu_device *adev) adev->irq.msi_enabled = false; if (!amdgpu_msi_ok(adev)) - flags = PCI_IRQ_LEGACY; + flags = PCI_IRQ_INTX; else flags = PCI_IRQ_ALL_TYPES; @@ -445,6 +445,14 @@ void amdgpu_irq_dispatch(struct amdgpu_device *adev, entry.ih = ih; entry.iv_entry = (const uint32_t *)&ih->ring[ring_index]; + + /* + * timestamp is not supported on some legacy SOCs (cik, cz, iceland, + * si and tonga), so initialize timestamp and timestamp_src to 0 + */ + entry.timestamp = 0; + entry.timestamp_src = 0; + amdgpu_ih_decode_iv(adev, &entry); trace_amdgpu_iv(ih - &adev->irq.ih, &entry); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c index 4b3000c21ef2..e4742b65032d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c @@ -304,12 +304,15 @@ static struct dma_fence *amdgpu_job_run(struct drm_sched_job *sched_job) dma_fence_set_error(finished, -ECANCELED); if (finished->error < 0) { - DRM_INFO("Skip scheduling IBs!\n"); + dev_dbg(adev->dev, "Skip scheduling IBs in ring(%s)", + ring->name); } else { r = amdgpu_ib_schedule(ring, job->num_ibs, job->ibs, job, &fence); if (r) - DRM_ERROR("Error scheduling IBs (%d)\n", r); + dev_err(adev->dev, + "Error scheduling IBs (%d) in ring(%s)", r, + ring->name); } job->job_run_counter++; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c index a2df3025a754..a0ea6fe8d060 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c @@ -149,38 +149,7 @@ int amdgpu_driver_load_kms(struct amdgpu_device *adev, unsigned long flags) goto out; } - adev->pm.rpm_mode = AMDGPU_RUNPM_NONE; - if (amdgpu_device_supports_px(dev) && - (amdgpu_runtime_pm != 0)) { /* enable PX as runtime mode */ - adev->pm.rpm_mode = AMDGPU_RUNPM_PX; - dev_info(adev->dev, "Using ATPX for runtime pm\n"); - } else if (amdgpu_device_supports_boco(dev) && - (amdgpu_runtime_pm != 0)) { /* enable boco as runtime mode */ - adev->pm.rpm_mode = AMDGPU_RUNPM_BOCO; - dev_info(adev->dev, "Using BOCO for runtime pm\n"); - } else if (amdgpu_device_supports_baco(dev) && - (amdgpu_runtime_pm != 0)) { - switch (adev->asic_type) { - case CHIP_VEGA20: - case CHIP_ARCTURUS: - /* enable BACO as runpm mode if runpm=1 */ - if (amdgpu_runtime_pm > 0) - adev->pm.rpm_mode = AMDGPU_RUNPM_BACO; - break; - case CHIP_VEGA10: - /* enable BACO as runpm mode if noretry=0 */ - if (!adev->gmc.noretry) - adev->pm.rpm_mode = AMDGPU_RUNPM_BACO; - break; - default: - /* enable BACO as runpm mode on CI+ */ - adev->pm.rpm_mode = AMDGPU_RUNPM_BACO; - break; - } - - if (adev->pm.rpm_mode == AMDGPU_RUNPM_BACO) - dev_info(adev->dev, "Using BACO for runtime pm\n"); - } + amdgpu_device_detect_runtime_pm_mode(adev); /* Call ACPI methods: require modeset init * but failure is not fatal diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mca.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_mca.c index 24ad4b97177b..0734490347db 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mca.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mca.c @@ -210,22 +210,26 @@ int amdgpu_mca_smu_set_debug_mode(struct amdgpu_device *adev, bool enable) return -EOPNOTSUPP; } -static void amdgpu_mca_smu_mca_bank_dump(struct amdgpu_device *adev, int idx, struct mca_bank_entry *entry) +static void amdgpu_mca_smu_mca_bank_dump(struct amdgpu_device *adev, int idx, struct mca_bank_entry *entry, + struct ras_query_context *qctx) { - dev_info(adev->dev, HW_ERR "Accelerator Check Architecture events logged\n"); - dev_info(adev->dev, HW_ERR "aca entry[%02d].STATUS=0x%016llx\n", - idx, entry->regs[MCA_REG_IDX_STATUS]); - dev_info(adev->dev, HW_ERR "aca entry[%02d].ADDR=0x%016llx\n", - idx, entry->regs[MCA_REG_IDX_ADDR]); - dev_info(adev->dev, HW_ERR "aca entry[%02d].MISC0=0x%016llx\n", - idx, entry->regs[MCA_REG_IDX_MISC0]); - dev_info(adev->dev, HW_ERR "aca entry[%02d].IPID=0x%016llx\n", - idx, entry->regs[MCA_REG_IDX_IPID]); - dev_info(adev->dev, HW_ERR "aca entry[%02d].SYND=0x%016llx\n", - idx, entry->regs[MCA_REG_IDX_SYND]); + u64 event_id = qctx->event_id; + + RAS_EVENT_LOG(adev, event_id, HW_ERR "Accelerator Check Architecture events logged\n"); + RAS_EVENT_LOG(adev, event_id, HW_ERR "aca entry[%02d].STATUS=0x%016llx\n", + idx, entry->regs[MCA_REG_IDX_STATUS]); + RAS_EVENT_LOG(adev, event_id, HW_ERR "aca entry[%02d].ADDR=0x%016llx\n", + idx, entry->regs[MCA_REG_IDX_ADDR]); + RAS_EVENT_LOG(adev, event_id, HW_ERR "aca entry[%02d].MISC0=0x%016llx\n", + idx, entry->regs[MCA_REG_IDX_MISC0]); + RAS_EVENT_LOG(adev, event_id, HW_ERR "aca entry[%02d].IPID=0x%016llx\n", + idx, entry->regs[MCA_REG_IDX_IPID]); + RAS_EVENT_LOG(adev, event_id, HW_ERR "aca entry[%02d].SYND=0x%016llx\n", + idx, entry->regs[MCA_REG_IDX_SYND]); } -int amdgpu_mca_smu_log_ras_error(struct amdgpu_device *adev, enum amdgpu_ras_block blk, enum amdgpu_mca_error_type type, struct ras_err_data *err_data) +int amdgpu_mca_smu_log_ras_error(struct amdgpu_device *adev, enum amdgpu_ras_block blk, enum amdgpu_mca_error_type type, + struct ras_err_data *err_data, struct ras_query_context *qctx) { struct amdgpu_smuio_mcm_config_info mcm_info; struct ras_err_addr err_addr = {0}; @@ -244,7 +248,7 @@ int amdgpu_mca_smu_log_ras_error(struct amdgpu_device *adev, enum amdgpu_ras_blo list_for_each_entry(node, &mca_set.list, node) { entry = &node->entry; - amdgpu_mca_smu_mca_bank_dump(adev, i++, entry); + amdgpu_mca_smu_mca_bank_dump(adev, i++, entry, qctx); count = 0; ret = amdgpu_mca_smu_parse_mca_error_count(adev, blk, type, entry, &count); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mca.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mca.h index b964110ed1e0..e5bf07ce3451 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mca.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mca.h @@ -169,6 +169,7 @@ void amdgpu_mca_smu_debugfs_init(struct amdgpu_device *adev, struct dentry *root void amdgpu_mca_bank_set_init(struct mca_bank_set *mca_set); int amdgpu_mca_bank_set_add_entry(struct mca_bank_set *mca_set, struct mca_bank_entry *entry); void amdgpu_mca_bank_set_release(struct mca_bank_set *mca_set); -int amdgpu_mca_smu_log_ras_error(struct amdgpu_device *adev, enum amdgpu_ras_block blk, enum amdgpu_mca_error_type type, struct ras_err_data *err_data); +int amdgpu_mca_smu_log_ras_error(struct amdgpu_device *adev, enum amdgpu_ras_block blk, enum amdgpu_mca_error_type type, + struct ras_err_data *err_data, struct ras_query_context *qctx); #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.c index a98e03e0a51f..5ca5c47ab54e 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.c @@ -32,6 +32,18 @@ #define AMDGPU_MES_MAX_NUM_OF_QUEUES_PER_PROCESS 1024 #define AMDGPU_ONE_DOORBELL_SIZE 8 +signed long amdgpu_mes_fence_wait_polling(u64 *fence, + u64 wait_seq, + signed long timeout) +{ + + while ((s64)(wait_seq - *fence) > 0 && timeout > 0) { + udelay(2); + timeout -= 2; + } + return timeout > 0 ? timeout : 0; +} + int amdgpu_mes_doorbell_process_slice(struct amdgpu_device *adev) { return roundup(AMDGPU_ONE_DOORBELL_SIZE * @@ -40,7 +52,6 @@ int amdgpu_mes_doorbell_process_slice(struct amdgpu_device *adev) } static int amdgpu_mes_kernel_doorbell_get(struct amdgpu_device *adev, - struct amdgpu_mes_process *process, int ip_type, uint64_t *doorbell_index) { unsigned int offset, found; @@ -65,7 +76,6 @@ static int amdgpu_mes_kernel_doorbell_get(struct amdgpu_device *adev, } static void amdgpu_mes_kernel_doorbell_free(struct amdgpu_device *adev, - struct amdgpu_mes_process *process, uint32_t doorbell_index) { unsigned int old, rel_index; @@ -102,7 +112,10 @@ static int amdgpu_mes_event_log_init(struct amdgpu_device *adev) { int r; - r = amdgpu_bo_create_kernel(adev, PAGE_SIZE, PAGE_SIZE, + if (!amdgpu_mes_log_enable) + return 0; + + r = amdgpu_bo_create_kernel(adev, AMDGPU_MES_LOG_BUFFER_SIZE, PAGE_SIZE, AMDGPU_GEM_DOMAIN_GTT, &adev->mes.event_log_gpu_obj, &adev->mes.event_log_gpu_addr, @@ -653,7 +666,7 @@ int amdgpu_mes_add_hw_queue(struct amdgpu_device *adev, int gang_id, *queue_id = queue->queue_id = r; /* allocate a doorbell index for the queue */ - r = amdgpu_mes_kernel_doorbell_get(adev, gang->process, + r = amdgpu_mes_kernel_doorbell_get(adev, qprops->queue_type, &qprops->doorbell_off); if (r) @@ -711,8 +724,7 @@ int amdgpu_mes_add_hw_queue(struct amdgpu_device *adev, int gang_id, return 0; clean_up_doorbell: - amdgpu_mes_kernel_doorbell_free(adev, gang->process, - qprops->doorbell_off); + amdgpu_mes_kernel_doorbell_free(adev, qprops->doorbell_off); clean_up_queue_id: spin_lock_irqsave(&adev->mes.queue_id_lock, flags); idr_remove(&adev->mes.queue_id_idr, queue->queue_id); @@ -766,8 +778,7 @@ int amdgpu_mes_remove_hw_queue(struct amdgpu_device *adev, int queue_id) queue_id); list_del(&queue->list); - amdgpu_mes_kernel_doorbell_free(adev, gang->process, - queue->doorbell_off); + amdgpu_mes_kernel_doorbell_free(adev, queue->doorbell_off); amdgpu_mes_unlock(&adev->mes); amdgpu_mes_queue_free_mqd(queue); @@ -775,6 +786,28 @@ int amdgpu_mes_remove_hw_queue(struct amdgpu_device *adev, int queue_id) return 0; } +int amdgpu_mes_map_legacy_queue(struct amdgpu_device *adev, + struct amdgpu_ring *ring) +{ + struct mes_map_legacy_queue_input queue_input; + int r; + + memset(&queue_input, 0, sizeof(queue_input)); + + queue_input.queue_type = ring->funcs->type; + queue_input.doorbell_offset = ring->doorbell_index; + queue_input.pipe_id = ring->pipe; + queue_input.queue_id = ring->queue; + queue_input.mqd_addr = amdgpu_bo_gpu_offset(ring->mqd_obj); + queue_input.wptr_addr = ring->wptr_gpu_addr; + + r = adev->mes.funcs->map_legacy_queue(&adev->mes, &queue_input); + if (r) + DRM_ERROR("failed to map legacy queue\n"); + + return r; +} + int amdgpu_mes_unmap_legacy_queue(struct amdgpu_device *adev, struct amdgpu_ring *ring, enum amdgpu_unmap_queues_action action, @@ -1129,6 +1162,7 @@ void amdgpu_mes_remove_ring(struct amdgpu_device *adev, return; amdgpu_mes_remove_hw_queue(adev, ring->hw_queue_id); + del_timer_sync(&ring->fence_drv.fallback_timer); amdgpu_ring_fini(ring); kfree(ring); } @@ -1471,7 +1505,7 @@ int amdgpu_mes_init_microcode(struct amdgpu_device *adev, int pipe) const struct mes_firmware_header_v1_0 *mes_hdr; struct amdgpu_firmware_info *info; char ucode_prefix[30]; - char fw_name[40]; + char fw_name[50]; bool need_retry = false; int r; @@ -1549,12 +1583,11 @@ static int amdgpu_debugfs_mes_event_log_show(struct seq_file *m, void *unused) uint32_t *mem = (uint32_t *)(adev->mes.event_log_cpu_addr); seq_hex_dump(m, "", DUMP_PREFIX_OFFSET, 32, 4, - mem, PAGE_SIZE, false); + mem, AMDGPU_MES_LOG_BUFFER_SIZE, false); return 0; } - DEFINE_SHOW_ATTRIBUTE(amdgpu_debugfs_mes_event_log); #endif @@ -1565,7 +1598,7 @@ void amdgpu_debugfs_mes_event_log_init(struct amdgpu_device *adev) #if defined(CONFIG_DEBUG_FS) struct drm_minor *minor = adev_to_drm(adev)->primary; struct dentry *root = minor->debugfs_root; - if (adev->enable_mes) + if (adev->enable_mes && amdgpu_mes_log_enable) debugfs_create_file("amdgpu_mes_event_log", 0444, root, adev, &amdgpu_debugfs_mes_event_log_fops); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.h index 7d4f93fea937..df9f0404d842 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.h @@ -52,6 +52,7 @@ enum amdgpu_mes_priority_level { #define AMDGPU_MES_PROC_CTX_SIZE 0x1000 /* one page area */ #define AMDGPU_MES_GANG_CTX_SIZE 0x1000 /* one page area */ +#define AMDGPU_MES_LOG_BUFFER_SIZE 0x4000 /* Maximu log buffer size for MES */ struct amdgpu_mes_funcs; @@ -140,6 +141,12 @@ struct amdgpu_mes { /* ip specific functions */ const struct amdgpu_mes_funcs *funcs; + + /* mes resource_1 bo*/ + struct amdgpu_bo *resource_1; + uint64_t resource_1_gpu_addr; + void *resource_1_addr; + }; struct amdgpu_mes_process { @@ -241,6 +248,15 @@ struct mes_remove_queue_input { uint64_t gang_context_addr; }; +struct mes_map_legacy_queue_input { + uint32_t queue_type; + uint32_t doorbell_offset; + uint32_t pipe_id; + uint32_t queue_id; + uint64_t mqd_addr; + uint64_t wptr_addr; +}; + struct mes_unmap_legacy_queue_input { enum amdgpu_unmap_queues_action action; uint32_t queue_type; @@ -317,6 +333,9 @@ struct amdgpu_mes_funcs { int (*remove_hw_queue)(struct amdgpu_mes *mes, struct mes_remove_queue_input *input); + int (*map_legacy_queue)(struct amdgpu_mes *mes, + struct mes_map_legacy_queue_input *input); + int (*unmap_legacy_queue)(struct amdgpu_mes *mes, struct mes_unmap_legacy_queue_input *input); @@ -333,6 +352,10 @@ struct amdgpu_mes_funcs { #define amdgpu_mes_kiq_hw_init(adev) (adev)->mes.kiq_hw_init((adev)) #define amdgpu_mes_kiq_hw_fini(adev) (adev)->mes.kiq_hw_fini((adev)) +signed long amdgpu_mes_fence_wait_polling(u64 *fence, + u64 wait_seq, + signed long timeout); + int amdgpu_mes_ctx_get_offs(struct amdgpu_ring *ring, unsigned int id_offs); int amdgpu_mes_init_microcode(struct amdgpu_device *adev, int pipe); @@ -356,6 +379,8 @@ int amdgpu_mes_add_hw_queue(struct amdgpu_device *adev, int gang_id, int *queue_id); int amdgpu_mes_remove_hw_queue(struct amdgpu_device *adev, int queue_id); +int amdgpu_mes_map_legacy_queue(struct amdgpu_device *adev, + struct amdgpu_ring *ring); int amdgpu_mes_unmap_legacy_queue(struct amdgpu_device *adev, struct amdgpu_ring *ring, enum amdgpu_unmap_queues_action action, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mmhub.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mmhub.h index 1ca9d4ed8063..95d676ee207f 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mmhub.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mmhub.h @@ -63,6 +63,8 @@ struct amdgpu_mmhub_funcs { uint64_t page_table_base); void (*update_power_gating)(struct amdgpu_device *adev, bool enable); + bool (*query_utcl2_poison_status)(struct amdgpu_device *adev, + int hub_inst); }; struct amdgpu_mmhub { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c index 010b0cb7693c..8d8c39be6129 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c @@ -39,6 +39,7 @@ #include "amdgpu.h" #include "amdgpu_trace.h" #include "amdgpu_amdkfd.h" +#include "amdgpu_vram_mgr.h" /** * DOC: amdgpu_object @@ -153,8 +154,10 @@ void amdgpu_bo_placement_from_domain(struct amdgpu_bo *abo, u32 domain) else places[c].flags |= TTM_PL_FLAG_TOPDOWN; - if (flags & AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS) + if (abo->tbo.type == ttm_bo_type_kernel && + flags & AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS) places[c].flags |= TTM_PL_FLAG_CONTIGUOUS; + c++; } @@ -173,6 +176,12 @@ void amdgpu_bo_placement_from_domain(struct amdgpu_bo *abo, u32 domain) abo->flags & AMDGPU_GEM_CREATE_PREEMPTIBLE ? AMDGPU_PL_PREEMPT : TTM_PL_TT; places[c].flags = 0; + /* + * When GTT is just an alternative to VRAM make sure that we + * only use it as fallback and still try to fill up VRAM first. + */ + if (domain & abo->preferred_domains & AMDGPU_GEM_DOMAIN_VRAM) + places[c].flags |= TTM_PL_FLAG_FALLBACK; c++; } @@ -595,8 +604,7 @@ int amdgpu_bo_create(struct amdgpu_device *adev, if (!amdgpu_bo_support_uswc(bo->flags)) bo->flags &= ~AMDGPU_GEM_CREATE_CPU_GTT_USWC; - if (adev->ras_enabled) - bo->flags |= AMDGPU_GEM_CREATE_VRAM_WIPE_ON_RELEASE; + bo->flags |= AMDGPU_GEM_CREATE_VRAM_WIPE_ON_RELEASE; bo->tbo.bdev = &adev->mman.bdev; if (bp->domain & (AMDGPU_GEM_DOMAIN_GWS | AMDGPU_GEM_DOMAIN_OA | @@ -605,6 +613,8 @@ int amdgpu_bo_create(struct amdgpu_device *adev, else amdgpu_bo_placement_from_domain(bo, bp->domain); if (bp->type == ttm_bo_type_kernel) + bo->tbo.priority = 2; + else if (!(bp->flags & AMDGPU_GEM_CREATE_DISCARDABLE)) bo->tbo.priority = 1; if (!bp->destroy) @@ -617,8 +627,7 @@ int amdgpu_bo_create(struct amdgpu_device *adev, return r; if (!amdgpu_gmc_vram_full_visible(&adev->gmc) && - bo->tbo.resource->mem_type == TTM_PL_VRAM && - amdgpu_bo_in_cpu_visible_vram(bo)) + amdgpu_res_cpu_visible(adev, bo->tbo.resource)) amdgpu_cs_report_moved_bytes(adev, ctx.bytes_moved, ctx.bytes_moved); else @@ -628,7 +637,7 @@ int amdgpu_bo_create(struct amdgpu_device *adev, bo->tbo.resource->mem_type == TTM_PL_VRAM) { struct dma_fence *fence; - r = amdgpu_fill_buffer(bo, 0, bo->tbo.base.resv, &fence, true); + r = amdgpu_ttm_clear_buffer(bo, bo->tbo.base.resv, &fence); if (unlikely(r)) goto fail_unreserve; @@ -758,7 +767,7 @@ int amdgpu_bo_restore_shadow(struct amdgpu_bo *shadow, struct dma_fence **fence) return amdgpu_copy_buffer(ring, shadow_addr, parent_addr, amdgpu_bo_size(shadow), NULL, fence, - true, false, false); + true, false, 0); } /** @@ -960,6 +969,10 @@ int amdgpu_bo_pin_restricted(struct amdgpu_bo *bo, u32 domain, if (!bo->placements[i].lpfn || (lpfn && lpfn < bo->placements[i].lpfn)) bo->placements[i].lpfn = lpfn; + + if (bo->flags & AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS && + bo->placements[i].mem_type == TTM_PL_VRAM) + bo->placements[i].flags |= TTM_PL_FLAG_CONTIGUOUS; } r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx); @@ -1242,14 +1255,18 @@ int amdgpu_bo_get_metadata(struct amdgpu_bo *bo, void *buffer, * amdgpu_bo_move_notify - notification about a memory move * @bo: pointer to a buffer object * @evict: if this move is evicting the buffer from the graphics address space + * @new_mem: new resource for backing the BO * * Marks the corresponding &amdgpu_bo buffer object as invalid, also performs * bookkeeping. * TTM driver callback which is called when ttm moves a buffer. */ -void amdgpu_bo_move_notify(struct ttm_buffer_object *bo, bool evict) +void amdgpu_bo_move_notify(struct ttm_buffer_object *bo, + bool evict, + struct ttm_resource *new_mem) { struct amdgpu_device *adev = amdgpu_ttm_adev(bo->bdev); + struct ttm_resource *old_mem = bo->resource; struct amdgpu_bo *abo; if (!amdgpu_bo_is_amdgpu_bo(bo)) @@ -1261,34 +1278,36 @@ void amdgpu_bo_move_notify(struct ttm_buffer_object *bo, bool evict) amdgpu_bo_kunmap(abo); if (abo->tbo.base.dma_buf && !abo->tbo.base.import_attach && - bo->resource->mem_type != TTM_PL_SYSTEM) + old_mem && old_mem->mem_type != TTM_PL_SYSTEM) dma_buf_move_notify(abo->tbo.base.dma_buf); - /* remember the eviction */ - if (evict) - atomic64_inc(&adev->num_evictions); + /* move_notify is called before move happens */ + trace_amdgpu_bo_move(abo, new_mem ? new_mem->mem_type : -1, + old_mem ? old_mem->mem_type : -1); } void amdgpu_bo_get_memory(struct amdgpu_bo *bo, struct amdgpu_mem_stats *stats) { + struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev); + struct ttm_resource *res = bo->tbo.resource; uint64_t size = amdgpu_bo_size(bo); struct drm_gem_object *obj; unsigned int domain; bool shared; /* Abort if the BO doesn't currently have a backing store */ - if (!bo->tbo.resource) + if (!res) return; obj = &bo->tbo.base; shared = drm_gem_object_is_shared_for_memory_stats(obj); - domain = amdgpu_mem_type_to_domain(bo->tbo.resource->mem_type); + domain = amdgpu_mem_type_to_domain(res->mem_type); switch (domain) { case AMDGPU_GEM_DOMAIN_VRAM: stats->vram += size; - if (amdgpu_bo_in_cpu_visible_vram(bo)) + if (amdgpu_res_cpu_visible(adev, bo->tbo.resource)) stats->visible_vram += size; if (shared) stats->vram_shared += size; @@ -1359,8 +1378,9 @@ void amdgpu_bo_release_notify(struct ttm_buffer_object *bo) if (WARN_ON_ONCE(!dma_resv_trylock(bo->base.resv))) return; - r = amdgpu_fill_buffer(abo, AMDGPU_POISON, bo->base.resv, &fence, true); + r = amdgpu_fill_buffer(abo, 0, bo->base.resv, &fence, true); if (!WARN_ON(r)) { + amdgpu_vram_mgr_set_cleared(bo->resource); amdgpu_bo_fence(abo, fence, false); dma_fence_put(fence); } @@ -1389,10 +1409,7 @@ vm_fault_t amdgpu_bo_fault_reserve_notify(struct ttm_buffer_object *bo) /* Remember that this BO was accessed by the CPU */ abo->flags |= AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED; - if (bo->resource->mem_type != TTM_PL_VRAM) - return 0; - - if (amdgpu_bo_in_cpu_visible_vram(abo)) + if (amdgpu_res_cpu_visible(adev, bo->resource)) return 0; /* Can't move a pinned BO to visible VRAM */ @@ -1415,7 +1432,7 @@ vm_fault_t amdgpu_bo_fault_reserve_notify(struct ttm_buffer_object *bo) /* this should never happen */ if (bo->resource->mem_type == TTM_PL_VRAM && - !amdgpu_bo_in_cpu_visible_vram(abo)) + !amdgpu_res_cpu_visible(adev, bo->resource)) return VM_FAULT_SIGBUS; ttm_bo_move_to_lru_tail_unlocked(bo); @@ -1579,6 +1596,7 @@ uint32_t amdgpu_bo_get_preferred_domain(struct amdgpu_device *adev, */ u64 amdgpu_bo_print_info(int id, struct amdgpu_bo *bo, struct seq_file *m) { + struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev); struct dma_buf_attachment *attachment; struct dma_buf *dma_buf; const char *placement; @@ -1587,10 +1605,11 @@ u64 amdgpu_bo_print_info(int id, struct amdgpu_bo *bo, struct seq_file *m) if (dma_resv_trylock(bo->tbo.base.resv)) { unsigned int domain; + domain = amdgpu_mem_type_to_domain(bo->tbo.resource->mem_type); switch (domain) { case AMDGPU_GEM_DOMAIN_VRAM: - if (amdgpu_bo_in_cpu_visible_vram(bo)) + if (amdgpu_res_cpu_visible(adev, bo->tbo.resource)) placement = "VRAM VISIBLE"; else placement = "VRAM"; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h index be679c42b0b8..bc42ccbde659 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h @@ -251,28 +251,6 @@ static inline u64 amdgpu_bo_mmap_offset(struct amdgpu_bo *bo) } /** - * amdgpu_bo_in_cpu_visible_vram - check if BO is (partly) in visible VRAM - */ -static inline bool amdgpu_bo_in_cpu_visible_vram(struct amdgpu_bo *bo) -{ - struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev); - struct amdgpu_res_cursor cursor; - - if (!bo->tbo.resource || bo->tbo.resource->mem_type != TTM_PL_VRAM) - return false; - - amdgpu_res_first(bo->tbo.resource, 0, amdgpu_bo_size(bo), &cursor); - while (cursor.remaining) { - if (cursor.start < adev->gmc.visible_vram_size) - return true; - - amdgpu_res_next(&cursor, cursor.size); - } - - return false; -} - -/** * amdgpu_bo_explicit_sync - return whether the bo is explicitly synced */ static inline bool amdgpu_bo_explicit_sync(struct amdgpu_bo *bo) @@ -350,7 +328,9 @@ int amdgpu_bo_set_metadata (struct amdgpu_bo *bo, void *metadata, int amdgpu_bo_get_metadata(struct amdgpu_bo *bo, void *buffer, size_t buffer_size, uint32_t *metadata_size, uint64_t *flags); -void amdgpu_bo_move_notify(struct ttm_buffer_object *bo, bool evict); +void amdgpu_bo_move_notify(struct ttm_buffer_object *bo, + bool evict, + struct ttm_resource *new_mem); void amdgpu_bo_release_notify(struct ttm_buffer_object *bo); vm_fault_t amdgpu_bo_fault_reserve_notify(struct ttm_buffer_object *bo); void amdgpu_bo_fence(struct amdgpu_bo *bo, struct dma_fence *fence, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c index 94b310fdb719..4bd4602d11b1 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c @@ -1053,6 +1053,11 @@ static int psp_asd_initialize(struct psp_context *psp) if (amdgpu_sriov_vf(psp->adev) || !psp->asd_context.bin_desc.size_bytes) return 0; + /* bypass asd if display hardware is not available */ + if (!amdgpu_device_has_display_hardware(psp->adev) && + amdgpu_ip_version(psp->adev, MP0_HWIP, 0) >= IP_VERSION(13, 0, 10)) + return 0; + psp->asd_context.mem_context.shared_mc_addr = 0; psp->asd_context.mem_context.shared_mem_size = PSP_ASD_SHARED_MEM_SIZE; psp->asd_context.ta_load_type = GFX_CMD_ID_LOAD_ASD; @@ -2260,6 +2265,15 @@ static int psp_hw_start(struct psp_context *psp) } } + if ((is_psp_fw_valid(psp->ipkeymgr_drv)) && + (psp->funcs->bootloader_load_ipkeymgr_drv != NULL)) { + ret = psp_bootloader_load_ipkeymgr_drv(psp); + if (ret) { + dev_err(adev->dev, "PSP load ipkeymgr_drv failed!\n"); + return ret; + } + } + if ((is_psp_fw_valid(psp->sos)) && (psp->funcs->bootloader_load_sos != NULL)) { ret = psp_bootloader_load_sos(psp); @@ -2617,7 +2631,8 @@ static int psp_load_p2s_table(struct psp_context *psp) struct amdgpu_firmware_info *ucode = &adev->firmware.ucode[AMDGPU_UCODE_ID_P2S_TABLE]; - if (adev->in_runpm && (adev->pm.rpm_mode == AMDGPU_RUNPM_BACO)) + if (adev->in_runpm && ((adev->pm.rpm_mode == AMDGPU_RUNPM_BACO) || + (adev->pm.rpm_mode == AMDGPU_RUNPM_BAMACO))) return 0; if (amdgpu_ip_version(adev, MP0_HWIP, 0) == IP_VERSION(13, 0, 6)) { @@ -2647,7 +2662,8 @@ static int psp_load_smu_fw(struct psp_context *psp) * Skip SMU FW reloading in case of using BACO for runpm only, * as SMU is always alive. */ - if (adev->in_runpm && (adev->pm.rpm_mode == AMDGPU_RUNPM_BACO)) + if (adev->in_runpm && ((adev->pm.rpm_mode == AMDGPU_RUNPM_BACO) || + (adev->pm.rpm_mode == AMDGPU_RUNPM_BAMACO))) return 0; if (!ucode->fw || amdgpu_sriov_vf(psp->adev)) @@ -3273,6 +3289,12 @@ static int parse_sos_bin_descriptor(struct psp_context *psp, psp->ras_drv.size_bytes = le32_to_cpu(desc->size_bytes); psp->ras_drv.start_addr = ucode_start_addr; break; + case PSP_FW_TYPE_PSP_IPKEYMGR_DRV: + psp->ipkeymgr_drv.fw_version = le32_to_cpu(desc->fw_version); + psp->ipkeymgr_drv.feature_version = le32_to_cpu(desc->fw_version); + psp->ipkeymgr_drv.size_bytes = le32_to_cpu(desc->size_bytes); + psp->ipkeymgr_drv.start_addr = ucode_start_addr; + break; default: dev_warn(psp->adev->dev, "Unsupported PSP FW type: %d\n", desc->fw_type); break; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h index ee16f134ae92..3635303e6548 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h @@ -73,8 +73,10 @@ enum psp_bootloader_cmd { PSP_BL__LOAD_KEY_DATABASE = 0x80000, PSP_BL__LOAD_SOCDRV = 0xB0000, PSP_BL__LOAD_DBGDRV = 0xC0000, + PSP_BL__LOAD_HADDRV = PSP_BL__LOAD_DBGDRV, PSP_BL__LOAD_INTFDRV = 0xD0000, - PSP_BL__LOAD_RASDRV = 0xE0000, + PSP_BL__LOAD_RASDRV = 0xE0000, + PSP_BL__LOAD_IPKEYMGRDRV = 0xF0000, PSP_BL__DRAM_LONG_TRAIN = 0x100000, PSP_BL__DRAM_SHORT_TRAIN = 0x200000, PSP_BL__LOAD_TOS_SPL_TABLE = 0x10000000, @@ -117,6 +119,7 @@ struct psp_funcs { int (*bootloader_load_intf_drv)(struct psp_context *psp); int (*bootloader_load_dbg_drv)(struct psp_context *psp); int (*bootloader_load_ras_drv)(struct psp_context *psp); + int (*bootloader_load_ipkeymgr_drv)(struct psp_context *psp); int (*bootloader_load_sos)(struct psp_context *psp); int (*ring_create)(struct psp_context *psp, enum psp_ring_type ring_type); @@ -336,6 +339,7 @@ struct psp_context { struct psp_bin_desc intf_drv; struct psp_bin_desc dbg_drv; struct psp_bin_desc ras_drv; + struct psp_bin_desc ipkeymgr_drv; /* tmr buffer */ struct amdgpu_bo *tmr_bo; @@ -424,6 +428,9 @@ struct amdgpu_psp_funcs { #define psp_bootloader_load_ras_drv(psp) \ ((psp)->funcs->bootloader_load_ras_drv ? \ (psp)->funcs->bootloader_load_ras_drv((psp)) : 0) +#define psp_bootloader_load_ipkeymgr_drv(psp) \ + ((psp)->funcs->bootloader_load_ipkeymgr_drv ? \ + (psp)->funcs->bootloader_load_ipkeymgr_drv((psp)) : 0) #define psp_bootloader_load_sos(psp) \ ((psp)->funcs->bootloader_load_sos ? (psp)->funcs->bootloader_load_sos((psp)) : 0) #define psp_smu_reload_quirk(psp) \ diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c index 8ebab6f22e5a..1adc81a55734 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c @@ -122,6 +122,8 @@ const char *get_ras_block_str(struct ras_common_if *ras_block) #define MAX_UMC_POISON_POLLING_TIME_ASYNC 100 //ms +#define AMDGPU_RAS_RETIRE_PAGE_INTERVAL 100 //ms + enum amdgpu_ras_retire_page_reservation { AMDGPU_RAS_RETIRE_PAGE_RESERVED, AMDGPU_RAS_RETIRE_PAGE_PENDING, @@ -1045,6 +1047,7 @@ static void amdgpu_ras_get_ecc_info(struct amdgpu_device *adev, struct ras_err_d static void amdgpu_ras_error_print_error_data(struct amdgpu_device *adev, struct ras_manager *ras_mgr, struct ras_err_data *err_data, + struct ras_query_context *qctx, const char *blk_name, bool is_ue, bool is_de) @@ -1052,27 +1055,28 @@ static void amdgpu_ras_error_print_error_data(struct amdgpu_device *adev, struct amdgpu_smuio_mcm_config_info *mcm_info; struct ras_err_node *err_node; struct ras_err_info *err_info; + u64 event_id = qctx->event_id; if (is_ue) { for_each_ras_error(err_node, err_data) { err_info = &err_node->err_info; mcm_info = &err_info->mcm_info; if (err_info->ue_count) { - dev_info(adev->dev, "socket: %d, die: %d, " - "%lld new uncorrectable hardware errors detected in %s block\n", - mcm_info->socket_id, - mcm_info->die_id, - err_info->ue_count, - blk_name); + RAS_EVENT_LOG(adev, event_id, "socket: %d, die: %d, " + "%lld new uncorrectable hardware errors detected in %s block\n", + mcm_info->socket_id, + mcm_info->die_id, + err_info->ue_count, + blk_name); } } for_each_ras_error(err_node, &ras_mgr->err_data) { err_info = &err_node->err_info; mcm_info = &err_info->mcm_info; - dev_info(adev->dev, "socket: %d, die: %d, " - "%lld uncorrectable hardware errors detected in total in %s block\n", - mcm_info->socket_id, mcm_info->die_id, err_info->ue_count, blk_name); + RAS_EVENT_LOG(adev, event_id, "socket: %d, die: %d, " + "%lld uncorrectable hardware errors detected in total in %s block\n", + mcm_info->socket_id, mcm_info->die_id, err_info->ue_count, blk_name); } } else { @@ -1081,44 +1085,44 @@ static void amdgpu_ras_error_print_error_data(struct amdgpu_device *adev, err_info = &err_node->err_info; mcm_info = &err_info->mcm_info; if (err_info->de_count) { - dev_info(adev->dev, "socket: %d, die: %d, " - "%lld new deferred hardware errors detected in %s block\n", - mcm_info->socket_id, - mcm_info->die_id, - err_info->de_count, - blk_name); + RAS_EVENT_LOG(adev, event_id, "socket: %d, die: %d, " + "%lld new deferred hardware errors detected in %s block\n", + mcm_info->socket_id, + mcm_info->die_id, + err_info->de_count, + blk_name); } } for_each_ras_error(err_node, &ras_mgr->err_data) { err_info = &err_node->err_info; mcm_info = &err_info->mcm_info; - dev_info(adev->dev, "socket: %d, die: %d, " - "%lld deferred hardware errors detected in total in %s block\n", - mcm_info->socket_id, mcm_info->die_id, - err_info->de_count, blk_name); + RAS_EVENT_LOG(adev, event_id, "socket: %d, die: %d, " + "%lld deferred hardware errors detected in total in %s block\n", + mcm_info->socket_id, mcm_info->die_id, + err_info->de_count, blk_name); } } else { for_each_ras_error(err_node, err_data) { err_info = &err_node->err_info; mcm_info = &err_info->mcm_info; if (err_info->ce_count) { - dev_info(adev->dev, "socket: %d, die: %d, " - "%lld new correctable hardware errors detected in %s block\n", - mcm_info->socket_id, - mcm_info->die_id, - err_info->ce_count, - blk_name); + RAS_EVENT_LOG(adev, event_id, "socket: %d, die: %d, " + "%lld new correctable hardware errors detected in %s block\n", + mcm_info->socket_id, + mcm_info->die_id, + err_info->ce_count, + blk_name); } } for_each_ras_error(err_node, &ras_mgr->err_data) { err_info = &err_node->err_info; mcm_info = &err_info->mcm_info; - dev_info(adev->dev, "socket: %d, die: %d, " - "%lld correctable hardware errors detected in total in %s block\n", - mcm_info->socket_id, mcm_info->die_id, - err_info->ce_count, blk_name); + RAS_EVENT_LOG(adev, event_id, "socket: %d, die: %d, " + "%lld correctable hardware errors detected in total in %s block\n", + mcm_info->socket_id, mcm_info->die_id, + err_info->ce_count, blk_name); } } } @@ -1131,77 +1135,79 @@ static inline bool err_data_has_source_info(struct ras_err_data *data) static void amdgpu_ras_error_generate_report(struct amdgpu_device *adev, struct ras_query_if *query_if, - struct ras_err_data *err_data) + struct ras_err_data *err_data, + struct ras_query_context *qctx) { struct ras_manager *ras_mgr = amdgpu_ras_find_obj(adev, &query_if->head); const char *blk_name = get_ras_block_str(&query_if->head); + u64 event_id = qctx->event_id; if (err_data->ce_count) { if (err_data_has_source_info(err_data)) { - amdgpu_ras_error_print_error_data(adev, ras_mgr, err_data, + amdgpu_ras_error_print_error_data(adev, ras_mgr, err_data, qctx, blk_name, false, false); } else if (!adev->aid_mask && adev->smuio.funcs && adev->smuio.funcs->get_socket_id && adev->smuio.funcs->get_die_id) { - dev_info(adev->dev, "socket: %d, die: %d " - "%ld correctable hardware errors " - "detected in %s block\n", - adev->smuio.funcs->get_socket_id(adev), - adev->smuio.funcs->get_die_id(adev), - ras_mgr->err_data.ce_count, - blk_name); + RAS_EVENT_LOG(adev, event_id, "socket: %d, die: %d " + "%ld correctable hardware errors " + "detected in %s block\n", + adev->smuio.funcs->get_socket_id(adev), + adev->smuio.funcs->get_die_id(adev), + ras_mgr->err_data.ce_count, + blk_name); } else { - dev_info(adev->dev, "%ld correctable hardware errors " - "detected in %s block\n", - ras_mgr->err_data.ce_count, - blk_name); + RAS_EVENT_LOG(adev, event_id, "%ld correctable hardware errors " + "detected in %s block\n", + ras_mgr->err_data.ce_count, + blk_name); } } if (err_data->ue_count) { if (err_data_has_source_info(err_data)) { - amdgpu_ras_error_print_error_data(adev, ras_mgr, err_data, + amdgpu_ras_error_print_error_data(adev, ras_mgr, err_data, qctx, blk_name, true, false); } else if (!adev->aid_mask && adev->smuio.funcs && adev->smuio.funcs->get_socket_id && adev->smuio.funcs->get_die_id) { - dev_info(adev->dev, "socket: %d, die: %d " - "%ld uncorrectable hardware errors " - "detected in %s block\n", - adev->smuio.funcs->get_socket_id(adev), - adev->smuio.funcs->get_die_id(adev), - ras_mgr->err_data.ue_count, - blk_name); + RAS_EVENT_LOG(adev, event_id, "socket: %d, die: %d " + "%ld uncorrectable hardware errors " + "detected in %s block\n", + adev->smuio.funcs->get_socket_id(adev), + adev->smuio.funcs->get_die_id(adev), + ras_mgr->err_data.ue_count, + blk_name); } else { - dev_info(adev->dev, "%ld uncorrectable hardware errors " - "detected in %s block\n", - ras_mgr->err_data.ue_count, - blk_name); + RAS_EVENT_LOG(adev, event_id, "%ld uncorrectable hardware errors " + "detected in %s block\n", + ras_mgr->err_data.ue_count, + blk_name); } } if (err_data->de_count) { if (err_data_has_source_info(err_data)) { - amdgpu_ras_error_print_error_data(adev, ras_mgr, err_data, + amdgpu_ras_error_print_error_data(adev, ras_mgr, err_data, qctx, blk_name, false, true); } else if (!adev->aid_mask && adev->smuio.funcs && adev->smuio.funcs->get_socket_id && adev->smuio.funcs->get_die_id) { - dev_info(adev->dev, "socket: %d, die: %d " - "%ld deferred hardware errors " - "detected in %s block\n", - adev->smuio.funcs->get_socket_id(adev), - adev->smuio.funcs->get_die_id(adev), - ras_mgr->err_data.de_count, - blk_name); + RAS_EVENT_LOG(adev, event_id, "socket: %d, die: %d " + "%ld deferred hardware errors " + "detected in %s block\n", + adev->smuio.funcs->get_socket_id(adev), + adev->smuio.funcs->get_die_id(adev), + ras_mgr->err_data.de_count, + blk_name); } else { - dev_info(adev->dev, "%ld deferred hardware errors " - "detected in %s block\n", - ras_mgr->err_data.de_count, - blk_name); + RAS_EVENT_LOG(adev, event_id, "%ld deferred hardware errors " + "detected in %s block\n", + ras_mgr->err_data.de_count, + blk_name); } } } @@ -1244,6 +1250,10 @@ int amdgpu_ras_bind_aca(struct amdgpu_device *adev, enum amdgpu_ras_block blk, { struct ras_manager *obj; + /* in resume phase, no need to create aca fs node */ + if (adev->in_suspend || amdgpu_in_reset(adev)) + return 0; + obj = get_ras_manager(adev, blk); if (!obj) return -EINVAL; @@ -1265,7 +1275,8 @@ int amdgpu_ras_unbind_aca(struct amdgpu_device *adev, enum amdgpu_ras_block blk) } static int amdgpu_aca_log_ras_error_data(struct amdgpu_device *adev, enum amdgpu_ras_block blk, - enum aca_error_type type, struct ras_err_data *err_data) + enum aca_error_type type, struct ras_err_data *err_data, + struct ras_query_context *qctx) { struct ras_manager *obj; @@ -1273,7 +1284,7 @@ static int amdgpu_aca_log_ras_error_data(struct amdgpu_device *adev, enum amdgpu if (!obj) return -EINVAL; - return amdgpu_aca_get_error_data(adev, &obj->aca_handle, type, err_data); + return amdgpu_aca_get_error_data(adev, &obj->aca_handle, type, err_data, qctx); } ssize_t amdgpu_ras_aca_sysfs_read(struct device *dev, struct device_attribute *attr, @@ -1287,13 +1298,14 @@ ssize_t amdgpu_ras_aca_sysfs_read(struct device *dev, struct device_attribute *a if (amdgpu_ras_query_error_status(obj->adev, &info)) return -EINVAL; - return sysfs_emit(buf, "%s: %lu\n%s: %lu\n", "ue", info.ue_count, - "ce", info.ce_count); + return sysfs_emit(buf, "%s: %lu\n%s: %lu\n%s: %lu\n", "ue", info.ue_count, + "ce", info.ce_count, "de", info.ue_count); } static int amdgpu_ras_query_error_status_helper(struct amdgpu_device *adev, struct ras_query_if *info, struct ras_err_data *err_data, + struct ras_query_context *qctx, unsigned int error_query_mode) { enum amdgpu_ras_block blk = info ? info->head.block : AMDGPU_RAS_BLOCK_COUNT; @@ -1329,17 +1341,21 @@ static int amdgpu_ras_query_error_status_helper(struct amdgpu_device *adev, } } else { if (amdgpu_aca_is_enabled(adev)) { - ret = amdgpu_aca_log_ras_error_data(adev, blk, ACA_ERROR_TYPE_UE, err_data); + ret = amdgpu_aca_log_ras_error_data(adev, blk, ACA_ERROR_TYPE_UE, err_data, qctx); + if (ret) + return ret; + + ret = amdgpu_aca_log_ras_error_data(adev, blk, ACA_ERROR_TYPE_CE, err_data, qctx); if (ret) return ret; - ret = amdgpu_aca_log_ras_error_data(adev, blk, ACA_ERROR_TYPE_CE, err_data); + ret = amdgpu_aca_log_ras_error_data(adev, blk, ACA_ERROR_TYPE_DEFERRED, err_data, qctx); if (ret) return ret; } else { /* FIXME: add code to check return value later */ - amdgpu_mca_smu_log_ras_error(adev, blk, AMDGPU_MCA_ERROR_TYPE_UE, err_data); - amdgpu_mca_smu_log_ras_error(adev, blk, AMDGPU_MCA_ERROR_TYPE_CE, err_data); + amdgpu_mca_smu_log_ras_error(adev, blk, AMDGPU_MCA_ERROR_TYPE_UE, err_data, qctx); + amdgpu_mca_smu_log_ras_error(adev, blk, AMDGPU_MCA_ERROR_TYPE_CE, err_data, qctx); } } @@ -1351,6 +1367,7 @@ int amdgpu_ras_query_error_status(struct amdgpu_device *adev, struct ras_query_i { struct ras_manager *obj = amdgpu_ras_find_obj(adev, &info->head); struct ras_err_data err_data; + struct ras_query_context qctx; unsigned int error_query_mode; int ret; @@ -1364,8 +1381,12 @@ int amdgpu_ras_query_error_status(struct amdgpu_device *adev, struct ras_query_i if (!amdgpu_ras_get_error_query_mode(adev, &error_query_mode)) return -EINVAL; + memset(&qctx, 0, sizeof(qctx)); + qctx.event_id = amdgpu_ras_acquire_event_id(adev, amdgpu_ras_intr_triggered() ? + RAS_EVENT_TYPE_ISR : RAS_EVENT_TYPE_INVALID); ret = amdgpu_ras_query_error_status_helper(adev, info, &err_data, + &qctx, error_query_mode); if (ret) goto out_fini_err_data; @@ -1376,7 +1397,7 @@ int amdgpu_ras_query_error_status(struct amdgpu_device *adev, struct ras_query_i info->ce_count = obj->err_data.ce_count; info->de_count = obj->err_data.de_count; - amdgpu_ras_error_generate_report(adev, info, &err_data); + amdgpu_ras_error_generate_report(adev, info, &err_data, &qctx); out_fini_err_data: amdgpu_ras_error_data_fini(&err_data); @@ -2041,7 +2062,7 @@ static void amdgpu_ras_interrupt_poison_consumption_handler(struct ras_manager * } } - amdgpu_umc_poison_handler(adev, obj->head.block, false); + amdgpu_umc_poison_handler(adev, obj->head.block, 0); if (block_obj->hw_ops && block_obj->hw_ops->handle_poison_consumption) poison_stat = block_obj->hw_ops->handle_poison_consumption(adev); @@ -2061,6 +2082,17 @@ static void amdgpu_ras_interrupt_poison_creation_handler(struct ras_manager *obj { dev_info(obj->adev->dev, "Poison is created\n"); + + if (amdgpu_ip_version(obj->adev, UMC_HWIP, 0) >= IP_VERSION(12, 0, 0)) { + struct amdgpu_ras *con = amdgpu_ras_get_context(obj->adev); + + amdgpu_ras_put_poison_req(obj->adev, + AMDGPU_RAS_BLOCK__UMC, 0, NULL, NULL, false); + + atomic_inc(&con->page_retirement_req_cnt); + + wake_up(&con->page_retirement_wq); + } } static void amdgpu_ras_interrupt_umc_handler(struct ras_manager *obj, @@ -2371,7 +2403,7 @@ static int amdgpu_ras_badpages_read(struct amdgpu_device *adev, .flags = AMDGPU_RAS_RETIRE_PAGE_RESERVED, }; status = amdgpu_vram_mgr_query_page_status(&adev->mman.vram_mgr, - data->bps[i].retired_page); + data->bps[i].retired_page << AMDGPU_GPU_PAGE_SHIFT); if (status == -EBUSY) (*bps)[i].flags = AMDGPU_RAS_RETIRE_PAGE_PENDING; else if (status == -ENOENT) @@ -2384,6 +2416,19 @@ out: return ret; } +static void amdgpu_ras_set_fed_all(struct amdgpu_device *adev, + struct amdgpu_hive_info *hive, bool status) +{ + struct amdgpu_device *tmp_adev; + + if (hive) { + list_for_each_entry(tmp_adev, &hive->device_list, gmc.xgmi.head) + amdgpu_ras_set_fed(tmp_adev, status); + } else { + amdgpu_ras_set_fed(adev, status); + } +} + static void amdgpu_ras_do_recovery(struct work_struct *work) { struct amdgpu_ras *ras = @@ -2393,8 +2438,21 @@ static void amdgpu_ras_do_recovery(struct work_struct *work) struct list_head device_list, *device_list_handle = NULL; struct amdgpu_hive_info *hive = amdgpu_get_xgmi_hive(adev); - if (hive) + if (hive) { atomic_set(&hive->ras_recovery, 1); + + /* If any device which is part of the hive received RAS fatal + * error interrupt, set fatal error status on all. This + * condition will need a recovery, and flag will be cleared + * as part of recovery. + */ + list_for_each_entry(remote_adev, &hive->device_list, + gmc.xgmi.head) + if (amdgpu_ras_get_fed_status(remote_adev)) { + amdgpu_ras_set_fed_all(adev, hive, true); + break; + } + } if (!ras->disable_ras_err_cnt_harvest) { /* Build list of devices to query RAS related errors */ @@ -2439,18 +2497,6 @@ static void amdgpu_ras_do_recovery(struct work_struct *work) ras->gpu_reset_flags &= ~AMDGPU_RAS_GPU_RESET_MODE1_RESET; set_bit(AMDGPU_NEED_FULL_RESET, &reset_context.flags); - /* For any RAS error that needs a full reset to - * recover, set the fatal error status - */ - if (hive) { - list_for_each_entry(remote_adev, - &hive->device_list, - gmc.xgmi.head) - amdgpu_ras_set_fed(remote_adev, - true); - } else { - amdgpu_ras_set_fed(adev, true); - } psp_fatal_error_recovery_quirk(&adev->psp); } } @@ -2516,9 +2562,7 @@ int amdgpu_ras_add_bad_pages(struct amdgpu_device *adev, goto out; } - amdgpu_vram_mgr_reserve_range(&adev->mman.vram_mgr, - bps[i].retired_page << AMDGPU_GPU_PAGE_SHIFT, - AMDGPU_GPU_PAGE_SIZE); + amdgpu_ras_reserve_page(adev, bps[i].retired_page); memcpy(&data->bps[data->count], &bps[i], sizeof(*data->bps)); data->count++; @@ -2674,10 +2718,167 @@ static void amdgpu_ras_validate_threshold(struct amdgpu_device *adev, } } +int amdgpu_ras_put_poison_req(struct amdgpu_device *adev, + enum amdgpu_ras_block block, uint16_t pasid, + pasid_notify pasid_fn, void *data, uint32_t reset) +{ + int ret = 0; + struct ras_poison_msg poison_msg; + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + + memset(&poison_msg, 0, sizeof(poison_msg)); + poison_msg.block = block; + poison_msg.pasid = pasid; + poison_msg.reset = reset; + poison_msg.pasid_fn = pasid_fn; + poison_msg.data = data; + + ret = kfifo_put(&con->poison_fifo, poison_msg); + if (!ret) { + dev_err(adev->dev, "Poison message fifo is full!\n"); + return -ENOSPC; + } + + return 0; +} + +static int amdgpu_ras_get_poison_req(struct amdgpu_device *adev, + struct ras_poison_msg *poison_msg) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + + return kfifo_get(&con->poison_fifo, poison_msg); +} + +static void amdgpu_ras_ecc_log_init(struct ras_ecc_log_info *ecc_log) +{ + mutex_init(&ecc_log->lock); + + /* Set any value as siphash key */ + memset(&ecc_log->ecc_key, 0xad, sizeof(ecc_log->ecc_key)); + + INIT_RADIX_TREE(&ecc_log->de_page_tree, GFP_KERNEL); + ecc_log->de_updated = false; +} + +static void amdgpu_ras_ecc_log_fini(struct ras_ecc_log_info *ecc_log) +{ + struct radix_tree_iter iter; + void __rcu **slot; + struct ras_ecc_err *ecc_err; + + mutex_lock(&ecc_log->lock); + radix_tree_for_each_slot(slot, &ecc_log->de_page_tree, &iter, 0) { + ecc_err = radix_tree_deref_slot(slot); + kfree(ecc_err->err_pages.pfn); + kfree(ecc_err); + radix_tree_iter_delete(&ecc_log->de_page_tree, &iter, slot); + } + mutex_unlock(&ecc_log->lock); + + mutex_destroy(&ecc_log->lock); + ecc_log->de_updated = false; +} + +static void amdgpu_ras_do_page_retirement(struct work_struct *work) +{ + struct amdgpu_ras *con = container_of(work, struct amdgpu_ras, + page_retirement_dwork.work); + struct amdgpu_device *adev = con->adev; + struct ras_err_data err_data; + + if (amdgpu_in_reset(adev) || atomic_read(&con->in_recovery)) + return; + + amdgpu_ras_error_data_init(&err_data); + + amdgpu_umc_handle_bad_pages(adev, &err_data); + + amdgpu_ras_error_data_fini(&err_data); + + mutex_lock(&con->umc_ecc_log.lock); + if (radix_tree_tagged(&con->umc_ecc_log.de_page_tree, + UMC_ECC_NEW_DETECTED_TAG)) + schedule_delayed_work(&con->page_retirement_dwork, + msecs_to_jiffies(AMDGPU_RAS_RETIRE_PAGE_INTERVAL)); + mutex_unlock(&con->umc_ecc_log.lock); +} + +static int amdgpu_ras_query_ecc_status(struct amdgpu_device *adev, + enum amdgpu_ras_block ras_block, uint32_t timeout_ms) +{ + int ret = 0; + struct ras_ecc_log_info *ecc_log; + struct ras_query_if info; + uint32_t timeout = timeout_ms; + struct amdgpu_ras *ras = amdgpu_ras_get_context(adev); + + memset(&info, 0, sizeof(info)); + info.head.block = ras_block; + + ecc_log = &ras->umc_ecc_log; + ecc_log->de_updated = false; + do { + ret = amdgpu_ras_query_error_status(adev, &info); + if (ret) { + dev_err(adev->dev, "Failed to query ras error! ret:%d\n", ret); + return ret; + } + + if (timeout && !ecc_log->de_updated) { + msleep(1); + timeout--; + } + } while (timeout && !ecc_log->de_updated); + + if (timeout_ms && !timeout) { + dev_warn(adev->dev, "Can't find deferred error\n"); + return -ETIMEDOUT; + } + + return 0; +} + +static void amdgpu_ras_poison_creation_handler(struct amdgpu_device *adev, + uint32_t timeout) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + int ret; + + ret = amdgpu_ras_query_ecc_status(adev, AMDGPU_RAS_BLOCK__UMC, timeout); + if (!ret) + schedule_delayed_work(&con->page_retirement_dwork, 0); +} + +static int amdgpu_ras_poison_consumption_handler(struct amdgpu_device *adev, + struct ras_poison_msg *poison_msg) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + uint32_t reset = poison_msg->reset; + uint16_t pasid = poison_msg->pasid; + + kgd2kfd_set_sram_ecc_flag(adev->kfd.dev); + + if (poison_msg->pasid_fn) + poison_msg->pasid_fn(adev, pasid, poison_msg->data); + + if (reset) { + flush_delayed_work(&con->page_retirement_dwork); + + con->gpu_reset_flags |= reset; + amdgpu_ras_reset_gpu(adev); + } + + return 0; +} + static int amdgpu_ras_page_retirement_thread(void *param) { struct amdgpu_device *adev = (struct amdgpu_device *)param; struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + struct ras_poison_msg poison_msg; + enum amdgpu_ras_block ras_block; + bool poison_creation_is_handled = false; while (!kthread_should_stop()) { @@ -2688,13 +2889,34 @@ static int amdgpu_ras_page_retirement_thread(void *param) if (kthread_should_stop()) break; - dev_info(adev->dev, "Start processing page retirement. request:%d\n", - atomic_read(&con->page_retirement_req_cnt)); - atomic_dec(&con->page_retirement_req_cnt); - amdgpu_umc_bad_page_polling_timeout(adev, - false, MAX_UMC_POISON_POLLING_TIME_ASYNC); + if (!amdgpu_ras_get_poison_req(adev, &poison_msg)) + continue; + + ras_block = poison_msg.block; + + dev_info(adev->dev, "Start processing ras block %s(%d)\n", + ras_block_str(ras_block), ras_block); + + if (ras_block == AMDGPU_RAS_BLOCK__UMC) { + amdgpu_ras_poison_creation_handler(adev, + MAX_UMC_POISON_POLLING_TIME_ASYNC); + poison_creation_is_handled = true; + } else { + /* poison_creation_is_handled: + * false: no poison creation interrupt, but it has poison + * consumption interrupt. + * true: It has poison creation interrupt at the beginning, + * but it has no poison creation interrupt later. + */ + amdgpu_ras_poison_creation_handler(adev, + poison_creation_is_handled ? + 0 : MAX_UMC_POISON_POLLING_TIME_ASYNC); + + amdgpu_ras_poison_consumption_handler(adev, &poison_msg); + poison_creation_is_handled = false; + } } return 0; @@ -2763,6 +2985,8 @@ int amdgpu_ras_recovery_init(struct amdgpu_device *adev) } } + mutex_init(&con->page_rsv_lock); + INIT_KFIFO(con->poison_fifo); mutex_init(&con->page_retirement_lock); init_waitqueue_head(&con->page_retirement_wq); atomic_set(&con->page_retirement_req_cnt, 0); @@ -2773,6 +2997,8 @@ int amdgpu_ras_recovery_init(struct amdgpu_device *adev) dev_warn(adev->dev, "Failed to create umc_page_retirement thread!!!\n"); } + INIT_DELAYED_WORK(&con->page_retirement_dwork, amdgpu_ras_do_page_retirement); + amdgpu_ras_ecc_log_init(&con->umc_ecc_log); #ifdef CONFIG_X86_MCE_AMD if ((adev->asic_type == CHIP_ALDEBARAN) && (adev->gmc.xgmi.connected_to_cpu)) @@ -2813,8 +3039,14 @@ static int amdgpu_ras_recovery_fini(struct amdgpu_device *adev) atomic_set(&con->page_retirement_req_cnt, 0); + mutex_destroy(&con->page_rsv_lock); + cancel_work_sync(&con->recovery_work); + cancel_delayed_work_sync(&con->page_retirement_dwork); + + amdgpu_ras_ecc_log_fini(&con->umc_ecc_log); + mutex_lock(&con->recovery_lock); con->eh_data = NULL; kfree(data->bps); @@ -3036,6 +3268,35 @@ static int amdgpu_get_ras_schema(struct amdgpu_device *adev) AMDGPU_RAS_ERROR__PARITY; } +static void ras_event_mgr_init(struct ras_event_manager *mgr) +{ + int i; + + for (i = 0; i < ARRAY_SIZE(mgr->seqnos); i++) + atomic64_set(&mgr->seqnos[i], 0); +} + +static void amdgpu_ras_event_mgr_init(struct amdgpu_device *adev) +{ + struct amdgpu_ras *ras = amdgpu_ras_get_context(adev); + struct amdgpu_hive_info *hive; + + if (!ras) + return; + + hive = amdgpu_get_xgmi_hive(adev); + ras->event_mgr = hive ? &hive->event_mgr : &ras->__event_mgr; + + /* init event manager with node 0 on xgmi system */ + if (!amdgpu_in_reset(adev)) { + if (!hive || adev->gmc.xgmi.node_id == 0) + ras_event_mgr_init(ras->event_mgr); + } + + if (hive) + amdgpu_put_xgmi_hive(hive); +} + int amdgpu_ras_init(struct amdgpu_device *adev) { struct amdgpu_ras *con = amdgpu_ras_get_context(adev); @@ -3356,6 +3617,8 @@ int amdgpu_ras_late_init(struct amdgpu_device *adev) if (amdgpu_sriov_vf(adev)) return 0; + amdgpu_ras_event_mgr_init(adev); + if (amdgpu_aca_is_enabled(adev)) { if (amdgpu_in_reset(adev)) r = amdgpu_aca_reset(adev); @@ -3472,14 +3735,39 @@ void amdgpu_ras_set_fed(struct amdgpu_device *adev, bool status) atomic_set(&ras->fed, !!status); } +bool amdgpu_ras_event_id_is_valid(struct amdgpu_device *adev, u64 id) +{ + return !(id & BIT_ULL(63)); +} + +u64 amdgpu_ras_acquire_event_id(struct amdgpu_device *adev, enum ras_event_type type) +{ + struct amdgpu_ras *ras = amdgpu_ras_get_context(adev); + u64 id; + + switch (type) { + case RAS_EVENT_TYPE_ISR: + id = (u64)atomic64_read(&ras->event_mgr->seqnos[type]); + break; + case RAS_EVENT_TYPE_INVALID: + default: + id = BIT_ULL(63) | 0ULL; + break; + } + + return id; +} + void amdgpu_ras_global_ras_isr(struct amdgpu_device *adev) { if (atomic_cmpxchg(&amdgpu_ras_in_intr, 0, 1) == 0) { struct amdgpu_ras *ras = amdgpu_ras_get_context(adev); + u64 event_id = (u64)atomic64_inc_return(&ras->event_mgr->seqnos[RAS_EVENT_TYPE_ISR]); - dev_info(adev->dev, "uncorrectable hardware error" - "(ERREVENT_ATHUB_INTERRUPT) detected!\n"); + RAS_EVENT_LOG(adev, event_id, "uncorrectable hardware error" + "(ERREVENT_ATHUB_INTERRUPT) detected!\n"); + amdgpu_ras_set_fed(adev, true); ras->gpu_reset_flags |= AMDGPU_RAS_GPU_RESET_MODE1_RESET; amdgpu_ras_reset_gpu(adev); } @@ -3998,6 +4286,8 @@ void amdgpu_ras_add_mca_err_addr(struct ras_err_info *err_info, struct ras_err_a { struct ras_err_addr *mca_err_addr; + /* This function will be retired. */ + return; mca_err_addr = kzalloc(sizeof(*mca_err_addr), GFP_KERNEL); if (!mca_err_addr) return; @@ -4195,3 +4485,19 @@ void amdgpu_ras_query_boot_status(struct amdgpu_device *adev, u32 num_instances) amdgpu_ras_boot_time_error_reporting(adev, i, boot_error); } } + +int amdgpu_ras_reserve_page(struct amdgpu_device *adev, uint64_t pfn) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + struct amdgpu_vram_mgr *mgr = &adev->mman.vram_mgr; + uint64_t start = pfn << AMDGPU_GPU_PAGE_SHIFT; + int ret = 0; + + mutex_lock(&con->page_rsv_lock); + ret = amdgpu_vram_mgr_query_page_status(mgr, start); + if (ret == -ENOENT) + ret = amdgpu_vram_mgr_reserve_range(mgr, start, AMDGPU_GPU_PAGE_SIZE); + mutex_unlock(&con->page_rsv_lock); + + return ret; +} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h index e0f8ce9d8440..7021c4a66fb5 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h @@ -26,6 +26,9 @@ #include <linux/debugfs.h> #include <linux/list.h> +#include <linux/kfifo.h> +#include <linux/radix-tree.h> +#include <linux/siphash.h> #include "ta_ras_if.h" #include "amdgpu_ras_eeprom.h" #include "amdgpu_smuio.h" @@ -43,7 +46,7 @@ struct amdgpu_iv_entry; #define AMDGPU_RAS_GPU_ERR_HBM_BIST_TEST(x) AMDGPU_GET_REG_FIELD(x, 7, 7) #define AMDGPU_RAS_GPU_ERR_SOCKET_ID(x) AMDGPU_GET_REG_FIELD(x, 10, 8) #define AMDGPU_RAS_GPU_ERR_AID_ID(x) AMDGPU_GET_REG_FIELD(x, 12, 11) -#define AMDGPU_RAS_GPU_ERR_HBM_ID(x) AMDGPU_GET_REG_FIELD(x, 13, 13) +#define AMDGPU_RAS_GPU_ERR_HBM_ID(x) AMDGPU_GET_REG_FIELD(x, 14, 13) #define AMDGPU_RAS_GPU_ERR_BOOT_STATUS(x) AMDGPU_GET_REG_FIELD(x, 31, 31) #define AMDGPU_RAS_BOOT_STATUS_POLLING_LIMIT 1000 @@ -64,6 +67,14 @@ struct amdgpu_iv_entry; /* The high three bits indicates socketid */ #define AMDGPU_RAS_GET_FEATURES(val) ((val) & ~AMDGPU_RAS_FEATURES_SOCKETID_MASK) +#define RAS_EVENT_LOG(_adev, _id, _fmt, ...) \ +do { \ + if (amdgpu_ras_event_id_is_valid((_adev), (_id))) \ + dev_info((_adev)->dev, "{%llu}" _fmt, (_id), ##__VA_ARGS__); \ + else \ + dev_info((_adev)->dev, _fmt, ##__VA_ARGS__); \ +} while (0) + enum amdgpu_ras_block { AMDGPU_RAS_BLOCK__UMC = 0, AMDGPU_RAS_BLOCK__SDMA, @@ -419,6 +430,52 @@ struct umc_ecc_info { int record_ce_addr_supported; }; +enum ras_event_type { + RAS_EVENT_TYPE_INVALID = -1, + RAS_EVENT_TYPE_ISR = 0, + RAS_EVENT_TYPE_COUNT, +}; + +struct ras_event_manager { + atomic64_t seqnos[RAS_EVENT_TYPE_COUNT]; +}; + +struct ras_query_context { + enum ras_event_type type; + u64 event_id; +}; + +typedef int (*pasid_notify)(struct amdgpu_device *adev, + uint16_t pasid, void *data); + +struct ras_poison_msg { + enum amdgpu_ras_block block; + uint16_t pasid; + uint32_t reset; + pasid_notify pasid_fn; + void *data; +}; + +struct ras_err_pages { + uint32_t count; + uint64_t *pfn; +}; + +struct ras_ecc_err { + u64 hash_index; + uint64_t status; + uint64_t ipid; + uint64_t addr; + struct ras_err_pages err_pages; +}; + +struct ras_ecc_log_info { + struct mutex lock; + siphash_key_t ecc_key; + struct radix_tree_root de_page_tree; + bool de_updated; +}; + struct amdgpu_ras { /* ras infrastructure */ /* for ras itself. */ @@ -477,8 +534,18 @@ struct amdgpu_ras { wait_queue_head_t page_retirement_wq; struct mutex page_retirement_lock; atomic_t page_retirement_req_cnt; + struct mutex page_rsv_lock; + DECLARE_KFIFO(poison_fifo, struct ras_poison_msg, 128); + struct ras_ecc_log_info umc_ecc_log; + struct delayed_work page_retirement_dwork; + /* Fatal error detected flag */ atomic_t fed; + + /* RAS event manager */ + struct ras_event_manager __event_mgr; + struct ras_event_manager *event_mgr; + }; struct ras_fs_data { @@ -512,6 +579,7 @@ struct ras_err_data { unsigned long de_count; unsigned long err_addr_cnt; struct eeprom_table_record *err_addr; + unsigned long err_addr_len; u32 err_list_count; struct list_head err_node_list; }; @@ -879,4 +947,13 @@ void amdgpu_ras_del_mca_err_addr(struct ras_err_info *err_info, void amdgpu_ras_set_fed(struct amdgpu_device *adev, bool status); bool amdgpu_ras_get_fed_status(struct amdgpu_device *adev); +bool amdgpu_ras_event_id_is_valid(struct amdgpu_device *adev, u64 id); +u64 amdgpu_ras_acquire_event_id(struct amdgpu_device *adev, enum ras_event_type type); + +int amdgpu_ras_reserve_page(struct amdgpu_device *adev, uint64_t pfn); + +int amdgpu_ras_put_poison_req(struct amdgpu_device *adev, + enum amdgpu_ras_block block, uint16_t pasid, + pasid_notify pasid_fn, void *data, uint32_t reset); + #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c index b12808c0c331..06a62a8a992e 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c @@ -404,6 +404,22 @@ static int amdgpu_ras_eeprom_correct_header_tag( return res; } +static void amdgpu_ras_set_eeprom_table_version(struct amdgpu_ras_eeprom_control *control) +{ + struct amdgpu_device *adev = to_amdgpu_device(control); + struct amdgpu_ras_eeprom_table_header *hdr = &control->tbl_hdr; + + switch (amdgpu_ip_version(adev, UMC_HWIP, 0)) { + case IP_VERSION(8, 10, 0): + case IP_VERSION(12, 0, 0): + hdr->version = RAS_TABLE_VER_V2_1; + return; + default: + hdr->version = RAS_TABLE_VER_V1; + return; + } +} + /** * amdgpu_ras_eeprom_reset_table -- Reset the RAS EEPROM table * @control: pointer to control structure @@ -423,11 +439,7 @@ int amdgpu_ras_eeprom_reset_table(struct amdgpu_ras_eeprom_control *control) mutex_lock(&control->ras_tbl_mutex); hdr->header = RAS_TABLE_HDR_VAL; - if (adev->umc.ras && - adev->umc.ras->set_eeprom_table_version) - adev->umc.ras->set_eeprom_table_version(hdr); - else - hdr->version = RAS_TABLE_VER_V1; + amdgpu_ras_set_eeprom_table_version(control); if (hdr->version == RAS_TABLE_VER_V2_1) { hdr->first_rec_offset = RAS_RECORD_START_V2_1; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_res_cursor.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_res_cursor.h index 381101d2bf05..50fcd86e1033 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_res_cursor.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_res_cursor.h @@ -164,4 +164,29 @@ static inline void amdgpu_res_next(struct amdgpu_res_cursor *cur, uint64_t size) } } +/** + * amdgpu_res_cleared - check if blocks are cleared + * + * @cur: the cursor to extract the block + * + * Check if the @cur block is cleared + */ +static inline bool amdgpu_res_cleared(struct amdgpu_res_cursor *cur) +{ + struct drm_buddy_block *block; + + switch (cur->mem_type) { + case TTM_PL_VRAM: + block = cur->node; + + if (!amdgpu_vram_mgr_is_cleared(block)) + return false; + break; + default: + return false; + } + + return true; +} + #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_reset.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_reset.c index 147100c27c2d..ea4873f6ccd1 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_reset.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_reset.c @@ -21,9 +21,6 @@ * */ -#include <linux/devcoredump.h> -#include <generated/utsrelease.h> - #include "amdgpu_reset.h" #include "aldebaran.h" #include "sienna_cichlid.h" @@ -161,105 +158,3 @@ void amdgpu_device_unlock_reset_domain(struct amdgpu_reset_domain *reset_domain) atomic_set(&reset_domain->in_gpu_reset, 0); up_write(&reset_domain->sem); } - -#ifndef CONFIG_DEV_COREDUMP -void amdgpu_coredump(struct amdgpu_device *adev, bool vram_lost, - struct amdgpu_reset_context *reset_context) -{ -} -#else -static ssize_t -amdgpu_devcoredump_read(char *buffer, loff_t offset, size_t count, - void *data, size_t datalen) -{ - struct drm_printer p; - struct amdgpu_coredump_info *coredump = data; - struct drm_print_iterator iter; - int i; - - iter.data = buffer; - iter.offset = 0; - iter.start = offset; - iter.remain = count; - - p = drm_coredump_printer(&iter); - - drm_printf(&p, "**** AMDGPU Device Coredump ****\n"); - drm_printf(&p, "version: " AMDGPU_COREDUMP_VERSION "\n"); - drm_printf(&p, "kernel: " UTS_RELEASE "\n"); - drm_printf(&p, "module: " KBUILD_MODNAME "\n"); - drm_printf(&p, "time: %lld.%09ld\n", coredump->reset_time.tv_sec, - coredump->reset_time.tv_nsec); - - if (coredump->reset_task_info.pid) - drm_printf(&p, "process_name: %s PID: %d\n", - coredump->reset_task_info.process_name, - coredump->reset_task_info.pid); - - if (coredump->ring) { - drm_printf(&p, "\nRing timed out details\n"); - drm_printf(&p, "IP Type: %d Ring Name: %s\n", - coredump->ring->funcs->type, - coredump->ring->name); - } - - if (coredump->reset_vram_lost) - drm_printf(&p, "VRAM is lost due to GPU reset!\n"); - if (coredump->adev->reset_info.num_regs) { - drm_printf(&p, "AMDGPU register dumps:\nOffset: Value:\n"); - - for (i = 0; i < coredump->adev->reset_info.num_regs; i++) - drm_printf(&p, "0x%08x: 0x%08x\n", - coredump->adev->reset_info.reset_dump_reg_list[i], - coredump->adev->reset_info.reset_dump_reg_value[i]); - } - - return count - iter.remain; -} - -static void amdgpu_devcoredump_free(void *data) -{ - kfree(data); -} - -void amdgpu_coredump(struct amdgpu_device *adev, bool vram_lost, - struct amdgpu_reset_context *reset_context) -{ - struct amdgpu_coredump_info *coredump; - struct drm_device *dev = adev_to_drm(adev); - struct amdgpu_job *job = reset_context->job; - struct drm_sched_job *s_job; - - coredump = kzalloc(sizeof(*coredump), GFP_NOWAIT); - - if (!coredump) { - DRM_ERROR("%s: failed to allocate memory for coredump\n", __func__); - return; - } - - coredump->reset_vram_lost = vram_lost; - - if (reset_context->job && reset_context->job->vm) { - struct amdgpu_task_info *ti; - struct amdgpu_vm *vm = reset_context->job->vm; - - ti = amdgpu_vm_get_task_info_vm(vm); - if (ti) { - coredump->reset_task_info = *ti; - amdgpu_vm_put_task_info(ti); - } - } - - if (job) { - s_job = &job->base; - coredump->ring = to_amdgpu_ring(s_job->sched); - } - - coredump->adev = adev; - - ktime_get_ts64(&coredump->reset_time); - - dev_coredumpm(dev->dev, THIS_MODULE, coredump, 0, GFP_NOWAIT, - amdgpu_devcoredump_read, amdgpu_devcoredump_free); -} -#endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_reset.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_reset.h index 60522963aaca..b11d190ece53 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_reset.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_reset.h @@ -32,6 +32,7 @@ enum AMDGPU_RESET_FLAGS { AMDGPU_NEED_FULL_RESET = 0, AMDGPU_SKIP_HW_RESET = 1, + AMDGPU_SKIP_COREDUMP = 2, }; struct amdgpu_reset_context { @@ -88,19 +89,6 @@ struct amdgpu_reset_domain { atomic_t reset_res; }; -#ifdef CONFIG_DEV_COREDUMP - -#define AMDGPU_COREDUMP_VERSION "1" - -struct amdgpu_coredump_info { - struct amdgpu_device *adev; - struct amdgpu_task_info reset_task_info; - struct timespec64 reset_time; - bool reset_vram_lost; - struct amdgpu_ring *ring; -}; -#endif - int amdgpu_reset_init(struct amdgpu_device *adev); int amdgpu_reset_fini(struct amdgpu_device *adev); @@ -141,9 +129,6 @@ void amdgpu_device_lock_reset_domain(struct amdgpu_reset_domain *reset_domain); void amdgpu_device_unlock_reset_domain(struct amdgpu_reset_domain *reset_domain); -void amdgpu_coredump(struct amdgpu_device *adev, bool vram_lost, - struct amdgpu_reset_context *reset_context); - #define for_each_handler(i, handler, reset_ctl) \ for (i = 0; (i < AMDGPU_RESET_MAX_HANDLERS) && \ (handler = (*reset_ctl->reset_handlers)[i]); \ diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c index 5505d646f43a..06f0a6534a94 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c @@ -524,46 +524,58 @@ static ssize_t amdgpu_debugfs_mqd_read(struct file *f, char __user *buf, { struct amdgpu_ring *ring = file_inode(f)->i_private; volatile u32 *mqd; - int r; + u32 *kbuf; + int r, i; uint32_t value, result; if (*pos & 3 || size & 3) return -EINVAL; - result = 0; + kbuf = kmalloc(ring->mqd_size, GFP_KERNEL); + if (!kbuf) + return -ENOMEM; r = amdgpu_bo_reserve(ring->mqd_obj, false); if (unlikely(r != 0)) - return r; + goto err_free; r = amdgpu_bo_kmap(ring->mqd_obj, (void **)&mqd); - if (r) { - amdgpu_bo_unreserve(ring->mqd_obj); - return r; - } + if (r) + goto err_unreserve; + /* + * Copy to local buffer to avoid put_user(), which might fault + * and acquire mmap_sem, under reservation_ww_class_mutex. + */ + for (i = 0; i < ring->mqd_size/sizeof(u32); i++) + kbuf[i] = mqd[i]; + + amdgpu_bo_kunmap(ring->mqd_obj); + amdgpu_bo_unreserve(ring->mqd_obj); + + result = 0; while (size) { if (*pos >= ring->mqd_size) - goto done; + break; - value = mqd[*pos/4]; + value = kbuf[*pos/4]; r = put_user(value, (uint32_t *)buf); if (r) - goto done; + goto err_free; buf += 4; result += 4; size -= 4; *pos += 4; } -done: - amdgpu_bo_kunmap(ring->mqd_obj); - mqd = NULL; - amdgpu_bo_unreserve(ring->mqd_obj); - if (r) - return r; - + kfree(kbuf); return result; + +err_unreserve: + amdgpu_bo_unreserve(ring->mqd_obj); +err_free: + kfree(kbuf); + return r; } static const struct file_operations amdgpu_debugfs_mqd_fops = { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.h index 173a2a308078..b51a82e711df 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.h @@ -132,7 +132,7 @@ struct amdgpu_buffer_funcs { uint64_t dst_offset, /* number of byte to transfer */ uint32_t byte_count, - bool tmz); + uint32_t copy_flags); /* maximum bytes in a single operation */ uint32_t fill_max_bytes; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_smuio.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_smuio.h index ff4435181055..ec9d12f85f39 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_smuio.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_smuio.h @@ -44,6 +44,7 @@ struct amdgpu_smuio_funcs { u32 (*get_socket_id)(struct amdgpu_device *adev); enum amdgpu_pkg_type (*get_pkg_type)(struct amdgpu_device *adev); bool (*is_host_gpu_xgmi_supported)(struct amdgpu_device *adev); + u64 (*get_gpu_clock_counter)(struct amdgpu_device *adev); }; struct amdgpu_smuio { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h index f539b1d00234..7aafeb763e5d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h @@ -178,10 +178,10 @@ TRACE_EVENT(amdgpu_cs_ioctl, TP_fast_assign( __entry->sched_job_id = job->base.id; - __assign_str(timeline, AMDGPU_JOB_GET_TIMELINE_NAME(job)); + __assign_str(timeline); __entry->context = job->base.s_fence->finished.context; __entry->seqno = job->base.s_fence->finished.seqno; - __assign_str(ring, to_amdgpu_ring(job->base.sched)->name); + __assign_str(ring); __entry->num_ibs = job->num_ibs; ), TP_printk("sched_job=%llu, timeline=%s, context=%u, seqno=%u, ring_name=%s, num_ibs=%u", @@ -203,10 +203,10 @@ TRACE_EVENT(amdgpu_sched_run_job, TP_fast_assign( __entry->sched_job_id = job->base.id; - __assign_str(timeline, AMDGPU_JOB_GET_TIMELINE_NAME(job)); + __assign_str(timeline); __entry->context = job->base.s_fence->finished.context; __entry->seqno = job->base.s_fence->finished.seqno; - __assign_str(ring, to_amdgpu_ring(job->base.sched)->name); + __assign_str(ring); __entry->num_ibs = job->num_ibs; ), TP_printk("sched_job=%llu, timeline=%s, context=%u, seqno=%u, ring_name=%s, num_ibs=%u", @@ -231,7 +231,7 @@ TRACE_EVENT(amdgpu_vm_grab_id, TP_fast_assign( __entry->pasid = vm->pasid; - __assign_str(ring, ring->name); + __assign_str(ring); __entry->vmid = job->vmid; __entry->vm_hub = ring->vm_hub, __entry->pd_addr = job->vm_pd_addr; @@ -425,7 +425,7 @@ TRACE_EVENT(amdgpu_vm_flush, ), TP_fast_assign( - __assign_str(ring, ring->name); + __assign_str(ring); __entry->vmid = vmid; __entry->vm_hub = ring->vm_hub; __entry->pd_addr = pd_addr; @@ -526,7 +526,7 @@ TRACE_EVENT(amdgpu_ib_pipe_sync, ), TP_fast_assign( - __assign_str(ring, sched_job->base.sched->name); + __assign_str(ring); __entry->id = sched_job->base.id; __entry->fence = fence; __entry->ctx = fence->context; @@ -563,7 +563,7 @@ TRACE_EVENT(amdgpu_runpm_reference_dumps, ), TP_fast_assign( __entry->index = index; - __assign_str(func, func); + __assign_str(func); ), TP_printk("amdgpu runpm reference dump 0x%x: 0x%s\n", __entry->index, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c index fc418e670fda..e785f128411d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c @@ -133,7 +133,7 @@ static void amdgpu_evict_flags(struct ttm_buffer_object *bo, } else if (!amdgpu_gmc_vram_full_visible(&adev->gmc) && !(abo->flags & AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED) && - amdgpu_bo_in_cpu_visible_vram(abo)) { + amdgpu_res_cpu_visible(adev, bo->resource)) { /* Try evicting to the CPU inaccessible part of VRAM * first, but only set GTT as busy placement, so this @@ -236,7 +236,7 @@ static int amdgpu_ttm_map_buffer(struct ttm_buffer_object *bo, dst_addr = amdgpu_bo_gpu_offset(adev->gart.bo); dst_addr += window * AMDGPU_GTT_MAX_TRANSFER_SIZE * 8; amdgpu_emit_copy_buffer(adev, &job->ibs[0], src_addr, - dst_addr, num_bytes, false); + dst_addr, num_bytes, 0); amdgpu_ring_pad_ib(ring, &job->ibs[0]); WARN_ON(job->ibs[0].length_dw > num_dw); @@ -296,6 +296,8 @@ int amdgpu_ttm_copy_mem_to_mem(struct amdgpu_device *adev, struct dma_fence *fence = NULL; int r = 0; + uint32_t copy_flags = 0; + if (!adev->mman.buffer_funcs_enabled) { DRM_ERROR("Trying to move memory with ring turned off.\n"); return -EINVAL; @@ -323,8 +325,11 @@ int amdgpu_ttm_copy_mem_to_mem(struct amdgpu_device *adev, if (r) goto error; - r = amdgpu_copy_buffer(ring, from, to, cur_size, - resv, &next, false, true, tmz); + if (tmz) + copy_flags |= AMDGPU_COPY_FLAGS_TMZ; + + r = amdgpu_copy_buffer(ring, from, to, cur_size, resv, + &next, false, true, copy_flags); if (r) goto error; @@ -378,11 +383,12 @@ static int amdgpu_move_blit(struct ttm_buffer_object *bo, (abo->flags & AMDGPU_GEM_CREATE_VRAM_WIPE_ON_RELEASE)) { struct dma_fence *wipe_fence = NULL; - r = amdgpu_fill_buffer(abo, AMDGPU_POISON, NULL, &wipe_fence, - false); + r = amdgpu_fill_buffer(abo, 0, NULL, &wipe_fence, + false); if (r) { goto error; } else if (wipe_fence) { + amdgpu_vram_mgr_set_cleared(bo->resource); dma_fence_put(fence); fence = wipe_fence; } @@ -403,40 +409,55 @@ error: return r; } -/* - * amdgpu_mem_visible - Check that memory can be accessed by ttm_bo_move_memcpy +/** + * amdgpu_res_cpu_visible - Check that resource can be accessed by CPU + * @adev: amdgpu device + * @res: the resource to check * - * Called by amdgpu_bo_move() + * Returns: true if the full resource is CPU visible, false otherwise. */ -static bool amdgpu_mem_visible(struct amdgpu_device *adev, - struct ttm_resource *mem) +bool amdgpu_res_cpu_visible(struct amdgpu_device *adev, + struct ttm_resource *res) { - u64 mem_size = (u64)mem->size; struct amdgpu_res_cursor cursor; - u64 end; - if (mem->mem_type == TTM_PL_SYSTEM || - mem->mem_type == TTM_PL_TT) + if (!res) + return false; + + if (res->mem_type == TTM_PL_SYSTEM || res->mem_type == TTM_PL_TT || + res->mem_type == AMDGPU_PL_PREEMPT || res->mem_type == AMDGPU_PL_DOORBELL) return true; - if (mem->mem_type != TTM_PL_VRAM) + + if (res->mem_type != TTM_PL_VRAM) return false; - amdgpu_res_first(mem, 0, mem_size, &cursor); - end = cursor.start + cursor.size; + amdgpu_res_first(res, 0, res->size, &cursor); while (cursor.remaining) { + if ((cursor.start + cursor.size) > adev->gmc.visible_vram_size) + return false; amdgpu_res_next(&cursor, cursor.size); + } - if (!cursor.remaining) - break; + return true; +} - /* ttm_resource_ioremap only supports contiguous memory */ - if (end != cursor.start) - return false; +/* + * amdgpu_res_copyable - Check that memory can be accessed by ttm_bo_move_memcpy + * + * Called by amdgpu_bo_move() + */ +static bool amdgpu_res_copyable(struct amdgpu_device *adev, + struct ttm_resource *mem) +{ + if (!amdgpu_res_cpu_visible(adev, mem)) + return false; - end = cursor.start + cursor.size; - } + /* ttm_resource_ioremap only supports contiguous memory */ + if (mem->mem_type == TTM_PL_VRAM && + !(mem->placement & TTM_PL_FLAG_CONTIGUOUS)) + return false; - return end <= adev->gmc.visible_vram_size; + return true; } /* @@ -466,14 +487,16 @@ static int amdgpu_bo_move(struct ttm_buffer_object *bo, bool evict, if (!old_mem || (old_mem->mem_type == TTM_PL_SYSTEM && bo->ttm == NULL)) { + amdgpu_bo_move_notify(bo, evict, new_mem); ttm_bo_move_null(bo, new_mem); - goto out; + return 0; } if (old_mem->mem_type == TTM_PL_SYSTEM && (new_mem->mem_type == TTM_PL_TT || new_mem->mem_type == AMDGPU_PL_PREEMPT)) { + amdgpu_bo_move_notify(bo, evict, new_mem); ttm_bo_move_null(bo, new_mem); - goto out; + return 0; } if ((old_mem->mem_type == TTM_PL_TT || old_mem->mem_type == AMDGPU_PL_PREEMPT) && @@ -483,9 +506,10 @@ static int amdgpu_bo_move(struct ttm_buffer_object *bo, bool evict, return r; amdgpu_ttm_backend_unbind(bo->bdev, bo->ttm); + amdgpu_bo_move_notify(bo, evict, new_mem); ttm_resource_free(bo, &bo->resource); ttm_bo_assign_mem(bo, new_mem); - goto out; + return 0; } if (old_mem->mem_type == AMDGPU_PL_GDS || @@ -497,8 +521,9 @@ static int amdgpu_bo_move(struct ttm_buffer_object *bo, bool evict, new_mem->mem_type == AMDGPU_PL_OA || new_mem->mem_type == AMDGPU_PL_DOORBELL) { /* Nothing to save here */ + amdgpu_bo_move_notify(bo, evict, new_mem); ttm_bo_move_null(bo, new_mem); - goto out; + return 0; } if (bo->type == ttm_bo_type_device && @@ -510,27 +535,28 @@ static int amdgpu_bo_move(struct ttm_buffer_object *bo, bool evict, abo->flags &= ~AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED; } - if (adev->mman.buffer_funcs_enabled) { - if (((old_mem->mem_type == TTM_PL_SYSTEM && - new_mem->mem_type == TTM_PL_VRAM) || - (old_mem->mem_type == TTM_PL_VRAM && - new_mem->mem_type == TTM_PL_SYSTEM))) { - hop->fpfn = 0; - hop->lpfn = 0; - hop->mem_type = TTM_PL_TT; - hop->flags = TTM_PL_FLAG_TEMPORARY; - return -EMULTIHOP; - } + if (adev->mman.buffer_funcs_enabled && + ((old_mem->mem_type == TTM_PL_SYSTEM && + new_mem->mem_type == TTM_PL_VRAM) || + (old_mem->mem_type == TTM_PL_VRAM && + new_mem->mem_type == TTM_PL_SYSTEM))) { + hop->fpfn = 0; + hop->lpfn = 0; + hop->mem_type = TTM_PL_TT; + hop->flags = TTM_PL_FLAG_TEMPORARY; + return -EMULTIHOP; + } + amdgpu_bo_move_notify(bo, evict, new_mem); + if (adev->mman.buffer_funcs_enabled) r = amdgpu_move_blit(bo, evict, new_mem, old_mem); - } else { + else r = -ENODEV; - } if (r) { /* Check that all memory is CPU accessible */ - if (!amdgpu_mem_visible(adev, old_mem) || - !amdgpu_mem_visible(adev, new_mem)) { + if (!amdgpu_res_copyable(adev, old_mem) || + !amdgpu_res_copyable(adev, new_mem)) { pr_err("Move buffer fallback to memcpy unavailable\n"); return r; } @@ -540,11 +566,10 @@ static int amdgpu_bo_move(struct ttm_buffer_object *bo, bool evict, return r; } - trace_amdgpu_bo_move(abo, new_mem->mem_type, old_mem->mem_type); -out: - /* update statistics */ + /* update statistics after the move */ + if (evict) + atomic64_inc(&adev->num_evictions); atomic64_add(bo->base.size, &adev->num_bytes_moved); - amdgpu_bo_move_notify(bo, evict); return 0; } @@ -557,7 +582,6 @@ static int amdgpu_ttm_io_mem_reserve(struct ttm_device *bdev, struct ttm_resource *mem) { struct amdgpu_device *adev = amdgpu_ttm_adev(bdev); - size_t bus_size = (size_t)mem->size; switch (mem->mem_type) { case TTM_PL_SYSTEM: @@ -568,9 +592,6 @@ static int amdgpu_ttm_io_mem_reserve(struct ttm_device *bdev, break; case TTM_PL_VRAM: mem->bus.offset = mem->start << PAGE_SHIFT; - /* check if it's visible */ - if ((mem->bus.offset + bus_size) > adev->gmc.visible_vram_size) - return -EINVAL; if (adev->mman.aper_base_kaddr && mem->placement & TTM_PL_FLAG_CONTIGUOUS) @@ -1477,7 +1498,7 @@ static int amdgpu_ttm_access_memory_sdma(struct ttm_buffer_object *bo, swap(src_addr, dst_addr); amdgpu_emit_copy_buffer(adev, &job->ibs[0], src_addr, dst_addr, - PAGE_SIZE, false); + PAGE_SIZE, 0); amdgpu_ring_pad_ib(adev->mman.buffer_funcs_ring, &job->ibs[0]); WARN_ON(job->ibs[0].length_dw > num_dw); @@ -1548,7 +1569,7 @@ static int amdgpu_ttm_access_memory(struct ttm_buffer_object *bo, static void amdgpu_bo_delete_mem_notify(struct ttm_buffer_object *bo) { - amdgpu_bo_move_notify(bo, false); + amdgpu_bo_move_notify(bo, false, NULL); } static struct ttm_device_funcs amdgpu_bo_driver = { @@ -2128,7 +2149,7 @@ int amdgpu_copy_buffer(struct amdgpu_ring *ring, uint64_t src_offset, uint64_t dst_offset, uint32_t byte_count, struct dma_resv *resv, struct dma_fence **fence, bool direct_submit, - bool vm_needs_flush, bool tmz) + bool vm_needs_flush, uint32_t copy_flags) { struct amdgpu_device *adev = ring->adev; unsigned int num_loops, num_dw; @@ -2154,8 +2175,7 @@ int amdgpu_copy_buffer(struct amdgpu_ring *ring, uint64_t src_offset, uint32_t cur_size_in_bytes = min(byte_count, max_bytes); amdgpu_emit_copy_buffer(adev, &job->ibs[0], src_offset, - dst_offset, cur_size_in_bytes, tmz); - + dst_offset, cur_size_in_bytes, copy_flags); src_offset += cur_size_in_bytes; dst_offset += cur_size_in_bytes; byte_count -= cur_size_in_bytes; @@ -2215,6 +2235,71 @@ static int amdgpu_ttm_fill_mem(struct amdgpu_ring *ring, uint32_t src_data, return 0; } +/** + * amdgpu_ttm_clear_buffer - clear memory buffers + * @bo: amdgpu buffer object + * @resv: reservation object + * @fence: dma_fence associated with the operation + * + * Clear the memory buffer resource. + * + * Returns: + * 0 for success or a negative error code on failure. + */ +int amdgpu_ttm_clear_buffer(struct amdgpu_bo *bo, + struct dma_resv *resv, + struct dma_fence **fence) +{ + struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev); + struct amdgpu_ring *ring = adev->mman.buffer_funcs_ring; + struct amdgpu_res_cursor cursor; + u64 addr; + int r; + + if (!adev->mman.buffer_funcs_enabled) + return -EINVAL; + + if (!fence) + return -EINVAL; + + *fence = dma_fence_get_stub(); + + amdgpu_res_first(bo->tbo.resource, 0, amdgpu_bo_size(bo), &cursor); + + mutex_lock(&adev->mman.gtt_window_lock); + while (cursor.remaining) { + struct dma_fence *next = NULL; + u64 size; + + if (amdgpu_res_cleared(&cursor)) { + amdgpu_res_next(&cursor, cursor.size); + continue; + } + + /* Never clear more than 256MiB at once to avoid timeouts */ + size = min(cursor.size, 256ULL << 20); + + r = amdgpu_ttm_map_buffer(&bo->tbo, bo->tbo.resource, &cursor, + 1, ring, false, &size, &addr); + if (r) + goto err; + + r = amdgpu_ttm_fill_mem(ring, 0, addr, size, resv, + &next, true, true); + if (r) + goto err; + + dma_fence_put(*fence); + *fence = next; + + amdgpu_res_next(&cursor, size); + } +err: + mutex_unlock(&adev->mman.gtt_window_lock); + + return r; +} + int amdgpu_fill_buffer(struct amdgpu_bo *bo, uint32_t src_data, struct dma_resv *resv, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h index 65ec82141a8e..b6f53129dea3 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h @@ -38,8 +38,6 @@ #define AMDGPU_GTT_MAX_TRANSFER_SIZE 512 #define AMDGPU_GTT_NUM_TRANSFER_WINDOWS 2 -#define AMDGPU_POISON 0xd0bed0be - extern const struct attribute_group amdgpu_vram_mgr_attr_group; extern const struct attribute_group amdgpu_gtt_mgr_attr_group; @@ -111,6 +109,8 @@ struct amdgpu_copy_mem { unsigned long offset; }; +#define AMDGPU_COPY_FLAGS_TMZ (1 << 0) + int amdgpu_gtt_mgr_init(struct amdgpu_device *adev, uint64_t gtt_size); void amdgpu_gtt_mgr_fini(struct amdgpu_device *adev); int amdgpu_preempt_mgr_init(struct amdgpu_device *adev); @@ -139,6 +139,9 @@ int amdgpu_vram_mgr_reserve_range(struct amdgpu_vram_mgr *mgr, int amdgpu_vram_mgr_query_page_status(struct amdgpu_vram_mgr *mgr, uint64_t start); +bool amdgpu_res_cpu_visible(struct amdgpu_device *adev, + struct ttm_resource *res); + int amdgpu_ttm_init(struct amdgpu_device *adev); void amdgpu_ttm_fini(struct amdgpu_device *adev); void amdgpu_ttm_set_buffer_funcs_status(struct amdgpu_device *adev, @@ -148,13 +151,16 @@ int amdgpu_copy_buffer(struct amdgpu_ring *ring, uint64_t src_offset, uint64_t dst_offset, uint32_t byte_count, struct dma_resv *resv, struct dma_fence **fence, bool direct_submit, - bool vm_needs_flush, bool tmz); + bool vm_needs_flush, uint32_t copy_flags); int amdgpu_ttm_copy_mem_to_mem(struct amdgpu_device *adev, const struct amdgpu_copy_mem *src, const struct amdgpu_copy_mem *dst, uint64_t size, bool tmz, struct dma_resv *resv, struct dma_fence **f); +int amdgpu_ttm_clear_buffer(struct amdgpu_bo *bo, + struct dma_resv *resv, + struct dma_fence **fence); int amdgpu_fill_buffer(struct amdgpu_bo *bo, uint32_t src_data, struct dma_resv *resv, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h index 619445760037..105d4de0613a 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h @@ -125,6 +125,7 @@ enum psp_fw_type { PSP_FW_TYPE_PSP_INTF_DRV, PSP_FW_TYPE_PSP_DBG_DRV, PSP_FW_TYPE_PSP_RAS_DRV, + PSP_FW_TYPE_PSP_IPKEYMGR_DRV, PSP_FW_TYPE_MAX_INDEX, }; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.c index 20436f81856a..540e0f066b26 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.c @@ -21,10 +21,13 @@ * */ +#include <linux/sort.h> #include "amdgpu.h" #include "umc_v6_7.h" #define MAX_UMC_POISON_POLLING_TIME_SYNC 20 //ms +#define MAX_UMC_HASH_STRING_SIZE 256 + static int amdgpu_umc_convert_error_address(struct amdgpu_device *adev, struct ras_err_data *err_data, uint64_t err_addr, uint32_t ch_inst, uint32_t umc_inst) @@ -63,6 +66,8 @@ int amdgpu_umc_page_retirement_mca(struct amdgpu_device *adev, goto out_fini_err_data; } + err_data.err_addr_len = adev->umc.max_ras_err_cnt_per_query; + /* * Translate UMC channel address to Physical address */ @@ -86,7 +91,7 @@ out_fini_err_data: return ret; } -static void amdgpu_umc_handle_bad_pages(struct amdgpu_device *adev, +void amdgpu_umc_handle_bad_pages(struct amdgpu_device *adev, void *ras_error_status) { struct ras_err_data *err_data = (struct ras_err_data *)ras_error_status; @@ -118,6 +123,8 @@ static void amdgpu_umc_handle_bad_pages(struct amdgpu_device *adev, if(!err_data->err_addr) dev_warn(adev->dev, "Failed to alloc memory for " "umc error address record!\n"); + else + err_data->err_addr_len = adev->umc.max_ras_err_cnt_per_query; /* umc query_ras_error_address is also responsible for clearing * error status @@ -143,6 +150,8 @@ static void amdgpu_umc_handle_bad_pages(struct amdgpu_device *adev, if(!err_data->err_addr) dev_warn(adev->dev, "Failed to alloc memory for " "umc error address record!\n"); + else + err_data->err_addr_len = adev->umc.max_ras_err_cnt_per_query; /* umc query_ras_error_address is also responsible for clearing * error status @@ -170,6 +179,7 @@ static void amdgpu_umc_handle_bad_pages(struct amdgpu_device *adev, } kfree(err_data->err_addr); + err_data->err_addr = NULL; mutex_unlock(&con->page_retirement_lock); } @@ -177,7 +187,7 @@ static void amdgpu_umc_handle_bad_pages(struct amdgpu_device *adev, static int amdgpu_umc_do_page_retirement(struct amdgpu_device *adev, void *ras_error_status, struct amdgpu_iv_entry *entry, - bool reset) + uint32_t reset) { struct ras_err_data *err_data = (struct ras_err_data *)ras_error_status; struct amdgpu_ras *con = amdgpu_ras_get_context(adev); @@ -186,9 +196,7 @@ static int amdgpu_umc_do_page_retirement(struct amdgpu_device *adev, amdgpu_umc_handle_bad_pages(adev, ras_error_status); if (err_data->ue_count && reset) { - /* use mode-2 reset for poison consumption */ - if (!entry) - con->gpu_reset_flags |= AMDGPU_RAS_GPU_RESET_MODE2_RESET; + con->gpu_reset_flags |= reset; amdgpu_ras_reset_gpu(adev); } @@ -196,7 +204,7 @@ static int amdgpu_umc_do_page_retirement(struct amdgpu_device *adev, } int amdgpu_umc_bad_page_polling_timeout(struct amdgpu_device *adev, - bool reset, uint32_t timeout_ms) + uint32_t reset, uint32_t timeout_ms) { struct ras_err_data err_data; struct ras_common_if head = { @@ -238,16 +246,16 @@ int amdgpu_umc_bad_page_polling_timeout(struct amdgpu_device *adev, if (reset) { struct amdgpu_ras *con = amdgpu_ras_get_context(adev); - /* use mode-2 reset for poison consumption */ - con->gpu_reset_flags |= AMDGPU_RAS_GPU_RESET_MODE2_RESET; + con->gpu_reset_flags |= reset; amdgpu_ras_reset_gpu(adev); } return 0; } -int amdgpu_umc_poison_handler(struct amdgpu_device *adev, - enum amdgpu_ras_block block, bool reset) +int amdgpu_umc_pasid_poison_handler(struct amdgpu_device *adev, + enum amdgpu_ras_block block, uint16_t pasid, + pasid_notify pasid_fn, void *data, uint32_t reset) { int ret = AMDGPU_RAS_SUCCESS; @@ -285,16 +293,14 @@ int amdgpu_umc_poison_handler(struct amdgpu_device *adev, amdgpu_ras_error_data_fini(&err_data); } else { - if (reset) { - amdgpu_umc_bad_page_polling_timeout(adev, - reset, MAX_UMC_POISON_POLLING_TIME_SYNC); - } else { struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + amdgpu_ras_put_poison_req(adev, + block, pasid, pasid_fn, data, reset); + atomic_inc(&con->page_retirement_req_cnt); wake_up(&con->page_retirement_wq); - } } } else { if (adev->virt.ops && adev->virt.ops->ras_poison_handler) @@ -307,11 +313,19 @@ int amdgpu_umc_poison_handler(struct amdgpu_device *adev, return ret; } +int amdgpu_umc_poison_handler(struct amdgpu_device *adev, + enum amdgpu_ras_block block, uint32_t reset) +{ + return amdgpu_umc_pasid_poison_handler(adev, + block, 0, NULL, NULL, reset); +} + int amdgpu_umc_process_ras_data_cb(struct amdgpu_device *adev, void *ras_error_status, struct amdgpu_iv_entry *entry) { - return amdgpu_umc_do_page_retirement(adev, ras_error_status, entry, true); + return amdgpu_umc_do_page_retirement(adev, ras_error_status, entry, + AMDGPU_RAS_GPU_RESET_MODE1_RESET); } int amdgpu_umc_ras_sw_init(struct amdgpu_device *adev) @@ -388,14 +402,20 @@ int amdgpu_umc_process_ecc_irq(struct amdgpu_device *adev, return 0; } -void amdgpu_umc_fill_error_record(struct ras_err_data *err_data, +int amdgpu_umc_fill_error_record(struct ras_err_data *err_data, uint64_t err_addr, uint64_t retired_page, uint32_t channel_index, uint32_t umc_inst) { - struct eeprom_table_record *err_rec = - &err_data->err_addr[err_data->err_addr_cnt]; + struct eeprom_table_record *err_rec; + + if (!err_data || + !err_data->err_addr || + (err_data->err_addr_cnt >= err_data->err_addr_len)) + return -EINVAL; + + err_rec = &err_data->err_addr[err_data->err_addr_cnt]; err_rec->address = err_addr; /* page frame address is saved */ @@ -407,6 +427,8 @@ void amdgpu_umc_fill_error_record(struct ras_err_data *err_data, err_rec->mcumc_id = umc_inst; err_data->err_addr_cnt++; + + return 0; } int amdgpu_umc_loop_channels(struct amdgpu_device *adev, @@ -439,3 +461,76 @@ int amdgpu_umc_loop_channels(struct amdgpu_device *adev, return 0; } + +int amdgpu_umc_update_ecc_status(struct amdgpu_device *adev, + uint64_t status, uint64_t ipid, uint64_t addr) +{ + if (adev->umc.ras->update_ecc_status) + return adev->umc.ras->update_ecc_status(adev, + status, ipid, addr); + return 0; +} + +static int amdgpu_umc_uint64_cmp(const void *a, const void *b) +{ + uint64_t *addr_a = (uint64_t *)a; + uint64_t *addr_b = (uint64_t *)b; + + if (*addr_a > *addr_b) + return 1; + else if (*addr_a < *addr_b) + return -1; + else + return 0; +} + +/* Use string hash to avoid logging the same bad pages repeatedly */ +int amdgpu_umc_build_pages_hash(struct amdgpu_device *adev, + uint64_t *pfns, int len, uint64_t *val) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + char buf[MAX_UMC_HASH_STRING_SIZE] = {0}; + int offset = 0, i = 0; + uint64_t hash_val; + + if (!pfns || !len) + return -EINVAL; + + sort(pfns, len, sizeof(uint64_t), amdgpu_umc_uint64_cmp, NULL); + + for (i = 0; i < len; i++) + offset += snprintf(&buf[offset], sizeof(buf) - offset, "%llx", pfns[i]); + + hash_val = siphash(buf, offset, &con->umc_ecc_log.ecc_key); + + *val = hash_val; + + return 0; +} + +int amdgpu_umc_logs_ecc_err(struct amdgpu_device *adev, + struct radix_tree_root *ecc_tree, struct ras_ecc_err *ecc_err) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + struct ras_ecc_log_info *ecc_log; + int ret; + + ecc_log = &con->umc_ecc_log; + + mutex_lock(&ecc_log->lock); + ret = radix_tree_insert(ecc_tree, ecc_err->hash_index, ecc_err); + if (!ret) { + struct ras_err_pages *err_pages = &ecc_err->err_pages; + int i; + + /* Reserve memory */ + for (i = 0; i < err_pages->count; i++) + amdgpu_ras_reserve_page(adev, err_pages->pfn[i]); + + radix_tree_tag_set(ecc_tree, + ecc_err->hash_index, UMC_ECC_NEW_DETECTED_TAG); + } + mutex_unlock(&ecc_log->lock); + + return ret; +} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.h index 26d2ae498daf..5f50c69c3cec 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.h @@ -52,6 +52,8 @@ #define LOOP_UMC_EACH_NODE_INST_AND_CH(node_inst, umc_inst, ch_inst) \ LOOP_UMC_NODE_INST((node_inst)) LOOP_UMC_INST_AND_CH((umc_inst), (ch_inst)) +/* Page retirement tag */ +#define UMC_ECC_NEW_DETECTED_TAG 0x1 typedef int (*umc_func)(struct amdgpu_device *adev, uint32_t node_inst, uint32_t umc_inst, uint32_t ch_inst, void *data); @@ -66,8 +68,8 @@ struct amdgpu_umc_ras { void *ras_error_status); bool (*check_ecc_err_status)(struct amdgpu_device *adev, enum amdgpu_mca_error_type type, void *ras_error_status); - /* support different eeprom table version for different asic */ - void (*set_eeprom_table_version)(struct amdgpu_ras_eeprom_table_header *hdr); + int (*update_ecc_status)(struct amdgpu_device *adev, + uint64_t status, uint64_t ipid, uint64_t addr); }; struct amdgpu_umc_funcs { @@ -103,11 +105,14 @@ struct amdgpu_umc { int amdgpu_umc_ras_sw_init(struct amdgpu_device *adev); int amdgpu_umc_ras_late_init(struct amdgpu_device *adev, struct ras_common_if *ras_block); int amdgpu_umc_poison_handler(struct amdgpu_device *adev, - enum amdgpu_ras_block block, bool reset); + enum amdgpu_ras_block block, uint32_t reset); +int amdgpu_umc_pasid_poison_handler(struct amdgpu_device *adev, + enum amdgpu_ras_block block, uint16_t pasid, + pasid_notify pasid_fn, void *data, uint32_t reset); int amdgpu_umc_process_ecc_irq(struct amdgpu_device *adev, struct amdgpu_irq_src *source, struct amdgpu_iv_entry *entry); -void amdgpu_umc_fill_error_record(struct ras_err_data *err_data, +int amdgpu_umc_fill_error_record(struct ras_err_data *err_data, uint64_t err_addr, uint64_t retired_page, uint32_t channel_index, @@ -123,5 +128,15 @@ int amdgpu_umc_loop_channels(struct amdgpu_device *adev, umc_func func, void *data); int amdgpu_umc_bad_page_polling_timeout(struct amdgpu_device *adev, - bool reset, uint32_t timeout_ms); + uint32_t reset, uint32_t timeout_ms); + +int amdgpu_umc_update_ecc_status(struct amdgpu_device *adev, + uint64_t status, uint64_t ipid, uint64_t addr); +int amdgpu_umc_build_pages_hash(struct amdgpu_device *adev, + uint64_t *pfns, int len, uint64_t *val); +int amdgpu_umc_logs_ecc_err(struct amdgpu_device *adev, + struct radix_tree_root *ecc_tree, struct ras_ecc_err *ecc_err); + +void amdgpu_umc_handle_bad_pages(struct amdgpu_device *adev, + void *ras_error_status); #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_umsch_mm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_umsch_mm.c index ab820cf52668..e01c1c8e64c4 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_umsch_mm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_umsch_mm.c @@ -189,10 +189,13 @@ static void setup_vpe_queue(struct amdgpu_device *adev, mqd->rptr_val = 0; mqd->unmapped = 1; + if (adev->vpe.collaborate_mode) + memcpy(++mqd, test->mqd_data_cpu_addr, sizeof(struct MQD_INFO)); + qinfo->mqd_addr = test->mqd_data_gpu_addr; qinfo->csa_addr = test->ctx_data_gpu_addr + offsetof(struct umsch_mm_test_ctx_data, vpe_ctx_csa); - qinfo->doorbell_offset_0 = (adev->doorbell_index.vpe_ring + 1) << 1; + qinfo->doorbell_offset_0 = 0; qinfo->doorbell_offset_1 = 0; } @@ -287,7 +290,10 @@ static int submit_vpe_queue(struct amdgpu_device *adev, struct umsch_mm_test *te ring[5] = 0; mqd->wptr_val = (6 << 2); - // WDOORBELL32(adev->umsch_mm.agdb_index[CONTEXT_PRIORITY_LEVEL_NORMAL], mqd->wptr_val); + if (adev->vpe.collaborate_mode) + (++mqd)->wptr_val = (6 << 2); + + WDOORBELL32(adev->umsch_mm.agdb_index[CONTEXT_PRIORITY_LEVEL_NORMAL], mqd->wptr_val); for (i = 0; i < adev->usec_timeout; i++) { if (*fence == test_pattern) @@ -571,6 +577,7 @@ int amdgpu_umsch_mm_init_microcode(struct amdgpu_umsch_mm *umsch) switch (amdgpu_ip_version(adev, VCN_HWIP, 0)) { case IP_VERSION(4, 0, 5): + case IP_VERSION(4, 0, 6): fw_name = "amdgpu/umsch_mm_4_0_0.bin"; break; default: @@ -750,6 +757,7 @@ static int umsch_mm_early_init(void *handle) switch (amdgpu_ip_version(adev, VCN_HWIP, 0)) { case IP_VERSION(4, 0, 5): + case IP_VERSION(4, 0, 6): umsch_mm_v4_0_set_funcs(&adev->umsch_mm); break; default: @@ -766,6 +774,9 @@ static int umsch_mm_late_init(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; + if (amdgpu_in_reset(adev) || adev->in_s0ix || adev->in_suspend) + return 0; + return umsch_mm_test(adev); } @@ -867,6 +878,8 @@ static const struct amd_ip_funcs umsch_mm_v4_0_ip_funcs = { .hw_fini = umsch_mm_hw_fini, .suspend = umsch_mm_suspend, .resume = umsch_mm_resume, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; const struct amdgpu_ip_block_version umsch_mm_v4_0_ip_block = { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_umsch_mm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_umsch_mm.h index 8258a43a6236..5014b5af95fd 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_umsch_mm.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_umsch_mm.h @@ -33,13 +33,6 @@ enum UMSCH_SWIP_ENGINE_TYPE { UMSCH_SWIP_ENGINE_TYPE_MAX }; -enum UMSCH_SWIP_AFFINITY_TYPE { - UMSCH_SWIP_AFFINITY_TYPE_ANY = 0, - UMSCH_SWIP_AFFINITY_TYPE_VCN0 = 1, - UMSCH_SWIP_AFFINITY_TYPE_VCN1 = 2, - UMSCH_SWIP_AFFINITY_TYPE_MAX -}; - enum UMSCH_CONTEXT_PRIORITY_LEVEL { CONTEXT_PRIORITY_LEVEL_IDLE = 0, CONTEXT_PRIORITY_LEVEL_NORMAL = 1, @@ -51,13 +44,15 @@ enum UMSCH_CONTEXT_PRIORITY_LEVEL { struct umsch_mm_set_resource_input { uint32_t vmid_mask_mm_vcn; uint32_t vmid_mask_mm_vpe; + uint32_t collaboration_mask_vpe; uint32_t logging_vmid; uint32_t engine_mask; union { struct { uint32_t disable_reset : 1; uint32_t disable_umsch_mm_log : 1; - uint32_t reserved : 30; + uint32_t use_rs64mem_for_proc_ctx_csa : 1; + uint32_t reserved : 29; }; uint32_t uint32_all; }; @@ -78,15 +73,18 @@ struct umsch_mm_add_queue_input { uint32_t doorbell_offset_1; enum UMSCH_SWIP_ENGINE_TYPE engine_type; uint32_t affinity; - enum UMSCH_SWIP_AFFINITY_TYPE affinity_type; uint64_t mqd_addr; uint64_t h_context; uint64_t h_queue; uint32_t vm_context_cntl; + uint32_t process_csa_array_index; + uint32_t context_csa_array_index; + struct { uint32_t is_context_suspended : 1; - uint32_t reserved : 31; + uint32_t collaboration_mode : 1; + uint32_t reserved : 30; }; }; @@ -94,6 +92,7 @@ struct umsch_mm_remove_queue_input { uint32_t doorbell_offset_0; uint32_t doorbell_offset_1; uint64_t context_csa_addr; + uint32_t context_csa_array_index; }; struct MQD_INFO { @@ -103,6 +102,7 @@ struct MQD_INFO { uint32_t wptr_val; uint32_t rptr_val; uint32_t unmapped; + uint32_t vmid; }; struct amdgpu_umsch_mm; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c index 59acf424a078..968ca2c84ef7 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c @@ -743,7 +743,8 @@ int amdgpu_vce_ring_parse_cs(struct amdgpu_cs_parser *p, uint32_t created = 0; uint32_t allocated = 0; uint32_t tmp, handle = 0; - uint32_t *size = &tmp; + uint32_t dummy = 0xffffffff; + uint32_t *size = &dummy; unsigned int idx; int i, r = 0; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c index 9c514a606a2f..677eb141554e 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c @@ -93,7 +93,7 @@ static void amdgpu_vcn_idle_work_handler(struct work_struct *work); int amdgpu_vcn_early_init(struct amdgpu_device *adev) { - char ucode_prefix[30]; + char ucode_prefix[25]; char fw_name[40]; int r, i; @@ -185,7 +185,10 @@ int amdgpu_vcn_sw_init(struct amdgpu_device *adev) if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP) bo_size += AMDGPU_GPU_PAGE_ALIGN(le32_to_cpu(hdr->ucode_size_bytes) + 8); - if (amdgpu_ip_version(adev, UVD_HWIP, 0) >= IP_VERSION(4, 0, 0)) { + if (amdgpu_ip_version(adev, UVD_HWIP, 0) >= IP_VERSION(5, 0, 0)) { + fw_shared_size = AMDGPU_GPU_PAGE_ALIGN(sizeof(struct amdgpu_vcn5_fw_shared)); + log_offset = offsetof(struct amdgpu_vcn5_fw_shared, fw_log); + } else if (amdgpu_ip_version(adev, UVD_HWIP, 0) >= IP_VERSION(4, 0, 0)) { fw_shared_size = AMDGPU_GPU_PAGE_ALIGN(sizeof(struct amdgpu_vcn4_fw_shared)); log_offset = offsetof(struct amdgpu_vcn4_fw_shared, fw_log); } else { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h index a418393d89ec..9f06def236fd 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h @@ -454,6 +454,16 @@ struct amdgpu_vcn_rb_metadata { uint8_t pad[26]; }; +struct amdgpu_vcn5_fw_shared { + uint32_t present_flag_0; + uint8_t pad[12]; + struct amdgpu_fw_shared_unified_queue_struct sq; + uint8_t pad1[8]; + struct amdgpu_fw_shared_fw_logging fw_log; + struct amdgpu_fw_shared_rb_setup rb_setup; + uint8_t pad2[4]; +}; + #define VCN_BLOCK_ENCODE_DISABLE_MASK 0x80 #define VCN_BLOCK_DECODE_DISABLE_MASK 0x40 #define VCN_BLOCK_QUEUE_DISABLE_MASK 0xC0 diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c index 7a4eae36778a..54ab51a4ada7 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c @@ -32,6 +32,7 @@ #include "amdgpu.h" #include "amdgpu_ras.h" +#include "amdgpu_reset.h" #include "vi.h" #include "soc15.h" #include "nv.h" @@ -424,7 +425,7 @@ static int amdgpu_virt_read_pf2vf_data(struct amdgpu_device *adev) return -EINVAL; if (pf2vf_info->size > 1024) { - DRM_ERROR("invalid pf2vf message size\n"); + dev_err(adev->dev, "invalid pf2vf message size: 0x%x\n", pf2vf_info->size); return -EINVAL; } @@ -435,7 +436,9 @@ static int amdgpu_virt_read_pf2vf_data(struct amdgpu_device *adev) adev->virt.fw_reserve.p_pf2vf, pf2vf_info->size, adev->virt.fw_reserve.checksum_key, checksum); if (checksum != checkval) { - DRM_ERROR("invalid pf2vf message\n"); + dev_err(adev->dev, + "invalid pf2vf message: header checksum=0x%x calculated checksum=0x%x\n", + checksum, checkval); return -EINVAL; } @@ -449,7 +452,9 @@ static int amdgpu_virt_read_pf2vf_data(struct amdgpu_device *adev) adev->virt.fw_reserve.p_pf2vf, pf2vf_info->size, 0, checksum); if (checksum != checkval) { - DRM_ERROR("invalid pf2vf message\n"); + dev_err(adev->dev, + "invalid pf2vf message: header checksum=0x%x calculated checksum=0x%x\n", + checksum, checkval); return -EINVAL; } @@ -485,7 +490,7 @@ static int amdgpu_virt_read_pf2vf_data(struct amdgpu_device *adev) ((struct amd_sriov_msg_pf2vf_info *)pf2vf_info)->uuid; break; default: - DRM_ERROR("invalid pf2vf version\n"); + dev_err(adev->dev, "invalid pf2vf version: 0x%x\n", pf2vf_info->version); return -EINVAL; } @@ -571,6 +576,11 @@ static int amdgpu_virt_write_vf2pf_data(struct amdgpu_device *adev) vf2pf_info->decode_usage = 0; vf2pf_info->dummy_page_addr = (uint64_t)adev->dummy_page_addr; + vf2pf_info->mes_info_addr = (uint64_t)adev->mes.resource_1_gpu_addr; + + if (adev->mes.resource_1) { + vf2pf_info->mes_info_size = adev->mes.resource_1->tbo.base.size; + } vf2pf_info->checksum = amd_sriov_msg_checksum( vf2pf_info, vf2pf_info->header.size, 0, 0); @@ -584,8 +594,22 @@ static void amdgpu_virt_update_vf2pf_work_item(struct work_struct *work) int ret; ret = amdgpu_virt_read_pf2vf_data(adev); - if (ret) + if (ret) { + adev->virt.vf2pf_update_retry_cnt++; + if ((adev->virt.vf2pf_update_retry_cnt >= AMDGPU_VF2PF_UPDATE_MAX_RETRY_LIMIT) && + amdgpu_sriov_runtime(adev) && !amdgpu_in_reset(adev)) { + amdgpu_ras_set_fed(adev, true); + if (amdgpu_reset_domain_schedule(adev->reset_domain, + &adev->virt.flr_work)) + return; + else + dev_err(adev->dev, "Failed to queue work! at %s", __func__); + } + goto out; + } + + adev->virt.vf2pf_update_retry_cnt = 0; amdgpu_virt_write_vf2pf_data(adev); out: @@ -606,6 +630,7 @@ void amdgpu_virt_init_data_exchange(struct amdgpu_device *adev) adev->virt.fw_reserve.p_pf2vf = NULL; adev->virt.fw_reserve.p_vf2pf = NULL; adev->virt.vf2pf_update_interval_ms = 0; + adev->virt.vf2pf_update_retry_cnt = 0; if (adev->mman.fw_vram_usage_va && adev->mman.drv_vram_usage_va) { DRM_WARN("Currently fw_vram and drv_vram should not have values at the same time!"); @@ -705,12 +730,6 @@ void amdgpu_detect_virtualization(struct amdgpu_device *adev) adev->virt.caps |= AMDGPU_PASSTHROUGH_MODE; } - if (amdgpu_sriov_vf(adev) && adev->asic_type == CHIP_SIENNA_CICHLID) - /* VF MMIO access (except mailbox range) from CPU - * will be blocked during sriov runtime - */ - adev->virt.caps |= AMDGPU_VF_MMIO_ACCESS_PROTECT; - /* we have the ability to check now */ if (amdgpu_sriov_vf(adev)) { switch (adev->asic_type) { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h index 3f59b7b5523f..642f1fd287d8 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h @@ -52,6 +52,8 @@ /* tonga/fiji use this offset */ #define mmBIF_IOV_FUNC_IDENTIFIER 0x1503 +#define AMDGPU_VF2PF_UPDATE_MAX_RETRY_LIMIT 5 + enum amdgpu_sriov_vf_mode { SRIOV_VF_MODE_BARE_METAL = 0, SRIOV_VF_MODE_ONE_VF, @@ -130,6 +132,8 @@ enum AMDGIM_FEATURE_FLAG { AMDGIM_FEATURE_AV1_SUPPORT = (1 << 6), /* VCN RB decouple */ AMDGIM_FEATURE_VCN_RB_DECOUPLE = (1 << 7), + /* MES info */ + AMDGIM_FEATURE_MES_INFO_ENABLE = (1 << 8), }; enum AMDGIM_REG_ACCESS_FLAG { @@ -257,6 +261,7 @@ struct amdgpu_virt { /* vf2pf message */ struct delayed_work vf2pf_work; uint32_t vf2pf_update_interval_ms; + int vf2pf_update_retry_cnt; /* multimedia bandwidth config */ bool is_mm_bw_enabled; @@ -332,6 +337,8 @@ static inline bool is_virtual_machine(void) ((adev)->virt.gim_feature & AMDGIM_FEATURE_AV1_SUPPORT) #define amdgpu_sriov_is_vcn_rb_decouple(adev) \ ((adev)->virt.gim_feature & AMDGIM_FEATURE_VCN_RB_DECOUPLE) +#define amdgpu_sriov_is_mes_info_enable(adev) \ + ((adev)->virt.gim_feature & AMDGIM_FEATURE_MES_INFO_ENABLE) bool amdgpu_virt_mmio_blocked(struct amdgpu_device *adev); void amdgpu_virt_init_setting(struct amdgpu_device *adev); int amdgpu_virt_request_full_gpu(struct amdgpu_device *adev, bool init); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vkms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vkms.c index 8baa2e0935cc..e30eecd02ae1 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vkms.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vkms.c @@ -658,6 +658,8 @@ static const struct amd_ip_funcs amdgpu_vkms_ip_funcs = { .soft_reset = amdgpu_vkms_soft_reset, .set_clockgating_state = amdgpu_vkms_set_clockgating_state, .set_powergating_state = amdgpu_vkms_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; const struct amdgpu_ip_block_version amdgpu_vkms_ip_block = { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index 4299ce386322..4e2391c83d7c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -886,6 +886,44 @@ static void amdgpu_vm_tlb_seq_cb(struct dma_fence *fence, } /** + * amdgpu_vm_tlb_flush - prepare TLB flush + * + * @params: parameters for update + * @fence: input fence to sync TLB flush with + * @tlb_cb: the callback structure + * + * Increments the tlb sequence to make sure that future CS execute a VM flush. + */ +static void +amdgpu_vm_tlb_flush(struct amdgpu_vm_update_params *params, + struct dma_fence **fence, + struct amdgpu_vm_tlb_seq_struct *tlb_cb) +{ + struct amdgpu_vm *vm = params->vm; + + if (!fence || !*fence) + return; + + tlb_cb->vm = vm; + if (!dma_fence_add_callback(*fence, &tlb_cb->cb, + amdgpu_vm_tlb_seq_cb)) { + dma_fence_put(vm->last_tlb_flush); + vm->last_tlb_flush = dma_fence_get(*fence); + } else { + amdgpu_vm_tlb_seq_cb(NULL, &tlb_cb->cb); + } + + /* Prepare a TLB flush fence to be attached to PTs */ + if (!params->unlocked && vm->is_compute_context) { + amdgpu_vm_tlb_fence_create(params->adev, vm, fence); + + /* Makes sure no PD/PT is freed before the flush */ + dma_resv_add_fence(vm->root.bo->tbo.base.resv, *fence, + DMA_RESV_USAGE_BOOKKEEP); + } +} + +/** * amdgpu_vm_update_range - update a range in the vm page table * * @adev: amdgpu_device pointer to use for commands @@ -916,8 +954,8 @@ int amdgpu_vm_update_range(struct amdgpu_device *adev, struct amdgpu_vm *vm, struct ttm_resource *res, dma_addr_t *pages_addr, struct dma_fence **fence) { - struct amdgpu_vm_update_params params; struct amdgpu_vm_tlb_seq_struct *tlb_cb; + struct amdgpu_vm_update_params params; struct amdgpu_res_cursor cursor; enum amdgpu_sync_mode sync_mode; int r, idx; @@ -927,8 +965,8 @@ int amdgpu_vm_update_range(struct amdgpu_device *adev, struct amdgpu_vm *vm, tlb_cb = kmalloc(sizeof(*tlb_cb), GFP_KERNEL); if (!tlb_cb) { - r = -ENOMEM; - goto error_unlock; + drm_dev_exit(idx); + return -ENOMEM; } /* Vega20+XGMI where PTEs get inadvertently cached in L2 texture cache, @@ -948,7 +986,9 @@ int amdgpu_vm_update_range(struct amdgpu_device *adev, struct amdgpu_vm *vm, params.immediate = immediate; params.pages_addr = pages_addr; params.unlocked = unlocked; + params.needs_flush = flush_tlb; params.allow_override = allow_override; + INIT_LIST_HEAD(¶ms.tlb_flush_waitlist); /* Implicitly sync to command submissions in the same VM before * unmapping. Sync to moving fences before mapping. @@ -1031,24 +1071,18 @@ int amdgpu_vm_update_range(struct amdgpu_device *adev, struct amdgpu_vm *vm, } r = vm->update_funcs->commit(¶ms, fence); + if (r) + goto error_free; - if (flush_tlb || params.table_freed) { - tlb_cb->vm = vm; - if (fence && *fence && - !dma_fence_add_callback(*fence, &tlb_cb->cb, - amdgpu_vm_tlb_seq_cb)) { - dma_fence_put(vm->last_tlb_flush); - vm->last_tlb_flush = dma_fence_get(*fence); - } else { - amdgpu_vm_tlb_seq_cb(NULL, &tlb_cb->cb); - } + if (params.needs_flush) { + amdgpu_vm_tlb_flush(¶ms, fence, tlb_cb); tlb_cb = NULL; } + amdgpu_vm_pt_free_list(adev, ¶ms); + error_free: kfree(tlb_cb); - -error_unlock: amdgpu_vm_eviction_unlock(vm); drm_dev_exit(idx); return r; @@ -1613,6 +1647,37 @@ static void amdgpu_vm_bo_insert_map(struct amdgpu_device *adev, trace_amdgpu_vm_bo_map(bo_va, mapping); } +/* Validate operation parameters to prevent potential abuse */ +static int amdgpu_vm_verify_parameters(struct amdgpu_device *adev, + struct amdgpu_bo *bo, + uint64_t saddr, + uint64_t offset, + uint64_t size) +{ + uint64_t tmp, lpfn; + + if (saddr & AMDGPU_GPU_PAGE_MASK + || offset & AMDGPU_GPU_PAGE_MASK + || size & AMDGPU_GPU_PAGE_MASK) + return -EINVAL; + + if (check_add_overflow(saddr, size, &tmp) + || check_add_overflow(offset, size, &tmp) + || size == 0 /* which also leads to end < begin */) + return -EINVAL; + + /* make sure object fit at this offset */ + if (bo && offset + size > amdgpu_bo_size(bo)) + return -EINVAL; + + /* Ensure last pfn not exceed max_pfn */ + lpfn = (saddr + size - 1) >> AMDGPU_GPU_PAGE_SHIFT; + if (lpfn >= adev->vm_manager.max_pfn) + return -EINVAL; + + return 0; +} + /** * amdgpu_vm_bo_map - map bo inside a vm * @@ -1639,21 +1704,14 @@ int amdgpu_vm_bo_map(struct amdgpu_device *adev, struct amdgpu_bo *bo = bo_va->base.bo; struct amdgpu_vm *vm = bo_va->base.vm; uint64_t eaddr; + int r; - /* validate the parameters */ - if (saddr & ~PAGE_MASK || offset & ~PAGE_MASK || size & ~PAGE_MASK) - return -EINVAL; - if (saddr + size <= saddr || offset + size <= offset) - return -EINVAL; - - /* make sure object fit at this offset */ - eaddr = saddr + size - 1; - if ((bo && offset + size > amdgpu_bo_size(bo)) || - (eaddr >= adev->vm_manager.max_pfn << AMDGPU_GPU_PAGE_SHIFT)) - return -EINVAL; + r = amdgpu_vm_verify_parameters(adev, bo, saddr, offset, size); + if (r) + return r; saddr /= AMDGPU_GPU_PAGE_SIZE; - eaddr /= AMDGPU_GPU_PAGE_SIZE; + eaddr = saddr + (size - 1) / AMDGPU_GPU_PAGE_SIZE; tmp = amdgpu_vm_it_iter_first(&vm->va, saddr, eaddr); if (tmp) { @@ -1706,17 +1764,9 @@ int amdgpu_vm_bo_replace_map(struct amdgpu_device *adev, uint64_t eaddr; int r; - /* validate the parameters */ - if (saddr & ~PAGE_MASK || offset & ~PAGE_MASK || size & ~PAGE_MASK) - return -EINVAL; - if (saddr + size <= saddr || offset + size <= offset) - return -EINVAL; - - /* make sure object fit at this offset */ - eaddr = saddr + size - 1; - if ((bo && offset + size > amdgpu_bo_size(bo)) || - (eaddr >= adev->vm_manager.max_pfn << AMDGPU_GPU_PAGE_SHIFT)) - return -EINVAL; + r = amdgpu_vm_verify_parameters(adev, bo, saddr, offset, size); + if (r) + return r; /* Allocate all the needed memory */ mapping = kmalloc(sizeof(*mapping), GFP_KERNEL); @@ -1730,7 +1780,7 @@ int amdgpu_vm_bo_replace_map(struct amdgpu_device *adev, } saddr /= AMDGPU_GPU_PAGE_SIZE; - eaddr /= AMDGPU_GPU_PAGE_SIZE; + eaddr = saddr + (size - 1) / AMDGPU_GPU_PAGE_SIZE; mapping->start = saddr; mapping->last = eaddr; @@ -1817,10 +1867,14 @@ int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev, struct amdgpu_bo_va_mapping *before, *after, *tmp, *next; LIST_HEAD(removed); uint64_t eaddr; + int r; + + r = amdgpu_vm_verify_parameters(adev, NULL, saddr, 0, size); + if (r) + return r; - eaddr = saddr + size - 1; saddr /= AMDGPU_GPU_PAGE_SIZE; - eaddr /= AMDGPU_GPU_PAGE_SIZE; + eaddr = saddr + (size - 1) / AMDGPU_GPU_PAGE_SIZE; /* Allocate all the needed memory */ before = kzalloc(sizeof(*before), GFP_KERNEL); @@ -2391,6 +2445,7 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm, mutex_init(&vm->eviction_lock); vm->evicting = false; + vm->tlb_fence_context = dma_fence_context_alloc(1); r = amdgpu_vm_pt_create(adev, vm, adev->vm_manager.root_level, false, &root, xcp_id); @@ -2924,6 +2979,14 @@ void amdgpu_vm_update_fault_cache(struct amdgpu_device *adev, if (vm && status) { vm->fault_info.addr = addr; vm->fault_info.status = status; + /* + * Update the fault information globally for later usage + * when vm could be stale or freed. + */ + adev->vm_manager.fault_info.addr = addr; + adev->vm_manager.fault_info.vmhub = vmhub; + adev->vm_manager.fault_info.status = status; + if (AMDGPU_IS_GFXHUB(vmhub)) { vm->fault_info.vmhub = AMDGPU_VMHUB_TYPE_GFX; vm->fault_info.vmhub |= diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h index 047ec1930d12..54d7da396de0 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h @@ -257,15 +257,20 @@ struct amdgpu_vm_update_params { unsigned int num_dw_left; /** - * @table_freed: return true if page table is freed when updating + * @needs_flush: true whenever we need to invalidate the TLB */ - bool table_freed; + bool needs_flush; /** * @allow_override: true for memory that is not uncached: allows MTYPE * to be overridden for NUMA local memory. */ bool allow_override; + + /** + * @tlb_flush_waitlist: temporary storage for BOs until tlb_flush + */ + struct list_head tlb_flush_waitlist; }; struct amdgpu_vm_update_funcs { @@ -342,6 +347,7 @@ struct amdgpu_vm { atomic64_t tlb_seq; struct dma_fence *last_tlb_flush; atomic64_t kfd_last_flushed_seq; + uint64_t tlb_fence_context; /* How many times we had to re-generate the page tables */ uint64_t generation; @@ -422,6 +428,8 @@ struct amdgpu_vm_manager { * look up VM of a page fault */ struct xarray pasids; + /* Global registration of recent page fault information */ + struct amdgpu_vm_fault_info fault_info; }; struct amdgpu_bo_va_mapping; @@ -544,6 +552,8 @@ int amdgpu_vm_ptes_update(struct amdgpu_vm_update_params *params, uint64_t start, uint64_t end, uint64_t dst, uint64_t flags); void amdgpu_vm_pt_free_work(struct work_struct *work); +void amdgpu_vm_pt_free_list(struct amdgpu_device *adev, + struct amdgpu_vm_update_params *params); #if defined(CONFIG_DEBUG_FS) void amdgpu_debugfs_vm_bo_info(struct amdgpu_vm *vm, struct seq_file *m); @@ -609,5 +619,8 @@ void amdgpu_vm_update_fault_cache(struct amdgpu_device *adev, uint64_t addr, uint32_t status, unsigned int vmhub); +void amdgpu_vm_tlb_fence_create(struct amdgpu_device *adev, + struct amdgpu_vm *vm, + struct dma_fence **fence); #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_cpu.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_cpu.c index 6e31621452de..3895bd7d176a 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_cpu.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_cpu.c @@ -108,7 +108,9 @@ static int amdgpu_vm_cpu_update(struct amdgpu_vm_update_params *p, static int amdgpu_vm_cpu_commit(struct amdgpu_vm_update_params *p, struct dma_fence **fence) { - /* Flush HDP */ + if (p->needs_flush) + atomic64_inc(&p->vm->tlb_seq); + mb(); amdgpu_device_flush_hdp(p->adev, NULL); return 0; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_pt.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_pt.c index 124389a6bf48..f07647a9a9d9 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_pt.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_pt.c @@ -622,40 +622,58 @@ void amdgpu_vm_pt_free_work(struct work_struct *work) } /** - * amdgpu_vm_pt_free_dfs - free PD/PT levels + * amdgpu_vm_pt_free_list - free PD/PT levels * * @adev: amdgpu device structure - * @vm: amdgpu vm structure - * @start: optional cursor where to start freeing PDs/PTs - * @unlocked: vm resv unlock status + * @params: see amdgpu_vm_update_params definition * - * Free the page directory or page table level and all sub levels. + * Free the page directory objects saved in the flush list */ -static void amdgpu_vm_pt_free_dfs(struct amdgpu_device *adev, - struct amdgpu_vm *vm, - struct amdgpu_vm_pt_cursor *start, - bool unlocked) +void amdgpu_vm_pt_free_list(struct amdgpu_device *adev, + struct amdgpu_vm_update_params *params) { - struct amdgpu_vm_pt_cursor cursor; - struct amdgpu_vm_bo_base *entry; + struct amdgpu_vm_bo_base *entry, *next; + struct amdgpu_vm *vm = params->vm; + bool unlocked = params->unlocked; + + if (list_empty(¶ms->tlb_flush_waitlist)) + return; if (unlocked) { spin_lock(&vm->status_lock); - for_each_amdgpu_vm_pt_dfs_safe(adev, vm, start, cursor, entry) - list_move(&entry->vm_status, &vm->pt_freed); - - if (start) - list_move(&start->entry->vm_status, &vm->pt_freed); + list_splice_init(¶ms->tlb_flush_waitlist, &vm->pt_freed); spin_unlock(&vm->status_lock); schedule_work(&vm->pt_free_work); return; } - for_each_amdgpu_vm_pt_dfs_safe(adev, vm, start, cursor, entry) + list_for_each_entry_safe(entry, next, ¶ms->tlb_flush_waitlist, vm_status) amdgpu_vm_pt_free(entry); +} - if (start) - amdgpu_vm_pt_free(start->entry); +/** + * amdgpu_vm_pt_add_list - add PD/PT level to the flush list + * + * @params: parameters for the update + * @cursor: first PT entry to start DF search from, non NULL + * + * This list will be freed after TLB flush. + */ +static void amdgpu_vm_pt_add_list(struct amdgpu_vm_update_params *params, + struct amdgpu_vm_pt_cursor *cursor) +{ + struct amdgpu_vm_pt_cursor seek; + struct amdgpu_vm_bo_base *entry; + + spin_lock(¶ms->vm->status_lock); + for_each_amdgpu_vm_pt_dfs_safe(params->adev, params->vm, cursor, seek, entry) { + if (entry && entry->bo) + list_move(&entry->vm_status, ¶ms->tlb_flush_waitlist); + } + + /* enter start node now */ + list_move(&cursor->entry->vm_status, ¶ms->tlb_flush_waitlist); + spin_unlock(¶ms->vm->status_lock); } /** @@ -667,7 +685,13 @@ static void amdgpu_vm_pt_free_dfs(struct amdgpu_device *adev, */ void amdgpu_vm_pt_free_root(struct amdgpu_device *adev, struct amdgpu_vm *vm) { - amdgpu_vm_pt_free_dfs(adev, vm, NULL, false); + struct amdgpu_vm_pt_cursor cursor; + struct amdgpu_vm_bo_base *entry; + + for_each_amdgpu_vm_pt_dfs_safe(adev, vm, NULL, cursor, entry) { + if (entry) + amdgpu_vm_pt_free(entry); + } } /** @@ -682,11 +706,15 @@ int amdgpu_vm_pde_update(struct amdgpu_vm_update_params *params, struct amdgpu_vm_bo_base *entry) { struct amdgpu_vm_bo_base *parent = amdgpu_vm_pt_parent(entry); - struct amdgpu_bo *bo = parent->bo, *pbo; + struct amdgpu_bo *bo, *pbo; struct amdgpu_vm *vm = params->vm; uint64_t pde, pt, flags; unsigned int level; + if (WARN_ON(!parent)) + return -EINVAL; + + bo = parent->bo; for (level = 0, pbo = bo->parent; pbo; ++level) pbo = pbo->parent; @@ -972,10 +1000,8 @@ int amdgpu_vm_ptes_update(struct amdgpu_vm_update_params *params, while (cursor.pfn < frag_start) { /* Make sure previous mapping is freed */ if (cursor.entry->bo) { - params->table_freed = true; - amdgpu_vm_pt_free_dfs(adev, params->vm, - &cursor, - params->unlocked); + params->needs_flush = true; + amdgpu_vm_pt_add_list(params, &cursor); } amdgpu_vm_pt_next(adev, &cursor); } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_sdma.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_sdma.c index 349416e176a1..66e8a016126b 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_sdma.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_sdma.c @@ -126,6 +126,10 @@ static int amdgpu_vm_sdma_commit(struct amdgpu_vm_update_params *p, WARN_ON(ib->length_dw == 0); amdgpu_ring_pad_ib(ring, ib); + + if (p->needs_flush) + atomic64_inc(&p->vm->tlb_seq); + WARN_ON(ib->length_dw > p->num_dw_left); f = amdgpu_job_submit(p->job); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_tlb_fence.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_tlb_fence.c new file mode 100644 index 000000000000..51cddfa3f1e8 --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_tlb_fence.c @@ -0,0 +1,112 @@ +// SPDX-License-Identifier: GPL-2.0 OR MIT +/* + * Copyright 2023 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + */ + +#include <linux/dma-fence.h> +#include <linux/workqueue.h> + +#include "amdgpu.h" +#include "amdgpu_vm.h" +#include "amdgpu_gmc.h" + +struct amdgpu_tlb_fence { + struct dma_fence base; + struct amdgpu_device *adev; + struct dma_fence *dependency; + struct work_struct work; + spinlock_t lock; + uint16_t pasid; + +}; + +static const char *amdgpu_tlb_fence_get_driver_name(struct dma_fence *fence) +{ + return "amdgpu tlb fence"; +} + +static const char *amdgpu_tlb_fence_get_timeline_name(struct dma_fence *f) +{ + return "amdgpu tlb timeline"; +} + +static void amdgpu_tlb_fence_work(struct work_struct *work) +{ + struct amdgpu_tlb_fence *f = container_of(work, typeof(*f), work); + int r; + + if (f->dependency) { + dma_fence_wait(f->dependency, false); + dma_fence_put(f->dependency); + f->dependency = NULL; + } + + r = amdgpu_gmc_flush_gpu_tlb_pasid(f->adev, f->pasid, 2, true, 0); + if (r) { + dev_err(f->adev->dev, "TLB flush failed for PASID %d.\n", + f->pasid); + dma_fence_set_error(&f->base, r); + } + + dma_fence_signal(&f->base); + dma_fence_put(&f->base); +} + +static const struct dma_fence_ops amdgpu_tlb_fence_ops = { + .use_64bit_seqno = true, + .get_driver_name = amdgpu_tlb_fence_get_driver_name, + .get_timeline_name = amdgpu_tlb_fence_get_timeline_name +}; + +void amdgpu_vm_tlb_fence_create(struct amdgpu_device *adev, struct amdgpu_vm *vm, + struct dma_fence **fence) +{ + struct amdgpu_tlb_fence *f; + + f = kmalloc(sizeof(*f), GFP_KERNEL); + if (!f) { + /* + * We can't fail since the PDEs and PTEs are already updated, so + * just block for the dependency and execute the TLB flush + */ + if (*fence) + dma_fence_wait(*fence, false); + + amdgpu_gmc_flush_gpu_tlb_pasid(adev, vm->pasid, 2, true, 0); + *fence = dma_fence_get_stub(); + return; + } + + f->adev = adev; + f->dependency = *fence; + f->pasid = vm->pasid; + INIT_WORK(&f->work, amdgpu_tlb_fence_work); + spin_lock_init(&f->lock); + + dma_fence_init(&f->base, &amdgpu_tlb_fence_ops, &f->lock, + vm->tlb_fence_context, atomic64_read(&vm->tlb_seq)); + + /* TODO: We probably need a separate wq here */ + dma_fence_get(&f->base); + schedule_work(&f->work); + + *fence = &f->base; +} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vpe.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vpe.c index 7a65a2b128ec..c23d97d34b7e 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vpe.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vpe.c @@ -205,7 +205,7 @@ disable_dpm: dpm_ctl &= 0xfffffffe; /* Disable DPM */ WREG32(vpe_get_reg_offset(vpe, 0, vpe->regs.dpm_enable), dpm_ctl); dev_dbg(adev->dev, "%s: disable vpe dpm\n", __func__); - return 0; + return -EINVAL; } int amdgpu_vpe_psp_update_sram(struct amdgpu_device *adev) @@ -396,6 +396,12 @@ static int vpe_hw_init(void *handle) struct amdgpu_vpe *vpe = &adev->vpe; int ret; + /* Power on VPE */ + ret = amdgpu_device_ip_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_VPE, + AMD_PG_STATE_UNGATE); + if (ret) + return ret; + ret = vpe_load_microcode(vpe); if (ret) return ret; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c index 8db880244324..6c30eceec896 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c @@ -450,6 +450,7 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man, { struct amdgpu_vram_mgr *mgr = to_vram_mgr(man); struct amdgpu_device *adev = to_amdgpu_device(mgr); + struct amdgpu_bo *bo = ttm_to_amdgpu_bo(tbo); u64 vis_usage = 0, max_bytes, min_block_size; struct amdgpu_vram_mgr_resource *vres; u64 size, remaining_size, lpfn, fpfn; @@ -468,7 +469,7 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man, if (tbo->type != ttm_bo_type_kernel) max_bytes -= AMDGPU_VM_RESERVED_VRAM; - if (place->flags & TTM_PL_FLAG_CONTIGUOUS) { + if (bo->flags & AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS) { pages_per_block = ~0ul; } else { #ifdef CONFIG_TRANSPARENT_HUGEPAGE @@ -477,7 +478,7 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man, /* default to 2MB */ pages_per_block = 2UL << (20UL - PAGE_SHIFT); #endif - pages_per_block = max_t(uint32_t, pages_per_block, + pages_per_block = max_t(u32, pages_per_block, tbo->page_alignment); } @@ -498,9 +499,12 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man, if (place->flags & TTM_PL_FLAG_TOPDOWN) vres->flags |= DRM_BUDDY_TOPDOWN_ALLOCATION; - if (place->flags & TTM_PL_FLAG_CONTIGUOUS) + if (bo->flags & AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS) vres->flags |= DRM_BUDDY_CONTIGUOUS_ALLOCATION; + if (bo->flags & AMDGPU_GEM_CREATE_VRAM_CLEARED) + vres->flags |= DRM_BUDDY_CLEAR_ALLOCATION; + if (fpfn || lpfn != mgr->mm.size) /* Allocate blocks in desired range */ vres->flags |= DRM_BUDDY_RANGE_ALLOCATION; @@ -514,21 +518,31 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man, else min_block_size = mgr->default_page_size; - BUG_ON(min_block_size < mm->chunk_size); - /* Limit maximum size to 2GiB due to SG table limitations */ size = min(remaining_size, 2ULL << 30); if ((size >= (u64)pages_per_block << PAGE_SHIFT) && - !(size & (((u64)pages_per_block << PAGE_SHIFT) - 1))) + !(size & (((u64)pages_per_block << PAGE_SHIFT) - 1))) min_block_size = (u64)pages_per_block << PAGE_SHIFT; + BUG_ON(min_block_size < mm->chunk_size); + r = drm_buddy_alloc_blocks(mm, fpfn, lpfn, size, min_block_size, &vres->blocks, vres->flags); + + if (unlikely(r == -ENOSPC) && pages_per_block == ~0ul && + !(place->flags & TTM_PL_FLAG_CONTIGUOUS)) { + vres->flags &= ~DRM_BUDDY_CONTIGUOUS_ALLOCATION; + pages_per_block = max_t(u32, 2UL << (20UL - PAGE_SHIFT), + tbo->page_alignment); + + continue; + } + if (unlikely(r)) goto error_free_blocks; @@ -571,7 +585,7 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man, return 0; error_free_blocks: - drm_buddy_free_list(mm, &vres->blocks); + drm_buddy_free_list(mm, &vres->blocks, 0); mutex_unlock(&mgr->lock); error_fini: ttm_resource_fini(man, &vres->base); @@ -604,7 +618,7 @@ static void amdgpu_vram_mgr_del(struct ttm_resource_manager *man, amdgpu_vram_mgr_do_reserve(man); - drm_buddy_free_list(mm, &vres->blocks); + drm_buddy_free_list(mm, &vres->blocks, vres->flags); mutex_unlock(&mgr->lock); atomic64_sub(vis_usage, &mgr->vis_usage); @@ -912,7 +926,7 @@ void amdgpu_vram_mgr_fini(struct amdgpu_device *adev) kfree(rsv); list_for_each_entry_safe(rsv, temp, &mgr->reserved_pages, blocks) { - drm_buddy_free_list(&mgr->mm, &rsv->allocated); + drm_buddy_free_list(&mgr->mm, &rsv->allocated, 0); kfree(rsv); } if (!adev->gmc.is_app_apu) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.h index 0e04e42cf809..b256cbc2bc27 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.h @@ -53,10 +53,20 @@ static inline u64 amdgpu_vram_mgr_block_size(struct drm_buddy_block *block) return (u64)PAGE_SIZE << drm_buddy_block_order(block); } +static inline bool amdgpu_vram_mgr_is_cleared(struct drm_buddy_block *block) +{ + return drm_buddy_block_is_clear(block); +} + static inline struct amdgpu_vram_mgr_resource * to_amdgpu_vram_mgr_resource(struct ttm_resource *res) { return container_of(res, struct amdgpu_vram_mgr_resource, base); } +static inline void amdgpu_vram_mgr_set_cleared(struct ttm_resource *res) +{ + to_amdgpu_vram_mgr_resource(res)->flags |= DRM_BUDDY_CLEARED; +} + #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c index 20d51f6c9bb8..dd2ec48cf5c2 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c @@ -1035,15 +1035,16 @@ int amdgpu_xgmi_remove_device(struct amdgpu_device *adev) return 0; } -static int xgmi_v6_4_0_aca_bank_generate_report(struct aca_handle *handle, struct aca_bank *bank, enum aca_error_type type, - struct aca_bank_report *report, void *data) +static int xgmi_v6_4_0_aca_bank_parser(struct aca_handle *handle, struct aca_bank *bank, + enum aca_smu_type type, void *data) { struct amdgpu_device *adev = handle->adev; + struct aca_bank_info info; const char *error_str; - u64 status; + u64 status, count; int ret, ext_error_code; - ret = aca_bank_info_decode(bank, &report->info); + ret = aca_bank_info_decode(bank, &info); if (ret) return ret; @@ -1055,15 +1056,28 @@ static int xgmi_v6_4_0_aca_bank_generate_report(struct aca_handle *handle, struc if (error_str) dev_info(adev->dev, "%s detected\n", error_str); - if ((type == ACA_ERROR_TYPE_UE && ext_error_code == 0) || - (type == ACA_ERROR_TYPE_CE && ext_error_code == 6)) - report->count[type] = ACA_REG__MISC0__ERRCNT(bank->regs[ACA_REG_IDX_MISC0]); + count = ACA_REG__MISC0__ERRCNT(bank->regs[ACA_REG_IDX_MISC0]); - return 0; + switch (type) { + case ACA_SMU_TYPE_UE: + if (ext_error_code != 0 && ext_error_code != 9) + count = 0ULL; + + ret = aca_error_cache_log_bank_error(handle, &info, ACA_ERROR_TYPE_UE, count); + break; + case ACA_SMU_TYPE_CE: + count = ext_error_code == 6 ? count : 0ULL; + ret = aca_error_cache_log_bank_error(handle, &info, ACA_ERROR_TYPE_CE, count); + break; + default: + return -EINVAL; + } + + return ret; } static const struct aca_bank_ops xgmi_v6_4_0_aca_bank_ops = { - .aca_bank_generate_report = xgmi_v6_4_0_aca_bank_generate_report, + .aca_bank_parser = xgmi_v6_4_0_aca_bank_parser, }; static const struct aca_info xgmi_v6_4_0_aca_info = { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.h index 1592c63b3099..a3bfc16de6d4 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.h @@ -44,6 +44,7 @@ struct amdgpu_hive_info { struct amdgpu_reset_domain *reset_domain; atomic_t ras_recovery; + struct ras_event_manager event_mgr; }; struct amdgpu_pcs_ras_field { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h b/drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h index 51a14f6d93bd..fb2b394bb9c5 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h @@ -94,7 +94,8 @@ union amd_sriov_msg_feature_flags { uint32_t reg_indirect_acc : 1; uint32_t av1_support : 1; uint32_t vcn_rb_decouple : 1; - uint32_t reserved : 24; + uint32_t mes_info_enable : 1; + uint32_t reserved : 23; } flags; uint32_t all; }; @@ -157,7 +158,7 @@ struct amd_sriov_msg_pf2vf_info_header { uint32_t reserved[2]; }; -#define AMD_SRIOV_MSG_PF2VF_INFO_FILLED_SIZE (48) +#define AMD_SRIOV_MSG_PF2VF_INFO_FILLED_SIZE (49) struct amd_sriov_msg_pf2vf_info { /* header contains size and version */ struct amd_sriov_msg_pf2vf_info_header header; @@ -208,6 +209,8 @@ struct amd_sriov_msg_pf2vf_info { struct amd_sriov_msg_uuid_info uuid_info; /* PCIE atomic ops support flag */ uint32_t pcie_atomic_ops_support_flags; + /* Portion of GPU memory occupied by VF. MAX value is 65535, but set to uint32_t to maintain alignment with reserved size */ + uint32_t gpu_capacity; /* reserved */ uint32_t reserved[256 - AMD_SRIOV_MSG_PF2VF_INFO_FILLED_SIZE]; }; @@ -221,7 +224,7 @@ struct amd_sriov_msg_vf2pf_info_header { uint32_t reserved[2]; }; -#define AMD_SRIOV_MSG_VF2PF_INFO_FILLED_SIZE (70) +#define AMD_SRIOV_MSG_VF2PF_INFO_FILLED_SIZE (73) struct amd_sriov_msg_vf2pf_info { /* header contains size and version */ struct amd_sriov_msg_vf2pf_info_header header; @@ -265,7 +268,9 @@ struct amd_sriov_msg_vf2pf_info { uint32_t version; } ucode_info[AMD_SRIOV_MSG_RESERVE_UCODE]; uint64_t dummy_page_addr; - + /* FB allocated for guest MES to record UQ info */ + uint64_t mes_info_addr; + uint32_t mes_info_size; /* reserved */ uint32_t reserved[256 - AMD_SRIOV_MSG_VF2PF_INFO_FILLED_SIZE]; }; diff --git a/drivers/gpu/drm/amd/amdgpu/aqua_vanjaram.c b/drivers/gpu/drm/amd/amdgpu/aqua_vanjaram.c index d6f808acfb17..d4e2aed2efa3 100644 --- a/drivers/gpu/drm/amd/amdgpu/aqua_vanjaram.c +++ b/drivers/gpu/drm/amd/amdgpu/aqua_vanjaram.c @@ -62,6 +62,11 @@ void aqua_vanjaram_doorbell_index_init(struct amdgpu_device *adev) adev->doorbell_index.max_assignment = AMDGPU_DOORBELL_LAYOUT1_MAX_ASSIGNMENT << 1; } +static bool aqua_vanjaram_xcp_vcn_shared(struct amdgpu_device *adev) +{ + return (adev->xcp_mgr->num_xcps > adev->vcn.num_vcn_inst); +} + static void aqua_vanjaram_set_xcp_id(struct amdgpu_device *adev, uint32_t inst_idx, struct amdgpu_ring *ring) { @@ -87,7 +92,7 @@ static void aqua_vanjaram_set_xcp_id(struct amdgpu_device *adev, case AMDGPU_RING_TYPE_VCN_ENC: case AMDGPU_RING_TYPE_VCN_JPEG: ip_blk = AMDGPU_XCP_VCN; - if (adev->xcp_mgr->mode == AMDGPU_CPX_PARTITION_MODE) + if (aqua_vanjaram_xcp_vcn_shared(adev)) inst_mask = 1 << (inst_idx * 2); break; default: @@ -140,10 +145,12 @@ static int aqua_vanjaram_xcp_sched_list_update( aqua_vanjaram_xcp_gpu_sched_update(adev, ring, ring->xcp_id); - /* VCN is shared by two partitions under CPX MODE */ + /* VCN may be shared by two partitions under CPX MODE in certain + * configs. + */ if ((ring->funcs->type == AMDGPU_RING_TYPE_VCN_ENC || - ring->funcs->type == AMDGPU_RING_TYPE_VCN_JPEG) && - adev->xcp_mgr->mode == AMDGPU_CPX_PARTITION_MODE) + ring->funcs->type == AMDGPU_RING_TYPE_VCN_JPEG) && + aqua_vanjaram_xcp_vcn_shared(adev)) aqua_vanjaram_xcp_gpu_sched_update(adev, ring, ring->xcp_id + 1); } @@ -415,7 +422,7 @@ __aqua_vanjaram_get_auto_mode(struct amdgpu_xcp_mgr *xcp_mgr) if (adev->gmc.num_mem_partitions == num_xcc / 2) return (adev->flags & AMD_IS_APU) ? AMDGPU_TPX_PARTITION_MODE : - AMDGPU_QPX_PARTITION_MODE; + AMDGPU_CPX_PARTITION_MODE; if (adev->gmc.num_mem_partitions == 2 && !(adev->flags & AMD_IS_APU)) return AMDGPU_DPX_PARTITION_MODE; @@ -623,7 +630,7 @@ static int aqua_vanjaram_xcp_mgr_init(struct amdgpu_device *adev) int aqua_vanjaram_init_soc_config(struct amdgpu_device *adev) { - u32 mask, inst_mask = adev->sdma.sdma_mask; + u32 mask, avail_inst, inst_mask = adev->sdma.sdma_mask; int ret, i; /* generally 1 AID supports 4 instances */ @@ -635,7 +642,9 @@ int aqua_vanjaram_init_soc_config(struct amdgpu_device *adev) for (mask = (1 << adev->sdma.num_inst_per_aid) - 1; inst_mask; inst_mask >>= adev->sdma.num_inst_per_aid, ++i) { - if ((inst_mask & mask) == mask) + avail_inst = inst_mask & mask; + if (avail_inst == mask || avail_inst == 0x3 || + avail_inst == 0xc) adev->aid_mask |= (1 << i); } diff --git a/drivers/gpu/drm/amd/amdgpu/atom.c b/drivers/gpu/drm/amd/amdgpu/atom.c index 72362df352f6..d552e013354c 100644 --- a/drivers/gpu/drm/amd/amdgpu/atom.c +++ b/drivers/gpu/drm/amd/amdgpu/atom.c @@ -1243,6 +1243,7 @@ static int amdgpu_atom_execute_table_locked(struct atom_context *ctx, int index, ectx.ps_size = params_size; ectx.abort = false; ectx.last_jump = 0; + ectx.last_jump_jiffies = 0; if (ws) { ectx.ws = kcalloc(4, ws, GFP_KERNEL); ectx.ws_size = ws; diff --git a/drivers/gpu/drm/amd/amdgpu/cik.c b/drivers/gpu/drm/amd/amdgpu/cik.c index a3a643254d7a..cf1d5d462b67 100644 --- a/drivers/gpu/drm/amd/amdgpu/cik.c +++ b/drivers/gpu/drm/amd/amdgpu/cik.c @@ -1375,14 +1375,14 @@ static int cik_asic_pci_config_reset(struct amdgpu_device *adev) return r; } -static bool cik_asic_supports_baco(struct amdgpu_device *adev) +static int cik_asic_supports_baco(struct amdgpu_device *adev) { switch (adev->asic_type) { case CHIP_BONAIRE: case CHIP_HAWAII: return amdgpu_dpm_is_baco_supported(adev); default: - return false; + return 0; } } @@ -2210,6 +2210,8 @@ static const struct amd_ip_funcs cik_common_ip_funcs = { .soft_reset = cik_common_soft_reset, .set_clockgating_state = cik_common_set_clockgating_state, .set_powergating_state = cik_common_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_ip_block_version cik_common_ip_block = diff --git a/drivers/gpu/drm/amd/amdgpu/cik_ih.c b/drivers/gpu/drm/amd/amdgpu/cik_ih.c index f24e34dc33d1..576baa9dbb0e 100644 --- a/drivers/gpu/drm/amd/amdgpu/cik_ih.c +++ b/drivers/gpu/drm/amd/amdgpu/cik_ih.c @@ -435,6 +435,8 @@ static const struct amd_ip_funcs cik_ih_ip_funcs = { .soft_reset = cik_ih_soft_reset, .set_clockgating_state = cik_ih_set_clockgating_state, .set_powergating_state = cik_ih_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_ih_funcs cik_ih_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/cik_sdma.c b/drivers/gpu/drm/amd/amdgpu/cik_sdma.c index a3fccc4c1f43..6948ebda0fa2 100644 --- a/drivers/gpu/drm/amd/amdgpu/cik_sdma.c +++ b/drivers/gpu/drm/amd/amdgpu/cik_sdma.c @@ -1228,6 +1228,8 @@ static const struct amd_ip_funcs cik_sdma_ip_funcs = { .soft_reset = cik_sdma_soft_reset, .set_clockgating_state = cik_sdma_set_clockgating_state, .set_powergating_state = cik_sdma_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs cik_sdma_ring_funcs = { @@ -1290,7 +1292,7 @@ static void cik_sdma_set_irq_funcs(struct amdgpu_device *adev) * @src_offset: src GPU address * @dst_offset: dst GPU address * @byte_count: number of bytes to xfer - * @tmz: is this a secure operation + * @copy_flags: unused * * Copy GPU buffers using the DMA engine (CIK). * Used by the amdgpu ttm implementation to move pages if @@ -1300,7 +1302,7 @@ static void cik_sdma_emit_copy_buffer(struct amdgpu_ib *ib, uint64_t src_offset, uint64_t dst_offset, uint32_t byte_count, - bool tmz) + uint32_t copy_flags) { ib->ptr[ib->length_dw++] = SDMA_PACKET(SDMA_OPCODE_COPY, SDMA_COPY_SUB_OPCODE_LINEAR, 0); ib->ptr[ib->length_dw++] = byte_count; diff --git a/drivers/gpu/drm/amd/amdgpu/cz_ih.c b/drivers/gpu/drm/amd/amdgpu/cz_ih.c index c19681492efa..072643787384 100644 --- a/drivers/gpu/drm/amd/amdgpu/cz_ih.c +++ b/drivers/gpu/drm/amd/amdgpu/cz_ih.c @@ -433,6 +433,8 @@ static const struct amd_ip_funcs cz_ih_ip_funcs = { .soft_reset = cz_ih_soft_reset, .set_clockgating_state = cz_ih_set_clockgating_state, .set_powergating_state = cz_ih_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_ih_funcs cz_ih_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c index 221af054d874..b44fce44c066 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c @@ -3333,6 +3333,8 @@ static const struct amd_ip_funcs dce_v10_0_ip_funcs = { .soft_reset = dce_v10_0_soft_reset, .set_clockgating_state = dce_v10_0_set_clockgating_state, .set_powergating_state = dce_v10_0_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static void diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c index 69e8b0db6cf7..80b2e7f79acf 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c @@ -3464,6 +3464,8 @@ static const struct amd_ip_funcs dce_v11_0_ip_funcs = { .soft_reset = dce_v11_0_soft_reset, .set_clockgating_state = dce_v11_0_set_clockgating_state, .set_powergating_state = dce_v11_0_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static void diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c index 60d40201fdd1..db20012600f5 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c @@ -3154,6 +3154,8 @@ static const struct amd_ip_funcs dce_v6_0_ip_funcs = { .soft_reset = dce_v6_0_soft_reset, .set_clockgating_state = dce_v6_0_set_clockgating_state, .set_powergating_state = dce_v6_0_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static void diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c index 5a5fcc45e452..5b56100ec902 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c @@ -3242,6 +3242,8 @@ static const struct amd_ip_funcs dce_v8_0_ip_funcs = { .soft_reset = dce_v8_0_soft_reset, .set_clockgating_state = dce_v8_0_set_clockgating_state, .set_powergating_state = dce_v8_0_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static void diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c index f90905ef32c7..536287ddd2ec 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c @@ -276,6 +276,99 @@ MODULE_FIRMWARE("amdgpu/gc_10_3_7_mec.bin"); MODULE_FIRMWARE("amdgpu/gc_10_3_7_mec2.bin"); MODULE_FIRMWARE("amdgpu/gc_10_3_7_rlc.bin"); +static const struct amdgpu_hwip_reg_entry gc_reg_list_10_1[] = { + SOC15_REG_ENTRY_STR(GC, 0, mmGRBM_STATUS), + SOC15_REG_ENTRY_STR(GC, 0, mmGRBM_STATUS2), + SOC15_REG_ENTRY_STR(GC, 0, mmGRBM_STATUS3), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_STALLED_STAT1), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_STALLED_STAT2), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_CPC_STALLED_STAT1), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_CPF_STALLED_STAT1), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_BUSY_STAT), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_CPC_BUSY_STAT), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_CPF_BUSY_STAT), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_CPC_BUSY_STAT2), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_CPF_BUSY_STAT2), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_CPF_STATUS), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_GFX_ERROR), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_GFX_HPD_STATUS0), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_RB_BASE), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_RB_RPTR), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_RB_WPTR), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_RB0_BASE), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_RB0_RPTR), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_RB0_WPTR), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_RB1_BASE), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_RB1_RPTR), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_RB1_WPTR), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_RB2_BASE), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_RB2_WPTR), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_RB2_WPTR), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_CE_IB1_CMD_BUFSZ), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_CE_IB2_CMD_BUFSZ), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_IB1_CMD_BUFSZ), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_IB2_CMD_BUFSZ), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_CE_IB1_BASE_LO), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_CE_IB1_BASE_HI), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_CE_IB1_BUFSZ), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_CE_IB2_BASE_LO), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_CE_IB2_BASE_HI), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_CE_IB2_BUFSZ), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_IB1_BASE_LO), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_IB1_BASE_HI), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_IB1_BUFSZ), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_IB2_BASE_LO), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_IB2_BASE_HI), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_IB2_BUFSZ), + SOC15_REG_ENTRY_STR(GC, 0, mmCPF_UTCL1_STATUS), + SOC15_REG_ENTRY_STR(GC, 0, mmCPC_UTCL1_STATUS), + SOC15_REG_ENTRY_STR(GC, 0, mmCPG_UTCL1_STATUS), + SOC15_REG_ENTRY_STR(GC, 0, mmGDS_PROTECTION_FAULT), + SOC15_REG_ENTRY_STR(GC, 0, mmGDS_VM_PROTECTION_FAULT), + SOC15_REG_ENTRY_STR(GC, 0, mmIA_UTCL1_STATUS), + SOC15_REG_ENTRY_STR(GC, 0, mmIA_UTCL1_STATUS_2), + SOC15_REG_ENTRY_STR(GC, 0, mmPA_CL_CNTL_STATUS), + SOC15_REG_ENTRY_STR(GC, 0, mmRLC_UTCL1_STATUS), + SOC15_REG_ENTRY_STR(GC, 0, mmRMI_UTCL1_STATUS), + SOC15_REG_ENTRY_STR(GC, 0, mmSQC_DCACHE_UTCL0_STATUS), + SOC15_REG_ENTRY_STR(GC, 0, mmSQC_ICACHE_UTCL0_STATUS), + SOC15_REG_ENTRY_STR(GC, 0, mmSQG_UTCL0_STATUS), + SOC15_REG_ENTRY_STR(GC, 0, mmTCP_UTCL0_STATUS), + SOC15_REG_ENTRY_STR(GC, 0, mmWD_UTCL1_STATUS), + SOC15_REG_ENTRY_STR(GC, 0, mmGCVM_L2_PROTECTION_FAULT_CNTL), + SOC15_REG_ENTRY_STR(GC, 0, mmGCVM_L2_PROTECTION_FAULT_STATUS), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_DEBUG), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_MEC_CNTL), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_MES_CNTL), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_CE_INSTR_PNTR), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_MEC1_INSTR_PNTR), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_MEC2_INSTR_PNTR), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_MES_DEBUG_INTERRUPT_INSTR_PNTR), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_MES_INSTR_PNTR), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_ME_INSTR_PNTR), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_PFP_INSTR_PNTR), + SOC15_REG_ENTRY_STR(GC, 0, mmCP_CPC_STATUS), + SOC15_REG_ENTRY_STR(GC, 0, mmRLC_STAT), + SOC15_REG_ENTRY_STR(GC, 0, mmRLC_SMU_COMMAND), + SOC15_REG_ENTRY_STR(GC, 0, mmRLC_SMU_MESSAGE), + SOC15_REG_ENTRY_STR(GC, 0, mmRLC_SMU_ARGUMENT_1), + SOC15_REG_ENTRY_STR(GC, 0, mmRLC_SMU_ARGUMENT_2), + SOC15_REG_ENTRY_STR(GC, 0, mmRLC_SMU_ARGUMENT_3), + SOC15_REG_ENTRY_STR(GC, 0, mmRLC_SMU_ARGUMENT_4), + SOC15_REG_ENTRY_STR(GC, 0, mmSMU_RLC_RESPONSE), + SOC15_REG_ENTRY_STR(GC, 0, mmRLC_SAFE_MODE), + SOC15_REG_ENTRY_STR(GC, 0, mmRLC_SMU_SAFE_MODE), + SOC15_REG_ENTRY_STR(GC, 0, mmRLC_RLCS_GPM_STAT_2), + SOC15_REG_ENTRY_STR(GC, 0, mmRLC_SPP_STATUS), + SOC15_REG_ENTRY_STR(GC, 0, mmRLC_RLCS_BOOTLOAD_STATUS), + SOC15_REG_ENTRY_STR(GC, 0, mmRLC_INT_STAT), + SOC15_REG_ENTRY_STR(GC, 0, mmRLC_GPM_GENERAL_6), + SOC15_REG_ENTRY_STR(GC, 0, mmRLC_GPM_DEBUG_INST_A), + SOC15_REG_ENTRY_STR(GC, 0, mmRLC_GPM_DEBUG_INST_B), + SOC15_REG_ENTRY_STR(GC, 0, mmRLC_GPM_DEBUG_INST_ADDR), + SOC15_REG_ENTRY_STR(GC, 0, mmRLC_LX6_CORE_PDEBUG_INST) +}; + static const struct soc15_reg_golden golden_settings_gc_10_1[] = { SOC15_REG_GOLDEN_VALUE(GC, 0, mmCB_HW_CONTROL_4, 0xffffffff, 0x00400014), SOC15_REG_GOLDEN_VALUE(GC, 0, mmCGTT_CPF_CLK_CTRL, 0xfcff8fff, 0xf8000100), @@ -3964,7 +4057,7 @@ static void gfx_v10_0_check_gfxoff_flag(struct amdgpu_device *adev) static int gfx_v10_0_init_microcode(struct amdgpu_device *adev) { - char fw_name[40]; + char fw_name[53]; char ucode_prefix[30]; const char *wks = ""; int err; @@ -4490,6 +4583,22 @@ static int gfx_v10_0_compute_ring_init(struct amdgpu_device *adev, int ring_id, hw_prio, NULL); } +static void gfx_v10_0_alloc_dump_mem(struct amdgpu_device *adev) +{ + uint32_t reg_count = ARRAY_SIZE(gc_reg_list_10_1); + uint32_t *ptr; + + ptr = kcalloc(reg_count, sizeof(uint32_t), GFP_KERNEL); + if (ptr == NULL) { + DRM_ERROR("Failed to allocate memory for IP Dump\n"); + adev->gfx.ip_dump = NULL; + adev->gfx.reg_count = 0; + } else { + adev->gfx.ip_dump = ptr; + adev->gfx.reg_count = reg_count; + } +} + static int gfx_v10_0_sw_init(void *handle) { int i, j, k, r, ring_id = 0; @@ -4518,7 +4627,7 @@ static int gfx_v10_0_sw_init(void *handle) case IP_VERSION(10, 3, 3): case IP_VERSION(10, 3, 7): adev->gfx.me.num_me = 1; - adev->gfx.me.num_pipe_per_me = 1; + adev->gfx.me.num_pipe_per_me = 2; adev->gfx.me.num_queue_per_pipe = 1; adev->gfx.mec.num_mec = 2; adev->gfx.mec.num_pipe_per_mec = 4; @@ -4642,6 +4751,8 @@ static int gfx_v10_0_sw_init(void *handle) gfx_v10_0_gpu_early_init(adev); + gfx_v10_0_alloc_dump_mem(adev); + return 0; } @@ -4694,6 +4805,8 @@ static int gfx_v10_0_sw_fini(void *handle) gfx_v10_0_free_microcode(adev); + kfree(adev->gfx.ip_dump); + return 0; } @@ -8317,7 +8430,7 @@ static void gfx_v10_0_ring_emit_hdp_flush(struct amdgpu_ring *ring) } reg_mem_engine = 0; } else { - ref_and_mask = nbio_hf_reg->ref_and_mask_cp0; + ref_and_mask = nbio_hf_reg->ref_and_mask_cp0 << ring->pipe; reg_mem_engine = 1; /* pfp */ } @@ -9154,6 +9267,36 @@ static void gfx_v10_0_emit_mem_sync(struct amdgpu_ring *ring) amdgpu_ring_write(ring, gcr_cntl); /* GCR_CNTL */ } +static void gfx_v10_ip_print(void *handle, struct drm_printer *p) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + uint32_t i; + uint32_t reg_count = ARRAY_SIZE(gc_reg_list_10_1); + + if (!adev->gfx.ip_dump) + return; + + for (i = 0; i < reg_count; i++) + drm_printf(p, "%-50s \t 0x%08x\n", + gc_reg_list_10_1[i].reg_name, + adev->gfx.ip_dump[i]); +} + +static void gfx_v10_ip_dump(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + uint32_t i; + uint32_t reg_count = ARRAY_SIZE(gc_reg_list_10_1); + + if (!adev->gfx.ip_dump) + return; + + amdgpu_gfx_off_ctrl(adev, false); + for (i = 0; i < reg_count; i++) + adev->gfx.ip_dump[i] = RREG32(SOC15_REG_ENTRY_OFFSET(gc_reg_list_10_1[i])); + amdgpu_gfx_off_ctrl(adev, true); +} + static const struct amd_ip_funcs gfx_v10_0_ip_funcs = { .name = "gfx_v10_0", .early_init = gfx_v10_0_early_init, @@ -9170,6 +9313,8 @@ static const struct amd_ip_funcs gfx_v10_0_ip_funcs = { .set_clockgating_state = gfx_v10_0_set_clockgating_state, .set_powergating_state = gfx_v10_0_set_powergating_state, .get_clockgating_state = gfx_v10_0_get_clockgating_state, + .dump_ip_state = gfx_v10_ip_dump, + .print_ip_state = gfx_v10_ip_print, }; static const struct amdgpu_ring_funcs gfx_v10_0_ring_funcs_gfx = { @@ -9186,7 +9331,7 @@ static const struct amdgpu_ring_funcs gfx_v10_0_ring_funcs_gfx = { 7 + /* PIPELINE_SYNC */ SOC15_FLUSH_GPU_TLB_NUM_WREG * 5 + SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 7 + - 2 + /* VM_FLUSH */ + 4 + /* VM_FLUSH */ 8 + /* FENCE for VM_FLUSH */ 20 + /* GDS switch */ 4 + /* double SWITCH_BUFFER, @@ -9276,7 +9421,6 @@ static const struct amdgpu_ring_funcs gfx_v10_0_ring_funcs_kiq = { 7 + /* gfx_v10_0_ring_emit_pipeline_sync */ SOC15_FLUSH_GPU_TLB_NUM_WREG * 5 + SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 7 + - 2 + /* gfx_v10_0_ring_emit_vm_flush */ 8 + 8 + 8, /* gfx_v10_0_ring_emit_fence_kiq x3 for user fence, vm fence */ .emit_ib_size = 7, /* gfx_v10_0_ring_emit_ib_compute */ .emit_ib = gfx_v10_0_ring_emit_ib_compute, diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c index 1770e496c1b7..ad6431013c73 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c @@ -510,7 +510,7 @@ static void gfx_v11_0_check_fw_cp_gfx_shadow(struct amdgpu_device *adev) static int gfx_v11_0_init_microcode(struct amdgpu_device *adev) { char fw_name[40]; - char ucode_prefix[30]; + char ucode_prefix[25]; int err; const struct rlc_firmware_header_v2_0 *rlc_hdr; uint16_t version_major; @@ -1635,7 +1635,7 @@ static void gfx_v11_0_setup_rb(struct amdgpu_device *adev) active_rb_bitmap |= (0x3 << (i * rb_bitmap_width_per_sa)); } - active_rb_bitmap |= global_active_rb_bitmap; + active_rb_bitmap &= global_active_rb_bitmap; adev->gfx.config.backend_enable_mask = active_rb_bitmap; adev->gfx.config.num_rbs = hweight32(active_rb_bitmap); } @@ -4506,14 +4506,11 @@ static int gfx_v11_0_soft_reset(void *handle) gfx_v11_0_set_safe_mode(adev, 0); + mutex_lock(&adev->srbm_mutex); for (i = 0; i < adev->gfx.mec.num_mec; ++i) { for (j = 0; j < adev->gfx.mec.num_queue_per_pipe; j++) { for (k = 0; k < adev->gfx.mec.num_pipe_per_mec; k++) { - tmp = RREG32_SOC15(GC, 0, regGRBM_GFX_CNTL); - tmp = REG_SET_FIELD(tmp, GRBM_GFX_CNTL, MEID, i); - tmp = REG_SET_FIELD(tmp, GRBM_GFX_CNTL, QUEUEID, j); - tmp = REG_SET_FIELD(tmp, GRBM_GFX_CNTL, PIPEID, k); - WREG32_SOC15(GC, 0, regGRBM_GFX_CNTL, tmp); + soc21_grbm_select(adev, i, k, j, 0); WREG32_SOC15(GC, 0, regCP_HQD_DEQUEUE_REQUEST, 0x2); WREG32_SOC15(GC, 0, regSPI_COMPUTE_QUEUE_RESET, 0x1); @@ -4523,16 +4520,14 @@ static int gfx_v11_0_soft_reset(void *handle) for (i = 0; i < adev->gfx.me.num_me; ++i) { for (j = 0; j < adev->gfx.me.num_queue_per_pipe; j++) { for (k = 0; k < adev->gfx.me.num_pipe_per_me; k++) { - tmp = RREG32_SOC15(GC, 0, regGRBM_GFX_CNTL); - tmp = REG_SET_FIELD(tmp, GRBM_GFX_CNTL, MEID, i); - tmp = REG_SET_FIELD(tmp, GRBM_GFX_CNTL, QUEUEID, j); - tmp = REG_SET_FIELD(tmp, GRBM_GFX_CNTL, PIPEID, k); - WREG32_SOC15(GC, 0, regGRBM_GFX_CNTL, tmp); + soc21_grbm_select(adev, i, k, j, 0); WREG32_SOC15(GC, 0, regCP_GFX_HQD_DEQUEUE_REQUEST, 0x1); } } } + soc21_grbm_select(adev, 0, 0, 0, 0); + mutex_unlock(&adev->srbm_mutex); /* Try to acquire the gfx mutex before access to CP_VMID_RESET */ r = gfx_v11_0_request_gfx_index_mutex(adev, 1); @@ -5465,6 +5460,7 @@ static void gfx_v11_0_ring_emit_vm_flush(struct amdgpu_ring *ring, /* Make sure that we can't skip the SET_Q_MODE packets when the VM * changed in any way. */ + ring->set_q_mode_offs = 0; ring->set_q_mode_ptr = NULL; } @@ -6173,6 +6169,8 @@ static const struct amd_ip_funcs gfx_v11_0_ip_funcs = { .set_clockgating_state = gfx_v11_0_set_clockgating_state, .set_powergating_state = gfx_v11_0_set_powergating_state, .get_clockgating_state = gfx_v11_0_get_clockgating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs gfx_v11_0_ring_funcs_gfx = { @@ -6191,7 +6189,7 @@ static const struct amdgpu_ring_funcs gfx_v11_0_ring_funcs_gfx = { 7 + /* PIPELINE_SYNC */ SOC15_FLUSH_GPU_TLB_NUM_WREG * 5 + SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 7 + - 2 + /* VM_FLUSH */ + 4 + /* VM_FLUSH */ 8 + /* FENCE for VM_FLUSH */ 20 + /* GDS switch */ 5 + /* COND_EXEC */ @@ -6277,7 +6275,6 @@ static const struct amdgpu_ring_funcs gfx_v11_0_ring_funcs_kiq = { 7 + /* gfx_v11_0_ring_emit_pipeline_sync */ SOC15_FLUSH_GPU_TLB_NUM_WREG * 5 + SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 7 + - 2 + /* gfx_v11_0_ring_emit_vm_flush */ 8 + 8 + 8, /* gfx_v11_0_ring_emit_fence_kiq x3 for user fence, vm fence */ .emit_ib_size = 7, /* gfx_v11_0_ring_emit_ib_compute */ .emit_ib = gfx_v11_0_ring_emit_ib_compute, diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c index 34f9211b2679..d0992ce9fb47 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c @@ -3457,6 +3457,8 @@ static const struct amd_ip_funcs gfx_v6_0_ip_funcs = { .soft_reset = gfx_v6_0_soft_reset, .set_clockgating_state = gfx_v6_0_set_clockgating_state, .set_powergating_state = gfx_v6_0_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs gfx_v6_0_ring_funcs_gfx = { diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c index 86a4865b1ae5..541dbd70d8c7 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c @@ -4977,6 +4977,8 @@ static const struct amd_ip_funcs gfx_v7_0_ip_funcs = { .soft_reset = gfx_v7_0_soft_reset, .set_clockgating_state = gfx_v7_0_set_clockgating_state, .set_powergating_state = gfx_v7_0_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs gfx_v7_0_ring_funcs_gfx = { diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c index 202ddda57f98..2f0e72caee1a 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c @@ -6878,6 +6878,8 @@ static const struct amd_ip_funcs gfx_v8_0_ip_funcs = { .set_clockgating_state = gfx_v8_0_set_clockgating_state, .set_powergating_state = gfx_v8_0_set_powergating_state, .get_clockgating_state = gfx_v8_0_get_clockgating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs gfx_v8_0_ring_funcs_gfx = { diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c index 6f97a6d0e6d0..3c8c5abf35ab 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c @@ -1249,7 +1249,7 @@ static void gfx_v9_0_check_if_need_gfxoff(struct amdgpu_device *adev) static int gfx_v9_0_init_cp_gfx_microcode(struct amdgpu_device *adev, char *chip_name) { - char fw_name[30]; + char fw_name[50]; int err; snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_pfp.bin", chip_name); @@ -1282,7 +1282,7 @@ out: static int gfx_v9_0_init_rlc_microcode(struct amdgpu_device *adev, char *chip_name) { - char fw_name[30]; + char fw_name[53]; int err; const struct rlc_firmware_header_v2_0 *rlc_hdr; uint16_t version_major; @@ -1337,7 +1337,7 @@ static bool gfx_v9_0_load_mec2_fw_bin_support(struct amdgpu_device *adev) static int gfx_v9_0_init_cp_compute_microcode(struct amdgpu_device *adev, char *chip_name) { - char fw_name[30]; + char fw_name[50]; int err; if (amdgpu_sriov_vf(adev) && (adev->asic_type == CHIP_ALDEBARAN)) @@ -6856,6 +6856,8 @@ static const struct amd_ip_funcs gfx_v9_0_ip_funcs = { .set_clockgating_state = gfx_v9_0_set_clockgating_state, .set_powergating_state = gfx_v9_0_set_powergating_state, .get_clockgating_state = gfx_v9_0_get_clockgating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs gfx_v9_0_ring_funcs_gfx = { @@ -6981,7 +6983,6 @@ static const struct amdgpu_ring_funcs gfx_v9_0_ring_funcs_compute = { 7 + /* gfx_v9_0_ring_emit_pipeline_sync */ SOC15_FLUSH_GPU_TLB_NUM_WREG * 5 + SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 7 + - 2 + /* gfx_v9_0_ring_emit_vm_flush */ 8 + 8 + 8 + /* gfx_v9_0_ring_emit_fence x3 for user fence, vm fence */ 7 + /* gfx_v9_0_emit_mem_sync */ 5 + /* gfx_v9_0_emit_wave_limit for updating mmSPI_WCL_PIPE_PERCENT_GFX register */ @@ -7019,7 +7020,6 @@ static const struct amdgpu_ring_funcs gfx_v9_0_ring_funcs_kiq = { 7 + /* gfx_v9_0_ring_emit_pipeline_sync */ SOC15_FLUSH_GPU_TLB_NUM_WREG * 5 + SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 7 + - 2 + /* gfx_v9_0_ring_emit_vm_flush */ 8 + 8 + 8, /* gfx_v9_0_ring_emit_fence_kiq x3 for user fence, vm fence */ .emit_ib_size = 7, /* gfx_v9_0_ring_emit_ib_compute */ .emit_fence = gfx_v9_0_ring_emit_fence_kiq, diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_2.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_2.c index 065b2bd5f5a6..3f4fd2f08163 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_2.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_2.c @@ -1909,18 +1909,7 @@ static void gfx_v9_4_2_reset_sq_timeout_status(struct amdgpu_device *adev) mutex_unlock(&adev->grbm_idx_mutex); } -static bool gfx_v9_4_2_query_uctl2_poison_status(struct amdgpu_device *adev) -{ - u32 status = 0; - struct amdgpu_vmhub *hub; - - hub = &adev->vmhub[AMDGPU_GFXHUB(0)]; - status = RREG32(hub->vm_l2_pro_fault_status); - /* reset page fault status */ - WREG32_P(hub->vm_l2_pro_fault_cntl, 1, ~1); - return REG_GET_FIELD(status, VM_L2_PROTECTION_FAULT_STATUS, FED); -} struct amdgpu_ras_block_hw_ops gfx_v9_4_2_ras_ops = { .query_ras_error_count = &gfx_v9_4_2_query_ras_error_count, @@ -1934,5 +1923,4 @@ struct amdgpu_gfx_ras gfx_v9_4_2_ras = { .hw_ops = &gfx_v9_4_2_ras_ops, }, .enable_watchdog_timer = &gfx_v9_4_2_enable_watchdog_timer, - .query_utcl2_poison_status = gfx_v9_4_2_query_uctl2_poison_status, }; diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_3.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_3.c index b53c8fd4e8cf..7b16e8cca86a 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_3.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_3.c @@ -431,16 +431,16 @@ out: static int gfx_v9_4_3_init_microcode(struct amdgpu_device *adev) { - const char *chip_name; + char ucode_prefix[15]; int r; - chip_name = "gc_9_4_3"; + amdgpu_ucode_ip_version_decode(adev, GC_HWIP, ucode_prefix, sizeof(ucode_prefix)); - r = gfx_v9_4_3_init_rlc_microcode(adev, chip_name); + r = gfx_v9_4_3_init_rlc_microcode(adev, ucode_prefix); if (r) return r; - r = gfx_v9_4_3_init_cp_compute_microcode(adev, chip_name); + r = gfx_v9_4_3_init_cp_compute_microcode(adev, ucode_prefix); if (r) return r; @@ -680,38 +680,44 @@ static const struct amdgpu_gfx_funcs gfx_v9_4_3_gfx_funcs = { .ih_node_to_logical_xcc = &gfx_v9_4_3_ih_to_xcc_inst, }; -static int gfx_v9_4_3_aca_bank_generate_report(struct aca_handle *handle, - struct aca_bank *bank, enum aca_error_type type, - struct aca_bank_report *report, void *data) +static int gfx_v9_4_3_aca_bank_parser(struct aca_handle *handle, + struct aca_bank *bank, enum aca_smu_type type, + void *data) { - u64 status, misc0; + struct aca_bank_info info; + u64 misc0; u32 instlo; int ret; - status = bank->regs[ACA_REG_IDX_STATUS]; - if ((type == ACA_ERROR_TYPE_UE && - ACA_REG__STATUS__ERRORCODEEXT(status) == ACA_EXTERROR_CODE_FAULT) || - (type == ACA_ERROR_TYPE_CE && - ACA_REG__STATUS__ERRORCODEEXT(status) == ACA_EXTERROR_CODE_CE)) { + ret = aca_bank_info_decode(bank, &info); + if (ret) + return ret; - ret = aca_bank_info_decode(bank, &report->info); - if (ret) - return ret; + /* NOTE: overwrite info.die_id with xcd id for gfx */ + instlo = ACA_REG__IPID__INSTANCEIDLO(bank->regs[ACA_REG_IDX_IPID]); + instlo &= GENMASK(31, 1); + info.die_id = instlo == mmSMNAID_XCD0_MCA_SMU ? 0 : 1; - /* NOTE: overwrite info.die_id with xcd id for gfx */ - instlo = ACA_REG__IPID__INSTANCEIDLO(bank->regs[ACA_REG_IDX_IPID]); - instlo &= GENMASK(31, 1); - report->info.die_id = instlo == mmSMNAID_XCD0_MCA_SMU ? 0 : 1; + misc0 = bank->regs[ACA_REG_IDX_MISC0]; - misc0 = bank->regs[ACA_REG_IDX_MISC0]; - report->count[type] = ACA_REG__MISC0__ERRCNT(misc0); + switch (type) { + case ACA_SMU_TYPE_UE: + ret = aca_error_cache_log_bank_error(handle, &info, + ACA_ERROR_TYPE_UE, 1ULL); + break; + case ACA_SMU_TYPE_CE: + ret = aca_error_cache_log_bank_error(handle, &info, + ACA_ERROR_TYPE_CE, ACA_REG__MISC0__ERRCNT(misc0)); + break; + default: + return -EINVAL; } - return 0; + return ret; } static bool gfx_v9_4_3_aca_bank_is_valid(struct aca_handle *handle, struct aca_bank *bank, - enum aca_error_type type, void *data) + enum aca_smu_type type, void *data) { u32 instlo; @@ -730,7 +736,7 @@ static bool gfx_v9_4_3_aca_bank_is_valid(struct aca_handle *handle, struct aca_b } static const struct aca_bank_ops gfx_v9_4_3_aca_bank_ops = { - .aca_bank_generate_report = gfx_v9_4_3_aca_bank_generate_report, + .aca_bank_parser = gfx_v9_4_3_aca_bank_parser, .aca_bank_is_valid = gfx_v9_4_3_aca_bank_is_valid, }; @@ -2398,10 +2404,10 @@ gfx_v9_4_3_xcc_update_coarse_grain_clock_gating(struct amdgpu_device *adev, if (def != data) WREG32_SOC15(GC, GET_INST(GC, xcc_id), regRLC_CGTT_MGCG_OVERRIDE, data); - /* enable cgcg FSM(0x0000363F) */ + /* CGCG Hysteresis: 400us */ def = RREG32_SOC15(GC, GET_INST(GC, xcc_id), regRLC_CGCG_CGLS_CTRL); - data = (0x36 + data = (0x2710 << RLC_CGCG_CGLS_CTRL__CGCG_GFX_IDLE_THRESHOLD__SHIFT) | RLC_CGCG_CGLS_CTRL__CGCG_EN_MASK; if (adev->cg_flags & AMD_CG_SUPPORT_GFX_CGLS) @@ -2410,10 +2416,10 @@ gfx_v9_4_3_xcc_update_coarse_grain_clock_gating(struct amdgpu_device *adev, if (def != data) WREG32_SOC15(GC, GET_INST(GC, xcc_id), regRLC_CGCG_CGLS_CTRL, data); - /* set IDLE_POLL_COUNT(0x00900100) */ + /* set IDLE_POLL_COUNT(0x33450100)*/ def = RREG32_SOC15(GC, GET_INST(GC, xcc_id), regCP_RB_WPTR_POLL_CNTL); data = (0x0100 << CP_RB_WPTR_POLL_CNTL__POLL_FREQUENCY__SHIFT) | - (0x0090 << CP_RB_WPTR_POLL_CNTL__IDLE_POLL_COUNT__SHIFT); + (0x3345 << CP_RB_WPTR_POLL_CNTL__IDLE_POLL_COUNT__SHIFT); if (def != data) WREG32_SOC15(GC, GET_INST(GC, xcc_id), regCP_RB_WPTR_POLL_CNTL, data); } else { @@ -4010,6 +4016,8 @@ static const struct amd_ip_funcs gfx_v9_4_3_ip_funcs = { .set_clockgating_state = gfx_v9_4_3_set_clockgating_state, .set_powergating_state = gfx_v9_4_3_set_powergating_state, .get_clockgating_state = gfx_v9_4_3_get_clockgating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs gfx_v9_4_3_ring_funcs_compute = { diff --git a/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c b/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c index 22175da0e16a..d200310d1731 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c @@ -443,6 +443,22 @@ static void gfxhub_v1_0_init(struct amdgpu_device *adev) mmVM_INVALIDATE_ENG0_ADDR_RANGE_LO32; } +static bool gfxhub_v1_0_query_utcl2_poison_status(struct amdgpu_device *adev, + int xcc_id) +{ + u32 status = 0; + struct amdgpu_vmhub *hub; + + if (amdgpu_ip_version(adev, GC_HWIP, 0) != IP_VERSION(9, 4, 2)) + return false; + + hub = &adev->vmhub[AMDGPU_GFXHUB(0)]; + status = RREG32(hub->vm_l2_pro_fault_status); + /* reset page fault status */ + WREG32_P(hub->vm_l2_pro_fault_cntl, 1, ~1); + + return REG_GET_FIELD(status, VM_L2_PROTECTION_FAULT_STATUS, FED); +} const struct amdgpu_gfxhub_funcs gfxhub_v1_0_funcs = { .get_mc_fb_offset = gfxhub_v1_0_get_mc_fb_offset, @@ -452,4 +468,5 @@ const struct amdgpu_gfxhub_funcs gfxhub_v1_0_funcs = { .set_fault_enable_default = gfxhub_v1_0_set_fault_enable_default, .init = gfxhub_v1_0_init, .get_xgmi_info = gfxhub_v1_1_get_xgmi_info, + .query_utcl2_poison_status = gfxhub_v1_0_query_utcl2_poison_status, }; diff --git a/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_2.c b/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_2.c index 49aecdcee006..77df8c9cbad2 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_2.c +++ b/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_2.c @@ -620,6 +620,20 @@ static int gfxhub_v1_2_get_xgmi_info(struct amdgpu_device *adev) return 0; } +static bool gfxhub_v1_2_query_utcl2_poison_status(struct amdgpu_device *adev, + int xcc_id) +{ + u32 fed, status; + + status = RREG32_SOC15(GC, GET_INST(GC, xcc_id), regVM_L2_PROTECTION_FAULT_STATUS); + fed = REG_GET_FIELD(status, VM_L2_PROTECTION_FAULT_STATUS, FED); + /* reset page fault status */ + WREG32_P(SOC15_REG_OFFSET(GC, GET_INST(GC, xcc_id), + regVM_L2_PROTECTION_FAULT_STATUS), 1, ~1); + + return fed; +} + const struct amdgpu_gfxhub_funcs gfxhub_v1_2_funcs = { .get_mc_fb_offset = gfxhub_v1_2_get_mc_fb_offset, .setup_vm_pt_regs = gfxhub_v1_2_setup_vm_pt_regs, @@ -628,6 +642,7 @@ const struct amdgpu_gfxhub_funcs gfxhub_v1_2_funcs = { .set_fault_enable_default = gfxhub_v1_2_set_fault_enable_default, .init = gfxhub_v1_2_init, .get_xgmi_info = gfxhub_v1_2_get_xgmi_info, + .query_utcl2_poison_status = gfxhub_v1_2_query_utcl2_poison_status, }; static int gfxhub_v1_2_xcp_resume(void *handle, uint32_t inst_mask) diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c index 23b478639921..3e38d8bfcb69 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c @@ -1115,6 +1115,8 @@ static const struct amd_ip_funcs gmc_v6_0_ip_funcs = { .soft_reset = gmc_v6_0_soft_reset, .set_clockgating_state = gmc_v6_0_set_clockgating_state, .set_powergating_state = gmc_v6_0_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_gmc_funcs gmc_v6_0_gmc_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c index 3da7b6a2b00d..85df8fc81065 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c @@ -1354,6 +1354,8 @@ static const struct amd_ip_funcs gmc_v7_0_ip_funcs = { .soft_reset = gmc_v7_0_soft_reset, .set_clockgating_state = gmc_v7_0_set_clockgating_state, .set_powergating_state = gmc_v7_0_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_gmc_funcs gmc_v7_0_gmc_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c index d20e5f20ee31..fc97757e33d9 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c @@ -1717,6 +1717,8 @@ static const struct amd_ip_funcs gmc_v8_0_ip_funcs = { .set_clockgating_state = gmc_v8_0_set_clockgating_state, .set_powergating_state = gmc_v8_0_set_powergating_state, .get_clockgating_state = gmc_v8_0_get_clockgating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_gmc_funcs gmc_v8_0_gmc_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c index 47b63a4ce68b..c4ec1358f3aa 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c @@ -548,7 +548,7 @@ static int gmc_v9_0_process_interrupt(struct amdgpu_device *adev, { bool retry_fault = !!(entry->src_data[1] & 0x80); bool write_fault = !!(entry->src_data[1] & 0x20); - uint32_t status = 0, cid = 0, rw = 0; + uint32_t status = 0, cid = 0, rw = 0, fed = 0; struct amdgpu_task_info *task_info; struct amdgpu_vmhub *hub; const char *mmhub_cid; @@ -664,6 +664,13 @@ static int gmc_v9_0_process_interrupt(struct amdgpu_device *adev, status = RREG32(hub->vm_l2_pro_fault_status); cid = REG_GET_FIELD(status, VM_L2_PROTECTION_FAULT_STATUS, CID); rw = REG_GET_FIELD(status, VM_L2_PROTECTION_FAULT_STATUS, RW); + fed = REG_GET_FIELD(status, VM_L2_PROTECTION_FAULT_STATUS, FED); + + /* for fed error, kfd will handle it, return directly */ + if (fed && amdgpu_ras_is_poison_mode_supported(adev) && + (amdgpu_ip_version(adev, GC_HWIP, 0) >= IP_VERSION(9, 4, 2))) + return 0; + WREG32_P(hub->vm_l2_pro_fault_cntl, 1, ~1); amdgpu_vm_update_fault_cache(adev, entry->pasid, addr, status, vmhub); @@ -1450,7 +1457,6 @@ static void gmc_v9_0_set_umc_funcs(struct amdgpu_device *adev) adev->umc.channel_offs = UMC_V12_0_PER_CHANNEL_OFFSET; adev->umc.active_mask = adev->aid_mask; adev->umc.retire_unit = UMC_V12_0_BAD_PAGE_NUM_PER_CHANNEL; - adev->umc.channel_idx_tbl = &umc_v12_0_channel_idx_tbl[0][0][0]; if (!adev->gmc.xgmi.connected_to_cpu && !adev->gmc.is_app_apu) adev->umc.ras = &umc_v12_0_ras; break; diff --git a/drivers/gpu/drm/amd/amdgpu/iceland_ih.c b/drivers/gpu/drm/amd/amdgpu/iceland_ih.c index 2c02ae69883d..07984f7c3ae7 100644 --- a/drivers/gpu/drm/amd/amdgpu/iceland_ih.c +++ b/drivers/gpu/drm/amd/amdgpu/iceland_ih.c @@ -425,6 +425,8 @@ static const struct amd_ip_funcs iceland_ih_ip_funcs = { .soft_reset = iceland_ih_soft_reset, .set_clockgating_state = iceland_ih_set_clockgating_state, .set_powergating_state = iceland_ih_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_ih_funcs iceland_ih_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/ih_v6_0.c b/drivers/gpu/drm/amd/amdgpu/ih_v6_0.c index ad4ad39f128f..3cb64c8f7175 100644 --- a/drivers/gpu/drm/amd/amdgpu/ih_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/ih_v6_0.c @@ -346,6 +346,21 @@ static int ih_v6_0_irq_init(struct amdgpu_device *adev) DELAY, 3); WREG32_SOC15(OSSSYS, 0, regIH_MSI_STORM_CTRL, tmp); + /* Redirect the interrupts to IH RB1 for dGPU */ + if (adev->irq.ih1.ring_size) { + tmp = RREG32_SOC15(OSSSYS, 0, regIH_RING1_CLIENT_CFG_INDEX); + tmp = REG_SET_FIELD(tmp, IH_RING1_CLIENT_CFG_INDEX, INDEX, 0); + WREG32_SOC15(OSSSYS, 0, regIH_RING1_CLIENT_CFG_INDEX, tmp); + + tmp = RREG32_SOC15(OSSSYS, 0, regIH_RING1_CLIENT_CFG_DATA); + tmp = REG_SET_FIELD(tmp, IH_RING1_CLIENT_CFG_DATA, CLIENT_ID, 0xa); + tmp = REG_SET_FIELD(tmp, IH_RING1_CLIENT_CFG_DATA, SOURCE_ID, 0x0); + tmp = REG_SET_FIELD(tmp, IH_RING1_CLIENT_CFG_DATA, + SOURCE_ID_MATCH_ENABLE, 0x1); + + WREG32_SOC15(OSSSYS, 0, regIH_RING1_CLIENT_CFG_DATA, tmp); + } + pci_set_master(adev->pdev); /* enable interrupts */ @@ -549,8 +564,15 @@ static int ih_v6_0_sw_init(void *handle) adev->irq.ih.use_doorbell = true; adev->irq.ih.doorbell_index = adev->doorbell_index.ih << 1; - adev->irq.ih1.ring_size = 0; - adev->irq.ih2.ring_size = 0; + if (!(adev->flags & AMD_IS_APU)) { + r = amdgpu_ih_ring_init(adev, &adev->irq.ih1, IH_RING_SIZE, + use_bus_addr); + if (r) + return r; + + adev->irq.ih1.use_doorbell = true; + adev->irq.ih1.doorbell_index = (adev->doorbell_index.ih + 1) << 1; + } /* initialize ih control register offset */ ih_v6_0_init_register_offset(adev); @@ -748,6 +770,8 @@ static const struct amd_ip_funcs ih_v6_0_ip_funcs = { .set_clockgating_state = ih_v6_0_set_clockgating_state, .set_powergating_state = ih_v6_0_set_powergating_state, .get_clockgating_state = ih_v6_0_get_clockgating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_ih_funcs ih_v6_0_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/ih_v6_1.c b/drivers/gpu/drm/amd/amdgpu/ih_v6_1.c index b8da0fc29378..0fbf5fa7b0f8 100644 --- a/drivers/gpu/drm/amd/amdgpu/ih_v6_1.c +++ b/drivers/gpu/drm/amd/amdgpu/ih_v6_1.c @@ -346,6 +346,21 @@ static int ih_v6_1_irq_init(struct amdgpu_device *adev) DELAY, 3); WREG32_SOC15(OSSSYS, 0, regIH_MSI_STORM_CTRL, tmp); + /* Redirect the interrupts to IH RB1 for dGPU */ + if (adev->irq.ih1.ring_size) { + tmp = RREG32_SOC15(OSSSYS, 0, regIH_RING1_CLIENT_CFG_INDEX); + tmp = REG_SET_FIELD(tmp, IH_RING1_CLIENT_CFG_INDEX, INDEX, 0); + WREG32_SOC15(OSSSYS, 0, regIH_RING1_CLIENT_CFG_INDEX, tmp); + + tmp = RREG32_SOC15(OSSSYS, 0, regIH_RING1_CLIENT_CFG_DATA); + tmp = REG_SET_FIELD(tmp, IH_RING1_CLIENT_CFG_DATA, CLIENT_ID, 0xa); + tmp = REG_SET_FIELD(tmp, IH_RING1_CLIENT_CFG_DATA, SOURCE_ID, 0x0); + tmp = REG_SET_FIELD(tmp, IH_RING1_CLIENT_CFG_DATA, + SOURCE_ID_MATCH_ENABLE, 0x1); + + WREG32_SOC15(OSSSYS, 0, regIH_RING1_CLIENT_CFG_DATA, tmp); + } + pci_set_master(adev->pdev); /* enable interrupts */ @@ -550,8 +565,15 @@ static int ih_v6_1_sw_init(void *handle) adev->irq.ih.use_doorbell = true; adev->irq.ih.doorbell_index = adev->doorbell_index.ih << 1; - adev->irq.ih1.ring_size = 0; - adev->irq.ih2.ring_size = 0; + if (!(adev->flags & AMD_IS_APU)) { + r = amdgpu_ih_ring_init(adev, &adev->irq.ih1, IH_RING_SIZE, + use_bus_addr); + if (r) + return r; + + adev->irq.ih1.use_doorbell = true; + adev->irq.ih1.doorbell_index = (adev->doorbell_index.ih + 1) << 1; + } /* initialize ih control register offset */ ih_v6_1_init_register_offset(adev); @@ -753,6 +775,8 @@ static const struct amd_ip_funcs ih_v6_1_ip_funcs = { .set_clockgating_state = ih_v6_1_set_clockgating_state, .set_powergating_state = ih_v6_1_set_powergating_state, .get_clockgating_state = ih_v6_1_get_clockgating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_ih_funcs ih_v6_1_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/ih_v7_0.c b/drivers/gpu/drm/amd/amdgpu/ih_v7_0.c index 7aed96fa10a9..aa6235dd4f2b 100644 --- a/drivers/gpu/drm/amd/amdgpu/ih_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/ih_v7_0.c @@ -749,6 +749,8 @@ static const struct amd_ip_funcs ih_v7_0_ip_funcs = { .set_clockgating_state = ih_v7_0_set_clockgating_state, .set_powergating_state = ih_v7_0_set_powergating_state, .get_clockgating_state = ih_v7_0_get_clockgating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_ih_funcs ih_v7_0_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/jpeg_v2_0.c b/drivers/gpu/drm/amd/amdgpu/jpeg_v2_0.c index 1c8116d75f63..ef3e42f6b841 100644 --- a/drivers/gpu/drm/amd/amdgpu/jpeg_v2_0.c +++ b/drivers/gpu/drm/amd/amdgpu/jpeg_v2_0.c @@ -759,6 +759,8 @@ static const struct amd_ip_funcs jpeg_v2_0_ip_funcs = { .post_soft_reset = NULL, .set_clockgating_state = jpeg_v2_0_set_clockgating_state, .set_powergating_state = jpeg_v2_0_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs jpeg_v2_0_dec_ring_vm_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/jpeg_v2_5.c b/drivers/gpu/drm/amd/amdgpu/jpeg_v2_5.c index 99cd49ee8ef6..afeaf3c64e27 100644 --- a/drivers/gpu/drm/amd/amdgpu/jpeg_v2_5.c +++ b/drivers/gpu/drm/amd/amdgpu/jpeg_v2_5.c @@ -632,6 +632,8 @@ static const struct amd_ip_funcs jpeg_v2_5_ip_funcs = { .post_soft_reset = NULL, .set_clockgating_state = jpeg_v2_5_set_clockgating_state, .set_powergating_state = jpeg_v2_5_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amd_ip_funcs jpeg_v2_6_ip_funcs = { @@ -652,6 +654,8 @@ static const struct amd_ip_funcs jpeg_v2_6_ip_funcs = { .post_soft_reset = NULL, .set_clockgating_state = jpeg_v2_5_set_clockgating_state, .set_powergating_state = jpeg_v2_5_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs jpeg_v2_5_dec_ring_vm_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/jpeg_v3_0.c b/drivers/gpu/drm/amd/amdgpu/jpeg_v3_0.c index a92481da60cd..1c7cf4800bf7 100644 --- a/drivers/gpu/drm/amd/amdgpu/jpeg_v3_0.c +++ b/drivers/gpu/drm/amd/amdgpu/jpeg_v3_0.c @@ -557,6 +557,8 @@ static const struct amd_ip_funcs jpeg_v3_0_ip_funcs = { .post_soft_reset = NULL, .set_clockgating_state = jpeg_v3_0_set_clockgating_state, .set_powergating_state = jpeg_v3_0_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs jpeg_v3_0_dec_ring_vm_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0.c b/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0.c index 88ea58d5c4ab..237fe5df5a8f 100644 --- a/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0.c +++ b/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0.c @@ -719,6 +719,8 @@ static const struct amd_ip_funcs jpeg_v4_0_ip_funcs = { .post_soft_reset = NULL, .set_clockgating_state = jpeg_v4_0_set_clockgating_state, .set_powergating_state = jpeg_v4_0_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs jpeg_v4_0_dec_ring_vm_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0_3.c b/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0_3.c index 32caeb37cef9..d66af11aa66c 100644 --- a/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0_3.c +++ b/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0_3.c @@ -1053,6 +1053,8 @@ static const struct amd_ip_funcs jpeg_v4_0_3_ip_funcs = { .post_soft_reset = NULL, .set_clockgating_state = jpeg_v4_0_3_set_clockgating_state, .set_powergating_state = jpeg_v4_0_3_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs jpeg_v4_0_3_dec_ring_vm_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0_5.c b/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0_5.c index edf5bcdd2bc9..da6bb9022b80 100644 --- a/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0_5.c +++ b/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0_5.c @@ -762,6 +762,8 @@ static const struct amd_ip_funcs jpeg_v4_0_5_ip_funcs = { .post_soft_reset = NULL, .set_clockgating_state = jpeg_v4_0_5_set_clockgating_state, .set_powergating_state = jpeg_v4_0_5_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs jpeg_v4_0_5_dec_ring_vm_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/jpeg_v5_0_0.c b/drivers/gpu/drm/amd/amdgpu/jpeg_v5_0_0.c index e70200f97555..64c856bfe0cb 100644 --- a/drivers/gpu/drm/amd/amdgpu/jpeg_v5_0_0.c +++ b/drivers/gpu/drm/amd/amdgpu/jpeg_v5_0_0.c @@ -513,6 +513,8 @@ static const struct amd_ip_funcs jpeg_v5_0_0_ip_funcs = { .post_soft_reset = NULL, .set_clockgating_state = jpeg_v5_0_0_set_clockgating_state, .set_powergating_state = jpeg_v5_0_0_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs jpeg_v5_0_0_dec_ring_vm_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/mes_v10_1.c b/drivers/gpu/drm/amd/amdgpu/mes_v10_1.c index 1e5ad1e08d2a..a626bf904926 100644 --- a/drivers/gpu/drm/amd/amdgpu/mes_v10_1.c +++ b/drivers/gpu/drm/amd/amdgpu/mes_v10_1.c @@ -1176,6 +1176,8 @@ static const struct amd_ip_funcs mes_v10_1_ip_funcs = { .hw_fini = mes_v10_1_hw_fini, .suspend = mes_v10_1_suspend, .resume = mes_v10_1_resume, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; const struct amdgpu_ip_block_version mes_v10_1_ip_block = { diff --git a/drivers/gpu/drm/amd/amdgpu/mes_v11_0.c b/drivers/gpu/drm/amd/amdgpu/mes_v11_0.c index 072c478665ad..0d1407f25005 100644 --- a/drivers/gpu/drm/amd/amdgpu/mes_v11_0.c +++ b/drivers/gpu/drm/amd/amdgpu/mes_v11_0.c @@ -100,18 +100,76 @@ static const struct amdgpu_ring_funcs mes_v11_0_ring_funcs = { .insert_nop = amdgpu_ring_insert_nop, }; +static const char *mes_v11_0_opcodes[] = { + "SET_HW_RSRC", + "SET_SCHEDULING_CONFIG", + "ADD_QUEUE", + "REMOVE_QUEUE", + "PERFORM_YIELD", + "SET_GANG_PRIORITY_LEVEL", + "SUSPEND", + "RESUME", + "RESET", + "SET_LOG_BUFFER", + "CHANGE_GANG_PRORITY", + "QUERY_SCHEDULER_STATUS", + "PROGRAM_GDS", + "SET_DEBUG_VMID", + "MISC", + "UPDATE_ROOT_PAGE_TABLE", + "AMD_LOG", +}; + +static const char *mes_v11_0_misc_opcodes[] = { + "WRITE_REG", + "INV_GART", + "QUERY_STATUS", + "READ_REG", + "WAIT_REG_MEM", + "SET_SHADER_DEBUGGER", +}; + +static const char *mes_v11_0_get_op_string(union MESAPI__MISC *x_pkt) +{ + const char *op_str = NULL; + + if (x_pkt->header.opcode < ARRAY_SIZE(mes_v11_0_opcodes)) + op_str = mes_v11_0_opcodes[x_pkt->header.opcode]; + + return op_str; +} + +static const char *mes_v11_0_get_misc_op_string(union MESAPI__MISC *x_pkt) +{ + const char *op_str = NULL; + + if ((x_pkt->header.opcode == MES_SCH_API_MISC) && + (x_pkt->opcode < ARRAY_SIZE(mes_v11_0_misc_opcodes))) + op_str = mes_v11_0_misc_opcodes[x_pkt->opcode]; + + return op_str; +} + static int mes_v11_0_submit_pkt_and_poll_completion(struct amdgpu_mes *mes, void *pkt, int size, int api_status_off) { int ndw = size / 4; signed long r; - union MESAPI__ADD_QUEUE *x_pkt = pkt; + union MESAPI__MISC *x_pkt = pkt; struct MES_API_STATUS *api_status; struct amdgpu_device *adev = mes->adev; struct amdgpu_ring *ring = &mes->ring; unsigned long flags; - signed long timeout = adev->usec_timeout; + signed long timeout = 3000000; /* 3000 ms */ + const char *op_str, *misc_op_str; + u32 fence_offset; + u64 fence_gpu_addr; + u64 *fence_ptr; + int ret; + + if (x_pkt->header.opcode >= MES_SCH_API_MAX) + return -EINVAL; if (amdgpu_emu_mode) { timeout *= 100; @@ -121,27 +179,52 @@ static int mes_v11_0_submit_pkt_and_poll_completion(struct amdgpu_mes *mes, } BUG_ON(size % 4 != 0); + ret = amdgpu_device_wb_get(adev, &fence_offset); + if (ret) + return ret; + fence_gpu_addr = + adev->wb.gpu_addr + (fence_offset * 4); + fence_ptr = (u64 *)&adev->wb.wb[fence_offset]; + *fence_ptr = 0; + spin_lock_irqsave(&mes->ring_lock, flags); if (amdgpu_ring_alloc(ring, ndw)) { spin_unlock_irqrestore(&mes->ring_lock, flags); + amdgpu_device_wb_free(adev, fence_offset); return -ENOMEM; } api_status = (struct MES_API_STATUS *)((char *)pkt + api_status_off); - api_status->api_completion_fence_addr = mes->ring.fence_drv.gpu_addr; - api_status->api_completion_fence_value = ++mes->ring.fence_drv.sync_seq; + api_status->api_completion_fence_addr = fence_gpu_addr; + api_status->api_completion_fence_value = 1; amdgpu_ring_write_multiple(ring, pkt, ndw); amdgpu_ring_commit(ring); spin_unlock_irqrestore(&mes->ring_lock, flags); - DRM_DEBUG("MES msg=%d was emitted\n", x_pkt->header.opcode); + op_str = mes_v11_0_get_op_string(x_pkt); + misc_op_str = mes_v11_0_get_misc_op_string(x_pkt); + + if (misc_op_str) + dev_dbg(adev->dev, "MES msg=%s (%s) was emitted\n", op_str, misc_op_str); + else if (op_str) + dev_dbg(adev->dev, "MES msg=%s was emitted\n", op_str); + else + dev_dbg(adev->dev, "MES msg=%d was emitted\n", x_pkt->header.opcode); - r = amdgpu_fence_wait_polling(ring, ring->fence_drv.sync_seq, - timeout); + r = amdgpu_mes_fence_wait_polling(fence_ptr, (u64)1, timeout); + amdgpu_device_wb_free(adev, fence_offset); if (r < 1) { - DRM_ERROR("MES failed to response msg=%d\n", - x_pkt->header.opcode); + + if (misc_op_str) + dev_err(adev->dev, "MES failed to respond to msg=%s (%s)\n", + op_str, misc_op_str); + else if (op_str) + dev_err(adev->dev, "MES failed to respond to msg=%s\n", + op_str); + else + dev_err(adev->dev, "MES failed to respond to msg=%d\n", + x_pkt->header.opcode); while (halt_if_hws_hang) schedule(); @@ -411,14 +494,47 @@ static int mes_v11_0_set_hw_resources(struct amdgpu_mes *mes) mes_set_hw_res_pkt.enable_reg_active_poll = 1; mes_set_hw_res_pkt.enable_level_process_quantum_check = 1; mes_set_hw_res_pkt.oversubscription_timer = 50; - mes_set_hw_res_pkt.enable_mes_event_int_logging = 1; - mes_set_hw_res_pkt.event_intr_history_gpu_mc_ptr = mes->event_log_gpu_addr; + if (amdgpu_mes_log_enable) { + mes_set_hw_res_pkt.enable_mes_event_int_logging = 1; + mes_set_hw_res_pkt.event_intr_history_gpu_mc_ptr = + mes->event_log_gpu_addr; + } return mes_v11_0_submit_pkt_and_poll_completion(mes, &mes_set_hw_res_pkt, sizeof(mes_set_hw_res_pkt), offsetof(union MESAPI_SET_HW_RESOURCES, api_status)); } +static int mes_v11_0_set_hw_resources_1(struct amdgpu_mes *mes) +{ + int size = 128 * PAGE_SIZE; + int ret = 0; + struct amdgpu_device *adev = mes->adev; + union MESAPI_SET_HW_RESOURCES_1 mes_set_hw_res_pkt; + memset(&mes_set_hw_res_pkt, 0, sizeof(mes_set_hw_res_pkt)); + + mes_set_hw_res_pkt.header.type = MES_API_TYPE_SCHEDULER; + mes_set_hw_res_pkt.header.opcode = MES_SCH_API_SET_HW_RSRC_1; + mes_set_hw_res_pkt.header.dwsize = API_FRAME_SIZE_IN_DWORDS; + mes_set_hw_res_pkt.enable_mes_info_ctx = 1; + + ret = amdgpu_bo_create_kernel(adev, size, PAGE_SIZE, + AMDGPU_GEM_DOMAIN_VRAM, + &mes->resource_1, + &mes->resource_1_gpu_addr, + &mes->resource_1_addr); + if (ret) { + dev_err(adev->dev, "(%d) failed to create mes resource_1 bo\n", ret); + return ret; + } + + mes_set_hw_res_pkt.mes_info_ctx_mc_addr = mes->resource_1_gpu_addr; + mes_set_hw_res_pkt.mes_info_ctx_size = mes->resource_1->tbo.base.size; + return mes_v11_0_submit_pkt_and_poll_completion(mes, + &mes_set_hw_res_pkt, sizeof(mes_set_hw_res_pkt), + offsetof(union MESAPI_SET_HW_RESOURCES_1, api_status)); +} + static const struct amdgpu_mes_funcs mes_v11_0_funcs = { .add_hw_queue = mes_v11_0_add_hw_queue, .remove_hw_queue = mes_v11_0_remove_hw_queue, @@ -1200,6 +1316,14 @@ static int mes_v11_0_hw_init(void *handle) if (r) goto failure; + if (amdgpu_sriov_is_mes_info_enable(adev)) { + r = mes_v11_0_set_hw_resources_1(&adev->mes); + if (r) { + DRM_ERROR("failed mes_v11_0_set_hw_resources_1, r=%d\n", r); + goto failure; + } + } + r = mes_v11_0_query_sched_status(&adev->mes); if (r) { DRM_ERROR("MES is busy\n"); @@ -1223,6 +1347,11 @@ failure: static int mes_v11_0_hw_fini(void *handle) { + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + if (amdgpu_sriov_is_mes_info_enable(adev)) { + amdgpu_bo_free_kernel(&adev->mes.resource_1, &adev->mes.resource_1_gpu_addr, + &adev->mes.resource_1_addr); + } return 0; } @@ -1288,6 +1417,8 @@ static const struct amd_ip_funcs mes_v11_0_ip_funcs = { .hw_fini = mes_v11_0_hw_fini, .suspend = mes_v11_0_suspend, .resume = mes_v11_0_resume, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; const struct amdgpu_ip_block_version mes_v11_0_ip_block = { diff --git a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_8.c b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_8.c index c0fc44cdd658..7a1ff298417a 100644 --- a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_8.c +++ b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_8.c @@ -559,6 +559,20 @@ static void mmhub_v1_8_get_clockgating(struct amdgpu_device *adev, u64 *flags) } +static bool mmhub_v1_8_query_utcl2_poison_status(struct amdgpu_device *adev, + int hub_inst) +{ + u32 fed, status; + + status = RREG32_SOC15(MMHUB, hub_inst, regVM_L2_PROTECTION_FAULT_STATUS); + fed = REG_GET_FIELD(status, VM_L2_PROTECTION_FAULT_STATUS, FED); + /* reset page fault status */ + WREG32_P(SOC15_REG_OFFSET(MMHUB, hub_inst, + regVM_L2_PROTECTION_FAULT_STATUS), 1, ~1); + + return fed; +} + const struct amdgpu_mmhub_funcs mmhub_v1_8_funcs = { .get_fb_location = mmhub_v1_8_get_fb_location, .init = mmhub_v1_8_init, @@ -568,6 +582,7 @@ const struct amdgpu_mmhub_funcs mmhub_v1_8_funcs = { .setup_vm_pt_regs = mmhub_v1_8_setup_vm_pt_regs, .set_clockgating = mmhub_v1_8_set_clockgating, .get_clockgating = mmhub_v1_8_get_clockgating, + .query_utcl2_poison_status = mmhub_v1_8_query_utcl2_poison_status, }; static const struct amdgpu_ras_err_status_reg_entry mmhub_v1_8_ce_reg_list[] = { @@ -706,28 +721,32 @@ static const struct amdgpu_ras_block_hw_ops mmhub_v1_8_ras_hw_ops = { .reset_ras_error_count = mmhub_v1_8_reset_ras_error_count, }; -static int mmhub_v1_8_aca_bank_generate_report(struct aca_handle *handle, - struct aca_bank *bank, enum aca_error_type type, - struct aca_bank_report *report, void *data) +static int mmhub_v1_8_aca_bank_parser(struct aca_handle *handle, struct aca_bank *bank, + enum aca_smu_type type, void *data) { - u64 status, misc0; + struct aca_bank_info info; + u64 misc0; int ret; - status = bank->regs[ACA_REG_IDX_STATUS]; - if ((type == ACA_ERROR_TYPE_UE && - ACA_REG__STATUS__ERRORCODEEXT(status) == ACA_EXTERROR_CODE_FAULT) || - (type == ACA_ERROR_TYPE_CE && - ACA_REG__STATUS__ERRORCODEEXT(status) == ACA_EXTERROR_CODE_CE)) { - - ret = aca_bank_info_decode(bank, &report->info); - if (ret) - return ret; - - misc0 = bank->regs[ACA_REG_IDX_MISC0]; - report->count[type] = ACA_REG__MISC0__ERRCNT(misc0); + ret = aca_bank_info_decode(bank, &info); + if (ret) + return ret; + + misc0 = bank->regs[ACA_REG_IDX_MISC0]; + switch (type) { + case ACA_SMU_TYPE_UE: + ret = aca_error_cache_log_bank_error(handle, &info, ACA_ERROR_TYPE_UE, + 1ULL); + break; + case ACA_SMU_TYPE_CE: + ret = aca_error_cache_log_bank_error(handle, &info, ACA_ERROR_TYPE_CE, + ACA_REG__MISC0__ERRCNT(misc0)); + break; + default: + return -EINVAL; } - return 0; + return ret; } /* reference to smu driver if header file */ @@ -741,7 +760,7 @@ static int mmhub_v1_8_err_codes[] = { }; static bool mmhub_v1_8_aca_bank_is_valid(struct aca_handle *handle, struct aca_bank *bank, - enum aca_error_type type, void *data) + enum aca_smu_type type, void *data) { u32 instlo; @@ -760,7 +779,7 @@ static bool mmhub_v1_8_aca_bank_is_valid(struct aca_handle *handle, struct aca_b } static const struct aca_bank_ops mmhub_v1_8_aca_bank_ops = { - .aca_bank_generate_report = mmhub_v1_8_aca_bank_generate_report, + .aca_bank_parser = mmhub_v1_8_aca_bank_parser, .aca_bank_is_valid = mmhub_v1_8_aca_bank_is_valid, }; diff --git a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c index a2bd2c3b1ef9..0c7275bca8f7 100644 --- a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c +++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c @@ -276,6 +276,8 @@ static void xgpu_ai_mailbox_flr_work(struct work_struct *work) timeout -= 10; } while (timeout > 1); + dev_warn(adev->dev, "waiting IDH_FLR_NOTIFICATION_CMPL timeout\n"); + flr_done: atomic_set(&adev->reset_domain->in_gpu_reset, 0); up_write(&adev->reset_domain->sem); diff --git a/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c b/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c index 77f5b55decf9..aba00d961627 100644 --- a/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c +++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c @@ -309,6 +309,8 @@ static void xgpu_nv_mailbox_flr_work(struct work_struct *work) timeout -= 10; } while (timeout > 1); + dev_warn(adev->dev, "waiting IDH_FLR_NOTIFICATION_CMPL timeout\n"); + flr_done: atomic_set(&adev->reset_domain->in_gpu_reset, 0); up_write(&adev->reset_domain->sem); @@ -444,7 +446,6 @@ static void xgpu_nv_ras_poison_handler(struct amdgpu_device *adev, amdgpu_virt_fini_data_exchange(adev); xgpu_nv_send_access_requests_with_param(adev, IDH_RAS_POISON, block, 0, 0); - amdgpu_virt_init_data_exchange(adev); } } diff --git a/drivers/gpu/drm/amd/amdgpu/navi10_ih.c b/drivers/gpu/drm/amd/amdgpu/navi10_ih.c index 4178f4e5dad7..b281462093f1 100644 --- a/drivers/gpu/drm/amd/amdgpu/navi10_ih.c +++ b/drivers/gpu/drm/amd/amdgpu/navi10_ih.c @@ -713,6 +713,8 @@ static const struct amd_ip_funcs navi10_ih_ip_funcs = { .set_clockgating_state = navi10_ih_set_clockgating_state, .set_powergating_state = navi10_ih_set_powergating_state, .get_clockgating_state = navi10_ih_get_clockgating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_ih_funcs navi10_ih_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/nv.c b/drivers/gpu/drm/amd/amdgpu/nv.c index 4d7976b77767..12e54047bf79 100644 --- a/drivers/gpu/drm/amd/amdgpu/nv.c +++ b/drivers/gpu/drm/amd/amdgpu/nv.c @@ -110,7 +110,7 @@ static const struct amdgpu_video_codec_info sc_video_codecs_decode_array_vcn0[] {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_MPEG4_AVC, 4096, 4096, 52)}, {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_VC1, 4096, 4096, 4)}, {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_HEVC, 8192, 4352, 186)}, - {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_JPEG, 4096, 4096, 0)}, + {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_JPEG, 16384, 16384, 0)}, {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_VP9, 8192, 4352, 0)}, {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_AV1, 8192, 4352, 0)}, }; @@ -121,7 +121,7 @@ static const struct amdgpu_video_codec_info sc_video_codecs_decode_array_vcn1[] {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_MPEG4_AVC, 4096, 4096, 52)}, {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_VC1, 4096, 4096, 4)}, {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_HEVC, 8192, 4352, 186)}, - {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_JPEG, 4096, 4096, 0)}, + {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_JPEG, 16384, 16384, 0)}, {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_VP9, 8192, 4352, 0)}, }; @@ -199,7 +199,7 @@ static const struct amdgpu_video_codec_info yc_video_codecs_decode_array[] = { {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_MPEG4_AVC, 4096, 4096, 52)}, {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_HEVC, 8192, 4352, 186)}, {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_VP9, 8192, 4352, 0)}, - {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_JPEG, 4096, 4096, 0)}, + {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_JPEG, 16384, 16384, 0)}, {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_AV1, 8192, 4352, 0)}, }; @@ -1131,4 +1131,6 @@ static const struct amd_ip_funcs nv_common_ip_funcs = { .set_clockgating_state = nv_common_set_clockgating_state, .set_powergating_state = nv_common_set_powergating_state, .get_clockgating_state = nv_common_get_clockgating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v14_0.c b/drivers/gpu/drm/amd/amdgpu/psp_v14_0.c index 78a95f8f370b..f08a32c18694 100644 --- a/drivers/gpu/drm/amd/amdgpu/psp_v14_0.c +++ b/drivers/gpu/drm/amd/amdgpu/psp_v14_0.c @@ -169,7 +169,8 @@ static int psp_v14_0_bootloader_load_intf_drv(struct psp_context *psp) static int psp_v14_0_bootloader_load_dbg_drv(struct psp_context *psp) { - return psp_v14_0_bootloader_load_component(psp, &psp->dbg_drv, PSP_BL__LOAD_DBGDRV); + /* dbg_drv was renamed to had_drv in psp v14 */ + return psp_v14_0_bootloader_load_component(psp, &psp->dbg_drv, PSP_BL__LOAD_HADDRV); } static int psp_v14_0_bootloader_load_ras_drv(struct psp_context *psp) @@ -177,6 +178,10 @@ static int psp_v14_0_bootloader_load_ras_drv(struct psp_context *psp) return psp_v14_0_bootloader_load_component(psp, &psp->ras_drv, PSP_BL__LOAD_RASDRV); } +static int psp_v14_0_bootloader_load_ipkeymgr_drv(struct psp_context *psp) +{ + return psp_v14_0_bootloader_load_component(psp, &psp->ipkeymgr_drv, PSP_BL__LOAD_IPKEYMGRDRV); +} static int psp_v14_0_bootloader_load_sos(struct psp_context *psp) { @@ -653,6 +658,7 @@ static const struct psp_funcs psp_v14_0_funcs = { .bootloader_load_intf_drv = psp_v14_0_bootloader_load_intf_drv, .bootloader_load_dbg_drv = psp_v14_0_bootloader_load_dbg_drv, .bootloader_load_ras_drv = psp_v14_0_bootloader_load_ras_drv, + .bootloader_load_ipkeymgr_drv = psp_v14_0_bootloader_load_ipkeymgr_drv, .bootloader_load_sos = psp_v14_0_bootloader_load_sos, .ring_create = psp_v14_0_ring_create, .ring_stop = psp_v14_0_ring_stop, diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c b/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c index 07e19caf2bc1..ac8a9b9b3e52 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c @@ -1113,6 +1113,8 @@ static const struct amd_ip_funcs sdma_v2_4_ip_funcs = { .soft_reset = sdma_v2_4_soft_reset, .set_clockgating_state = sdma_v2_4_set_clockgating_state, .set_powergating_state = sdma_v2_4_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs sdma_v2_4_ring_funcs = { @@ -1176,7 +1178,7 @@ static void sdma_v2_4_set_irq_funcs(struct amdgpu_device *adev) * @src_offset: src GPU address * @dst_offset: dst GPU address * @byte_count: number of bytes to xfer - * @tmz: unused + * @copy_flags: unused * * Copy GPU buffers using the DMA engine (VI). * Used by the amdgpu ttm implementation to move pages if @@ -1186,7 +1188,7 @@ static void sdma_v2_4_emit_copy_buffer(struct amdgpu_ib *ib, uint64_t src_offset, uint64_t dst_offset, uint32_t byte_count, - bool tmz) + uint32_t copy_flags) { ib->ptr[ib->length_dw++] = SDMA_PKT_HEADER_OP(SDMA_OP_COPY) | SDMA_PKT_HEADER_SUB_OP(SDMA_SUBOP_COPY_LINEAR); diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c index 2ad615be4bb3..b8ebdc4ae6f6 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c @@ -1553,6 +1553,8 @@ static const struct amd_ip_funcs sdma_v3_0_ip_funcs = { .set_clockgating_state = sdma_v3_0_set_clockgating_state, .set_powergating_state = sdma_v3_0_set_powergating_state, .get_clockgating_state = sdma_v3_0_get_clockgating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs sdma_v3_0_ring_funcs = { @@ -1616,7 +1618,7 @@ static void sdma_v3_0_set_irq_funcs(struct amdgpu_device *adev) * @src_offset: src GPU address * @dst_offset: dst GPU address * @byte_count: number of bytes to xfer - * @tmz: unused + * @copy_flags: unused * * Copy GPU buffers using the DMA engine (VI). * Used by the amdgpu ttm implementation to move pages if @@ -1626,7 +1628,7 @@ static void sdma_v3_0_emit_copy_buffer(struct amdgpu_ib *ib, uint64_t src_offset, uint64_t dst_offset, uint32_t byte_count, - bool tmz) + uint32_t copy_flags) { ib->ptr[ib->length_dw++] = SDMA_PKT_HEADER_OP(SDMA_OP_COPY) | SDMA_PKT_HEADER_SUB_OP(SDMA_SUBOP_COPY_LINEAR); diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c index 43775cb67ff5..101038395c3b 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c @@ -2021,6 +2021,9 @@ static int sdma_v4_0_process_trap_irq(struct amdgpu_device *adev, DRM_DEBUG("IH: SDMA trap\n"); instance = sdma_v4_0_irq_id_to_seq(entry->client_id); + if (instance < 0) + return instance; + switch (entry->ring_id) { case 0: amdgpu_fence_process(&adev->sdma.instance[instance].ring); @@ -2448,7 +2451,7 @@ static void sdma_v4_0_set_irq_funcs(struct amdgpu_device *adev) * @src_offset: src GPU address * @dst_offset: dst GPU address * @byte_count: number of bytes to xfer - * @tmz: if a secure copy should be used + * @copy_flags: copy flags for the buffers * * Copy GPU buffers using the DMA engine (VEGA10/12). * Used by the amdgpu ttm implementation to move pages if @@ -2458,11 +2461,11 @@ static void sdma_v4_0_emit_copy_buffer(struct amdgpu_ib *ib, uint64_t src_offset, uint64_t dst_offset, uint32_t byte_count, - bool tmz) + uint32_t copy_flags) { ib->ptr[ib->length_dw++] = SDMA_PKT_HEADER_OP(SDMA_OP_COPY) | SDMA_PKT_HEADER_SUB_OP(SDMA_SUBOP_COPY_LINEAR) | - SDMA_PKT_COPY_LINEAR_HEADER_TMZ(tmz ? 1 : 0); + SDMA_PKT_COPY_LINEAR_HEADER_TMZ((copy_flags & AMDGPU_COPY_FLAGS_TMZ) ? 1 : 0); ib->ptr[ib->length_dw++] = byte_count - 1; ib->ptr[ib->length_dw++] = 0; /* src/dst endian swap */ ib->ptr[ib->length_dw++] = lower_32_bits(src_offset); diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v4_4_2.c b/drivers/gpu/drm/amd/amdgpu/sdma_v4_4_2.c index 34237a1b1f2e..341b24d8320b 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v4_4_2.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v4_4_2.c @@ -368,7 +368,8 @@ static void sdma_v4_4_2_ring_emit_hdp_flush(struct amdgpu_ring *ring) u32 ref_and_mask = 0; const struct nbio_hdp_flush_reg *nbio_hf_reg = adev->nbio.hdp_flush_reg; - ref_and_mask = nbio_hf_reg->ref_and_mask_sdma0 << ring->me; + ref_and_mask = nbio_hf_reg->ref_and_mask_sdma0 + << (ring->me % adev->sdma.num_inst_per_aid); sdma_v4_4_2_wait_reg_mem(ring, 0, 1, adev->nbio.funcs->get_hdp_flush_done_offset(adev), @@ -1602,19 +1603,9 @@ static int sdma_v4_4_2_set_ecc_irq_state(struct amdgpu_device *adev, u32 sdma_cntl; sdma_cntl = RREG32_SDMA(type, regSDMA_CNTL); - switch (state) { - case AMDGPU_IRQ_STATE_DISABLE: - sdma_cntl = REG_SET_FIELD(sdma_cntl, SDMA_CNTL, - DRAM_ECC_INT_ENABLE, 0); - WREG32_SDMA(type, regSDMA_CNTL, sdma_cntl); - break; - /* sdma ecc interrupt is enabled by default - * driver doesn't need to do anything to - * enable the interrupt */ - case AMDGPU_IRQ_STATE_ENABLE: - default: - break; - } + sdma_cntl = REG_SET_FIELD(sdma_cntl, SDMA_CNTL, DRAM_ECC_INT_ENABLE, + state == AMDGPU_IRQ_STATE_ENABLE ? 1 : 0); + WREG32_SDMA(type, regSDMA_CNTL, sdma_cntl); return 0; } @@ -1954,7 +1945,7 @@ static void sdma_v4_4_2_set_irq_funcs(struct amdgpu_device *adev) * @src_offset: src GPU address * @dst_offset: dst GPU address * @byte_count: number of bytes to xfer - * @tmz: if a secure copy should be used + * @copy_flags: copy flags for the buffers * * Copy GPU buffers using the DMA engine. * Used by the amdgpu ttm implementation to move pages if @@ -1964,11 +1955,11 @@ static void sdma_v4_4_2_emit_copy_buffer(struct amdgpu_ib *ib, uint64_t src_offset, uint64_t dst_offset, uint32_t byte_count, - bool tmz) + uint32_t copy_flags) { ib->ptr[ib->length_dw++] = SDMA_PKT_HEADER_OP(SDMA_OP_COPY) | SDMA_PKT_HEADER_SUB_OP(SDMA_SUBOP_COPY_LINEAR) | - SDMA_PKT_COPY_LINEAR_HEADER_TMZ(tmz ? 1 : 0); + SDMA_PKT_COPY_LINEAR_HEADER_TMZ((copy_flags & AMDGPU_COPY_FLAGS_TMZ) ? 1 : 0); ib->ptr[ib->length_dw++] = byte_count - 1; ib->ptr[ib->length_dw++] = 0; /* src/dst endian swap */ ib->ptr[ib->length_dw++] = lower_32_bits(src_offset); @@ -2189,35 +2180,39 @@ static const struct amdgpu_ras_block_hw_ops sdma_v4_4_2_ras_hw_ops = { .reset_ras_error_count = sdma_v4_4_2_reset_ras_error_count, }; -static int sdma_v4_4_2_aca_bank_generate_report(struct aca_handle *handle, - struct aca_bank *bank, enum aca_error_type type, - struct aca_bank_report *report, void *data) +static int sdma_v4_4_2_aca_bank_parser(struct aca_handle *handle, struct aca_bank *bank, + enum aca_smu_type type, void *data) { - u64 status, misc0; + struct aca_bank_info info; + u64 misc0; int ret; - status = bank->regs[ACA_REG_IDX_STATUS]; - if ((type == ACA_ERROR_TYPE_UE && - ACA_REG__STATUS__ERRORCODEEXT(status) == ACA_EXTERROR_CODE_FAULT) || - (type == ACA_ERROR_TYPE_CE && - ACA_REG__STATUS__ERRORCODEEXT(status) == ACA_EXTERROR_CODE_CE)) { - - ret = aca_bank_info_decode(bank, &report->info); - if (ret) - return ret; + ret = aca_bank_info_decode(bank, &info); + if (ret) + return ret; - misc0 = bank->regs[ACA_REG_IDX_MISC0]; - report->count[type] = ACA_REG__MISC0__ERRCNT(misc0); + misc0 = bank->regs[ACA_REG_IDX_MISC0]; + switch (type) { + case ACA_SMU_TYPE_UE: + ret = aca_error_cache_log_bank_error(handle, &info, ACA_ERROR_TYPE_UE, + 1ULL); + break; + case ACA_SMU_TYPE_CE: + ret = aca_error_cache_log_bank_error(handle, &info, ACA_ERROR_TYPE_CE, + ACA_REG__MISC0__ERRCNT(misc0)); + break; + default: + return -EINVAL; } - return 0; + return ret; } /* CODE_SDMA0 - CODE_SDMA4, reference to smu driver if header file */ static int sdma_v4_4_2_err_codes[] = { 33, 34, 35, 36 }; static bool sdma_v4_4_2_aca_bank_is_valid(struct aca_handle *handle, struct aca_bank *bank, - enum aca_error_type type, void *data) + enum aca_smu_type type, void *data) { u32 instlo; @@ -2236,7 +2231,7 @@ static bool sdma_v4_4_2_aca_bank_is_valid(struct aca_handle *handle, struct aca_ } static const struct aca_bank_ops sdma_v4_4_2_aca_bank_ops = { - .aca_bank_generate_report = sdma_v4_4_2_aca_bank_generate_report, + .aca_bank_parser = sdma_v4_4_2_aca_bank_parser, .aca_bank_is_valid = sdma_v4_4_2_aca_bank_is_valid, }; diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c index 883e8a1b8a40..b7d33d78bce0 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c @@ -999,7 +999,8 @@ static int sdma_v5_0_ring_test_ring(struct amdgpu_ring *ring) r = amdgpu_ring_alloc(ring, 20); if (r) { DRM_ERROR("amdgpu: dma failed to lock ring %d (%d).\n", ring->idx, r); - amdgpu_device_wb_free(adev, index); + if (!ring->is_mes_queue) + amdgpu_device_wb_free(adev, index); return r; } @@ -1805,7 +1806,7 @@ static void sdma_v5_0_set_irq_funcs(struct amdgpu_device *adev) * @src_offset: src GPU address * @dst_offset: dst GPU address * @byte_count: number of bytes to xfer - * @tmz: if a secure copy should be used + * @copy_flags: copy flags for the buffers * * Copy GPU buffers using the DMA engine (NAVI10). * Used by the amdgpu ttm implementation to move pages if @@ -1815,11 +1816,11 @@ static void sdma_v5_0_emit_copy_buffer(struct amdgpu_ib *ib, uint64_t src_offset, uint64_t dst_offset, uint32_t byte_count, - bool tmz) + uint32_t copy_flags) { ib->ptr[ib->length_dw++] = SDMA_PKT_HEADER_OP(SDMA_OP_COPY) | SDMA_PKT_HEADER_SUB_OP(SDMA_SUBOP_COPY_LINEAR) | - SDMA_PKT_COPY_LINEAR_HEADER_TMZ(tmz ? 1 : 0); + SDMA_PKT_COPY_LINEAR_HEADER_TMZ((copy_flags & AMDGPU_COPY_FLAGS_TMZ) ? 1 : 0); ib->ptr[ib->length_dw++] = byte_count - 1; ib->ptr[ib->length_dw++] = 0; /* src/dst endian swap */ ib->ptr[ib->length_dw++] = lower_32_bits(src_offset); diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c b/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c index 42f4bd250def..cc9e961f0078 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c @@ -280,17 +280,21 @@ static void sdma_v5_2_ring_emit_hdp_flush(struct amdgpu_ring *ring) u32 ref_and_mask = 0; const struct nbio_hdp_flush_reg *nbio_hf_reg = adev->nbio.hdp_flush_reg; - ref_and_mask = nbio_hf_reg->ref_and_mask_sdma0 << ring->me; - - amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_POLL_REGMEM) | - SDMA_PKT_POLL_REGMEM_HEADER_HDP_FLUSH(1) | - SDMA_PKT_POLL_REGMEM_HEADER_FUNC(3)); /* == */ - amdgpu_ring_write(ring, (adev->nbio.funcs->get_hdp_flush_done_offset(adev)) << 2); - amdgpu_ring_write(ring, (adev->nbio.funcs->get_hdp_flush_req_offset(adev)) << 2); - amdgpu_ring_write(ring, ref_and_mask); /* reference */ - amdgpu_ring_write(ring, ref_and_mask); /* mask */ - amdgpu_ring_write(ring, SDMA_PKT_POLL_REGMEM_DW5_RETRY_COUNT(0xfff) | - SDMA_PKT_POLL_REGMEM_DW5_INTERVAL(10)); /* retry count, poll interval */ + if (ring->me > 1) { + amdgpu_asic_flush_hdp(adev, ring); + } else { + ref_and_mask = nbio_hf_reg->ref_and_mask_sdma0 << ring->me; + + amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_POLL_REGMEM) | + SDMA_PKT_POLL_REGMEM_HEADER_HDP_FLUSH(1) | + SDMA_PKT_POLL_REGMEM_HEADER_FUNC(3)); /* == */ + amdgpu_ring_write(ring, (adev->nbio.funcs->get_hdp_flush_done_offset(adev)) << 2); + amdgpu_ring_write(ring, (adev->nbio.funcs->get_hdp_flush_req_offset(adev)) << 2); + amdgpu_ring_write(ring, ref_and_mask); /* reference */ + amdgpu_ring_write(ring, ref_and_mask); /* mask */ + amdgpu_ring_write(ring, SDMA_PKT_POLL_REGMEM_DW5_RETRY_COUNT(0xfff) | + SDMA_PKT_POLL_REGMEM_DW5_INTERVAL(10)); /* retry count, poll interval */ + } } /** @@ -835,7 +839,8 @@ static int sdma_v5_2_ring_test_ring(struct amdgpu_ring *ring) r = amdgpu_ring_alloc(ring, 20); if (r) { DRM_ERROR("amdgpu: dma failed to lock ring %d (%d).\n", ring->idx, r); - amdgpu_device_wb_free(adev, index); + if (!ring->is_mes_queue) + amdgpu_device_wb_free(adev, index); return r; } @@ -1747,7 +1752,7 @@ static void sdma_v5_2_set_irq_funcs(struct amdgpu_device *adev) * @src_offset: src GPU address * @dst_offset: dst GPU address * @byte_count: number of bytes to xfer - * @tmz: if a secure copy should be used + * @copy_flags: copy flags for the buffers * * Copy GPU buffers using the DMA engine. * Used by the amdgpu ttm implementation to move pages if @@ -1757,11 +1762,11 @@ static void sdma_v5_2_emit_copy_buffer(struct amdgpu_ib *ib, uint64_t src_offset, uint64_t dst_offset, uint32_t byte_count, - bool tmz) + uint32_t copy_flags) { ib->ptr[ib->length_dw++] = SDMA_PKT_HEADER_OP(SDMA_OP_COPY) | SDMA_PKT_HEADER_SUB_OP(SDMA_SUBOP_COPY_LINEAR) | - SDMA_PKT_COPY_LINEAR_HEADER_TMZ(tmz ? 1 : 0); + SDMA_PKT_COPY_LINEAR_HEADER_TMZ((copy_flags & AMDGPU_COPY_FLAGS_TMZ) ? 1 : 0); ib->ptr[ib->length_dw++] = byte_count - 1; ib->ptr[ib->length_dw++] = 0; /* src/dst endian swap */ ib->ptr[ib->length_dw++] = lower_32_bits(src_offset); diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v6_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v6_0.c index 361835a61f2e..c833b6b8373b 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v6_0.c @@ -507,6 +507,13 @@ static int sdma_v6_0_gfx_resume(struct amdgpu_device *adev) /* set minor_ptr_update to 0 after wptr programed */ WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_MINOR_PTR_UPDATE), 0); + /* Set up sdma hang watchdog */ + temp = RREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_WATCHDOG_CNTL)); + /* 100ms per unit */ + temp = REG_SET_FIELD(temp, SDMA0_WATCHDOG_CNTL, QUEUE_HANG_COUNT, + max(adev->usec_timeout/100000, 1)); + WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_WATCHDOG_CNTL), temp); + /* Set up RESP_MODE to non-copy addresses */ temp = RREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_UTCL1_CNTL)); temp = REG_SET_FIELD(temp, SDMA0_UTCL1_CNTL, RESP_MODE, 3); @@ -854,7 +861,8 @@ static int sdma_v6_0_ring_test_ring(struct amdgpu_ring *ring) r = amdgpu_ring_alloc(ring, 5); if (r) { DRM_ERROR("amdgpu: dma failed to lock ring %d (%d).\n", ring->idx, r); - amdgpu_device_wb_free(adev, index); + if (!ring->is_mes_queue) + amdgpu_device_wb_free(adev, index); return r; } @@ -1567,7 +1575,7 @@ static void sdma_v6_0_set_irq_funcs(struct amdgpu_device *adev) * @src_offset: src GPU address * @dst_offset: dst GPU address * @byte_count: number of bytes to xfer - * @tmz: if a secure copy should be used + * @copy_flags: copy flags for the buffers * * Copy GPU buffers using the DMA engine. * Used by the amdgpu ttm implementation to move pages if @@ -1577,11 +1585,11 @@ static void sdma_v6_0_emit_copy_buffer(struct amdgpu_ib *ib, uint64_t src_offset, uint64_t dst_offset, uint32_t byte_count, - bool tmz) + uint32_t copy_flags) { ib->ptr[ib->length_dw++] = SDMA_PKT_COPY_LINEAR_HEADER_OP(SDMA_OP_COPY) | SDMA_PKT_COPY_LINEAR_HEADER_SUB_OP(SDMA_SUBOP_COPY_LINEAR) | - SDMA_PKT_COPY_LINEAR_HEADER_TMZ(tmz ? 1 : 0); + SDMA_PKT_COPY_LINEAR_HEADER_TMZ((copy_flags & AMDGPU_COPY_FLAGS_TMZ) ? 1 : 0); ib->ptr[ib->length_dw++] = byte_count - 1; ib->ptr[ib->length_dw++] = 0; /* src/dst endian swap */ ib->ptr[ib->length_dw++] = lower_32_bits(src_offset); diff --git a/drivers/gpu/drm/amd/amdgpu/si.c b/drivers/gpu/drm/amd/amdgpu/si.c index 23e4ef4fff7c..85235470e872 100644 --- a/drivers/gpu/drm/amd/amdgpu/si.c +++ b/drivers/gpu/drm/amd/amdgpu/si.c @@ -1409,9 +1409,9 @@ static int si_gpu_pci_config_reset(struct amdgpu_device *adev) return r; } -static bool si_asic_supports_baco(struct amdgpu_device *adev) +static int si_asic_supports_baco(struct amdgpu_device *adev) { - return false; + return 0; } static enum amd_reset_method @@ -2706,6 +2706,8 @@ static const struct amd_ip_funcs si_common_ip_funcs = { .soft_reset = si_common_soft_reset, .set_clockgating_state = si_common_set_clockgating_state, .set_powergating_state = si_common_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_ip_block_version si_common_ip_block = diff --git a/drivers/gpu/drm/amd/amdgpu/si_dma.c b/drivers/gpu/drm/amd/amdgpu/si_dma.c index 9aa0e11ee673..11db5b755832 100644 --- a/drivers/gpu/drm/amd/amdgpu/si_dma.c +++ b/drivers/gpu/drm/amd/amdgpu/si_dma.c @@ -708,6 +708,8 @@ static const struct amd_ip_funcs si_dma_ip_funcs = { .soft_reset = si_dma_soft_reset, .set_clockgating_state = si_dma_set_clockgating_state, .set_powergating_state = si_dma_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs si_dma_ring_funcs = { @@ -761,7 +763,7 @@ static void si_dma_set_irq_funcs(struct amdgpu_device *adev) * @src_offset: src GPU address * @dst_offset: dst GPU address * @byte_count: number of bytes to xfer - * @tmz: is this a secure operation + * @copy_flags: unused * * Copy GPU buffers using the DMA engine (VI). * Used by the amdgpu ttm implementation to move pages if @@ -771,7 +773,7 @@ static void si_dma_emit_copy_buffer(struct amdgpu_ib *ib, uint64_t src_offset, uint64_t dst_offset, uint32_t byte_count, - bool tmz) + uint32_t copy_flags) { ib->ptr[ib->length_dw++] = DMA_PACKET(DMA_PACKET_COPY, 1, 0, 0, byte_count); diff --git a/drivers/gpu/drm/amd/amdgpu/si_ih.c b/drivers/gpu/drm/amd/amdgpu/si_ih.c index cada9f300a7f..5237395e4fab 100644 --- a/drivers/gpu/drm/amd/amdgpu/si_ih.c +++ b/drivers/gpu/drm/amd/amdgpu/si_ih.c @@ -296,6 +296,8 @@ static const struct amd_ip_funcs si_ih_ip_funcs = { .soft_reset = si_ih_soft_reset, .set_clockgating_state = si_ih_set_clockgating_state, .set_powergating_state = si_ih_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_ih_funcs si_ih_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/sienna_cichlid.c b/drivers/gpu/drm/amd/amdgpu/sienna_cichlid.c index 93f6772d1b24..481217c32d85 100644 --- a/drivers/gpu/drm/amd/amdgpu/sienna_cichlid.c +++ b/drivers/gpu/drm/amd/amdgpu/sienna_cichlid.c @@ -92,7 +92,7 @@ static int sienna_cichlid_mode2_suspend_ip(struct amdgpu_device *adev) adev->ip_blocks[i].status.hw = false; } - return r; + return 0; } static int diff --git a/drivers/gpu/drm/amd/amdgpu/smuio_v14_0_2.c b/drivers/gpu/drm/amd/amdgpu/smuio_v14_0_2.c new file mode 100644 index 000000000000..2a51a70d4846 --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/smuio_v14_0_2.c @@ -0,0 +1,62 @@ +/* + * Copyright 2023 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ +#include "amdgpu.h" +#include "smuio_v14_0_2.h" +#include "smuio/smuio_14_0_2_offset.h" +#include "smuio/smuio_14_0_2_sh_mask.h" +#include <linux/preempt.h> + +static u32 smuio_v14_0_2_get_rom_index_offset(struct amdgpu_device *adev) +{ + return SOC15_REG_OFFSET(SMUIO, 0, regROM_INDEX); +} + +static u32 smuio_v14_0_2_get_rom_data_offset(struct amdgpu_device *adev) +{ + return SOC15_REG_OFFSET(SMUIO, 0, regROM_DATA); +} + +static u64 smuio_v14_0_2_get_gpu_clock_counter(struct amdgpu_device *adev) +{ + u64 clock; + u64 clock_counter_lo, clock_counter_hi_pre, clock_counter_hi_after; + + preempt_disable(); + clock_counter_hi_pre = (u64)RREG32_SOC15(SMUIO, 0, regGOLDEN_TSC_COUNT_UPPER); + clock_counter_lo = (u64)RREG32_SOC15(SMUIO, 0, regGOLDEN_TSC_COUNT_LOWER); + /* the clock counter may be udpated during polling the counters */ + clock_counter_hi_after = (u64)RREG32_SOC15(SMUIO, 0, regGOLDEN_TSC_COUNT_UPPER); + if (clock_counter_hi_pre != clock_counter_hi_after) + clock_counter_lo = (u64)RREG32_SOC15(SMUIO, 0, regGOLDEN_TSC_COUNT_LOWER); + preempt_enable(); + + clock = clock_counter_lo | (clock_counter_hi_after << 32ULL); + + return clock; +} + +const struct amdgpu_smuio_funcs smuio_v14_0_2_funcs = { + .get_rom_index_offset = smuio_v14_0_2_get_rom_index_offset, + .get_rom_data_offset = smuio_v14_0_2_get_rom_data_offset, + .get_gpu_clock_counter = smuio_v14_0_2_get_gpu_clock_counter, +}; diff --git a/drivers/gpu/drm/amd/amdgpu/smuio_v14_0_2.h b/drivers/gpu/drm/amd/amdgpu/smuio_v14_0_2.h new file mode 100644 index 000000000000..6e617f832d90 --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/smuio_v14_0_2.h @@ -0,0 +1,30 @@ +/* + * Copyright 2023 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ +#ifndef __SMUIO_V14_0_2_H__ +#define __SMUIO_V14_0_2_H__ + +#include "soc15_common.h" + +extern const struct amdgpu_smuio_funcs smuio_v14_0_2_funcs; + +#endif /* __SMUIO_V14_0_2_H__ */ diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c index dec81ccf6240..170f02e96717 100644 --- a/drivers/gpu/drm/amd/amdgpu/soc15.c +++ b/drivers/gpu/drm/amd/amdgpu/soc15.c @@ -143,7 +143,7 @@ static const struct amdgpu_video_codec_info rn_video_codecs_decode_array[] = {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_MPEG4_AVC, 4096, 4096, 52)}, {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_VC1, 4096, 4096, 4)}, {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_HEVC, 8192, 4352, 186)}, - {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_JPEG, 4096, 4096, 0)}, + {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_JPEG, 16384, 16384, 0)}, {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_VP9, 8192, 4352, 0)}, }; @@ -156,7 +156,7 @@ static const struct amdgpu_video_codecs rn_video_codecs_decode = static const struct amdgpu_video_codec_info vcn_4_0_3_video_codecs_decode_array[] = { {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_MPEG4_AVC, 4096, 4096, 52)}, {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_HEVC, 8192, 4352, 186)}, - {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_JPEG, 4096, 4096, 0)}, + {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_JPEG, 16384, 16384, 0)}, {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_VP9, 8192, 4352, 0)}, {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_AV1, 8192, 4352, 0)}, }; @@ -502,7 +502,7 @@ static int soc15_asic_baco_reset(struct amdgpu_device *adev) static enum amd_reset_method soc15_asic_reset_method(struct amdgpu_device *adev) { - bool baco_reset = false; + int baco_reset = 0; bool connected_to_cpu = false; struct amdgpu_ras *ras = amdgpu_ras_get_context(adev); @@ -540,7 +540,7 @@ soc15_asic_reset_method(struct amdgpu_device *adev) */ if (ras && adev->ras_enabled && adev->pm.fw_version <= 0x283400) - baco_reset = false; + baco_reset = 0; } else { baco_reset = amdgpu_dpm_is_baco_supported(adev); } @@ -620,7 +620,7 @@ static int soc15_asic_reset(struct amdgpu_device *adev) } } -static bool soc15_supports_baco(struct amdgpu_device *adev) +static int soc15_supports_baco(struct amdgpu_device *adev) { switch (amdgpu_ip_version(adev, MP1_HWIP, 0)) { case IP_VERSION(9, 0, 0): @@ -628,13 +628,13 @@ static bool soc15_supports_baco(struct amdgpu_device *adev) if (adev->asic_type == CHIP_VEGA20) { if (adev->psp.sos.fw_version >= 0x80067) return amdgpu_dpm_is_baco_supported(adev); - return false; + return 0; } else { return amdgpu_dpm_is_baco_supported(adev); } break; default: - return false; + return 0; } } @@ -1501,4 +1501,6 @@ static const struct amd_ip_funcs soc15_common_ip_funcs = { .set_clockgating_state = soc15_common_set_clockgating_state, .set_powergating_state = soc15_common_set_powergating_state, .get_clockgating_state= soc15_common_get_clockgating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.h b/drivers/gpu/drm/amd/amdgpu/soc15.h index 1444b7765e4b..282584a48be0 100644 --- a/drivers/gpu/drm/amd/amdgpu/soc15.h +++ b/drivers/gpu/drm/amd/amdgpu/soc15.h @@ -88,6 +88,8 @@ struct soc15_ras_field_entry { }; #define SOC15_REG_ENTRY(ip, inst, reg) ip##_HWIP, inst, reg##_BASE_IDX, reg +#define SOC15_REG_ENTRY_STR(ip, inst, reg) \ + { ip##_HWIP, inst, reg##_BASE_IDX, reg, #reg } #define SOC15_REG_ENTRY_OFFSET(entry) (adev->reg_offset[entry.hwip][entry.inst][entry.seg] + entry.reg_offset) diff --git a/drivers/gpu/drm/amd/amdgpu/soc21.c b/drivers/gpu/drm/amd/amdgpu/soc21.c index 581a3bd11481..fb6797467571 100644 --- a/drivers/gpu/drm/amd/amdgpu/soc21.c +++ b/drivers/gpu/drm/amd/amdgpu/soc21.c @@ -72,7 +72,7 @@ static const struct amdgpu_video_codecs vcn_4_0_0_video_codecs_encode_vcn1 = { static const struct amdgpu_video_codec_info vcn_4_0_0_video_codecs_decode_array_vcn0[] = { {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_MPEG4_AVC, 4096, 4096, 52)}, {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_HEVC, 8192, 4352, 186)}, - {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_JPEG, 4096, 4096, 0)}, + {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_JPEG, 16384, 16384, 0)}, {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_VP9, 8192, 4352, 0)}, {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_AV1, 8192, 4352, 0)}, }; @@ -80,7 +80,7 @@ static const struct amdgpu_video_codec_info vcn_4_0_0_video_codecs_decode_array_ static const struct amdgpu_video_codec_info vcn_4_0_0_video_codecs_decode_array_vcn1[] = { {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_MPEG4_AVC, 4096, 4096, 52)}, {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_HEVC, 8192, 4352, 186)}, - {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_JPEG, 4096, 4096, 0)}, + {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_JPEG, 16384, 16384, 0)}, {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_VP9, 8192, 4352, 0)}, }; @@ -457,10 +457,8 @@ static bool soc21_need_full_reset(struct amdgpu_device *adev) { switch (amdgpu_ip_version(adev, GC_HWIP, 0)) { case IP_VERSION(11, 0, 0): - return amdgpu_ras_is_supported(adev, AMDGPU_RAS_BLOCK__UMC); case IP_VERSION(11, 0, 2): case IP_VERSION(11, 0, 3): - return false; default: return true; } @@ -722,7 +720,10 @@ static int soc21_common_early_init(void *handle) AMD_PG_SUPPORT_VCN | AMD_PG_SUPPORT_JPEG | AMD_PG_SUPPORT_GFX_PG; - adev->external_rev_id = adev->rev_id + 0x1; + if (adev->rev_id == 0) + adev->external_rev_id = 0x1; + else + adev->external_rev_id = adev->rev_id + 0x10; break; case IP_VERSION(11, 5, 1): adev->cg_flags = @@ -869,10 +870,35 @@ static int soc21_common_suspend(void *handle) return soc21_common_hw_fini(adev); } +static bool soc21_need_reset_on_resume(struct amdgpu_device *adev) +{ + u32 sol_reg1, sol_reg2; + + /* Will reset for the following suspend abort cases. + * 1) Only reset dGPU side. + * 2) S3 suspend got aborted and TOS is active. + */ + if (!(adev->flags & AMD_IS_APU) && adev->in_s3 && + !adev->suspend_complete) { + sol_reg1 = RREG32_SOC15(MP0, 0, regMP0_SMN_C2PMSG_81); + msleep(100); + sol_reg2 = RREG32_SOC15(MP0, 0, regMP0_SMN_C2PMSG_81); + + return (sol_reg1 != sol_reg2); + } + + return false; +} + static int soc21_common_resume(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; + if (soc21_need_reset_on_resume(adev)) { + dev_info(adev->dev, "S3 suspend aborted, resetting..."); + soc21_asic_reset(adev); + } + return soc21_common_hw_init(adev); } @@ -959,4 +985,6 @@ static const struct amd_ip_funcs soc21_common_ip_funcs = { .set_clockgating_state = soc21_common_set_clockgating_state, .set_powergating_state = soc21_common_set_powergating_state, .get_clockgating_state = soc21_common_get_clockgating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; diff --git a/drivers/gpu/drm/amd/amdgpu/ta_ras_if.h b/drivers/gpu/drm/amd/amdgpu/ta_ras_if.h index 056d4df8fa1f..3ac56a9645eb 100644 --- a/drivers/gpu/drm/amd/amdgpu/ta_ras_if.h +++ b/drivers/gpu/drm/amd/amdgpu/ta_ras_if.h @@ -146,6 +146,7 @@ struct ta_ras_mca_addr { uint32_t ch_inst; uint32_t umc_inst; uint32_t node_inst; + uint32_t socket_id; }; struct ta_ras_phy_addr { diff --git a/drivers/gpu/drm/amd/amdgpu/tonga_ih.c b/drivers/gpu/drm/amd/amdgpu/tonga_ih.c index 450b6e831509..24d49d813607 100644 --- a/drivers/gpu/drm/amd/amdgpu/tonga_ih.c +++ b/drivers/gpu/drm/amd/amdgpu/tonga_ih.c @@ -486,6 +486,8 @@ static const struct amd_ip_funcs tonga_ih_ip_funcs = { .post_soft_reset = tonga_ih_post_soft_reset, .set_clockgating_state = tonga_ih_set_clockgating_state, .set_powergating_state = tonga_ih_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_ih_funcs tonga_ih_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/umc_v12_0.c b/drivers/gpu/drm/amd/amdgpu/umc_v12_0.c index 77af4e25ff46..bfe61d86ee6c 100644 --- a/drivers/gpu/drm/amd/amdgpu/umc_v12_0.c +++ b/drivers/gpu/drm/amd/amdgpu/umc_v12_0.c @@ -28,27 +28,7 @@ #include "umc/umc_12_0_0_sh_mask.h" #include "mp/mp_13_0_6_sh_mask.h" -const uint32_t - umc_v12_0_channel_idx_tbl[] - [UMC_V12_0_UMC_INSTANCE_NUM] - [UMC_V12_0_CHANNEL_INSTANCE_NUM] = { - {{3, 7, 11, 15, 2, 6, 10, 14}, {1, 5, 9, 13, 0, 4, 8, 12}, - {19, 23, 27, 31, 18, 22, 26, 30}, {17, 21, 25, 29, 16, 20, 24, 28}}, - {{47, 43, 39, 35, 46, 42, 38, 34}, {45, 41, 37, 33, 44, 40, 36, 32}, - {63, 59, 55, 51, 62, 58, 54, 50}, {61, 57, 53, 49, 60, 56, 52, 48}}, - {{79, 75, 71, 67, 78, 74, 70, 66}, {77, 73, 69, 65, 76, 72, 68, 64}, - {95, 91, 87, 83, 94, 90, 86, 82}, {93, 89, 85, 81, 92, 88, 84, 80}}, - {{99, 103, 107, 111, 98, 102, 106, 110}, {97, 101, 105, 109, 96, 100, 104, 108}, - {115, 119, 123, 127, 114, 118, 122, 126}, {113, 117, 121, 125, 112, 116, 120, 124}} - }; - -/* mapping of MCA error address to normalized address */ -static const uint32_t umc_v12_0_ma2na_mapping[] = { - 0, 5, 6, 8, 9, 14, 12, 13, - 10, 11, 15, 16, 17, 18, 19, 20, - 21, 22, 23, 24, 25, 26, 27, 28, - 24, 7, 29, 30, -}; +#define MAX_ECC_NUM_PER_RETIREMENT 32 static inline uint64_t get_umc_v12_0_reg_offset(struct amdgpu_device *adev, uint32_t node_inst, @@ -192,99 +172,74 @@ static void umc_v12_0_query_ras_error_count(struct amdgpu_device *adev, umc_v12_0_reset_error_count(adev); } -static bool umc_v12_0_bit_wise_xor(uint32_t val) +static void umc_v12_0_convert_error_address(struct amdgpu_device *adev, + struct ras_err_data *err_data, + struct ta_ras_query_address_input *addr_in) { - bool result = 0; - int i; + uint32_t col, row, row_xor, bank, channel_index; + uint64_t soc_pa, retired_page, column, err_addr; + struct ta_ras_query_address_output addr_out; - for (i = 0; i < 32; i++) - result = result ^ ((val >> i) & 0x1); + err_addr = addr_in->ma.err_addr; + addr_in->addr_type = TA_RAS_MCA_TO_PA; + if (psp_ras_query_address(&adev->psp, addr_in, &addr_out)) { + dev_warn(adev->dev, "Failed to query RAS physical address for 0x%llx", + err_addr); - return result; -} + return; + } + + soc_pa = addr_out.pa.pa; + bank = addr_out.pa.bank; + channel_index = addr_out.pa.channel_idx; -static void umc_v12_0_mca_addr_to_pa(struct amdgpu_device *adev, - uint64_t err_addr, uint32_t ch_inst, uint32_t umc_inst, - uint32_t node_inst, - struct ta_ras_query_address_output *addr_out) -{ - uint32_t channel_index, i; - uint64_t na, soc_pa; - uint32_t bank_hash0, bank_hash1, bank_hash2, bank_hash3, col, row; - uint32_t bank0, bank1, bank2, bank3, bank; - - bank_hash0 = (err_addr >> UMC_V12_0_MCA_B0_BIT) & 0x1ULL; - bank_hash1 = (err_addr >> UMC_V12_0_MCA_B1_BIT) & 0x1ULL; - bank_hash2 = (err_addr >> UMC_V12_0_MCA_B2_BIT) & 0x1ULL; - bank_hash3 = (err_addr >> UMC_V12_0_MCA_B3_BIT) & 0x1ULL; col = (err_addr >> 1) & 0x1fULL; row = (err_addr >> 10) & 0x3fffULL; + row_xor = row ^ (0x1ULL << 13); + /* clear [C3 C2] in soc physical address */ + soc_pa &= ~(0x3ULL << UMC_V12_0_PA_C2_BIT); + /* clear [C4] in soc physical address */ + soc_pa &= ~(0x1ULL << UMC_V12_0_PA_C4_BIT); + + /* loop for all possibilities of [C4 C3 C2] */ + for (column = 0; column < UMC_V12_0_NA_MAP_PA_NUM; column++) { + retired_page = soc_pa | ((column & 0x3) << UMC_V12_0_PA_C2_BIT); + retired_page |= (((column & 0x4) >> 2) << UMC_V12_0_PA_C4_BIT); + /* include column bit 0 and 1 */ + col &= 0x3; + col |= (column << 2); + dev_info(adev->dev, + "Error Address(PA):0x%-10llx Row:0x%-4x Col:0x%-2x Bank:0x%x Channel:0x%x\n", + retired_page, row, col, bank, channel_index); + amdgpu_umc_fill_error_record(err_data, err_addr, + retired_page, channel_index, addr_in->ma.umc_inst); - /* apply bank hash algorithm */ - bank0 = - bank_hash0 ^ (UMC_V12_0_XOR_EN0 & - (umc_v12_0_bit_wise_xor(col & UMC_V12_0_COL_XOR0) ^ - (umc_v12_0_bit_wise_xor(row & UMC_V12_0_ROW_XOR0)))); - bank1 = - bank_hash1 ^ (UMC_V12_0_XOR_EN1 & - (umc_v12_0_bit_wise_xor(col & UMC_V12_0_COL_XOR1) ^ - (umc_v12_0_bit_wise_xor(row & UMC_V12_0_ROW_XOR1)))); - bank2 = - bank_hash2 ^ (UMC_V12_0_XOR_EN2 & - (umc_v12_0_bit_wise_xor(col & UMC_V12_0_COL_XOR2) ^ - (umc_v12_0_bit_wise_xor(row & UMC_V12_0_ROW_XOR2)))); - bank3 = - bank_hash3 ^ (UMC_V12_0_XOR_EN3 & - (umc_v12_0_bit_wise_xor(col & UMC_V12_0_COL_XOR3) ^ - (umc_v12_0_bit_wise_xor(row & UMC_V12_0_ROW_XOR3)))); - - bank = bank0 | (bank1 << 1) | (bank2 << 2) | (bank3 << 3); - err_addr &= ~0x3c0ULL; - err_addr |= (bank << UMC_V12_0_MCA_B0_BIT); - - na = 0x0; - /* convert mca error address to normalized address */ - for (i = 1; i < ARRAY_SIZE(umc_v12_0_ma2na_mapping); i++) - na |= ((err_addr >> i) & 0x1ULL) << umc_v12_0_ma2na_mapping[i]; - - channel_index = - adev->umc.channel_idx_tbl[node_inst * adev->umc.umc_inst_num * - adev->umc.channel_inst_num + - umc_inst * adev->umc.channel_inst_num + - ch_inst]; - /* translate umc channel address to soc pa, 3 parts are included */ - soc_pa = ADDR_OF_32KB_BLOCK(na) | - ADDR_OF_256B_BLOCK(channel_index) | - OFFSET_IN_256B_BLOCK(na); - - /* the umc channel bits are not original values, they are hashed */ - UMC_V12_0_SET_CHANNEL_HASH(channel_index, soc_pa); - - addr_out->pa.pa = soc_pa; - addr_out->pa.bank = bank; - addr_out->pa.channel_idx = channel_index; + /* shift R13 bit */ + retired_page ^= (0x1ULL << UMC_V12_0_PA_R13_BIT); + dev_info(adev->dev, + "Error Address(PA):0x%-10llx Row:0x%-4x Col:0x%-2x Bank:0x%x Channel:0x%x\n", + retired_page, row_xor, col, bank, channel_index); + amdgpu_umc_fill_error_record(err_data, err_addr, + retired_page, channel_index, addr_in->ma.umc_inst); + } } -static void umc_v12_0_convert_error_address(struct amdgpu_device *adev, - struct ras_err_data *err_data, uint64_t err_addr, - uint32_t ch_inst, uint32_t umc_inst, - uint32_t node_inst) +static int umc_v12_0_convert_err_addr(struct amdgpu_device *adev, + struct ta_ras_query_address_input *addr_in, + uint64_t *pfns, int len) { uint32_t col, row, row_xor, bank, channel_index; - uint64_t soc_pa, retired_page, column; - struct ta_ras_query_address_input addr_in; + uint64_t soc_pa, retired_page, column, err_addr; struct ta_ras_query_address_output addr_out; + uint32_t pos = 0; - addr_in.addr_type = TA_RAS_MCA_TO_PA; - addr_in.ma.err_addr = err_addr; - addr_in.ma.ch_inst = ch_inst; - addr_in.ma.umc_inst = umc_inst; - addr_in.ma.node_inst = node_inst; - - if (psp_ras_query_address(&adev->psp, &addr_in, &addr_out)) - /* fallback to old path if fail to get pa from psp */ - umc_v12_0_mca_addr_to_pa(adev, err_addr, ch_inst, umc_inst, - node_inst, &addr_out); + err_addr = addr_in->ma.err_addr; + addr_in->addr_type = TA_RAS_MCA_TO_PA; + if (psp_ras_query_address(&adev->psp, addr_in, &addr_out)) { + dev_warn(adev->dev, "Failed to query RAS physical address for 0x%llx", + err_addr); + return 0; + } soc_pa = addr_out.pa.pa; bank = addr_out.pa.bank; @@ -302,33 +257,42 @@ static void umc_v12_0_convert_error_address(struct amdgpu_device *adev, for (column = 0; column < UMC_V12_0_NA_MAP_PA_NUM; column++) { retired_page = soc_pa | ((column & 0x3) << UMC_V12_0_PA_C2_BIT); retired_page |= (((column & 0x4) >> 2) << UMC_V12_0_PA_C4_BIT); + + if (pos >= len) + return 0; + pfns[pos++] = retired_page >> AMDGPU_GPU_PAGE_SHIFT; + /* include column bit 0 and 1 */ col &= 0x3; col |= (column << 2); dev_info(adev->dev, "Error Address(PA):0x%-10llx Row:0x%-4x Col:0x%-2x Bank:0x%x Channel:0x%x\n", retired_page, row, col, bank, channel_index); - amdgpu_umc_fill_error_record(err_data, err_addr, - retired_page, channel_index, umc_inst); /* shift R13 bit */ retired_page ^= (0x1ULL << UMC_V12_0_PA_R13_BIT); + + if (pos >= len) + return 0; + pfns[pos++] = retired_page >> AMDGPU_GPU_PAGE_SHIFT; + dev_info(adev->dev, "Error Address(PA):0x%-10llx Row:0x%-4x Col:0x%-2x Bank:0x%x Channel:0x%x\n", retired_page, row_xor, col, bank, channel_index); - amdgpu_umc_fill_error_record(err_data, err_addr, - retired_page, channel_index, umc_inst); } + + return pos; } static int umc_v12_0_query_error_address(struct amdgpu_device *adev, uint32_t node_inst, uint32_t umc_inst, uint32_t ch_inst, void *data) { + struct ras_err_data *err_data = (struct ras_err_data *)data; + struct ta_ras_query_address_input addr_in; uint64_t mc_umc_status_addr; uint64_t mc_umc_status, err_addr; uint64_t mc_umc_addrt0; - struct ras_err_data *err_data = (struct ras_err_data *)data; uint64_t umc_reg_offset = get_umc_v12_0_reg_offset(adev, node_inst, umc_inst, ch_inst); @@ -357,8 +321,19 @@ static int umc_v12_0_query_error_address(struct amdgpu_device *adev, err_addr = REG_GET_FIELD(err_addr, MCA_UMC_UMC0_MCUMC_ADDRT0, ErrorAddr); - umc_v12_0_convert_error_address(adev, err_data, err_addr, - ch_inst, umc_inst, node_inst); + if (!adev->aid_mask && + adev->smuio.funcs && + adev->smuio.funcs->get_socket_id) + addr_in.ma.socket_id = adev->smuio.funcs->get_socket_id(adev); + else + addr_in.ma.socket_id = 0; + + addr_in.ma.err_addr = err_addr; + addr_in.ma.ch_inst = ch_inst; + addr_in.ma.umc_inst = umc_inst; + addr_in.ma.node_inst = node_inst; + + umc_v12_0_convert_error_address(adev, err_data, &addr_in); } /* clear umc status */ @@ -401,13 +376,20 @@ static int umc_v12_0_err_cnt_init_per_channel(struct amdgpu_device *adev, return 0; } +#ifdef TO_BE_REMOVED static void umc_v12_0_ecc_info_query_ras_error_count(struct amdgpu_device *adev, void *ras_error_status) { + struct ras_query_context qctx; + + memset(&qctx, 0, sizeof(qctx)); + qctx.event_id = amdgpu_ras_acquire_event_id(adev, amdgpu_ras_intr_triggered() ? + RAS_EVENT_TYPE_ISR : RAS_EVENT_TYPE_INVALID); + amdgpu_mca_smu_log_ras_error(adev, - AMDGPU_RAS_BLOCK__UMC, AMDGPU_MCA_ERROR_TYPE_CE, ras_error_status); + AMDGPU_RAS_BLOCK__UMC, AMDGPU_MCA_ERROR_TYPE_CE, ras_error_status, &qctx); amdgpu_mca_smu_log_ras_error(adev, - AMDGPU_RAS_BLOCK__UMC, AMDGPU_MCA_ERROR_TYPE_UE, ras_error_status); + AMDGPU_RAS_BLOCK__UMC, AMDGPU_MCA_ERROR_TYPE_UE, ras_error_status, &qctx); } static void umc_v12_0_ecc_info_query_ras_error_address(struct amdgpu_device *adev, @@ -418,12 +400,16 @@ static void umc_v12_0_ecc_info_query_ras_error_address(struct amdgpu_device *ade struct ras_err_info *err_info; struct ras_err_addr *mca_err_addr, *tmp; struct ras_err_data *err_data = (struct ras_err_data *)ras_error_status; + struct ta_ras_query_address_input addr_in; for_each_ras_error(err_node, err_data) { err_info = &err_node->err_info; if (list_empty(&err_info->err_addr_list)) continue; + addr_in.ma.node_inst = err_info->mcm_info.die_id; + addr_in.ma.socket_id = err_info->mcm_info.socket_id; + list_for_each_entry_safe(mca_err_addr, tmp, &err_info->err_addr_list, node) { mc_umc_status = mca_err_addr->err_status; if (mc_umc_status && @@ -439,6 +425,10 @@ static void umc_v12_0_ecc_info_query_ras_error_address(struct amdgpu_device *ade MCA_UMC_UMC0_MCUMC_ADDRT0, ErrorAddr); InstanceIdLo = REG_GET_FIELD(mca_ipid, MCMP1_IPIDT0, InstanceIdLo); + addr_in.ma.err_addr = err_addr; + addr_in.ma.ch_inst = MCA_IPID_LO_2_UMC_CH(InstanceIdLo); + addr_in.ma.umc_inst = MCA_IPID_LO_2_UMC_INST(InstanceIdLo); + dev_info(adev->dev, "UMC:IPID:0x%llx, aid:%d, inst:%d, ch:%d, err_addr:0x%llx\n", mca_ipid, err_info->mcm_info.die_id, @@ -447,10 +437,7 @@ static void umc_v12_0_ecc_info_query_ras_error_address(struct amdgpu_device *ade err_addr); umc_v12_0_convert_error_address(adev, - err_data, err_addr, - MCA_IPID_LO_2_UMC_CH(InstanceIdLo), - MCA_IPID_LO_2_UMC_INST(InstanceIdLo), - err_info->mcm_info.die_id); + err_data, &addr_in); } /* Delete error address node from list and free memory */ @@ -458,6 +445,7 @@ static void umc_v12_0_ecc_info_query_ras_error_address(struct amdgpu_device *ade } } } +#endif static bool umc_v12_0_check_ecc_err_status(struct amdgpu_device *adev, enum amdgpu_mca_error_type type, void *ras_error_status) @@ -498,43 +486,49 @@ const struct amdgpu_ras_block_hw_ops umc_v12_0_ras_hw_ops = { .query_ras_error_address = umc_v12_0_query_ras_error_address, }; -static int umc_v12_0_aca_bank_generate_report(struct aca_handle *handle, struct aca_bank *bank, enum aca_error_type type, - struct aca_bank_report *report, void *data) +static int umc_v12_0_aca_bank_parser(struct aca_handle *handle, struct aca_bank *bank, + enum aca_smu_type type, void *data) { struct amdgpu_device *adev = handle->adev; - u64 status; + struct aca_bank_info info; + enum aca_error_type err_type; + u64 status, count; + u32 ext_error_code; int ret; - ret = aca_bank_info_decode(bank, &report->info); + status = bank->regs[ACA_REG_IDX_STATUS]; + if (umc_v12_0_is_deferred_error(adev, status)) + err_type = ACA_ERROR_TYPE_DEFERRED; + else if (umc_v12_0_is_uncorrectable_error(adev, status)) + err_type = ACA_ERROR_TYPE_UE; + else if (umc_v12_0_is_correctable_error(adev, status)) + err_type = ACA_ERROR_TYPE_CE; + else + return 0; + + ret = aca_bank_info_decode(bank, &info); if (ret) return ret; - status = bank->regs[ACA_REG_IDX_STATUS]; - switch (type) { - case ACA_ERROR_TYPE_UE: - if (umc_v12_0_is_uncorrectable_error(adev, status)) { - report->count[type] = 1; - } - break; - case ACA_ERROR_TYPE_CE: - if (umc_v12_0_is_correctable_error(adev, status)) { - report->count[type] = 1; - } - break; - default: - return -EINVAL; - } + amdgpu_umc_update_ecc_status(adev, + bank->regs[ACA_REG_IDX_STATUS], + bank->regs[ACA_REG_IDX_IPID], + bank->regs[ACA_REG_IDX_ADDR]); - return 0; + ext_error_code = ACA_REG__STATUS__ERRORCODEEXT(status); + count = ext_error_code == 0 ? + ACA_REG__MISC0__ERRCNT(bank->regs[ACA_REG_IDX_MISC0]) : 1ULL; + + return aca_error_cache_log_bank_error(handle, &info, err_type, count); } static const struct aca_bank_ops umc_v12_0_aca_bank_ops = { - .aca_bank_generate_report = umc_v12_0_aca_bank_generate_report, + .aca_bank_parser = umc_v12_0_aca_bank_parser, }; const struct aca_info umc_v12_0_aca_info = { .hwip = ACA_HWIP_TYPE_UMC, - .mask = ACA_ERROR_UE_MASK | ACA_ERROR_CE_MASK, + .mask = ACA_ERROR_UE_MASK | ACA_ERROR_CE_MASK | ACA_ERROR_DEFERRED_MASK, .bank_ops = &umc_v12_0_aca_bank_ops, }; @@ -554,6 +548,152 @@ static int umc_v12_0_ras_late_init(struct amdgpu_device *adev, struct ras_common return 0; } +static int umc_v12_0_update_ecc_status(struct amdgpu_device *adev, + uint64_t status, uint64_t ipid, uint64_t addr) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + uint16_t hwid, mcatype; + struct ta_ras_query_address_input addr_in; + uint64_t page_pfn[UMC_V12_0_BAD_PAGE_NUM_PER_CHANNEL]; + uint64_t err_addr, hash_val = 0; + struct ras_ecc_err *ecc_err; + int count; + int ret; + + hwid = REG_GET_FIELD(ipid, MCMP1_IPIDT0, HardwareID); + mcatype = REG_GET_FIELD(ipid, MCMP1_IPIDT0, McaType); + + if ((hwid != MCA_UMC_HWID_V12_0) || (mcatype != MCA_UMC_MCATYPE_V12_0)) + return 0; + + if (!status) + return 0; + + if (!umc_v12_0_is_deferred_error(adev, status)) + return 0; + + err_addr = REG_GET_FIELD(addr, + MCA_UMC_UMC0_MCUMC_ADDRT0, ErrorAddr); + + dev_info(adev->dev, + "UMC:IPID:0x%llx, socket:%llu, aid:%llu, inst:%llu, ch:%llu, err_addr:0x%llx\n", + ipid, + MCA_IPID_2_SOCKET_ID(ipid), + MCA_IPID_2_DIE_ID(ipid), + MCA_IPID_2_UMC_INST(ipid), + MCA_IPID_2_UMC_CH(ipid), + err_addr); + + memset(page_pfn, 0, sizeof(page_pfn)); + + memset(&addr_in, 0, sizeof(addr_in)); + addr_in.ma.err_addr = err_addr; + addr_in.ma.ch_inst = MCA_IPID_2_UMC_CH(ipid); + addr_in.ma.umc_inst = MCA_IPID_2_UMC_INST(ipid); + addr_in.ma.node_inst = MCA_IPID_2_DIE_ID(ipid); + addr_in.ma.socket_id = MCA_IPID_2_SOCKET_ID(ipid); + + count = umc_v12_0_convert_err_addr(adev, + &addr_in, page_pfn, ARRAY_SIZE(page_pfn)); + if (count <= 0) { + dev_warn(adev->dev, "Fail to convert error address! count:%d\n", count); + return 0; + } + + ret = amdgpu_umc_build_pages_hash(adev, + page_pfn, count, &hash_val); + if (ret) { + dev_err(adev->dev, "Fail to build error pages hash\n"); + return ret; + } + + ecc_err = kzalloc(sizeof(*ecc_err), GFP_KERNEL); + if (!ecc_err) + return -ENOMEM; + + ecc_err->err_pages.pfn = kcalloc(count, sizeof(*ecc_err->err_pages.pfn), GFP_KERNEL); + if (!ecc_err->err_pages.pfn) { + kfree(ecc_err); + return -ENOMEM; + } + + memcpy(ecc_err->err_pages.pfn, page_pfn, count * sizeof(*ecc_err->err_pages.pfn)); + ecc_err->err_pages.count = count; + + ecc_err->hash_index = hash_val; + ecc_err->status = status; + ecc_err->ipid = ipid; + ecc_err->addr = addr; + + ret = amdgpu_umc_logs_ecc_err(adev, &con->umc_ecc_log.de_page_tree, ecc_err); + if (ret) { + if (ret == -EEXIST) + con->umc_ecc_log.de_updated = true; + else + dev_err(adev->dev, "Fail to log ecc error! ret:%d\n", ret); + + kfree(ecc_err->err_pages.pfn); + kfree(ecc_err); + return ret; + } + + con->umc_ecc_log.de_updated = true; + + return 0; +} + +static int umc_v12_0_fill_error_record(struct amdgpu_device *adev, + struct ras_ecc_err *ecc_err, void *ras_error_status) +{ + struct ras_err_data *err_data = (struct ras_err_data *)ras_error_status; + uint32_t i = 0; + int ret = 0; + + if (!err_data || !ecc_err) + return -EINVAL; + + for (i = 0; i < ecc_err->err_pages.count; i++) { + ret = amdgpu_umc_fill_error_record(err_data, + ecc_err->addr, + ecc_err->err_pages.pfn[i] << AMDGPU_GPU_PAGE_SHIFT, + MCA_IPID_2_UMC_CH(ecc_err->ipid), + MCA_IPID_2_UMC_INST(ecc_err->ipid)); + if (ret) + break; + } + + err_data->de_count++; + + return ret; +} + +static void umc_v12_0_query_ras_ecc_err_addr(struct amdgpu_device *adev, + void *ras_error_status) +{ + struct amdgpu_ras *con = amdgpu_ras_get_context(adev); + struct ras_ecc_err *entries[MAX_ECC_NUM_PER_RETIREMENT]; + struct radix_tree_root *ecc_tree; + int new_detected, ret, i; + + ecc_tree = &con->umc_ecc_log.de_page_tree; + + mutex_lock(&con->umc_ecc_log.lock); + new_detected = radix_tree_gang_lookup_tag(ecc_tree, (void **)entries, + 0, ARRAY_SIZE(entries), UMC_ECC_NEW_DETECTED_TAG); + for (i = 0; i < new_detected; i++) { + if (!entries[i]) + continue; + + ret = umc_v12_0_fill_error_record(adev, entries[i], ras_error_status); + if (ret) { + dev_err(adev->dev, "Fail to fill umc error record, ret:%d\n", ret); + break; + } + radix_tree_tag_clear(ecc_tree, entries[i]->hash_index, UMC_ECC_NEW_DETECTED_TAG); + } + mutex_unlock(&con->umc_ecc_log.lock); +} + struct amdgpu_umc_ras umc_v12_0_ras = { .ras_block = { .hw_ops = &umc_v12_0_ras_hw_ops, @@ -561,8 +701,8 @@ struct amdgpu_umc_ras umc_v12_0_ras = { }, .err_cnt_init = umc_v12_0_err_cnt_init, .query_ras_poison_mode = umc_v12_0_query_ras_poison_mode, - .ecc_info_query_ras_error_count = umc_v12_0_ecc_info_query_ras_error_count, - .ecc_info_query_ras_error_address = umc_v12_0_ecc_info_query_ras_error_address, + .ecc_info_query_ras_error_address = umc_v12_0_query_ras_ecc_err_addr, .check_ecc_err_status = umc_v12_0_check_ecc_err_status, + .update_ecc_status = umc_v12_0_update_ecc_status, }; diff --git a/drivers/gpu/drm/amd/amdgpu/umc_v12_0.h b/drivers/gpu/drm/amd/amdgpu/umc_v12_0.h index 5973bfb14fce..b4974793850b 100644 --- a/drivers/gpu/drm/amd/amdgpu/umc_v12_0.h +++ b/drivers/gpu/drm/amd/amdgpu/umc_v12_0.h @@ -55,83 +55,38 @@ #define UMC_V12_0_NA_MAP_PA_NUM 8 /* R13 bit shift should be considered, double the number */ #define UMC_V12_0_BAD_PAGE_NUM_PER_CHANNEL (UMC_V12_0_NA_MAP_PA_NUM * 2) -/* bank bits in MCA error address */ -#define UMC_V12_0_MCA_B0_BIT 6 -#define UMC_V12_0_MCA_B1_BIT 7 -#define UMC_V12_0_MCA_B2_BIT 8 -#define UMC_V12_0_MCA_B3_BIT 9 + /* column bits in SOC physical address */ #define UMC_V12_0_PA_C2_BIT 15 #define UMC_V12_0_PA_C4_BIT 21 /* row bits in SOC physical address */ #define UMC_V12_0_PA_R13_BIT 35 -/* channel index bits in SOC physical address */ -#define UMC_V12_0_PA_CH4_BIT 12 -#define UMC_V12_0_PA_CH5_BIT 13 -#define UMC_V12_0_PA_CH6_BIT 14 - -/* bank hash settings */ -#define UMC_V12_0_XOR_EN0 1 -#define UMC_V12_0_XOR_EN1 1 -#define UMC_V12_0_XOR_EN2 1 -#define UMC_V12_0_XOR_EN3 1 -#define UMC_V12_0_COL_XOR0 0x0 -#define UMC_V12_0_COL_XOR1 0x0 -#define UMC_V12_0_COL_XOR2 0x800 -#define UMC_V12_0_COL_XOR3 0x1000 -#define UMC_V12_0_ROW_XOR0 0x11111 -#define UMC_V12_0_ROW_XOR1 0x22222 -#define UMC_V12_0_ROW_XOR2 0x4444 -#define UMC_V12_0_ROW_XOR3 0x8888 - -/* channel hash settings */ -#define UMC_V12_0_HASH_4K 0 -#define UMC_V12_0_HASH_64K 1 -#define UMC_V12_0_HASH_2M 1 -#define UMC_V12_0_HASH_1G 1 -#define UMC_V12_0_HASH_1T 1 - -/* XOR some bits of PA into CH4~CH6 bits (bits 12~14 of PA), - * hash bit is only effective when related setting is enabled - */ -#define UMC_V12_0_CHANNEL_HASH_CH4(channel_idx, pa) ((((channel_idx) >> 5) & 0x1) ^ \ - (((pa) >> 20) & 0x1ULL & UMC_V12_0_HASH_64K) ^ \ - (((pa) >> 27) & 0x1ULL & UMC_V12_0_HASH_2M) ^ \ - (((pa) >> 34) & 0x1ULL & UMC_V12_0_HASH_1G) ^ \ - (((pa) >> 41) & 0x1ULL & UMC_V12_0_HASH_1T)) -#define UMC_V12_0_CHANNEL_HASH_CH5(channel_idx, pa) ((((channel_idx) >> 6) & 0x1) ^ \ - (((pa) >> 21) & 0x1ULL & UMC_V12_0_HASH_64K) ^ \ - (((pa) >> 28) & 0x1ULL & UMC_V12_0_HASH_2M) ^ \ - (((pa) >> 35) & 0x1ULL & UMC_V12_0_HASH_1G) ^ \ - (((pa) >> 42) & 0x1ULL & UMC_V12_0_HASH_1T)) -#define UMC_V12_0_CHANNEL_HASH_CH6(channel_idx, pa) ((((channel_idx) >> 4) & 0x1) ^ \ - (((pa) >> 19) & 0x1ULL & UMC_V12_0_HASH_64K) ^ \ - (((pa) >> 26) & 0x1ULL & UMC_V12_0_HASH_2M) ^ \ - (((pa) >> 33) & 0x1ULL & UMC_V12_0_HASH_1G) ^ \ - (((pa) >> 40) & 0x1ULL & UMC_V12_0_HASH_1T) ^ \ - (((pa) >> 47) & 0x1ULL & UMC_V12_0_HASH_4K)) -#define UMC_V12_0_SET_CHANNEL_HASH(channel_idx, pa) do { \ - (pa) &= ~(0x7ULL << UMC_V12_0_PA_CH4_BIT); \ - (pa) |= (UMC_V12_0_CHANNEL_HASH_CH4(channel_idx, pa) << UMC_V12_0_PA_CH4_BIT); \ - (pa) |= (UMC_V12_0_CHANNEL_HASH_CH5(channel_idx, pa) << UMC_V12_0_PA_CH5_BIT); \ - (pa) |= (UMC_V12_0_CHANNEL_HASH_CH6(channel_idx, pa) << UMC_V12_0_PA_CH6_BIT); \ - } while (0) + +#define MCA_UMC_HWID_V12_0 0x96 +#define MCA_UMC_MCATYPE_V12_0 0x0 #define MCA_IPID_LO_2_UMC_CH(_ipid_lo) (((((_ipid_lo) >> 20) & 0x1) * 4) + \ (((_ipid_lo) >> 12) & 0xF)) #define MCA_IPID_LO_2_UMC_INST(_ipid_lo) (((_ipid_lo) >> 21) & 0x7) +#define MCA_IPID_2_DIE_ID(ipid) ((REG_GET_FIELD(ipid, MCMP1_IPIDT0, InstanceIdHi) >> 2) & 0x03) + +#define MCA_IPID_2_UMC_CH(ipid) \ + (MCA_IPID_LO_2_UMC_CH(REG_GET_FIELD(ipid, MCMP1_IPIDT0, InstanceIdLo))) + +#define MCA_IPID_2_UMC_INST(ipid) \ + (MCA_IPID_LO_2_UMC_INST(REG_GET_FIELD(ipid, MCMP1_IPIDT0, InstanceIdLo))) + +#define MCA_IPID_2_SOCKET_ID(ipid) \ + (((REG_GET_FIELD(ipid, MCMP1_IPIDT0, InstanceIdLo) & 0x1) << 2) | \ + (REG_GET_FIELD(ipid, MCMP1_IPIDT0, InstanceIdHi) & 0x03)) + bool umc_v12_0_is_deferred_error(struct amdgpu_device *adev, uint64_t mc_umc_status); bool umc_v12_0_is_uncorrectable_error(struct amdgpu_device *adev, uint64_t mc_umc_status); bool umc_v12_0_is_correctable_error(struct amdgpu_device *adev, uint64_t mc_umc_status); typedef bool (*check_error_type_func)(struct amdgpu_device *adev, uint64_t mc_umc_status); -extern const uint32_t - umc_v12_0_channel_idx_tbl[] - [UMC_V12_0_UMC_INSTANCE_NUM] - [UMC_V12_0_CHANNEL_INSTANCE_NUM]; - extern struct amdgpu_umc_ras umc_v12_0_ras; #endif diff --git a/drivers/gpu/drm/amd/amdgpu/umc_v8_10.c b/drivers/gpu/drm/amd/amdgpu/umc_v8_10.c index c4c77257710c..a32f87992f20 100644 --- a/drivers/gpu/drm/amd/amdgpu/umc_v8_10.c +++ b/drivers/gpu/drm/amd/amdgpu/umc_v8_10.c @@ -442,11 +442,6 @@ static void umc_v8_10_ecc_info_query_ras_error_address(struct amdgpu_device *ade umc_v8_10_ecc_info_query_error_address, ras_error_status); } -static void umc_v8_10_set_eeprom_table_version(struct amdgpu_ras_eeprom_table_header *hdr) -{ - hdr->version = RAS_TABLE_VER_V2_1; -} - const struct amdgpu_ras_block_hw_ops umc_v8_10_ras_hw_ops = { .query_ras_error_count = umc_v8_10_query_ras_error_count, .query_ras_error_address = umc_v8_10_query_ras_error_address, @@ -460,5 +455,4 @@ struct amdgpu_umc_ras umc_v8_10_ras = { .query_ras_poison_mode = umc_v8_10_query_ras_poison_mode, .ecc_info_query_ras_error_count = umc_v8_10_ecc_info_query_ras_error_count, .ecc_info_query_ras_error_address = umc_v8_10_ecc_info_query_ras_error_address, - .set_eeprom_table_version = umc_v8_10_set_eeprom_table_version, }; diff --git a/drivers/gpu/drm/amd/amdgpu/umsch_mm_v4_0.c b/drivers/gpu/drm/amd/amdgpu/umsch_mm_v4_0.c index 8e7b763cfdb7..bd57896ab85d 100644 --- a/drivers/gpu/drm/amd/amdgpu/umsch_mm_v4_0.c +++ b/drivers/gpu/drm/amd/amdgpu/umsch_mm_v4_0.c @@ -60,7 +60,7 @@ static int umsch_mm_v4_0_load_microcode(struct amdgpu_umsch_mm *umsch) umsch->cmd_buf_curr_ptr = umsch->cmd_buf_ptr; - if (amdgpu_ip_version(adev, VCN_HWIP, 0) == IP_VERSION(4, 0, 5)) { + if (amdgpu_ip_version(adev, VCN_HWIP, 0) >= IP_VERSION(4, 0, 5)) { WREG32_SOC15(VCN, 0, regUVD_IPX_DLDO_CONFIG, 1 << UVD_IPX_DLDO_CONFIG__ONO0_PWR_CONFIG__SHIFT); SOC15_WAIT_ON_RREG(VCN, 0, regUVD_IPX_DLDO_STATUS, @@ -225,6 +225,8 @@ static int umsch_mm_v4_0_ring_start(struct amdgpu_umsch_mm *umsch) WREG32_SOC15(VCN, 0, regVCN_UMSCH_RB_SIZE, ring->ring_size); + ring->wptr = 0; + data = RREG32_SOC15(VCN, 0, regVCN_RB_ENABLE); data &= ~(VCN_RB_ENABLE__AUDIO_RB_EN_MASK); WREG32_SOC15(VCN, 0, regVCN_RB_ENABLE, data); @@ -248,7 +250,7 @@ static int umsch_mm_v4_0_ring_stop(struct amdgpu_umsch_mm *umsch) data = REG_SET_FIELD(data, VCN_UMSCH_RB_DB_CTRL, EN, 0); WREG32_SOC15(VCN, 0, regVCN_UMSCH_RB_DB_CTRL, data); - if (amdgpu_ip_version(adev, VCN_HWIP, 0) == IP_VERSION(4, 0, 5)) { + if (amdgpu_ip_version(adev, VCN_HWIP, 0) >= IP_VERSION(4, 0, 5)) { WREG32_SOC15(VCN, 0, regUVD_IPX_DLDO_CONFIG, 2 << UVD_IPX_DLDO_CONFIG__ONO0_PWR_CONFIG__SHIFT); SOC15_WAIT_ON_RREG(VCN, 0, regUVD_IPX_DLDO_STATUS, @@ -271,6 +273,8 @@ static int umsch_mm_v4_0_set_hw_resources(struct amdgpu_umsch_mm *umsch) set_hw_resources.vmid_mask_mm_vcn = umsch->vmid_mask_mm_vcn; set_hw_resources.vmid_mask_mm_vpe = umsch->vmid_mask_mm_vpe; + set_hw_resources.collaboration_mask_vpe = + adev->vpe.collaborate_mode ? 0x3 : 0x0; set_hw_resources.engine_mask = umsch->engine_mask; set_hw_resources.vcn0_hqd_mask[0] = umsch->vcn0_hqd_mask; @@ -346,6 +350,7 @@ static int umsch_mm_v4_0_add_queue(struct amdgpu_umsch_mm *umsch, add_queue.h_queue = input_ptr->h_queue; add_queue.vm_context_cntl = input_ptr->vm_context_cntl; add_queue.is_context_suspended = input_ptr->is_context_suspended; + add_queue.collaboration_mode = adev->vpe.collaborate_mode ? 1 : 0; add_queue.api_status.api_completion_fence_addr = umsch->ring.fence_drv.gpu_addr; add_queue.api_status.api_completion_fence_value = ++umsch->ring.fence_drv.sync_seq; diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v3_1.c b/drivers/gpu/drm/amd/amdgpu/uvd_v3_1.c index a6006f231c65..805d6662c88b 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v3_1.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v3_1.c @@ -819,6 +819,8 @@ static const struct amd_ip_funcs uvd_v3_1_ip_funcs = { .soft_reset = uvd_v3_1_soft_reset, .set_clockgating_state = uvd_v3_1_set_clockgating_state, .set_powergating_state = uvd_v3_1_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; const struct amdgpu_ip_block_version uvd_v3_1_ip_block = { diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c b/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c index 1aa09ad7bbe3..3f19c606f4de 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c @@ -769,6 +769,8 @@ static const struct amd_ip_funcs uvd_v4_2_ip_funcs = { .soft_reset = uvd_v4_2_soft_reset, .set_clockgating_state = uvd_v4_2_set_clockgating_state, .set_powergating_state = uvd_v4_2_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs uvd_v4_2_ring_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c index f8b229b75435..efd903c21d48 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c @@ -877,6 +877,8 @@ static const struct amd_ip_funcs uvd_v5_0_ip_funcs = { .set_clockgating_state = uvd_v5_0_set_clockgating_state, .set_powergating_state = uvd_v5_0_set_powergating_state, .get_clockgating_state = uvd_v5_0_get_clockgating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs uvd_v5_0_ring_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c index a9a6880f44e3..495de5068455 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c @@ -1545,6 +1545,8 @@ static const struct amd_ip_funcs uvd_v6_0_ip_funcs = { .set_clockgating_state = uvd_v6_0_set_clockgating_state, .set_powergating_state = uvd_v6_0_set_powergating_state, .get_clockgating_state = uvd_v6_0_get_clockgating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs uvd_v6_0_ring_phys_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c index a08e7abca423..66fada199bda 100644 --- a/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c @@ -626,6 +626,8 @@ static const struct amd_ip_funcs vce_v2_0_ip_funcs = { .soft_reset = vce_v2_0_soft_reset, .set_clockgating_state = vce_v2_0_set_clockgating_state, .set_powergating_state = vce_v2_0_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs vce_v2_0_ring_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c index f4760748d349..32517c364cf7 100644 --- a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c @@ -913,6 +913,8 @@ static const struct amd_ip_funcs vce_v3_0_ip_funcs = { .set_clockgating_state = vce_v3_0_set_clockgating_state, .set_powergating_state = vce_v3_0_set_powergating_state, .get_clockgating_state = vce_v3_0_get_clockgating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs vce_v3_0_ring_phys_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c index aaceecd558cf..cb253bd3a2a2 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c @@ -1902,6 +1902,8 @@ static const struct amd_ip_funcs vcn_v1_0_ip_funcs = { .post_soft_reset = NULL /* vcn_v1_0_post_soft_reset */, .set_clockgating_state = vcn_v1_0_set_clockgating_state, .set_powergating_state = vcn_v1_0_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; /* diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v2_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v2_0.c index e357d8cf0c01..f18fd61c435e 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v2_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v2_0.c @@ -2008,6 +2008,8 @@ static const struct amd_ip_funcs vcn_v2_0_ip_funcs = { .post_soft_reset = NULL, .set_clockgating_state = vcn_v2_0_set_clockgating_state, .set_powergating_state = vcn_v2_0_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_ring_funcs vcn_v2_0_dec_ring_vm_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c b/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c index 1cd8a94b0fbc..baec14bde2a2 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c @@ -1901,6 +1901,8 @@ static const struct amd_ip_funcs vcn_v2_5_ip_funcs = { .post_soft_reset = NULL, .set_clockgating_state = vcn_v2_5_set_clockgating_state, .set_powergating_state = vcn_v2_5_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amd_ip_funcs vcn_v2_6_ip_funcs = { @@ -1921,6 +1923,8 @@ static const struct amd_ip_funcs vcn_v2_6_ip_funcs = { .post_soft_reset = NULL, .set_clockgating_state = vcn_v2_5_set_clockgating_state, .set_powergating_state = vcn_v2_5_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; const struct amdgpu_ip_block_version vcn_v2_5_ip_block = diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v3_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v3_0.c index 8f82fb887e9c..6b31cf4b8aac 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v3_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v3_0.c @@ -359,6 +359,7 @@ static int vcn_v3_0_hw_init(void *handle) } } + return 0; done: if (!r) DRM_INFO("VCN decode and encode initialized successfully(under %s).\n", @@ -2230,6 +2231,8 @@ static const struct amd_ip_funcs vcn_v3_0_ip_funcs = { .post_soft_reset = NULL, .set_clockgating_state = vcn_v3_0_set_clockgating_state, .set_powergating_state = vcn_v3_0_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; const struct amdgpu_ip_block_version vcn_v3_0_ip_block = { diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v4_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v4_0.c index 832d15f7b5f6..ac1b8ead03b3 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v4_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v4_0.c @@ -288,6 +288,7 @@ static int vcn_v4_0_hw_init(void *handle) } } + return 0; done: if (!r) DRM_INFO("VCN decode and encode initialized successfully(under %s).\n", @@ -2130,6 +2131,8 @@ static const struct amd_ip_funcs vcn_v4_0_ip_funcs = { .post_soft_reset = NULL, .set_clockgating_state = vcn_v4_0_set_clockgating_state, .set_powergating_state = vcn_v4_0_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; const struct amdgpu_ip_block_version vcn_v4_0_ip_block = { diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_3.c b/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_3.c index 203fa988322b..2279d8fce03d 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_3.c +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_3.c @@ -1660,6 +1660,8 @@ static const struct amd_ip_funcs vcn_v4_0_3_ip_funcs = { .post_soft_reset = NULL, .set_clockgating_state = vcn_v4_0_3_set_clockgating_state, .set_powergating_state = vcn_v4_0_3_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; const struct amdgpu_ip_block_version vcn_v4_0_3_ip_block = { diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_5.c b/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_5.c index 501e53e69f2a..81fb99729f37 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_5.c +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_5.c @@ -237,6 +237,7 @@ static int vcn_v4_0_5_hw_init(void *handle) goto done; } + return 0; done: if (!r) DRM_INFO("VCN decode and encode initialized successfully(under %s).\n", @@ -1752,6 +1753,8 @@ static const struct amd_ip_funcs vcn_v4_0_5_ip_funcs = { .post_soft_reset = NULL, .set_clockgating_state = vcn_v4_0_5_set_clockgating_state, .set_powergating_state = vcn_v4_0_5_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; const struct amdgpu_ip_block_version vcn_v4_0_5_ip_block = { diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v5_0_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v5_0_0.c index bc60c554eb32..851975b5ce29 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v5_0_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v5_0_0.c @@ -95,7 +95,7 @@ static int vcn_v5_0_0_sw_init(void *handle) return r; for (i = 0; i < adev->vcn.num_vcn_inst; i++) { - volatile struct amdgpu_vcn4_fw_shared *fw_shared; + volatile struct amdgpu_vcn5_fw_shared *fw_shared; if (adev->vcn.harvest_config & (1 << i)) continue; @@ -154,7 +154,7 @@ static int vcn_v5_0_0_sw_fini(void *handle) if (drm_dev_enter(adev_to_drm(adev), &idx)) { for (i = 0; i < adev->vcn.num_vcn_inst; i++) { - volatile struct amdgpu_vcn4_fw_shared *fw_shared; + volatile struct amdgpu_vcn5_fw_shared *fw_shared; if (adev->vcn.harvest_config & (1 << i)) continue; @@ -203,6 +203,7 @@ static int vcn_v5_0_0_hw_init(void *handle) goto done; } + return 0; done: if (!r) DRM_INFO("VCN decode and encode initialized successfully(under %s).\n", @@ -334,7 +335,7 @@ static void vcn_v5_0_0_mc_resume(struct amdgpu_device *adev, int inst) upper_32_bits(adev->vcn.inst[inst].fw_shared.gpu_addr)); WREG32_SOC15(VCN, inst, regUVD_VCPU_NONCACHE_OFFSET0, 0); WREG32_SOC15(VCN, inst, regUVD_VCPU_NONCACHE_SIZE0, - AMDGPU_GPU_PAGE_ALIGN(sizeof(struct amdgpu_vcn4_fw_shared))); + AMDGPU_GPU_PAGE_ALIGN(sizeof(struct amdgpu_vcn5_fw_shared))); } /** @@ -438,7 +439,7 @@ static void vcn_v5_0_0_mc_resume_dpg_mode(struct amdgpu_device *adev, int inst_i VCN, inst_idx, regUVD_VCPU_NONCACHE_OFFSET0), 0, 0, indirect); WREG32_SOC24_DPG_MODE(inst_idx, SOC24_DPG_MODE_OFFSET( VCN, inst_idx, regUVD_VCPU_NONCACHE_SIZE0), - AMDGPU_GPU_PAGE_ALIGN(sizeof(struct amdgpu_vcn4_fw_shared)), 0, indirect); + AMDGPU_GPU_PAGE_ALIGN(sizeof(struct amdgpu_vcn5_fw_shared)), 0, indirect); /* VCN global tiling registers */ WREG32_SOC24_DPG_MODE(inst_idx, SOC24_DPG_MODE_OFFSET( @@ -615,7 +616,7 @@ static void vcn_v5_0_0_enable_clock_gating(struct amdgpu_device *adev, int inst) */ static int vcn_v5_0_0_start_dpg_mode(struct amdgpu_device *adev, int inst_idx, bool indirect) { - volatile struct amdgpu_vcn4_fw_shared *fw_shared = adev->vcn.inst[inst_idx].fw_shared.cpu_addr; + volatile struct amdgpu_vcn5_fw_shared *fw_shared = adev->vcn.inst[inst_idx].fw_shared.cpu_addr; struct amdgpu_ring *ring; uint32_t tmp; @@ -712,7 +713,7 @@ static int vcn_v5_0_0_start_dpg_mode(struct amdgpu_device *adev, int inst_idx, b */ static int vcn_v5_0_0_start(struct amdgpu_device *adev) { - volatile struct amdgpu_vcn4_fw_shared *fw_shared; + volatile struct amdgpu_vcn5_fw_shared *fw_shared; struct amdgpu_ring *ring; uint32_t tmp; int i, j, k, r; @@ -893,7 +894,7 @@ static void vcn_v5_0_0_stop_dpg_mode(struct amdgpu_device *adev, int inst_idx) */ static int vcn_v5_0_0_stop(struct amdgpu_device *adev) { - volatile struct amdgpu_vcn4_fw_shared *fw_shared; + volatile struct amdgpu_vcn5_fw_shared *fw_shared; uint32_t tmp; int i, r = 0; @@ -1328,6 +1329,8 @@ static const struct amd_ip_funcs vcn_v5_0_0_ip_funcs = { .post_soft_reset = NULL, .set_clockgating_state = vcn_v5_0_0_set_clockgating_state, .set_powergating_state = vcn_v5_0_0_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; const struct amdgpu_ip_block_version vcn_v5_0_0_ip_block = { diff --git a/drivers/gpu/drm/amd/amdgpu/vi.c b/drivers/gpu/drm/amd/amdgpu/vi.c index 1a98812981f4..d39c670f6220 100644 --- a/drivers/gpu/drm/amd/amdgpu/vi.c +++ b/drivers/gpu/drm/amd/amdgpu/vi.c @@ -897,7 +897,7 @@ static int vi_asic_pci_config_reset(struct amdgpu_device *adev) return r; } -static bool vi_asic_supports_baco(struct amdgpu_device *adev) +static int vi_asic_supports_baco(struct amdgpu_device *adev) { switch (adev->asic_type) { case CHIP_FIJI: @@ -908,14 +908,14 @@ static bool vi_asic_supports_baco(struct amdgpu_device *adev) case CHIP_TOPAZ: return amdgpu_dpm_is_baco_supported(adev); default: - return false; + return 0; } } static enum amd_reset_method vi_asic_reset_method(struct amdgpu_device *adev) { - bool baco_reset; + int baco_reset; if (amdgpu_reset_method == AMD_RESET_METHOD_LEGACY || amdgpu_reset_method == AMD_RESET_METHOD_BACO) @@ -935,7 +935,7 @@ vi_asic_reset_method(struct amdgpu_device *adev) baco_reset = amdgpu_dpm_is_baco_supported(adev); break; default: - baco_reset = false; + baco_reset = 0; break; } @@ -2058,6 +2058,8 @@ static const struct amd_ip_funcs vi_common_ip_funcs = { .set_clockgating_state = vi_common_set_clockgating_state, .set_powergating_state = vi_common_set_powergating_state, .get_clockgating_state = vi_common_get_clockgating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; static const struct amdgpu_ip_block_version vi_common_ip_block = diff --git a/drivers/gpu/drm/amd/amdgpu/vpe_v6_1.c b/drivers/gpu/drm/amd/amdgpu/vpe_v6_1.c index 769eb8f7bb3c..09315dd5a1ec 100644 --- a/drivers/gpu/drm/amd/amdgpu/vpe_v6_1.c +++ b/drivers/gpu/drm/amd/amdgpu/vpe_v6_1.c @@ -144,6 +144,12 @@ static int vpe_v6_1_load_microcode(struct amdgpu_vpe *vpe) WREG32(vpe_get_reg_offset(vpe, j, regVPEC_CNTL), ret); } + /* setup collaborate mode */ + vpe_v6_1_set_collaborate_mode(vpe, true); + /* setup DPM */ + if (amdgpu_vpe_configure_dpm(vpe)) + dev_warn(adev->dev, "VPE failed to enable DPM\n"); + /* * For VPE 6.1.1, still only need to add master's offset, and psp will apply it to slave as well. * Here use instance 0 as master. @@ -159,11 +165,7 @@ static int vpe_v6_1_load_microcode(struct amdgpu_vpe *vpe) adev->vpe.cmdbuf_cpu_addr[0] = f32_offset; adev->vpe.cmdbuf_cpu_addr[1] = f32_cntl; - amdgpu_vpe_psp_update_sram(adev); - vpe_v6_1_set_collaborate_mode(vpe, true); - amdgpu_vpe_configure_dpm(vpe); - - return 0; + return amdgpu_vpe_psp_update_sram(adev); } vpe_hdr = (const struct vpe_firmware_header_v1_0 *)adev->vpe.fw->data; @@ -196,8 +198,6 @@ static int vpe_v6_1_load_microcode(struct amdgpu_vpe *vpe) } vpe_v6_1_halt(vpe, false); - vpe_v6_1_set_collaborate_mode(vpe, true); - amdgpu_vpe_configure_dpm(vpe); return 0; } diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c index dfa8c69532d4..fdf171ad4a3c 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c @@ -371,6 +371,11 @@ static int kfd_ioctl_create_queue(struct file *filep, struct kfd_process *p, err = -EINVAL; goto err_wptr_map_gart; } + if (dev->adev != amdgpu_ttm_adev(wptr_bo->tbo.bdev)) { + pr_err("Queue memory allocated to wrong device\n"); + err = -EINVAL; + goto err_wptr_map_gart; + } err = amdgpu_amdkfd_map_gtt_bo_to_gart(wptr_bo); if (err) { @@ -779,8 +784,8 @@ static int kfd_ioctl_get_process_apertures_new(struct file *filp, * nodes, but not more than args->num_of_nodes as that is * the amount of memory allocated by user */ - pa = kzalloc((sizeof(struct kfd_process_device_apertures) * - args->num_of_nodes), GFP_KERNEL); + pa = kcalloc(args->num_of_nodes, sizeof(struct kfd_process_device_apertures), + GFP_KERNEL); if (!pa) return -ENOMEM; @@ -1139,7 +1144,7 @@ static int kfd_ioctl_alloc_memory_of_gpu(struct file *filep, goto err_unlock; } offset = dev->adev->rmmio_remap.bus_addr; - if (!offset) { + if (!offset || (PAGE_SIZE > 4096)) { err = -ENOMEM; goto err_unlock; } @@ -1523,7 +1528,7 @@ static int kfd_ioctl_get_dmabuf_info(struct file *filep, /* Find a KFD GPU device that supports the get_dmabuf_info query */ for (i = 0; kfd_topology_enum_kfd_devices(i, &dev) == 0; i++) - if (dev) + if (dev && !kfd_devcgroup_check_permission(dev)) break; if (!dev) return -EINVAL; @@ -1545,7 +1550,7 @@ static int kfd_ioctl_get_dmabuf_info(struct file *filep, if (xcp_id >= 0) args->gpu_id = dmabuf_adev->kfd.dev->nodes[xcp_id]->id; else - args->gpu_id = dmabuf_adev->kfd.dev->nodes[0]->id; + args->gpu_id = dev->id; args->flags = flags; /* Copy metadata buffer to user mode */ @@ -2307,7 +2312,7 @@ static int criu_restore_memory_of_gpu(struct kfd_process_device *pdd, return -EINVAL; } offset = pdd->dev->adev->rmmio_remap.bus_addr; - if (!offset) { + if (!offset || (PAGE_SIZE > 4096)) { pr_err("amdgpu_amdkfd_get_mmio_remap_phys_addr failed\n"); return -ENOMEM; } @@ -3349,6 +3354,9 @@ static int kfd_mmio_mmap(struct kfd_node *dev, struct kfd_process *process, if (vma->vm_end - vma->vm_start != PAGE_SIZE) return -EINVAL; + if (PAGE_SIZE > 4096) + return -EINVAL; + address = dev->adev->rmmio_remap.bus_addr; vm_flags_set(vma, VM_IO | VM_DONTCOPY | VM_DONTEXPAND | VM_NORESERVE | diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device.c b/drivers/gpu/drm/amd/amdkfd/kfd_device.c index 041ec3de55e7..afc57df421cd 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_device.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device.c @@ -408,15 +408,8 @@ struct kfd_dev *kgd2kfd_probe(struct amdgpu_device *adev, bool vf) f2g = &gfx_v11_kfd2kgd; break; case IP_VERSION(11, 0, 3): - if ((adev->pdev->device == 0x7460 && - adev->pdev->revision == 0x00) || - (adev->pdev->device == 0x7461 && - adev->pdev->revision == 0x00)) - /* Note: Compiler version is 11.0.5 while HW version is 11.0.3 */ - gfx_target_version = 110005; - else - /* Note: Compiler version is 11.0.1 while HW version is 11.0.3 */ - gfx_target_version = 110001; + /* Note: Compiler version is 11.0.1 while HW version is 11.0.3 */ + gfx_target_version = 110001; f2g = &gfx_v11_kfd2kgd; break; case IP_VERSION(11, 5, 0): @@ -435,12 +428,12 @@ struct kfd_dev *kgd2kfd_probe(struct amdgpu_device *adev, bool vf) if (!f2g) { if (amdgpu_ip_version(adev, GC_HWIP, 0)) - dev_err(kfd_device, + dev_info(kfd_device, "GC IP %06x %s not supported in kfd\n", amdgpu_ip_version(adev, GC_HWIP, 0), vf ? "VF" : ""); else - dev_err(kfd_device, "%s %s not supported in kfd\n", + dev_info(kfd_device, "%s %s not supported in kfd\n", amdgpu_asic_name[adev->asic_type], vf ? "VF" : ""); return NULL; } @@ -960,7 +953,6 @@ void kgd2kfd_suspend(struct kfd_dev *kfd, bool run_pm) { struct kfd_node *node; int i; - int count; if (!kfd->init_complete) return; @@ -968,12 +960,10 @@ void kgd2kfd_suspend(struct kfd_dev *kfd, bool run_pm) /* for runtime suspend, skip locking kfd */ if (!run_pm) { mutex_lock(&kfd_processes_mutex); - count = ++kfd_locked; - mutex_unlock(&kfd_processes_mutex); - /* For first KFD device suspend all the KFD processes */ - if (count == 1) + if (++kfd_locked == 1) kfd_suspend_all_processes(); + mutex_unlock(&kfd_processes_mutex); } for (i = 0; i < kfd->num_nodes; i++) { @@ -984,7 +974,7 @@ void kgd2kfd_suspend(struct kfd_dev *kfd, bool run_pm) int kgd2kfd_resume(struct kfd_dev *kfd, bool run_pm) { - int ret, count, i; + int ret, i; if (!kfd->init_complete) return 0; @@ -998,12 +988,10 @@ int kgd2kfd_resume(struct kfd_dev *kfd, bool run_pm) /* for runtime resume, skip unlocking kfd */ if (!run_pm) { mutex_lock(&kfd_processes_mutex); - count = --kfd_locked; - mutex_unlock(&kfd_processes_mutex); - - WARN_ONCE(count < 0, "KFD suspend / resume ref. error"); - if (count == 0) + if (--kfd_locked == 0) ret = kfd_resume_all_processes(); + WARN_ONCE(kfd_locked < 0, "KFD suspend / resume ref. error"); + mutex_unlock(&kfd_processes_mutex); } return ret; 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 f4d395e38683..c08b6ee25289 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c @@ -1997,10 +1997,10 @@ static int unmap_queues_cpsch(struct device_queue_manager *dqm, * check those fields */ mqd_mgr = dqm->mqd_mgrs[KFD_MQD_TYPE_HIQ]; - if (mqd_mgr->read_doorbell_id(dqm->packet_mgr.priv_queue->queue->mqd)) { - dev_err(dev, "HIQ MQD's queue_doorbell_id0 is not 0, Queue preemption time out\n"); + if (mqd_mgr->check_preemption_failed(mqd_mgr, dqm->packet_mgr.priv_queue->queue->mqd)) { while (halt_if_hws_hang) schedule(); + kfd_hws_hang(dqm); return -ETIME; } diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v10.c b/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v10.c index 9a06c6fb6605..8e0d0356e810 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v10.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v10.c @@ -134,6 +134,7 @@ static void event_interrupt_poison_consumption(struct kfd_node *dev, { enum amdgpu_ras_block block = 0; int old_poison, ret = -EINVAL; + uint32_t reset = 0; struct kfd_process *p = kfd_lookup_process_by_pasid(pasid); if (!p) @@ -153,6 +154,8 @@ static void event_interrupt_poison_consumption(struct kfd_node *dev, case SOC15_IH_CLIENTID_UTCL2: ret = kfd_dqm_evict_pasid(dev->dqm, pasid); block = AMDGPU_RAS_BLOCK__GFX; + if (ret) + reset = AMDGPU_RAS_GPU_RESET_MODE2_RESET; break; case SOC15_IH_CLIENTID_SDMA0: case SOC15_IH_CLIENTID_SDMA1: @@ -160,6 +163,7 @@ static void event_interrupt_poison_consumption(struct kfd_node *dev, case SOC15_IH_CLIENTID_SDMA3: case SOC15_IH_CLIENTID_SDMA4: block = AMDGPU_RAS_BLOCK__SDMA; + reset = AMDGPU_RAS_GPU_RESET_MODE2_RESET; break; default: break; @@ -170,17 +174,16 @@ static void event_interrupt_poison_consumption(struct kfd_node *dev, /* resetting queue passes, do page retirement without gpu reset * resetting queue fails, fallback to gpu reset solution */ - if (!ret) { + if (!ret) dev_warn(dev->adev->dev, "RAS poison consumption, unmap queue flow succeeded: client id %d\n", client_id); - amdgpu_amdkfd_ras_poison_consumption_handler(dev->adev, block, false); - } else { + else dev_warn(dev->adev->dev, "RAS poison consumption, fall back to gpu reset flow: client id %d\n", client_id); - amdgpu_amdkfd_ras_poison_consumption_handler(dev->adev, block, true); - } + + amdgpu_amdkfd_ras_poison_consumption_handler(dev->adev, block, reset); } static bool event_interrupt_isr_v10(struct kfd_node *dev, @@ -339,7 +342,8 @@ static void event_interrupt_wq_v10(struct kfd_node *dev, break; } kfd_signal_event_interrupt(pasid, context_id0 & 0x7fffff, 23); - } else if (source_id == SOC15_INTSRC_CP_BAD_OPCODE) { + } else if (source_id == SOC15_INTSRC_CP_BAD_OPCODE && + KFD_DBG_EC_TYPE_IS_PACKET(KFD_DEBUG_CP_BAD_OP_ECODE(context_id0))) { kfd_set_dbg_ev_from_interrupt(dev, pasid, KFD_DEBUG_DOORBELL_ID(context_id0), KFD_EC_MASK(KFD_DEBUG_CP_BAD_OP_ECODE(context_id0)), @@ -367,10 +371,25 @@ static void event_interrupt_wq_v10(struct kfd_node *dev, client_id == SOC15_IH_CLIENTID_UTCL2) { struct kfd_vm_fault_info info = {0}; uint16_t ring_id = SOC15_RING_ID_FROM_IH_ENTRY(ih_ring_entry); + uint32_t node_id = SOC15_NODEID_FROM_IH_ENTRY(ih_ring_entry); + uint32_t vmid_type = SOC15_VMID_TYPE_FROM_IH_ENTRY(ih_ring_entry); + int hub_inst = 0; struct kfd_hsa_memory_exception_data exception_data; - if (client_id == SOC15_IH_CLIENTID_UTCL2 && - amdgpu_amdkfd_ras_query_utcl2_poison_status(dev->adev)) { + /* gfxhub */ + if (!vmid_type && dev->adev->gfx.funcs->ih_node_to_logical_xcc) { + hub_inst = dev->adev->gfx.funcs->ih_node_to_logical_xcc(dev->adev, + node_id); + if (hub_inst < 0) + hub_inst = 0; + } + + /* mmhub */ + if (vmid_type && client_id == SOC15_IH_CLIENTID_VMC) + hub_inst = node_id / 4; + + if (amdgpu_amdkfd_ras_query_utcl2_poison_status(dev->adev, + hub_inst, vmid_type)) { event_interrupt_poison_consumption(dev, pasid, client_id); return; } diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v11.c b/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v11.c index 7e2859736a55..f524a55eee11 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v11.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v11.c @@ -193,6 +193,7 @@ static void event_interrupt_poison_consumption_v11(struct kfd_node *dev, { enum amdgpu_ras_block block = 0; int ret = -EINVAL; + uint32_t reset = 0; struct kfd_process *p = kfd_lookup_process_by_pasid(pasid); if (!p) @@ -212,10 +213,13 @@ static void event_interrupt_poison_consumption_v11(struct kfd_node *dev, if (dev->dqm->ops.reset_queues) ret = dev->dqm->ops.reset_queues(dev->dqm, pasid); block = AMDGPU_RAS_BLOCK__GFX; + if (ret) + reset = AMDGPU_RAS_GPU_RESET_MODE2_RESET; break; case SOC21_INTSRC_SDMA_ECC: default: block = AMDGPU_RAS_BLOCK__GFX; + reset = AMDGPU_RAS_GPU_RESET_MODE2_RESET; break; } @@ -223,10 +227,7 @@ static void event_interrupt_poison_consumption_v11(struct kfd_node *dev, /* resetting queue passes, do page retirement without gpu reset resetting queue fails, fallback to gpu reset solution */ - if (!ret) - amdgpu_amdkfd_ras_poison_consumption_handler(dev->adev, block, false); - else - amdgpu_amdkfd_ras_poison_consumption_handler(dev->adev, block, true); + amdgpu_amdkfd_ras_poison_consumption_handler(dev->adev, block, reset); } static bool event_interrupt_isr_v11(struct kfd_node *dev, @@ -328,7 +329,8 @@ static void event_interrupt_wq_v11(struct kfd_node *dev, /* CP */ if (source_id == SOC15_INTSRC_CP_END_OF_PIPE) kfd_signal_event_interrupt(pasid, context_id0, 32); - else if (source_id == SOC15_INTSRC_CP_BAD_OPCODE) + else if (source_id == SOC15_INTSRC_CP_BAD_OPCODE && + KFD_DBG_EC_TYPE_IS_PACKET(KFD_CTXID0_CP_BAD_OP_ECODE(context_id0))) kfd_set_dbg_ev_from_interrupt(dev, pasid, KFD_CTXID0_DOORBELL_ID(context_id0), KFD_EC_MASK(KFD_CTXID0_CP_BAD_OP_ECODE(context_id0)), diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c b/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c index 91dd5e045b51..e1c21d250611 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c @@ -144,7 +144,8 @@ static void event_interrupt_poison_consumption_v9(struct kfd_node *dev, uint16_t pasid, uint16_t client_id) { enum amdgpu_ras_block block = 0; - int old_poison, ret = -EINVAL; + int old_poison; + uint32_t reset = 0; struct kfd_process *p = kfd_lookup_process_by_pasid(pasid); if (!p) @@ -162,8 +163,13 @@ static void event_interrupt_poison_consumption_v9(struct kfd_node *dev, case SOC15_IH_CLIENTID_SE2SH: case SOC15_IH_CLIENTID_SE3SH: case SOC15_IH_CLIENTID_UTCL2: - ret = kfd_dqm_evict_pasid(dev->dqm, pasid); block = AMDGPU_RAS_BLOCK__GFX; + reset = AMDGPU_RAS_GPU_RESET_MODE2_RESET; + break; + case SOC15_IH_CLIENTID_VMC: + case SOC15_IH_CLIENTID_VMC1: + block = AMDGPU_RAS_BLOCK__MMHUB; + reset = AMDGPU_RAS_GPU_RESET_MODE1_RESET; break; case SOC15_IH_CLIENTID_SDMA0: case SOC15_IH_CLIENTID_SDMA1: @@ -171,27 +177,21 @@ static void event_interrupt_poison_consumption_v9(struct kfd_node *dev, case SOC15_IH_CLIENTID_SDMA3: case SOC15_IH_CLIENTID_SDMA4: block = AMDGPU_RAS_BLOCK__SDMA; + reset = AMDGPU_RAS_GPU_RESET_MODE2_RESET; break; default: - break; + dev_warn(dev->adev->dev, + "client %d does not support poison consumption\n", client_id); + return; } kfd_signal_poison_consumed_event(dev, pasid); - /* resetting queue passes, do page retirement without gpu reset - * resetting queue fails, fallback to gpu reset solution - */ - if (!ret) { - dev_warn(dev->adev->dev, - "RAS poison consumption, unmap queue flow succeeded: client id %d\n", - client_id); - amdgpu_amdkfd_ras_poison_consumption_handler(dev->adev, block, false); - } else { - dev_warn(dev->adev->dev, - "RAS poison consumption, fall back to gpu reset flow: client id %d\n", - client_id); - amdgpu_amdkfd_ras_poison_consumption_handler(dev->adev, block, true); - } + dev_warn(dev->adev->dev, + "poison is consumed by client %d, kick off gpu reset flow\n", client_id); + + amdgpu_amdkfd_ras_pasid_poison_consumption_handler(dev->adev, + block, pasid, NULL, NULL, reset); } static bool context_id_expected(struct kfd_dev *dev) @@ -388,7 +388,8 @@ static void event_interrupt_wq_v9(struct kfd_node *dev, break; } kfd_signal_event_interrupt(pasid, sq_int_data, 24); - } else if (source_id == SOC15_INTSRC_CP_BAD_OPCODE) { + } else if (source_id == SOC15_INTSRC_CP_BAD_OPCODE && + KFD_DBG_EC_TYPE_IS_PACKET(KFD_DEBUG_CP_BAD_OP_ECODE(context_id0))) { kfd_set_dbg_ev_from_interrupt(dev, pasid, KFD_DEBUG_DOORBELL_ID(context_id0), KFD_EC_MASK(KFD_DEBUG_CP_BAD_OP_ECODE(context_id0)), @@ -413,10 +414,25 @@ static void event_interrupt_wq_v9(struct kfd_node *dev, client_id == SOC15_IH_CLIENTID_UTCL2) { struct kfd_vm_fault_info info = {0}; uint16_t ring_id = SOC15_RING_ID_FROM_IH_ENTRY(ih_ring_entry); + uint32_t node_id = SOC15_NODEID_FROM_IH_ENTRY(ih_ring_entry); + uint32_t vmid_type = SOC15_VMID_TYPE_FROM_IH_ENTRY(ih_ring_entry); + int hub_inst = 0; struct kfd_hsa_memory_exception_data exception_data; - if (client_id == SOC15_IH_CLIENTID_UTCL2 && - amdgpu_amdkfd_ras_query_utcl2_poison_status(dev->adev)) { + /* gfxhub */ + if (!vmid_type && dev->adev->gfx.funcs->ih_node_to_logical_xcc) { + hub_inst = dev->adev->gfx.funcs->ih_node_to_logical_xcc(dev->adev, + node_id); + if (hub_inst < 0) + hub_inst = 0; + } + + /* mmhub */ + if (vmid_type && client_id == SOC15_IH_CLIENTID_VMC) + hub_inst = node_id / 4; + + if (amdgpu_amdkfd_ras_query_utcl2_poison_status(dev->adev, + hub_inst, vmid_type)) { event_interrupt_poison_consumption_v9(dev, pasid, client_id); return; } diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_interrupt.c b/drivers/gpu/drm/amd/amdkfd/kfd_interrupt.c index dd3c43c1ad70..9b6b6e882593 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_interrupt.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_interrupt.c @@ -104,6 +104,8 @@ void kfd_interrupt_exit(struct kfd_node *node) */ flush_workqueue(node->ih_wq); + destroy_workqueue(node->ih_wq); + kfifo_free(&node->ih_fifo); } diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c b/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c index bdc01ca9609a..8ee3d07ffbdf 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c @@ -77,7 +77,7 @@ svm_migrate_gart_map(struct amdgpu_ring *ring, uint64_t npages, dst_addr = amdgpu_bo_gpu_offset(adev->gart.bo); amdgpu_emit_copy_buffer(adev, &job->ibs[0], src_addr, - dst_addr, num_bytes, false); + dst_addr, num_bytes, 0); amdgpu_ring_pad_ib(ring, &job->ibs[0]); WARN_ON(job->ibs[0].length_dw > num_dw); @@ -153,7 +153,7 @@ svm_migrate_copy_memory_gart(struct amdgpu_device *adev, dma_addr_t *sys, } r = amdgpu_copy_buffer(ring, gart_s, gart_d, size * PAGE_SIZE, - NULL, &next, false, true, false); + NULL, &next, false, true, 0); if (r) { dev_err(adev->dev, "fail %d to copy memory\n", r); goto out_unlock; @@ -509,10 +509,19 @@ svm_migrate_ram_to_vram(struct svm_range *prange, uint32_t best_loc, start = start_mgr << PAGE_SHIFT; end = (last_mgr + 1) << PAGE_SHIFT; + r = amdgpu_amdkfd_reserve_mem_limit(node->adev, + prange->npages * PAGE_SIZE, + KFD_IOC_ALLOC_MEM_FLAGS_VRAM, + node->xcp ? node->xcp->id : 0); + if (r) { + dev_dbg(node->adev->dev, "failed to reserve VRAM, r: %ld\n", r); + return -ENOSPC; + } + r = svm_range_vram_node_new(node, prange, true); if (r) { dev_dbg(node->adev->dev, "fail %ld to alloc vram\n", r); - return r; + goto out; } ttm_res_offset = (start_mgr - prange->start + prange->offset) << PAGE_SHIFT; @@ -545,6 +554,11 @@ svm_migrate_ram_to_vram(struct svm_range *prange, uint32_t best_loc, svm_range_vram_node_free(prange); } +out: + amdgpu_amdkfd_unreserve_mem_limit(node->adev, + prange->npages * PAGE_SIZE, + KFD_IOC_ALLOC_MEM_FLAGS_VRAM, + node->xcp ? node->xcp->id : 0); return r < 0 ? r : 0; } @@ -1009,7 +1023,7 @@ int kgd2kfd_init_zone_device(struct amdgpu_device *adev) if (amdgpu_ip_version(adev, GC_HWIP, 0) < IP_VERSION(9, 0, 1)) return -EINVAL; - if (adev->gmc.is_app_apu) + if (adev->flags & AMD_IS_APU) return 0; pgmap = &kfddev->pgmap; diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.c index 050a6936ff84..8746a61a852d 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.c @@ -290,3 +290,21 @@ uint64_t kfd_mqd_stride(struct mqd_manager *mm, { return mm->mqd_size; } + +bool kfd_check_hiq_mqd_doorbell_id(struct kfd_node *node, uint32_t doorbell_id, + uint32_t inst) +{ + if (doorbell_id) { + struct device *dev = node->adev->dev; + + if (node->adev->xcp_mgr && node->adev->xcp_mgr->num_xcps > 0) + dev_err(dev, "XCC %d: Queue preemption failed for queue with doorbell_id: %x\n", + inst, doorbell_id); + else + dev_err(dev, "Queue preemption failed for queue with doorbell_id: %x\n", + doorbell_id); + return true; + } + + return false; +} diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.h b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.h index e5cc697a3ca8..17cc1f25c8d0 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.h @@ -119,7 +119,7 @@ struct mqd_manager { #if defined(CONFIG_DEBUG_FS) int (*debugfs_show_mqd)(struct seq_file *m, void *data); #endif - uint32_t (*read_doorbell_id)(void *mqd); + bool (*check_preemption_failed)(struct mqd_manager *mm, void *mqd); uint64_t (*mqd_stride)(struct mqd_manager *mm, struct queue_properties *p); @@ -198,4 +198,6 @@ void kfd_get_hiq_xcc_mqd(struct kfd_node *dev, uint64_t kfd_hiq_mqd_stride(struct kfd_node *dev); uint64_t kfd_mqd_stride(struct mqd_manager *mm, struct queue_properties *q); +bool kfd_check_hiq_mqd_doorbell_id(struct kfd_node *node, uint32_t doorbell_id, + uint32_t inst); #endif /* KFD_MQD_MANAGER_H_ */ diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c index 1a4a69943c71..05f3ac2eaef9 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c @@ -206,11 +206,11 @@ static void __update_mqd(struct mqd_manager *mm, void *mqd, q->is_active = QUEUE_IS_ACTIVE(*q); } -static uint32_t read_doorbell_id(void *mqd) +static bool check_preemption_failed(struct mqd_manager *mm, void *mqd) { struct cik_mqd *m = (struct cik_mqd *)mqd; - return m->queue_doorbell_id0; + return kfd_check_hiq_mqd_doorbell_id(mm->dev, m->queue_doorbell_id0, 0); } static void update_mqd(struct mqd_manager *mm, void *mqd, @@ -423,7 +423,7 @@ struct mqd_manager *mqd_manager_init_cik(enum KFD_MQD_TYPE type, #if defined(CONFIG_DEBUG_FS) mqd->debugfs_show_mqd = debugfs_show_mqd; #endif - mqd->read_doorbell_id = read_doorbell_id; + mqd->check_preemption_failed = check_preemption_failed; break; case KFD_MQD_TYPE_DIQ: mqd->allocate_mqd = allocate_mqd; diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v10.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v10.c index 22cbfa1bdadd..2eff37aaf827 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v10.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v10.c @@ -224,11 +224,11 @@ static void update_mqd(struct mqd_manager *mm, void *mqd, q->is_active = QUEUE_IS_ACTIVE(*q); } -static uint32_t read_doorbell_id(void *mqd) +static bool check_preemption_failed(struct mqd_manager *mm, void *mqd) { struct v10_compute_mqd *m = (struct v10_compute_mqd *)mqd; - return m->queue_doorbell_id0; + return kfd_check_hiq_mqd_doorbell_id(mm->dev, m->queue_doorbell_id0, 0); } static int get_wave_state(struct mqd_manager *mm, void *mqd, @@ -488,7 +488,7 @@ struct mqd_manager *mqd_manager_init_v10(enum KFD_MQD_TYPE type, #if defined(CONFIG_DEBUG_FS) mqd->debugfs_show_mqd = debugfs_show_mqd; #endif - mqd->read_doorbell_id = read_doorbell_id; + mqd->check_preemption_failed = check_preemption_failed; pr_debug("%s@%i\n", __func__, __LINE__); break; case KFD_MQD_TYPE_DIQ: diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v11.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v11.c index 826bc4f6c8a7..68dbc0399c87 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v11.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v11.c @@ -278,11 +278,11 @@ static void update_mqd(struct mqd_manager *mm, void *mqd, q->is_active = QUEUE_IS_ACTIVE(*q); } -static uint32_t read_doorbell_id(void *mqd) +static bool check_preemption_failed(struct mqd_manager *mm, void *mqd) { struct v11_compute_mqd *m = (struct v11_compute_mqd *)mqd; - return m->queue_doorbell_id0; + return kfd_check_hiq_mqd_doorbell_id(mm->dev, m->queue_doorbell_id0, 0); } static int get_wave_state(struct mqd_manager *mm, void *mqd, @@ -517,7 +517,7 @@ struct mqd_manager *mqd_manager_init_v11(enum KFD_MQD_TYPE type, #if defined(CONFIG_DEBUG_FS) mqd->debugfs_show_mqd = debugfs_show_mqd; #endif - mqd->read_doorbell_id = read_doorbell_id; + mqd->check_preemption_failed = check_preemption_failed; pr_debug("%s@%i\n", __func__, __LINE__); break; case KFD_MQD_TYPE_DIQ: diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c index 697b6d530d12..6bddc16808d7 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c @@ -316,11 +316,11 @@ static void update_mqd(struct mqd_manager *mm, void *mqd, } -static uint32_t read_doorbell_id(void *mqd) +static bool check_preemption_failed(struct mqd_manager *mm, void *mqd) { struct v9_mqd *m = (struct v9_mqd *)mqd; - return m->queue_doorbell_id0; + return kfd_check_hiq_mqd_doorbell_id(mm->dev, m->queue_doorbell_id0, 0); } static int get_wave_state(struct mqd_manager *mm, void *mqd, @@ -607,6 +607,24 @@ static int destroy_hiq_mqd_v9_4_3(struct mqd_manager *mm, void *mqd, return err; } +static bool check_preemption_failed_v9_4_3(struct mqd_manager *mm, void *mqd) +{ + uint64_t hiq_mqd_size = kfd_hiq_mqd_stride(mm->dev); + uint32_t xcc_mask = mm->dev->xcc_mask; + int inst = 0, xcc_id; + struct v9_mqd *m; + bool ret = false; + + for_each_inst(xcc_id, xcc_mask) { + m = get_mqd(mqd + hiq_mqd_size * inst); + ret |= kfd_check_hiq_mqd_doorbell_id(mm->dev, + m->queue_doorbell_id0, inst); + ++inst; + } + + return ret; +} + static void get_xcc_mqd(struct kfd_mem_obj *mqd_mem_obj, struct kfd_mem_obj *xcc_mqd_mem_obj, uint64_t offset) @@ -881,15 +899,16 @@ struct mqd_manager *mqd_manager_init_v9(enum KFD_MQD_TYPE type, #if defined(CONFIG_DEBUG_FS) mqd->debugfs_show_mqd = debugfs_show_mqd; #endif - mqd->read_doorbell_id = read_doorbell_id; if (KFD_GC_VERSION(dev) == IP_VERSION(9, 4, 3)) { mqd->init_mqd = init_mqd_hiq_v9_4_3; mqd->load_mqd = hiq_load_mqd_kiq_v9_4_3; mqd->destroy_mqd = destroy_hiq_mqd_v9_4_3; + mqd->check_preemption_failed = check_preemption_failed_v9_4_3; } else { mqd->init_mqd = init_mqd_hiq; mqd->load_mqd = kfd_hiq_load_mqd_kiq; mqd->destroy_mqd = destroy_hiq_mqd; + mqd->check_preemption_failed = check_preemption_failed; } break; case KFD_MQD_TYPE_DIQ: diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c index 3e1a574d4ea6..c1fafc502515 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c @@ -237,11 +237,11 @@ static void __update_mqd(struct mqd_manager *mm, void *mqd, q->is_active = QUEUE_IS_ACTIVE(*q); } -static uint32_t read_doorbell_id(void *mqd) +static bool check_preemption_failed(struct mqd_manager *mm, void *mqd) { struct vi_mqd *m = (struct vi_mqd *)mqd; - return m->queue_doorbell_id0; + return kfd_check_hiq_mqd_doorbell_id(mm->dev, m->queue_doorbell_id0, 0); } static void update_mqd(struct mqd_manager *mm, void *mqd, @@ -482,7 +482,7 @@ struct mqd_manager *mqd_manager_init_vi(enum KFD_MQD_TYPE type, #if defined(CONFIG_DEBUG_FS) mqd->debugfs_show_mqd = debugfs_show_mqd; #endif - mqd->read_doorbell_id = read_doorbell_id; + mqd->check_preemption_failed = check_preemption_failed; break; case KFD_MQD_TYPE_DIQ: mqd->allocate_mqd = allocate_mqd; diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h index 42d40560cd30..a81ef232fdef 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h @@ -1473,7 +1473,7 @@ static inline void kfd_flush_tlb(struct kfd_process_device *pdd, static inline bool kfd_flush_tlb_after_unmap(struct kfd_dev *dev) { - return KFD_GC_VERSION(dev) > IP_VERSION(9, 4, 2) || + return KFD_GC_VERSION(dev) >= IP_VERSION(9, 4, 2) || (KFD_GC_VERSION(dev) == IP_VERSION(9, 4, 1) && dev->sdma_fw_version >= 18) || KFD_GC_VERSION(dev) == IP_VERSION(9, 4, 0); } diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process.c b/drivers/gpu/drm/amd/amdkfd/kfd_process.c index 717a60d7a4ea..451bb058cc62 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_process.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_process.c @@ -819,9 +819,9 @@ struct kfd_process *kfd_create_process(struct task_struct *thread) mutex_lock(&kfd_processes_mutex); if (kfd_is_locked()) { - mutex_unlock(&kfd_processes_mutex); pr_debug("KFD is locked! Cannot create process"); - return ERR_PTR(-EINVAL); + process = ERR_PTR(-EINVAL); + goto out; } /* A prior open of /dev/kfd could have already created the process. */ @@ -829,6 +829,14 @@ struct kfd_process *kfd_create_process(struct task_struct *thread) if (process) { pr_debug("Process already found\n"); } else { + /* If the process just called exec(3), it is possible that the + * cleanup of the kfd_process (following the release of the mm + * of the old process image) is still in the cleanup work queue. + * Make sure to drain any job before trying to recreate any + * resource for this process. + */ + flush_workqueue(kfd_process_wq); + process = create_process(thread); if (IS_ERR(process)) goto out; @@ -1922,6 +1930,8 @@ static int signal_eviction_fence(struct kfd_process *p) rcu_read_lock(); ef = dma_fence_get_rcu_safe(&p->ef); rcu_read_unlock(); + if (!ef) + return -EINVAL; ret = dma_fence_signal(ef); dma_fence_put(ef); @@ -1949,10 +1959,9 @@ static void evict_process_worker(struct work_struct *work) * they are responsible stopping the queues and scheduling * the restore work. */ - if (!signal_eviction_fence(p)) - queue_delayed_work(kfd_restore_wq, &p->restore_work, - msecs_to_jiffies(PROCESS_RESTORE_TIME_MS)); - else + if (signal_eviction_fence(p) || + mod_delayed_work(kfd_restore_wq, &p->restore_work, + msecs_to_jiffies(PROCESS_RESTORE_TIME_MS))) kfd_process_restore_queues(p); pr_debug("Finished evicting pasid 0x%x\n", p->pasid); @@ -2011,9 +2020,9 @@ static void restore_process_worker(struct work_struct *work) if (ret) { pr_debug("Failed to restore BOs of pasid 0x%x, retry after %d ms\n", p->pasid, PROCESS_BACK_OFF_TIME_MS); - ret = queue_delayed_work(kfd_restore_wq, &p->restore_work, - msecs_to_jiffies(PROCESS_BACK_OFF_TIME_MS)); - WARN(!ret, "reschedule restore work failed\n"); + if (mod_delayed_work(kfd_restore_wq, &p->restore_work, + msecs_to_jiffies(PROCESS_RESTORE_TIME_MS))) + kfd_process_restore_queues(p); } } diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c index f0f7f48af413..31e500859ab0 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c @@ -2619,7 +2619,7 @@ svm_range_best_restore_location(struct svm_range *prange, return -1; } - if (node->adev->gmc.is_app_apu) + if (node->adev->flags & AMD_IS_APU) return 0; if (prange->preferred_loc == gpuid || @@ -3337,7 +3337,7 @@ svm_range_best_prefetch_location(struct svm_range *prange) goto out; } - if (bo_node->adev->gmc.is_app_apu) { + if (bo_node->adev->flags & AMD_IS_APU) { best_loc = 0; goto out; } @@ -3426,7 +3426,7 @@ svm_range_trigger_migration(struct mm_struct *mm, struct svm_range *prange, mm, KFD_MIGRATE_TRIGGER_PREFETCH); *migrated = !r; - return r; + return 0; } int svm_range_schedule_evict_svm_bo(struct amdgpu_amdkfd_fence *fence) diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_svm.h b/drivers/gpu/drm/amd/amdkfd/kfd_svm.h index 026863a0abcd..70c1776611c4 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_svm.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_svm.h @@ -201,7 +201,7 @@ void svm_range_list_lock_and_flush_work(struct svm_range_list *svms, struct mm_s * is initialized to not 0 when page migration register device memory. */ #define KFD_IS_SVM_API_SUPPORTED(adev) ((adev)->kfd.pgmap.type != 0 ||\ - (adev)->gmc.is_app_apu) + ((adev)->flags & AMD_IS_APU)) void svm_range_bo_unref_async(struct svm_range_bo *svm_bo); diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c index c51f131eaa2f..bc9eb847ecfe 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c @@ -1997,9 +1997,8 @@ int kfd_topology_add_device(struct kfd_node *gpu) HSA_CAP_ASIC_REVISION_MASK); dev->node_props.location_id = pci_dev_id(gpu->adev->pdev); - /* On multi-partition nodes, node id = location_id[31:28] */ - if (gpu->kfd->num_nodes > 1) - dev->node_props.location_id |= (dev->gpu->node_id << 28); + if (KFD_GC_VERSION(dev->gpu->kfd) == IP_VERSION(9, 4, 3)) + dev->node_props.location_id |= dev->gpu->node_id; dev->node_props.domain = pci_domain_nr(gpu->adev->pdev->bus); dev->node_props.max_engine_clk_fcompute = diff --git a/drivers/gpu/drm/amd/display/Kconfig b/drivers/gpu/drm/amd/display/Kconfig index 901d1961b739..5fcd4f778dc3 100644 --- a/drivers/gpu/drm/amd/display/Kconfig +++ b/drivers/gpu/drm/amd/display/Kconfig @@ -8,7 +8,7 @@ config DRM_AMD_DC depends on BROKEN || !CC_IS_CLANG || ARM64 || RISCV || SPARC64 || X86_64 select SND_HDA_COMPONENT if SND_HDA_CORE # !CC_IS_CLANG: https://github.com/ClangBuiltLinux/linux/issues/1752 - select DRM_AMD_DC_FP if (X86 || LOONGARCH || (PPC64 && ALTIVEC) || (ARM64 && KERNEL_MODE_NEON && !CC_IS_CLANG)) + select DRM_AMD_DC_FP if ARCH_HAS_KERNEL_FPU_SUPPORT && (!ARM64 || !CC_IS_CLANG) help Choose this option if you want to use the new display engine support for AMDGPU. This adds required support for Vega and diff --git a/drivers/gpu/drm/amd/display/Makefile b/drivers/gpu/drm/amd/display/Makefile index 92a5c5efcf92..9a5bcafbf730 100644 --- a/drivers/gpu/drm/amd/display/Makefile +++ b/drivers/gpu/drm/amd/display/Makefile @@ -33,6 +33,7 @@ subdir-ccflags-y += -I$(FULL_AMD_DISPLAY_PATH)/dc/hwss subdir-ccflags-y += -I$(FULL_AMD_DISPLAY_PATH)/dc/resource subdir-ccflags-y += -I$(FULL_AMD_DISPLAY_PATH)/dc/dsc subdir-ccflags-y += -I$(FULL_AMD_DISPLAY_PATH)/dc/optc +subdir-ccflags-y += -I$(FULL_AMD_DISPLAY_PATH)/dc/dpp subdir-ccflags-y += -I$(FULL_AMD_DISPLAY_PATH)/modules/inc subdir-ccflags-y += -I$(FULL_AMD_DISPLAY_PATH)/modules/freesync subdir-ccflags-y += -I$(FULL_AMD_DISPLAY_PATH)/modules/color 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 2851719d7121..f1d67c6f4b98 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -148,6 +148,9 @@ MODULE_FIRMWARE(FIRMWARE_NAVI12_DMCU); #define FIRMWARE_DCN_35_DMUB "amdgpu/dcn_3_5_dmcub.bin" MODULE_FIRMWARE(FIRMWARE_DCN_35_DMUB); +#define FIRMWARE_DCN_351_DMUB "amdgpu/dcn_3_5_1_dmcub.bin" +MODULE_FIRMWARE(FIRMWARE_DCN_351_DMUB); + /* Number of bytes in PSP header for firmware. */ #define PSP_HEADER_BYTES 0x100 @@ -271,7 +274,7 @@ static u32 dm_vblank_get_counter(struct amdgpu_device *adev, int crtc) static int dm_crtc_get_scanoutpos(struct amdgpu_device *adev, int crtc, u32 *vbl, u32 *position) { - u32 v_blank_start, v_blank_end, h_position, v_position; + u32 v_blank_start = 0, v_blank_end = 0, h_position = 0, v_position = 0; struct amdgpu_crtc *acrtc = NULL; struct dc *dc = adev->dm.dc; @@ -845,7 +848,7 @@ static void dm_handle_hpd_work(struct work_struct *work) */ static void dm_dmub_outbox1_low_irq(void *interrupt_params) { - struct dmub_notification notify; + struct dmub_notification notify = {0}; struct common_irq_params *irq_params = interrupt_params; struct amdgpu_device *adev = irq_params->adev; struct amdgpu_display_manager *dm = &adev->dm; @@ -1227,6 +1230,15 @@ static int dm_dmub_hw_init(struct amdgpu_device *adev) break; } + switch (amdgpu_ip_version(adev, DCE_HWIP, 0)) { + case IP_VERSION(3, 5, 0): + case IP_VERSION(3, 5, 1): + hw_params.ips_sequential_ono = adev->external_rev_id > 0x10; + break; + default: + break; + } + status = dmub_srv_hw_init(dmub_srv, &hw_params); if (status != DMUB_STATUS_OK) { DRM_ERROR("Error initializing DMUB HW: %d\n", status); @@ -1723,8 +1735,10 @@ static int amdgpu_dm_init(struct amdgpu_device *adev) if (amdgpu_dc_debug_mask & DC_DISABLE_IPS) init_data.flags.disable_ips = DMUB_IPS_DISABLE_ALL; + else + init_data.flags.disable_ips = DMUB_IPS_RCG_IN_ACTIVE_IPS2_IN_OFF; - init_data.flags.disable_ips_in_vpb = 1; + init_data.flags.disable_ips_in_vpb = 0; /* Enable DWB for tested platforms only */ if (amdgpu_ip_version(adev, DCE_HWIP, 0) >= IP_VERSION(3, 0, 0)) @@ -2626,6 +2640,7 @@ static enum dc_status amdgpu_dm_commit_zero_streams(struct dc *dc) int i; struct dc_stream_state *del_streams[MAX_PIPES]; int del_streams_count = 0; + struct dc_commit_streams_params params = {}; memset(del_streams, 0, sizeof(del_streams)); @@ -2652,7 +2667,9 @@ static enum dc_status amdgpu_dm_commit_zero_streams(struct dc *dc) goto fail; } - res = dc_commit_streams(dc, context->streams, context->stream_count); + params.streams = context->streams; + params.stream_count = context->stream_count; + res = dc_commit_streams(dc, ¶ms); fail: dc_state_release(context); @@ -2874,6 +2891,7 @@ static int dm_resume(void *handle) struct dc_state *dc_state; int i, r, j, ret; bool need_hotplug = false; + struct dc_commit_streams_params commit_params = {}; if (dm->dc->caps.ips_support) { dc_dmub_srv_apply_idle_power_optimizations(dm->dc, false); @@ -2923,7 +2941,9 @@ static int dm_resume(void *handle) dc_enable_dmub_outbox(adev->dm.dc); } - WARN_ON(!dc_commit_streams(dm->dc, dc_state->streams, dc_state->stream_count)); + commit_params.streams = dc_state->streams; + commit_params.stream_count = dc_state->stream_count; + WARN_ON(!dc_commit_streams(dm->dc, &commit_params)); dm_gpureset_commit_state(dm->cached_dc_state, dm); @@ -2940,7 +2960,7 @@ static int dm_resume(void *handle) } /* Recreate dc_state - DC invalidates it when setting power state to S3. */ dc_state_release(dm_state->context); - dm_state->context = dc_state_create(dm->dc); + dm_state->context = dc_state_create(dm->dc, NULL); /* TODO: Remove dc_state->dccg, use dc->dccg directly. */ /* Before powering on DC we need to re-initialize DMUB. */ @@ -3026,6 +3046,7 @@ static int dm_resume(void *handle) dc_stream_release(dm_new_crtc_state->stream); dm_new_crtc_state->stream = NULL; } + dm_new_crtc_state->base.color_mgmt_changed = true; } for_each_new_plane_in_state(dm->cached_state, plane, new_plane_state, i) { @@ -3044,6 +3065,10 @@ static int dm_resume(void *handle) /* Do mst topology probing after resuming cached state*/ drm_connector_list_iter_begin(ddev, &iter); drm_for_each_connector_iter(connector, &iter) { + + if (connector->connector_type == DRM_MODE_CONNECTOR_WRITEBACK) + continue; + aconnector = to_amdgpu_dm_connector(connector); if (aconnector->dc_link->type != dc_connection_mst_branch || aconnector->mst_root) @@ -3096,6 +3121,8 @@ static const struct amd_ip_funcs amdgpu_dm_funcs = { .soft_reset = dm_soft_reset, .set_clockgating_state = dm_set_clockgating_state, .set_powergating_state = dm_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; const struct amdgpu_ip_block_version dm_ip_block = { @@ -4529,15 +4556,18 @@ static int amdgpu_dm_initialize_drm_device(struct amdgpu_device *adev) /* Determine whether to enable Replay support by default. */ if (!(amdgpu_dc_debug_mask & DC_DISABLE_REPLAY)) { switch (amdgpu_ip_version(adev, DCE_HWIP, 0)) { - case IP_VERSION(3, 1, 4): - case IP_VERSION(3, 1, 5): - case IP_VERSION(3, 1, 6): - case IP_VERSION(3, 2, 0): - case IP_VERSION(3, 2, 1): - case IP_VERSION(3, 5, 0): - case IP_VERSION(3, 5, 1): - replay_feature_enabled = true; - break; +/* + * Disabled by default due to https://gitlab.freedesktop.org/drm/amd/-/issues/3344 + * case IP_VERSION(3, 1, 4): + * case IP_VERSION(3, 1, 5): + * case IP_VERSION(3, 1, 6): + * case IP_VERSION(3, 2, 0): + * case IP_VERSION(3, 2, 1): + * case IP_VERSION(3, 5, 0): + * case IP_VERSION(3, 5, 1): + * replay_feature_enabled = true; + * break; + */ default: replay_feature_enabled = amdgpu_dc_feature_mask & DC_REPLAY_MASK; break; @@ -4820,9 +4850,11 @@ static int dm_init_microcode(struct amdgpu_device *adev) fw_name_dmub = FIRMWARE_DCN_V3_2_1_DMCUB; break; case IP_VERSION(3, 5, 0): - case IP_VERSION(3, 5, 1): fw_name_dmub = FIRMWARE_DCN_35_DMUB; break; + case IP_VERSION(3, 5, 1): + fw_name_dmub = FIRMWARE_DCN_351_DMUB; + break; default: /* ASIC doesn't support DMUB. */ return 0; @@ -5700,8 +5732,8 @@ static void fill_stream_properties_from_drm_display_mode( timing_out->aspect_ratio = get_aspect_ratio(mode_in); - stream->out_transfer_func->type = TF_TYPE_PREDEFINED; - stream->out_transfer_func->tf = TRANSFER_FUNCTION_SRGB; + stream->out_transfer_func.type = TF_TYPE_PREDEFINED; + stream->out_transfer_func.tf = TRANSFER_FUNCTION_SRGB; if (stream->signal == SIGNAL_TYPE_HDMI_TYPE_A) { if (!adjust_colour_depth_from_display_info(timing_out, info) && drm_mode_is_420_also(info, mode_in) && @@ -5921,6 +5953,9 @@ get_highest_refresh_rate_mode(struct amdgpu_dm_connector *aconnector, &aconnector->base.probed_modes : &aconnector->base.modes; + if (aconnector->base.connector_type == DRM_MODE_CONNECTOR_WRITEBACK) + return NULL; + if (aconnector->freesync_vid_base.clock != 0) return &aconnector->freesync_vid_base; @@ -6305,27 +6340,22 @@ create_stream_for_sink(struct drm_connector *connector, if (stream->signal == SIGNAL_TYPE_HDMI_TYPE_A) mod_build_hf_vsif_infopacket(stream, &stream->vsp_infopacket); - else if (stream->signal == SIGNAL_TYPE_DISPLAY_PORT || - stream->signal == SIGNAL_TYPE_DISPLAY_PORT_MST || - stream->signal == SIGNAL_TYPE_EDP) { + + if (stream->signal == SIGNAL_TYPE_DISPLAY_PORT || + stream->signal == SIGNAL_TYPE_DISPLAY_PORT_MST || + stream->signal == SIGNAL_TYPE_EDP) { // // should decide stream support vsc sdp colorimetry capability // before building vsc info packet // - stream->use_vsc_sdp_for_colorimetry = false; - if (aconnector->dc_sink->sink_signal == SIGNAL_TYPE_DISPLAY_PORT_MST) { - stream->use_vsc_sdp_for_colorimetry = - aconnector->dc_sink->is_vsc_sdp_colorimetry_supported; - } else { - if (stream->link->dpcd_caps.dprx_feature.bits.VSC_SDP_COLORIMETRY_SUPPORTED) - stream->use_vsc_sdp_for_colorimetry = true; - } - if (stream->out_transfer_func->tf == TRANSFER_FUNCTION_GAMMA22) + stream->use_vsc_sdp_for_colorimetry = stream->link->dpcd_caps.dpcd_rev.raw >= 0x14 && + stream->link->dpcd_caps.dprx_feature.bits.VSC_SDP_COLORIMETRY_SUPPORTED; + + if (stream->out_transfer_func.tf == TRANSFER_FUNCTION_GAMMA22) tf = TRANSFER_FUNC_GAMMA_22; mod_build_vsc_infopacket(stream, &stream->vsc_infopacket, stream->output_color_space, tf); + aconnector->psr_skip_count = AMDGPU_DM_PSR_ENTRY_DELAY; - if (stream->link->psr_settings.psr_feature_enabled) - aconnector->psr_skip_count = AMDGPU_DM_PSR_ENTRY_DELAY; } finish: dc_sink_release(sink); @@ -6792,7 +6822,7 @@ static enum dc_status dm_validate_stream_and_context(struct dc *dc, if (!dc_plane_state) goto cleanup; - dc_state = dc_state_create(dc); + dc_state = dc_state_create(dc, NULL); if (!dc_state) goto cleanup; @@ -7181,7 +7211,7 @@ static int dm_update_mst_vcpi_slots_for_dsc(struct drm_atomic_state *state, struct amdgpu_dm_connector *aconnector; struct dm_connector_state *dm_conn_state; int i, j, ret; - int vcpi, pbn_div, pbn, slot_num = 0; + int vcpi, pbn_div, pbn = 0, slot_num = 0; for_each_new_connector_in_state(state, connector, new_con_state, i) { @@ -8394,13 +8424,13 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state, bundle->surface_updates[planes_count].surface = dc_plane; if (new_pcrtc_state->color_mgmt_changed) { - bundle->surface_updates[planes_count].gamma = dc_plane->gamma_correction; - bundle->surface_updates[planes_count].in_transfer_func = dc_plane->in_transfer_func; + bundle->surface_updates[planes_count].gamma = &dc_plane->gamma_correction; + bundle->surface_updates[planes_count].in_transfer_func = &dc_plane->in_transfer_func; bundle->surface_updates[planes_count].gamut_remap_matrix = &dc_plane->gamut_remap_matrix; bundle->surface_updates[planes_count].hdr_mult = dc_plane->hdr_mult; - bundle->surface_updates[planes_count].func_shaper = dc_plane->in_shaper_func; - bundle->surface_updates[planes_count].lut3d_func = dc_plane->lut3d_func; - bundle->surface_updates[planes_count].blend_tf = dc_plane->blend_tf; + bundle->surface_updates[planes_count].func_shaper = &dc_plane->in_shaper_func; + bundle->surface_updates[planes_count].lut3d_func = &dc_plane->lut3d_func; + bundle->surface_updates[planes_count].blend_tf = &dc_plane->blend_tf; } amdgpu_dm_plane_fill_dc_scaling_info(dm->adev, new_plane_state, @@ -8613,7 +8643,7 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state, bundle->stream_update.output_csc_transform = &acrtc_state->stream->csc_color_matrix; bundle->stream_update.out_transfer_func = - acrtc_state->stream->out_transfer_func; + &acrtc_state->stream->out_transfer_func; bundle->stream_update.lut3d_func = (struct dc_3dlut *) acrtc_state->stream->lut3d_func; bundle->stream_update.func_shaper = @@ -8764,10 +8794,10 @@ static void amdgpu_dm_commit_audio(struct drm_device *dev, if (!drm_atomic_crtc_needs_modeset(new_crtc_state)) continue; +notify: if (connector->connector_type == DRM_MODE_CONNECTOR_WRITEBACK) continue; -notify: aconnector = to_amdgpu_dm_connector(connector); mutex_lock(&adev->dm.audio_lock); @@ -8847,6 +8877,7 @@ static void amdgpu_dm_commit_streams(struct drm_atomic_state *state, struct drm_connector *connector; bool mode_set_reset_required = false; u32 i; + struct dc_commit_streams_params params = {dc_state->streams, dc_state->stream_count}; /* Disable writeback */ for_each_old_connector_in_state(state, connector, old_con_state, i) { @@ -8983,7 +9014,7 @@ static void amdgpu_dm_commit_streams(struct drm_atomic_state *state, dm_enable_per_frame_crtc_master_sync(dc_state); mutex_lock(&dm->dc_lock); - WARN_ON(!dc_commit_streams(dm->dc, dc_state->streams, dc_state->stream_count)); + WARN_ON(!dc_commit_streams(dm->dc, ¶ms)); /* Allow idle optimization when vblank count is 0 for display off */ if (dm->active_vblank_irq_count == 0) @@ -10587,7 +10618,7 @@ static int amdgpu_dm_atomic_check(struct drm_device *dev, struct dm_crtc_state *dm_old_crtc_state, *dm_new_crtc_state; struct drm_dp_mst_topology_mgr *mgr; struct drm_dp_mst_topology_state *mst_state; - struct dsc_mst_fairness_vars vars[MAX_PIPES]; + struct dsc_mst_fairness_vars vars[MAX_PIPES] = {0}; trace_amdgpu_dm_atomic_check_begin(state); diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c index c87b64e464ed..ebabfe3a512f 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c @@ -571,7 +571,7 @@ static int amdgpu_dm_set_atomic_regamma(struct dc_stream_state *stream, uint32_t regamma_size, bool has_rom, enum dc_transfer_func_predefined tf) { - struct dc_transfer_func *out_tf = stream->out_transfer_func; + struct dc_transfer_func *out_tf = &stream->out_transfer_func; int ret = 0; if (regamma_size || tf != TRANSFER_FUNCTION_LINEAR) { @@ -954,8 +954,8 @@ int amdgpu_dm_update_crtc_color_mgmt(struct dm_crtc_state *crtc) * inverse color ramp in legacy userspace. */ crtc->cm_is_degamma_srgb = true; - stream->out_transfer_func->type = TF_TYPE_DISTRIBUTED_POINTS; - stream->out_transfer_func->tf = TRANSFER_FUNCTION_SRGB; + stream->out_transfer_func.type = TF_TYPE_DISTRIBUTED_POINTS; + stream->out_transfer_func.tf = TRANSFER_FUNCTION_SRGB; /* * Note: although we pass has_rom as parameter here, we never * actually use ROM because the color module only takes the ROM @@ -963,7 +963,7 @@ int amdgpu_dm_update_crtc_color_mgmt(struct dm_crtc_state *crtc) * * See more in mod_color_calculate_regamma_params() */ - r = __set_legacy_tf(stream->out_transfer_func, regamma_lut, + r = __set_legacy_tf(&stream->out_transfer_func, regamma_lut, regamma_size, has_rom); if (r) return r; @@ -1034,7 +1034,7 @@ map_crtc_degamma_to_dc_plane(struct dm_crtc_state *crtc, °amma_size); ASSERT(degamma_size == MAX_COLOR_LUT_ENTRIES); - dc_plane_state->in_transfer_func->type = TF_TYPE_DISTRIBUTED_POINTS; + dc_plane_state->in_transfer_func.type = TF_TYPE_DISTRIBUTED_POINTS; /* * This case isn't fully correct, but also fairly @@ -1061,12 +1061,12 @@ map_crtc_degamma_to_dc_plane(struct dm_crtc_state *crtc, * map these to the atomic one instead. */ if (crtc->cm_is_degamma_srgb) - dc_plane_state->in_transfer_func->tf = tf; + dc_plane_state->in_transfer_func.tf = tf; else - dc_plane_state->in_transfer_func->tf = + dc_plane_state->in_transfer_func.tf = TRANSFER_FUNCTION_LINEAR; - r = __set_input_tf(caps, dc_plane_state->in_transfer_func, + r = __set_input_tf(caps, &dc_plane_state->in_transfer_func, degamma_lut, degamma_size); if (r) return r; @@ -1075,12 +1075,12 @@ map_crtc_degamma_to_dc_plane(struct dm_crtc_state *crtc, * For legacy gamma support we need the regamma input * in linear space. Assume that the input is sRGB. */ - dc_plane_state->in_transfer_func->type = TF_TYPE_PREDEFINED; - dc_plane_state->in_transfer_func->tf = tf; + dc_plane_state->in_transfer_func.type = TF_TYPE_PREDEFINED; + dc_plane_state->in_transfer_func.tf = tf; if (tf != TRANSFER_FUNCTION_SRGB && !mod_color_calculate_degamma_params(caps, - dc_plane_state->in_transfer_func, + &dc_plane_state->in_transfer_func, NULL, false)) return -ENOMEM; } @@ -1114,24 +1114,24 @@ __set_dm_plane_degamma(struct drm_plane_state *plane_state, if (!has_degamma_lut && tf == AMDGPU_TRANSFER_FUNCTION_DEFAULT) return -EINVAL; - dc_plane_state->in_transfer_func->tf = amdgpu_tf_to_dc_tf(tf); + dc_plane_state->in_transfer_func.tf = amdgpu_tf_to_dc_tf(tf); if (has_degamma_lut) { ASSERT(degamma_size == MAX_COLOR_LUT_ENTRIES); - dc_plane_state->in_transfer_func->type = + dc_plane_state->in_transfer_func.type = TF_TYPE_DISTRIBUTED_POINTS; - ret = __set_input_tf(color_caps, dc_plane_state->in_transfer_func, + ret = __set_input_tf(color_caps, &dc_plane_state->in_transfer_func, degamma_lut, degamma_size); if (ret) return ret; } else { - dc_plane_state->in_transfer_func->type = + dc_plane_state->in_transfer_func.type = TF_TYPE_PREDEFINED; if (!mod_color_calculate_degamma_params(color_caps, - dc_plane_state->in_transfer_func, NULL, false)) + &dc_plane_state->in_transfer_func, NULL, false)) return -ENOMEM; } return 0; @@ -1156,11 +1156,11 @@ amdgpu_dm_plane_set_color_properties(struct drm_plane_state *plane_state, lut3d = __extract_blob_lut(dm_plane_state->lut3d, &lut3d_size); lut3d_size = lut3d != NULL ? lut3d_size : 0; - amdgpu_dm_atomic_lut3d(lut3d, lut3d_size, dc_plane_state->lut3d_func); + amdgpu_dm_atomic_lut3d(lut3d, lut3d_size, &dc_plane_state->lut3d_func); ret = amdgpu_dm_atomic_shaper_lut(shaper_lut, false, amdgpu_tf_to_dc_tf(shaper_tf), shaper_size, - dc_plane_state->in_shaper_func); + &dc_plane_state->in_shaper_func); if (ret) { drm_dbg_kms(plane_state->plane->dev, "setting plane %d shaper LUT failed.\n", @@ -1175,7 +1175,7 @@ amdgpu_dm_plane_set_color_properties(struct drm_plane_state *plane_state, ret = amdgpu_dm_atomic_blend_lut(blend_lut, false, amdgpu_tf_to_dc_tf(blend_tf), - blend_size, dc_plane_state->blend_tf); + blend_size, &dc_plane_state->blend_tf); if (ret) { drm_dbg_kms(plane_state->plane->dev, "setting plane %d gamma lut failed.\n", @@ -1221,8 +1221,8 @@ int amdgpu_dm_update_plane_color_mgmt(struct dm_crtc_state *crtc, color_caps = &dc_plane_state->ctx->dc->caps.color; /* Initially, we can just bypass the DGM block. */ - dc_plane_state->in_transfer_func->type = TF_TYPE_BYPASS; - dc_plane_state->in_transfer_func->tf = TRANSFER_FUNCTION_LINEAR; + dc_plane_state->in_transfer_func.type = TF_TYPE_BYPASS; + dc_plane_state->in_transfer_func.tf = TRANSFER_FUNCTION_LINEAR; /* After, we start to update values according to color props */ has_crtc_cm_degamma = (crtc->cm_has_degamma || crtc->cm_is_degamma_srgb); 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 eee4945653e2..4d7a5d470b1e 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 @@ -1249,7 +1249,7 @@ static ssize_t dp_sdp_message_debugfs_write(struct file *f, const char __user *b size_t size, loff_t *pos) { int r; - uint8_t data[36]; + uint8_t data[36] = {0}; struct amdgpu_dm_connector *connector = file_inode(f)->i_private; struct dm_crtc_state *acrtc_state; uint32_t write_size = 36; @@ -1495,7 +1495,9 @@ static ssize_t dp_dsc_clock_en_read(struct file *f, char __user *buf, for (i = 0; i < MAX_PIPES; i++) { pipe_ctx = &aconnector->dc_link->dc->current_state->res_ctx.pipe_ctx[i]; if (pipe_ctx->stream && - pipe_ctx->stream->link == aconnector->dc_link) + pipe_ctx->stream->link == aconnector->dc_link && + pipe_ctx->stream->sink && + pipe_ctx->stream->sink == aconnector->dc_sink) break; } @@ -1596,7 +1598,9 @@ static ssize_t dp_dsc_clock_en_write(struct file *f, const char __user *buf, for (i = 0; i < MAX_PIPES; i++) { pipe_ctx = &aconnector->dc_link->dc->current_state->res_ctx.pipe_ctx[i]; if (pipe_ctx->stream && - pipe_ctx->stream->link == aconnector->dc_link) + pipe_ctx->stream->link == aconnector->dc_link && + pipe_ctx->stream->sink && + pipe_ctx->stream->sink == aconnector->dc_sink) break; } @@ -1681,7 +1685,9 @@ static ssize_t dp_dsc_slice_width_read(struct file *f, char __user *buf, for (i = 0; i < MAX_PIPES; i++) { pipe_ctx = &aconnector->dc_link->dc->current_state->res_ctx.pipe_ctx[i]; if (pipe_ctx->stream && - pipe_ctx->stream->link == aconnector->dc_link) + pipe_ctx->stream->link == aconnector->dc_link && + pipe_ctx->stream->sink && + pipe_ctx->stream->sink == aconnector->dc_sink) break; } @@ -1780,7 +1786,9 @@ static ssize_t dp_dsc_slice_width_write(struct file *f, const char __user *buf, for (i = 0; i < MAX_PIPES; i++) { pipe_ctx = &aconnector->dc_link->dc->current_state->res_ctx.pipe_ctx[i]; if (pipe_ctx->stream && - pipe_ctx->stream->link == aconnector->dc_link) + pipe_ctx->stream->link == aconnector->dc_link && + pipe_ctx->stream->sink && + pipe_ctx->stream->sink == aconnector->dc_sink) break; } @@ -1865,7 +1873,9 @@ static ssize_t dp_dsc_slice_height_read(struct file *f, char __user *buf, for (i = 0; i < MAX_PIPES; i++) { pipe_ctx = &aconnector->dc_link->dc->current_state->res_ctx.pipe_ctx[i]; if (pipe_ctx->stream && - pipe_ctx->stream->link == aconnector->dc_link) + pipe_ctx->stream->link == aconnector->dc_link && + pipe_ctx->stream->sink && + pipe_ctx->stream->sink == aconnector->dc_sink) break; } @@ -1964,7 +1974,9 @@ static ssize_t dp_dsc_slice_height_write(struct file *f, const char __user *buf, for (i = 0; i < MAX_PIPES; i++) { pipe_ctx = &aconnector->dc_link->dc->current_state->res_ctx.pipe_ctx[i]; if (pipe_ctx->stream && - pipe_ctx->stream->link == aconnector->dc_link) + pipe_ctx->stream->link == aconnector->dc_link && + pipe_ctx->stream->sink && + pipe_ctx->stream->sink == aconnector->dc_sink) break; } @@ -2045,7 +2057,9 @@ static ssize_t dp_dsc_bits_per_pixel_read(struct file *f, char __user *buf, for (i = 0; i < MAX_PIPES; i++) { pipe_ctx = &aconnector->dc_link->dc->current_state->res_ctx.pipe_ctx[i]; if (pipe_ctx->stream && - pipe_ctx->stream->link == aconnector->dc_link) + pipe_ctx->stream->link == aconnector->dc_link && + pipe_ctx->stream->sink && + pipe_ctx->stream->sink == aconnector->dc_sink) break; } @@ -2141,7 +2155,9 @@ static ssize_t dp_dsc_bits_per_pixel_write(struct file *f, const char __user *bu for (i = 0; i < MAX_PIPES; i++) { pipe_ctx = &aconnector->dc_link->dc->current_state->res_ctx.pipe_ctx[i]; if (pipe_ctx->stream && - pipe_ctx->stream->link == aconnector->dc_link) + pipe_ctx->stream->link == aconnector->dc_link && + pipe_ctx->stream->sink && + pipe_ctx->stream->sink == aconnector->dc_sink) break; } @@ -2220,7 +2236,9 @@ static ssize_t dp_dsc_pic_width_read(struct file *f, char __user *buf, for (i = 0; i < MAX_PIPES; i++) { pipe_ctx = &aconnector->dc_link->dc->current_state->res_ctx.pipe_ctx[i]; if (pipe_ctx->stream && - pipe_ctx->stream->link == aconnector->dc_link) + pipe_ctx->stream->link == aconnector->dc_link && + pipe_ctx->stream->sink && + pipe_ctx->stream->sink == aconnector->dc_sink) break; } @@ -2276,7 +2294,9 @@ static ssize_t dp_dsc_pic_height_read(struct file *f, char __user *buf, for (i = 0; i < MAX_PIPES; i++) { pipe_ctx = &aconnector->dc_link->dc->current_state->res_ctx.pipe_ctx[i]; if (pipe_ctx->stream && - pipe_ctx->stream->link == aconnector->dc_link) + pipe_ctx->stream->link == aconnector->dc_link && + pipe_ctx->stream->sink && + pipe_ctx->stream->sink == aconnector->dc_sink) break; } @@ -2347,7 +2367,9 @@ static ssize_t dp_dsc_chunk_size_read(struct file *f, char __user *buf, for (i = 0; i < MAX_PIPES; i++) { pipe_ctx = &aconnector->dc_link->dc->current_state->res_ctx.pipe_ctx[i]; if (pipe_ctx->stream && - pipe_ctx->stream->link == aconnector->dc_link) + pipe_ctx->stream->link == aconnector->dc_link && + pipe_ctx->stream->sink && + pipe_ctx->stream->sink == aconnector->dc_sink) break; } @@ -2418,7 +2440,9 @@ static ssize_t dp_dsc_slice_bpg_offset_read(struct file *f, char __user *buf, for (i = 0; i < MAX_PIPES; i++) { pipe_ctx = &aconnector->dc_link->dc->current_state->res_ctx.pipe_ctx[i]; if (pipe_ctx->stream && - pipe_ctx->stream->link == aconnector->dc_link) + pipe_ctx->stream->link == aconnector->dc_link && + pipe_ctx->stream->sink && + pipe_ctx->stream->sink == aconnector->dc_sink) break; } @@ -2936,7 +2960,7 @@ static int psr_read_residency(void *data, u64 *val) { struct amdgpu_dm_connector *connector = data; struct dc_link *link = connector->dc_link; - u32 residency; + u32 residency = 0; link->dc->link_srv->edp_get_psr_residency(link, &residency); 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 c27063305a13..2c36f3d00ca2 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 @@ -363,7 +363,7 @@ void dm_helpers_dp_mst_send_payload_allocation( mst_state = to_drm_dp_mst_topology_state(mst_mgr->base.state); new_payload = drm_atomic_get_mst_payload_state(mst_state, aconnector->mst_output_port); - ret = drm_dp_add_payload_part2(mst_mgr, mst_state->base.state, new_payload); + ret = drm_dp_add_payload_part2(mst_mgr, new_payload); if (ret) { amdgpu_dm_set_mst_status(&aconnector->mst_status, diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c index 941e96f100f4..a5e1a93ddaea 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c @@ -613,6 +613,9 @@ dm_dp_add_mst_connector(struct drm_dp_mst_topology_mgr *mgr, &connector->base, dev->mode_config.tile_property, 0); + connector->colorspace_property = master->base.colorspace_property; + if (connector->colorspace_property) + drm_connector_attach_colorspace_property(connector); drm_connector_set_path_property(connector, pathprop); @@ -791,25 +794,12 @@ struct dsc_mst_fairness_params { struct amdgpu_dm_connector *aconnector; }; -static uint16_t get_fec_overhead_multiplier(struct dc_link *dc_link) -{ - u8 link_coding_cap; - uint16_t fec_overhead_multiplier_x1000 = PBN_FEC_OVERHEAD_MULTIPLIER_8B_10B; - - link_coding_cap = dc_link_dp_mst_decide_link_encoding_format(dc_link); - if (link_coding_cap == DP_128b_132b_ENCODING) - fec_overhead_multiplier_x1000 = PBN_FEC_OVERHEAD_MULTIPLIER_128B_132B; - - return fec_overhead_multiplier_x1000; -} - -static int kbps_to_peak_pbn(int kbps, uint16_t fec_overhead_multiplier_x1000) +static int kbps_to_peak_pbn(int kbps) { u64 peak_kbps = kbps; peak_kbps *= 1006; - peak_kbps *= fec_overhead_multiplier_x1000; - peak_kbps = div_u64(peak_kbps, 1000 * 1000); + peak_kbps = div_u64(peak_kbps, 1000); return (int) DIV64_U64_ROUND_UP(peak_kbps * 64, (54 * 8 * 1000)); } @@ -910,12 +900,11 @@ static int increase_dsc_bpp(struct drm_atomic_state *state, int link_timeslots_used; int fair_pbn_alloc; int ret = 0; - uint16_t fec_overhead_multiplier_x1000 = get_fec_overhead_multiplier(dc_link); for (i = 0; i < count; i++) { if (vars[i + k].dsc_enabled) { initial_slack[i] = - kbps_to_peak_pbn(params[i].bw_range.max_kbps, fec_overhead_multiplier_x1000) - vars[i + k].pbn; + kbps_to_peak_pbn(params[i].bw_range.max_kbps) - vars[i + k].pbn; bpp_increased[i] = false; remaining_to_increase += 1; } else { @@ -1011,7 +1000,6 @@ static int try_disable_dsc(struct drm_atomic_state *state, int next_index; int remaining_to_try = 0; int ret; - uint16_t fec_overhead_multiplier_x1000 = get_fec_overhead_multiplier(dc_link); for (i = 0; i < count; i++) { if (vars[i + k].dsc_enabled @@ -1041,7 +1029,7 @@ static int try_disable_dsc(struct drm_atomic_state *state, if (next_index == -1) break; - vars[next_index].pbn = kbps_to_peak_pbn(params[next_index].bw_range.stream_kbps, fec_overhead_multiplier_x1000); + vars[next_index].pbn = kbps_to_peak_pbn(params[next_index].bw_range.stream_kbps); ret = drm_dp_atomic_find_time_slots(state, params[next_index].port->mgr, params[next_index].port, @@ -1054,7 +1042,8 @@ static int try_disable_dsc(struct drm_atomic_state *state, vars[next_index].dsc_enabled = false; vars[next_index].bpp_x16 = 0; } else { - vars[next_index].pbn = kbps_to_peak_pbn(params[next_index].bw_range.max_kbps, fec_overhead_multiplier_x1000); + vars[next_index].pbn = kbps_to_peak_pbn( + params[next_index].bw_range.max_kbps); ret = drm_dp_atomic_find_time_slots(state, params[next_index].port->mgr, params[next_index].port, @@ -1083,7 +1072,6 @@ static int compute_mst_dsc_configs_for_link(struct drm_atomic_state *state, int count = 0; int i, k, ret; bool debugfs_overwrite = false; - uint16_t fec_overhead_multiplier_x1000 = get_fec_overhead_multiplier(dc_link); memset(params, 0, sizeof(params)); @@ -1148,7 +1136,7 @@ static int compute_mst_dsc_configs_for_link(struct drm_atomic_state *state, /* Try no compression */ for (i = 0; i < count; i++) { vars[i + k].aconnector = params[i].aconnector; - vars[i + k].pbn = kbps_to_peak_pbn(params[i].bw_range.stream_kbps, fec_overhead_multiplier_x1000); + vars[i + k].pbn = kbps_to_peak_pbn(params[i].bw_range.stream_kbps); vars[i + k].dsc_enabled = false; vars[i + k].bpp_x16 = 0; ret = drm_dp_atomic_find_time_slots(state, params[i].port->mgr, params[i].port, @@ -1167,7 +1155,7 @@ static int compute_mst_dsc_configs_for_link(struct drm_atomic_state *state, /* Try max compression */ for (i = 0; i < count; i++) { if (params[i].compression_possible && params[i].clock_force_enable != DSC_CLK_FORCE_DISABLE) { - vars[i + k].pbn = kbps_to_peak_pbn(params[i].bw_range.min_kbps, fec_overhead_multiplier_x1000); + vars[i + k].pbn = kbps_to_peak_pbn(params[i].bw_range.min_kbps); vars[i + k].dsc_enabled = true; vars[i + k].bpp_x16 = params[i].bw_range.min_target_bpp_x16; ret = drm_dp_atomic_find_time_slots(state, params[i].port->mgr, @@ -1175,7 +1163,7 @@ static int compute_mst_dsc_configs_for_link(struct drm_atomic_state *state, if (ret < 0) return ret; } else { - vars[i + k].pbn = kbps_to_peak_pbn(params[i].bw_range.stream_kbps, fec_overhead_multiplier_x1000); + vars[i + k].pbn = kbps_to_peak_pbn(params[i].bw_range.stream_kbps); vars[i + k].dsc_enabled = false; vars[i + k].bpp_x16 = 0; ret = drm_dp_atomic_find_time_slots(state, params[i].port->mgr, @@ -1219,8 +1207,10 @@ static bool is_dsc_need_re_compute( if (dc_link->type != dc_connection_mst_branch) return false; - if (!(dc_link->dpcd_caps.dsc_caps.dsc_basic_caps.fields.dsc_support.DSC_SUPPORT || - dc_link->dpcd_caps.dsc_caps.dsc_basic_caps.fields.dsc_support.DSC_PASSTHROUGH_SUPPORT)) + /* add a check for older MST DSC with no virtual DPCDs */ + if (needs_dsc_aux_workaround(dc_link) && + (!(dc_link->dpcd_caps.dsc_caps.dsc_basic_caps.fields.dsc_support.DSC_SUPPORT || + dc_link->dpcd_caps.dsc_caps.dsc_basic_caps.fields.dsc_support.DSC_PASSTHROUGH_SUPPORT))) return false; for (i = 0; i < MAX_PIPES; i++) @@ -1240,7 +1230,15 @@ static bool is_dsc_need_re_compute( continue; aconnector = (struct amdgpu_dm_connector *) stream->dm_stream_context; - if (!aconnector) + if (!aconnector || !aconnector->dsc_aux) + continue; + + /* + * check if cached virtual MST DSC caps are available and DSC is supported + * as per specifications in their Virtual DPCD registers. + */ + if (!(aconnector->dc_sink->dsc_caps.dsc_dec_caps.is_dsc_supported || + aconnector->dc_link->dpcd_caps.dsc_caps.dsc_basic_caps.fields.dsc_support.DSC_PASSTHROUGH_SUPPORT)) continue; stream_on_link[new_stream_on_link_num] = aconnector; @@ -1601,7 +1599,7 @@ enum dc_status dm_dp_mst_is_port_support_mode( struct amdgpu_dm_connector *aconnector, struct dc_stream_state *stream) { - int bpp, pbn, branch_max_throughput_mps = 0; + int pbn, branch_max_throughput_mps = 0; struct dc_link_settings cur_link_settings; unsigned int end_to_end_bw_in_kbps = 0; unsigned int upper_link_bw_in_kbps = 0, down_link_bw_in_kbps = 0; @@ -1651,11 +1649,34 @@ enum dc_status dm_dp_mst_is_port_support_mode( } } } else { - /* check if mode could be supported within full_pbn */ - bpp = convert_dc_color_depth_into_bpc(stream->timing.display_color_depth) * 3; - pbn = drm_dp_calc_pbn_mode(stream->timing.pix_clk_100hz / 10, bpp << 4); - if (pbn > aconnector->mst_output_port->full_pbn) + /* Check if mode could be supported within max slot + * number of current mst link and full_pbn of mst links. + */ + int pbn_div, slot_num, max_slot_num; + enum dc_link_encoding_format link_encoding; + uint32_t stream_kbps = + dc_bandwidth_in_kbps_from_timing(&stream->timing, + dc_link_get_highest_encoding_format(stream->link)); + + pbn = kbps_to_peak_pbn(stream_kbps); + pbn_div = dm_mst_get_pbn_divider(stream->link); + slot_num = DIV_ROUND_UP(pbn, pbn_div); + + link_encoding = dc_link_get_highest_encoding_format(stream->link); + if (link_encoding == DC_LINK_ENCODING_DP_8b_10b) + max_slot_num = 63; + else if (link_encoding == DC_LINK_ENCODING_DP_128b_132b) + max_slot_num = 64; + else { + DRM_DEBUG_DRIVER("Invalid link encoding format\n"); return DC_FAIL_BANDWIDTH_VALIDATE; + } + + if (slot_num > max_slot_num || + pbn > aconnector->mst_output_port->full_pbn) { + DRM_DEBUG_DRIVER("Mode can not be supported within mst links!"); + return DC_FAIL_BANDWIDTH_VALIDATE; + } } /* check is mst dsc output bandwidth branch_overall_throughput_0_mps */ diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.h b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.h index 37c820ab0fdb..fa84d34b7373 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.h +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.h @@ -46,9 +46,6 @@ #define SYNAPTICS_CASCADED_HUB_ID 0x5A #define IS_SYNAPTICS_CASCADED_PANAMERA(devName, data) ((IS_SYNAPTICS_PANAMERA(devName) && ((int)data[2] == SYNAPTICS_CASCADED_HUB_ID)) ? 1 : 0) -#define PBN_FEC_OVERHEAD_MULTIPLIER_8B_10B 1031 -#define PBN_FEC_OVERHEAD_MULTIPLIER_128B_132B 1000 - enum mst_msg_ready_type { NONE_MSG_RDY_EVENT = 0, DOWN_REP_MSG_RDY_EVENT = 1, diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.c index 1f08c6564c3b..bfa090432ce2 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.c @@ -141,9 +141,8 @@ bool amdgpu_dm_link_setup_psr(struct dc_stream_state *stream) * amdgpu_dm_psr_enable() - enable psr f/w * @stream: stream state * - * Return: true if success */ -bool amdgpu_dm_psr_enable(struct dc_stream_state *stream) +void amdgpu_dm_psr_enable(struct dc_stream_state *stream) { struct dc_link *link = stream->link; unsigned int vsync_rate_hz = 0; @@ -190,7 +189,10 @@ bool amdgpu_dm_psr_enable(struct dc_stream_state *stream) if (link->psr_settings.psr_version < DC_PSR_VERSION_SU_1) power_opt |= psr_power_opt_z10_static_screen; - return dc_link_set_psr_allow_active(link, &psr_enable, false, false, &power_opt); + dc_link_set_psr_allow_active(link, &psr_enable, false, false, &power_opt); + + if (link->ctx->dc->caps.ips_support) + dc_allow_idle_optimizations(link->ctx->dc, true); } /* @@ -210,7 +212,7 @@ bool amdgpu_dm_psr_disable(struct dc_stream_state *stream) } /* - * amdgpu_dm_psr_disable() - disable psr f/w + * amdgpu_dm_psr_disable_all() - disable psr f/w for all streams * if psr is enabled on any stream * * Return: true if success diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.h b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.h index 6806b3c9c84b..1fdfd183c0d9 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.h +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.h @@ -32,7 +32,7 @@ #define AMDGPU_DM_PSR_ENTRY_DELAY 5 void amdgpu_dm_set_psr_caps(struct dc_link *link); -bool amdgpu_dm_psr_enable(struct dc_stream_state *stream); +void amdgpu_dm_psr_enable(struct dc_stream_state *stream); bool amdgpu_dm_link_setup_psr(struct dc_stream_state *stream); bool amdgpu_dm_psr_disable(struct dc_stream_state *stream); bool amdgpu_dm_psr_disable_all(struct amdgpu_display_manager *dm); diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c index d9e33c6bccd9..0005f5f8f34f 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c @@ -52,4 +52,12 @@ void dm_perf_trace_timestamp(const char *func_name, unsigned int line, struct dc func_name, line); } +void dm_trace_smu_msg(uint32_t msg_id, uint32_t param_in, struct dc_context *ctx) +{ +} + +void dm_trace_smu_delay(uint32_t delay, struct dc_context *ctx) +{ +} + /**** power component interfaces ****/ diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_trace.h b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_trace.h index 133af994a08c..4686d4b0cbad 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_trace.h +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_trace.h @@ -87,7 +87,7 @@ TRACE_EVENT(amdgpu_dc_performance, __entry->writes = write_count; __entry->read_delta = read_count - *last_read; __entry->write_delta = write_count - *last_write; - __assign_str(func, func); + __assign_str(func); __entry->line = line; *last_read = read_count; *last_write = write_count; diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_wb.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_wb.c index 16e72d623630..08c494a7a21b 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_wb.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_wb.c @@ -76,10 +76,8 @@ static int amdgpu_dm_wb_encoder_atomic_check(struct drm_encoder *encoder, static int amdgpu_dm_wb_connector_get_modes(struct drm_connector *connector) { - struct drm_device *dev = connector->dev; - - return drm_add_modes_noedid(connector, dev->mode_config.max_width, - dev->mode_config.max_height); + /* Maximum resolution supported by DWB */ + return drm_add_modes_noedid(connector, 3840, 2160); } static int amdgpu_dm_wb_prepare_job(struct drm_writeback_connector *wb_connector, diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/dc_fpu.c b/drivers/gpu/drm/amd/display/amdgpu_dm/dc_fpu.c index 4ae4720535a5..e46f8ce41d87 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/dc_fpu.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/dc_fpu.c @@ -26,16 +26,7 @@ #include "dc_trace.h" -#if defined(CONFIG_X86) -#include <asm/fpu/api.h> -#elif defined(CONFIG_PPC64) -#include <asm/switch_to.h> -#include <asm/cputable.h> -#elif defined(CONFIG_ARM64) -#include <asm/neon.h> -#elif defined(CONFIG_LOONGARCH) -#include <asm/fpu.h> -#endif +#include <linux/fpu.h> /** * DOC: DC FPU manipulation overview @@ -87,20 +78,9 @@ void dc_fpu_begin(const char *function_name, const int line) WARN_ON_ONCE(!in_task()); preempt_disable(); depth = __this_cpu_inc_return(fpu_recursion_depth); - if (depth == 1) { -#if defined(CONFIG_X86) || defined(CONFIG_LOONGARCH) + BUG_ON(!kernel_fpu_available()); kernel_fpu_begin(); -#elif defined(CONFIG_PPC64) - if (cpu_has_feature(CPU_FTR_VSX_COMP)) - enable_kernel_vsx(); - else if (cpu_has_feature(CPU_FTR_ALTIVEC_COMP)) - enable_kernel_altivec(); - else if (!cpu_has_feature(CPU_FTR_FPU_UNAVAILABLE)) - enable_kernel_fp(); -#elif defined(CONFIG_ARM64) - kernel_neon_begin(); -#endif } TRACE_DCN_FPU(true, function_name, line, depth); @@ -122,18 +102,7 @@ void dc_fpu_end(const char *function_name, const int line) depth = __this_cpu_dec_return(fpu_recursion_depth); if (depth == 0) { -#if defined(CONFIG_X86) || defined(CONFIG_LOONGARCH) kernel_fpu_end(); -#elif defined(CONFIG_PPC64) - if (cpu_has_feature(CPU_FTR_VSX_COMP)) - disable_kernel_vsx(); - else if (cpu_has_feature(CPU_FTR_ALTIVEC_COMP)) - disable_kernel_altivec(); - else if (!cpu_has_feature(CPU_FTR_FPU_UNAVAILABLE)) - disable_kernel_fp(); -#elif defined(CONFIG_ARM64) - kernel_neon_end(); -#endif } else { WARN_ON_ONCE(depth < 0); } diff --git a/drivers/gpu/drm/amd/display/dc/Makefile b/drivers/gpu/drm/amd/display/dc/Makefile index 7991ae468f75..4e9fb1742877 100644 --- a/drivers/gpu/drm/amd/display/dc/Makefile +++ b/drivers/gpu/drm/amd/display/dc/Makefile @@ -22,7 +22,7 @@ # # Makefile for Display Core (dc) component. -DC_LIBS = basics bios dml clk_mgr dce gpio hwss irq link virtual dsc resource optc +DC_LIBS = basics bios dml clk_mgr dce gpio hwss irq link virtual dsc resource optc dpp ifdef CONFIG_DRM_AMD_DC_FP diff --git a/drivers/gpu/drm/amd/display/dc/bios/bios_parser.c b/drivers/gpu/drm/amd/display/dc/bios/bios_parser.c index 6450853fea94..bc16db69a663 100644 --- a/drivers/gpu/drm/amd/display/dc/bios/bios_parser.c +++ b/drivers/gpu/drm/amd/display/dc/bios/bios_parser.c @@ -44,8 +44,6 @@ #include "bios_parser_common.h" -#include "dc.h" - #define THREE_PERCENT_OF_10000 300 #define LAST_RECORD_TYPE 0xff @@ -1731,6 +1729,7 @@ static uint32_t get_ss_entry_number_from_internal_ss_info_tbl_v2_1( return 0; } + /** * get_ss_entry_number_from_internal_ss_info_tbl_V3_1 * Get Number of SpreadSpectrum Entry from the ASIC_InternalSS_Info table of diff --git a/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c b/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c index 05f392501c0a..9fe0020bcb9c 100644 --- a/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c +++ b/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c @@ -1594,8 +1594,6 @@ static bool bios_parser_is_device_id_supported( return (le16_to_cpu(bp->object_info_tbl.v1_5->supporteddevices) & mask) != 0; break; } - - return false; } static uint32_t bios_parser_get_ss_entry_number( @@ -2948,6 +2946,7 @@ static enum bp_result construct_integrated_info( result = get_integrated_info_v2_1(bp, info); break; case 2: + case 3: result = get_integrated_info_v2_2(bp, info); break; default: diff --git a/drivers/gpu/drm/amd/display/dc/bios/command_table.c b/drivers/gpu/drm/amd/display/dc/bios/command_table.c index 86f9198e7501..2bcae0643e61 100644 --- a/drivers/gpu/drm/amd/display/dc/bios/command_table.c +++ b/drivers/gpu/drm/amd/display/dc/bios/command_table.c @@ -399,7 +399,7 @@ static enum bp_result transmitter_control_v1_6( static void init_transmitter_control(struct bios_parser *bp) { uint8_t frev; - uint8_t crev; + uint8_t crev = 0; if (BIOS_CMD_TABLE_REVISION(UNIPHYTransmitterControl, frev, crev) == false) diff --git a/drivers/gpu/drm/amd/display/dc/bios/command_table2.c b/drivers/gpu/drm/amd/display/dc/bios/command_table2.c index cbae1be7b009..cc000833d300 100644 --- a/drivers/gpu/drm/amd/display/dc/bios/command_table2.c +++ b/drivers/gpu/drm/amd/display/dc/bios/command_table2.c @@ -225,7 +225,7 @@ static enum bp_result transmitter_control_fallback( static void init_transmitter_control(struct bios_parser *bp) { uint8_t frev; - uint8_t crev; + uint8_t crev = 0; BIOS_CMD_TABLE_REVISION(dig1transmittercontrol, frev, crev); diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/clk_mgr.c index 9f0f25aee426..a2b4ff2cff16 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/clk_mgr.c @@ -272,7 +272,7 @@ struct clk_mgr *dc_clk_mgr_create(struct dc_context *ctx, struct pp_smu_funcs *p dcn3_clk_mgr_construct(ctx, clk_mgr, pp_smu, dccg); return &clk_mgr->base; } - if (asic_id.chip_id == DEVICE_ID_NV_13FE) { + if (ctx->dce_version == DCN_VERSION_2_01) { dcn201_clk_mgr_construct(ctx, clk_mgr, pp_smu, dccg); return &clk_mgr->base; } @@ -329,15 +329,14 @@ struct clk_mgr *dc_clk_mgr_create(struct dc_context *ctx, struct pp_smu_funcs *p } break; case AMDGPU_FAMILY_GC_11_0_0: { - struct clk_mgr_internal *clk_mgr = kzalloc(sizeof(*clk_mgr), GFP_KERNEL); - - if (clk_mgr == NULL) { - BREAK_TO_DEBUGGER(); - return NULL; - } + struct clk_mgr_internal *clk_mgr = kzalloc(sizeof(*clk_mgr), GFP_KERNEL); - dcn32_clk_mgr_construct(ctx, clk_mgr, pp_smu, dccg); - return &clk_mgr->base; + if (clk_mgr == NULL) { + BREAK_TO_DEBUGGER(); + return NULL; + } + dcn32_clk_mgr_construct(ctx, clk_mgr, pp_smu, dccg); + return &clk_mgr->base; } case AMDGPU_FAMILY_GC_11_0_1: { diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dce100/dce_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dce100/dce_clk_mgr.c index b77804cfde0f..2a5dd3a296b2 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dce100/dce_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dce100/dce_clk_mgr.c @@ -131,8 +131,8 @@ int dce_get_dp_ref_freq_khz(struct clk_mgr *clk_mgr_base) struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base); int dprefclk_wdivider; int dprefclk_src_sel; - int dp_ref_clk_khz; - int target_div = 600000; + int dp_ref_clk_khz = 600000; + int target_div; /* ASSERT DP Reference Clock source is from DFS*/ REG_GET(DPREFCLK_CNTL, DPREFCLK_SRC_SEL, &dprefclk_src_sel); diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr.c index 2a74e2d74909..369421e46c52 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr.c @@ -23,7 +23,6 @@ * */ -#include "reg_helper.h" #include "core_types.h" #include "clk_mgr_internal.h" #include "rv1_clk_mgr.h" diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr_vbios_smu.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr_vbios_smu.c index 89b79dd39628..19897fa52e7e 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr_vbios_smu.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr_vbios_smu.c @@ -26,7 +26,6 @@ #include "core_types.h" #include "clk_mgr_internal.h" #include "reg_helper.h" -#include <linux/delay.h> #include "rv1_clk_mgr_vbios_smu.h" diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn20/dcn20_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn20/dcn20_clk_mgr.c index 5ee87965a078..bb4f3bd7532e 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn20/dcn20_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn20/dcn20_clk_mgr.c @@ -503,7 +503,7 @@ static void dcn2_notify_link_rate_change(struct clk_mgr *clk_mgr_base, struct dc clk_mgr->cur_phyclk_req_table[link->link_index] = link->cur_link_settings.link_rate * LINK_RATE_REF_FREQ_IN_KHZ; - for (i = 0; i < MAX_PIPES * 2; i++) { + for (i = 0; i < MAX_LINKS; i++) { if (clk_mgr->cur_phyclk_req_table[i] > max_phyclk_req) max_phyclk_req = clk_mgr->cur_phyclk_req_table[i]; } diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn201/dcn201_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn201/dcn201_clk_mgr.c index 9c90090e7351..f77840dd051e 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn201/dcn201_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn201/dcn201_clk_mgr.c @@ -100,7 +100,15 @@ static void dcn201_update_clocks(struct clk_mgr *clk_mgr_base, if (clk_mgr_base->clks.dispclk_khz == 0 || dc->debug.force_clock_mode & 0x1) { + /* this is from resume or boot up, if forced_clock cfg option + * used, we bypass program dispclk and DPPCLK, but need set them + * for S3. + */ + force_reset = true; + /* force_clock_mode 0x1: force reset the clock even it is the + * same clock as long as it is in Passive level. + */ dcn2_read_clocks_from_hw_dentist(clk_mgr_base); } @@ -150,11 +158,14 @@ static void dcn201_update_clocks(struct clk_mgr *clk_mgr_base, if (dc->config.forced_clocks == false || (force_reset && safe_to_lower)) { if (dpp_clock_lowered) { + // if clock is being lowered, increase DTO before lowering refclk dcn20_update_clocks_update_dpp_dto(clk_mgr, context, safe_to_lower); dcn20_update_clocks_update_dentist(clk_mgr, context); } else { + // if clock is being raised, increase refclk before lowering DTO if (update_dppclk || update_dispclk) dcn20_update_clocks_update_dentist(clk_mgr, context); + // always update dtos unless clock is lowered and not safe to lower dcn20_update_clocks_update_dpp_dto(clk_mgr, context, safe_to_lower); } } diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr.c index e3e1940198a9..5ef0879f6ad9 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr.c @@ -548,7 +548,7 @@ static void rn_notify_link_rate_change(struct clk_mgr *clk_mgr_base, struct dc_l clk_mgr->cur_phyclk_req_table[link->link_index] = link->cur_link_settings.link_rate * LINK_RATE_REF_FREQ_IN_KHZ; - for (i = 0; i < MAX_PIPES * 2; i++) { + for (i = 0; i < MAX_LINKS; i++) { if (clk_mgr->cur_phyclk_req_table[i] > max_phyclk_req) max_phyclk_req = clk_mgr->cur_phyclk_req_table[i]; } @@ -642,7 +642,8 @@ static void rn_clk_mgr_helper_populate_bw_params(struct clk_bw_params *bw_params j = -1; - ASSERT(PP_SMU_NUM_FCLK_DPM_LEVELS <= MAX_NUM_DPM_LVL); + static_assert(PP_SMU_NUM_FCLK_DPM_LEVELS <= MAX_NUM_DPM_LVL, + "number of reported FCLK DPM levels exceed maximum"); /* Find lowest DPM, FCLK is filled in reverse order*/ diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30_clk_mgr.c index 3271c8c7905d..8083a553c60e 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30_clk_mgr.c @@ -474,7 +474,7 @@ static void dcn30_notify_link_rate_change(struct clk_mgr *clk_mgr_base, struct d clk_mgr->cur_phyclk_req_table[link->link_index] = link->cur_link_settings.link_rate * LINK_RATE_REF_FREQ_IN_KHZ; - for (i = 0; i < MAX_PIPES * 2; i++) { + for (i = 0; i < MAX_LINKS; i++) { if (clk_mgr->cur_phyclk_req_table[i] > max_phyclk_req) max_phyclk_req = clk_mgr->cur_phyclk_req_table[i]; } @@ -560,11 +560,19 @@ void dcn3_clk_mgr_construct( dce_clock_read_ss_info(clk_mgr); clk_mgr->base.bw_params = kzalloc(sizeof(*clk_mgr->base.bw_params), GFP_KERNEL); + if (!clk_mgr->base.bw_params) { + BREAK_TO_DEBUGGER(); + return; + } /* need physical address of table to give to PMFW */ clk_mgr->wm_range_table = dm_helpers_allocate_gpu_mem(clk_mgr->base.ctx, DC_MEM_ALLOC_TYPE_GART, sizeof(WatermarksExternal_t), &clk_mgr->wm_range_table_addr); + if (!clk_mgr->wm_range_table) { + BREAK_TO_DEBUGGER(); + return; + } } void dcn3_clk_mgr_destroy(struct clk_mgr_internal *clk_mgr) diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30_clk_mgr_smu_msg.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30_clk_mgr_smu_msg.c index bdbf18306698..3253115a153d 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30_clk_mgr_smu_msg.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30_clk_mgr_smu_msg.c @@ -23,7 +23,6 @@ * */ -#include <linux/delay.h> #include "dcn30_clk_mgr_smu_msg.h" #include "clk_mgr_internal.h" @@ -54,6 +53,7 @@ */ static uint32_t dcn30_smu_wait_for_response(struct clk_mgr_internal *clk_mgr, unsigned int delay_us, unsigned int max_retries) { + const uint32_t initial_max_retries = max_retries; uint32_t reg = 0; do { @@ -69,7 +69,7 @@ static uint32_t dcn30_smu_wait_for_response(struct clk_mgr_internal *clk_mgr, un /* handle DALSMC_Result_CmdRejectedBusy? */ - /* Log? */ + TRACE_SMU_DELAY(delay_us * (initial_max_retries - max_retries), clk_mgr->base.ctx); return reg; } @@ -89,6 +89,8 @@ static bool dcn30_smu_send_msg_with_param(struct clk_mgr_internal *clk_mgr, uint /* Trigger the message transaction by writing the message ID */ REG_WRITE(DAL_MSG_REG, msg_id); + TRACE_SMU_MSG(msg_id, param_in, clk_mgr->base.ctx); + result = dcn30_smu_wait_for_response(clk_mgr, 10, 200000); if (IS_SMU_TIMEOUT(result)) { diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn301/vg_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn301/vg_clk_mgr.c index aa9fd1dc550a..191d8b969d19 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn301/vg_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn301/vg_clk_mgr.c @@ -566,7 +566,8 @@ static void vg_clk_mgr_helper_populate_bw_params( j = -1; - ASSERT(VG_NUM_FCLK_DPM_LEVELS <= MAX_NUM_DPM_LVL); + static_assert(VG_NUM_FCLK_DPM_LEVELS <= MAX_NUM_DPM_LVL, + "number of reported FCLK DPM levels exceeds maximum"); /* Find lowest DPM, FCLK is filled in reverse order*/ diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn31/dcn31_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn31/dcn31_clk_mgr.c index ce1386e22576..12a7752758b8 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn31/dcn31_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn31/dcn31_clk_mgr.c @@ -562,7 +562,8 @@ static void dcn31_clk_mgr_helper_populate_bw_params(struct clk_mgr_internal *clk j = -1; - ASSERT(NUM_DF_PSTATE_LEVELS <= MAX_NUM_DPM_LVL); + static_assert(NUM_DF_PSTATE_LEVELS <= MAX_NUM_DPM_LVL, + "number of reported pstate levels exceeds maximum"); /* Find lowest DPM, FCLK is filled in reverse order*/ diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn31/dcn31_smu.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn31/dcn31_smu.c index 6904e95113c1..f201628e4e98 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn31/dcn31_smu.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn31/dcn31_smu.c @@ -23,7 +23,6 @@ * */ -#include <linux/delay.h> #include "core_types.h" #include "clk_mgr_internal.h" #include "reg_helper.h" diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn314/dcn314_smu.h b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn314/dcn314_smu.h index 047d19ea919c..78ca1e5c5e9e 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn314/dcn314_smu.h +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn314/dcn314_smu.h @@ -37,34 +37,34 @@ typedef enum { } WCK_RATIO_e; typedef struct { - uint32_t FClk; - uint32_t MemClk; - uint32_t Voltage; - uint8_t WckRatio; - uint8_t Spare[3]; + uint32_t FClk; + uint32_t MemClk; + uint32_t Voltage; + uint8_t WckRatio; + uint8_t Spare[3]; } DfPstateTable314_t; //Freq in MHz //Voltage in milli volts with 2 fractional bits typedef struct { - uint32_t DcfClocks[NUM_DCFCLK_DPM_LEVELS]; - uint32_t DispClocks[NUM_DISPCLK_DPM_LEVELS]; - uint32_t DppClocks[NUM_DPPCLK_DPM_LEVELS]; - uint32_t SocClocks[NUM_SOCCLK_DPM_LEVELS]; - uint32_t VClocks[NUM_VCN_DPM_LEVELS]; - uint32_t DClocks[NUM_VCN_DPM_LEVELS]; - uint32_t SocVoltage[NUM_SOC_VOLTAGE_LEVELS]; - DfPstateTable314_t DfPstateTable[NUM_DF_PSTATE_LEVELS]; + uint32_t DcfClocks[NUM_DCFCLK_DPM_LEVELS]; + uint32_t DispClocks[NUM_DISPCLK_DPM_LEVELS]; + uint32_t DppClocks[NUM_DPPCLK_DPM_LEVELS]; + uint32_t SocClocks[NUM_SOCCLK_DPM_LEVELS]; + uint32_t VClocks[NUM_VCN_DPM_LEVELS]; + uint32_t DClocks[NUM_VCN_DPM_LEVELS]; + uint32_t SocVoltage[NUM_SOC_VOLTAGE_LEVELS]; + DfPstateTable314_t DfPstateTable[NUM_DF_PSTATE_LEVELS]; - uint8_t NumDcfClkLevelsEnabled; - uint8_t NumDispClkLevelsEnabled; //Applies to both Dispclk and Dppclk - uint8_t NumSocClkLevelsEnabled; - uint8_t VcnClkLevelsEnabled; //Applies to both Vclk and Dclk - uint8_t NumDfPstatesEnabled; - uint8_t spare[3]; + uint8_t NumDcfClkLevelsEnabled; + uint8_t NumDispClkLevelsEnabled; //Applies to both Dispclk and Dppclk + uint8_t NumSocClkLevelsEnabled; + uint8_t VcnClkLevelsEnabled; //Applies to both Vclk and Dclk + uint8_t NumDfPstatesEnabled; + uint8_t spare[3]; - uint32_t MinGfxClk; - uint32_t MaxGfxClk; + uint32_t MinGfxClk; + uint32_t MaxGfxClk; } DpmClocks314_t; struct dcn314_watermarks { diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn315/dcn315_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn315/dcn315_clk_mgr.c index 644da4637320..5506cf9b3672 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn315/dcn315_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn315/dcn315_clk_mgr.c @@ -145,6 +145,10 @@ static void dcn315_update_clocks(struct clk_mgr *clk_mgr_base, */ clk_mgr_base->clks.zstate_support = new_clocks->zstate_support; if (safe_to_lower) { + if (clk_mgr_base->clks.dtbclk_en && !new_clocks->dtbclk_en) { + dcn315_smu_set_dtbclk(clk_mgr, false); + clk_mgr_base->clks.dtbclk_en = new_clocks->dtbclk_en; + } /* check that we're not already in lower */ if (clk_mgr_base->clks.pwr_state != DCN_PWR_STATE_LOW_POWER) { display_count = dcn315_get_active_display_cnt_wa(dc, context); @@ -160,6 +164,10 @@ static void dcn315_update_clocks(struct clk_mgr *clk_mgr_base, } } } else { + if (!clk_mgr_base->clks.dtbclk_en && new_clocks->dtbclk_en) { + dcn315_smu_set_dtbclk(clk_mgr, true); + clk_mgr_base->clks.dtbclk_en = new_clocks->dtbclk_en; + } /* check that we're not already in D0 */ if (clk_mgr_base->clks.pwr_state != DCN_PWR_STATE_MISSION_MODE) { union display_idle_optimization_u idle_info = { 0 }; diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn315/dcn315_smu.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn315/dcn315_smu.c index 879f1494c4cd..2d14346b680e 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn315/dcn315_smu.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn315/dcn315_smu.c @@ -29,6 +29,7 @@ #include "dm_helpers.h" #include "dcn315_smu.h" #include "mp/mp_13_0_5_offset.h" +#include "logger_types.h" #define MAX_INSTANCE 6 #define MAX_SEGMENT 6 @@ -69,7 +70,6 @@ static const struct IP_BASE NBIO_BASE = { { { { 0x00000000, 0x00000014, 0x00000D #define REG_NBIO(reg_name) \ (NBIO_BASE.instance[0].segment[regBIF_BX_PF2_ ## reg_name ## _BASE_IDX] + regBIF_BX_PF2_ ## reg_name) -#include "logger_types.h" #undef DC_LOGGER #define DC_LOGGER \ CTX->logger diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn316/dcn316_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn316/dcn316_clk_mgr.c index 12f3e8aa46d8..20ca7afa9cb4 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn316/dcn316_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn316/dcn316_clk_mgr.c @@ -99,20 +99,25 @@ static int dcn316_get_active_display_cnt_wa( return display_count; } -static void dcn316_disable_otg_wa(struct clk_mgr *clk_mgr_base, struct dc_state *context, bool disable) +static void dcn316_disable_otg_wa(struct clk_mgr *clk_mgr_base, struct dc_state *context, + bool safe_to_lower, bool disable) { struct dc *dc = clk_mgr_base->ctx->dc; int i; for (i = 0; i < dc->res_pool->pipe_count; ++i) { - struct pipe_ctx *pipe = &dc->current_state->res_ctx.pipe_ctx[i]; + struct pipe_ctx *pipe = safe_to_lower + ? &context->res_ctx.pipe_ctx[i] + : &dc->current_state->res_ctx.pipe_ctx[i]; if (pipe->top_pipe || pipe->prev_odm_pipe) continue; - if (pipe->stream && (pipe->stream->dpms_off || pipe->plane_state == NULL || - dc_is_virtual_signal(pipe->stream->signal))) { + if (pipe->stream && (pipe->stream->dpms_off || dc_is_virtual_signal(pipe->stream->signal) || + !pipe->stream->link_enc)) { if (disable) { - pipe->stream_res.tg->funcs->immediate_disable_crtc(pipe->stream_res.tg); + if (pipe->stream_res.tg && pipe->stream_res.tg->funcs->immediate_disable_crtc) + pipe->stream_res.tg->funcs->immediate_disable_crtc(pipe->stream_res.tg); + reset_sync_context_for_pipe(dc, context, i); } else pipe->stream_res.tg->funcs->enable_crtc(pipe->stream_res.tg); @@ -207,11 +212,11 @@ static void dcn316_update_clocks(struct clk_mgr *clk_mgr_base, } if (should_set_clock(safe_to_lower, new_clocks->dispclk_khz, clk_mgr_base->clks.dispclk_khz)) { - dcn316_disable_otg_wa(clk_mgr_base, context, true); + dcn316_disable_otg_wa(clk_mgr_base, context, safe_to_lower, true); clk_mgr_base->clks.dispclk_khz = new_clocks->dispclk_khz; dcn316_smu_set_dispclk(clk_mgr, clk_mgr_base->clks.dispclk_khz); - dcn316_disable_otg_wa(clk_mgr_base, context, false); + dcn316_disable_otg_wa(clk_mgr_base, context, safe_to_lower, false); update_dispclk = true; } @@ -480,7 +485,8 @@ static void dcn316_clk_mgr_helper_populate_bw_params( j = -1; - ASSERT(NUM_DF_PSTATE_LEVELS <= MAX_NUM_DPM_LVL); + static_assert(NUM_DF_PSTATE_LEVELS <= MAX_NUM_DPM_LVL, + "number of reported pstate levels exceeds maximum"); /* Find lowest DPM, FCLK is filled in reverse order*/ diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn32/dcn32_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn32/dcn32_clk_mgr.c index bec252e1dd27..ff5fdc7b1198 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn32/dcn32_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn32/dcn32_clk_mgr.c @@ -29,6 +29,7 @@ #include "dcn20/dcn20_clk_mgr.h" #include "dce100/dce_clk_mgr.h" #include "dcn31/dcn31_clk_mgr.h" +#include "dcn32/dcn32_clk_mgr.h" #include "reg_helper.h" #include "core_types.h" #include "dm_helpers.h" @@ -40,7 +41,6 @@ #include "dcn/dcn_3_2_0_offset.h" #include "dcn/dcn_3_2_0_sh_mask.h" -#include "dcn32/dcn32_clk_mgr.h" #include "dml/dcn32/dcn32_fpu.h" #define DCN_BASE__INST0_SEG1 0x000000C0 @@ -712,8 +712,12 @@ static void dcn32_update_clocks(struct clk_mgr *clk_mgr_base, * since we calculate mode support based on softmax being the max UCLK * frequency. */ - dcn32_smu_set_hard_min_by_freq(clk_mgr, PPCLK_UCLK, - dc->clk_mgr->bw_params->dc_mode_softmax_memclk); + if (dc->debug.disable_dc_mode_overwrite) { + dcn30_smu_set_hard_max_by_freq(clk_mgr, PPCLK_UCLK, dc->clk_mgr->bw_params->max_memclk_mhz); + dcn32_smu_set_hard_min_by_freq(clk_mgr, PPCLK_UCLK, dc->clk_mgr->bw_params->max_memclk_mhz); + } else + dcn32_smu_set_hard_min_by_freq(clk_mgr, PPCLK_UCLK, + dc->clk_mgr->bw_params->dc_mode_softmax_memclk); } else { dcn32_smu_set_hard_min_by_freq(clk_mgr, PPCLK_UCLK, dc->clk_mgr->bw_params->max_memclk_mhz); } @@ -746,8 +750,13 @@ static void dcn32_update_clocks(struct clk_mgr *clk_mgr_base, /* set UCLK to requested value if P-State switching is supported, or to re-enable P-State switching */ if (clk_mgr_base->clks.p_state_change_support && (update_uclk || !clk_mgr_base->clks.prev_p_state_change_support) && - !dc->work_arounds.clock_update_disable_mask.uclk) + !dc->work_arounds.clock_update_disable_mask.uclk) { + if (dc->clk_mgr->dc_mode_softmax_enabled && dc->debug.disable_dc_mode_overwrite) + dcn30_smu_set_hard_max_by_freq(clk_mgr, PPCLK_UCLK, + max((int)dc->clk_mgr->bw_params->dc_mode_softmax_memclk, khz_to_mhz_ceil(clk_mgr_base->clks.dramclk_khz))); + dcn32_smu_set_hard_min_by_freq(clk_mgr, PPCLK_UCLK, khz_to_mhz_ceil(clk_mgr_base->clks.dramclk_khz)); + } if (clk_mgr_base->clks.num_ways != new_clocks->num_ways && clk_mgr_base->clks.num_ways > new_clocks->num_ways) { @@ -829,7 +838,7 @@ static void dcn32_update_clocks(struct clk_mgr *clk_mgr_base, dmcu->funcs->set_psr_wait_loop(dmcu, clk_mgr_base->clks.dispclk_khz / 1000 / 7); - if (dc->config.enable_auto_dpm_test_logs && safe_to_lower) { + if (dc->config.enable_auto_dpm_test_logs) { dcn32_auto_dpm_test_log(new_clocks, clk_mgr, context); } } @@ -1199,11 +1208,19 @@ void dcn32_clk_mgr_construct( clk_mgr->smu_present = false; clk_mgr->base.bw_params = kzalloc(sizeof(*clk_mgr->base.bw_params), GFP_KERNEL); + if (!clk_mgr->base.bw_params) { + BREAK_TO_DEBUGGER(); + return; + } /* need physical address of table to give to PMFW */ clk_mgr->wm_range_table = dm_helpers_allocate_gpu_mem(clk_mgr->base.ctx, DC_MEM_ALLOC_TYPE_GART, sizeof(WatermarksExternal_t), &clk_mgr->wm_range_table_addr); + if (!clk_mgr->wm_range_table) { + BREAK_TO_DEBUGGER(); + return; + } } void dcn32_clk_mgr_destroy(struct clk_mgr_internal *clk_mgr) diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn32/dcn32_clk_mgr_smu_msg.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn32/dcn32_clk_mgr_smu_msg.c index df244b175fdb..f2f60478b1a6 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn32/dcn32_clk_mgr_smu_msg.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn32/dcn32_clk_mgr_smu_msg.c @@ -49,6 +49,7 @@ */ static uint32_t dcn32_smu_wait_for_response(struct clk_mgr_internal *clk_mgr, unsigned int delay_us, unsigned int max_retries) { + const uint32_t initial_max_retries = max_retries; uint32_t reg = 0; do { @@ -62,6 +63,8 @@ static uint32_t dcn32_smu_wait_for_response(struct clk_mgr_internal *clk_mgr, un udelay(delay_us); } while (max_retries--); + TRACE_SMU_DELAY(delay_us * (initial_max_retries - max_retries), clk_mgr->base.ctx); + return reg; } @@ -79,6 +82,8 @@ static bool dcn32_smu_send_msg_with_param(struct clk_mgr_internal *clk_mgr, uint /* Trigger the message transaction by writing the message ID */ REG_WRITE(DAL_MSG_REG, msg_id); + TRACE_SMU_MSG(msg_id, param_in, clk_mgr->base.ctx); + /* Wait for response */ if (dcn32_smu_wait_for_response(clk_mgr, 10, 200000) == DALSMC_Result_OK) { if (param_out) @@ -115,6 +120,8 @@ static uint32_t dcn32_smu_wait_for_response_delay(struct clk_mgr_internal *clk_m *total_delay_us += delay_us; } while (max_retries--); + TRACE_SMU_DELAY(*total_delay_us, clk_mgr->base.ctx); + return reg; } @@ -135,6 +142,8 @@ static bool dcn32_smu_send_msg_with_param_delay(struct clk_mgr_internal *clk_mgr /* Trigger the message transaction by writing the message ID */ REG_WRITE(DAL_MSG_REG, msg_id); + TRACE_SMU_MSG(msg_id, param_in, clk_mgr->base.ctx); + /* Wait for response */ if (dcn32_smu_wait_for_response_delay(clk_mgr, 10, 200000, &delay2_us) == DALSMC_Result_OK) { if (param_out) diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn32/dcn32_clk_mgr_smu_msg.h b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn32/dcn32_clk_mgr_smu_msg.h index c76352a817de..5c44ab0e8667 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn32/dcn32_clk_mgr_smu_msg.h +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn32/dcn32_clk_mgr_smu_msg.h @@ -37,10 +37,9 @@ #define DALSMC_Result_OK 0x1 void dcn32_smu_send_fclk_pstate_message(struct clk_mgr_internal *clk_mgr, bool enable); -void dcn32_smu_transfer_wm_table_dram_2_smu(struct clk_mgr_internal *clk_mgr); -void dcn32_smu_set_pme_workaround(struct clk_mgr_internal *clk_mgr); void dcn32_smu_send_cab_for_uclk_message(struct clk_mgr_internal *clk_mgr, unsigned int num_ways); void dcn32_smu_transfer_wm_table_dram_2_smu(struct clk_mgr_internal *clk_mgr); +void dcn32_smu_set_pme_workaround(struct clk_mgr_internal *clk_mgr); unsigned int dcn32_smu_set_hard_min_by_freq(struct clk_mgr_internal *clk_mgr, uint32_t clk, uint16_t freq_mhz); void dcn32_smu_wait_for_dmub_ack_mclk(struct clk_mgr_internal *clk_mgr, bool enable); diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_clk_mgr.c index c378b879c76d..6c9b4e6491a5 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_clk_mgr.c @@ -73,6 +73,14 @@ #define CLK1_CLK2_BYPASS_CNTL__CLK2_BYPASS_SEL_MASK 0x00000007L #define CLK1_CLK2_BYPASS_CNTL__CLK2_BYPASS_DIV_MASK 0x000F0000L +#define regCLK5_0_CLK5_spll_field_8 0x464b +#define regCLK5_0_CLK5_spll_field_8_BASE_IDX 0 + +#define CLK5_0_CLK5_spll_field_8__spll_ssc_en__SHIFT 0xd +#define CLK5_0_CLK5_spll_field_8__spll_ssc_en_MASK 0x00002000L + +#define SMU_VER_THRESHOLD 0x5D4A00 //93.74.0 + #define REG(reg_name) \ (ctx->clk_reg_offsets[reg ## reg_name ## _BASE_IDX] + reg ## reg_name) @@ -244,7 +252,8 @@ void dcn35_update_clocks(struct clk_mgr *clk_mgr_base, } if (clk_mgr_base->clks.dtbclk_en && !new_clocks->dtbclk_en) { - dcn35_smu_set_dtbclk(clk_mgr, false); + if (clk_mgr->base.ctx->dc->config.allow_0_dtb_clk) + dcn35_smu_set_dtbclk(clk_mgr, false); clk_mgr_base->clks.dtbclk_en = new_clocks->dtbclk_en; } /* check that we're not already in lower */ @@ -409,11 +418,25 @@ static void dcn35_dump_clk_registers(struct clk_state_registers_and_bypass *regs { } +static bool dcn35_is_spll_ssc_enabled(struct clk_mgr *clk_mgr_base) +{ + struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base); + struct dc_context *ctx = clk_mgr->base.ctx; + uint32_t ssc_enable; + + REG_GET(CLK5_0_CLK5_spll_field_8, spll_ssc_en, &ssc_enable); + + return ssc_enable == 1; +} + static void init_clk_states(struct clk_mgr *clk_mgr) { + struct clk_mgr_internal *clk_mgr_int = TO_CLK_MGR_INTERNAL(clk_mgr); uint32_t ref_dtbclk = clk_mgr->clks.ref_dtbclk_khz; memset(&(clk_mgr->clks), 0, sizeof(struct dc_clocks)); + if (clk_mgr_int->smu_ver >= SMU_VER_THRESHOLD) + clk_mgr->clks.dtbclk_en = true; // request DTBCLK disable on first commit clk_mgr->clks.ref_dtbclk_khz = ref_dtbclk; // restore ref_dtbclk clk_mgr->clks.p_state_change_support = true; clk_mgr->clks.prev_p_state_change_support = true; @@ -423,7 +446,16 @@ static void init_clk_states(struct clk_mgr *clk_mgr) void dcn35_init_clocks(struct clk_mgr *clk_mgr) { + struct clk_mgr_internal *clk_mgr_int = TO_CLK_MGR_INTERNAL(clk_mgr); init_clk_states(clk_mgr); + + // to adjust dp_dto reference clock if ssc is enable otherwise to apply dprefclk + if (dcn35_is_spll_ssc_enabled(clk_mgr)) + clk_mgr->dp_dto_source_clock_in_khz = + dce_adjust_dp_ref_freq_for_ss(clk_mgr_int, clk_mgr->dprefclk_khz); + else + clk_mgr->dp_dto_source_clock_in_khz = clk_mgr->dprefclk_khz; + } static struct clk_bw_params dcn35_bw_params = { .vram_type = Ddr4MemType, @@ -512,6 +544,28 @@ static DpmClocks_t_dcn35 dummy_clocks; static struct dcn35_watermarks dummy_wms = { 0 }; +static struct dcn35_ss_info_table ss_info_table = { + .ss_divider = 1000, + .ss_percentage = {0, 0, 375, 375, 375} +}; + +static void dcn35_read_ss_info_from_lut(struct clk_mgr_internal *clk_mgr) +{ + struct dc_context *ctx = clk_mgr->base.ctx; + uint32_t clock_source; + + REG_GET(CLK1_CLK2_BYPASS_CNTL, CLK2_BYPASS_SEL, &clock_source); + // If it's DFS mode, clock_source is 0. + if (dcn35_is_spll_ssc_enabled(&clk_mgr->base) && (clock_source < ARRAY_SIZE(ss_info_table.ss_percentage))) { + clk_mgr->dprefclk_ss_percentage = ss_info_table.ss_percentage[clock_source]; + + if (clk_mgr->dprefclk_ss_percentage != 0) { + clk_mgr->ss_on_dprefclk = true; + clk_mgr->dprefclk_ss_divider = ss_info_table.ss_divider; + } + } +} + static void dcn35_build_watermark_ranges(struct clk_bw_params *bw_params, struct dcn35_watermarks *table) { int i, num_valid_sets; @@ -709,7 +763,7 @@ static void dcn35_clk_mgr_helper_populate_bw_params(struct clk_mgr_internal *clk clock_table->NumFclkLevelsEnabled; max_fclk = find_max_clk_value(clock_table->FclkClocks_Freq, num_fclk); - num_dcfclk = (clock_table->NumFclkLevelsEnabled > NUM_DCFCLK_DPM_LEVELS) ? NUM_DCFCLK_DPM_LEVELS : + num_dcfclk = (clock_table->NumDcfClkLevelsEnabled > NUM_DCFCLK_DPM_LEVELS) ? NUM_DCFCLK_DPM_LEVELS : clock_table->NumDcfClkLevelsEnabled; for (i = 0; i < num_dcfclk; i++) { int j; @@ -836,35 +890,6 @@ static void dcn35_set_low_power_state(struct clk_mgr *clk_mgr_base) } } -static void dcn35_set_ips_idle_state(struct clk_mgr *clk_mgr_base, bool allow_idle) -{ - struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base); - struct dc *dc = clk_mgr_base->ctx->dc; - uint32_t val = dcn35_smu_read_ips_scratch(clk_mgr); - - if (dc->config.disable_ips == DMUB_IPS_ENABLE || - dc->config.disable_ips == DMUB_IPS_DISABLE_DYNAMIC) { - val = val & ~DMUB_IPS1_ALLOW_MASK; - val = val & ~DMUB_IPS2_ALLOW_MASK; - } else if (dc->config.disable_ips == DMUB_IPS_DISABLE_IPS1) { - val |= DMUB_IPS1_ALLOW_MASK; - val |= DMUB_IPS2_ALLOW_MASK; - } else if (dc->config.disable_ips == DMUB_IPS_DISABLE_IPS2) { - val = val & ~DMUB_IPS1_ALLOW_MASK; - val |= DMUB_IPS2_ALLOW_MASK; - } else if (dc->config.disable_ips == DMUB_IPS_DISABLE_IPS2_Z10) { - val = val & ~DMUB_IPS1_ALLOW_MASK; - val = val & ~DMUB_IPS2_ALLOW_MASK; - } - - if (!allow_idle) { - val |= DMUB_IPS1_ALLOW_MASK; - val |= DMUB_IPS2_ALLOW_MASK; - } - - dcn35_smu_write_ips_scratch(clk_mgr, val); -} - static void dcn35_exit_low_power_state(struct clk_mgr *clk_mgr_base) { struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base); @@ -884,13 +909,6 @@ static bool dcn35_is_ips_supported(struct clk_mgr *clk_mgr_base) return ips_supported; } -static uint32_t dcn35_get_ips_idle_state(struct clk_mgr *clk_mgr_base) -{ - struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base); - - return dcn35_smu_read_ips_scratch(clk_mgr); -} - static void dcn35_init_clocks_fpga(struct clk_mgr *clk_mgr) { init_clk_states(clk_mgr); @@ -978,8 +996,6 @@ static struct clk_mgr_funcs dcn35_funcs = { .set_low_power_state = dcn35_set_low_power_state, .exit_low_power_state = dcn35_exit_low_power_state, .is_ips_supported = dcn35_is_ips_supported, - .set_idle_state = dcn35_set_ips_idle_state, - .get_idle_state = dcn35_get_ips_idle_state }; struct clk_mgr_funcs dcn35_fpga_funcs = { @@ -1056,6 +1072,8 @@ void dcn35_clk_mgr_construct( dce_clock_read_ss_info(&clk_mgr->base); /*when clk src is from FCH, it could have ss, same clock src as DPREF clk*/ + dcn35_read_ss_info_from_lut(&clk_mgr->base); + clk_mgr->base.base.bw_params = &dcn35_bw_params; if (clk_mgr->base.base.ctx->dc->debug.pstate_enabled) { diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_smu.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_smu.c index 9e588c56c570..1399b41dfd1c 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_smu.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_smu.c @@ -487,24 +487,3 @@ int dcn35_smu_get_ips_supported(struct clk_mgr_internal *clk_mgr) //smu_print("%s: VBIOSSMC_MSG_QueryIPS2Support return = %x\n", __func__, retv); return retv; } - -void dcn35_smu_write_ips_scratch(struct clk_mgr_internal *clk_mgr, uint32_t param) -{ - if (!clk_mgr->smu_present) - return; - - REG_WRITE(MP1_SMN_C2PMSG_71, param); - //smu_print("%s: write_ips_scratch = %x\n", __func__, param); -} - -uint32_t dcn35_smu_read_ips_scratch(struct clk_mgr_internal *clk_mgr) -{ - uint32_t retv; - - if (!clk_mgr->smu_present) - return 0; - - retv = REG_READ(MP1_SMN_C2PMSG_71); - //smu_print("%s: dcn35_smu_read_ips_scratch = %x\n", __func__, retv); - return retv; -} diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_smu.h b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_smu.h index 2b8e6959a03d..06cd3cc6d36e 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_smu.h +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_smu.h @@ -198,6 +198,4 @@ int dcn35_smu_exit_low_power_state(struct clk_mgr_internal *clk_mgr); int dcn35_smu_get_ips_supported(struct clk_mgr_internal *clk_mgr); int dcn35_smu_get_dtbclk(struct clk_mgr_internal *clk_mgr); int dcn35_smu_get_dprefclk(struct clk_mgr_internal *clk_mgr); -void dcn35_smu_write_ips_scratch(struct clk_mgr_internal *clk_mgr, uint32_t param); -uint32_t dcn35_smu_read_ips_scratch(struct clk_mgr_internal *clk_mgr); #endif /* DAL_DC_35_SMU_H_ */ diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c index e7dc128f6284..236876d95185 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc.c @@ -36,6 +36,7 @@ #include "resource.h" #include "dc_state.h" #include "dc_state_priv.h" +#include "dc_plane_priv.h" #include "gpio_service_interface.h" #include "clk_mgr.h" @@ -212,7 +213,8 @@ static bool create_links( connectors_num, num_virtual_links); - for (i = 0; i < connectors_num; i++) { + // condition loop on link_count to allow skipping invalid indices + for (i = 0; dc->link_count < connectors_num && i < MAX_LINKS; i++) { struct link_init_data link_init_params = {0}; struct dc_link *link; @@ -386,6 +388,30 @@ static void dc_perf_trace_destroy(struct dc_perf_trace **perf_trace) *perf_trace = NULL; } +static bool set_long_vtotal(struct dc *dc, struct dc_stream_state *stream, struct dc_crtc_timing_adjust *adjust) +{ + if (!dc || !stream || !adjust) + return false; + + if (!dc->current_state) + return false; + + int i; + + for (i = 0; i < MAX_PIPES; i++) { + struct pipe_ctx *pipe = &dc->current_state->res_ctx.pipe_ctx[i]; + + if (pipe->stream == stream && pipe->stream_res.tg) { + if (dc->hwss.set_long_vtotal) + dc->hwss.set_long_vtotal(&pipe, 1, adjust->v_total_min, adjust->v_total_max); + + return true; + } + } + + return false; +} + /** * dc_stream_adjust_vmin_vmax - look up pipe context & update parts of DRR * @dc: dc reference @@ -420,6 +446,15 @@ bool dc_stream_adjust_vmin_vmax(struct dc *dc, stream->adjust.v_total_mid = adjust->v_total_mid; stream->adjust.v_total_mid_frame_num = adjust->v_total_mid_frame_num; stream->adjust.v_total_min = adjust->v_total_min; + stream->adjust.allow_otg_v_count_halt = adjust->allow_otg_v_count_halt; + + if (dc->caps.max_v_total != 0 && + (adjust->v_total_max > dc->caps.max_v_total || adjust->v_total_min > dc->caps.max_v_total)) { + if (adjust->allow_otg_v_count_halt) + return set_long_vtotal(dc, stream, adjust); + else + return false; + } for (i = 0; i < MAX_PIPES; i++) { struct pipe_ctx *pipe = &dc->current_state->res_ctx.pipe_ctx[i]; @@ -1055,8 +1090,7 @@ static bool dc_construct(struct dc *dc, * is initialized in dc_create_resource_pool because * on creation it copies the contents of dc->dml */ - - dc->current_state = dc_state_create(dc); + dc->current_state = dc_state_create(dc, NULL); if (!dc->current_state) { dm_error("%s: failed to create validate ctx\n", __func__); @@ -1272,7 +1306,7 @@ static void disable_vbios_mode_if_required( if (link != NULL && link->link_enc->funcs->is_dig_enabled(link->link_enc)) { unsigned int enc_inst, tg_inst = 0; - unsigned int pix_clk_100hz; + unsigned int pix_clk_100hz = 0; enc_inst = link->link_enc->funcs->get_dig_frontend(link->link_enc); if (enc_inst != ENGINE_ID_UNKNOWN) { @@ -1759,7 +1793,7 @@ bool dc_validate_boot_timing(const struct dc *dc, return false; if (dc_is_dp_signal(link->connector_signal)) { - unsigned int pix_clk_100hz; + unsigned int pix_clk_100hz = 0; uint32_t numOdmPipes = 1; uint32_t id_src[4] = {0}; @@ -1801,6 +1835,9 @@ bool dc_validate_boot_timing(const struct dc *dc, return false; } + if (link->dpcd_caps.channel_coding_cap.bits.DP_128b_132b_SUPPORTED) + return false; + if (dc->link_srv->edp_is_ilr_optimization_required(link, crtc_timing)) { DC_LOG_EVENT_LINK_TRAINING("Seamless boot disabled to optimize eDP link rate\n"); return false; @@ -2085,15 +2122,14 @@ static enum dc_status dc_commit_state_no_check(struct dc *dc, struct dc_state *c return result; } -static bool commit_minimal_transition_state_legacy(struct dc *dc, +static bool commit_minimal_transition_state(struct dc *dc, struct dc_state *transition_base_context); /** * dc_commit_streams - Commit current stream state * * @dc: DC object with the commit state to be configured in the hardware - * @streams: Array with a list of stream state - * @stream_count: Total of streams + * @params: Parameters for the commit, including the streams to be committed * * Function responsible for commit streams change to the hardware. * @@ -2101,9 +2137,7 @@ static bool commit_minimal_transition_state_legacy(struct dc *dc, * Return DC_OK if everything work as expected, otherwise, return a dc_status * code. */ -enum dc_status dc_commit_streams(struct dc *dc, - struct dc_stream_state *streams[], - uint8_t stream_count) +enum dc_status dc_commit_streams(struct dc *dc, struct dc_commit_streams_params *params) { int i, j; struct dc_state *context; @@ -2112,18 +2146,22 @@ enum dc_status dc_commit_streams(struct dc *dc, struct pipe_ctx *pipe; bool handle_exit_odm2to1 = false; + if (!params) + return DC_ERROR_UNEXPECTED; + if (dc->ctx->dce_environment == DCE_ENV_VIRTUAL_HW) return res; - if (!streams_changed(dc, streams, stream_count)) + if (!streams_changed(dc, params->streams, params->stream_count) && + dc->current_state->power_source == params->power_source) return res; dc_exit_ips_for_hw_access(dc); - DC_LOG_DC("%s: %d streams\n", __func__, stream_count); + DC_LOG_DC("%s: %d streams\n", __func__, params->stream_count); - for (i = 0; i < stream_count; i++) { - struct dc_stream_state *stream = streams[i]; + for (i = 0; i < params->stream_count; i++) { + struct dc_stream_state *stream = params->streams[i]; struct dc_stream_status *status = dc_stream_get_status(stream); dc_stream_log(dc, stream); @@ -2141,7 +2179,7 @@ enum dc_status dc_commit_streams(struct dc *dc, * scenario, it uses extra pipes than needed to reduce power consumption * We need to switch off this feature to make room for new streams. */ - if (stream_count > dc->current_state->stream_count && + if (params->stream_count > dc->current_state->stream_count && dc->current_state->stream_count == 1) { for (i = 0; i < dc->res_pool->pipe_count; i++) { pipe = &dc->current_state->res_ctx.pipe_ctx[i]; @@ -2151,13 +2189,15 @@ enum dc_status dc_commit_streams(struct dc *dc, } if (handle_exit_odm2to1) - res = commit_minimal_transition_state_legacy(dc, dc->current_state); + res = commit_minimal_transition_state(dc, dc->current_state); context = dc_state_create_current_copy(dc); if (!context) goto context_alloc_fail; - res = dc_validate_with_context(dc, set, stream_count, context, false); + context->power_source = params->power_source; + + res = dc_validate_with_context(dc, set, params->stream_count, context, false); if (res != DC_OK) { BREAK_TO_DEBUGGER(); goto fail; @@ -2165,16 +2205,16 @@ enum dc_status dc_commit_streams(struct dc *dc, res = dc_commit_state_no_check(dc, context); - for (i = 0; i < stream_count; i++) { + for (i = 0; i < params->stream_count; i++) { for (j = 0; j < context->stream_count; j++) { - if (streams[i]->stream_id == context->streams[j]->stream_id) - streams[i]->out.otg_offset = context->stream_status[j].primary_otg_inst; + if (params->streams[i]->stream_id == context->streams[j]->stream_id) + params->streams[i]->out.otg_offset = context->stream_status[j].primary_otg_inst; - if (dc_is_embedded_signal(streams[i]->signal)) { - struct dc_stream_status *status = dc_state_get_stream_status(context, streams[i]); + if (dc_is_embedded_signal(params->streams[i]->signal)) { + struct dc_stream_status *status = dc_state_get_stream_status(context, params->streams[i]); if (dc->hwss.is_abm_supported) - status->is_abm_supported = dc->hwss.is_abm_supported(dc, context, streams[i]); + status->is_abm_supported = dc->hwss.is_abm_supported(dc, context, params->streams[i]); else status->is_abm_supported = true; } @@ -2818,55 +2858,45 @@ static void copy_surface_update_to_plane( srf_update->plane_info->layer_index; } - if (srf_update->gamma && - (surface->gamma_correction != - srf_update->gamma)) { - memcpy(&surface->gamma_correction->entries, + if (srf_update->gamma) { + memcpy(&surface->gamma_correction.entries, &srf_update->gamma->entries, sizeof(struct dc_gamma_entries)); - surface->gamma_correction->is_identity = + surface->gamma_correction.is_identity = srf_update->gamma->is_identity; - surface->gamma_correction->num_entries = + surface->gamma_correction.num_entries = srf_update->gamma->num_entries; - surface->gamma_correction->type = + surface->gamma_correction.type = srf_update->gamma->type; } - if (srf_update->in_transfer_func && - (surface->in_transfer_func != - srf_update->in_transfer_func)) { - surface->in_transfer_func->sdr_ref_white_level = + if (srf_update->in_transfer_func) { + surface->in_transfer_func.sdr_ref_white_level = srf_update->in_transfer_func->sdr_ref_white_level; - surface->in_transfer_func->tf = + surface->in_transfer_func.tf = srf_update->in_transfer_func->tf; - surface->in_transfer_func->type = + surface->in_transfer_func.type = srf_update->in_transfer_func->type; - memcpy(&surface->in_transfer_func->tf_pts, + memcpy(&surface->in_transfer_func.tf_pts, &srf_update->in_transfer_func->tf_pts, sizeof(struct dc_transfer_func_distributed_points)); } - if (srf_update->func_shaper && - (surface->in_shaper_func != - srf_update->func_shaper)) - memcpy(surface->in_shaper_func, srf_update->func_shaper, - sizeof(*surface->in_shaper_func)); + if (srf_update->func_shaper) + memcpy(&surface->in_shaper_func, srf_update->func_shaper, + sizeof(surface->in_shaper_func)); - if (srf_update->lut3d_func && - (surface->lut3d_func != - srf_update->lut3d_func)) - memcpy(surface->lut3d_func, srf_update->lut3d_func, - sizeof(*surface->lut3d_func)); + if (srf_update->lut3d_func) + memcpy(&surface->lut3d_func, srf_update->lut3d_func, + sizeof(surface->lut3d_func)); if (srf_update->hdr_mult.value) surface->hdr_mult = srf_update->hdr_mult; - if (srf_update->blend_tf && - (surface->blend_tf != - srf_update->blend_tf)) - memcpy(surface->blend_tf, srf_update->blend_tf, - sizeof(*surface->blend_tf)); + if (srf_update->blend_tf) + memcpy(&surface->blend_tf, srf_update->blend_tf, + sizeof(surface->blend_tf)); if (srf_update->input_csc_color_matrix) surface->input_csc_color_matrix = @@ -2897,14 +2927,13 @@ static void copy_stream_update_to_stream(struct dc *dc, if (update->dst.height && update->dst.width) stream->dst = update->dst; - if (update->out_transfer_func && - stream->out_transfer_func != update->out_transfer_func) { - stream->out_transfer_func->sdr_ref_white_level = + if (update->out_transfer_func) { + stream->out_transfer_func.sdr_ref_white_level = update->out_transfer_func->sdr_ref_white_level; - stream->out_transfer_func->tf = update->out_transfer_func->tf; - stream->out_transfer_func->type = + stream->out_transfer_func.tf = update->out_transfer_func->tf; + stream->out_transfer_func.type = update->out_transfer_func->type; - memcpy(&stream->out_transfer_func->tf_pts, + memcpy(&stream->out_transfer_func.tf_pts, &update->out_transfer_func->tf_pts, sizeof(struct dc_transfer_func_distributed_points)); } @@ -3017,14 +3046,8 @@ static void backup_planes_and_stream_state( for (i = 0; i < status->plane_count; i++) { scratch->plane_states[i] = *status->plane_states[i]; - scratch->gamma_correction[i] = *status->plane_states[i]->gamma_correction; - scratch->in_transfer_func[i] = *status->plane_states[i]->in_transfer_func; - scratch->lut3d_func[i] = *status->plane_states[i]->lut3d_func; - scratch->in_shaper_func[i] = *status->plane_states[i]->in_shaper_func; - scratch->blend_tf[i] = *status->plane_states[i]->blend_tf; } scratch->stream_state = *stream; - scratch->out_transfer_func = *stream->out_transfer_func; } static void restore_planes_and_stream_state( @@ -3039,16 +3062,67 @@ static void restore_planes_and_stream_state( for (i = 0; i < status->plane_count; i++) { *status->plane_states[i] = scratch->plane_states[i]; - *status->plane_states[i]->gamma_correction = scratch->gamma_correction[i]; - *status->plane_states[i]->in_transfer_func = scratch->in_transfer_func[i]; - *status->plane_states[i]->lut3d_func = scratch->lut3d_func[i]; - *status->plane_states[i]->in_shaper_func = scratch->in_shaper_func[i]; - *status->plane_states[i]->blend_tf = scratch->blend_tf[i]; } *stream = scratch->stream_state; - *stream->out_transfer_func = scratch->out_transfer_func; } +/** + * update_seamless_boot_flags() - Helper function for updating seamless boot flags + * + * @dc: Current DC state + * @context: New DC state to be programmed + * @surface_count: Number of surfaces that have an updated + * @stream: Corresponding stream to be updated in the current flip + * + * Updating seamless boot flags do not need to be part of the commit sequence. This + * helper function will update the seamless boot flags on each flip (if required) + * outside of the HW commit sequence (fast or slow). + * + * Return: void + */ +static void update_seamless_boot_flags(struct dc *dc, + struct dc_state *context, + int surface_count, + struct dc_stream_state *stream) +{ + if (get_seamless_boot_stream_count(context) > 0 && surface_count > 0) { + /* Optimize seamless boot flag keeps clocks and watermarks high until + * first flip. After first flip, optimization is required to lower + * bandwidth. Important to note that it is expected UEFI will + * only light up a single display on POST, therefore we only expect + * one stream with seamless boot flag set. + */ + if (stream->apply_seamless_boot_optimization) { + stream->apply_seamless_boot_optimization = false; + + if (get_seamless_boot_stream_count(context) == 0) + dc->optimized_required = true; + } + } +} + +/** + * update_planes_and_stream_state() - The function takes planes and stream + * updates as inputs and determines the appropriate update type. If update type + * is FULL, the function allocates a new context, populates and validates it. + * Otherwise, it updates current dc context. The function will return both + * new_context and new_update_type back to the caller. The function also backs + * up both current and new contexts into corresponding dc state scratch memory. + * TODO: The function does too many things, and even conditionally allocates dc + * context memory implicitly. We should consider to break it down. + * + * @dc: Current DC state + * @srf_updates: an array of surface updates + * @surface_count: surface update count + * @stream: Corresponding stream to be updated + * @stream_update: stream update + * @new_update_type: [out] determined update type by the function + * @new_context: [out] new context allocated and validated if update type is + * FULL, reference to current context if update type is less than FULL. + * + * Return: true if a valid update is populated into new_context, false + * otherwise. + */ static bool update_planes_and_stream_state(struct dc *dc, struct dc_surface_update *srf_updates, int surface_count, struct dc_stream_state *stream, @@ -3072,9 +3146,10 @@ static bool update_planes_and_stream_state(struct dc *dc, } context = dc->current_state; - backup_planes_and_stream_state(&dc->current_state->scratch, stream); update_type = dc_check_update_surfaces_for_stream( dc, srf_updates, surface_count, stream_update, stream_status); + if (update_type == UPDATE_TYPE_FULL) + backup_planes_and_stream_state(&dc->scratch.current_state, stream); /* update current stream with the new updates */ copy_stream_update_to_stream(dc, context, stream, stream_update); @@ -3143,7 +3218,10 @@ static bool update_planes_and_stream_state(struct dc *dc, for (i = 0; i < surface_count; i++) { struct dc_plane_state *surface = srf_updates[i].surface; - if (update_type >= UPDATE_TYPE_MED) { + if (update_type != UPDATE_TYPE_MED) + continue; + if (surface->update_flags.bits.clip_size_change || + surface->update_flags.bits.position_change) { for (j = 0; j < dc->res_pool->pipe_count; j++) { struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[j]; @@ -3160,19 +3238,13 @@ static bool update_planes_and_stream_state(struct dc *dc, BREAK_TO_DEBUGGER(); goto fail; } - - for (i = 0; i < context->stream_count; i++) { - struct pipe_ctx *otg_master = resource_get_otg_master_for_stream(&context->res_ctx, - context->streams[i]); - - if (otg_master && otg_master->stream->test_pattern.type != DP_TEST_PATTERN_VIDEO_MODE) - resource_build_test_pattern_params(&context->res_ctx, otg_master); - } } + update_seamless_boot_flags(dc, context, surface_count, stream); *new_context = context; *new_update_type = update_type; - backup_planes_and_stream_state(&context->scratch, stream); + if (update_type == UPDATE_TYPE_FULL) + backup_planes_and_stream_state(&dc->scratch.new_state, stream); return true; @@ -3261,12 +3333,26 @@ static void commit_planes_do_stream_update(struct dc *dc, } if (stream_update->pending_test_pattern) { - dc_link_dp_set_test_pattern(stream->link, + /* + * test pattern params depends on ODM topology + * changes that we could be applying to front + * end. Since at the current stage front end + * changes are not yet applied. We can only + * apply test pattern in hw based on current + * state and populate the final test pattern + * params in new state. If current and new test + * pattern params are different as result of + * different ODM topology being used, it will be + * detected and handle during front end + * programming update. + */ + dc->link_srv->dp_set_test_pattern(stream->link, stream->test_pattern.type, stream->test_pattern.color_space, stream->test_pattern.p_link_settings, stream->test_pattern.p_custom_pattern, stream->test_pattern.cust_pattern_size); + resource_build_test_pattern_params(&context->res_ctx, pipe_ctx); } if (stream_update->dpms_off) { @@ -3363,6 +3449,7 @@ void dc_dmub_update_dirty_rect(struct dc *dc, if (srf_updates[i].surface->flip_immediate) continue; + update_dirty_rect->cmd_version = DMUB_CMD_PSR_CONTROL_VERSION_1; update_dirty_rect->dirty_rect_count = flip_addr->dirty_rect_count; memcpy(update_dirty_rect->src_dirty_rects, flip_addr->dirty_rects, sizeof(flip_addr->dirty_rects)); @@ -3479,6 +3566,7 @@ static void commit_planes_for_stream_fast(struct dc *dc, int i, j; struct pipe_ctx *top_pipe_to_program = NULL; struct dc_stream_status *stream_status = NULL; + dc_exit_ips_for_hw_access(dc); dc_z10_restore(dc); @@ -3536,7 +3624,8 @@ static void commit_planes_for_stream_fast(struct dc *dc, context->block_sequence, &(context->block_sequence_steps), top_pipe_to_program, - stream_status); + stream_status, + context); hwss_execute_sequence(dc, context->block_sequence, context->block_sequence_steps); @@ -4065,24 +4154,14 @@ struct pipe_split_policy_backup { bool dynamic_odm_policy; bool subvp_policy; enum pipe_split_policy mpc_policy; + char force_odm[MAX_PIPES]; }; -static void release_minimal_transition_state(struct dc *dc, - struct dc_state *context, struct pipe_split_policy_backup *policy) -{ - dc_state_release(context); - /* restore previous pipe split and odm policy */ - if (!dc->config.is_vmin_only_asic) - dc->debug.pipe_split_policy = policy->mpc_policy; - dc->debug.enable_single_display_2to1_odm_policy = policy->dynamic_odm_policy; - dc->debug.force_disable_subvp = policy->subvp_policy; -} - -static struct dc_state *create_minimal_transition_state(struct dc *dc, - struct dc_state *base_context, struct pipe_split_policy_backup *policy) +static void backup_and_set_minimal_pipe_split_policy(struct dc *dc, + struct dc_state *context, + struct pipe_split_policy_backup *policy) { - struct dc_state *minimal_transition_context = NULL; - unsigned int i, j; + int i; if (!dc->config.is_vmin_only_asic) { policy->mpc_policy = dc->debug.pipe_split_policy; @@ -4092,97 +4171,257 @@ static struct dc_state *create_minimal_transition_state(struct dc *dc, dc->debug.enable_single_display_2to1_odm_policy = false; policy->subvp_policy = dc->debug.force_disable_subvp; dc->debug.force_disable_subvp = true; + for (i = 0; i < context->stream_count; i++) { + policy->force_odm[i] = context->streams[i]->debug.force_odm_combine_segments; + context->streams[i]->debug.force_odm_combine_segments = 0; + } +} + +static void restore_minimal_pipe_split_policy(struct dc *dc, + struct dc_state *context, + struct pipe_split_policy_backup *policy) +{ + uint8_t i; + + if (!dc->config.is_vmin_only_asic) + dc->debug.pipe_split_policy = policy->mpc_policy; + dc->debug.enable_single_display_2to1_odm_policy = + policy->dynamic_odm_policy; + dc->debug.force_disable_subvp = policy->subvp_policy; + for (i = 0; i < context->stream_count; i++) + context->streams[i]->debug.force_odm_combine_segments = policy->force_odm[i]; +} + +static void release_minimal_transition_state(struct dc *dc, + struct dc_state *minimal_transition_context, + struct dc_state *base_context, + struct pipe_split_policy_backup *policy) +{ + restore_minimal_pipe_split_policy(dc, base_context, policy); + dc_state_release(minimal_transition_context); +} + +static void force_vsync_flip_in_minimal_transition_context(struct dc_state *context) +{ + uint8_t i; + int j; + struct dc_stream_status *stream_status; + + for (i = 0; i < context->stream_count; i++) { + stream_status = &context->stream_status[i]; + + for (j = 0; j < stream_status->plane_count; j++) + stream_status->plane_states[j]->flip_immediate = false; + } +} + +static struct dc_state *create_minimal_transition_state(struct dc *dc, + struct dc_state *base_context, struct pipe_split_policy_backup *policy) +{ + struct dc_state *minimal_transition_context = NULL; minimal_transition_context = dc_state_create_copy(base_context); if (!minimal_transition_context) return NULL; + backup_and_set_minimal_pipe_split_policy(dc, base_context, policy); /* commit minimal state */ if (dc->res_pool->funcs->validate_bandwidth(dc, minimal_transition_context, false)) { - for (i = 0; i < minimal_transition_context->stream_count; i++) { - struct dc_stream_status *stream_status = &minimal_transition_context->stream_status[i]; - - for (j = 0; j < stream_status->plane_count; j++) { - struct dc_plane_state *plane_state = stream_status->plane_states[j]; - - /* force vsync flip when reconfiguring pipes to prevent underflow - * and corruption - */ - plane_state->flip_immediate = false; - } - } + /* prevent underflow and corruption when reconfiguring pipes */ + force_vsync_flip_in_minimal_transition_context(minimal_transition_context); } else { - /* this should never happen */ - release_minimal_transition_state(dc, minimal_transition_context, policy); + /* + * This should never happen, minimal transition state should + * always be validated first before adding pipe split features. + */ + release_minimal_transition_state(dc, minimal_transition_context, base_context, policy); BREAK_TO_DEBUGGER(); minimal_transition_context = NULL; } return minimal_transition_context; } +static bool is_pipe_topology_transition_seamless_with_intermediate_step( + struct dc *dc, + struct dc_state *initial_state, + struct dc_state *intermediate_state, + struct dc_state *final_state) +{ + return dc->hwss.is_pipe_topology_transition_seamless(dc, initial_state, + intermediate_state) && + dc->hwss.is_pipe_topology_transition_seamless(dc, + intermediate_state, final_state); +} + +static void swap_and_release_current_context(struct dc *dc, + struct dc_state *new_context, struct dc_stream_state *stream) +{ + + int i; + struct dc_state *old = dc->current_state; + struct pipe_ctx *pipe_ctx; + + /* Since memory free requires elevated IRQ, an interrupt + * request is generated by mem free. If this happens + * between freeing and reassigning the context, our vsync + * interrupt will call into dc and cause a memory + * corruption. Hence, we first reassign the context, + * then free the old context. + */ + dc->current_state = new_context; + dc_state_release(old); + + // clear any forced full updates + for (i = 0; i < dc->res_pool->pipe_count; i++) { + pipe_ctx = &new_context->res_ctx.pipe_ctx[i]; + + if (pipe_ctx->plane_state && pipe_ctx->stream == stream) + pipe_ctx->plane_state->force_full_update = false; + } +} + +static int initialize_empty_surface_updates( + struct dc_stream_state *stream, + struct dc_surface_update *srf_updates) +{ + struct dc_stream_status *status = dc_stream_get_status(stream); + int i; + + if (!status) + return 0; + + for (i = 0; i < status->plane_count; i++) + srf_updates[i].surface = status->plane_states[i]; + + return status->plane_count; +} + +static bool commit_minimal_transition_based_on_new_context(struct dc *dc, + struct dc_state *new_context, + struct dc_stream_state *stream, + struct dc_surface_update *srf_updates, + int surface_count) +{ + bool success = false; + struct pipe_split_policy_backup policy; + struct dc_state *intermediate_context = + create_minimal_transition_state(dc, new_context, + &policy); + + if (intermediate_context) { + if (is_pipe_topology_transition_seamless_with_intermediate_step( + dc, + dc->current_state, + intermediate_context, + new_context)) { + DC_LOG_DC("commit minimal transition state: base = new state\n"); + commit_planes_for_stream(dc, srf_updates, + surface_count, stream, NULL, + UPDATE_TYPE_FULL, intermediate_context); + swap_and_release_current_context( + dc, intermediate_context, stream); + dc_state_retain(dc->current_state); + success = true; + } + release_minimal_transition_state( + dc, intermediate_context, new_context, &policy); + } + return success; +} + +static bool commit_minimal_transition_based_on_current_context(struct dc *dc, + struct dc_state *new_context, struct dc_stream_state *stream) +{ + bool success = false; + struct pipe_split_policy_backup policy; + struct dc_state *intermediate_context; + struct dc_state *old_current_state = dc->current_state; + struct dc_surface_update srf_updates[MAX_SURFACE_NUM] = {0}; + int surface_count; + + /* + * Both current and new contexts share the same stream and plane state + * pointers. When new context is validated, stream and planes get + * populated with new updates such as new plane addresses. This makes + * the current context no longer valid because stream and planes are + * modified from the original. We backup current stream and plane states + * into scratch space whenever we are populating new context. So we can + * restore the original values back by calling the restore function now. + * This restores back the original stream and plane states associated + * with the current state. + */ + restore_planes_and_stream_state(&dc->scratch.current_state, stream); + dc_state_retain(old_current_state); + intermediate_context = create_minimal_transition_state(dc, + old_current_state, &policy); + + if (intermediate_context) { + if (is_pipe_topology_transition_seamless_with_intermediate_step( + dc, + dc->current_state, + intermediate_context, + new_context)) { + DC_LOG_DC("commit minimal transition state: base = current state\n"); + surface_count = initialize_empty_surface_updates( + stream, srf_updates); + commit_planes_for_stream(dc, srf_updates, + surface_count, stream, NULL, + UPDATE_TYPE_FULL, intermediate_context); + swap_and_release_current_context( + dc, intermediate_context, stream); + dc_state_retain(dc->current_state); + success = true; + } + release_minimal_transition_state(dc, intermediate_context, + old_current_state, &policy); + } + dc_state_release(old_current_state); + /* + * Restore stream and plane states back to the values associated with + * new context. + */ + restore_planes_and_stream_state(&dc->scratch.new_state, stream); + return success; +} /** - * commit_minimal_transition_state - Commit a minimal state based on current or new context + * commit_minimal_transition_state_in_dc_update - Commit a minimal state based + * on current or new context * * @dc: DC structure, used to get the current state - * @context: New context + * @new_context: New context * @stream: Stream getting the update for the flip + * @srf_updates: Surface updates + * @surface_count: Number of surfaces * - * The function takes in current state and new state and determine a minimal transition state - * as the intermediate step which could make the transition between current and new states - * seamless. If found, it will commit the minimal transition state and update current state to - * this minimal transition state and return true, if not, it will return false. + * The function takes in current state and new state and determine a minimal + * transition state as the intermediate step which could make the transition + * between current and new states seamless. If found, it will commit the minimal + * transition state and update current state to this minimal transition state + * and return true, if not, it will return false. * * Return: * Return True if the minimal transition succeeded, false otherwise */ -static bool commit_minimal_transition_state(struct dc *dc, - struct dc_state *context, - struct dc_stream_state *stream) -{ - bool success = false; - struct dc_state *minimal_transition_context; - struct pipe_split_policy_backup policy; - - /* commit based on new context */ - minimal_transition_context = create_minimal_transition_state(dc, - context, &policy); - if (minimal_transition_context) { - if (dc->hwss.is_pipe_topology_transition_seamless( - dc, dc->current_state, minimal_transition_context) && - dc->hwss.is_pipe_topology_transition_seamless( - dc, minimal_transition_context, context)) { - DC_LOG_DC("%s base = new state\n", __func__); - - success = dc_commit_state_no_check(dc, minimal_transition_context) == DC_OK; - } - release_minimal_transition_state(dc, minimal_transition_context, &policy); - } - - if (!success) { - /* commit based on current context */ - restore_planes_and_stream_state(&dc->current_state->scratch, stream); - minimal_transition_context = create_minimal_transition_state(dc, - dc->current_state, &policy); - if (minimal_transition_context) { - if (dc->hwss.is_pipe_topology_transition_seamless( - dc, dc->current_state, minimal_transition_context) && - dc->hwss.is_pipe_topology_transition_seamless( - dc, minimal_transition_context, context)) { - DC_LOG_DC("%s base = current state\n", __func__); - success = dc_commit_state_no_check(dc, minimal_transition_context) == DC_OK; - } - release_minimal_transition_state(dc, minimal_transition_context, &policy); - } - restore_planes_and_stream_state(&context->scratch, stream); - } - - ASSERT(success); +static bool commit_minimal_transition_state_in_dc_update(struct dc *dc, + struct dc_state *new_context, + struct dc_stream_state *stream, + struct dc_surface_update *srf_updates, + int surface_count) +{ + bool success = commit_minimal_transition_based_on_new_context( + dc, new_context, stream, srf_updates, + surface_count); + if (!success) + success = commit_minimal_transition_based_on_current_context(dc, + new_context, stream); + if (!success) + DC_LOG_ERROR("Fail to commit a seamless minimal transition state between current and new states.\nThis pipe topology update is non-seamless!\n"); return success; } /** - * commit_minimal_transition_state_legacy - Create a transition pipe split state + * commit_minimal_transition_state - Create a transition pipe split state * * @dc: Used to get the current state status * @transition_base_context: New transition state @@ -4199,7 +4438,7 @@ static bool commit_minimal_transition_state(struct dc *dc, * Return: * Return false if something is wrong in the transition state. */ -static bool commit_minimal_transition_state_legacy(struct dc *dc, +static bool commit_minimal_transition_state(struct dc *dc, struct dc_state *transition_base_context) { struct dc_state *transition_context; @@ -4260,12 +4499,14 @@ static bool commit_minimal_transition_state_legacy(struct dc *dc, dc->debug.pipe_split_policy != MPC_SPLIT_AVOID ? "MPC in Use" : "Unknown"); + dc_state_retain(transition_base_context); transition_context = create_minimal_transition_state(dc, transition_base_context, &policy); if (transition_context) { ret = dc_commit_state_no_check(dc, transition_context); - release_minimal_transition_state(dc, transition_context, &policy); + release_minimal_transition_state(dc, transition_context, transition_base_context, &policy); } + dc_state_release(transition_base_context); if (ret != DC_OK) { /* this should never happen */ @@ -4283,41 +4524,6 @@ static bool commit_minimal_transition_state_legacy(struct dc *dc, return true; } -/** - * update_seamless_boot_flags() - Helper function for updating seamless boot flags - * - * @dc: Current DC state - * @context: New DC state to be programmed - * @surface_count: Number of surfaces that have an updated - * @stream: Corresponding stream to be updated in the current flip - * - * Updating seamless boot flags do not need to be part of the commit sequence. This - * helper function will update the seamless boot flags on each flip (if required) - * outside of the HW commit sequence (fast or slow). - * - * Return: void - */ -static void update_seamless_boot_flags(struct dc *dc, - struct dc_state *context, - int surface_count, - struct dc_stream_state *stream) -{ - if (get_seamless_boot_stream_count(context) > 0 && surface_count > 0) { - /* Optimize seamless boot flag keeps clocks and watermarks high until - * first flip. After first flip, optimization is required to lower - * bandwidth. Important to note that it is expected UEFI will - * only light up a single display on POST, therefore we only expect - * one stream with seamless boot flag set. - */ - if (stream->apply_seamless_boot_optimization) { - stream->apply_seamless_boot_optimization = false; - - if (get_seamless_boot_stream_count(context) == 0) - dc->optimized_required = true; - } - } -} - static void populate_fast_updates(struct dc_fast_update *fast_update, struct dc_surface_update *srf_updates, int surface_count, @@ -4437,123 +4643,9 @@ static bool fast_update_only(struct dc *dc, && !full_update_required(dc, srf_updates, surface_count, stream_update, stream); } -bool dc_update_planes_and_stream(struct dc *dc, +static bool update_planes_and_stream_v1(struct dc *dc, struct dc_surface_update *srf_updates, int surface_count, struct dc_stream_state *stream, - struct dc_stream_update *stream_update) -{ - struct dc_state *context; - enum surface_update_type update_type; - int i; - struct dc_fast_update fast_update[MAX_SURFACES] = {0}; - - /* In cases where MPO and split or ODM are used transitions can - * cause underflow. Apply stream configuration with minimal pipe - * split first to avoid unsupported transitions for active pipes. - */ - bool force_minimal_pipe_splitting = 0; - bool is_plane_addition = 0; - bool is_fast_update_only; - - dc_exit_ips_for_hw_access(dc); - - populate_fast_updates(fast_update, srf_updates, surface_count, stream_update); - is_fast_update_only = fast_update_only(dc, fast_update, srf_updates, - surface_count, stream_update, stream); - force_minimal_pipe_splitting = could_mpcc_tree_change_for_active_pipes( - dc, - stream, - srf_updates, - surface_count, - &is_plane_addition); - - /* on plane addition, minimal state is the current one */ - if (force_minimal_pipe_splitting && is_plane_addition && - !commit_minimal_transition_state_legacy(dc, dc->current_state)) - return false; - - if (!update_planes_and_stream_state( - dc, - srf_updates, - surface_count, - stream, - stream_update, - &update_type, - &context)) - return false; - - /* on plane removal, minimal state is the new one */ - if (force_minimal_pipe_splitting && !is_plane_addition) { - if (!commit_minimal_transition_state_legacy(dc, context)) { - dc_state_release(context); - return false; - } - update_type = UPDATE_TYPE_FULL; - } - - if (dc->hwss.is_pipe_topology_transition_seamless && - !dc->hwss.is_pipe_topology_transition_seamless( - dc, dc->current_state, context)) { - commit_minimal_transition_state(dc, - context, stream); - } - update_seamless_boot_flags(dc, context, surface_count, stream); - if (is_fast_update_only && !dc->debug.enable_legacy_fast_update) { - commit_planes_for_stream_fast(dc, - srf_updates, - surface_count, - stream, - stream_update, - update_type, - context); - } else { - if (!stream_update && - dc->hwss.is_pipe_topology_transition_seamless && - !dc->hwss.is_pipe_topology_transition_seamless( - dc, dc->current_state, context)) { - DC_LOG_ERROR("performing non-seamless pipe topology transition with surface only update!\n"); - BREAK_TO_DEBUGGER(); - } - commit_planes_for_stream( - dc, - srf_updates, - surface_count, - stream, - stream_update, - update_type, - context); - } - - if (dc->current_state != context) { - - /* Since memory free requires elevated IRQL, an interrupt - * request is generated by mem free. If this happens - * between freeing and reassigning the context, our vsync - * interrupt will call into dc and cause a memory - * corruption BSOD. Hence, we first reassign the context, - * then free the old context. - */ - - struct dc_state *old = dc->current_state; - - dc->current_state = context; - dc_state_release(old); - - // clear any forced full updates - for (i = 0; i < dc->res_pool->pipe_count; i++) { - struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i]; - - if (pipe_ctx->plane_state && pipe_ctx->stream == stream) - pipe_ctx->plane_state->force_full_update = false; - } - } - return true; -} - -void dc_commit_updates_for_stream(struct dc *dc, - struct dc_surface_update *srf_updates, - int surface_count, - struct dc_stream_state *stream, struct dc_stream_update *stream_update, struct dc_state *state) { @@ -4573,35 +4665,13 @@ void dc_commit_updates_for_stream(struct dc *dc, update_type = dc_check_update_surfaces_for_stream( dc, srf_updates, surface_count, stream_update, stream_status); - /* TODO: Since change commit sequence can have a huge impact, - * we decided to only enable it for DCN3x. However, as soon as - * we get more confident about this change we'll need to enable - * the new sequence for all ASICs. - */ - if (dc->ctx->dce_version >= DCN_VERSION_3_2) { - /* - * Previous frame finished and HW is ready for optimization. - */ - if (update_type == UPDATE_TYPE_FAST) - dc_post_update_surfaces_to_stream(dc); - - dc_update_planes_and_stream(dc, srf_updates, - surface_count, stream, - stream_update); - return; - } - - if (update_type >= update_surface_trace_level) - update_surface_trace(dc, srf_updates, surface_count); - - if (update_type >= UPDATE_TYPE_FULL) { /* initialize scratch memory for building context */ context = dc_state_create_copy(state); if (context == NULL) { DC_ERROR("Failed to allocate new validate context!\n"); - return; + return false; } for (i = 0; i < dc->res_pool->pipe_count; i++) { @@ -4618,7 +4688,6 @@ void dc_commit_updates_for_stream(struct dc *dc, dc_post_update_surfaces_to_stream(dc); } - for (i = 0; i < surface_count; i++) { struct dc_plane_state *surface = srf_updates[i].surface; @@ -4643,13 +4712,12 @@ void dc_commit_updates_for_stream(struct dc *dc, if (!dc->res_pool->funcs->validate_bandwidth(dc, context, false)) { DC_ERROR("Mode validation failed for stream update!\n"); dc_state_release(context); - return; + return false; } } TRACE_DC_PIPE_STATE(pipe_ctx, i, MAX_PIPES); - update_seamless_boot_flags(dc, context, surface_count, stream); if (fast_update_only(dc, fast_update, srf_updates, surface_count, stream_update, stream) && !dc->debug.enable_legacy_fast_update) { commit_planes_for_stream_fast(dc, @@ -4690,9 +4758,252 @@ void dc_commit_updates_for_stream(struct dc *dc, dc_post_update_surfaces_to_stream(dc); TRACE_DCE_CLOCK_STATE(&context->bw_ctx.bw.dce); } + return true; +} + +static bool update_planes_and_stream_v2(struct dc *dc, + struct dc_surface_update *srf_updates, int surface_count, + struct dc_stream_state *stream, + struct dc_stream_update *stream_update) +{ + struct dc_state *context; + enum surface_update_type update_type; + struct dc_fast_update fast_update[MAX_SURFACES] = {0}; + + /* In cases where MPO and split or ODM are used transitions can + * cause underflow. Apply stream configuration with minimal pipe + * split first to avoid unsupported transitions for active pipes. + */ + bool force_minimal_pipe_splitting = 0; + bool is_plane_addition = 0; + bool is_fast_update_only; + + populate_fast_updates(fast_update, srf_updates, surface_count, stream_update); + is_fast_update_only = fast_update_only(dc, fast_update, srf_updates, + surface_count, stream_update, stream); + force_minimal_pipe_splitting = could_mpcc_tree_change_for_active_pipes( + dc, + stream, + srf_updates, + surface_count, + &is_plane_addition); + + /* on plane addition, minimal state is the current one */ + if (force_minimal_pipe_splitting && is_plane_addition && + !commit_minimal_transition_state(dc, dc->current_state)) + return false; + + if (!update_planes_and_stream_state( + dc, + srf_updates, + surface_count, + stream, + stream_update, + &update_type, + &context)) + return false; + + /* on plane removal, minimal state is the new one */ + if (force_minimal_pipe_splitting && !is_plane_addition) { + if (!commit_minimal_transition_state(dc, context)) { + dc_state_release(context); + return false; + } + update_type = UPDATE_TYPE_FULL; + } - return; + if (dc->hwss.is_pipe_topology_transition_seamless && + !dc->hwss.is_pipe_topology_transition_seamless( + dc, dc->current_state, context)) + commit_minimal_transition_state_in_dc_update(dc, context, stream, + srf_updates, surface_count); + if (is_fast_update_only && !dc->debug.enable_legacy_fast_update) { + commit_planes_for_stream_fast(dc, + srf_updates, + surface_count, + stream, + stream_update, + update_type, + context); + } else { + if (!stream_update && + dc->hwss.is_pipe_topology_transition_seamless && + !dc->hwss.is_pipe_topology_transition_seamless( + dc, dc->current_state, context)) { + DC_LOG_ERROR("performing non-seamless pipe topology transition with surface only update!\n"); + BREAK_TO_DEBUGGER(); + } + commit_planes_for_stream( + dc, + srf_updates, + surface_count, + stream, + stream_update, + update_type, + context); + } + if (dc->current_state != context) + swap_and_release_current_context(dc, context, stream); + return true; +} + +static void commit_planes_and_stream_update_on_current_context(struct dc *dc, + struct dc_surface_update *srf_updates, int surface_count, + struct dc_stream_state *stream, + struct dc_stream_update *stream_update, + enum surface_update_type update_type) +{ + struct dc_fast_update fast_update[MAX_SURFACES] = {0}; + + ASSERT(update_type < UPDATE_TYPE_FULL); + populate_fast_updates(fast_update, srf_updates, surface_count, + stream_update); + if (fast_update_only(dc, fast_update, srf_updates, surface_count, + stream_update, stream) && + !dc->debug.enable_legacy_fast_update) + commit_planes_for_stream_fast(dc, + srf_updates, + surface_count, + stream, + stream_update, + update_type, + dc->current_state); + else + commit_planes_for_stream( + dc, + srf_updates, + surface_count, + stream, + stream_update, + update_type, + dc->current_state); +} + +static void commit_planes_and_stream_update_with_new_context(struct dc *dc, + struct dc_surface_update *srf_updates, int surface_count, + struct dc_stream_state *stream, + struct dc_stream_update *stream_update, + enum surface_update_type update_type, + struct dc_state *new_context) +{ + ASSERT(update_type >= UPDATE_TYPE_FULL); + if (!dc->hwss.is_pipe_topology_transition_seamless(dc, + dc->current_state, new_context)) + /* + * It is required by the feature design that all pipe topologies + * using extra free pipes for power saving purposes such as + * dynamic ODM or SubVp shall only be enabled when it can be + * transitioned seamlessly to AND from its minimal transition + * state. A minimal transition state is defined as the same dc + * state but with all power saving features disabled. So it uses + * the minimum pipe topology. When we can't seamlessly + * transition from state A to state B, we will insert the + * minimal transition state A' or B' in between so seamless + * transition between A and B can be made possible. + */ + commit_minimal_transition_state_in_dc_update(dc, new_context, + stream, srf_updates, surface_count); + + commit_planes_for_stream( + dc, + srf_updates, + surface_count, + stream, + stream_update, + update_type, + new_context); +} + +static bool update_planes_and_stream_v3(struct dc *dc, + struct dc_surface_update *srf_updates, int surface_count, + struct dc_stream_state *stream, + struct dc_stream_update *stream_update) +{ + struct dc_state *new_context; + enum surface_update_type update_type; + + /* + * When this function returns true and new_context is not equal to + * current state, the function allocates and validates a new dc state + * and assigns it to new_context. The function expects that the caller + * is responsible to free this memory when new_context is no longer + * used. We swap current with new context and free current instead. So + * new_context's memory will live until the next full update after it is + * replaced by a newer context. Refer to the use of + * swap_and_free_current_context below. + */ + if (!update_planes_and_stream_state(dc, srf_updates, surface_count, + stream, stream_update, &update_type, + &new_context)) + return false; + + if (new_context == dc->current_state) { + commit_planes_and_stream_update_on_current_context(dc, + srf_updates, surface_count, stream, + stream_update, update_type); + } else { + commit_planes_and_stream_update_with_new_context(dc, + srf_updates, surface_count, stream, + stream_update, update_type, new_context); + swap_and_release_current_context(dc, new_context, stream); + } + + return true; +} + +bool dc_update_planes_and_stream(struct dc *dc, + struct dc_surface_update *srf_updates, int surface_count, + struct dc_stream_state *stream, + struct dc_stream_update *stream_update) +{ + dc_exit_ips_for_hw_access(dc); + /* + * update planes and stream version 3 separates FULL and FAST updates + * to their own sequences. It aims to clean up frequent checks for + * update type resulting unnecessary branching in logic flow. It also + * adds a new commit minimal transition sequence, which detects the need + * for minimal transition based on the actual comparison of current and + * new states instead of "predicting" it based on per feature software + * policy.i.e could_mpcc_tree_change_for_active_pipes. The new commit + * minimal transition sequence is made universal to any power saving + * features that would use extra free pipes such as Dynamic ODM/MPC + * Combine, MPO or SubVp. Therefore there is no longer a need to + * specially handle compatibility problems with transitions among those + * features as they are now transparent to the new sequence. + */ + if (dc->ctx->dce_version > DCN_VERSION_3_51) + return update_planes_and_stream_v3(dc, srf_updates, + surface_count, stream, stream_update); + return update_planes_and_stream_v2(dc, srf_updates, + surface_count, stream, stream_update); +} + +void dc_commit_updates_for_stream(struct dc *dc, + struct dc_surface_update *srf_updates, + int surface_count, + struct dc_stream_state *stream, + struct dc_stream_update *stream_update, + struct dc_state *state) +{ + dc_exit_ips_for_hw_access(dc); + /* TODO: Since change commit sequence can have a huge impact, + * we decided to only enable it for DCN3x. However, as soon as + * we get more confident about this change we'll need to enable + * the new sequence for all ASICs. + */ + if (dc->ctx->dce_version > DCN_VERSION_3_51) { + update_planes_and_stream_v3(dc, srf_updates, surface_count, + stream, stream_update); + return; + } + if (dc->ctx->dce_version >= DCN_VERSION_3_2) { + update_planes_and_stream_v2(dc, srf_updates, surface_count, + stream, stream_update); + return; + } + update_planes_and_stream_v1(dc, srf_updates, surface_count, stream, + stream_update, state); } uint8_t dc_get_current_stream_count(struct dc *dc) @@ -4735,8 +5046,13 @@ void dc_interrupt_ack(struct dc *dc, enum dc_irq_source src) void dc_power_down_on_boot(struct dc *dc) { if (dc->ctx->dce_environment != DCE_ENV_VIRTUAL_HW && - dc->hwss.power_down_on_boot) + dc->hwss.power_down_on_boot) { + + if (dc->caps.ips_support) + dc_exit_ips_for_hw_access(dc); + dc->hwss.power_down_on_boot(dc); + } } void dc_set_power_state( @@ -4874,11 +5190,15 @@ bool dc_set_replay_allow_active(struct dc *dc, bool active) return true; } -void dc_allow_idle_optimizations(struct dc *dc, bool allow) +void dc_allow_idle_optimizations_internal(struct dc *dc, bool allow, char const *caller_name) { if (dc->debug.disable_idle_power_optimizations) return; + if (allow != dc->idle_optimizations_allowed) + DC_LOG_IPS("%s: allow_idle old=%d new=%d (caller=%s)\n", __func__, + dc->idle_optimizations_allowed, allow, caller_name); + if (dc->caps.ips_support && (dc->config.disable_ips == DMUB_IPS_DISABLE_ALL)) return; @@ -4893,10 +5213,10 @@ void dc_allow_idle_optimizations(struct dc *dc, bool allow) dc->idle_optimizations_allowed = allow; } -void dc_exit_ips_for_hw_access(struct dc *dc) +void dc_exit_ips_for_hw_access_internal(struct dc *dc, const char *caller_name) { if (dc->caps.ips_support) - dc_allow_idle_optimizations(dc, false); + dc_allow_idle_optimizations_internal(dc, false, caller_name); } bool dc_dmub_is_ips_idle_state(struct dc *dc) @@ -5030,10 +5350,13 @@ void dc_enable_dcmode_clk_limit(struct dc *dc, bool enable) } dc->clk_mgr->dc_mode_softmax_enabled = enable; } -bool dc_is_plane_eligible_for_idle_optimizations(struct dc *dc, struct dc_plane_state *plane, +bool dc_is_plane_eligible_for_idle_optimizations(struct dc *dc, + unsigned int pitch, + unsigned int height, + enum surface_pixel_format format, struct dc_cursor_attributes *cursor_attr) { - if (dc->hwss.does_plane_fit_in_mall && dc->hwss.does_plane_fit_in_mall(dc, plane, cursor_attr)) + if (dc->hwss.does_plane_fit_in_mall && dc->hwss.does_plane_fit_in_mall(dc, pitch, height, format, cursor_attr)) return true; return false; } diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/core/dc_hw_sequencer.c index 9c05b1a07142..5c1d3017aefd 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_hw_sequencer.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_hw_sequencer.c @@ -392,10 +392,10 @@ void get_hdr_visual_confirm_color( switch (top_pipe_ctx->plane_res.scl_data.format) { case PIXEL_FORMAT_ARGB2101010: - if (top_pipe_ctx->stream->out_transfer_func->tf == TRANSFER_FUNCTION_PQ) { + if (top_pipe_ctx->stream->out_transfer_func.tf == TRANSFER_FUNCTION_PQ) { /* HDR10, ARGB2101010 - set border color to red */ color->color_r_cr = color_value; - } else if (top_pipe_ctx->stream->out_transfer_func->tf == TRANSFER_FUNCTION_GAMMA22) { + } else if (top_pipe_ctx->stream->out_transfer_func.tf == TRANSFER_FUNCTION_GAMMA22) { /* FreeSync 2 ARGB2101010 - set border color to pink */ color->color_r_cr = color_value; color->color_b_cb = color_value; @@ -403,10 +403,10 @@ void get_hdr_visual_confirm_color( is_sdr = true; break; case PIXEL_FORMAT_FP16: - if (top_pipe_ctx->stream->out_transfer_func->tf == TRANSFER_FUNCTION_PQ) { + if (top_pipe_ctx->stream->out_transfer_func.tf == TRANSFER_FUNCTION_PQ) { /* HDR10, FP16 - set border color to blue */ color->color_b_cb = color_value; - } else if (top_pipe_ctx->stream->out_transfer_func->tf == TRANSFER_FUNCTION_GAMMA22) { + } else if (top_pipe_ctx->stream->out_transfer_func.tf == TRANSFER_FUNCTION_GAMMA22) { /* FreeSync 2 HDR - set border color to green */ color->color_g_y = color_value; } else @@ -558,9 +558,10 @@ void hwss_build_fast_sequence(struct dc *dc, struct dc_dmub_cmd *dc_dmub_cmd, unsigned int dmub_cmd_count, struct block_sequence block_sequence[], - int *num_steps, + unsigned int *num_steps, struct pipe_ctx *pipe_ctx, - struct dc_stream_status *stream_status) + struct dc_stream_status *stream_status, + struct dc_state *context) { struct dc_plane_state *plane = pipe_ctx->plane_state; struct dc_stream_state *stream = pipe_ctx->stream; 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 ec4bf9432bdb..15819416a2f3 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c @@ -340,7 +340,7 @@ struct resource_pool *dc_create_resource_pool(struct dc *dc, return res_pool; } -void dc_destroy_resource_pool(struct dc *dc) +void dc_destroy_resource_pool(struct dc *dc) { if (dc) { if (dc->res_pool) @@ -1457,6 +1457,9 @@ void resource_build_test_pattern_params(struct resource_context *res_ctx, controller_color_space = convert_dp_to_controller_color_space( otg_master->stream->test_pattern.color_space); + if (controller_test_pattern == CONTROLLER_DP_TEST_PATTERN_VIDEOMODE) + return; + odm_cnt = resource_get_opp_heads_for_otg_master(otg_master, res_ctx, opp_heads); odm_slice_width = h_active / odm_cnt; @@ -1485,6 +1488,7 @@ bool resource_build_scaling_params(struct pipe_ctx *pipe_ctx) struct dc_crtc_timing *timing = &pipe_ctx->stream->timing; const struct rect odm_slice_rec = calculate_odm_slice_in_timing_active(pipe_ctx); bool res = false; + DC_LOGGER_INIT(pipe_ctx->stream->ctx->logger); /* Invalid input */ @@ -1496,9 +1500,6 @@ bool resource_build_scaling_params(struct pipe_ctx *pipe_ctx) return false; } - pipe_ctx->plane_res.scl_data.format = convert_pixel_format_to_dalsurface( - pipe_ctx->plane_state->format); - /* Timing borders are part of vactive that we are also supposed to skip in addition * to any stream dst offset. Since dm logic assumes dst is in addressable * space we need to add the left and top borders to dst offsets temporarily. @@ -1510,6 +1511,8 @@ bool resource_build_scaling_params(struct pipe_ctx *pipe_ctx) /* Calculate H and V active size */ pipe_ctx->plane_res.scl_data.h_active = odm_slice_rec.width; pipe_ctx->plane_res.scl_data.v_active = odm_slice_rec.height; + pipe_ctx->plane_res.scl_data.format = convert_pixel_format_to_dalsurface( + pipe_ctx->plane_state->format); /* depends on h_active */ calculate_recout(pipe_ctx); @@ -1794,6 +1797,30 @@ int recource_find_free_pipe_used_as_otg_master_in_cur_res_ctx( return free_pipe_idx; } +int resource_find_free_pipe_used_as_cur_sec_dpp( + const struct resource_context *cur_res_ctx, + struct resource_context *new_res_ctx, + const struct resource_pool *pool) +{ + int free_pipe_idx = FREE_PIPE_INDEX_NOT_FOUND; + const struct pipe_ctx *new_pipe, *cur_pipe; + int i; + + for (i = 0; i < pool->pipe_count; i++) { + cur_pipe = &cur_res_ctx->pipe_ctx[i]; + new_pipe = &new_res_ctx->pipe_ctx[i]; + + if (resource_is_pipe_type(cur_pipe, DPP_PIPE) && + !resource_is_pipe_type(cur_pipe, OPP_HEAD) && + resource_is_pipe_type(new_pipe, FREE_PIPE)) { + free_pipe_idx = i; + break; + } + } + + return free_pipe_idx; +} + int resource_find_free_pipe_used_as_cur_sec_dpp_in_mpcc_combine( const struct resource_context *cur_res_ctx, struct resource_context *new_res_ctx, @@ -2168,50 +2195,91 @@ static void resource_log_pipe(struct dc *dc, struct pipe_ctx *pipe, } } -void resource_log_pipe_topology_update(struct dc *dc, struct dc_state *state) +static void resource_log_pipe_for_stream(struct dc *dc, struct dc_state *state, + struct pipe_ctx *otg_master, int stream_idx) { - struct pipe_ctx *otg_master; struct pipe_ctx *opp_heads[MAX_PIPES]; struct pipe_ctx *dpp_pipes[MAX_PIPES]; - int stream_idx, slice_idx, dpp_idx, plane_idx, slice_count, dpp_count; + int slice_idx, dpp_idx, plane_idx, slice_count, dpp_count; bool is_primary; DC_LOGGER_INIT(dc->ctx->logger); + slice_count = resource_get_opp_heads_for_otg_master(otg_master, + &state->res_ctx, opp_heads); + for (slice_idx = 0; slice_idx < slice_count; slice_idx++) { + plane_idx = -1; + if (opp_heads[slice_idx]->plane_state) { + dpp_count = resource_get_dpp_pipes_for_opp_head( + opp_heads[slice_idx], + &state->res_ctx, + dpp_pipes); + for (dpp_idx = 0; dpp_idx < dpp_count; dpp_idx++) { + is_primary = !dpp_pipes[dpp_idx]->top_pipe || + dpp_pipes[dpp_idx]->top_pipe->plane_state != dpp_pipes[dpp_idx]->plane_state; + if (is_primary) + plane_idx++; + resource_log_pipe(dc, dpp_pipes[dpp_idx], + stream_idx, slice_idx, + plane_idx, slice_count, + is_primary); + } + } else { + resource_log_pipe(dc, opp_heads[slice_idx], + stream_idx, slice_idx, plane_idx, + slice_count, true); + } + + } +} + +static int resource_stream_to_stream_idx(struct dc_state *state, + struct dc_stream_state *stream) +{ + int i, stream_idx = -1; + + for (i = 0; i < state->stream_count; i++) + if (state->streams[i] == stream) { + stream_idx = i; + break; + } + + /* never return negative array index */ + if (stream_idx == -1) { + ASSERT(0); + return 0; + } + + return stream_idx; +} + +void resource_log_pipe_topology_update(struct dc *dc, struct dc_state *state) +{ + struct pipe_ctx *otg_master; + int stream_idx, phantom_stream_idx; + DC_LOGGER_INIT(dc->ctx->logger); + DC_LOG_DC(" pipe topology update"); DC_LOG_DC(" ________________________"); for (stream_idx = 0; stream_idx < state->stream_count; stream_idx++) { + if (state->streams[stream_idx]->is_phantom) + continue; + otg_master = resource_get_otg_master_for_stream( &state->res_ctx, state->streams[stream_idx]); - if (!otg_master || otg_master->stream_res.tg == NULL) { - DC_LOG_DC("topology update: otg_master NULL stream_idx %d!\n", stream_idx); - return; - } - slice_count = resource_get_opp_heads_for_otg_master(otg_master, - &state->res_ctx, opp_heads); - for (slice_idx = 0; slice_idx < slice_count; slice_idx++) { - plane_idx = -1; - if (opp_heads[slice_idx]->plane_state) { - dpp_count = resource_get_dpp_pipes_for_opp_head( - opp_heads[slice_idx], - &state->res_ctx, - dpp_pipes); - for (dpp_idx = 0; dpp_idx < dpp_count; dpp_idx++) { - is_primary = !dpp_pipes[dpp_idx]->top_pipe || - dpp_pipes[dpp_idx]->top_pipe->plane_state != dpp_pipes[dpp_idx]->plane_state; - if (is_primary) - plane_idx++; - resource_log_pipe(dc, dpp_pipes[dpp_idx], - stream_idx, slice_idx, - plane_idx, slice_count, - is_primary); - } - } else { - resource_log_pipe(dc, opp_heads[slice_idx], - stream_idx, slice_idx, plane_idx, - slice_count, true); - } + resource_log_pipe_for_stream(dc, state, otg_master, stream_idx); + } + if (state->phantom_stream_count > 0) { + DC_LOG_DC(" | (phantom pipes) |"); + for (stream_idx = 0; stream_idx < state->stream_count; stream_idx++) { + if (state->stream_status[stream_idx].mall_stream_config.type != SUBVP_MAIN) + continue; + phantom_stream_idx = resource_stream_to_stream_idx(state, + state->stream_status[stream_idx].mall_stream_config.paired_stream); + otg_master = resource_get_otg_master_for_stream( + &state->res_ctx, state->streams[phantom_stream_idx]); + resource_log_pipe_for_stream(dc, state, otg_master, stream_idx); } } DC_LOG_DC(" |________________________|\n"); @@ -2266,6 +2334,9 @@ static bool update_pipe_params_after_odm_slice_count_change( if (pool->funcs->build_pipe_pix_clk_params) pool->funcs->build_pipe_pix_clk_params(otg_master); + + resource_build_test_pattern_params(&context->res_ctx, otg_master); + return result; } @@ -2624,13 +2695,19 @@ bool resource_append_dpp_pipes_for_plane_composition( struct pipe_ctx *otg_master_pipe, struct dc_plane_state *plane_state) { + bool success; if (otg_master_pipe->plane_state == NULL) - return add_plane_to_opp_head_pipes(otg_master_pipe, + success = add_plane_to_opp_head_pipes(otg_master_pipe, plane_state, new_ctx); else - return acquire_secondary_dpp_pipes_and_add_plane( + success = acquire_secondary_dpp_pipes_and_add_plane( otg_master_pipe, plane_state, new_ctx, cur_ctx, pool); + if (success) + /* when appending a plane mpc slice count changes from 0 to 1 */ + success = update_pipe_params_after_mpc_slice_count_change( + plane_state, new_ctx, pool); + return success; } void resource_remove_dpp_pipes_for_plane_composition( @@ -2965,7 +3042,7 @@ bool resource_update_pipes_for_plane_with_slice_count( int i; int dpp_pipe_count; int cur_slice_count; - struct pipe_ctx *dpp_pipes[MAX_PIPES]; + struct pipe_ctx *dpp_pipes[MAX_PIPES] = {0}; bool result = true; dpp_pipe_count = resource_get_dpp_pipes_for_plane(plane, @@ -3117,6 +3194,9 @@ static struct audio *find_first_free_audio( { int i, available_audio_count; + if (id == ENGINE_ID_UNKNOWN) + return NULL; + available_audio_count = pool->audio_count; for (i = 0; i < available_audio_count; i++) { @@ -3371,11 +3451,31 @@ static bool acquire_otg_master_pipe_for_stream( * any free pipes already used in current context as this could tear * down exiting ODM/MPC/MPO configuration unnecessarily. */ + + /* + * Try to acquire the same OTG master already in use. This is not + * optimal because resetting an enabled OTG master pipe for a new stream + * requires an extra frame of wait. However there are test automation + * and eDP assumptions that rely on reusing the same OTG master pipe + * during mode change. We have to keep this logic as is for now. + */ pipe_idx = recource_find_free_pipe_used_as_otg_master_in_cur_res_ctx( &cur_ctx->res_ctx, &new_ctx->res_ctx, pool); + /* + * Try to acquire a pipe not used in current resource context to avoid + * pipe swapping. + */ if (pipe_idx == FREE_PIPE_INDEX_NOT_FOUND) pipe_idx = recource_find_free_pipe_not_used_in_cur_res_ctx( &cur_ctx->res_ctx, &new_ctx->res_ctx, pool); + /* + * If pipe swapping is unavoidable, try to acquire pipe used as + * secondary DPP pipe in current state as we prioritize to support more + * streams over supporting MPO planes. + */ + if (pipe_idx == FREE_PIPE_INDEX_NOT_FOUND) + pipe_idx = resource_find_free_pipe_used_as_cur_sec_dpp( + &cur_ctx->res_ctx, &new_ctx->res_ctx, pool); if (pipe_idx == FREE_PIPE_INDEX_NOT_FOUND) pipe_idx = resource_find_any_free_pipe(&new_ctx->res_ctx, pool); if (pipe_idx != FREE_PIPE_INDEX_NOT_FOUND) { @@ -3990,7 +4090,7 @@ static void set_avi_info_frame( } if (pixel_encoding && color_space == COLOR_SPACE_2020_YCBCR && - stream->out_transfer_func->tf == TRANSFER_FUNCTION_GAMMA22) { + stream->out_transfer_func.tf == TRANSFER_FUNCTION_GAMMA22) { hdmi_info.bits.EC0_EC2 = 0; hdmi_info.bits.C0_C1 = COLORIMETRY_ITU709; } @@ -4992,3 +5092,39 @@ bool check_subvp_sw_cursor_fallback_req(const struct dc *dc, struct dc_stream_st return false; } + +void resource_init_common_dml2_callbacks(struct dc *dc, struct dml2_configuration_options *dml2_options) +{ + dml2_options->callbacks.dc = dc; + dml2_options->callbacks.build_scaling_params = &resource_build_scaling_params; + dml2_options->callbacks.build_test_pattern_params = &resource_build_test_pattern_params; + dml2_options->callbacks.acquire_secondary_pipe_for_mpc_odm = &dc_resource_acquire_secondary_pipe_for_mpc_odm_legacy; + dml2_options->callbacks.update_pipes_for_stream_with_slice_count = &resource_update_pipes_for_stream_with_slice_count; + dml2_options->callbacks.update_pipes_for_plane_with_slice_count = &resource_update_pipes_for_plane_with_slice_count; + dml2_options->callbacks.get_mpc_slice_index = &resource_get_mpc_slice_index; + dml2_options->callbacks.get_mpc_slice_count = &resource_get_mpc_slice_count; + dml2_options->callbacks.get_odm_slice_index = &resource_get_odm_slice_index; + dml2_options->callbacks.get_odm_slice_count = &resource_get_odm_slice_count; + dml2_options->callbacks.get_opp_head = &resource_get_opp_head; + dml2_options->callbacks.get_otg_master_for_stream = &resource_get_otg_master_for_stream; + dml2_options->callbacks.get_opp_heads_for_otg_master = &resource_get_opp_heads_for_otg_master; + dml2_options->callbacks.get_dpp_pipes_for_plane = &resource_get_dpp_pipes_for_plane; + dml2_options->callbacks.get_stream_status = &dc_state_get_stream_status; + dml2_options->callbacks.get_stream_from_id = &dc_state_get_stream_from_id; + + dml2_options->svp_pstate.callbacks.dc = dc; + dml2_options->svp_pstate.callbacks.add_phantom_plane = &dc_state_add_phantom_plane; + dml2_options->svp_pstate.callbacks.add_phantom_stream = &dc_state_add_phantom_stream; + dml2_options->svp_pstate.callbacks.build_scaling_params = &resource_build_scaling_params; + dml2_options->svp_pstate.callbacks.create_phantom_plane = &dc_state_create_phantom_plane; + dml2_options->svp_pstate.callbacks.remove_phantom_plane = &dc_state_remove_phantom_plane; + dml2_options->svp_pstate.callbacks.remove_phantom_stream = &dc_state_remove_phantom_stream; + dml2_options->svp_pstate.callbacks.create_phantom_stream = &dc_state_create_phantom_stream; + dml2_options->svp_pstate.callbacks.release_phantom_plane = &dc_state_release_phantom_plane; + dml2_options->svp_pstate.callbacks.release_phantom_stream = &dc_state_release_phantom_stream; + dml2_options->svp_pstate.callbacks.get_pipe_subvp_type = &dc_state_get_pipe_subvp_type; + dml2_options->svp_pstate.callbacks.get_stream_subvp_type = &dc_state_get_stream_subvp_type; + dml2_options->svp_pstate.callbacks.get_paired_subvp_stream = &dc_state_get_paired_subvp_stream; + dml2_options->svp_pstate.callbacks.remove_phantom_streams_and_planes = &dc_state_remove_phantom_streams_and_planes; + dml2_options->svp_pstate.callbacks.release_phantom_streams_and_planes = &dc_state_release_phantom_streams_and_planes; +} diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_stat.c b/drivers/gpu/drm/amd/display/dc/core/dc_stat.c index 5f6392ae31a6..cd6570a1e20e 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_stat.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_stat.c @@ -61,7 +61,7 @@ void dc_stat_get_dmub_notification(const struct dc *dc, struct dmub_notification /* For HPD/HPD RX, convert dpia port index into link index */ if (notify->type == DMUB_NOTIFICATION_HPD || notify->type == DMUB_NOTIFICATION_HPD_IRQ || - notify->type == DMUB_NOTIFICATION_DPIA_NOTIFICATION || + notify->type == DMUB_NOTIFICATION_DPIA_NOTIFICATION || notify->type == DMUB_NOTIFICATION_SET_CONFIG_REPLY) { notify->link_index = get_link_index_from_dpia_port_index(dc, notify->link_index); diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_state.c b/drivers/gpu/drm/amd/display/dc/core/dc_state.c index 5cc7f8da209c..76bb05f4d6bf 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_state.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_state.c @@ -188,8 +188,11 @@ static void init_state(struct dc *dc, struct dc_state *state) } /* Public dc_state functions */ -struct dc_state *dc_state_create(struct dc *dc) +struct dc_state *dc_state_create(struct dc *dc, struct dc_state_create_params *params) { +#ifdef CONFIG_DRM_AMD_DC_FP + struct dml2_configuration_options *dml2_opt = &dc->dml2_options; +#endif struct dc_state *state = kvzalloc(sizeof(struct dc_state), GFP_KERNEL); @@ -198,10 +201,16 @@ struct dc_state *dc_state_create(struct dc *dc) init_state(dc, state); dc_state_construct(dc, state); + state->power_source = params ? params->power_source : DC_POWER_SOURCE_AC; #ifdef CONFIG_DRM_AMD_DC_FP - if (dc->debug.using_dml2) - dml2_create(dc, &dc->dml2_options, &state->bw_ctx.dml2); + if (dc->debug.using_dml2) { + dml2_opt->use_clock_dc_limits = false; + dml2_create(dc, dml2_opt, &state->bw_ctx.dml2); + + dml2_opt->use_clock_dc_limits = true; + dml2_create(dc, dml2_opt, &state->bw_ctx.dml2_dc_power_source); + } #endif kref_init(&state->refcount); @@ -214,6 +223,7 @@ void dc_state_copy(struct dc_state *dst_state, struct dc_state *src_state) struct kref refcount = dst_state->refcount; #ifdef CONFIG_DRM_AMD_DC_FP struct dml2_context *dst_dml2 = dst_state->bw_ctx.dml2; + struct dml2_context *dst_dml2_dc_power_source = dst_state->bw_ctx.dml2_dc_power_source; #endif dc_state_copy_internal(dst_state, src_state); @@ -222,6 +232,10 @@ void dc_state_copy(struct dc_state *dst_state, struct dc_state *src_state) dst_state->bw_ctx.dml2 = dst_dml2; if (src_state->bw_ctx.dml2) dml2_copy(dst_state->bw_ctx.dml2, src_state->bw_ctx.dml2); + + dst_state->bw_ctx.dml2_dc_power_source = dst_dml2_dc_power_source; + if (src_state->bw_ctx.dml2_dc_power_source) + dml2_copy(dst_state->bw_ctx.dml2_dc_power_source, src_state->bw_ctx.dml2_dc_power_source); #endif /* context refcount should not be overridden */ @@ -245,6 +259,12 @@ struct dc_state *dc_state_create_copy(struct dc_state *src_state) dc_state_release(new_state); return NULL; } + + if (src_state->bw_ctx.dml2_dc_power_source && + !dml2_create_copy(&new_state->bw_ctx.dml2_dc_power_source, src_state->bw_ctx.dml2_dc_power_source)) { + dc_state_release(new_state); + return NULL; + } #endif kref_init(&new_state->refcount); @@ -310,7 +330,6 @@ void dc_state_destruct(struct dc_state *state) memset(state->dc_dmub_cmd, 0, sizeof(state->dc_dmub_cmd)); state->dmub_cmd_count = 0; memset(&state->perf_params, 0, sizeof(state->perf_params)); - memset(&state->scratch, 0, sizeof(state->scratch)); } void dc_state_retain(struct dc_state *state) @@ -327,6 +346,9 @@ static void dc_state_free(struct kref *kref) #ifdef CONFIG_DRM_AMD_DC_FP dml2_destroy(state->bw_ctx.dml2); state->bw_ctx.dml2 = 0; + + dml2_destroy(state->bw_ctx.dml2_dc_power_source); + state->bw_ctx.dml2_dc_power_source = 0; #endif kvfree(state); @@ -341,7 +363,7 @@ void dc_state_release(struct dc_state *state) * dc_state_add_stream() - Add a new dc_stream_state to a dc_state. */ enum dc_status dc_state_add_stream( - struct dc *dc, + const struct dc *dc, struct dc_state *state, struct dc_stream_state *stream) { @@ -370,7 +392,7 @@ enum dc_status dc_state_add_stream( * dc_state_remove_stream() - Remove a stream from a dc_state. */ enum dc_status dc_state_remove_stream( - struct dc *dc, + const struct dc *dc, struct dc_state *state, struct dc_stream_state *stream) { @@ -436,6 +458,15 @@ bool dc_state_add_plane( goto out; } + if (stream_status->plane_count == 0 && dc->config.enable_windowed_mpo_odm) + /* ODM combine could prevent us from supporting more planes + * we will reset ODM slice count back to 1 when all planes have + * been removed to maximize the amount of planes supported when + * new planes are added. + */ + resource_update_pipes_for_stream_with_slice_count( + state, dc->current_state, dc->res_pool, stream, 1); + otg_master_pipe = resource_get_otg_master_for_stream( &state->res_ctx, stream); if (otg_master_pipe) @@ -586,7 +617,7 @@ bool dc_state_add_all_planes_for_stream( */ struct dc_stream_status *dc_state_get_stream_status( struct dc_state *state, - struct dc_stream_state *stream) + const struct dc_stream_state *stream) { uint8_t i; @@ -680,7 +711,7 @@ void dc_state_release_phantom_stream(const struct dc *dc, dc_stream_release(phantom_stream); } -struct dc_plane_state *dc_state_create_phantom_plane(struct dc *dc, +struct dc_plane_state *dc_state_create_phantom_plane(const struct dc *dc, struct dc_state *state, struct dc_plane_state *main_plane) { @@ -716,7 +747,7 @@ void dc_state_release_phantom_plane(const struct dc *dc, } /* add phantom streams to context and generate correct meta inside dc_state */ -enum dc_status dc_state_add_phantom_stream(struct dc *dc, +enum dc_status dc_state_add_phantom_stream(const struct dc *dc, struct dc_state *state, struct dc_stream_state *phantom_stream, struct dc_stream_state *main_stream) @@ -742,7 +773,7 @@ enum dc_status dc_state_add_phantom_stream(struct dc *dc, return res; } -enum dc_status dc_state_remove_phantom_stream(struct dc *dc, +enum dc_status dc_state_remove_phantom_stream(const struct dc *dc, struct dc_state *state, struct dc_stream_state *phantom_stream) { @@ -836,7 +867,7 @@ bool dc_state_add_all_phantom_planes_for_stream( } bool dc_state_remove_phantom_streams_and_planes( - struct dc *dc, + const struct dc *dc, struct dc_state *state) { int i; @@ -858,7 +889,7 @@ bool dc_state_remove_phantom_streams_and_planes( } void dc_state_release_phantom_streams_and_planes( - struct dc *dc, + const struct dc *dc, struct dc_state *state) { int i; @@ -869,3 +900,19 @@ void dc_state_release_phantom_streams_and_planes( for (i = 0; i < state->phantom_plane_count; i++) dc_state_release_phantom_plane(dc, state, state->phantom_planes[i]); } + +struct dc_stream_state *dc_state_get_stream_from_id(const struct dc_state *state, unsigned int id) +{ + struct dc_stream_state *stream = NULL; + int i; + + for (i = 0; i < state->stream_count; i++) { + if (state->streams[i] && state->streams[i]->stream_id == id) { + stream = state->streams[i]; + break; + } + } + + return stream; +} + diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_stream.c b/drivers/gpu/drm/amd/display/dc/core/dc_stream.c index 51a970fcb5d0..5c7e4884cac2 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_stream.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_stream.c @@ -116,12 +116,7 @@ bool dc_stream_construct(struct dc_stream_state *stream, update_stream_signal(stream, dc_sink_data); - stream->out_transfer_func = dc_create_transfer_func(); - if (stream->out_transfer_func == NULL) { - dc_sink_release(dc_sink_data); - return false; - } - stream->out_transfer_func->type = TF_TYPE_BYPASS; + stream->out_transfer_func.type = TF_TYPE_BYPASS; dc_stream_assign_stream_id(stream); @@ -131,10 +126,6 @@ bool dc_stream_construct(struct dc_stream_state *stream, void dc_stream_destruct(struct dc_stream_state *stream) { dc_sink_release(stream->sink); - if (stream->out_transfer_func != NULL) { - dc_transfer_func_release(stream->out_transfer_func); - stream->out_transfer_func = NULL; - } } void dc_stream_assign_stream_id(struct dc_stream_state *stream) @@ -201,9 +192,6 @@ struct dc_stream_state *dc_copy_stream(const struct dc_stream_state *stream) if (new_stream->sink) dc_sink_retain(new_stream->sink); - if (new_stream->out_transfer_func) - dc_transfer_func_retain(new_stream->out_transfer_func); - dc_stream_assign_stream_id(new_stream); /* If using dynamic encoder assignment, wait till stream committed to assign encoder. */ @@ -319,7 +307,7 @@ bool dc_stream_set_cursor_attributes( program_cursor_attributes(dc, stream, attributes); /* re-enable idle optimizations if necessary */ - if (reset_idle_optimizations) + if (reset_idle_optimizations && !dc->debug.disable_dmub_reallow_idle) dc_allow_idle_optimizations(dc, true); return true; @@ -394,7 +382,7 @@ bool dc_stream_set_cursor_position( program_cursor_position(dc, stream, position); /* re-enable idle optimizations if necessary */ - if (reset_idle_optimizations) + if (reset_idle_optimizations && !dc->debug.disable_dmub_reallow_idle) dc_allow_idle_optimizations(dc, true); return true; @@ -425,7 +413,7 @@ bool dc_stream_add_writeback(struct dc *dc, dc_exit_ips_for_hw_access(dc); - wb_info->dwb_params.out_transfer_func = stream->out_transfer_func; + wb_info->dwb_params.out_transfer_func = &stream->out_transfer_func; dwb = dc->res_pool->dwbc[wb_info->dwb_pipe_inst]; dwb->dwb_is_drc = false; @@ -507,7 +495,7 @@ bool dc_stream_remove_writeback(struct dc *dc, struct dc_stream_state *stream, uint32_t dwb_pipe_inst) { - int i = 0, j = 0; + unsigned int i, j; if (stream == NULL) { dm_error("DC: dc_stream is NULL!\n"); return false; diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_surface.c b/drivers/gpu/drm/amd/display/dc/core/dc_surface.c index 19140fb65787..067f6555cfdf 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_surface.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_surface.c @@ -41,25 +41,15 @@ void dc_plane_construct(struct dc_context *ctx, struct dc_plane_state *plane_sta { plane_state->ctx = ctx; - plane_state->gamma_correction = dc_create_gamma(); - if (plane_state->gamma_correction != NULL) - plane_state->gamma_correction->is_identity = true; + plane_state->gamma_correction.is_identity = true; - plane_state->in_transfer_func = dc_create_transfer_func(); - if (plane_state->in_transfer_func != NULL) { - plane_state->in_transfer_func->type = TF_TYPE_BYPASS; - } - plane_state->in_shaper_func = dc_create_transfer_func(); - if (plane_state->in_shaper_func != NULL) { - plane_state->in_shaper_func->type = TF_TYPE_BYPASS; - } + plane_state->in_transfer_func.type = TF_TYPE_BYPASS; - plane_state->lut3d_func = dc_create_3dlut_func(); + plane_state->in_shaper_func.type = TF_TYPE_BYPASS; - plane_state->blend_tf = dc_create_transfer_func(); - if (plane_state->blend_tf != NULL) { - plane_state->blend_tf->type = TF_TYPE_BYPASS; - } + plane_state->lut3d_func.state.raw = 0; + + plane_state->blend_tf.type = TF_TYPE_BYPASS; plane_state->pre_multiplied_alpha = true; @@ -67,30 +57,27 @@ void dc_plane_construct(struct dc_context *ctx, struct dc_plane_state *plane_sta void dc_plane_destruct(struct dc_plane_state *plane_state) { - if (plane_state->gamma_correction != NULL) { - dc_gamma_release(&plane_state->gamma_correction); - } - if (plane_state->in_transfer_func != NULL) { - dc_transfer_func_release( - plane_state->in_transfer_func); - plane_state->in_transfer_func = NULL; - } - if (plane_state->in_shaper_func != NULL) { - dc_transfer_func_release( - plane_state->in_shaper_func); - plane_state->in_shaper_func = NULL; - } - if (plane_state->lut3d_func != NULL) { - dc_3dlut_func_release( - plane_state->lut3d_func); - plane_state->lut3d_func = NULL; - } - if (plane_state->blend_tf != NULL) { - dc_transfer_func_release( - plane_state->blend_tf); - plane_state->blend_tf = NULL; + // no more pointers to free within dc_plane_state +} + + +/* dc_state is passed in separately since it may differ from the current dc state accessible from plane_state e.g. + * if the driver is doing an update from an old context to a new one and the caller wants the pipe mask for the new + * context rather than the existing one + */ +uint8_t dc_plane_get_pipe_mask(struct dc_state *dc_state, const struct dc_plane_state *plane_state) +{ + uint8_t pipe_mask = 0; + int i; + + for (i = 0; i < plane_state->ctx->dc->res_pool->pipe_count; i++) { + struct pipe_ctx *pipe_ctx = &dc_state->res_ctx.pipe_ctx[i]; + + if (pipe_ctx->plane_state == plane_state && pipe_ctx->plane_res.hubp) + pipe_mask |= 1 << pipe_ctx->plane_res.hubp->inst; } + return pipe_mask; } /******************************************************************************* @@ -103,7 +90,7 @@ void enable_surface_flip_reporting(struct dc_plane_state *plane_state, /*register_flip_interrupt(surface);*/ } -struct dc_plane_state *dc_create_plane_state(struct dc *dc) +struct dc_plane_state *dc_create_plane_state(const struct dc *dc) { struct dc_plane_state *plane_state = kvzalloc(sizeof(*plane_state), GFP_KERNEL); diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h index ee8453bf958f..3c33c3bcbe2c 100644 --- a/drivers/gpu/drm/amd/display/dc/dc.h +++ b/drivers/gpu/drm/amd/display/dc/dc.h @@ -44,6 +44,8 @@ #include "dml2/dml2_wrapper.h" +#include "dmub/inc/dmub_cmd.h" + struct abm_save_restore; /* forward declaration */ @@ -51,7 +53,7 @@ struct aux_payload; struct set_config_cmd_payload; struct dmub_notification; -#define DC_VER "3.2.273" +#define DC_VER "3.2.281" #define MAX_SURFACES 3 #define MAX_PLANES 6 @@ -219,6 +221,7 @@ struct dc_dmub_caps { bool mclk_sw; bool subvp_psr; bool gecc_enable; + uint8_t fams_ver; }; struct dc_caps { @@ -306,12 +309,12 @@ struct dc_dcc_setting { unsigned int max_compressed_blk_size; unsigned int max_uncompressed_blk_size; bool independent_64b_blks; - //These bitfields to be used starting with DCN + //These bitfields to be used starting with DCN 3.0 struct { - uint32_t dcc_256_64_64 : 1;//available in ASICs before DCN (the worst compression case) - uint32_t dcc_128_128_uncontrained : 1; //available in ASICs before DCN - uint32_t dcc_256_128_128 : 1; //available starting with DCN - uint32_t dcc_256_256_unconstrained : 1; //available in ASICs before DCN (the best compression case) + uint32_t dcc_256_64_64 : 1;//available in ASICs before DCN 3.0 (the worst compression case) + uint32_t dcc_128_128_uncontrained : 1; //available in ASICs before DCN 3.0 + uint32_t dcc_256_128_128 : 1; //available starting with DCN 3.0 + uint32_t dcc_256_256_unconstrained : 1; //available in ASICs before DCN 3.0 (the best compression case) } dcc_controls; }; @@ -435,6 +438,9 @@ struct dc_config { unsigned int disable_ips; unsigned int disable_ips_in_vpb; bool usb4_bw_alloc_support; + bool allow_0_dtb_clk; + bool use_assr_psp_message; + bool support_edp0_on_dp1; }; enum visual_confirm { @@ -693,6 +699,8 @@ enum pg_hw_pipe_resources { PG_MPCC, PG_OPP, PG_OPTC, + PG_DPSTREAM, + PG_HDMISTREAM, PG_HW_PIPE_RESOURCES_NUM_ELEMENT }; @@ -987,14 +995,17 @@ struct dc_debug_options { bool psp_disabled_wa; unsigned int ips2_eval_delay_us; unsigned int ips2_entry_delay_us; + bool optimize_ips_handshake; bool disable_dmub_reallow_idle; bool disable_timeout; bool disable_extblankadj; + bool enable_idle_reg_checks; unsigned int static_screen_wait_frames; bool force_chroma_subsampling_1tap; + bool disable_422_left_edge_pixel; + unsigned int force_cositing; }; -struct gpu_info_soc_bounding_box_v1_0; /* Generic structure that can be used to query properties of DC. More fields * can be added as required. @@ -1003,76 +1014,6 @@ struct dc_current_properties { unsigned int cursor_size_limit; }; -struct dc { - struct dc_debug_options debug; - struct dc_versions versions; - struct dc_caps caps; - struct dc_cap_funcs cap_funcs; - struct dc_config config; - struct dc_bounding_box_overrides bb_overrides; - struct dc_bug_wa work_arounds; - struct dc_context *ctx; - struct dc_phy_addr_space_config vm_pa_config; - - uint8_t link_count; - struct dc_link *links[MAX_PIPES * 2]; - struct link_service *link_srv; - - struct dc_state *current_state; - struct resource_pool *res_pool; - - struct clk_mgr *clk_mgr; - - /* Display Engine Clock levels */ - struct dm_pp_clock_levels sclk_lvls; - - /* Inputs into BW and WM calculations. */ - struct bw_calcs_dceip *bw_dceip; - struct bw_calcs_vbios *bw_vbios; - struct dcn_soc_bounding_box *dcn_soc; - struct dcn_ip_params *dcn_ip; - struct display_mode_lib dml; - - /* HW functions */ - struct hw_sequencer_funcs hwss; - struct dce_hwseq *hwseq; - - /* Require to optimize clocks and bandwidth for added/removed planes */ - bool optimized_required; - bool wm_optimized_required; - bool idle_optimizations_allowed; - bool enable_c20_dtm_b0; - - /* Require to maintain clocks and bandwidth for UEFI enabled HW */ - - /* FBC compressor */ - struct compressor *fbc_compressor; - - struct dc_debug_data debug_data; - struct dpcd_vendor_signature vendor_signature; - - const char *build_id; - struct vm_helper *vm_helper; - - uint32_t *dcn_reg_offsets; - uint32_t *nbio_reg_offsets; - uint32_t *clk_reg_offsets; - - /* Scratch memory */ - struct { - struct { - /* - * For matching clock_limits table in driver with table - * from PMFW. - */ - struct _vcs_dpi_voltage_scaling_st clock_limits[DC__VOLTAGE_STATES]; - } update_bw_bounding_box; - } scratch; - - struct dml2_configuration_options dml2_options; - enum dc_acpi_cm_power_state power_state; -}; - enum frame_buffer_mode { FRAME_BUFFER_MODE_LOCAL_ONLY = 0, FRAME_BUFFER_MODE_ZFB_ONLY, @@ -1277,6 +1218,8 @@ union surface_update_flags { uint32_t raw; }; +#define DC_REMOVE_PLANE_POINTERS 1 + struct dc_plane_state { struct dc_plane_address address; struct dc_plane_flip_time time; @@ -1291,8 +1234,8 @@ struct dc_plane_state { struct dc_plane_dcc_param dcc; - struct dc_gamma *gamma_correction; - struct dc_transfer_func *in_transfer_func; + struct dc_gamma gamma_correction; + struct dc_transfer_func in_transfer_func; struct dc_bias_and_scale *bias_and_scale; struct dc_csc_transform input_csc_color_matrix; struct fixed31_32 coeff_reduction_factor; @@ -1304,9 +1247,9 @@ struct dc_plane_state { enum dc_color_space color_space; - struct dc_3dlut *lut3d_func; - struct dc_transfer_func *in_shaper_func; - struct dc_transfer_func *blend_tf; + struct dc_3dlut lut3d_func; + struct dc_transfer_func in_shaper_func; + struct dc_transfer_func blend_tf; struct dc_transfer_func *gamcor_tf; enum surface_pixel_format format; @@ -1342,6 +1285,7 @@ struct dc_plane_state { struct tg_color visual_confirm_color; bool is_statically_allocated; + enum chroma_cositing cositing; }; struct dc_plane_info { @@ -1360,6 +1304,96 @@ struct dc_plane_info { int global_alpha_value; bool input_csc_enabled; int layer_index; + enum chroma_cositing cositing; +}; + +#include "dc_stream.h" + +struct dc_scratch_space { + /* used to temporarily backup plane states of a stream during + * dc update. The reason is that plane states are overwritten + * with surface updates in dc update. Once they are overwritten + * current state is no longer valid. We want to temporarily + * store current value in plane states so we can still recover + * a valid current state during dc update. + */ + struct dc_plane_state plane_states[MAX_SURFACE_NUM]; + + struct dc_stream_state stream_state; +}; + +struct dc { + struct dc_debug_options debug; + struct dc_versions versions; + struct dc_caps caps; + struct dc_cap_funcs cap_funcs; + struct dc_config config; + struct dc_bounding_box_overrides bb_overrides; + struct dc_bug_wa work_arounds; + struct dc_context *ctx; + struct dc_phy_addr_space_config vm_pa_config; + + uint8_t link_count; + struct dc_link *links[MAX_LINKS]; + struct link_service *link_srv; + + struct dc_state *current_state; + struct resource_pool *res_pool; + + struct clk_mgr *clk_mgr; + + /* Display Engine Clock levels */ + struct dm_pp_clock_levels sclk_lvls; + + /* Inputs into BW and WM calculations. */ + struct bw_calcs_dceip *bw_dceip; + struct bw_calcs_vbios *bw_vbios; + struct dcn_soc_bounding_box *dcn_soc; + struct dcn_ip_params *dcn_ip; + struct display_mode_lib dml; + + /* HW functions */ + struct hw_sequencer_funcs hwss; + struct dce_hwseq *hwseq; + + /* Require to optimize clocks and bandwidth for added/removed planes */ + bool optimized_required; + bool wm_optimized_required; + bool idle_optimizations_allowed; + bool enable_c20_dtm_b0; + + /* Require to maintain clocks and bandwidth for UEFI enabled HW */ + + /* FBC compressor */ + struct compressor *fbc_compressor; + + struct dc_debug_data debug_data; + struct dpcd_vendor_signature vendor_signature; + + const char *build_id; + struct vm_helper *vm_helper; + + uint32_t *dcn_reg_offsets; + uint32_t *nbio_reg_offsets; + uint32_t *clk_reg_offsets; + + /* Scratch memory */ + struct { + struct { + /* + * For matching clock_limits table in driver with table + * from PMFW. + */ + struct _vcs_dpi_voltage_scaling_st clock_limits[DC__VOLTAGE_STATES]; + } update_bw_bounding_box; + struct dc_scratch_space current_state; + struct dc_scratch_space new_state; + struct dc_stream_state temp_stream; // Used so we don't need to allocate stream on the stack + } scratch; + + struct dml2_configuration_options dml2_options; + enum dc_acpi_cm_power_state power_state; + }; struct dc_scaling_info { @@ -1476,10 +1510,15 @@ bool dc_acquire_release_mpc_3dlut( bool dc_resource_is_dsc_encoding_supported(const struct dc *dc); void get_audio_check(struct audio_info *aud_modes, struct audio_check *aud_chk); - -enum dc_status dc_commit_streams(struct dc *dc, - struct dc_stream_state *streams[], - uint8_t stream_count); +/* + * Set up streams and links associated to drive sinks + * The streams parameter is an absolute set of all active streams. + * + * After this call: + * Phy, Encoder, Timing Generator are programmed and enabled. + * New streams are enabled with blank stream; no memory read. + */ +enum dc_status dc_commit_streams(struct dc *dc, struct dc_commit_streams_params *params); struct dc_plane_state *dc_get_surface_for_mpcc(struct dc *dc, @@ -2335,11 +2374,17 @@ bool dc_is_dmcu_initialized(struct dc *dc); enum dc_status dc_set_clock(struct dc *dc, enum dc_clock_type clock_type, uint32_t clk_khz, uint32_t stepping); void dc_get_clock(struct dc *dc, enum dc_clock_type clock_type, struct dc_clock_config *clock_cfg); -bool dc_is_plane_eligible_for_idle_optimizations(struct dc *dc, struct dc_plane_state *plane, - struct dc_cursor_attributes *cursor_attr); +bool dc_is_plane_eligible_for_idle_optimizations(struct dc *dc, + unsigned int pitch, + unsigned int height, + enum surface_pixel_format format, + struct dc_cursor_attributes *cursor_attr); + +#define dc_allow_idle_optimizations(dc, allow) dc_allow_idle_optimizations_internal(dc, allow, __func__) +#define dc_exit_ips_for_hw_access(dc) dc_exit_ips_for_hw_access_internal(dc, __func__) -void dc_allow_idle_optimizations(struct dc *dc, bool allow); -void dc_exit_ips_for_hw_access(struct dc *dc); +void dc_allow_idle_optimizations_internal(struct dc *dc, bool allow, const char *caller_name); +void dc_exit_ips_for_hw_access_internal(struct dc *dc, const char *caller_name); bool dc_dmub_is_ips_idle_state(struct dc *dc); /* set min and max memory clock to lowest and highest DPM level, respectively */ 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 6083b1dcf050..2293a92df3be 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c +++ b/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c @@ -23,6 +23,7 @@ * */ +#include "dm_services.h" #include "dc.h" #include "dc_dmub_srv.h" #include "../dmub/dmub_srv.h" @@ -34,6 +35,7 @@ #include "resource.h" #include "clk_mgr.h" #include "dc_state_priv.h" +#include "dc_plane_priv.h" #define CTX dc_dmub_srv->ctx #define DC_LOGGER CTX->logger @@ -198,6 +200,11 @@ bool dc_dmub_srv_wait_for_idle(struct dc_dmub_srv *dc_dmub_srv, if (status != DMUB_STATUS_OK) { DC_LOG_DEBUG("No reply for DMUB command: status=%d\n", status); + if (!dmub->debug.timeout_occured) { + dmub->debug.timeout_occured = true; + dmub->debug.timeout_cmd = *cmd_list; + dmub->debug.timestamp = dm_get_timestamp(dc_dmub_srv->ctx); + } dc_dmub_srv_log_diagnostic_data(dc_dmub_srv); return false; } @@ -904,12 +911,15 @@ bool dc_dmub_srv_get_diagnostic_data(struct dc_dmub_srv *dc_dmub_srv, struct dmu void dc_dmub_srv_log_diagnostic_data(struct dc_dmub_srv *dc_dmub_srv) { struct dmub_diagnostic_data diag_data = {0}; + uint32_t i; if (!dc_dmub_srv || !dc_dmub_srv->dmub) { DC_LOG_ERROR("%s: invalid parameters.", __func__); return; } + DC_LOG_ERROR("%s: DMCUB error - collecting diagnostic data\n", __func__); + if (!dc_dmub_srv_get_diagnostic_data(dc_dmub_srv, &diag_data)) { DC_LOG_ERROR("%s: dc_dmub_srv_get_diagnostic_data failed.", __func__); return; @@ -933,7 +943,8 @@ void dc_dmub_srv_log_diagnostic_data(struct dc_dmub_srv *dc_dmub_srv) DC_LOG_DEBUG(" scratch [13] : %08x", diag_data.scratch[13]); DC_LOG_DEBUG(" scratch [14] : %08x", diag_data.scratch[14]); DC_LOG_DEBUG(" scratch [15] : %08x", diag_data.scratch[15]); - DC_LOG_DEBUG(" pc : %08x", diag_data.pc); + for (i = 0; i < DMUB_PC_SNAPSHOT_COUNT; i++) + DC_LOG_DEBUG(" pc[%d] : %08x", i, diag_data.pc[i]); DC_LOG_DEBUG(" unk_fault_addr : %08x", diag_data.undefined_address_fault_addr); DC_LOG_DEBUG(" inst_fault_addr : %08x", diag_data.inst_fetch_fault_addr); DC_LOG_DEBUG(" data_fault_addr : %08x", diag_data.data_write_fault_addr); @@ -1199,8 +1210,23 @@ bool dc_dmub_srv_is_hw_pwr_up(struct dc_dmub_srv *dc_dmub_srv, bool wait) return true; } +static int count_active_streams(const struct dc *dc) +{ + int i, count = 0; + + for (i = 0; i < dc->current_state->stream_count; ++i) { + struct dc_stream_state *stream = dc->current_state->streams[i]; + + if (stream && !stream->dpms_off) + count += 1; + } + + return count; +} + static void dc_dmub_srv_notify_idle(const struct dc *dc, bool allow_idle) { + volatile const struct dmub_shared_state_ips_fw *ips_fw; struct dc_dmub_srv *dc_dmub_srv; union dmub_rb_cmd cmd = {0}; @@ -1211,6 +1237,7 @@ static void dc_dmub_srv_notify_idle(const struct dc *dc, bool allow_idle) return; dc_dmub_srv = dc->ctx->dmub_srv; + ips_fw = &dc_dmub_srv->dmub->shared_state[DMUB_SHARED_SHARE_FEATURE__IPS_FW].data.ips_fw; memset(&cmd, 0, sizeof(cmd)); cmd.idle_opt_notify_idle.header.type = DMUB_CMD__IDLE_OPT; @@ -1226,6 +1253,12 @@ static void dc_dmub_srv_notify_idle(const struct dc *dc, bool allow_idle) &dc_dmub_srv->dmub->shared_state[DMUB_SHARED_SHARE_FEATURE__IPS_DRIVER].data.ips_driver; union dmub_shared_state_ips_driver_signals new_signals; + DC_LOG_IPS( + "%s wait idle (ips1_commit=%d ips2_commit=%d)", + __func__, + ips_fw->signals.bits.ips1_commit, + ips_fw->signals.bits.ips2_commit); + dc_dmub_srv_wait_idle(dc->ctx->dmub_srv); memset(&new_signals, 0, sizeof(new_signals)); @@ -1245,19 +1278,46 @@ static void dc_dmub_srv_notify_idle(const struct dc *dc, bool allow_idle) new_signals.bits.allow_pg = 1; new_signals.bits.allow_ips1 = 1; new_signals.bits.allow_ips2 = 1; + } else if (dc->config.disable_ips == DMUB_IPS_RCG_IN_ACTIVE_IPS2_IN_OFF) { + /* TODO: Move this logic out to hwseq */ + if (count_active_streams(dc) == 0) { + /* IPS2 - Display off */ + new_signals.bits.allow_pg = 1; + new_signals.bits.allow_ips1 = 1; + new_signals.bits.allow_ips2 = 1; + new_signals.bits.allow_z10 = 1; + } else { + /* RCG only */ + new_signals.bits.allow_pg = 0; + new_signals.bits.allow_ips1 = 1; + new_signals.bits.allow_ips2 = 0; + new_signals.bits.allow_z10 = 0; + } } ips_driver->signals = new_signals; } + DC_LOG_IPS( + "%s send allow_idle=%d (ips1_commit=%d ips2_commit=%d)", + __func__, + allow_idle, + ips_fw->signals.bits.ips1_commit, + ips_fw->signals.bits.ips2_commit); + /* NOTE: This does not use the "wake" interface since this is part of the wake path. */ /* We also do not perform a wait since DMCUB could enter idle after the notification. */ dm_execute_dmub_cmd(dc->ctx, &cmd, allow_idle ? DM_DMUB_WAIT_TYPE_NO_WAIT : DM_DMUB_WAIT_TYPE_WAIT); + + /* Register access should stop at this point. */ + if (allow_idle) + dc_dmub_srv->needs_idle_wake = true; } static void dc_dmub_srv_exit_low_power_state(const struct dc *dc) { struct dc_dmub_srv *dc_dmub_srv; + uint32_t rcg_exit_count = 0, ips1_exit_count = 0, ips2_exit_count = 0; if (dc->debug.dmcub_emulation) return; @@ -1274,40 +1334,113 @@ static void dc_dmub_srv_exit_low_power_state(const struct dc *dc) &dc_dmub_srv->dmub->shared_state[DMUB_SHARED_SHARE_FEATURE__IPS_DRIVER].data.ips_driver; union dmub_shared_state_ips_driver_signals prev_driver_signals = ips_driver->signals; + rcg_exit_count = ips_fw->rcg_exit_count; + ips1_exit_count = ips_fw->ips1_exit_count; + ips2_exit_count = ips_fw->ips2_exit_count; + ips_driver->signals.all = 0; - if (prev_driver_signals.bits.allow_ips2) { - udelay(dc->debug.ips2_eval_delay_us); + DC_LOG_IPS( + "%s (allow ips1=%d ips2=%d) (commit ips1=%d ips2=%d) (count rcg=%d ips1=%d ips2=%d)", + __func__, + ips_driver->signals.bits.allow_ips1, + ips_driver->signals.bits.allow_ips2, + ips_fw->signals.bits.ips1_commit, + ips_fw->signals.bits.ips2_commit, + ips_fw->rcg_entry_count, + ips_fw->ips1_entry_count, + ips_fw->ips2_entry_count); + + /* Note: register access has technically not resumed for DCN here, but we + * need to be message PMFW through our standard register interface. + */ + dc_dmub_srv->needs_idle_wake = false; + + if (prev_driver_signals.bits.allow_ips2 && + (!dc->debug.optimize_ips_handshake || + ips_fw->signals.bits.ips2_commit || !ips_fw->signals.bits.in_idle)) { + DC_LOG_IPS( + "wait IPS2 eval (ips1_commit=%d ips2_commit=%d)", + ips_fw->signals.bits.ips1_commit, + ips_fw->signals.bits.ips2_commit); + + if (!dc->debug.optimize_ips_handshake || !ips_fw->signals.bits.ips2_commit) + udelay(dc->debug.ips2_eval_delay_us); if (ips_fw->signals.bits.ips2_commit) { + DC_LOG_IPS( + "exit IPS2 #1 (ips1_commit=%d ips2_commit=%d)", + ips_fw->signals.bits.ips1_commit, + ips_fw->signals.bits.ips2_commit); + // Tell PMFW to exit low power state dc->clk_mgr->funcs->exit_low_power_state(dc->clk_mgr); + DC_LOG_IPS( + "wait IPS2 entry delay (ips1_commit=%d ips2_commit=%d)", + ips_fw->signals.bits.ips1_commit, + ips_fw->signals.bits.ips2_commit); + // Wait for IPS2 entry upper bound udelay(dc->debug.ips2_entry_delay_us); + DC_LOG_IPS( + "exit IPS2 #2 (ips1_commit=%d ips2_commit=%d)", + ips_fw->signals.bits.ips1_commit, + ips_fw->signals.bits.ips2_commit); + dc->clk_mgr->funcs->exit_low_power_state(dc->clk_mgr); + DC_LOG_IPS( + "wait IPS2 commit clear (ips1_commit=%d ips2_commit=%d)", + ips_fw->signals.bits.ips1_commit, + ips_fw->signals.bits.ips2_commit); + while (ips_fw->signals.bits.ips2_commit) udelay(1); + DC_LOG_IPS( + "wait hw_pwr_up (ips1_commit=%d ips2_commit=%d)", + ips_fw->signals.bits.ips1_commit, + ips_fw->signals.bits.ips2_commit); + if (!dc_dmub_srv_is_hw_pwr_up(dc->ctx->dmub_srv, true)) ASSERT(0); + DC_LOG_IPS( + "resync inbox1 (ips1_commit=%d ips2_commit=%d)", + ips_fw->signals.bits.ips1_commit, + ips_fw->signals.bits.ips2_commit); + dmub_srv_sync_inbox1(dc->ctx->dmub_srv->dmub); } } dc_dmub_srv_notify_idle(dc, false); if (prev_driver_signals.bits.allow_ips1) { + DC_LOG_IPS( + "wait for IPS1 commit clear (ips1_commit=%d ips2_commit=%d)", + ips_fw->signals.bits.ips1_commit, + ips_fw->signals.bits.ips2_commit); + while (ips_fw->signals.bits.ips1_commit) udelay(1); + DC_LOG_IPS( + "wait for IPS1 commit clear done (ips1_commit=%d ips2_commit=%d)", + ips_fw->signals.bits.ips1_commit, + ips_fw->signals.bits.ips2_commit); } } if (!dc_dmub_srv_is_hw_pwr_up(dc->ctx->dmub_srv, true)) ASSERT(0); + + DC_LOG_IPS("%s exit (count rcg=%d ips1=%d ips2=%d)", + __func__, + rcg_exit_count, + ips1_exit_count, + ips2_exit_count); } void dc_dmub_srv_set_power_state(struct dc_dmub_srv *dc_dmub_srv, enum dc_acpi_cm_power_state powerState) @@ -1335,21 +1468,42 @@ void dc_dmub_srv_apply_idle_power_optimizations(const struct dc *dc, bool allow_ if (dc_dmub_srv->idle_allowed == allow_idle) return; + DC_LOG_IPS("%s state change: old=%d new=%d", __func__, dc_dmub_srv->idle_allowed, allow_idle); + /* * Entering a low power state requires a driver notification. * Powering up the hardware requires notifying PMFW and DMCUB. * Clearing the driver idle allow requires a DMCUB command. * DMCUB commands requires the DMCUB to be powered up and restored. - * - * Exit out early to prevent an infinite loop of DMCUB commands - * triggering exit low power - use software state to track this. */ - dc_dmub_srv->idle_allowed = allow_idle; - if (!allow_idle) + if (!allow_idle) { + dc_dmub_srv->idle_exit_counter += 1; + dc_dmub_srv_exit_low_power_state(dc); - else + /* + * Idle is considered fully exited only after the sequence above + * fully completes. If we have a race of two threads exiting + * at the same time then it's safe to perform the sequence + * twice as long as we're not re-entering. + * + * Infinite command submission is avoided by using the + * dm_execute_dmub_cmd submission instead of the "wake" helpers. + */ + dc_dmub_srv->idle_allowed = false; + + dc_dmub_srv->idle_exit_counter -= 1; + if (dc_dmub_srv->idle_exit_counter < 0) { + ASSERT(0); + dc_dmub_srv->idle_exit_counter = 0; + } + } else { + /* Consider idle as notified prior to the actual submission to + * prevent multiple entries. */ + dc_dmub_srv->idle_allowed = true; + dc_dmub_srv_notify_idle(dc, allow_idle); + } } bool dc_wake_and_execute_dmub_cmd(const struct dc_context *ctx, union dmub_rb_cmd *cmd, @@ -1384,7 +1538,8 @@ bool dc_wake_and_execute_dmub_cmd_list(const struct dc_context *ctx, unsigned in else result = dm_execute_dmub_cmd(ctx, cmd, wait_type); - if (result && reallow_idle && !ctx->dc->debug.disable_dmub_reallow_idle) + if (result && reallow_idle && dc_dmub_srv->idle_exit_counter == 0 && + !ctx->dc->debug.disable_dmub_reallow_idle) dc_dmub_srv_apply_idle_power_optimizations(ctx->dc, true); return result; @@ -1433,8 +1588,10 @@ bool dc_wake_and_execute_gpint(const struct dc_context *ctx, enum dmub_gpint_com result = dc_dmub_execute_gpint(ctx, command_code, param, response, wait_type); - if (result && reallow_idle && !ctx->dc->debug.disable_dmub_reallow_idle) + if (result && reallow_idle && dc_dmub_srv->idle_exit_counter == 0 && + !ctx->dc->debug.disable_dmub_reallow_idle) dc_dmub_srv_apply_idle_power_optimizations(ctx->dc, true); return result; } + 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 952bfb368886..2c5866211f60 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.h +++ b/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.h @@ -35,6 +35,7 @@ struct pipe_ctx; struct dc_crtc_timing_adjust; struct dc_crtc_timing; struct dc_state; +struct dc_surface_update; struct dc_reg_helper_state { bool gather_in_progress; @@ -51,7 +52,9 @@ struct dc_dmub_srv { struct dc_context *ctx; void *dm; + int32_t idle_exit_counter; bool idle_allowed; + bool needs_idle_wake; }; void dc_dmub_srv_wait_idle(struct dc_dmub_srv *dc_dmub_srv); diff --git a/drivers/gpu/drm/amd/display/dc/dc_dp_types.h b/drivers/gpu/drm/amd/display/dc/dc_dp_types.h index 1cb7765f593a..519c3df78ee5 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_dp_types.h +++ b/drivers/gpu/drm/amd/display/dc/dc_dp_types.h @@ -137,8 +137,13 @@ enum dp_link_encoding { enum dp_test_link_rate { DP_TEST_LINK_RATE_RBR = 0x06, + DP_TEST_LINK_RATE_RATE_2 = 0x08, // Rate_2 - 2.16 Gbps/Lane + DP_TEST_LINK_RATE_RATE_3 = 0x09, // Rate_3 - 2.43 Gbps/Lane DP_TEST_LINK_RATE_HBR = 0x0A, + DP_TEST_LINK_RATE_RBR2 = 0x0C, // Rate_5 (RBR2) - 3.24 Gbps/Lane + DP_TEST_LINK_RATE_RATE_6 = 0x10, // Rate_6 - 4.32 Gbps/Lane DP_TEST_LINK_RATE_HBR2 = 0x14, + DP_TEST_LINK_RATE_RATE_8 = 0x19, // Rate_8 - 6.75 Gbps/Lane DP_TEST_LINK_RATE_HBR3 = 0x1E, DP_TEST_LINK_RATE_UHBR10 = 0x01, DP_TEST_LINK_RATE_UHBR20 = 0x02, @@ -917,16 +922,6 @@ struct dpcd_usb4_dp_tunneling_info { uint8_t usb4_topology_id[DPCD_USB4_TOPOLOGY_ID_LEN]; }; -#ifndef DP_DFP_CAPABILITY_EXTENSION_SUPPORT -#define DP_DFP_CAPABILITY_EXTENSION_SUPPORT 0x0A3 -#endif -#ifndef DP_TEST_264BIT_CUSTOM_PATTERN_7_0 -#define DP_TEST_264BIT_CUSTOM_PATTERN_7_0 0X2230 -#endif -#ifndef DP_TEST_264BIT_CUSTOM_PATTERN_263_256 -#define DP_TEST_264BIT_CUSTOM_PATTERN_263_256 0X2250 -#endif - union dp_main_line_channel_coding_cap { struct { uint8_t DP_8b_10b_SUPPORTED :1; @@ -1232,8 +1227,7 @@ union replay_enable_and_configuration { unsigned char FREESYNC_PANEL_REPLAY_MODE :1; unsigned char TIMING_DESYNC_ERROR_VERIFICATION :1; unsigned char STATE_TRANSITION_ERROR_DETECTION :1; - unsigned char RESERVED0 :1; - unsigned char RESERVED1 :4; + unsigned char RESERVED :5; } bits; unsigned char raw; }; diff --git a/drivers/gpu/drm/amd/display/dc/dc_hw_types.h b/drivers/gpu/drm/amd/display/dc/dc_hw_types.h index aae2f3a2660d..2ad7f60805f5 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_hw_types.h +++ b/drivers/gpu/drm/amd/display/dc/dc_hw_types.h @@ -738,6 +738,13 @@ enum scanning_type { SCANNING_TYPE_UNDEFINED }; +enum chroma_cositing { + CHROMA_COSITING_NONE, + CHROMA_COSITING_LEFT, + CHROMA_COSITING_TOPLEFT, + CHROMA_COSITING_COUNT +}; + struct dc_crtc_timing_flags { uint32_t INTERLACE :1; uint32_t HSYNC_POSITIVE_POLARITY :1; /* when set to 1, @@ -974,6 +981,7 @@ struct dc_crtc_timing_adjust { uint32_t v_total_max; uint32_t v_total_mid; uint32_t v_total_mid_frame_num; + uint32_t allow_otg_v_count_halt; }; diff --git a/drivers/gpu/drm/amd/display/dc/dc_plane.h b/drivers/gpu/drm/amd/display/dc/dc_plane.h index ef380cae816a..44afcd989224 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_plane.h +++ b/drivers/gpu/drm/amd/display/dc/dc_plane.h @@ -29,7 +29,7 @@ #include "dc.h" #include "dc_hw_types.h" -struct dc_plane_state *dc_create_plane_state(struct dc *dc); +struct dc_plane_state *dc_create_plane_state(const struct dc *dc); const struct dc_plane_status *dc_plane_get_status( const struct dc_plane_state *plane_state); void dc_plane_state_retain(struct dc_plane_state *plane_state); diff --git a/drivers/gpu/drm/amd/display/dc/dc_plane_priv.h b/drivers/gpu/drm/amd/display/dc/dc_plane_priv.h index 9ee184c1df00..ab13335f1d01 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_plane_priv.h +++ b/drivers/gpu/drm/amd/display/dc/dc_plane_priv.h @@ -30,5 +30,6 @@ void dc_plane_construct(struct dc_context *ctx, struct dc_plane_state *plane_state); void dc_plane_destruct(struct dc_plane_state *plane_state); +uint8_t dc_plane_get_pipe_mask(struct dc_state *dc_state, const struct dc_plane_state *plane_state); #endif /* _DC_PLANE_PRIV_H_ */ diff --git a/drivers/gpu/drm/amd/display/dc/dc_state.h b/drivers/gpu/drm/amd/display/dc/dc_state.h index d167fdbfa8a9..caa45db50232 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_state.h +++ b/drivers/gpu/drm/amd/display/dc/dc_state.h @@ -29,7 +29,7 @@ #include "dc.h" #include "inc/core_status.h" -struct dc_state *dc_state_create(struct dc *dc); +struct dc_state *dc_state_create(struct dc *dc, struct dc_state_create_params *params); void dc_state_copy(struct dc_state *dst_state, struct dc_state *src_state); struct dc_state *dc_state_create_copy(struct dc_state *src_state); void dc_state_copy_current(struct dc *dc, struct dc_state *dst_state); @@ -39,12 +39,12 @@ void dc_state_destruct(struct dc_state *state); void dc_state_retain(struct dc_state *state); void dc_state_release(struct dc_state *state); -enum dc_status dc_state_add_stream(struct dc *dc, +enum dc_status dc_state_add_stream(const struct dc *dc, struct dc_state *state, struct dc_stream_state *stream); enum dc_status dc_state_remove_stream( - struct dc *dc, + const struct dc *dc, struct dc_state *state, struct dc_stream_state *stream); @@ -74,5 +74,5 @@ bool dc_state_add_all_planes_for_stream( struct dc_stream_status *dc_state_get_stream_status( struct dc_state *state, - struct dc_stream_state *stream); + const struct dc_stream_state *stream); #endif /* _DC_STATE_H_ */ diff --git a/drivers/gpu/drm/amd/display/dc/dc_state_priv.h b/drivers/gpu/drm/amd/display/dc/dc_state_priv.h index c1f44e09a6c1..615086d74d32 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_state_priv.h +++ b/drivers/gpu/drm/amd/display/dc/dc_state_priv.h @@ -29,6 +29,8 @@ #include "dc_state.h" #include "dc_stream.h" +struct dc_stream_state *dc_state_get_stream_from_id(const struct dc_state *state, unsigned int id); + /* Get the type of the provided resource (none, phantom, main) based on the provided * context. If the context is unavailable, determine only if phantom or not. */ @@ -45,7 +47,7 @@ struct dc_stream_state *dc_state_get_paired_subvp_stream(const struct dc_state * struct dc_stream_state *dc_state_create_phantom_stream(const struct dc *dc, struct dc_state *state, struct dc_stream_state *main_stream); -struct dc_plane_state *dc_state_create_phantom_plane(struct dc *dc, +struct dc_plane_state *dc_state_create_phantom_plane(const struct dc *dc, struct dc_state *state, struct dc_plane_state *main_plane); @@ -58,11 +60,11 @@ void dc_state_release_phantom_plane(const struct dc *dc, struct dc_plane_state *phantom_plane); /* add/remove phantom stream to context and generate subvp meta data */ -enum dc_status dc_state_add_phantom_stream(struct dc *dc, +enum dc_status dc_state_add_phantom_stream(const struct dc *dc, struct dc_state *state, struct dc_stream_state *phantom_stream, struct dc_stream_state *main_stream); -enum dc_status dc_state_remove_phantom_stream(struct dc *dc, +enum dc_status dc_state_remove_phantom_stream(const struct dc *dc, struct dc_state *state, struct dc_stream_state *phantom_stream); @@ -92,11 +94,11 @@ bool dc_state_add_all_phantom_planes_for_stream( struct dc_state *state); bool dc_state_remove_phantom_streams_and_planes( - struct dc *dc, + const struct dc *dc, struct dc_state *state); void dc_state_release_phantom_streams_and_planes( - struct dc *dc, + const struct dc *dc, struct dc_state *state); #endif /* _DC_STATE_PRIV_H_ */ diff --git a/drivers/gpu/drm/amd/display/dc/dc_stream.h b/drivers/gpu/drm/amd/display/dc/dc_stream.h index ee10941caa59..e5dbbc6089a5 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_stream.h +++ b/drivers/gpu/drm/amd/display/dc/dc_stream.h @@ -190,7 +190,7 @@ struct dc_stream_state { PHYSICAL_ADDRESS_LOC dmdata_address; bool use_dynamic_meta; - struct dc_transfer_func *out_transfer_func; + struct dc_transfer_func out_transfer_func; struct colorspace_transform gamut_remap_matrix; struct dc_csc_transform csc_color_matrix; @@ -428,14 +428,6 @@ bool dc_stream_set_dynamic_metadata(struct dc *dc, enum dc_status dc_validate_stream(struct dc *dc, struct dc_stream_state *stream); /* - * Set up streams and links associated to drive sinks - * The streams parameter is an absolute set of all active streams. - * - * After this call: - * Phy, Encoder, Timing Generator are programmed and enabled. - * New streams are enabled with blank stream; no memory read. - */ -/* * Enable stereo when commit_streams is not required, * for example, frame alternate. */ diff --git a/drivers/gpu/drm/amd/display/dc/dc_types.h b/drivers/gpu/drm/amd/display/dc/dc_types.h index be2ac5c442a4..0f66d00ef80f 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_types.h +++ b/drivers/gpu/drm/amd/display/dc/dc_types.h @@ -422,7 +422,7 @@ struct dc_dwb_params { enum dwb_capture_rate capture_rate; /* controls the frame capture rate */ struct scaling_taps scaler_taps; /* Scaling taps */ enum dwb_subsample_position subsample_position; - struct dc_transfer_func *out_transfer_func; + const struct dc_transfer_func *out_transfer_func; }; /* audio*/ @@ -1050,6 +1050,8 @@ union replay_error_status { struct replay_config { /* Replay feature is supported */ bool replay_supported; + /* Replay caps support DPCD & EDID caps*/ + bool replay_cap_support; /* Power opt flags that are supported */ unsigned int replay_power_opt_supported; /* SMU optimization is supported */ @@ -1175,4 +1177,20 @@ enum mall_stream_type { SUBVP_MAIN, // subvp in use, this stream is main stream SUBVP_PHANTOM, // subvp in use, this stream is a phantom stream }; + +enum dc_power_source_type { + DC_POWER_SOURCE_AC, // wall power + DC_POWER_SOURCE_DC, // battery power +}; + +struct dc_state_create_params { + enum dc_power_source_type power_source; +}; + +struct dc_commit_streams_params { + struct dc_stream_state **streams; + uint8_t stream_count; + enum dc_power_source_type power_source; +}; + #endif /* DC_TYPES_H_ */ diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c b/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c index 970644b695cd..b5e0289d2fe8 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c @@ -976,7 +976,10 @@ static bool dcn31_program_pix_clk( struct bp_pixel_clock_parameters bp_pc_params = {0}; enum transmitter_color_depth bp_pc_colour_depth = TRANSMITTER_COLOR_DEPTH_24; - if (clock_source->ctx->dc->clk_mgr->dp_dto_source_clock_in_khz != 0) + // Apply ssed(spread spectrum) dpref clock for edp only. + if (clock_source->ctx->dc->clk_mgr->dp_dto_source_clock_in_khz != 0 + && pix_clk_params->signal_type == SIGNAL_TYPE_EDP + && encoding == DP_8b_10b_ENCODING) dp_dto_ref_khz = clock_source->ctx->dc->clk_mgr->dp_dto_source_clock_in_khz; // For these signal types Driver to program DP_DTO without calling VBIOS Command table if (dc_is_dp_signal(pix_clk_params->signal_type) || dc_is_virtual_signal(pix_clk_params->signal_type)) { @@ -1093,9 +1096,6 @@ static bool get_pixel_clk_frequency_100hz( unsigned int modulo_hz = 0; unsigned int dp_dto_ref_khz = clock_source->ctx->dc->clk_mgr->dprefclk_khz; - if (clock_source->ctx->dc->clk_mgr->dp_dto_source_clock_in_khz != 0) - dp_dto_ref_khz = clock_source->ctx->dc->clk_mgr->dp_dto_source_clock_in_khz; - if (clock_source->id == CLOCK_SOURCE_ID_DP_DTO) { clock_hz = REG_READ(PHASE[inst]); diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_i2c_hw.c b/drivers/gpu/drm/amd/display/dc/dce/dce_i2c_hw.c index a2f48d46d199..ee601a6897a1 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_i2c_hw.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_i2c_hw.c @@ -22,9 +22,6 @@ * Authors: AMD * */ - -#include <linux/delay.h> - #include "resource.h" #include "dce_i2c.h" #include "dce_i2c_hw.h" @@ -315,9 +312,6 @@ static bool setup_engine( /* we have checked I2c not used by DMCU, set SW use I2C REQ to 1 to indicate SW using it*/ REG_UPDATE(DC_I2C_ARBITRATION, DC_I2C_SW_USE_I2C_REG_REQ, 1); - /* we have checked I2c not used by DMCU, set SW use I2C REQ to 1 to indicate SW using it*/ - REG_UPDATE(DC_I2C_ARBITRATION, DC_I2C_SW_USE_I2C_REG_REQ, 1); - /*set SW requested I2c speed to default, if API calls in it will be override later*/ set_speed(dce_i2c_hw, dce_i2c_hw->ctx->dc->caps.i2c_speed_in_khz); diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.h b/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.h index f98400efdd9b..e34e445a4013 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.h +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.h @@ -181,6 +181,7 @@ struct dce_mem_input_registers { SFB(blk, GRPH_ENABLE, GRPH_ENABLE, mask_sh),\ SFB(blk, GRPH_CONTROL, GRPH_DEPTH, mask_sh),\ SFB(blk, GRPH_CONTROL, GRPH_FORMAT, mask_sh),\ + SFB(blk, GRPH_CONTROL, GRPH_NUM_BANKS, mask_sh),\ SFB(blk, GRPH_X_START, GRPH_X_START, mask_sh),\ SFB(blk, GRPH_Y_START, GRPH_Y_START, mask_sh),\ SFB(blk, GRPH_X_END, GRPH_X_END, mask_sh),\ diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_opp.h b/drivers/gpu/drm/amd/display/dc/dce/dce_opp.h index bf1ffc3629c7..3d9be87aae45 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_opp.h +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_opp.h @@ -111,6 +111,7 @@ enum dce110_opp_reg_type { OPP_SF(FMT_DITHER_RAND_R_SEED, FMT_RAND_R_SEED, mask_sh),\ OPP_SF(FMT_DITHER_RAND_G_SEED, FMT_RAND_G_SEED, mask_sh),\ OPP_SF(FMT_DITHER_RAND_B_SEED, FMT_RAND_B_SEED, mask_sh),\ + OPP_SF(FMT_BIT_DEPTH_CONTROL, FMT_TEMPORAL_DITHER_EN, mask_sh),\ OPP_SF(FMT_BIT_DEPTH_CONTROL, FMT_TEMPORAL_DITHER_RESET, mask_sh),\ OPP_SF(FMT_BIT_DEPTH_CONTROL, FMT_TEMPORAL_DITHER_OFFSET, mask_sh),\ OPP_SF(FMT_BIT_DEPTH_CONTROL, FMT_TEMPORAL_DITHER_DEPTH, mask_sh),\ diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_transform.c b/drivers/gpu/drm/amd/display/dc/dce/dce_transform.c index 670d5ab9d998..2b1673d69ea8 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_transform.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_transform.c @@ -1408,7 +1408,7 @@ void dce110_opp_set_csc_default( static void program_pwl(struct dce_transform *xfm_dce, const struct pwl_params *params) { - int retval; + uint32_t retval; uint8_t max_tries = 10; uint8_t counter = 0; uint32_t i = 0; diff --git a/drivers/gpu/drm/amd/display/dc/dce/dmub_abm_lcd.c b/drivers/gpu/drm/amd/display/dc/dce/dmub_abm_lcd.c index f9d6a181164a..b851fc65f5b7 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dmub_abm_lcd.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dmub_abm_lcd.c @@ -34,11 +34,7 @@ #include "reg_helper.h" #include "fixed31_32.h" -#ifdef _WIN32 -#include "atombios.h" -#else #include "atom.h" -#endif #define TO_DMUB_ABM(abm)\ container_of(abm, struct dce_abm, base) diff --git a/drivers/gpu/drm/amd/display/dc/dce/dmub_replay.c b/drivers/gpu/drm/amd/display/dc/dce/dmub_replay.c index b010814706fe..4f559a025cf0 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dmub_replay.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dmub_replay.c @@ -244,7 +244,7 @@ static void dmub_replay_residency(struct dmub_replay *dmub, uint8_t panel_inst, uint16_t param = (uint16_t)(panel_inst << 8); if (is_alpm) - param |= REPLAY_RESIDENCY_MODE_ALPM; + param |= REPLAY_RESIDENCY_FIELD_MODE_ALPM; if (is_start) param |= REPLAY_RESIDENCY_ENABLE; diff --git a/drivers/gpu/drm/amd/display/dc/dce110/Makefile b/drivers/gpu/drm/amd/display/dc/dce110/Makefile index f0777d61c2cb..c307f040e48f 100644 --- a/drivers/gpu/drm/amd/display/dc/dce110/Makefile +++ b/drivers/gpu/drm/amd/display/dc/dce110/Makefile @@ -23,7 +23,7 @@ # Makefile for the 'controller' sub-component of DAL. # It provides the control and status of HW CRTC block. -CFLAGS_$(AMDDALPATH)/dc/dce110/dce110_resource.o = $(call cc-disable-warning, override-init) +CFLAGS_$(AMDDALPATH)/dc/dce110/dce110_resource.o = -Wno-override-init DCE110 = dce110_timing_generator.o \ dce110_compressor.o dce110_opp_regamma_v.o \ diff --git a/drivers/gpu/drm/amd/display/dc/dce112/Makefile b/drivers/gpu/drm/amd/display/dc/dce112/Makefile index 7e92effec894..683866797709 100644 --- a/drivers/gpu/drm/amd/display/dc/dce112/Makefile +++ b/drivers/gpu/drm/amd/display/dc/dce112/Makefile @@ -23,7 +23,7 @@ # Makefile for the 'controller' sub-component of DAL. # It provides the control and status of HW CRTC block. -CFLAGS_$(AMDDALPATH)/dc/dce112/dce112_resource.o = $(call cc-disable-warning, override-init) +CFLAGS_$(AMDDALPATH)/dc/dce112/dce112_resource.o = -Wno-override-init DCE112 = dce112_compressor.o diff --git a/drivers/gpu/drm/amd/display/dc/dce120/Makefile b/drivers/gpu/drm/amd/display/dc/dce120/Makefile index 1e3ef68a452a..8f508e662748 100644 --- a/drivers/gpu/drm/amd/display/dc/dce120/Makefile +++ b/drivers/gpu/drm/amd/display/dc/dce120/Makefile @@ -24,7 +24,7 @@ # It provides the control and status of HW CRTC block. -CFLAGS_$(AMDDALPATH)/dc/dce120/dce120_resource.o = $(call cc-disable-warning, override-init) +CFLAGS_$(AMDDALPATH)/dc/dce120/dce120_resource.o = -Wno-override-init DCE120 = dce120_timing_generator.o diff --git a/drivers/gpu/drm/amd/display/dc/dce60/Makefile b/drivers/gpu/drm/amd/display/dc/dce60/Makefile index fee331accc0e..eede83ad91fa 100644 --- a/drivers/gpu/drm/amd/display/dc/dce60/Makefile +++ b/drivers/gpu/drm/amd/display/dc/dce60/Makefile @@ -23,7 +23,7 @@ # Makefile for the 'controller' sub-component of DAL. # It provides the control and status of HW CRTC block. -CFLAGS_$(AMDDALPATH)/dc/dce60/dce60_resource.o = $(call cc-disable-warning, override-init) +CFLAGS_$(AMDDALPATH)/dc/dce60/dce60_resource.o = -Wno-override-init DCE60 = dce60_timing_generator.o dce60_hw_sequencer.o \ dce60_resource.o diff --git a/drivers/gpu/drm/amd/display/dc/dce80/Makefile b/drivers/gpu/drm/amd/display/dc/dce80/Makefile index 7eefffbdc925..fba189d26652 100644 --- a/drivers/gpu/drm/amd/display/dc/dce80/Makefile +++ b/drivers/gpu/drm/amd/display/dc/dce80/Makefile @@ -23,7 +23,7 @@ # Makefile for the 'controller' sub-component of DAL. # It provides the control and status of HW CRTC block. -CFLAGS_$(AMDDALPATH)/dc/dce80/dce80_resource.o = $(call cc-disable-warning, override-init) +CFLAGS_$(AMDDALPATH)/dc/dce80/dce80_resource.o = -Wno-override-init DCE80 = dce80_timing_generator.o diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/Makefile b/drivers/gpu/drm/amd/display/dc/dcn10/Makefile index ae6a131be71b..8dc7938c36d8 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/Makefile +++ b/drivers/gpu/drm/amd/display/dc/dcn10/Makefile @@ -24,9 +24,9 @@ DCN10 = dcn10_ipp.o \ dcn10_hw_sequencer_debug.o \ - dcn10_dpp.o dcn10_opp.o \ + dcn10_opp.o \ dcn10_hubp.o dcn10_mpc.o \ - dcn10_dpp_dscl.o dcn10_dpp_cm.o dcn10_cm_common.o \ + dcn10_cm_common.o \ dcn10_hubbub.o dcn10_stream_encoder.o dcn10_link_encoder.o AMD_DAL_DCN10 = $(addprefix $(AMDDALPATH)/dc/dcn10/,$(DCN10)) diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_cm_common.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_cm_common.c index b7e57aa27361..0b49362f71b0 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_cm_common.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_cm_common.c @@ -24,7 +24,7 @@ */ #include "dc.h" #include "reg_helper.h" -#include "dcn10_dpp.h" +#include "dcn10/dcn10_dpp.h" #include "dcn10_cm_common.h" #include "custom_float.h" @@ -402,6 +402,11 @@ bool cm_helper_translate_curve_to_hw_format(struct dc_context *ctx, i += increment) { if (j == hw_points - 1) break; + if (i >= TRANSFER_FUNC_POINTS) { + DC_LOG_ERROR("Index out of bounds: i=%d, TRANSFER_FUNC_POINTS=%d\n", + i, TRANSFER_FUNC_POINTS); + return false; + } rgb_resulted[j].red = output_tf->tf_pts.red[i]; rgb_resulted[j].green = output_tf->tf_pts.green[i]; rgb_resulted[j].blue = output_tf->tf_pts.blue[i]; diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.c index d51f1ce02874..6dd355a03033 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.c @@ -130,7 +130,7 @@ bool hubbub1_verify_allow_pstate_change_high( static unsigned int max_sampled_pstate_wait_us; /* data collection */ static bool forced_pstate_allow; /* help with revert wa */ - unsigned int debug_data; + unsigned int debug_data = 0; unsigned int i; if (forced_pstate_allow) { @@ -242,7 +242,7 @@ void hubbub1_wm_change_req_wa(struct hubbub *hubbub) bool hubbub1_program_urgent_watermarks( struct hubbub *hubbub, - struct dcn_watermark_set *watermarks, + union dcn_watermark_set *watermarks, unsigned int refclk_mhz, bool safe_to_lower) { @@ -356,7 +356,7 @@ bool hubbub1_program_urgent_watermarks( bool hubbub1_program_stutter_watermarks( struct hubbub *hubbub, - struct dcn_watermark_set *watermarks, + union dcn_watermark_set *watermarks, unsigned int refclk_mhz, bool safe_to_lower) { @@ -501,7 +501,7 @@ bool hubbub1_program_stutter_watermarks( bool hubbub1_program_pstate_watermarks( struct hubbub *hubbub, - struct dcn_watermark_set *watermarks, + union dcn_watermark_set *watermarks, unsigned int refclk_mhz, bool safe_to_lower) { @@ -582,7 +582,7 @@ bool hubbub1_program_pstate_watermarks( bool hubbub1_program_watermarks( struct hubbub *hubbub, - struct dcn_watermark_set *watermarks, + union dcn_watermark_set *watermarks, unsigned int refclk_mhz, bool safe_to_lower) { diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.h index 4201b7627030..d1f9e63944c8 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.h +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.h @@ -409,7 +409,7 @@ struct dcn10_hubbub { const struct dcn_hubbub_shift *shifts; const struct dcn_hubbub_mask *masks; unsigned int debug_test_index_pstate; - struct dcn_watermark_set watermarks; + union dcn_watermark_set watermarks; }; void hubbub1_update_dchub( @@ -423,7 +423,7 @@ void hubbub1_wm_change_req_wa(struct hubbub *hubbub); bool hubbub1_program_watermarks( struct hubbub *hubbub, - struct dcn_watermark_set *watermarks, + union dcn_watermark_set *watermarks, unsigned int refclk_mhz, bool safe_to_lower); @@ -446,17 +446,17 @@ void hubbub1_construct(struct hubbub *hubbub, bool hubbub1_program_urgent_watermarks( struct hubbub *hubbub, - struct dcn_watermark_set *watermarks, + union dcn_watermark_set *watermarks, unsigned int refclk_mhz, bool safe_to_lower); bool hubbub1_program_stutter_watermarks( struct hubbub *hubbub, - struct dcn_watermark_set *watermarks, + union dcn_watermark_set *watermarks, unsigned int refclk_mhz, bool safe_to_lower); bool hubbub1_program_pstate_watermarks( struct hubbub *hubbub, - struct dcn_watermark_set *watermarks, + union dcn_watermark_set *watermarks, unsigned int refclk_mhz, bool safe_to_lower); diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h index 09784222cc03..69119b2fdce2 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h @@ -692,6 +692,7 @@ struct dcn_hubp_state { uint32_t primary_meta_addr_hi; uint32_t uclk_pstate_force; uint32_t hubp_cntl; + uint32_t flip_control; }; struct dcn10_hubp { diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer_debug.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer_debug.c index 9033b39e0e0c..c51b717e5622 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer_debug.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer_debug.c @@ -392,7 +392,7 @@ static unsigned int dcn10_get_mpcc_states(struct dc *dc, char *pBuf, unsigned in remaining_buffer -= chars_printed; pBuf += chars_printed; - for (i = 0; i < pool->pipe_count; i++) { + for (i = 0; i < pool->mpcc_count; i++) { struct mpcc_state s = {0}; pool->mpc->funcs->read_mpcc_state(pool->mpc, i, &s); diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.c index 377f1ba1a81b..4d0eed7598b2 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.c @@ -1439,7 +1439,6 @@ enum signal_type dcn10_get_dig_mode( default: return SIGNAL_TYPE_NONE; } - return SIGNAL_TYPE_NONE; } void dcn10_link_encoder_get_max_link_cap(struct link_encoder *enc, diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.h index d980e6bd6c66..b7a89c39f445 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.h +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.h @@ -167,7 +167,6 @@ struct dcn10_link_enc_registers { uint32_t DIO_LINKD_CNTL; uint32_t DIO_LINKE_CNTL; uint32_t DIO_LINKF_CNTL; - uint32_t DIG_FIFO_CTRL0; uint32_t DIO_CLK_CNTL; uint32_t DIG_BE_CLK_CNTL; }; @@ -475,9 +474,6 @@ struct dcn10_link_enc_registers { type HPO_DP_ENC_SEL;\ type HPO_HDMI_ENC_SEL -#define DCN32_LINK_ENCODER_REG_FIELD_LIST(type) \ - type DIG_FIFO_OUTPUT_PIXEL_MODE - #define DCN35_LINK_ENCODER_REG_FIELD_LIST(type) \ type DIG_BE_ENABLE;\ type DIG_RB_SWITCH_EN;\ @@ -512,7 +508,6 @@ struct dcn10_link_enc_shift { DCN20_LINK_ENCODER_REG_FIELD_LIST(uint8_t); DCN30_LINK_ENCODER_REG_FIELD_LIST(uint8_t); DCN31_LINK_ENCODER_REG_FIELD_LIST(uint8_t); - DCN32_LINK_ENCODER_REG_FIELD_LIST(uint8_t); DCN35_LINK_ENCODER_REG_FIELD_LIST(uint8_t); }; @@ -521,7 +516,6 @@ struct dcn10_link_enc_mask { DCN20_LINK_ENCODER_REG_FIELD_LIST(uint32_t); DCN30_LINK_ENCODER_REG_FIELD_LIST(uint32_t); DCN31_LINK_ENCODER_REG_FIELD_LIST(uint32_t); - DCN32_LINK_ENCODER_REG_FIELD_LIST(uint32_t); DCN35_LINK_ENCODER_REG_FIELD_LIST(uint32_t); }; diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_opp.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_opp.c index 5838a11efd00..71e9288d60ed 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_opp.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_opp.c @@ -168,6 +168,10 @@ static void opp1_set_pixel_encoding( case PIXEL_ENCODING_RGB: case PIXEL_ENCODING_YCBCR444: + REG_UPDATE_3(FMT_CONTROL, + FMT_PIXEL_ENCODING, 0, + FMT_SUBSAMPLING_MODE, 0, + FMT_CBCR_BIT_REDUCTION_BYPASS, 0); REG_UPDATE(FMT_CONTROL, FMT_PIXEL_ENCODING, 0); break; case PIXEL_ENCODING_YCBCR422: @@ -177,7 +181,10 @@ static void opp1_set_pixel_encoding( FMT_CBCR_BIT_REDUCTION_BYPASS, 0); break; case PIXEL_ENCODING_YCBCR420: - REG_UPDATE(FMT_CONTROL, FMT_PIXEL_ENCODING, 2); + REG_UPDATE_3(FMT_CONTROL, + FMT_PIXEL_ENCODING, 2, + FMT_SUBSAMPLING_MODE, 2, + FMT_CBCR_BIT_REDUCTION_BYPASS, 1); break; default: break; diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_opp.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_opp.h index 2c0ecfa5a643..c87de68a509e 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_opp.h +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_opp.h @@ -79,6 +79,8 @@ OPP_SF(FMT0_FMT_CONTROL, FMT_SPATIAL_DITHER_FRAME_COUNTER_MAX, mask_sh), \ OPP_SF(FMT0_FMT_CONTROL, FMT_SPATIAL_DITHER_FRAME_COUNTER_BIT_SWAP, mask_sh), \ OPP_SF(FMT0_FMT_CONTROL, FMT_PIXEL_ENCODING, mask_sh), \ + OPP_SF(FMT0_FMT_CONTROL, FMT_SUBSAMPLING_MODE, mask_sh), \ + OPP_SF(FMT0_FMT_CONTROL, FMT_CBCR_BIT_REDUCTION_BYPASS, mask_sh), \ OPP_SF(FMT0_FMT_CONTROL, FMT_STEREOSYNC_OVERRIDE, mask_sh), \ OPP_SF(FMT0_FMT_DITHER_RAND_R_SEED, FMT_RAND_R_SEED, mask_sh), \ OPP_SF(FMT0_FMT_DITHER_RAND_G_SEED, FMT_RAND_G_SEED, mask_sh), \ diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.h index c429590f1298..1b96972b9d0f 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.h +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.h @@ -127,7 +127,6 @@ struct dcn10_stream_enc_registers { uint32_t AFMT_60958_1; uint32_t AFMT_60958_2; uint32_t DIG_FE_CNTL; - uint32_t DIG_FE_CNTL2; uint32_t DIG_FIFO_STATUS; uint32_t DP_MSE_RATE_CNTL; uint32_t DP_MSE_RATE_UPDATE; @@ -570,7 +569,7 @@ struct dcn10_stream_enc_registers { type DP_SEC_GSP11_ENABLE;\ type DP_SEC_GSP11_LINE_NUM -#define SE_REG_FIELD_LIST_DCN3_2(type) \ +#define SE_REG_FIELD_LIST_DCN3_1_COMMON(type) \ type DIG_FIFO_OUTPUT_PIXEL_MODE;\ type DP_PIXEL_PER_CYCLE_PROCESSING_MODE;\ type DIG_SYMCLK_FE_ON;\ @@ -599,7 +598,7 @@ struct dcn10_stream_encoder_shift { uint8_t HDMI_ACP_SEND; SE_REG_FIELD_LIST_DCN2_0(uint8_t); SE_REG_FIELD_LIST_DCN3_0(uint8_t); - SE_REG_FIELD_LIST_DCN3_2(uint8_t); + SE_REG_FIELD_LIST_DCN3_1_COMMON(uint8_t); SE_REG_FIELD_LIST_DCN3_5_COMMON(uint8_t); }; @@ -608,7 +607,7 @@ struct dcn10_stream_encoder_mask { uint32_t HDMI_ACP_SEND; SE_REG_FIELD_LIST_DCN2_0(uint32_t); SE_REG_FIELD_LIST_DCN3_0(uint32_t); - SE_REG_FIELD_LIST_DCN3_2(uint32_t); + SE_REG_FIELD_LIST_DCN3_1_COMMON(uint32_t); SE_REG_FIELD_LIST_DCN3_5_COMMON(uint32_t); }; @@ -667,9 +666,6 @@ void enc1_stream_encoder_send_immediate_sdp_message( void enc1_stream_encoder_stop_dp_info_packets( struct stream_encoder *enc); -void enc1_stream_encoder_reset_fifo( - struct stream_encoder *enc); - void enc1_stream_encoder_dp_blank( struct dc_link *link, struct stream_encoder *enc); diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/Makefile b/drivers/gpu/drm/amd/display/dc/dcn20/Makefile index 3dae3943b056..9b6070c99794 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn20/Makefile +++ b/drivers/gpu/drm/amd/display/dc/dcn20/Makefile @@ -2,7 +2,7 @@ # # Makefile for DCN. -DCN20 = dcn20_dpp.o dcn20_dpp_cm.o dcn20_hubp.o \ +DCN20 = dcn20_hubp.o \ dcn20_mpc.o dcn20_opp.o dcn20_hubbub.o dcn20_mmhubbub.o \ dcn20_stream_encoder.o dcn20_link_encoder.o dcn20_dccg.o \ dcn20_vmid.o dcn20_dwb.o dcn20_dwb_scl.o diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dwb.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dwb.c index f8667be57046..80779e85e2c5 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dwb.c +++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dwb.c @@ -299,6 +299,17 @@ void dwb2_set_scaler(struct dwbc *dwbc, struct dc_dwb_params *params) } } + + if (dwbc20->dwbc_mask->WBSCL_COEF_RAM_SEL) { + /* Swap double buffered coefficient set */ + uint32_t wbscl_mode = REG_READ(WBSCL_MODE); + bool coef_ram_current = get_reg_field_value_ex( + wbscl_mode, dwbc20->dwbc_mask->WBSCL_COEF_RAM_SEL_CURRENT, + dwbc20->dwbc_shift->WBSCL_COEF_RAM_SEL_CURRENT); + + REG_UPDATE(WBSCL_MODE, WBSCL_COEF_RAM_SEL, !coef_ram_current); + } + } static const struct dwbc_funcs dcn20_dwbc_funcs = { diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubbub.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubbub.c index 6eebcb22e317..c6f859871d11 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubbub.c +++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubbub.c @@ -570,7 +570,7 @@ void hubbub2_get_dchub_ref_freq(struct hubbub *hubbub, static bool hubbub2_program_watermarks( struct hubbub *hubbub, - struct dcn_watermark_set *watermarks, + union dcn_watermark_set *watermarks, unsigned int refclk_mhz, bool safe_to_lower) { diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubbub.h b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubbub.h index 2f6146bf1d32..24a9c45988ed 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubbub.h +++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubbub.h @@ -85,7 +85,7 @@ struct dcn20_hubbub { const struct dcn_hubbub_shift *shifts; const struct dcn_hubbub_mask *masks; unsigned int debug_test_index_pstate; - struct dcn_watermark_set watermarks; + union dcn_watermark_set watermarks; int num_vmid; struct dcn20_vmid vmid[16]; unsigned int detile_buf_size; diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubp.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubp.c index 89c3bf0fe0c9..6bba020ad6fb 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubp.c +++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubp.c @@ -1331,6 +1331,12 @@ void hubp2_read_state(struct hubp *hubp) SWATH_HEIGHT_C, &rq_regs->rq_regs_c.swath_height, PTE_ROW_HEIGHT_LINEAR_C, &rq_regs->rq_regs_c.pte_row_height_linear); + if (REG(DCHUBP_CNTL)) + s->hubp_cntl = REG_READ(DCHUBP_CNTL); + + if (REG(DCSURF_FLIP_CONTROL)) + s->flip_control = REG_READ(DCSURF_FLIP_CONTROL); + } static void hubp2_validate_dml_output(struct hubp *hubp, diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubp.h b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubp.h index efa2adf4f83d..8da3084d933f 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubp.h +++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubp.h @@ -147,7 +147,7 @@ uint32_t DCN_CUR1_TTU_CNTL1;\ uint32_t VMID_SETTINGS_0 - +/*shared with dcn3.x*/ #define DCN21_HUBP_REG_COMMON_VARIABLE_LIST \ DCN2_HUBP_REG_COMMON_VARIABLE_LIST; \ uint32_t FLIP_PARAMETERS_3;\ diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_link_encoder.h b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_link_encoder.h index b2b266953d18..c34e04cac9a0 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_link_encoder.h +++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_link_encoder.h @@ -147,7 +147,8 @@ LE_SF(DPCSTX0_DPCSTX_TX_CNTL, DPCS_TX_DATA_SWAP, mask_sh),\ LE_SF(DPCSTX0_DPCSTX_TX_CNTL, DPCS_TX_DATA_ORDER_INVERT, mask_sh),\ LE_SF(DPCSTX0_DPCSTX_TX_CNTL, DPCS_TX_FIFO_EN, mask_sh),\ - LE_SF(DPCSTX0_DPCSTX_TX_CNTL, DPCS_TX_FIFO_RD_START_DELAY, mask_sh) + LE_SF(DPCSTX0_DPCSTX_TX_CNTL, DPCS_TX_FIFO_RD_START_DELAY, mask_sh),\ + LE_SF(DPCSTX0_DPCSTX_DEBUG_CONFIG, DPCS_DBG_CBUS_DIS, mask_sh) #define DPCS_DCN2_MASK_SH_LIST(mask_sh)\ DPCS_MASK_SH_LIST(mask_sh),\ @@ -231,6 +232,8 @@ SRI(RDPCSTX_PHY_FUSE3, RDPCSTX, id), \ SRI(DPCSTX_TX_CLOCK_CNTL, DPCSTX, id), \ SRI(DPCSTX_TX_CNTL, DPCSTX, id), \ + SRI(DPCSTX_DEBUG_CONFIG, DPCSTX, id), \ + SRI(RDPCSTX_DEBUG_CONFIG, RDPCSTX, id), \ SR(RDPCSTX0_RDPCSTX_SCRATCH) diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_mpc.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_mpc.c index 16b5ff208d14..ea73473b970a 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_mpc.c +++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_mpc.c @@ -395,9 +395,12 @@ static void mpc20_program_ogam_pwl( MPCC_OGAM_LUT_DATA, rgb[i].delta_green_reg); REG_SET(MPCC_OGAM_LUT_DATA[mpcc_id], 0, MPCC_OGAM_LUT_DATA, rgb[i].delta_blue_reg); - } + REG_SEQ_SUBMIT(); + PERF_TRACE(); + REG_SEQ_WAIT_DONE(); + PERF_TRACE(); } static void apply_DEDCN20_305_wa(struct mpc *mpc, int mpcc_id, @@ -501,11 +504,6 @@ void mpc2_assert_mpcc_idle_before_connect(struct mpc *mpc, int mpcc_id) ASSERT(!mpc_disabled); ASSERT(!mpc_idle); } - - REG_SEQ_SUBMIT(); - PERF_TRACE(); - REG_SEQ_WAIT_DONE(); - PERF_TRACE(); } static void mpc2_init_mpcc(struct mpcc *mpcc, int mpcc_inst) diff --git a/drivers/gpu/drm/amd/display/dc/dcn201/Makefile b/drivers/gpu/drm/amd/display/dc/dcn201/Makefile index 2b0b4f32e13b..3880db59e457 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn201/Makefile +++ b/drivers/gpu/drm/amd/display/dc/dcn201/Makefile @@ -2,7 +2,7 @@ # # Makefile for DCN. DCN201 = dcn201_hubbub.o\ - dcn201_mpc.o dcn201_hubp.o dcn201_opp.o dcn201_dpp.o \ + dcn201_mpc.o dcn201_hubp.o dcn201_opp.o \ dcn201_dccg.o dcn201_link_encoder.o AMD_DAL_DCN201 = $(addprefix $(AMDDALPATH)/dc/dcn201/,$(DCN201)) diff --git a/drivers/gpu/drm/amd/display/dc/dcn201/dcn201_hubbub.c b/drivers/gpu/drm/amd/display/dc/dcn201/dcn201_hubbub.c index 037d265431c6..63798132ed95 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn201/dcn201_hubbub.c +++ b/drivers/gpu/drm/amd/display/dc/dcn201/dcn201_hubbub.c @@ -52,7 +52,7 @@ static bool hubbub201_program_watermarks( struct hubbub *hubbub, - struct dcn_watermark_set *watermarks, + union dcn_watermark_set *watermarks, unsigned int refclk_mhz, bool safe_to_lower) { @@ -103,5 +103,5 @@ void hubbub201_construct(struct dcn20_hubbub *hubbub, hubbub->masks = hubbub_mask; hubbub->debug_test_index_pstate = 0xB; - hubbub->detile_buf_size = 164 * 1024; + hubbub->detile_buf_size = 164 * 1024; /* 164KB for DCN2.0 */ } diff --git a/drivers/gpu/drm/amd/display/dc/dcn201/dcn201_hubp.c b/drivers/gpu/drm/amd/display/dc/dcn201/dcn201_hubp.c index 35dd4bac242a..cd2bfcc51276 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn201/dcn201_hubp.c +++ b/drivers/gpu/drm/amd/display/dc/dcn201/dcn201_hubp.c @@ -77,6 +77,7 @@ static void hubp201_program_requestor(struct hubp *hubp, MRQ_EXPANSION_MODE, rq_regs->mrq_expansion_mode, CRQ_EXPANSION_MODE, rq_regs->crq_expansion_mode); + /* no need to program PTE */ REG_SET_5(DCHUBP_REQ_SIZE_CONFIG, 0, CHUNK_SIZE, rq_regs->rq_regs_l.chunk_size, MIN_CHUNK_SIZE, rq_regs->rq_regs_l.min_chunk_size, @@ -99,6 +100,10 @@ static void hubp201_setup( struct _vcs_dpi_display_rq_regs_st *rq_regs, struct _vcs_dpi_display_pipe_dest_params_st *pipe_dest) { + /* + * otg is locked when this func is called. Register are double buffered. + * disable the requestors is not needed + */ hubp2_vready_at_or_After_vsync(hubp, pipe_dest); hubp201_program_requestor(hubp, rq_regs); hubp201_program_deadline(hubp, dlg_attr, ttu_attr); diff --git a/drivers/gpu/drm/amd/display/dc/dcn201/dcn201_link_encoder.h b/drivers/gpu/drm/amd/display/dc/dcn201/dcn201_link_encoder.h index 8b95ef251332..be25e8dc0636 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn201/dcn201_link_encoder.h +++ b/drivers/gpu/drm/amd/display/dc/dcn201/dcn201_link_encoder.h @@ -30,6 +30,10 @@ #define DPCS_DCN201_MASK_SH_LIST(mask_sh)\ DPCS_MASK_SH_LIST(mask_sh),\ + LE_SF(DPCSSYS_CR0_RAWLANE0_DIG_PCS_XF_RX_OVRD_IN_2, VCO_LD_VAL_OVRD, mask_sh),\ + LE_SF(DPCSSYS_CR0_RAWLANE0_DIG_PCS_XF_RX_OVRD_IN_2, VCO_LD_VAL_OVRD_EN, mask_sh),\ + LE_SF(DPCSSYS_CR0_RAWLANE0_DIG_PCS_XF_RX_OVRD_IN_3, REF_LD_VAL_OVRD, mask_sh),\ + LE_SF(DPCSSYS_CR0_RAWLANE0_DIG_PCS_XF_RX_OVRD_IN_3, REF_LD_VAL_OVRD_EN, mask_sh),\ LE_SF(RDPCSTX0_RDPCSTX_PHY_CNTL2, RDPCS_PHY_DPALT_DISABLE_ACK, mask_sh),\ LE_SF(RDPCSTX0_RDPCSTX_PHY_CNTL2, RDPCS_PHY_DPALT_DISABLE, mask_sh),\ LE_SF(RDPCSTX0_RDPCSTX_PHY_CNTL2, RDPCS_PHY_DPALT_DP4, mask_sh),\ @@ -44,7 +48,15 @@ LE_SF(RDPCSTX0_RDPCSTX_PHY_CNTL11, RDPCS_PHY_DP_REF_CLK_EN, mask_sh) #define DPCS_DCN201_REG_LIST(id) \ - DPCS_DCN2_CMN_REG_LIST(id) + DPCS_DCN2_CMN_REG_LIST(id), \ + SRI_IX(RAWLANE0_DIG_PCS_XF_RX_OVRD_IN_2, DPCSSYS_CR, id), \ + SRI_IX(RAWLANE0_DIG_PCS_XF_RX_OVRD_IN_3, DPCSSYS_CR, id), \ + SRI_IX(RAWLANE1_DIG_PCS_XF_RX_OVRD_IN_2, DPCSSYS_CR, id), \ + SRI_IX(RAWLANE1_DIG_PCS_XF_RX_OVRD_IN_3, DPCSSYS_CR, id), \ + SRI_IX(RAWLANE2_DIG_PCS_XF_RX_OVRD_IN_2, DPCSSYS_CR, id), \ + SRI_IX(RAWLANE2_DIG_PCS_XF_RX_OVRD_IN_3, DPCSSYS_CR, id), \ + SRI_IX(RAWLANE3_DIG_PCS_XF_RX_OVRD_IN_2, DPCSSYS_CR, id), \ + SRI_IX(RAWLANE3_DIG_PCS_XF_RX_OVRD_IN_3, DPCSSYS_CR, id) void dcn201_link_encoder_construct( struct dcn20_link_encoder *enc20, diff --git a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hubbub.c b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hubbub.c index aeb0e0d9b70a..2546224b326a 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hubbub.c +++ b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hubbub.c @@ -140,7 +140,7 @@ int hubbub21_init_dchub(struct hubbub *hubbub, bool hubbub21_program_urgent_watermarks( struct hubbub *hubbub, - struct dcn_watermark_set *watermarks, + union dcn_watermark_set *watermarks, unsigned int refclk_mhz, bool safe_to_lower) { @@ -334,7 +334,7 @@ bool hubbub21_program_urgent_watermarks( bool hubbub21_program_stutter_watermarks( struct hubbub *hubbub, - struct dcn_watermark_set *watermarks, + union dcn_watermark_set *watermarks, unsigned int refclk_mhz, bool safe_to_lower) { @@ -487,7 +487,7 @@ bool hubbub21_program_stutter_watermarks( bool hubbub21_program_pstate_watermarks( struct hubbub *hubbub, - struct dcn_watermark_set *watermarks, + union dcn_watermark_set *watermarks, unsigned int refclk_mhz, bool safe_to_lower) { @@ -573,7 +573,7 @@ bool hubbub21_program_pstate_watermarks( bool hubbub21_program_watermarks( struct hubbub *hubbub, - struct dcn_watermark_set *watermarks, + union dcn_watermark_set *watermarks, unsigned int refclk_mhz, bool safe_to_lower) { diff --git a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hubbub.h b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hubbub.h index d8eb2bb7282c..ab2ce0313529 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hubbub.h +++ b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hubbub.h @@ -127,22 +127,22 @@ int hubbub21_init_dchub(struct hubbub *hubbub, struct dcn_hubbub_phys_addr_config *pa_config); bool hubbub21_program_watermarks( struct hubbub *hubbub, - struct dcn_watermark_set *watermarks, + union dcn_watermark_set *watermarks, unsigned int refclk_mhz, bool safe_to_lower); bool hubbub21_program_urgent_watermarks( struct hubbub *hubbub, - struct dcn_watermark_set *watermarks, + union dcn_watermark_set *watermarks, unsigned int refclk_mhz, bool safe_to_lower); bool hubbub21_program_stutter_watermarks( struct hubbub *hubbub, - struct dcn_watermark_set *watermarks, + union dcn_watermark_set *watermarks, unsigned int refclk_mhz, bool safe_to_lower); bool hubbub21_program_pstate_watermarks( struct hubbub *hubbub, - struct dcn_watermark_set *watermarks, + union dcn_watermark_set *watermarks, unsigned int refclk_mhz, bool safe_to_lower); diff --git a/drivers/gpu/drm/amd/display/dc/dcn30/Makefile b/drivers/gpu/drm/amd/display/dc/dcn30/Makefile index b5b2aa3b3783..c6ca70f3c061 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn30/Makefile +++ b/drivers/gpu/drm/amd/display/dc/dcn30/Makefile @@ -25,13 +25,11 @@ DCN30 := dcn30_hubbub.o \ dcn30_hubp.o \ - dcn30_dpp.o \ dcn30_dccg.o \ dcn30_mpc.o dcn30_vpg.o \ dcn30_afmt.o \ dcn30_dio_stream_encoder.o \ dcn30_dwb.o \ - dcn30_dpp_cm.o \ dcn30_dwb_cm.o \ dcn30_cm_common.o \ dcn30_mmhubbub.o \ diff --git a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_cm_common.c b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_cm_common.c index ddb344056d40..b8327237ed44 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_cm_common.c +++ b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_cm_common.c @@ -26,7 +26,7 @@ #include "dm_services.h" #include "core_types.h" #include "reg_helper.h" -#include "dcn30_dpp.h" +#include "dcn30/dcn30_dpp.h" #include "basics/conversion.h" #include "dcn30_cm_common.h" #include "custom_float.h" diff --git a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dccg.h b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dccg.h index 35a613bb08bf..3f1da7f3a91c 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dccg.h +++ b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dccg.h @@ -29,15 +29,9 @@ #include "dcn20/dcn20_dccg.h" -#define DCCG_REG_LIST_DCN3AG() \ - DCCG_COMMON_REG_LIST_DCN_BASE(),\ - SR(PHYASYMCLK_CLOCK_CNTL),\ - SR(PHYBSYMCLK_CLOCK_CNTL),\ - SR(PHYCSYMCLK_CLOCK_CNTL) - - #define DCCG_REG_LIST_DCN30() \ DCCG_REG_LIST_DCN2(),\ + DCCG_SRII(CLOCK_CNTL, HDMICHARCLK, 0),\ DCCG_SRII(PIXEL_RATE_CNTL, OTG, 2),\ DCCG_SRII(PIXEL_RATE_CNTL, OTG, 3),\ DCCG_SRII(PIXEL_RATE_CNTL, OTG, 4),\ @@ -46,19 +40,10 @@ SR(PHYBSYMCLK_CLOCK_CNTL),\ SR(PHYCSYMCLK_CLOCK_CNTL) -#define DCCG_MASK_SH_LIST_DCN3AG(mask_sh) \ - DCCG_MASK_SH_LIST_DCN2_1(mask_sh),\ - DCCG_SF(HDMICHARCLK0_CLOCK_CNTL, HDMICHARCLK0_EN, mask_sh),\ - DCCG_SF(HDMICHARCLK0_CLOCK_CNTL, HDMICHARCLK0_SRC_SEL, mask_sh),\ - DCCG_SF(PHYASYMCLK_CLOCK_CNTL, PHYASYMCLK_FORCE_EN, mask_sh),\ - DCCG_SF(PHYASYMCLK_CLOCK_CNTL, PHYASYMCLK_FORCE_SRC_SEL, mask_sh),\ - DCCG_SF(PHYBSYMCLK_CLOCK_CNTL, PHYBSYMCLK_FORCE_EN, mask_sh),\ - DCCG_SF(PHYBSYMCLK_CLOCK_CNTL, PHYBSYMCLK_FORCE_SRC_SEL, mask_sh),\ - DCCG_SF(PHYCSYMCLK_CLOCK_CNTL, PHYCSYMCLK_FORCE_EN, mask_sh),\ - DCCG_SF(PHYCSYMCLK_CLOCK_CNTL, PHYCSYMCLK_FORCE_SRC_SEL, mask_sh) - #define DCCG_MASK_SH_LIST_DCN3(mask_sh) \ DCCG_MASK_SH_LIST_DCN2(mask_sh),\ + DCCG_SF(HDMICHARCLK0_CLOCK_CNTL, HDMICHARCLK0_EN, mask_sh),\ + DCCG_SF(HDMICHARCLK0_CLOCK_CNTL, HDMICHARCLK0_SRC_SEL, mask_sh),\ DCCG_SF(PHYASYMCLK_CLOCK_CNTL, PHYASYMCLK_FORCE_EN, mask_sh),\ DCCG_SF(PHYASYMCLK_CLOCK_CNTL, PHYASYMCLK_FORCE_SRC_SEL, mask_sh),\ DCCG_SF(PHYBSYMCLK_CLOCK_CNTL, PHYBSYMCLK_FORCE_EN, mask_sh),\ diff --git a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dio_link_encoder.c b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dio_link_encoder.c index 1fb8fd7afc95..b8e31b5ea114 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dio_link_encoder.c +++ b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dio_link_encoder.c @@ -30,8 +30,6 @@ #include "dcn30_dio_link_encoder.h" #include "stream_encoder.h" #include "dc_bios_types.h" -/* #include "dcn3ag/dcn3ag_phy_fw.h" */ - #include "gpio_service_interface.h" #define CTX \ diff --git a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dio_link_encoder.h b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dio_link_encoder.h index f2d90f2b8bf1..5b6177c2ae98 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dio_link_encoder.h +++ b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dio_link_encoder.h @@ -55,7 +55,8 @@ SRI(DP_DPHY_HBR2_PATTERN_CONTROL, DP, id) #define LINK_ENCODER_MASK_SH_LIST_DCN30(mask_sh) \ - LINK_ENCODER_MASK_SH_LIST_DCN20(mask_sh) + LINK_ENCODER_MASK_SH_LIST_DCN20(mask_sh),\ + LE_SF(DIG0_TMDS_DCBALANCER_CONTROL, TMDS_SYNC_DCBAL_EN, mask_sh) #define DPCS_DCN3_MASK_SH_LIST(mask_sh)\ DPCS_DCN2_MASK_SH_LIST(mask_sh),\ diff --git a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dio_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dio_stream_encoder.c index 005dbe099a7a..425b830b88d2 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dio_stream_encoder.c +++ b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dio_stream_encoder.c @@ -29,9 +29,6 @@ #include "reg_helper.h" #include "hw_shared.h" #include "dc.h" -#include "core_types.h" -#include <linux/delay.h> - #define DC_LOGGER \ enc1->base.ctx->logger diff --git a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dwb.c b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dwb.c index 1b9d9495f76d..fae98cf52020 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dwb.c +++ b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dwb.c @@ -251,9 +251,7 @@ static const struct dwbc_funcs dcn30_dwbc_funcs = { .set_fc_enable = dwb3_set_fc_enable, .set_stereo = dwb3_set_stereo, .set_new_content = dwb3_set_new_content, - .dwb_program_output_csc = NULL, .dwb_ogam_set_input_transfer_func = dwb3_ogam_set_input_transfer_func, //TODO: rename - .dwb_set_scaler = NULL, }; void dcn30_dwbc_construct(struct dcn30_dwbc *dwbc30, diff --git a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dwb.h b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dwb.h index 332634b76aac..0f3f7c5fbaec 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dwb.h +++ b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dwb.h @@ -217,6 +217,7 @@ SF_DWB2(DWB_OGAM_LUT_DATA, DWBCP, 0, DWB_OGAM_LUT_DATA, mask_sh),\ SF_DWB2(DWB_OGAM_LUT_CONTROL, DWBCP, 0, DWB_OGAM_LUT_WRITE_COLOR_MASK, mask_sh),\ SF_DWB2(DWB_OGAM_LUT_CONTROL, DWBCP, 0, DWB_OGAM_LUT_READ_COLOR_SEL, mask_sh),\ + SF_DWB2(DWB_OGAM_LUT_CONTROL, DWBCP, 0, DWB_OGAM_LUT_READ_DBG, mask_sh),\ SF_DWB2(DWB_OGAM_LUT_CONTROL, DWBCP, 0, DWB_OGAM_LUT_HOST_SEL, mask_sh),\ SF_DWB2(DWB_OGAM_LUT_CONTROL, DWBCP, 0, DWB_OGAM_LUT_CONFIG_MODE, mask_sh),\ SF_DWB2(DWB_OGAM_RAMA_START_CNTL_B, DWBCP, 0, DWB_OGAM_RAMA_EXP_REGION_START_B, mask_sh),\ @@ -524,6 +525,7 @@ type DWB_OGAM_LUT_DATA;\ type DWB_OGAM_LUT_WRITE_COLOR_MASK;\ type DWB_OGAM_LUT_READ_COLOR_SEL;\ + type DWB_OGAM_LUT_READ_DBG;\ type DWB_OGAM_LUT_HOST_SEL;\ type DWB_OGAM_LUT_CONFIG_MODE;\ type DWB_OGAM_LUT_STATUS;\ @@ -710,7 +712,7 @@ type DWB_OGAM_RAMB_EXP_REGION32_LUT_OFFSET;\ type DWB_OGAM_RAMB_EXP_REGION32_NUM_SEGMENTS;\ type DWB_OGAM_RAMB_EXP_REGION33_LUT_OFFSET;\ - type DWB_OGAM_RAMB_EXP_REGION33_NUM_SEGMENTS; + type DWB_OGAM_RAMB_EXP_REGION33_NUM_SEGMENTS struct dcn30_dwbc_registers { /* DCN3AG */ @@ -733,6 +735,10 @@ struct dcn30_dwbc_registers { uint32_t DWB_MMHUBBUB_BACKPRESSURE_CNT; uint32_t DWB_HOST_READ_CONTROL; uint32_t DWB_SOFT_RESET; + uint32_t DWB_DEBUG_CTRL; + uint32_t DWB_DEBUG; + uint32_t DWB_TEST_DEBUG_INDEX; + uint32_t DWB_TEST_DEBUG_DATA; /* DWBSCL */ uint32_t DWBSCL_COEF_RAM_TAP_SELECT; @@ -747,6 +753,9 @@ struct dcn30_dwbc_registers { uint32_t DWBSCL_DEST_SIZE; uint32_t DWBSCL_OVERFLOW_STATUS; uint32_t DWBSCL_OVERFLOW_COUNTER; + uint32_t DWBSCL_DEBUG; + uint32_t DWBSCL_TEST_DEBUG_INDEX; + uint32_t DWBSCL_TEST_DEBUG_DATA; /* DWBCP */ uint32_t DWB_HDR_MULT_COEF; @@ -838,6 +847,9 @@ struct dcn30_dwbc_registers { uint32_t DWB_OGAM_RAMB_REGION_28_29; uint32_t DWB_OGAM_RAMB_REGION_30_31; uint32_t DWB_OGAM_RAMB_REGION_32_33; + uint32_t DWBCP_DEBUG; + uint32_t DWBCP_TEST_DEBUG_INDEX; + uint32_t DWBCP_TEST_DEBUG_DATA; }; /* Internal enums / structs */ diff --git a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hubbub.c b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hubbub.c index 152c9c5733f1..6a5af3da4b45 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hubbub.c +++ b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hubbub.c @@ -95,7 +95,7 @@ int hubbub3_init_dchub_sys_ctx(struct hubbub *hubbub, bool hubbub3_program_watermarks( struct hubbub *hubbub, - struct dcn_watermark_set *watermarks, + union dcn_watermark_set *watermarks, unsigned int refclk_mhz, bool safe_to_lower) { diff --git a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hubbub.h b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hubbub.h index 7b597908b937..ca6233e8f1f4 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hubbub.h +++ b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hubbub.h @@ -124,7 +124,7 @@ bool hubbub3_get_dcc_compression_cap(struct hubbub *hubbub, bool hubbub3_program_watermarks( struct hubbub *hubbub, - struct dcn_watermark_set *watermarks, + union dcn_watermark_set *watermarks, unsigned int refclk_mhz, bool safe_to_lower); diff --git a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hubp.c b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hubp.c index 75547ce86c09..60a64d290352 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hubp.c +++ b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hubp.c @@ -455,6 +455,9 @@ void hubp3_read_state(struct hubp *hubp) if (REG(DCHUBP_CNTL)) s->hubp_cntl = REG_READ(DCHUBP_CNTL); + if (REG(DCSURF_FLIP_CONTROL)) + s->flip_control = REG_READ(DCSURF_FLIP_CONTROL); + } void hubp3_setup( diff --git a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_mpc.c b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_mpc.c index bf3386cd444d..fca94e50ae93 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_mpc.c +++ b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_mpc.c @@ -44,6 +44,36 @@ #define NUM_ELEMENTS(a) (sizeof(a) / sizeof((a)[0])) +void mpc3_mpc_init(struct mpc *mpc) +{ + struct dcn30_mpc *mpc30 = TO_DCN30_MPC(mpc); + int opp_id; + + mpc1_mpc_init(mpc); + + for (opp_id = 0; opp_id < MAX_OPP; opp_id++) { + if (REG(MUX[opp_id])) + /* disable mpc out rate and flow control */ + REG_UPDATE_2(MUX[opp_id], MPC_OUT_RATE_CONTROL_DISABLE, + 1, MPC_OUT_FLOW_CONTROL_COUNT, 0); + } +} + +void mpc3_mpc_init_single_inst(struct mpc *mpc, unsigned int mpcc_id) +{ + struct dcn30_mpc *mpc30 = TO_DCN30_MPC(mpc); + + mpc1_mpc_init_single_inst(mpc, mpcc_id); + + /* assuming mpc out mux is connected to opp with the same index at this + * point in time (e.g. transitioning from vbios to driver) + */ + if (mpcc_id < MAX_OPP && REG(MUX[mpcc_id])) + /* disable mpc out rate and flow control */ + REG_UPDATE_2(MUX[mpcc_id], MPC_OUT_RATE_CONTROL_DISABLE, + 1, MPC_OUT_FLOW_CONTROL_COUNT, 0); +} + bool mpc3_is_dwb_idle( struct mpc *mpc, int dwb_id) @@ -80,25 +110,6 @@ void mpc3_disable_dwb_mux( MPC_DWB0_MUX, 0xf); } -void mpc3_set_out_rate_control( - struct mpc *mpc, - int opp_id, - bool enable, - bool rate_2x_mode, - struct mpc_dwb_flow_control *flow_control) -{ - struct dcn30_mpc *mpc30 = TO_DCN30_MPC(mpc); - - REG_UPDATE_2(MUX[opp_id], - MPC_OUT_RATE_CONTROL_DISABLE, !enable, - MPC_OUT_RATE_CONTROL, rate_2x_mode); - - if (flow_control) - REG_UPDATE_2(MUX[opp_id], - MPC_OUT_FLOW_CONTROL_MODE, flow_control->flow_ctrl_mode, - MPC_OUT_FLOW_CONTROL_COUNT, flow_control->flow_ctrl_cnt1); -} - enum dc_lut_mode mpc3_get_ogam_current(struct mpc *mpc, int mpcc_id) { /*Contrary to DCN2 and DCN1 wherein a single status register field holds this info; @@ -1172,7 +1183,7 @@ void mpc3_get_gamut_remap(struct mpc *mpc, struct mpc_grph_gamut_adjustment *adjust) { struct dcn30_mpc *mpc30 = TO_DCN30_MPC(mpc); - uint16_t arr_reg_val[12]; + uint16_t arr_reg_val[12] = {0}; int select; read_gamut_remap(mpc30, mpcc_id, arr_reg_val, &select); @@ -1490,8 +1501,8 @@ static const struct mpc_funcs dcn30_mpc_funcs = { .read_mpcc_state = mpc3_read_mpcc_state, .insert_plane = mpc1_insert_plane, .remove_mpcc = mpc1_remove_mpcc, - .mpc_init = mpc1_mpc_init, - .mpc_init_single_inst = mpc1_mpc_init_single_inst, + .mpc_init = mpc3_mpc_init, + .mpc_init_single_inst = mpc3_mpc_init_single_inst, .update_blending = mpc2_update_blending, .cursor_lock = mpc1_cursor_lock, .get_mpcc_for_dpp = mpc1_get_mpcc_for_dpp, @@ -1508,7 +1519,6 @@ static const struct mpc_funcs dcn30_mpc_funcs = { .set_dwb_mux = mpc3_set_dwb_mux, .disable_dwb_mux = mpc3_disable_dwb_mux, .is_dwb_idle = mpc3_is_dwb_idle, - .set_out_rate_control = mpc3_set_out_rate_control, .set_gamut_remap = mpc3_set_gamut_remap, .program_shaper = mpc3_program_shaper, .acquire_rmu = mpcc3_acquire_rmu, diff --git a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_mpc.h b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_mpc.h index 9cb96ae95a2f..ce93003dae01 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_mpc.h +++ b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_mpc.h @@ -1007,6 +1007,13 @@ void dcn30_mpc_construct(struct dcn30_mpc *mpc30, int num_mpcc, int num_rmu); +void mpc3_mpc_init( + struct mpc *mpc); + +void mpc3_mpc_init_single_inst( + struct mpc *mpc, + unsigned int mpcc_id); + bool mpc3_program_shaper( struct mpc *mpc, const struct pwl_params *params, @@ -1078,13 +1085,6 @@ bool mpc3_is_dwb_idle( struct mpc *mpc, int dwb_id); -void mpc3_set_out_rate_control( - struct mpc *mpc, - int opp_id, - bool enable, - bool rate_2x_mode, - struct mpc_dwb_flow_control *flow_control); - void mpc3_power_on_ogam_lut( struct mpc *mpc, int mpcc_id, bool power_on); diff --git a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_vpg.h b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_vpg.h index ed9a5549c389..466ba20b9c61 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_vpg.h +++ b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_vpg.h @@ -26,6 +26,7 @@ #ifndef __DAL_DCN30_VPG_H__ #define __DAL_DCN30_VPG_H__ +#include "vpg.h" #define DCN30_VPG_FROM_VPG(vpg)\ container_of(vpg, struct dcn30_vpg, base) @@ -132,28 +133,6 @@ struct dcn30_vpg_mask { VPG_DCN3_REG_FIELD_LIST(uint32_t); }; -struct vpg; - -struct vpg_funcs { - void (*update_generic_info_packet)( - struct vpg *vpg, - uint32_t packet_index, - const struct dc_info_packet *info_packet, - bool immediate_update); - - void (*vpg_poweron)( - struct vpg *vpg); - - void (*vpg_powerdown)( - struct vpg *vpg); -}; - -struct vpg { - const struct vpg_funcs *funcs; - struct dc_context *ctx; - int inst; -}; - struct dcn30_vpg { struct vpg base; const struct dcn30_vpg_registers *regs; diff --git a/drivers/gpu/drm/amd/display/dc/dcn301/dcn301_dccg.h b/drivers/gpu/drm/amd/display/dc/dcn301/dcn301_dccg.h index 73db962dbc03..067e49cb238e 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn301/dcn301_dccg.h +++ b/drivers/gpu/drm/amd/display/dc/dcn301/dcn301_dccg.h @@ -56,10 +56,4 @@ struct dccg *dccg301_create( const struct dccg_shift *dccg_shift, const struct dccg_mask *dccg_mask); -struct dccg *dccg301_create( - struct dc_context *ctx, - const struct dccg_registers *regs, - const struct dccg_shift *dccg_shift, - const struct dccg_mask *dccg_mask); - #endif //__DCN301_DCCG_H__ diff --git a/drivers/gpu/drm/amd/display/dc/dcn301/dcn301_hubbub.c b/drivers/gpu/drm/amd/display/dc/dcn301/dcn301_hubbub.c index a046664e2031..c1959672df50 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn301/dcn301_hubbub.c +++ b/drivers/gpu/drm/amd/display/dc/dcn301/dcn301_hubbub.c @@ -63,6 +63,7 @@ static const struct hubbub_funcs hubbub301_funcs = { .verify_allow_pstate_change_high = hubbub1_verify_allow_pstate_change_high, .force_wm_propagate_to_pipes = hubbub3_force_wm_propagate_to_pipes, .force_pstate_change_control = hubbub3_force_pstate_change_control, + .init_watermarks = hubbub3_init_watermarks, .hubbub_read_state = hubbub2_read_state, }; diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_dccg.h b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_dccg.h index e3caaacf7493..e3be0bab4007 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_dccg.h +++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_dccg.h @@ -34,12 +34,14 @@ DCCG_SRII(DTO_PARAM, DPPCLK, 1),\ DCCG_SRII(DTO_PARAM, DPPCLK, 2),\ DCCG_SRII(DTO_PARAM, DPPCLK, 3),\ + DCCG_SRII(CLOCK_CNTL, HDMICHARCLK, 0),\ SR(PHYASYMCLK_CLOCK_CNTL),\ SR(PHYBSYMCLK_CLOCK_CNTL),\ SR(PHYCSYMCLK_CLOCK_CNTL),\ SR(PHYDSYMCLK_CLOCK_CNTL),\ SR(PHYESYMCLK_CLOCK_CNTL),\ SR(DPSTREAMCLK_CNTL),\ + SR(HDMISTREAMCLK_CNTL),\ SR(SYMCLK32_SE_CNTL),\ SR(SYMCLK32_LE_CNTL),\ DCCG_SRII(PIXEL_RATE_CNTL, OTG, 0),\ @@ -78,6 +80,8 @@ DCCG_SFI(DPPCLK_DTO_CTRL, DTO_DB_EN, DPPCLK, 3, mask_sh),\ DCCG_SF(DPPCLK0_DTO_PARAM, DPPCLK0_DTO_PHASE, mask_sh),\ DCCG_SF(DPPCLK0_DTO_PARAM, DPPCLK0_DTO_MODULO, mask_sh),\ + DCCG_SF(HDMICHARCLK0_CLOCK_CNTL, HDMICHARCLK0_EN, mask_sh),\ + DCCG_SF(HDMICHARCLK0_CLOCK_CNTL, HDMICHARCLK0_SRC_SEL, mask_sh),\ DCCG_SF(PHYASYMCLK_CLOCK_CNTL, PHYASYMCLK_FORCE_EN, mask_sh),\ DCCG_SF(PHYASYMCLK_CLOCK_CNTL, PHYASYMCLK_FORCE_SRC_SEL, mask_sh),\ DCCG_SF(PHYBSYMCLK_CLOCK_CNTL, PHYBSYMCLK_FORCE_EN, mask_sh),\ @@ -92,6 +96,8 @@ DCCG_SF(DPSTREAMCLK_CNTL, DPSTREAMCLK_PIPE1_EN, mask_sh),\ DCCG_SF(DPSTREAMCLK_CNTL, DPSTREAMCLK_PIPE2_EN, mask_sh),\ DCCG_SF(DPSTREAMCLK_CNTL, DPSTREAMCLK_PIPE3_EN, mask_sh),\ + DCCG_SF(HDMISTREAMCLK_CNTL, HDMISTREAMCLK0_SRC_SEL, mask_sh),\ + DCCG_SF(HDMISTREAMCLK_CNTL, HDMISTREAMCLK0_DTO_FORCE_DIS, mask_sh),\ DCCG_SF(SYMCLK32_SE_CNTL, SYMCLK32_SE0_SRC_SEL, mask_sh),\ DCCG_SF(SYMCLK32_SE_CNTL, SYMCLK32_SE1_SRC_SEL, mask_sh),\ DCCG_SF(SYMCLK32_SE_CNTL, SYMCLK32_SE2_SRC_SEL, mask_sh),\ diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_dio_link_encoder.c b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_dio_link_encoder.c index 26be5fee7411..b2cea59ba5d4 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_dio_link_encoder.c +++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_dio_link_encoder.c @@ -205,7 +205,7 @@ void dcn31_link_encoder_set_dio_phy_mux( } } -static void enc31_hw_init(struct link_encoder *enc) +void enc31_hw_init(struct link_encoder *enc) { struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc); diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_dio_link_encoder.h b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_dio_link_encoder.h index 221671563a0b..ee78ba80797c 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_dio_link_encoder.h +++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_dio_link_encoder.h @@ -89,6 +89,7 @@ SRI(RDPCSTX_PHY_FUSE1, RDPCSTX, id), \ SRI(RDPCSTX_PHY_FUSE2, RDPCSTX, id), \ SRI(RDPCSTX_PHY_FUSE3, RDPCSTX, id), \ + SRI(RDPCSTX_DEBUG_CONFIG, RDPCSTX, id), \ SR(RDPCSTX0_RDPCSTX_SCRATCH), \ SRI(RDPCSTX_PHY_RX_LD_VAL, RDPCSTX, id),\ SRI(RDPCSTX_DMCU_DPALT_DIS_BLOCK_REG, RDPCSTX, id) @@ -222,6 +223,7 @@ SRI(RDPCSTX_PHY_FUSE1, RDPCSTX, id), \ SRI(RDPCSTX_PHY_FUSE2, RDPCSTX, id), \ SRI(RDPCSTX_PHY_FUSE3, RDPCSTX, id), \ + SRI(RDPCSTX_DEBUG_CONFIG, RDPCSTX, id), \ SR(RDPCSTX0_RDPCSTX_SCRATCH), \ SRI(RDPCSTX_PHY_RX_LD_VAL, RDPCSTX, id),\ SRI(RDPCSTX_DMCU_DPALT_DIS_BLOCK_REG, RDPCSTX, id) @@ -283,4 +285,6 @@ bool dcn31_link_encoder_is_in_alt_mode( void dcn31_link_encoder_get_max_link_cap(struct link_encoder *enc, struct dc_link_settings *link_settings); +void enc31_hw_init(struct link_encoder *enc); + #endif /* __DC_LINK_ENCODER__DCN31_H__ */ diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hpo_dp_link_encoder.c b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hpo_dp_link_encoder.c index 5b7ad38f85e0..03b4ac2f1991 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hpo_dp_link_encoder.c +++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hpo_dp_link_encoder.c @@ -377,7 +377,7 @@ void dcn31_hpo_dp_link_enc_update_stream_allocation_table( */ REG_WAIT(DP_DPHY_SYM32_STATUS, SAT_UPDATE_PENDING, 0, - 10, DP_SAT_UPDATE_MAX_RETRY); + 100, DP_SAT_UPDATE_MAX_RETRY); } void dcn31_hpo_dp_link_enc_set_throttled_vcp_size( @@ -395,6 +395,12 @@ void dcn31_hpo_dp_link_enc_set_throttled_vcp_size( x), 25)); + // If y rounds up to integer, carry it over to x. + if (y >> 25) { + x += 1; + y = 0; + } + switch (stream_encoder_inst) { case 0: REG_SET_2(DP_DPHY_SYM32_VC_RATE_CNTL0, 0, diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hpo_dp_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hpo_dp_stream_encoder.c index 45143459eedd..678db949cfe3 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hpo_dp_stream_encoder.c +++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hpo_dp_stream_encoder.c @@ -474,6 +474,10 @@ static void dcn31_hpo_dp_stream_enc_update_dp_info_packets( &info_frame->hdrsmd, true); + /* packetIndex 4 is used for send immediate sdp message, and please + * use other packetIndex (such as 5,6) for other info packet + */ + if (info_frame->adaptive_sync.valid) enc->vpg->funcs->update_generic_info_packet( enc->vpg, diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hubbub.c b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hubbub.c index 5b5b5e0775fa..b906db6e7355 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hubbub.c +++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hubbub.c @@ -172,7 +172,7 @@ static uint32_t convert_and_clamp( static bool hubbub31_program_urgent_watermarks( struct hubbub *hubbub, - struct dcn_watermark_set *watermarks, + union dcn_watermark_set *watermarks, unsigned int refclk_mhz, bool safe_to_lower) { @@ -362,7 +362,7 @@ static bool hubbub31_program_urgent_watermarks( static bool hubbub31_program_stutter_watermarks( struct hubbub *hubbub, - struct dcn_watermark_set *watermarks, + union dcn_watermark_set *watermarks, unsigned int refclk_mhz, bool safe_to_lower) { @@ -635,7 +635,7 @@ static bool hubbub31_program_stutter_watermarks( static bool hubbub31_program_pstate_watermarks( struct hubbub *hubbub, - struct dcn_watermark_set *watermarks, + union dcn_watermark_set *watermarks, unsigned int refclk_mhz, bool safe_to_lower) { @@ -717,7 +717,7 @@ static bool hubbub31_program_pstate_watermarks( static bool hubbub31_program_watermarks( struct hubbub *hubbub, - struct dcn_watermark_set *watermarks, + union dcn_watermark_set *watermarks, unsigned int refclk_mhz, bool safe_to_lower) { diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_panel_cntl.c b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_panel_cntl.c index 281be20b1a10..20c6fe48567f 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_panel_cntl.c +++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_panel_cntl.c @@ -173,5 +173,12 @@ void dcn31_panel_cntl_construct( break; } - dcn31_panel_cntl->base.pwrseq_inst = pwrseq_inst; + if (dcn31_panel_cntl->base.ctx->dc->config.support_edp0_on_dp1) + //If supported, power sequencer mapping shall follow the DIG instance + dcn31_panel_cntl->base.pwrseq_inst = pwrseq_inst; + else + /* If not supported, pwrseq will be assigned in order, + * so first pwrseq will be assigned to first panel instance (legacy behavior) + */ + dcn31_panel_cntl->base.pwrseq_inst = dcn31_panel_cntl->base.inst; } diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_vpg.c b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_vpg.c index f1deb1c3c363..cfb923d85630 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_vpg.c +++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_vpg.c @@ -63,7 +63,12 @@ void vpg31_poweron(struct vpg *vpg) { struct dcn31_vpg *vpg31 = DCN31_VPG_FROM_VPG(vpg); - if (vpg->ctx->dc->debug.enable_mem_low_power.bits.vpg == false) + uint32_t vpg_gsp_mem_pwr_state; + + REG_GET(VPG_MEM_PWR, VPG_GSP_MEM_PWR_STATE, &vpg_gsp_mem_pwr_state); + + if (vpg->ctx->dc->debug.enable_mem_low_power.bits.vpg == false && + vpg_gsp_mem_pwr_state == 0) return; REG_UPDATE_2(VPG_MEM_PWR, VPG_GSP_MEM_LIGHT_SLEEP_DIS, 1, VPG_GSP_LIGHT_SLEEP_FORCE, 0); diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_vpg.h b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_vpg.h index 0e76eabce441..609e58dbd056 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_vpg.h +++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_vpg.h @@ -26,6 +26,7 @@ #ifndef __DAL_DCN31_VPG_H__ #define __DAL_DCN31_VPG_H__ +#include "vpg.h" #define DCN31_VPG_FROM_VPG(vpg)\ container_of(vpg, struct dcn31_vpg, base) diff --git a/drivers/gpu/drm/amd/display/dc/dcn32/Makefile b/drivers/gpu/drm/amd/display/dc/dcn32/Makefile index 5314770fff1c..a58c37165f5a 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn32/Makefile +++ b/drivers/gpu/drm/amd/display/dc/dcn32/Makefile @@ -11,7 +11,7 @@ # Makefile for dcn32. DCN32 = dcn32_hubbub.o dcn32_dccg.o \ - dcn32_mmhubbub.o dcn32_dpp.o dcn32_hubp.o dcn32_mpc.o \ + dcn32_mmhubbub.o dcn32_hubp.o dcn32_mpc.o \ dcn32_dio_stream_encoder.o dcn32_dio_link_encoder.o dcn32_resource_helpers.o \ dcn32_hpo_dp_link_encoder.o diff --git a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dio_link_encoder.c b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dio_link_encoder.c index e224a028d68a..d9ff95cd2dbd 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dio_link_encoder.c +++ b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dio_link_encoder.c @@ -251,8 +251,6 @@ void dcn32_link_encoder_construct( enc10->base.preferred_engine = ENGINE_ID_UNKNOWN; enc10->base.features = *enc_features; - if (enc10->base.connector.id == CONNECTOR_ID_USBC) - enc10->base.features.flags.bits.DP_IS_USB_C = 1; if (enc10->base.connector.id == CONNECTOR_ID_USBC) enc10->base.features.flags.bits.DP_IS_USB_C = 1; diff --git a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dio_link_encoder.h b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dio_link_encoder.h index 2d5f25290ed1..35d23d9db45e 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dio_link_encoder.h +++ b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dio_link_encoder.h @@ -26,15 +26,7 @@ #ifndef __DC_LINK_ENCODER__DCN32_H__ #define __DC_LINK_ENCODER__DCN32_H__ -#include "dcn31/dcn31_dio_link_encoder.h" - -#define LE_DCN32_REG_LIST(id)\ - LE_DCN31_REG_LIST(id),\ - SRI(DIG_FIFO_CTRL0, DIG, id) - -#define LINK_ENCODER_MASK_SH_LIST_DCN32(mask_sh) \ - LINK_ENCODER_MASK_SH_LIST_DCN31(mask_sh),\ - LE_SF(DIG0_DIG_FIFO_CTRL0, DIG_FIFO_OUTPUT_PIXEL_MODE, mask_sh) +#include "dcn30/dcn30_dio_link_encoder.h" void dcn32_link_encoder_construct( struct dcn20_link_encoder *enc20, diff --git a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dio_stream_encoder.h b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dio_stream_encoder.h index 1be5410cce97..ca53d39561d2 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dio_stream_encoder.h +++ b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dio_stream_encoder.h @@ -177,11 +177,12 @@ SE_SF(DIG0_DIG_FE_CNTL, DIG_SYMCLK_FE_ON, mask_sh),\ SE_SF(DP0_DP_SEC_FRAMING4, DP_SST_SDP_SPLITTING, mask_sh),\ SE_SF(DIG0_DIG_CLOCK_PATTERN, DIG_CLOCK_PATTERN, mask_sh),\ + SE_SF(DIG0_DIG_FIFO_CTRL0, DIG_FIFO_OUTPUT_PIXEL_MODE, mask_sh),\ SE_SF(DIG0_DIG_FIFO_CTRL0, DIG_FIFO_READ_START_LEVEL, mask_sh),\ SE_SF(DIG0_DIG_FIFO_CTRL0, DIG_FIFO_ENABLE, mask_sh),\ SE_SF(DIG0_DIG_FIFO_CTRL0, DIG_FIFO_RESET, mask_sh),\ - SE_SF(DIG0_DIG_FIFO_CTRL0, DIG_FIFO_RESET_DONE, mask_sh),\ - SE_SF(DIG0_DIG_FIFO_CTRL0, DIG_FIFO_OUTPUT_PIXEL_MODE, mask_sh) + SE_SF(DIG0_DIG_FIFO_CTRL0, DIG_FIFO_RESET_DONE, mask_sh) + void dcn32_dio_stream_encoder_construct( struct dcn10_stream_encoder *enc1, diff --git a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_hubbub.c b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_hubbub.c index 88dfc907553d..515c4c2b4c21 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_hubbub.c +++ b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_hubbub.c @@ -167,7 +167,7 @@ static uint32_t convert_and_clamp( bool hubbub32_program_urgent_watermarks( struct hubbub *hubbub, - struct dcn_watermark_set *watermarks, + union dcn_watermark_set *watermarks, unsigned int refclk_mhz, bool safe_to_lower) { @@ -357,7 +357,7 @@ bool hubbub32_program_urgent_watermarks( bool hubbub32_program_stutter_watermarks( struct hubbub *hubbub, - struct dcn_watermark_set *watermarks, + union dcn_watermark_set *watermarks, unsigned int refclk_mhz, bool safe_to_lower) { @@ -503,7 +503,7 @@ bool hubbub32_program_stutter_watermarks( bool hubbub32_program_pstate_watermarks( struct hubbub *hubbub, - struct dcn_watermark_set *watermarks, + union dcn_watermark_set *watermarks, unsigned int refclk_mhz, bool safe_to_lower) { @@ -656,7 +656,7 @@ bool hubbub32_program_pstate_watermarks( bool hubbub32_program_usr_watermarks( struct hubbub *hubbub, - struct dcn_watermark_set *watermarks, + union dcn_watermark_set *watermarks, unsigned int refclk_mhz, bool safe_to_lower) { @@ -750,7 +750,7 @@ void hubbub32_force_usr_retraining_allow(struct hubbub *hubbub, bool allow) static bool hubbub32_program_watermarks( struct hubbub *hubbub, - struct dcn_watermark_set *watermarks, + union dcn_watermark_set *watermarks, unsigned int refclk_mhz, bool safe_to_lower) { diff --git a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_hubbub.h b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_hubbub.h index f073839a4b6d..e439ba0fa30f 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_hubbub.h +++ b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_hubbub.h @@ -118,25 +118,25 @@ bool hubbub32_program_urgent_watermarks( struct hubbub *hubbub, - struct dcn_watermark_set *watermarks, + union dcn_watermark_set *watermarks, unsigned int refclk_mhz, bool safe_to_lower); bool hubbub32_program_stutter_watermarks( struct hubbub *hubbub, - struct dcn_watermark_set *watermarks, + union dcn_watermark_set *watermarks, unsigned int refclk_mhz, bool safe_to_lower); bool hubbub32_program_pstate_watermarks( struct hubbub *hubbub, - struct dcn_watermark_set *watermarks, + union dcn_watermark_set *watermarks, unsigned int refclk_mhz, bool safe_to_lower); bool hubbub32_program_usr_watermarks( struct hubbub *hubbub, - struct dcn_watermark_set *watermarks, + union dcn_watermark_set *watermarks, unsigned int refclk_mhz, bool safe_to_lower); diff --git a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_mpc.c b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_mpc.c index e789e654c387..e408e859b355 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_mpc.c +++ b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_mpc.c @@ -47,7 +47,7 @@ void mpc32_mpc_init(struct mpc *mpc) struct dcn30_mpc *mpc30 = TO_DCN30_MPC(mpc); int mpcc_id; - mpc1_mpc_init(mpc); + mpc3_mpc_init(mpc); if (mpc->ctx->dc->debug.enable_mem_low_power.bits.mpc) { if (mpc30->mpc_mask->MPCC_MCM_SHAPER_MEM_LOW_PWR_MODE && mpc30->mpc_mask->MPCC_MCM_3DLUT_MEM_LOW_PWR_MODE) { @@ -991,7 +991,7 @@ static const struct mpc_funcs dcn32_mpc_funcs = { .insert_plane = mpc1_insert_plane, .remove_mpcc = mpc1_remove_mpcc, .mpc_init = mpc32_mpc_init, - .mpc_init_single_inst = mpc1_mpc_init_single_inst, + .mpc_init_single_inst = mpc3_mpc_init_single_inst, .update_blending = mpc2_update_blending, .cursor_lock = mpc1_cursor_lock, .get_mpcc_for_dpp = mpc1_get_mpcc_for_dpp, @@ -1008,7 +1008,6 @@ static const struct mpc_funcs dcn32_mpc_funcs = { .set_dwb_mux = mpc3_set_dwb_mux, .disable_dwb_mux = mpc3_disable_dwb_mux, .is_dwb_idle = mpc3_is_dwb_idle, - .set_out_rate_control = mpc3_set_out_rate_control, .set_gamut_remap = mpc3_set_gamut_remap, .program_shaper = mpc32_program_shaper, .program_3dlut = mpc32_program_3dlut, diff --git a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_resource_helpers.c b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_resource_helpers.c index f98def6c8c2d..fbcd6f7bc993 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_resource_helpers.c +++ b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_resource_helpers.c @@ -35,25 +35,6 @@ static bool is_dual_plane(enum surface_pixel_format format) return format >= SURFACE_PIXEL_FORMAT_VIDEO_BEGIN || format == SURFACE_PIXEL_FORMAT_GRPH_RGBE_ALPHA; } - -uint32_t dcn32_helper_mall_bytes_to_ways( - struct dc *dc, - uint32_t total_size_in_mall_bytes) -{ - uint32_t cache_lines_used, lines_per_way, total_cache_lines, num_ways; - - /* add 2 lines for worst case alignment */ - cache_lines_used = total_size_in_mall_bytes / dc->caps.cache_line_size + 2; - - total_cache_lines = dc->caps.max_cab_allocation_bytes / dc->caps.cache_line_size; - lines_per_way = total_cache_lines / dc->caps.cache_num_ways; - num_ways = cache_lines_used / lines_per_way; - if (cache_lines_used % lines_per_way > 0) - num_ways++; - - return num_ways; -} - uint32_t dcn32_helper_calculate_mall_bytes_for_cursor( struct dc *dc, struct pipe_ctx *pipe_ctx, @@ -112,8 +93,10 @@ uint32_t dcn32_helper_calculate_num_ways_for_subvp( if (context->bw_ctx.bw.dcn.mall_subvp_size_bytes > 0) { if (dc->debug.force_subvp_num_ways) { return dc->debug.force_subvp_num_ways; + } else if (dc->res_pool->funcs->calculate_mall_ways_from_bytes) { + return dc->res_pool->funcs->calculate_mall_ways_from_bytes(dc, context->bw_ctx.bw.dcn.mall_subvp_size_bytes); } else { - return dcn32_helper_mall_bytes_to_ways(dc, context->bw_ctx.bw.dcn.mall_subvp_size_bytes); + return 0; } } else { return 0; @@ -399,7 +382,7 @@ void dcn32_set_det_allocations(struct dc *dc, struct dc_state *context, { int i, pipe_cnt; struct resource_context *res_ctx = &context->res_ctx; - struct pipe_ctx *pipe; + struct pipe_ctx *pipe = 0; bool disable_unbounded_requesting = dc->debug.disable_z9_mpc || dc->debug.disable_unbounded_requesting; for (i = 0, pipe_cnt = 0; i < dc->res_pool->pipe_count; i++) { diff --git a/drivers/gpu/drm/amd/display/dc/dcn321/dcn321_dio_link_encoder.c b/drivers/gpu/drm/amd/display/dc/dcn321/dcn321_dio_link_encoder.c index 13be5f06d987..05783daa62ac 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn321/dcn321_dio_link_encoder.c +++ b/drivers/gpu/drm/amd/display/dc/dcn321/dcn321_dio_link_encoder.c @@ -127,11 +127,6 @@ void dcn321_link_encoder_construct( * while doing the DP sink detect */ -/* if (dal_adapter_service_is_feature_supported(as, - FEATURE_DP_SINK_DETECT_POLL_DATA_PIN)) - enc10->base.features.flags.bits. - DP_SINK_DETECT_POLL_DATA_PIN = true;*/ - enc10->base.output_signals = SIGNAL_TYPE_DVI_SINGLE_LINK | SIGNAL_TYPE_DVI_DUAL_LINK | @@ -191,7 +186,6 @@ void dcn321_link_encoder_construct( __func__, result); } - if (enc10->base.ctx->dc->debug.hdmi20_disable) { + if (enc10->base.ctx->dc->debug.hdmi20_disable) enc10->base.features.flags.bits.HDMI_6GB_EN = 0; - } } diff --git a/drivers/gpu/drm/amd/display/dc/dcn35/Makefile b/drivers/gpu/drm/amd/display/dc/dcn35/Makefile index 0e317e0c36a0..d5b4533d2f62 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn35/Makefile +++ b/drivers/gpu/drm/amd/display/dc/dcn35/Makefile @@ -13,7 +13,7 @@ DCN35 = dcn35_dio_stream_encoder.o \ dcn35_dio_link_encoder.o dcn35_dccg.o \ dcn35_hubp.o dcn35_hubbub.o \ - dcn35_mmhubbub.o dcn35_opp.o dcn35_dpp.o dcn35_pg_cntl.o dcn35_dwb.o + dcn35_mmhubbub.o dcn35_opp.o dcn35_pg_cntl.o dcn35_dwb.o AMD_DAL_DCN35 = $(addprefix $(AMDDALPATH)/dc/dcn35/,$(DCN35)) diff --git a/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_dccg.c b/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_dccg.c index f1ba7bb792ea..58dd3c5bbff0 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_dccg.c +++ b/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_dccg.c @@ -49,15 +49,23 @@ static void dcn35_set_dppclk_enable(struct dccg *dccg, switch (dpp_inst) { case 0: REG_UPDATE(DPPCLK_CTRL, DPPCLK0_EN, enable); + if (dccg->ctx->dc->debug.root_clock_optimization.bits.dpp) + REG_UPDATE(DCCG_GATE_DISABLE_CNTL6, DPPCLK0_ROOT_GATE_DISABLE, enable); break; case 1: REG_UPDATE(DPPCLK_CTRL, DPPCLK1_EN, enable); + if (dccg->ctx->dc->debug.root_clock_optimization.bits.dpp) + REG_UPDATE(DCCG_GATE_DISABLE_CNTL6, DPPCLK1_ROOT_GATE_DISABLE, enable); break; case 2: REG_UPDATE(DPPCLK_CTRL, DPPCLK2_EN, enable); + if (dccg->ctx->dc->debug.root_clock_optimization.bits.dpp) + REG_UPDATE(DCCG_GATE_DISABLE_CNTL6, DPPCLK2_ROOT_GATE_DISABLE, enable); break; case 3: REG_UPDATE(DPPCLK_CTRL, DPPCLK3_EN, enable); + if (dccg->ctx->dc->debug.root_clock_optimization.bits.dpp) + REG_UPDATE(DCCG_GATE_DISABLE_CNTL6, DPPCLK3_ROOT_GATE_DISABLE, enable); break; default: break; @@ -100,6 +108,32 @@ static void dccg35_update_dpp_dto(struct dccg *dccg, int dpp_inst, dccg->pipe_dppclk_khz[dpp_inst] = req_dppclk; } +static void dccg35_set_dppclk_root_clock_gating(struct dccg *dccg, + uint32_t dpp_inst, uint32_t enable) +{ + struct dcn_dccg *dccg_dcn = TO_DCN_DCCG(dccg); + + if (!dccg->ctx->dc->debug.root_clock_optimization.bits.dpp) + return; + + switch (dpp_inst) { + case 0: + REG_UPDATE(DCCG_GATE_DISABLE_CNTL6, DPPCLK0_ROOT_GATE_DISABLE, enable); + break; + case 1: + REG_UPDATE(DCCG_GATE_DISABLE_CNTL6, DPPCLK1_ROOT_GATE_DISABLE, enable); + break; + case 2: + REG_UPDATE(DCCG_GATE_DISABLE_CNTL6, DPPCLK2_ROOT_GATE_DISABLE, enable); + break; + case 3: + REG_UPDATE(DCCG_GATE_DISABLE_CNTL6, DPPCLK3_ROOT_GATE_DISABLE, enable); + break; + default: + break; + } +} + static void dccg35_get_pixel_rate_div( struct dccg *dccg, uint32_t otg_inst, @@ -333,21 +367,67 @@ static void dccg35_set_dpstreamclk( /* enabled to select one of the DTBCLKs for pipe */ switch (dp_hpo_inst) { case 0: - REG_UPDATE_2(DPSTREAMCLK_CNTL, - DPSTREAMCLK0_EN, + REG_UPDATE_2(DPSTREAMCLK_CNTL, DPSTREAMCLK0_EN, (src == REFCLK) ? 0 : 1, DPSTREAMCLK0_SRC_SEL, otg_inst); + if (dccg->ctx->dc->debug.root_clock_optimization.bits.dpstream) + REG_UPDATE(DCCG_GATE_DISABLE_CNTL5, DPSTREAMCLK0_ROOT_GATE_DISABLE, (src == REFCLK) ? 0 : 1); break; case 1: REG_UPDATE_2(DPSTREAMCLK_CNTL, DPSTREAMCLK1_EN, (src == REFCLK) ? 0 : 1, DPSTREAMCLK1_SRC_SEL, otg_inst); + if (dccg->ctx->dc->debug.root_clock_optimization.bits.dpstream) + REG_UPDATE(DCCG_GATE_DISABLE_CNTL5, DPSTREAMCLK1_ROOT_GATE_DISABLE, (src == REFCLK) ? 0 : 1); break; case 2: REG_UPDATE_2(DPSTREAMCLK_CNTL, DPSTREAMCLK2_EN, (src == REFCLK) ? 0 : 1, DPSTREAMCLK2_SRC_SEL, otg_inst); + if (dccg->ctx->dc->debug.root_clock_optimization.bits.dpstream) + REG_UPDATE(DCCG_GATE_DISABLE_CNTL5, DPSTREAMCLK2_ROOT_GATE_DISABLE, (src == REFCLK) ? 0 : 1); break; case 3: REG_UPDATE_2(DPSTREAMCLK_CNTL, DPSTREAMCLK3_EN, (src == REFCLK) ? 0 : 1, DPSTREAMCLK3_SRC_SEL, otg_inst); + if (dccg->ctx->dc->debug.root_clock_optimization.bits.dpstream) + REG_UPDATE(DCCG_GATE_DISABLE_CNTL5, DPSTREAMCLK3_ROOT_GATE_DISABLE, (src == REFCLK) ? 0 : 1); + break; + default: + BREAK_TO_DEBUGGER(); + return; + } +} + + +static void dccg35_set_dpstreamclk_root_clock_gating( + struct dccg *dccg, + int dp_hpo_inst, + bool enable) +{ + struct dcn_dccg *dccg_dcn = TO_DCN_DCCG(dccg); + + switch (dp_hpo_inst) { + case 0: + if (dccg->ctx->dc->debug.root_clock_optimization.bits.dpstream) { + REG_UPDATE(DCCG_GATE_DISABLE_CNTL5, DPSTREAMCLK0_ROOT_GATE_DISABLE, enable ? 1 : 0); + REG_UPDATE(DCCG_GATE_DISABLE_CNTL5, DPSTREAMCLK0_GATE_DISABLE, enable ? 1 : 0); + } + break; + case 1: + if (dccg->ctx->dc->debug.root_clock_optimization.bits.dpstream) { + REG_UPDATE(DCCG_GATE_DISABLE_CNTL5, DPSTREAMCLK1_ROOT_GATE_DISABLE, enable ? 1 : 0); + REG_UPDATE(DCCG_GATE_DISABLE_CNTL5, DPSTREAMCLK1_GATE_DISABLE, enable ? 1 : 0); + } + break; + case 2: + if (dccg->ctx->dc->debug.root_clock_optimization.bits.dpstream) { + REG_UPDATE(DCCG_GATE_DISABLE_CNTL5, DPSTREAMCLK2_ROOT_GATE_DISABLE, enable ? 1 : 0); + REG_UPDATE(DCCG_GATE_DISABLE_CNTL5, DPSTREAMCLK2_GATE_DISABLE, enable ? 1 : 0); + } + break; + case 3: + if (dccg->ctx->dc->debug.root_clock_optimization.bits.dpstream) { + REG_UPDATE(DCCG_GATE_DISABLE_CNTL5, DPSTREAMCLK3_ROOT_GATE_DISABLE, enable ? 1 : 0); + REG_UPDATE(DCCG_GATE_DISABLE_CNTL5, DPSTREAMCLK3_GATE_DISABLE, enable ? 1 : 0); + } break; default: BREAK_TO_DEBUGGER(); @@ -355,6 +435,8 @@ static void dccg35_set_dpstreamclk( } } + + static void dccg35_set_physymclk_root_clock_gating( struct dccg *dccg, int phy_inst, @@ -369,22 +451,32 @@ static void dccg35_set_physymclk_root_clock_gating( case 0: REG_UPDATE(DCCG_GATE_DISABLE_CNTL2, PHYASYMCLK_ROOT_GATE_DISABLE, enable ? 1 : 0); +// REG_UPDATE(DCCG_GATE_DISABLE_CNTL4, +// PHYA_REFCLK_ROOT_GATE_DISABLE, enable ? 1 : 0); break; case 1: REG_UPDATE(DCCG_GATE_DISABLE_CNTL2, PHYBSYMCLK_ROOT_GATE_DISABLE, enable ? 1 : 0); +// REG_UPDATE(DCCG_GATE_DISABLE_CNTL4, +// PHYB_REFCLK_ROOT_GATE_DISABLE, enable ? 1 : 0); break; case 2: REG_UPDATE(DCCG_GATE_DISABLE_CNTL2, PHYCSYMCLK_ROOT_GATE_DISABLE, enable ? 1 : 0); +// REG_UPDATE(DCCG_GATE_DISABLE_CNTL4, +// PHYC_REFCLK_ROOT_GATE_DISABLE, enable ? 1 : 0); break; case 3: REG_UPDATE(DCCG_GATE_DISABLE_CNTL2, PHYDSYMCLK_ROOT_GATE_DISABLE, enable ? 1 : 0); +// REG_UPDATE(DCCG_GATE_DISABLE_CNTL4, +// PHYD_REFCLK_ROOT_GATE_DISABLE, enable ? 1 : 0); break; case 4: REG_UPDATE(DCCG_GATE_DISABLE_CNTL2, PHYESYMCLK_ROOT_GATE_DISABLE, enable ? 1 : 0); +// REG_UPDATE(DCCG_GATE_DISABLE_CNTL4, +// PHYE_REFCLK_ROOT_GATE_DISABLE, enable ? 1 : 0); break; default: BREAK_TO_DEBUGGER(); @@ -407,10 +499,16 @@ static void dccg35_set_physymclk( REG_UPDATE_2(PHYASYMCLK_CLOCK_CNTL, PHYASYMCLK_EN, 1, PHYASYMCLK_SRC_SEL, clk_src); +// if (dccg->ctx->dc->debug.root_clock_optimization.bits.physymclk) +// REG_UPDATE(DCCG_GATE_DISABLE_CNTL4, +// PHYA_REFCLK_ROOT_GATE_DISABLE, 0); } else { REG_UPDATE_2(PHYASYMCLK_CLOCK_CNTL, PHYASYMCLK_EN, 0, PHYASYMCLK_SRC_SEL, 0); +// if (dccg->ctx->dc->debug.root_clock_optimization.bits.physymclk) +// REG_UPDATE(DCCG_GATE_DISABLE_CNTL4, +// PHYA_REFCLK_ROOT_GATE_DISABLE, 1); } break; case 1: @@ -418,10 +516,16 @@ static void dccg35_set_physymclk( REG_UPDATE_2(PHYBSYMCLK_CLOCK_CNTL, PHYBSYMCLK_EN, 1, PHYBSYMCLK_SRC_SEL, clk_src); +// if (dccg->ctx->dc->debug.root_clock_optimization.bits.physymclk) +// REG_UPDATE(DCCG_GATE_DISABLE_CNTL4, +// PHYB_REFCLK_ROOT_GATE_DISABLE, 0); } else { REG_UPDATE_2(PHYBSYMCLK_CLOCK_CNTL, PHYBSYMCLK_EN, 0, PHYBSYMCLK_SRC_SEL, 0); +// if (dccg->ctx->dc->debug.root_clock_optimization.bits.physymclk) +// REG_UPDATE(DCCG_GATE_DISABLE_CNTL4, +// PHYB_REFCLK_ROOT_GATE_DISABLE, 1); } break; case 2: @@ -429,10 +533,16 @@ static void dccg35_set_physymclk( REG_UPDATE_2(PHYCSYMCLK_CLOCK_CNTL, PHYCSYMCLK_EN, 1, PHYCSYMCLK_SRC_SEL, clk_src); +// if (dccg->ctx->dc->debug.root_clock_optimization.bits.physymclk) +// REG_UPDATE(DCCG_GATE_DISABLE_CNTL4, +// PHYC_REFCLK_ROOT_GATE_DISABLE, 0); } else { REG_UPDATE_2(PHYCSYMCLK_CLOCK_CNTL, PHYCSYMCLK_EN, 0, PHYCSYMCLK_SRC_SEL, 0); +// if (dccg->ctx->dc->debug.root_clock_optimization.bits.physymclk) +// REG_UPDATE(DCCG_GATE_DISABLE_CNTL4, +// PHYC_REFCLK_ROOT_GATE_DISABLE, 1); } break; case 3: @@ -440,10 +550,16 @@ static void dccg35_set_physymclk( REG_UPDATE_2(PHYDSYMCLK_CLOCK_CNTL, PHYDSYMCLK_EN, 1, PHYDSYMCLK_SRC_SEL, clk_src); +// if (dccg->ctx->dc->debug.root_clock_optimization.bits.physymclk) +// REG_UPDATE(DCCG_GATE_DISABLE_CNTL4, +// PHYD_REFCLK_ROOT_GATE_DISABLE, 0); } else { REG_UPDATE_2(PHYDSYMCLK_CLOCK_CNTL, PHYDSYMCLK_EN, 0, PHYDSYMCLK_SRC_SEL, 0); +// if (dccg->ctx->dc->debug.root_clock_optimization.bits.physymclk) +// REG_UPDATE(DCCG_GATE_DISABLE_CNTL4, +// PHYD_REFCLK_ROOT_GATE_DISABLE, 1); } break; case 4: @@ -451,10 +567,16 @@ static void dccg35_set_physymclk( REG_UPDATE_2(PHYESYMCLK_CLOCK_CNTL, PHYESYMCLK_EN, 1, PHYESYMCLK_SRC_SEL, clk_src); +// if (dccg->ctx->dc->debug.root_clock_optimization.bits.physymclk) +// REG_UPDATE(DCCG_GATE_DISABLE_CNTL4, +// PHYE_REFCLK_ROOT_GATE_DISABLE, 0); } else { REG_UPDATE_2(PHYESYMCLK_CLOCK_CNTL, PHYESYMCLK_EN, 0, PHYESYMCLK_SRC_SEL, 0); +// if (dccg->ctx->dc->debug.root_clock_optimization.bits.physymclk) +// REG_UPDATE(DCCG_GATE_DISABLE_CNTL4, +// PHYE_REFCLK_ROOT_GATE_DISABLE, 1); } break; default: @@ -491,12 +613,12 @@ static void dccg35_dpp_root_clock_control( if (clock_on) { /* turn off the DTO and leave phase/modulo at max */ - dcn35_set_dppclk_enable(dccg, dpp_inst, 0); + dcn35_set_dppclk_enable(dccg, dpp_inst, 1); REG_SET_2(DPPCLK_DTO_PARAM[dpp_inst], 0, DPPCLK0_DTO_PHASE, 0xFF, DPPCLK0_DTO_MODULO, 0xFF); } else { - dcn35_set_dppclk_enable(dccg, dpp_inst, 1); + dcn35_set_dppclk_enable(dccg, dpp_inst, 0); /* turn on the DTO to generate a 0hz clock */ REG_SET_2(DPPCLK_DTO_PARAM[dpp_inst], 0, DPPCLK0_DTO_PHASE, 0, @@ -575,18 +697,32 @@ void dccg35_init(struct dccg *dccg) dccg35_disable_symclk32_se(dccg, otg_inst); if (dccg->ctx->dc->debug.root_clock_optimization.bits.symclk32_le) - for (otg_inst = 0; otg_inst < 2; otg_inst++) + for (otg_inst = 0; otg_inst < 2; otg_inst++) { dccg31_disable_symclk32_le(dccg, otg_inst); + dccg31_set_symclk32_le_root_clock_gating(dccg, otg_inst, false); + } + +// if (dccg->ctx->dc->debug.root_clock_optimization.bits.symclk32_se) +// for (otg_inst = 0; otg_inst < 4; otg_inst++) +// dccg35_disable_symclk_se(dccg, otg_inst, otg_inst); + if (dccg->ctx->dc->debug.root_clock_optimization.bits.dpstream) - for (otg_inst = 0; otg_inst < 4; otg_inst++) - dccg314_set_dpstreamclk(dccg, REFCLK, otg_inst, + for (otg_inst = 0; otg_inst < 4; otg_inst++) { + dccg35_set_dpstreamclk(dccg, REFCLK, otg_inst, otg_inst); + dccg35_set_dpstreamclk_root_clock_gating(dccg, otg_inst, false); + } if (dccg->ctx->dc->debug.root_clock_optimization.bits.physymclk) for (otg_inst = 0; otg_inst < 5; otg_inst++) dccg35_set_physymclk_root_clock_gating(dccg, otg_inst, false); + + if (dccg->ctx->dc->debug.root_clock_optimization.bits.dpp) + for (otg_inst = 0; otg_inst < 4; otg_inst++) + dccg35_set_dppclk_root_clock_gating(dccg, otg_inst, 0); + /* dccg35_enable_global_fgcg_rep( dccg, dccg->ctx->dc->debug.enable_fine_grain_clock_gating.bits @@ -611,24 +747,32 @@ static void dccg35_enable_dscclk(struct dccg *dccg, int inst) DSCCLK0_DTO_PHASE, 0, DSCCLK0_DTO_MODULO, 0); REG_UPDATE(DSCCLK_DTO_CTRL, DSCCLK0_EN, 1); + if (dccg->ctx->dc->debug.root_clock_optimization.bits.dsc) + REG_UPDATE(DCCG_GATE_DISABLE_CNTL6, DSCCLK0_ROOT_GATE_DISABLE, 1); break; case 1: REG_UPDATE_2(DSCCLK1_DTO_PARAM, DSCCLK1_DTO_PHASE, 0, DSCCLK1_DTO_MODULO, 0); REG_UPDATE(DSCCLK_DTO_CTRL, DSCCLK1_EN, 1); + if (dccg->ctx->dc->debug.root_clock_optimization.bits.dsc) + REG_UPDATE(DCCG_GATE_DISABLE_CNTL6, DSCCLK1_ROOT_GATE_DISABLE, 1); break; case 2: REG_UPDATE_2(DSCCLK2_DTO_PARAM, DSCCLK2_DTO_PHASE, 0, DSCCLK2_DTO_MODULO, 0); REG_UPDATE(DSCCLK_DTO_CTRL, DSCCLK2_EN, 1); + if (dccg->ctx->dc->debug.root_clock_optimization.bits.dsc) + REG_UPDATE(DCCG_GATE_DISABLE_CNTL6, DSCCLK2_ROOT_GATE_DISABLE, 1); break; case 3: REG_UPDATE_2(DSCCLK3_DTO_PARAM, DSCCLK3_DTO_PHASE, 0, DSCCLK3_DTO_MODULO, 0); REG_UPDATE(DSCCLK_DTO_CTRL, DSCCLK3_EN, 1); + if (dccg->ctx->dc->debug.root_clock_optimization.bits.dsc) + REG_UPDATE(DCCG_GATE_DISABLE_CNTL6, DSCCLK3_ROOT_GATE_DISABLE, 1); break; default: BREAK_TO_DEBUGGER(); @@ -650,24 +794,32 @@ static void dccg35_disable_dscclk(struct dccg *dccg, REG_UPDATE_2(DSCCLK0_DTO_PARAM, DSCCLK0_DTO_PHASE, 0, DSCCLK0_DTO_MODULO, 1); + if (dccg->ctx->dc->debug.root_clock_optimization.bits.dsc) + REG_UPDATE(DCCG_GATE_DISABLE_CNTL6, DSCCLK0_ROOT_GATE_DISABLE, 0); break; case 1: REG_UPDATE(DSCCLK_DTO_CTRL, DSCCLK1_EN, 0); REG_UPDATE_2(DSCCLK1_DTO_PARAM, DSCCLK1_DTO_PHASE, 0, DSCCLK1_DTO_MODULO, 1); + if (dccg->ctx->dc->debug.root_clock_optimization.bits.dsc) + REG_UPDATE(DCCG_GATE_DISABLE_CNTL6, DSCCLK1_ROOT_GATE_DISABLE, 0); break; case 2: REG_UPDATE(DSCCLK_DTO_CTRL, DSCCLK2_EN, 0); REG_UPDATE_2(DSCCLK2_DTO_PARAM, DSCCLK2_DTO_PHASE, 0, DSCCLK2_DTO_MODULO, 1); + if (dccg->ctx->dc->debug.root_clock_optimization.bits.dsc) + REG_UPDATE(DCCG_GATE_DISABLE_CNTL6, DSCCLK2_ROOT_GATE_DISABLE, 0); break; case 3: REG_UPDATE(DSCCLK_DTO_CTRL, DSCCLK3_EN, 0); REG_UPDATE_2(DSCCLK3_DTO_PARAM, DSCCLK3_DTO_PHASE, 0, DSCCLK3_DTO_MODULO, 1); + if (dccg->ctx->dc->debug.root_clock_optimization.bits.dsc) + REG_UPDATE(DCCG_GATE_DISABLE_CNTL6, DSCCLK3_ROOT_GATE_DISABLE, 0); break; default: return; @@ -682,22 +834,32 @@ static void dccg35_enable_symclk_se(struct dccg *dccg, uint32_t stream_enc_inst, case 0: REG_UPDATE(SYMCLKA_CLOCK_ENABLE, SYMCLKA_CLOCK_ENABLE, 1); + if (dccg->ctx->dc->debug.root_clock_optimization.bits.symclk32_se) + REG_UPDATE(DCCG_GATE_DISABLE_CNTL5, SYMCLKA_ROOT_GATE_DISABLE, 1); break; case 1: REG_UPDATE(SYMCLKB_CLOCK_ENABLE, SYMCLKB_CLOCK_ENABLE, 1); + if (dccg->ctx->dc->debug.root_clock_optimization.bits.symclk32_se) + REG_UPDATE(DCCG_GATE_DISABLE_CNTL5, SYMCLKB_ROOT_GATE_DISABLE, 1); break; case 2: REG_UPDATE(SYMCLKC_CLOCK_ENABLE, SYMCLKC_CLOCK_ENABLE, 1); + if (dccg->ctx->dc->debug.root_clock_optimization.bits.symclk32_se) + REG_UPDATE(DCCG_GATE_DISABLE_CNTL5, SYMCLKC_ROOT_GATE_DISABLE, 1); break; case 3: REG_UPDATE(SYMCLKD_CLOCK_ENABLE, SYMCLKD_CLOCK_ENABLE, 1); + if (dccg->ctx->dc->debug.root_clock_optimization.bits.symclk32_se) + REG_UPDATE(DCCG_GATE_DISABLE_CNTL5, SYMCLKD_ROOT_GATE_DISABLE, 1); break; case 4: REG_UPDATE(SYMCLKE_CLOCK_ENABLE, SYMCLKE_CLOCK_ENABLE, 1); + if (dccg->ctx->dc->debug.root_clock_optimization.bits.symclk32_se) + REG_UPDATE(DCCG_GATE_DISABLE_CNTL5, SYMCLKE_ROOT_GATE_DISABLE, 1); break; } @@ -706,26 +868,36 @@ static void dccg35_enable_symclk_se(struct dccg *dccg, uint32_t stream_enc_inst, REG_UPDATE_2(SYMCLKA_CLOCK_ENABLE, SYMCLKA_FE_EN, 1, SYMCLKA_FE_SRC_SEL, link_enc_inst); + if (dccg->ctx->dc->debug.root_clock_optimization.bits.symclk32_se) + REG_UPDATE(DCCG_GATE_DISABLE_CNTL5, SYMCLKA_FE_ROOT_GATE_DISABLE, 1); break; case 1: REG_UPDATE_2(SYMCLKB_CLOCK_ENABLE, SYMCLKB_FE_EN, 1, SYMCLKB_FE_SRC_SEL, link_enc_inst); + if (dccg->ctx->dc->debug.root_clock_optimization.bits.symclk32_se) + REG_UPDATE(DCCG_GATE_DISABLE_CNTL5, SYMCLKB_FE_ROOT_GATE_DISABLE, 1); break; case 2: REG_UPDATE_2(SYMCLKC_CLOCK_ENABLE, SYMCLKC_FE_EN, 1, SYMCLKC_FE_SRC_SEL, link_enc_inst); + if (dccg->ctx->dc->debug.root_clock_optimization.bits.symclk32_se) + REG_UPDATE(DCCG_GATE_DISABLE_CNTL5, SYMCLKC_FE_ROOT_GATE_DISABLE, 1); break; case 3: REG_UPDATE_2(SYMCLKD_CLOCK_ENABLE, SYMCLKD_FE_EN, 1, SYMCLKD_FE_SRC_SEL, link_enc_inst); + if (dccg->ctx->dc->debug.root_clock_optimization.bits.symclk32_se) + REG_UPDATE(DCCG_GATE_DISABLE_CNTL5, SYMCLKD_FE_ROOT_GATE_DISABLE, 1); break; case 4: REG_UPDATE_2(SYMCLKE_CLOCK_ENABLE, SYMCLKE_FE_EN, 1, SYMCLKE_FE_SRC_SEL, link_enc_inst); + if (dccg->ctx->dc->debug.root_clock_optimization.bits.symclk32_se) + REG_UPDATE(DCCG_GATE_DISABLE_CNTL5, SYMCLKE_FE_ROOT_GATE_DISABLE, 1); break; } } @@ -786,26 +958,36 @@ static void dccg35_disable_symclk_se(struct dccg *dccg, uint32_t stream_enc_inst REG_UPDATE_2(SYMCLKA_CLOCK_ENABLE, SYMCLKA_FE_EN, 0, SYMCLKA_FE_SRC_SEL, 0); +// if (dccg->ctx->dc->debug.root_clock_optimization.bits.symclk32_se) +// REG_UPDATE(DCCG_GATE_DISABLE_CNTL5, SYMCLKA_FE_ROOT_GATE_DISABLE, 0); break; case 1: REG_UPDATE_2(SYMCLKB_CLOCK_ENABLE, SYMCLKB_FE_EN, 0, SYMCLKB_FE_SRC_SEL, 0); +// if (dccg->ctx->dc->debug.root_clock_optimization.bits.symclk32_se) +// REG_UPDATE(DCCG_GATE_DISABLE_CNTL5, SYMCLKB_FE_ROOT_GATE_DISABLE, 0); break; case 2: REG_UPDATE_2(SYMCLKC_CLOCK_ENABLE, SYMCLKC_FE_EN, 0, SYMCLKC_FE_SRC_SEL, 0); +// if (dccg->ctx->dc->debug.root_clock_optimization.bits.symclk32_se) +// REG_UPDATE(DCCG_GATE_DISABLE_CNTL5, SYMCLKC_FE_ROOT_GATE_DISABLE, 0); break; case 3: REG_UPDATE_2(SYMCLKD_CLOCK_ENABLE, SYMCLKD_FE_EN, 0, SYMCLKD_FE_SRC_SEL, 0); +// if (dccg->ctx->dc->debug.root_clock_optimization.bits.symclk32_se) +// REG_UPDATE(DCCG_GATE_DISABLE_CNTL5, SYMCLKD_FE_ROOT_GATE_DISABLE, 0); break; case 4: REG_UPDATE_2(SYMCLKE_CLOCK_ENABLE, SYMCLKE_FE_EN, 0, SYMCLKE_FE_SRC_SEL, 0); +// if (dccg->ctx->dc->debug.root_clock_optimization.bits.symclk32_se) +// REG_UPDATE(DCCG_GATE_DISABLE_CNTL5, SYMCLKE_FE_ROOT_GATE_DISABLE, 0); break; } @@ -818,22 +1000,32 @@ static void dccg35_disable_symclk_se(struct dccg *dccg, uint32_t stream_enc_inst case 0: REG_UPDATE(SYMCLKA_CLOCK_ENABLE, SYMCLKA_CLOCK_ENABLE, 0); +// if (dccg->ctx->dc->debug.root_clock_optimization.bits.symclk32_le) +// REG_UPDATE(DCCG_GATE_DISABLE_CNTL5, SYMCLKA_ROOT_GATE_DISABLE, 0); break; case 1: REG_UPDATE(SYMCLKB_CLOCK_ENABLE, SYMCLKB_CLOCK_ENABLE, 0); +// if (dccg->ctx->dc->debug.root_clock_optimization.bits.symclk32_le) +// REG_UPDATE(DCCG_GATE_DISABLE_CNTL5, SYMCLKB_ROOT_GATE_DISABLE, 0); break; case 2: REG_UPDATE(SYMCLKC_CLOCK_ENABLE, SYMCLKC_CLOCK_ENABLE, 0); +// if (dccg->ctx->dc->debug.root_clock_optimization.bits.symclk32_le) +// REG_UPDATE(DCCG_GATE_DISABLE_CNTL5, SYMCLKC_ROOT_GATE_DISABLE, 0); break; case 3: REG_UPDATE(SYMCLKD_CLOCK_ENABLE, SYMCLKD_CLOCK_ENABLE, 0); +// if (dccg->ctx->dc->debug.root_clock_optimization.bits.symclk32_le) +// REG_UPDATE(DCCG_GATE_DISABLE_CNTL5, SYMCLKD_ROOT_GATE_DISABLE, 0); break; case 4: REG_UPDATE(SYMCLKE_CLOCK_ENABLE, SYMCLKE_CLOCK_ENABLE, 0); +// if (dccg->ctx->dc->debug.root_clock_optimization.bits.symclk32_le) +// REG_UPDATE(DCCG_GATE_DISABLE_CNTL5, SYMCLKE_ROOT_GATE_DISABLE, 0); break; } } @@ -845,6 +1037,7 @@ static const struct dccg_funcs dccg35_funcs = { .get_dccg_ref_freq = dccg31_get_dccg_ref_freq, .dccg_init = dccg35_init, .set_dpstreamclk = dccg35_set_dpstreamclk, + .set_dpstreamclk_root_clock_gating = dccg35_set_dpstreamclk_root_clock_gating, .enable_symclk32_se = dccg31_enable_symclk32_se, .disable_symclk32_se = dccg35_disable_symclk32_se, .enable_symclk32_le = dccg31_enable_symclk32_le, diff --git a/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_dio_link_encoder.c b/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_dio_link_encoder.c index 81e349d5835b..20f810a6646c 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_dio_link_encoder.c +++ b/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_dio_link_encoder.c @@ -80,7 +80,6 @@ enum signal_type dcn35_get_dig_mode( default: return SIGNAL_TYPE_NONE; } - return SIGNAL_TYPE_NONE; } void dcn35_link_encoder_setup( @@ -119,7 +118,7 @@ void dcn35_link_encoder_setup( void dcn35_link_encoder_init(struct link_encoder *enc) { - enc32_hw_init(enc); + enc31_hw_init(enc); dcn35_link_encoder_set_fgcg(enc, enc->ctx->dc->debug.enable_fine_grain_clock_gating.bits.dio); } @@ -184,6 +183,8 @@ void dcn35_link_encoder_construct( enc10->base.hpd_source = init_data->hpd_source; enc10->base.connector = init_data->connector; + if (enc10->base.connector.id == CONNECTOR_ID_USBC) + enc10->base.features.flags.bits.DP_IS_USB_C = 1; enc10->base.preferred_engine = ENGINE_ID_UNKNOWN; @@ -238,8 +239,6 @@ void dcn35_link_encoder_construct( } enc10->base.features.flags.bits.HDMI_6GB_EN = 1; - if (enc10->base.connector.id == CONNECTOR_ID_USBC) - enc10->base.features.flags.bits.DP_IS_USB_C = 1; if (bp_funcs->get_connector_speed_cap_info) result = bp_funcs->get_connector_speed_cap_info(enc10->base.ctx->dc_bios, diff --git a/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_dio_link_encoder.h b/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_dio_link_encoder.h index e1e560732a9d..d546a3676304 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_dio_link_encoder.h +++ b/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_dio_link_encoder.h @@ -37,7 +37,9 @@ LE_SF(DIG0_DIG_BE_CLK_CNTL, DIG_BE_MODE, mask_sh),\ LE_SF(DIG0_DIG_BE_CLK_CNTL, DIG_BE_CLK_EN, mask_sh),\ LE_SF(DIG0_DIG_BE_CLK_CNTL, DIG_BE_SOFT_RESET, mask_sh),\ + LE_SF(DIG0_DIG_BE_CLK_CNTL, HDCP_SOFT_RESET, mask_sh),\ LE_SF(DIG0_DIG_BE_CLK_CNTL, DIG_BE_SYMCLK_G_CLOCK_ON, mask_sh),\ + LE_SF(DIG0_DIG_BE_CLK_CNTL, DIG_BE_SYMCLK_G_HDCP_CLOCK_ON, mask_sh),\ LE_SF(DIG0_DIG_BE_CLK_CNTL, DIG_BE_SYMCLK_G_TMDS_CLOCK_ON, mask_sh),\ LE_SF(DIG0_DIG_CLOCK_PATTERN, DIG_CLOCK_PATTERN, mask_sh),\ LE_SF(DIG0_TMDS_CTL_BITS, TMDS_CTL0, mask_sh), \ @@ -114,7 +116,15 @@ LE_SF(DIO_CLK_CNTL, SYMCLK_FE_G_GATE_DIS, mask_sh),\ LE_SF(DIO_CLK_CNTL, SYMCLK_R_GATE_DIS, mask_sh),\ LE_SF(DIO_CLK_CNTL, SYMCLK_G_GATE_DIS, mask_sh),\ - LE_SF(DIO_CLK_CNTL, DIO_FGCG_REP_DIS, mask_sh) + LE_SF(DIO_CLK_CNTL, DIO_FGCG_REP_DIS, mask_sh),\ + LE_SF(DIO_CLK_CNTL, DISPCLK_G_HDCP_GATE_DIS, mask_sh),\ + LE_SF(DIO_CLK_CNTL, SYMCLKA_G_HDCP_GATE_DIS, mask_sh),\ + LE_SF(DIO_CLK_CNTL, SYMCLKB_G_HDCP_GATE_DIS, mask_sh),\ + LE_SF(DIO_CLK_CNTL, SYMCLKC_G_HDCP_GATE_DIS, mask_sh),\ + LE_SF(DIO_CLK_CNTL, SYMCLKD_G_HDCP_GATE_DIS, mask_sh),\ + LE_SF(DIO_CLK_CNTL, SYMCLKE_G_HDCP_GATE_DIS, mask_sh),\ + LE_SF(DIO_CLK_CNTL, SYMCLKF_G_HDCP_GATE_DIS, mask_sh),\ + LE_SF(DIO_CLK_CNTL, SYMCLKG_G_HDCP_GATE_DIS, mask_sh) void dcn35_link_encoder_construct( diff --git a/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_dio_stream_encoder.h b/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_dio_stream_encoder.h index 499052329ebb..1212fcee38f2 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_dio_stream_encoder.h +++ b/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_dio_stream_encoder.h @@ -28,7 +28,6 @@ #include "dcn30/dcn30_vpg.h" #include "dcn30/dcn30_afmt.h" #include "stream_encoder.h" -#include "dcn10/dcn10_link_encoder.h" #include "dcn20/dcn20_stream_encoder.h" /* Register bit field name change */ diff --git a/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_hubbub.c b/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_hubbub.c index 339bf0c722dd..6293173ba2b9 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_hubbub.c +++ b/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_hubbub.c @@ -111,7 +111,7 @@ static uint32_t convert_and_clamp( static bool hubbub35_program_stutter_z8_watermarks( struct hubbub *hubbub, - struct dcn_watermark_set *watermarks, + union dcn_watermark_set *watermarks, unsigned int refclk_mhz, bool safe_to_lower) { @@ -297,7 +297,7 @@ static void hubbub35_get_dchub_ref_freq(struct hubbub *hubbub, static bool hubbub35_program_watermarks( struct hubbub *hubbub, - struct dcn_watermark_set *watermarks, + union dcn_watermark_set *watermarks, unsigned int refclk_mhz, bool safe_to_lower) { diff --git a/drivers/gpu/drm/amd/display/dc/dm_helpers.h b/drivers/gpu/drm/amd/display/dc/dm_helpers.h index 6d7a15dcf8a7..34adae7ab6e8 100644 --- a/drivers/gpu/drm/amd/display/dc/dm_helpers.h +++ b/drivers/gpu/drm/amd/display/dc/dm_helpers.h @@ -36,6 +36,7 @@ struct dc_dp_mst_stream_allocation_table; struct aux_payload; enum aux_return_code_type; +enum set_config_status; /* * Allocate memory accessible by the GPU @@ -200,7 +201,7 @@ int dm_helper_dmub_aux_transfer_sync( const struct dc_link *link, struct aux_payload *payload, enum aux_return_code_type *operation_result); -enum set_config_status; + int dm_helpers_dmub_set_config_sync(struct dc_context *ctx, const struct dc_link *link, struct set_config_cmd_payload *payload, diff --git a/drivers/gpu/drm/amd/display/dc/dm_services.h b/drivers/gpu/drm/amd/display/dc/dm_services.h index d0eed3b4771e..9405c47ee2a9 100644 --- a/drivers/gpu/drm/amd/display/dc/dm_services.h +++ b/drivers/gpu/drm/amd/display/dc/dm_services.h @@ -275,6 +275,16 @@ void dm_perf_trace_timestamp(const char *func_name, unsigned int line, struct dc #define PERF_TRACE_CTX(__CTX) dm_perf_trace_timestamp(__func__, __LINE__, __CTX) /* + * SMU message tracing + */ +void dm_trace_smu_msg(uint32_t msg_id, uint32_t param_in, struct dc_context *ctx); +void dm_trace_smu_delay(uint32_t delay, struct dc_context *ctx); + +#define TRACE_SMU_MSG(msg_id, param_in, ctx) dm_trace_smu_msg(msg_id, param_in, ctx) +#define TRACE_SMU_DELAY(response_delay, ctx) dm_trace_smu_delay(response_delay, ctx) + + +/* * DMUB Interfaces */ bool dm_execute_dmub_cmd(const struct dc_context *ctx, union dmub_rb_cmd *cmd, enum dm_dmub_wait_type wait_type); diff --git a/drivers/gpu/drm/amd/display/dc/dml/Makefile b/drivers/gpu/drm/amd/display/dc/dml/Makefile index c4a5efd2dda5..a94b6d546cd1 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/Makefile +++ b/drivers/gpu/drm/amd/display/dc/dml/Makefile @@ -25,40 +25,8 @@ # It provides the general basic services required by other DAL # subcomponents. -ifdef CONFIG_X86 -dml_ccflags-$(CONFIG_CC_IS_GCC) := -mhard-float -dml_ccflags := $(dml_ccflags-y) -msse -endif - -ifdef CONFIG_PPC64 -dml_ccflags := -mhard-float -maltivec -endif - -ifdef CONFIG_ARM64 -dml_rcflags := -mgeneral-regs-only -endif - -ifdef CONFIG_LOONGARCH -dml_ccflags := -mfpu=64 -dml_rcflags := -msoft-float -endif - -ifdef CONFIG_CC_IS_GCC -ifneq ($(call gcc-min-version, 70100),y) -IS_OLD_GCC = 1 -endif -endif - -ifdef CONFIG_X86 -ifdef IS_OLD_GCC -# Stack alignment mismatch, proceed with caution. -# GCC < 7.1 cannot compile code using `double` and -mpreferred-stack-boundary=3 -# (8B stack alignment). -dml_ccflags += -mpreferred-stack-boundary=4 -else -dml_ccflags += -msse2 -endif -endif +dml_ccflags := $(CC_FLAGS_FPU) +dml_rcflags := $(CC_FLAGS_NO_FPU) ifneq ($(CONFIG_FRAME_WARN),0) ifeq ($(filter y,$(CONFIG_KASAN)$(CONFIG_KCSAN)),y) diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c index 38ab9ad60ef8..74da9ebda016 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c @@ -1085,6 +1085,9 @@ static enum dcn_zstate_support_state decide_zstate_support(struct dc *dc, struc int minmum_z8_residency = dc->debug.minimum_z8_residency_time > 0 ? dc->debug.minimum_z8_residency_time : 1000; bool allow_z8 = context->bw_ctx.dml.vba.StutterPeriod > (double)minmum_z8_residency; bool is_pwrseq0 = link->link_index == 0; + bool is_psr = (link && (link->psr_settings.psr_version == DC_PSR_VERSION_1 || + link->psr_settings.psr_version == DC_PSR_VERSION_SU_1) && !link->panel_config.psr.disable_psr); + bool is_replay = link && link->replay_settings.replay_feature_enabled; /* Don't support multi-plane configurations */ if (stream_status->plane_count > 1) @@ -1092,8 +1095,8 @@ static enum dcn_zstate_support_state decide_zstate_support(struct dc *dc, struc if (is_pwrseq0 && context->bw_ctx.dml.vba.StutterPeriod > 5000.0) return DCN_ZSTATE_SUPPORT_ALLOW; - else if (is_pwrseq0 && link->psr_settings.psr_version == DC_PSR_VERSION_1 && !link->panel_config.psr.disable_psr) - return allow_z8 ? DCN_ZSTATE_SUPPORT_ALLOW_Z8_Z10_ONLY : DCN_ZSTATE_SUPPORT_ALLOW_Z10_ONLY; + else if (is_pwrseq0 && (is_psr || is_replay)) + return DCN_ZSTATE_SUPPORT_ALLOW_Z8_Z10_ONLY; else return allow_z8 ? DCN_ZSTATE_SUPPORT_ALLOW_Z8_ONLY : DCN_ZSTATE_SUPPORT_DISALLOW; } else { @@ -2369,7 +2372,7 @@ validate_out: static struct _vcs_dpi_voltage_scaling_st construct_low_pstate_lvl(struct clk_limit_table *clk_table, unsigned int high_voltage_lvl) { - struct _vcs_dpi_voltage_scaling_st low_pstate_lvl; + struct _vcs_dpi_voltage_scaling_st low_pstate_lvl = {0}; int i; low_pstate_lvl.state = 1; @@ -2474,7 +2477,7 @@ void dcn201_populate_dml_writeback_from_context_fpu(struct dc *dc, int pipe_cnt, i, j; double max_calc_writeback_dispclk; double writeback_dispclk; - struct writeback_st dout_wb; + struct writeback_st dout_wb = {0}; dc_assert_fp_enabled(); diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn30/dcn30_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn30/dcn30_fpu.c index ccb4ad78f667..81f7b90849ce 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn30/dcn30_fpu.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn30/dcn30_fpu.c @@ -260,7 +260,7 @@ void dcn30_fpu_populate_dml_writeback_from_context( int pipe_cnt, i, j; double max_calc_writeback_dispclk; double writeback_dispclk; - struct writeback_st dout_wb; + struct writeback_st dout_wb = {0}; dc_assert_fp_enabled(); diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn30/display_mode_vba_30.c b/drivers/gpu/drm/amd/display/dc/dml/dcn30/display_mode_vba_30.c index e7f4a2d491cc..e0b52db2c210 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn30/display_mode_vba_30.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn30/display_mode_vba_30.c @@ -3535,7 +3535,6 @@ static double TruncToValidBPP( return DesiredBPP; } } - return BPP_INVALID; } void dml30_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_lib) diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.c index deb6d162a2d5..94317b2e4a85 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.c @@ -291,6 +291,7 @@ static struct _vcs_dpi_soc_bounding_box_st dcn3_15_soc = { .do_urgent_latency_adjustment = false, .urgent_latency_adjustment_fabric_clock_component_us = 0, .urgent_latency_adjustment_fabric_clock_reference_mhz = 0, + .dispclk_dppclk_vco_speed_mhz = 2400.0, .num_chans = 4, .dummy_pstate_latency_us = 10.0 }; @@ -438,6 +439,7 @@ static struct _vcs_dpi_soc_bounding_box_st dcn3_16_soc = { .do_urgent_latency_adjustment = false, .urgent_latency_adjustment_fabric_clock_component_us = 0, .urgent_latency_adjustment_fabric_clock_reference_mhz = 0, + .dispclk_dppclk_vco_speed_mhz = 2500.0, }; void dcn31_zero_pipe_dcc_fraction(display_e2e_pipe_params_st *pipes, @@ -485,6 +487,7 @@ void dcn31_calculate_wm_and_dlg_fp( { int i, pipe_idx, total_det = 0, active_hubp_count = 0; double dcfclk = context->bw_ctx.dml.vba.DCFCLKState[vlevel][context->bw_ctx.dml.vba.maxMpcComb]; + uint32_t cstate_enter_plus_exit_z8_ns; dc_assert_fp_enabled(); @@ -504,6 +507,13 @@ void dcn31_calculate_wm_and_dlg_fp( pipes[0].clks_cfg.dcfclk_mhz = dcfclk; pipes[0].clks_cfg.socclk_mhz = context->bw_ctx.dml.soc.clock_limits[vlevel].socclk_mhz; + cstate_enter_plus_exit_z8_ns = + get_wm_z8_stutter_enter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + + if (get_stutter_period(&context->bw_ctx.dml, pipes, pipe_cnt) < dc->debug.minimum_z8_residency_time && + cstate_enter_plus_exit_z8_ns < dc->debug.minimum_z8_residency_time * 1000) + cstate_enter_plus_exit_z8_ns = dc->debug.minimum_z8_residency_time * 1000; + /* Set A: * All clocks min required * @@ -514,7 +524,7 @@ void dcn31_calculate_wm_and_dlg_fp( context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.cstate_enter_plus_exit_ns = get_wm_stutter_enter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.cstate_exit_ns = get_wm_stutter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.pstate_change_ns = get_wm_dram_clock_change(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.cstate_enter_plus_exit_z8_ns = get_wm_z8_stutter_enter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.cstate_enter_plus_exit_z8_ns = cstate_enter_plus_exit_z8_ns; context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.cstate_exit_z8_ns = get_wm_z8_stutter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; context->bw_ctx.bw.dcn.watermarks.a.pte_meta_urgent_ns = get_wm_memory_trip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; context->bw_ctx.bw.dcn.watermarks.a.frac_urg_bw_nom = get_fraction_of_urgent_bandwidth(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.h b/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.h index 8f9c8faed260..d2ae43a82ba5 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.h +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.h @@ -30,6 +30,7 @@ #define DCN3_15_DEFAULT_DET_SIZE 192 #define DCN3_15_MIN_COMPBUF_SIZE_KB 128 #define DCN3_16_DEFAULT_DET_SIZE 192 +#define DCN3_16_MIN_COMPBUF_SIZE_KB 128 void dcn31_zero_pipe_dcc_fraction(display_e2e_pipe_params_st *pipes, int pipe_cnt); diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn31/display_mode_vba_31.c b/drivers/gpu/drm/amd/display/dc/dml/dcn31/display_mode_vba_31.c index adea459e7d36..33cf824c5da1 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn31/display_mode_vba_31.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn31/display_mode_vba_31.c @@ -3679,7 +3679,6 @@ static double TruncToValidBPP( return DesiredBPP; } } - return BPP_INVALID; } static noinline void CalculatePrefetchSchedulePerPlane( diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn314/dcn314_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn314/dcn314_fpu.c index fb21572750e8..21f637ae4add 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn314/dcn314_fpu.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn314/dcn314_fpu.c @@ -310,7 +310,7 @@ int dcn314_populate_dml_pipes_from_context_fpu(struct dc *dc, struct dc_state *c { int i, pipe_cnt; struct resource_context *res_ctx = &context->res_ctx; - struct pipe_ctx *pipe; + struct pipe_ctx *pipe = 0; bool upscaled = false; const unsigned int max_allowed_vblank_nom = 1023; diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn314/display_mode_vba_314.c b/drivers/gpu/drm/amd/display/dc/dml/dcn314/display_mode_vba_314.c index 88e56889a68c..3242957d00c5 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn314/display_mode_vba_314.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn314/display_mode_vba_314.c @@ -3788,7 +3788,6 @@ static double TruncToValidBPP( return DesiredBPP; } } - return BPP_INVALID; } static noinline void CalculatePrefetchSchedulePerPlane( diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c index a0a65e099104..f6fe0a64beac 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c @@ -180,6 +180,9 @@ struct _vcs_dpi_soc_bounding_box_st dcn3_2_soc = { .urgent_latency_adjustment_fabric_clock_reference_mhz = 3000, }; +static bool dcn32_apply_merge_split_flags_helper(struct dc *dc, struct dc_state *context, + bool *repopulate_pipes, int *split, bool *merge); + void dcn32_build_wm_range_table_fpu(struct clk_mgr_internal *clk_mgr) { /* defaults */ @@ -622,7 +625,7 @@ static bool dcn32_assign_subvp_pipe(struct dc *dc, * to combine this with SubVP can cause issues with the scheduling). * - Not TMZ surface */ - if (pipe->plane_state && !pipe->top_pipe && !dcn32_is_center_timing(pipe) && + if (pipe->plane_state && !pipe->top_pipe && !pipe->prev_odm_pipe && !dcn32_is_center_timing(pipe) && !(pipe->stream->timing.pix_clk_100hz / 10000 > DCN3_2_MAX_SUBVP_PIXEL_RATE_MHZ) && (!dcn32_is_psr_capable(pipe) || (context->stream_count == 1 && dc->caps.dmub_caps.subvp_psr)) && dc_state_get_pipe_subvp_type(context, pipe) == SUBVP_NONE && @@ -720,7 +723,7 @@ static bool dcn32_enough_pipes_for_subvp(struct dc *dc, struct dc_state *context */ static bool subvp_subvp_schedulable(struct dc *dc, struct dc_state *context) { - struct pipe_ctx *subvp_pipes[2]; + struct pipe_ctx *subvp_pipes[2] = {0}; struct dc_stream_state *phantom = NULL; uint32_t microschedule_lines = 0; uint32_t index = 0; @@ -1425,13 +1428,14 @@ static bool is_test_pattern_enabled( return false; } -static void dcn32_full_validate_bw_helper(struct dc *dc, +static bool dcn32_full_validate_bw_helper(struct dc *dc, struct dc_state *context, display_e2e_pipe_params_st *pipes, int *vlevel, int *split, bool *merge, - int *pipe_cnt) + int *pipe_cnt, + bool *repopulate_pipes) { struct vba_vars_st *vba = &context->bw_ctx.dml.vba; unsigned int dc_pipe_idx = 0; @@ -1461,6 +1465,12 @@ static void dcn32_full_validate_bw_helper(struct dc *dc, vba->VoltageLevel = *vlevel; } + /* Apply split and merge flags before checking for subvp */ + if (!dcn32_apply_merge_split_flags_helper(dc, context, repopulate_pipes, split, merge)) + return false; + memset(split, 0, MAX_PIPES * sizeof(int)); + memset(merge, 0, MAX_PIPES * sizeof(bool)); + /* Conditions for setting up phantom pipes for SubVP: * 1. Not force disable SubVP * 2. Full update (i.e. !fast_validate) @@ -1475,19 +1485,7 @@ static void dcn32_full_validate_bw_helper(struct dc *dc, vba->DRAMClockChangeSupport[*vlevel][vba->maxMpcComb] == dm_dram_clock_change_unsupported || dc->debug.force_subvp_mclk_switch)) { - dcn32_merge_pipes_for_subvp(dc, context); - memset(merge, 0, MAX_PIPES * sizeof(bool)); - vlevel_temp = *vlevel; - /* to re-initialize viewport after the pipe merge */ - for (i = 0; i < dc->res_pool->pipe_count; i++) { - struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i]; - - if (!pipe_ctx->plane_state || !pipe_ctx->stream) - continue; - - resource_build_scaling_params(pipe_ctx); - } while (!found_supported_config && dcn32_enough_pipes_for_subvp(dc, context) && dcn32_assign_subvp_pipe(dc, context, &dc_pipe_idx)) { @@ -1576,8 +1574,6 @@ static void dcn32_full_validate_bw_helper(struct dc *dc, * add phantom pipes. If pipe split (ODM / MPC) is required, both the main * and phantom pipes will be split in the regular pipe splitting sequence. */ - memset(split, 0, MAX_PIPES * sizeof(int)); - memset(merge, 0, MAX_PIPES * sizeof(bool)); *vlevel = dcn20_validate_apply_pipe_split_flags(dc, context, *vlevel, split, merge); vba->VoltageLevel = *vlevel; // Note: We can't apply the phantom pipes to hardware at this time. We have to wait @@ -1590,6 +1586,7 @@ static void dcn32_full_validate_bw_helper(struct dc *dc, try_odm_power_optimization_and_revalidate( dc, context, pipes, split, merge, vlevel, *pipe_cnt); + return true; } static bool is_dtbclk_required(struct dc *dc, struct dc_state *context) @@ -1929,106 +1926,23 @@ static bool dcn32_split_stream_for_mpc_or_odm( return true; } -bool dcn32_internal_validate_bw(struct dc *dc, - struct dc_state *context, - display_e2e_pipe_params_st *pipes, - int *pipe_cnt_out, - int *vlevel_out, - bool fast_validate) +static bool dcn32_apply_merge_split_flags_helper( + struct dc *dc, + struct dc_state *context, + bool *repopulate_pipes, + int *split, + bool *merge) { - bool out = false; - bool repopulate_pipes = false; - int split[MAX_PIPES] = { 0 }; - bool merge[MAX_PIPES] = { false }; + int i, pipe_idx; bool newly_split[MAX_PIPES] = { false }; - int pipe_cnt, i, pipe_idx; - int vlevel = context->bw_ctx.dml.soc.num_states; struct vba_vars_st *vba = &context->bw_ctx.dml.vba; - dc_assert_fp_enabled(); - - ASSERT(pipes); - if (!pipes) - return false; - - // For each full update, remove all existing phantom pipes first - dc_state_remove_phantom_streams_and_planes(dc, context); - dc_state_release_phantom_streams_and_planes(dc, context); - - dc->res_pool->funcs->update_soc_for_wm_a(dc, context); - - pipe_cnt = dc->res_pool->funcs->populate_dml_pipes(dc, context, pipes, fast_validate); - - if (!pipe_cnt) { - out = true; - goto validate_out; - } - - dml_log_pipe_params(&context->bw_ctx.dml, pipes, pipe_cnt); - context->bw_ctx.dml.soc.max_vratio_pre = dcn32_determine_max_vratio_prefetch(dc, context); - - if (!fast_validate) - dcn32_full_validate_bw_helper(dc, context, pipes, &vlevel, split, merge, &pipe_cnt); - - if (fast_validate || - (dc->debug.dml_disallow_alternate_prefetch_modes && - (vlevel == context->bw_ctx.dml.soc.num_states || - vba->DRAMClockChangeSupport[vlevel][vba->maxMpcComb] == dm_dram_clock_change_unsupported))) { - /* - * If dml_disallow_alternate_prefetch_modes is false, then we have already - * tried alternate prefetch modes during full validation. - * - * If mode is unsupported or there is no p-state support, then - * fall back to favouring voltage. - * - * If Prefetch mode 0 failed for this config, or passed with Max UCLK, then try - * to support with Prefetch mode 1 (dm_prefetch_support_fclk_and_stutter == 2) - */ - context->bw_ctx.dml.soc.allow_for_pstate_or_stutter_in_vblank_final = - dm_prefetch_support_none; - - context->bw_ctx.dml.validate_max_state = fast_validate; - vlevel = dml_get_voltage_level(&context->bw_ctx.dml, pipes, pipe_cnt); - - context->bw_ctx.dml.validate_max_state = false; - - if (vlevel < context->bw_ctx.dml.soc.num_states) { - memset(split, 0, sizeof(split)); - memset(merge, 0, sizeof(merge)); - vlevel = dcn20_validate_apply_pipe_split_flags(dc, context, vlevel, split, merge); - // dcn20_validate_apply_pipe_split_flags can modify voltage level outside of DML - vba->VoltageLevel = vlevel; - } - } - - dml_log_mode_support_params(&context->bw_ctx.dml); - - if (vlevel == context->bw_ctx.dml.soc.num_states) - goto validate_fail; - - for (i = 0, pipe_idx = 0; i < dc->res_pool->pipe_count; i++) { - struct pipe_ctx *pipe = &context->res_ctx.pipe_ctx[i]; - struct pipe_ctx *mpo_pipe = pipe->bottom_pipe; - - if (!pipe->stream) - continue; - - if (vba->ODMCombineEnabled[vba->pipe_plane[pipe_idx]] != dm_odm_combine_mode_disabled - && !dc->config.enable_windowed_mpo_odm - && pipe->plane_state && mpo_pipe - && memcmp(&mpo_pipe->plane_state->clip_rect, - &pipe->stream->src, - sizeof(struct rect)) != 0) { - ASSERT(mpo_pipe->plane_state != pipe->plane_state); - goto validate_fail; - } - pipe_idx++; - } - if (dc->config.enable_windowed_mpo_odm) { - repopulate_pipes = update_pipes_with_split_flags( - dc, context, vba, split, merge); + if (update_pipes_with_split_flags( + dc, context, vba, split, merge)) + *repopulate_pipes = true; } else { + /* the code below will be removed once windowed mpo odm is fully * enabled. */ @@ -2085,7 +1999,7 @@ bool dcn32_internal_validate_bw(struct dc *dc, memset(&pipe->plane_res, 0, sizeof(pipe->plane_res)); memset(&pipe->stream_res, 0, sizeof(pipe->stream_res)); memset(&pipe->link_res, 0, sizeof(pipe->link_res)); - repopulate_pipes = true; + *repopulate_pipes = true; } else if (pipe->top_pipe && pipe->top_pipe->plane_state == pipe->plane_state) { struct pipe_ctx *top_pipe = pipe->top_pipe; struct pipe_ctx *bottom_pipe = pipe->bottom_pipe; @@ -2101,7 +2015,7 @@ bool dcn32_internal_validate_bw(struct dc *dc, memset(&pipe->plane_res, 0, sizeof(pipe->plane_res)); memset(&pipe->stream_res, 0, sizeof(pipe->stream_res)); memset(&pipe->link_res, 0, sizeof(pipe->link_res)); - repopulate_pipes = true; + *repopulate_pipes = true; } else ASSERT(0); /* Should never try to merge master pipe */ @@ -2140,15 +2054,15 @@ bool dcn32_internal_validate_bw(struct dc *dc, hsplit_pipe = dcn32_find_split_pipe(dc, context, old_index); ASSERT(hsplit_pipe); if (!hsplit_pipe) - goto validate_fail; + return false; if (!dcn32_split_stream_for_mpc_or_odm( dc, &context->res_ctx, pipe, hsplit_pipe, odm)) - goto validate_fail; + return false; newly_split[hsplit_pipe->pipe_idx] = true; - repopulate_pipes = true; + *repopulate_pipes = true; } if (split[i] == 4) { struct pipe_ctx *pipe_4to1; @@ -2163,11 +2077,11 @@ bool dcn32_internal_validate_bw(struct dc *dc, pipe_4to1 = dcn32_find_split_pipe(dc, context, old_index); ASSERT(pipe_4to1); if (!pipe_4to1) - goto validate_fail; + return false; if (!dcn32_split_stream_for_mpc_or_odm( dc, &context->res_ctx, pipe, pipe_4to1, odm)) - goto validate_fail; + return false; newly_split[pipe_4to1->pipe_idx] = true; if (odm && old_pipe->next_odm_pipe && old_pipe->next_odm_pipe->next_odm_pipe @@ -2182,11 +2096,11 @@ bool dcn32_internal_validate_bw(struct dc *dc, pipe_4to1 = dcn32_find_split_pipe(dc, context, old_index); ASSERT(pipe_4to1); if (!pipe_4to1) - goto validate_fail; + return false; if (!dcn32_split_stream_for_mpc_or_odm( dc, &context->res_ctx, hsplit_pipe, pipe_4to1, odm)) - goto validate_fail; + return false; newly_split[pipe_4to1->pipe_idx] = true; } if (odm) @@ -2198,11 +2112,122 @@ bool dcn32_internal_validate_bw(struct dc *dc, if (pipe->plane_state) { if (!resource_build_scaling_params(pipe)) - goto validate_fail; + return false; } } + + for (i = 0; i < context->stream_count; i++) { + struct pipe_ctx *otg_master = resource_get_otg_master_for_stream(&context->res_ctx, + context->streams[i]); + + if (otg_master) + resource_build_test_pattern_params(&context->res_ctx, otg_master); + } + } + return true; +} + +bool dcn32_internal_validate_bw(struct dc *dc, + struct dc_state *context, + display_e2e_pipe_params_st *pipes, + int *pipe_cnt_out, + int *vlevel_out, + bool fast_validate) +{ + bool out = false; + bool repopulate_pipes = false; + int split[MAX_PIPES] = { 0 }; + bool merge[MAX_PIPES] = { false }; + int pipe_cnt, i, pipe_idx; + int vlevel = context->bw_ctx.dml.soc.num_states; + struct vba_vars_st *vba = &context->bw_ctx.dml.vba; + + dc_assert_fp_enabled(); + + ASSERT(pipes); + if (!pipes) + return false; + + /* For each full update, remove all existing phantom pipes first */ + dc_state_remove_phantom_streams_and_planes(dc, context); + dc_state_release_phantom_streams_and_planes(dc, context); + + dc->res_pool->funcs->update_soc_for_wm_a(dc, context); + + pipe_cnt = dc->res_pool->funcs->populate_dml_pipes(dc, context, pipes, fast_validate); + + if (!pipe_cnt) { + out = true; + goto validate_out; + } + + dml_log_pipe_params(&context->bw_ctx.dml, pipes, pipe_cnt); + context->bw_ctx.dml.soc.max_vratio_pre = dcn32_determine_max_vratio_prefetch(dc, context); + + if (!fast_validate) { + if (!dcn32_full_validate_bw_helper(dc, context, pipes, &vlevel, split, merge, + &pipe_cnt, &repopulate_pipes)) + goto validate_fail; + } + + if (fast_validate || + (dc->debug.dml_disallow_alternate_prefetch_modes && + (vlevel == context->bw_ctx.dml.soc.num_states || + vba->DRAMClockChangeSupport[vlevel][vba->maxMpcComb] == dm_dram_clock_change_unsupported))) { + /* + * If dml_disallow_alternate_prefetch_modes is false, then we have already + * tried alternate prefetch modes during full validation. + * + * If mode is unsupported or there is no p-state support, then + * fall back to favouring voltage. + * + * If Prefetch mode 0 failed for this config, or passed with Max UCLK, then try + * to support with Prefetch mode 1 (dm_prefetch_support_fclk_and_stutter == 2) + */ + context->bw_ctx.dml.soc.allow_for_pstate_or_stutter_in_vblank_final = + dm_prefetch_support_none; + + context->bw_ctx.dml.validate_max_state = fast_validate; + vlevel = dml_get_voltage_level(&context->bw_ctx.dml, pipes, pipe_cnt); + + context->bw_ctx.dml.validate_max_state = false; + + if (vlevel < context->bw_ctx.dml.soc.num_states) { + memset(split, 0, sizeof(split)); + memset(merge, 0, sizeof(merge)); + vlevel = dcn20_validate_apply_pipe_split_flags(dc, context, vlevel, split, merge); + /* dcn20_validate_apply_pipe_split_flags can modify voltage level outside of DML */ + vba->VoltageLevel = vlevel; + } } + dml_log_mode_support_params(&context->bw_ctx.dml); + + if (vlevel == context->bw_ctx.dml.soc.num_states) + goto validate_fail; + + for (i = 0, pipe_idx = 0; i < dc->res_pool->pipe_count; i++) { + struct pipe_ctx *pipe = &context->res_ctx.pipe_ctx[i]; + struct pipe_ctx *mpo_pipe = pipe->bottom_pipe; + + if (!pipe->stream) + continue; + + if (vba->ODMCombineEnabled[vba->pipe_plane[pipe_idx]] != dm_odm_combine_mode_disabled + && !dc->config.enable_windowed_mpo_odm + && pipe->plane_state && mpo_pipe + && memcmp(&mpo_pipe->plane_state->clip_rect, + &pipe->stream->src, + sizeof(struct rect)) != 0) { + ASSERT(mpo_pipe->plane_state != pipe->plane_state); + goto validate_fail; + } + pipe_idx++; + } + + if (!dcn32_apply_merge_split_flags_helper(dc, context, &repopulate_pipes, split, merge)) + goto validate_fail; + /* Actual dsc count per stream dsc validation*/ if (!dcn20_validate_dsc(dc, context)) { vba->ValidationStatus[vba->soc.num_states] = DML_FAIL_DSC_VALIDATION_FAILURE; diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_util_32.c b/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_util_32.c index 80fccd4999a5..ba1310c8fd77 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_util_32.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_util_32.c @@ -1650,6 +1650,8 @@ double dml32_TruncToValidBPP( MaxLinkBPP = 2 * MaxLinkBPP; } + *RequiredSlots = dml_ceil(DesiredBPP / MaxLinkBPP * 64, 1); + if (DesiredBPP == 0) { if (DSCEnable) { if (MaxLinkBPP < MinDSCBPP) @@ -1676,10 +1678,6 @@ double dml32_TruncToValidBPP( else return DesiredBPP; } - - *RequiredSlots = dml_ceil(DesiredBPP / MaxLinkBPP * 64, 1); - - return BPP_INVALID; } // TruncToValidBPP double dml32_RequiredDTBCLK( @@ -1975,8 +1973,8 @@ void dml32_CalculateVMRowAndSwath( unsigned int PTEBufferSizeInRequestsForChroma[DC__NUM_DPP__MAX]; unsigned int PDEAndMetaPTEBytesFrameY; unsigned int PDEAndMetaPTEBytesFrameC; - unsigned int MetaRowByteY[DC__NUM_DPP__MAX]; - unsigned int MetaRowByteC[DC__NUM_DPP__MAX]; + unsigned int MetaRowByteY[DC__NUM_DPP__MAX] = {0}; + unsigned int MetaRowByteC[DC__NUM_DPP__MAX] = {0}; unsigned int PixelPTEBytesPerRowY[DC__NUM_DPP__MAX]; unsigned int PixelPTEBytesPerRowC[DC__NUM_DPP__MAX]; unsigned int PixelPTEBytesPerRowY_one_row_per_frame[DC__NUM_DPP__MAX]; @@ -4291,7 +4289,7 @@ void dml32_CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport( unsigned int i, j, k; unsigned int SurfaceWithMinActiveFCLKChangeMargin = 0; unsigned int DRAMClockChangeSupportNumber = 0; - unsigned int LastSurfaceWithoutMargin; + unsigned int LastSurfaceWithoutMargin = 0; unsigned int DRAMClockChangeMethod = 0; bool FoundFirstSurfaceWithMinActiveFCLKChangeMargin = false; double MinActiveFCLKChangeMargin = 0.; @@ -5656,9 +5654,9 @@ void dml32_CalculateStutterEfficiency( double LastZ8StutterPeriod = 0.0; double LastStutterPeriod = 0.0; unsigned int TotalNumberOfActiveOTG = 0; - double doublePixelClock; - unsigned int doubleHTotal; - unsigned int doubleVTotal; + double doublePixelClock = 0; + unsigned int doubleHTotal = 0; + unsigned int doubleVTotal = 0; bool SameTiming = true; double DETBufferingTimeY; double SwathWidthYCriticalSurface = 0.0; diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn35/dcn35_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn35/dcn35_fpu.c index 80bebfc268db..60f251cf973b 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn35/dcn35_fpu.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn35/dcn35_fpu.c @@ -166,8 +166,8 @@ struct _vcs_dpi_soc_bounding_box_st dcn3_5_soc = { .num_states = 5, .sr_exit_time_us = 28.0, .sr_enter_plus_exit_time_us = 30.0, - .sr_exit_z8_time_us = 210.0, - .sr_enter_plus_exit_z8_time_us = 320.0, + .sr_exit_z8_time_us = 250.0, + .sr_enter_plus_exit_z8_time_us = 350.0, .fclk_change_latency_us = 24.0, .usr_retraining_latency_us = 2, .writeback_latency_us = 12.0, @@ -195,9 +195,9 @@ struct _vcs_dpi_soc_bounding_box_st dcn3_5_soc = { .dcn_downspread_percent = 0.5, .gpuvm_min_page_size_bytes = 4096, .hostvm_min_page_size_bytes = 4096, - .do_urgent_latency_adjustment = 0, + .do_urgent_latency_adjustment = 1, .urgent_latency_adjustment_fabric_clock_component_us = 0, - .urgent_latency_adjustment_fabric_clock_reference_mhz = 0, + .urgent_latency_adjustment_fabric_clock_reference_mhz = 3000, }; void dcn35_build_wm_range_table_fpu(struct clk_mgr *clk_mgr) @@ -439,7 +439,7 @@ int dcn35_populate_dml_pipes_from_context_fpu(struct dc *dc, { int i, pipe_cnt; struct resource_context *res_ctx = &context->res_ctx; - struct pipe_ctx *pipe; + struct pipe_ctx *pipe = 0; bool upscaled = false; const unsigned int max_allowed_vblank_nom = 1023; @@ -577,6 +577,7 @@ void dcn35_decide_zstate_support(struct dc *dc, struct dc_state *context) { enum dcn_zstate_support_state support = DCN_ZSTATE_SUPPORT_DISALLOW; unsigned int i, plane_count = 0; + DC_LOGGER_INIT(dc->ctx->logger); for (i = 0; i < dc->res_pool->pipe_count; i++) { if (context->res_ctx.pipe_ctx[i].plane_state) @@ -602,11 +603,14 @@ void dcn35_decide_zstate_support(struct dc *dc, struct dc_state *context) if (is_pwrseq0 && allow_z10) support = DCN_ZSTATE_SUPPORT_ALLOW; else if (is_pwrseq0 && (is_psr || is_replay)) - support = allow_z8 ? DCN_ZSTATE_SUPPORT_ALLOW_Z8_Z10_ONLY : DCN_ZSTATE_SUPPORT_ALLOW_Z10_ONLY; + support = DCN_ZSTATE_SUPPORT_ALLOW_Z8_Z10_ONLY; else if (allow_z8) support = DCN_ZSTATE_SUPPORT_ALLOW_Z8_ONLY; } + DC_LOG_SMU("zstate_support: %d, StutterPeriod: %d\n", support, + (int)context->bw_ctx.dml.vba.StutterPeriod); + context->bw_ctx.bw.dcn.clk.zstate_support = support; } diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn351/dcn351_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn351/dcn351_fpu.c index dc9e1b758ed6..e4f333d4fb54 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn351/dcn351_fpu.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn351/dcn351_fpu.c @@ -98,55 +98,114 @@ struct _vcs_dpi_soc_bounding_box_st dcn3_51_soc = { .clock_limits = { { .state = 0, - .dispclk_mhz = 1200.0, - .dppclk_mhz = 1200.0, + .dcfclk_mhz = 400.0, + .fabricclk_mhz = 400.0, + .socclk_mhz = 600.0, + .dram_speed_mts = 3200.0, + .dispclk_mhz = 600.0, + .dppclk_mhz = 600.0, .phyclk_mhz = 600.0, .phyclk_d18_mhz = 667.0, - .dscclk_mhz = 186.0, + .dscclk_mhz = 200.0, .dtbclk_mhz = 600.0, }, { .state = 1, - .dispclk_mhz = 1200.0, - .dppclk_mhz = 1200.0, + .dcfclk_mhz = 600.0, + .fabricclk_mhz = 1000.0, + .socclk_mhz = 733.0, + .dram_speed_mts = 6400.0, + .dispclk_mhz = 800.0, + .dppclk_mhz = 800.0, .phyclk_mhz = 810.0, .phyclk_d18_mhz = 667.0, - .dscclk_mhz = 209.0, + .dscclk_mhz = 266.7, .dtbclk_mhz = 600.0, }, { .state = 2, - .dispclk_mhz = 1200.0, - .dppclk_mhz = 1200.0, + .dcfclk_mhz = 738.0, + .fabricclk_mhz = 1200.0, + .socclk_mhz = 880.0, + .dram_speed_mts = 7500.0, + .dispclk_mhz = 800.0, + .dppclk_mhz = 800.0, .phyclk_mhz = 810.0, .phyclk_d18_mhz = 667.0, - .dscclk_mhz = 209.0, + .dscclk_mhz = 266.7, .dtbclk_mhz = 600.0, }, { .state = 3, - .dispclk_mhz = 1200.0, - .dppclk_mhz = 1200.0, + .dcfclk_mhz = 800.0, + .fabricclk_mhz = 1400.0, + .socclk_mhz = 978.0, + .dram_speed_mts = 7500.0, + .dispclk_mhz = 960.0, + .dppclk_mhz = 960.0, .phyclk_mhz = 810.0, .phyclk_d18_mhz = 667.0, - .dscclk_mhz = 371.0, + .dscclk_mhz = 320.0, .dtbclk_mhz = 600.0, }, { .state = 4, + .dcfclk_mhz = 873.0, + .fabricclk_mhz = 1600.0, + .socclk_mhz = 1100.0, + .dram_speed_mts = 8533.0, + .dispclk_mhz = 1066.7, + .dppclk_mhz = 1066.7, + .phyclk_mhz = 810.0, + .phyclk_d18_mhz = 667.0, + .dscclk_mhz = 355.6, + .dtbclk_mhz = 600.0, + }, + { + .state = 5, + .dcfclk_mhz = 960.0, + .fabricclk_mhz = 1700.0, + .socclk_mhz = 1257.0, + .dram_speed_mts = 8533.0, .dispclk_mhz = 1200.0, .dppclk_mhz = 1200.0, .phyclk_mhz = 810.0, .phyclk_d18_mhz = 667.0, - .dscclk_mhz = 417.0, + .dscclk_mhz = 400.0, + .dtbclk_mhz = 600.0, + }, + { + .state = 6, + .dcfclk_mhz = 1067.0, + .fabricclk_mhz = 1850.0, + .socclk_mhz = 1257.0, + .dram_speed_mts = 8533.0, + .dispclk_mhz = 1371.4, + .dppclk_mhz = 1371.4, + .phyclk_mhz = 810.0, + .phyclk_d18_mhz = 667.0, + .dscclk_mhz = 457.1, + .dtbclk_mhz = 600.0, + }, + { + .state = 7, + .dcfclk_mhz = 1200.0, + .fabricclk_mhz = 2000.0, + .socclk_mhz = 1467.0, + .dram_speed_mts = 8533.0, + .dispclk_mhz = 1600.0, + .dppclk_mhz = 1600.0, + .phyclk_mhz = 810.0, + .phyclk_d18_mhz = 667.0, + .dscclk_mhz = 533.3, .dtbclk_mhz = 600.0, }, }, - .num_states = 5, + .num_states = 8, .sr_exit_time_us = 28.0, .sr_enter_plus_exit_time_us = 30.0, - .sr_exit_z8_time_us = 210.0, - .sr_enter_plus_exit_z8_time_us = 320.0, + .sr_exit_z8_time_us = 250.0, + .sr_enter_plus_exit_z8_time_us = 350.0, .fclk_change_latency_us = 24.0, .usr_retraining_latency_us = 2, .writeback_latency_us = 12.0, @@ -177,6 +236,8 @@ struct _vcs_dpi_soc_bounding_box_st dcn3_51_soc = { .do_urgent_latency_adjustment = 0, .urgent_latency_adjustment_fabric_clock_component_us = 0, .urgent_latency_adjustment_fabric_clock_reference_mhz = 0, + .num_chans = 4, + .dispclk_dppclk_vco_speed_mhz = 2400.0, }; /* @@ -340,6 +401,8 @@ void dcn351_update_bw_bounding_box_fpu(struct dc *dc, clock_limits[i].socclk_mhz; dc->dml2_options.bbox_overrides.clks_table.clk_entries[i].memclk_mhz = clk_table->entries[i].memclk_mhz * clk_table->entries[i].wck_ratio; + dc->dml2_options.bbox_overrides.clks_table.clk_entries[i].dtbclk_mhz = + clock_limits[i].dtbclk_mhz; dc->dml2_options.bbox_overrides.clks_table.num_entries_per_clk.num_dcfclk_levels = clk_table->num_entries; dc->dml2_options.bbox_overrides.clks_table.num_entries_per_clk.num_fclk_levels = @@ -352,6 +415,8 @@ void dcn351_update_bw_bounding_box_fpu(struct dc *dc, clk_table->num_entries; dc->dml2_options.bbox_overrides.clks_table.num_entries_per_clk.num_memclk_levels = clk_table->num_entries; + dc->dml2_options.bbox_overrides.clks_table.num_entries_per_clk.num_dtbclk_levels = + clk_table->num_entries; } } @@ -408,7 +473,7 @@ int dcn351_populate_dml_pipes_from_context_fpu(struct dc *dc, { int i, pipe_cnt; struct resource_context *res_ctx = &context->res_ctx; - struct pipe_ctx *pipe; + struct pipe_ctx *pipe = 0; bool upscaled = false; const unsigned int max_allowed_vblank_nom = 1023; @@ -551,6 +616,7 @@ void dcn351_decide_zstate_support(struct dc *dc, struct dc_state *context) if (context->res_ctx.pipe_ctx[i].plane_state) plane_count++; } + /*dcn351 does not support z9/z10*/ if (context->stream_count == 0 || plane_count == 0) { support = DCN_ZSTATE_SUPPORT_ALLOW_Z8_ONLY; @@ -564,11 +630,9 @@ void dcn351_decide_zstate_support(struct dc *dc, struct dc_state *context) dc->debug.minimum_z8_residency_time > 0 ? dc->debug.minimum_z8_residency_time : 1000; bool allow_z8 = context->bw_ctx.dml.vba.StutterPeriod > (double)minmum_z8_residency; - /*for psr1/psr-su, we allow z8 and z10 based on latency, for replay with IPS enabled, it will enter ips2*/ - if (is_pwrseq0 && (is_psr || is_replay)) + if (is_pwrseq0 && (is_psr || is_replay)) support = allow_z8 ? allow_z8 : DCN_ZSTATE_SUPPORT_DISALLOW; - } context->bw_ctx.bw.dcn.clk.zstate_support = support; } diff --git a/drivers/gpu/drm/amd/display/dc/dml2/Makefile b/drivers/gpu/drm/amd/display/dc/dml2/Makefile index acff3449b8d7..c576bb0c780f 100644 --- a/drivers/gpu/drm/amd/display/dc/dml2/Makefile +++ b/drivers/gpu/drm/amd/display/dc/dml2/Makefile @@ -24,40 +24,8 @@ # # Makefile for dml2. -ifdef CONFIG_X86 -dml2_ccflags-$(CONFIG_CC_IS_GCC) := -mhard-float -dml2_ccflags := $(dml2_ccflags-y) -msse -endif - -ifdef CONFIG_PPC64 -dml2_ccflags := -mhard-float -maltivec -endif - -ifdef CONFIG_ARM64 -dml2_rcflags := -mgeneral-regs-only -endif - -ifdef CONFIG_LOONGARCH -dml2_ccflags := -mfpu=64 -dml2_rcflags := -msoft-float -endif - -ifdef CONFIG_CC_IS_GCC -ifeq ($(call cc-ifversion, -lt, 0701, y), y) -IS_OLD_GCC = 1 -endif -endif - -ifdef CONFIG_X86 -ifdef IS_OLD_GCC -# Stack alignment mismatch, proceed with caution. -# GCC < 7.1 cannot compile code using `double` and -mpreferred-stack-boundary=3 -# (8B stack alignment). -dml2_ccflags += -mpreferred-stack-boundary=4 -else -dml2_ccflags += -msse2 -endif -endif +dml2_ccflags := $(CC_FLAGS_FPU) +dml2_rcflags := $(CC_FLAGS_NO_FPU) ifneq ($(CONFIG_FRAME_WARN),0) ifeq ($(filter y,$(CONFIG_KASAN)$(CONFIG_KCSAN)),y) @@ -67,6 +35,7 @@ frame_warn_flag := -Wframe-larger-than=2048 endif endif +subdir-ccflags-y += -I$(FULL_AMD_DISPLAY_PATH)/dc/dml2 CFLAGS_$(AMDDALPATH)/dc/dml2/display_mode_core.o := $(dml2_ccflags) $(frame_warn_flag) CFLAGS_$(AMDDALPATH)/dc/dml2/display_mode_util.o := $(dml2_ccflags) CFLAGS_$(AMDDALPATH)/dc/dml2/dml2_wrapper.o := $(dml2_ccflags) diff --git a/drivers/gpu/drm/amd/display/dc/dml2/display_mode_core.c b/drivers/gpu/drm/amd/display/dc/dml2/display_mode_core.c index 9be5ebf3a8c0..3e919f5c00ca 100644 --- a/drivers/gpu/drm/amd/display/dc/dml2/display_mode_core.c +++ b/drivers/gpu/drm/amd/display/dc/dml2/display_mode_core.c @@ -31,6 +31,8 @@ #include "dml_assert.h" #define DML2_MAX_FMT_420_BUFFER_WIDTH 4096 +#define TB_BORROWED_MAX 400 + // --------------------------- // Declaration Begins // --------------------------- @@ -2782,6 +2784,8 @@ static dml_float_t TruncToValidBPP( } } + *RequiredSlots = (dml_uint_t)(dml_ceil(DesiredBPP / MaxLinkBPP * 64, 1)); + if (DesiredBPP == 0) { if (DSCEnable) { if (MaxLinkBPP < MinDSCBPP) { @@ -2810,10 +2814,6 @@ static dml_float_t TruncToValidBPP( return DesiredBPP; } } - - *RequiredSlots = (dml_uint_t)(dml_ceil(DesiredBPP / MaxLinkBPP * 64, 1)); - - return __DML_DPP_INVALID__; } // TruncToValidBPP static void CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport( @@ -3790,9 +3790,9 @@ static void CalculateStutterEfficiency(struct display_mode_lib_scratch_st *scrat dml_bool_t FoundCriticalSurface = false; dml_uint_t TotalNumberOfActiveOTG = 0; - dml_float_t SinglePixelClock; - dml_uint_t SingleHTotal; - dml_uint_t SingleVTotal; + dml_float_t SinglePixelClock = 0; + dml_uint_t SingleHTotal = 0; + dml_uint_t SingleVTotal = 0; dml_bool_t SameTiming = true; dml_float_t LastStutterPeriod = 0.0; @@ -9460,8 +9460,10 @@ void dml_core_mode_programming(struct display_mode_lib_st *mode_lib, const struc /* Copy the calculated watermarks to mp.Watermark as the getter functions are * implemented by the DML team to copy the calculated values from the mp.Watermark interface. + * &mode_lib->mp.Watermark and &locals->Watermark are the same address, memcpy may lead to + * unexpected behavior. memmove should be used. */ - memcpy(&mode_lib->mp.Watermark, CalculateWatermarks_params->Watermark, sizeof(struct Watermarks)); + memmove(&mode_lib->mp.Watermark, CalculateWatermarks_params->Watermark, sizeof(struct Watermarks)); for (k = 0; k < mode_lib->ms.num_active_planes; ++k) { if (mode_lib->ms.cache_display_cfg.writeback.WritebackEnable[k] == true) { @@ -10214,6 +10216,7 @@ dml_get_var_func(fraction_of_urgent_bandwidth_imm_flip, dml_float_t, mode_lib->m dml_get_var_func(urgent_latency, dml_float_t, mode_lib->mp.UrgentLatency); dml_get_var_func(clk_dcf_deepsleep, dml_float_t, mode_lib->mp.DCFCLKDeepSleep); dml_get_var_func(wm_writeback_dram_clock_change, dml_float_t, mode_lib->mp.Watermark.WritebackDRAMClockChangeWatermark); +dml_get_var_func(wm_writeback_urgent, dml_float_t, mode_lib->mp.Watermark.WritebackUrgentWatermark); dml_get_var_func(stutter_efficiency, dml_float_t, mode_lib->mp.StutterEfficiency); dml_get_var_func(stutter_efficiency_no_vblank, dml_float_t, mode_lib->mp.StutterEfficiencyNotIncludingVBlank); dml_get_var_func(stutter_efficiency_z8, dml_float_t, mode_lib->mp.Z8StutterEfficiency); diff --git a/drivers/gpu/drm/amd/display/dc/dml2/display_mode_core.h b/drivers/gpu/drm/amd/display/dc/dml2/display_mode_core.h index 8452485684f5..3116b88e99dc 100644 --- a/drivers/gpu/drm/amd/display/dc/dml2/display_mode_core.h +++ b/drivers/gpu/drm/amd/display/dc/dml2/display_mode_core.h @@ -94,6 +94,7 @@ dml_get_var_decl(wm_usr_retraining, dml_float_t); dml_get_var_decl(urgent_latency, dml_float_t); dml_get_var_decl(wm_writeback_dram_clock_change, dml_float_t); +dml_get_var_decl(wm_writeback_urgent, dml_float_t); dml_get_var_decl(stutter_efficiency_no_vblank, dml_float_t); dml_get_var_decl(stutter_efficiency, dml_float_t); dml_get_var_decl(stutter_efficiency_z8, dml_float_t); diff --git a/drivers/gpu/drm/amd/display/dc/dml2/display_mode_lib_defines.h b/drivers/gpu/drm/amd/display/dc/dml2/display_mode_lib_defines.h index de63364be01d..14d389525296 100644 --- a/drivers/gpu/drm/amd/display/dc/dml2/display_mode_lib_defines.h +++ b/drivers/gpu/drm/amd/display/dc/dml2/display_mode_lib_defines.h @@ -41,6 +41,7 @@ #define DCN_DML__VM_PRESENT__1 1 #define DCN_DML__HOST_VM_PRESENT 1 #define DCN_DML__HOST_VM_PRESENT__1 1 +#define DCN_DML__DWB 1 #include "dml_depedencies.h" @@ -59,6 +60,7 @@ #define __DML_NUM_PLANES__ DCN_DML__NUM_PLANE #define __DML_NUM_CURSORS__ DCN_DML__NUM_CURSOR #define __DML_DPP_INVALID__ 0 +#define __DML_NUM_DMB__ DCN_DML__DWB #define __DML_PIPE_NO_PLANE__ 99 #define __DML_MAX_STATE_ARRAY_SIZE__ DCN_DML__NUM_PWR_STATE diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml2_dc_resource_mgmt.c b/drivers/gpu/drm/amd/display/dc/dml2/dml2_dc_resource_mgmt.c index a52c594e1ba4..ad2a6b4769fe 100644 --- a/drivers/gpu/drm/amd/display/dc/dml2/dml2_dc_resource_mgmt.c +++ b/drivers/gpu/drm/amd/display/dc/dml2/dml2_dc_resource_mgmt.c @@ -88,7 +88,8 @@ static int find_disp_cfg_idx_by_plane_id(struct dml2_dml_to_dc_pipe_mapping *map return i; } - return -1; + ASSERT(false); + return __DML2_WRAPPER_MAX_STREAMS_PLANES__; } static int find_disp_cfg_idx_by_stream_id(struct dml2_dml_to_dc_pipe_mapping *mapping, unsigned int stream_id) @@ -100,7 +101,8 @@ static int find_disp_cfg_idx_by_stream_id(struct dml2_dml_to_dc_pipe_mapping *ma return i; } - return -1; + ASSERT(false); + return __DML2_WRAPPER_MAX_STREAMS_PLANES__; } // The master pipe of a stream is defined as the top pipe in odm slice 0 @@ -793,8 +795,8 @@ static void map_pipes_for_plane(struct dml2_context *ctx, struct dc_state *state free_unused_pipes_for_plane(ctx, state, plane, &scratch->pipe_pool, stream->stream_id, plane_index); } -static unsigned int get_mpc_factor(struct dml2_context *ctx, - const struct dc_state *state, +static unsigned int get_target_mpc_factor(struct dml2_context *ctx, + struct dc_state *state, const struct dml_display_cfg_st *disp_cfg, struct dml2_dml_to_dc_pipe_mapping *mapping, const struct dc_stream_status *status, @@ -805,10 +807,10 @@ static unsigned int get_mpc_factor(struct dml2_context *ctx, unsigned int cfg_idx; unsigned int mpc_factor; - get_plane_id(ctx, state, status->plane_states[plane_idx], - stream->stream_id, plane_idx, &plane_id); - cfg_idx = find_disp_cfg_idx_by_plane_id(mapping, plane_id); if (ctx->architecture == dml2_architecture_20) { + get_plane_id(ctx, state, status->plane_states[plane_idx], + stream->stream_id, plane_idx, &plane_id); + cfg_idx = find_disp_cfg_idx_by_plane_id(mapping, plane_id); mpc_factor = (unsigned int)disp_cfg->hw.DPPPerSurface[cfg_idx]; } else { mpc_factor = 1; @@ -822,16 +824,18 @@ static unsigned int get_mpc_factor(struct dml2_context *ctx, return mpc_factor; } -static unsigned int get_odm_factor( +static unsigned int get_target_odm_factor( const struct dml2_context *ctx, + struct dc_state *state, const struct dml_display_cfg_st *disp_cfg, struct dml2_dml_to_dc_pipe_mapping *mapping, const struct dc_stream_state *stream) { - unsigned int cfg_idx = find_disp_cfg_idx_by_stream_id( - mapping, stream->stream_id); + unsigned int cfg_idx; - if (ctx->architecture == dml2_architecture_20) + if (ctx->architecture == dml2_architecture_20) { + cfg_idx = find_disp_cfg_idx_by_stream_id( + mapping, stream->stream_id); switch (disp_cfg->hw.ODMMode[cfg_idx]) { case dml_odm_mode_bypass: return 1; @@ -842,83 +846,122 @@ static unsigned int get_odm_factor( default: break; } + } ASSERT(false); return 1; } +static unsigned int get_source_odm_factor(const struct dml2_context *ctx, + struct dc_state *state, + const struct dc_stream_state *stream) +{ + struct pipe_ctx *otg_master = ctx->config.callbacks.get_otg_master_for_stream(&state->res_ctx, stream); + + return ctx->config.callbacks.get_odm_slice_count(otg_master); +} + +static unsigned int get_source_mpc_factor(const struct dml2_context *ctx, + struct dc_state *state, + const struct dc_plane_state *plane) +{ + struct pipe_ctx *dpp_pipes[MAX_PIPES] = {0}; + int dpp_pipe_count = ctx->config.callbacks.get_dpp_pipes_for_plane(plane, + &state->res_ctx, dpp_pipes); + + ASSERT(dpp_pipe_count > 0); + return ctx->config.callbacks.get_mpc_slice_count(dpp_pipes[0]); +} + + static void populate_mpc_factors_for_stream( struct dml2_context *ctx, const struct dml_display_cfg_st *disp_cfg, struct dml2_dml_to_dc_pipe_mapping *mapping, - const struct dc_state *state, + struct dc_state *state, unsigned int stream_idx, - unsigned int odm_factor, - unsigned int mpc_factors[MAX_PIPES]) + struct dml2_pipe_combine_factor odm_factor, + struct dml2_pipe_combine_factor mpc_factors[MAX_PIPES]) { const struct dc_stream_status *status = &state->stream_status[stream_idx]; int i; - for (i = 0; i < status->plane_count; i++) - if (odm_factor == 1) - mpc_factors[i] = get_mpc_factor( - ctx, state, disp_cfg, mapping, status, - state->streams[stream_idx], i); - else - mpc_factors[i] = 1; + for (i = 0; i < status->plane_count; i++) { + mpc_factors[i].source = get_source_mpc_factor(ctx, state, status->plane_states[i]); + mpc_factors[i].target = (odm_factor.target == 1) ? + get_target_mpc_factor(ctx, state, disp_cfg, mapping, status, state->streams[stream_idx], i) : 1; + } } static void populate_odm_factors(const struct dml2_context *ctx, const struct dml_display_cfg_st *disp_cfg, struct dml2_dml_to_dc_pipe_mapping *mapping, - const struct dc_state *state, - unsigned int odm_factors[MAX_PIPES]) + struct dc_state *state, + struct dml2_pipe_combine_factor odm_factors[MAX_PIPES]) { int i; - for (i = 0; i < state->stream_count; i++) - odm_factors[i] = get_odm_factor( - ctx, disp_cfg, mapping, state->streams[i]); + for (i = 0; i < state->stream_count; i++) { + odm_factors[i].source = get_source_odm_factor(ctx, state, state->streams[i]); + odm_factors[i].target = get_target_odm_factor( + ctx, state, disp_cfg, mapping, state->streams[i]); + } } -static bool map_dc_pipes_for_stream(struct dml2_context *ctx, +static bool unmap_dc_pipes_for_stream(struct dml2_context *ctx, struct dc_state *state, const struct dc_state *existing_state, const struct dc_stream_state *stream, const struct dc_stream_status *status, - unsigned int odm_factor, - unsigned int mpc_factors[MAX_PIPES]) + struct dml2_pipe_combine_factor odm_factor, + struct dml2_pipe_combine_factor mpc_factors[MAX_PIPES]) { int plane_idx; bool result = true; - if (odm_factor == 1) - /* - * ODM and MPC combines are by DML design mutually exclusive. - * ODM factor of 1 means MPC factors may be greater than 1. - * In this case, we want to set ODM factor to 1 first to free up - * pipe resources from previous ODM configuration before setting - * up MPC combine to acquire more pipe resources. - */ + for (plane_idx = 0; plane_idx < status->plane_count; plane_idx++) + if (mpc_factors[plane_idx].target < mpc_factors[plane_idx].source) + result &= ctx->config.callbacks.update_pipes_for_plane_with_slice_count( + state, + existing_state, + ctx->config.callbacks.dc->res_pool, + status->plane_states[plane_idx], + mpc_factors[plane_idx].target); + if (odm_factor.target < odm_factor.source) result &= ctx->config.callbacks.update_pipes_for_stream_with_slice_count( state, existing_state, ctx->config.callbacks.dc->res_pool, stream, - odm_factor); + odm_factor.target); + return result; +} + +static bool map_dc_pipes_for_stream(struct dml2_context *ctx, + struct dc_state *state, + const struct dc_state *existing_state, + const struct dc_stream_state *stream, + const struct dc_stream_status *status, + struct dml2_pipe_combine_factor odm_factor, + struct dml2_pipe_combine_factor mpc_factors[MAX_PIPES]) +{ + int plane_idx; + bool result = true; + for (plane_idx = 0; plane_idx < status->plane_count; plane_idx++) - result &= ctx->config.callbacks.update_pipes_for_plane_with_slice_count( - state, - existing_state, - ctx->config.callbacks.dc->res_pool, - status->plane_states[plane_idx], - mpc_factors[plane_idx]); - if (odm_factor > 1) + if (mpc_factors[plane_idx].target > mpc_factors[plane_idx].source) + result &= ctx->config.callbacks.update_pipes_for_plane_with_slice_count( + state, + existing_state, + ctx->config.callbacks.dc->res_pool, + status->plane_states[plane_idx], + mpc_factors[plane_idx].target); + if (odm_factor.target > odm_factor.source) result &= ctx->config.callbacks.update_pipes_for_stream_with_slice_count( state, existing_state, ctx->config.callbacks.dc->res_pool, stream, - odm_factor); + odm_factor.target); return result; } @@ -928,20 +971,20 @@ static bool map_dc_pipes_with_callbacks(struct dml2_context *ctx, struct dml2_dml_to_dc_pipe_mapping *mapping, const struct dc_state *existing_state) { - unsigned int odm_factors[MAX_PIPES]; - unsigned int mpc_factors_for_stream[MAX_PIPES]; int i; bool result = true; - populate_odm_factors(ctx, disp_cfg, mapping, state, odm_factors); - for (i = 0; i < state->stream_count; i++) { + populate_odm_factors(ctx, disp_cfg, mapping, state, ctx->pipe_combine_scratch.odm_factors); + for (i = 0; i < state->stream_count; i++) populate_mpc_factors_for_stream(ctx, disp_cfg, mapping, state, - i, odm_factors[i], mpc_factors_for_stream); - result &= map_dc_pipes_for_stream(ctx, state, existing_state, - state->streams[i], - &state->stream_status[i], - odm_factors[i], mpc_factors_for_stream); - } + i, ctx->pipe_combine_scratch.odm_factors[i], ctx->pipe_combine_scratch.mpc_factors[i]); + for (i = 0; i < state->stream_count; i++) + result &= unmap_dc_pipes_for_stream(ctx, state, existing_state, state->streams[i], + &state->stream_status[i], ctx->pipe_combine_scratch.odm_factors[i], ctx->pipe_combine_scratch.mpc_factors[i]); + for (i = 0; i < state->stream_count; i++) + result &= map_dc_pipes_for_stream(ctx, state, existing_state, state->streams[i], + &state->stream_status[i], ctx->pipe_combine_scratch.odm_factors[i], ctx->pipe_combine_scratch.mpc_factors[i]); + return result; } @@ -1037,6 +1080,12 @@ bool dml2_map_dc_pipes(struct dml2_context *ctx, struct dc_state *state, const s ASSERT(false); } } + + if (ctx->config.callbacks.build_test_pattern_params && + pipe->stream && + pipe->prev_odm_pipe == NULL && + pipe->top_pipe == NULL) + ctx->config.callbacks.build_test_pattern_params(&state->res_ctx, pipe); } return true; diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml2_dc_resource_mgmt.h b/drivers/gpu/drm/amd/display/dc/dml2/dml2_dc_resource_mgmt.h index 2f91244a7b01..1538b708d8be 100644 --- a/drivers/gpu/drm/amd/display/dc/dml2/dml2_dc_resource_mgmt.h +++ b/drivers/gpu/drm/amd/display/dc/dml2/dml2_dc_resource_mgmt.h @@ -30,6 +30,8 @@ #include "dml2_dc_types.h" struct dml2_context; +struct dml2_dml_to_dc_pipe_mapping; +struct dml_display_cfg_st; /* * dml2_map_dc_pipes - Creates a pipe linkage in dc_state based on current display config. diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml2_internal_types.h b/drivers/gpu/drm/amd/display/dc/dml2/dml2_internal_types.h index 1cf8a884c0fb..9dab4e43c511 100644 --- a/drivers/gpu/drm/amd/display/dc/dml2/dml2_internal_types.h +++ b/drivers/gpu/drm/amd/display/dc/dml2/dml2_internal_types.h @@ -109,10 +109,21 @@ enum dml2_architecture { dml2_architecture_20, }; +struct dml2_pipe_combine_factor { + unsigned int source; + unsigned int target; +}; + +struct dml2_pipe_combine_scratch { + struct dml2_pipe_combine_factor odm_factors[MAX_PIPES]; + struct dml2_pipe_combine_factor mpc_factors[MAX_PIPES][MAX_PIPES]; +}; + struct dml2_context { enum dml2_architecture architecture; struct dml2_configuration_options config; struct dml2_helper_det_policy_scratch det_helper_scratch; + struct dml2_pipe_combine_scratch pipe_combine_scratch; union { struct { struct display_mode_lib_st dml_core_ctx; diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml2_translation_helper.c b/drivers/gpu/drm/amd/display/dc/dml2/dml2_translation_helper.c index 17a58f41fc6a..a41812598ce8 100644 --- a/drivers/gpu/drm/amd/display/dc/dml2/dml2_translation_helper.c +++ b/drivers/gpu/drm/amd/display/dc/dml2/dml2_translation_helper.c @@ -29,6 +29,7 @@ #include "dml2_translation_helper.h" #define NUM_DCFCLK_STAS 5 +#define NUM_DCFCLK_STAS_NEW 8 void dml2_init_ip_params(struct dml2_context *dml2, const struct dc *in_dc, struct ip_params_st *out) { @@ -228,17 +229,13 @@ void dml2_init_socbb_params(struct dml2_context *dml2, const struct dc *in_dc, s break; case dml_project_dcn35: + case dml_project_dcn351: out->num_chans = 4; out->round_trip_ping_latency_dcfclk_cycles = 106; out->smn_latency_us = 2; out->dispclk_dppclk_vco_speed_mhz = 3600; break; - case dml_project_dcn351: - out->num_chans = 16; - out->round_trip_ping_latency_dcfclk_cycles = 1100; - out->smn_latency_us = 2; - break; } /* ---Overrides if available--- */ if (dml2->config.bbox_overrides.dram_num_chan) @@ -253,12 +250,21 @@ void dml2_init_soc_states(struct dml2_context *dml2, const struct dc *in_dc, { struct dml2_policy_build_synthetic_soc_states_scratch *s = &dml2->v20.scratch.create_scratch.build_synthetic_socbb_scratch; struct dml2_policy_build_synthetic_soc_states_params *p = &dml2->v20.scratch.build_synthetic_socbb_params; - unsigned int dcfclk_stas_mhz[NUM_DCFCLK_STAS]; + unsigned int dcfclk_stas_mhz[NUM_DCFCLK_STAS] = {0}; + unsigned int dcfclk_stas_mhz_new[NUM_DCFCLK_STAS_NEW] = {0}; + unsigned int dml_project = dml2->v20.dml_core_ctx.project; + unsigned int i = 0; unsigned int transactions_per_mem_clock = 16; // project specific, depends on used Memory type - p->dcfclk_stas_mhz = dcfclk_stas_mhz; - p->num_dcfclk_stas = NUM_DCFCLK_STAS; + if (dml_project == dml_project_dcn351) { + p->dcfclk_stas_mhz = dcfclk_stas_mhz_new; + p->num_dcfclk_stas = NUM_DCFCLK_STAS_NEW; + } else { + p->dcfclk_stas_mhz = dcfclk_stas_mhz; + p->num_dcfclk_stas = NUM_DCFCLK_STAS; + } + p->in_bbox = in_bbox; p->out_states = out; p->in_states = &dml2->v20.scratch.create_scratch.in_states; @@ -436,8 +442,7 @@ void dml2_init_soc_states(struct dml2_context *dml2, const struct dc *in_dc, } dml2_policy_build_synthetic_soc_states(s, p); - if (dml2->v20.dml_core_ctx.project == dml_project_dcn35 || - dml2->v20.dml_core_ctx.project == dml_project_dcn351) { + if (dml2->v20.dml_core_ctx.project == dml_project_dcn35) { // Override last out_state with data from last in_state // This will ensure that out_state contains max fclk memcpy(&p->out_states->state_array[p->out_states->num_states - 1], @@ -1056,7 +1061,46 @@ static void dml2_populate_pipe_to_plane_index_mapping(struct dml2_context *dml2, plane_index = 0; } } - +static void populate_dml_writeback_cfg_from_stream_state(struct dml_writeback_cfg_st *out, + unsigned int location, const struct dc_stream_state *in) +{ + if (in->num_wb_info > 0) { + for (int i = 0; i < __DML_NUM_DMB__; i++) { + const struct dc_writeback_info *wb_info = &in->writeback_info[i]; + /*current dml support 1 dwb per stream, limitation*/ + if (wb_info->wb_enabled) { + out->WritebackEnable[location] = wb_info->wb_enabled; + out->ActiveWritebacksPerSurface[location] = wb_info->dwb_params.cnv_params.src_width; + out->WritebackDestinationWidth[location] = wb_info->dwb_params.dest_width; + out->WritebackDestinationHeight[location] = wb_info->dwb_params.dest_height; + + out->WritebackSourceWidth[location] = wb_info->dwb_params.cnv_params.crop_en ? + wb_info->dwb_params.cnv_params.crop_width : + wb_info->dwb_params.cnv_params.src_width; + + out->WritebackSourceHeight[location] = wb_info->dwb_params.cnv_params.crop_en ? + wb_info->dwb_params.cnv_params.crop_height : + wb_info->dwb_params.cnv_params.src_height; + /*current design does not have chroma scaling, need to follow up*/ + out->WritebackHTaps[location] = wb_info->dwb_params.scaler_taps.h_taps > 0 ? + wb_info->dwb_params.scaler_taps.h_taps : 1; + out->WritebackVTaps[location] = wb_info->dwb_params.scaler_taps.v_taps > 0 ? + wb_info->dwb_params.scaler_taps.v_taps : 1; + + out->WritebackHRatio[location] = wb_info->dwb_params.cnv_params.crop_en ? + (double)wb_info->dwb_params.cnv_params.crop_width / + (double)wb_info->dwb_params.dest_width : + (double)wb_info->dwb_params.cnv_params.src_width / + (double)wb_info->dwb_params.dest_width; + out->WritebackVRatio[location] = wb_info->dwb_params.cnv_params.crop_en ? + (double)wb_info->dwb_params.cnv_params.crop_height / + (double)wb_info->dwb_params.dest_height : + (double)wb_info->dwb_params.cnv_params.src_height / + (double)wb_info->dwb_params.dest_height; + } + } + } +} void map_dc_state_into_dml_display_cfg(struct dml2_context *dml2, struct dc_state *context, struct dml_display_cfg_st *dml_dispcfg) { int i = 0, j = 0, k = 0; @@ -1101,6 +1145,10 @@ void map_dc_state_into_dml_display_cfg(struct dml2_context *dml2, struct dc_stat populate_dml_timing_cfg_from_stream_state(&dml_dispcfg->timing, disp_cfg_stream_location, context->streams[i]); populate_dml_output_cfg_from_stream_state(&dml_dispcfg->output, disp_cfg_stream_location, context->streams[i], current_pipe_context); + /*Call site for populate_dml_writeback_cfg_from_stream_state*/ + populate_dml_writeback_cfg_from_stream_state(&dml_dispcfg->writeback, + disp_cfg_stream_location, context->streams[i]); + switch (context->streams[i]->debug.force_odm_combine_segments) { case 2: dml2->v20.dml_core_ctx.policy.ODMUse[disp_cfg_stream_location] = dml_odm_use_policy_combine_2to1; diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml2_utils.c b/drivers/gpu/drm/amd/display/dc/dml2/dml2_utils.c index f15d1dbad6a9..0f8b3336e26d 100644 --- a/drivers/gpu/drm/amd/display/dc/dml2/dml2_utils.c +++ b/drivers/gpu/drm/amd/display/dc/dml2/dml2_utils.c @@ -224,7 +224,7 @@ static int find_dml_pipe_idx_by_plane_id(struct dml2_context *ctx, unsigned int static bool get_plane_id(struct dml2_context *dml2, const struct dc_state *state, const struct dc_plane_state *plane, unsigned int stream_id, unsigned int plane_index, unsigned int *plane_id) { - int i, j; + unsigned int i, j; bool is_plane_duplicate = dml2->v20.scratch.plane_duplicate_exists; if (!plane_id) @@ -327,6 +327,8 @@ void dml2_calculate_rq_and_dlg_params(const struct dc *dc, struct dc_state *cont dml_pipe_idx = dml2_helper_find_dml_pipe_idx_by_stream_id(in_ctx, context->res_ctx.pipe_ctx[dc_pipe_ctx_index].stream->stream_id); } + if (dml_pipe_idx == 0xFFFFFFFF) + continue; ASSERT(in_ctx->v20.scratch.dml_to_dc_pipe_mapping.dml_pipe_idx_to_stream_id_valid[dml_pipe_idx]); ASSERT(in_ctx->v20.scratch.dml_to_dc_pipe_mapping.dml_pipe_idx_to_stream_id[dml_pipe_idx] == context->res_ctx.pipe_ctx[dc_pipe_ctx_index].stream->stream_id); @@ -374,10 +376,16 @@ void dml2_calculate_rq_and_dlg_params(const struct dc *dc, struct dc_state *cont context->bw_ctx.bw.dcn.clk.bw_dppclk_khz = context->bw_ctx.bw.dcn.clk.dppclk_khz; context->bw_ctx.bw.dcn.clk.bw_dispclk_khz = context->bw_ctx.bw.dcn.clk.dispclk_khz; + context->bw_ctx.bw.dcn.clk.max_supported_dppclk_khz = in_ctx->v20.dml_core_ctx.states.state_array[in_ctx->v20.scratch.mode_support_params.out_lowest_state_idx].dppclk_mhz * 1000; context->bw_ctx.bw.dcn.clk.max_supported_dispclk_khz = in_ctx->v20.dml_core_ctx.states.state_array[in_ctx->v20.scratch.mode_support_params.out_lowest_state_idx].dispclk_mhz * 1000; + + if (dc->config.forced_clocks || dc->debug.max_disp_clk) { + context->bw_ctx.bw.dcn.clk.bw_dispclk_khz = context->bw_ctx.bw.dcn.clk.max_supported_dispclk_khz; + context->bw_ctx.bw.dcn.clk.bw_dppclk_khz = context->bw_ctx.bw.dcn.clk.max_supported_dppclk_khz ; + } } void dml2_extract_watermark_set(struct dcn_watermarks *watermark, struct display_mode_lib_st *dml_core_ctx) @@ -396,6 +404,71 @@ void dml2_extract_watermark_set(struct dcn_watermarks *watermark, struct display watermark->cstate_pstate.cstate_exit_z8_ns = dml_get_wm_z8_stutter(dml_core_ctx) * 1000; } +unsigned int dml2_calc_max_scaled_time( + unsigned int time_per_pixel, + enum mmhubbub_wbif_mode mode, + unsigned int urgent_watermark) +{ + unsigned int time_per_byte = 0; + unsigned int total_free_entry = 0xb40; + unsigned int buf_lh_capability; + unsigned int max_scaled_time; + + if (mode == PACKED_444) /* packed mode 32 bpp */ + time_per_byte = time_per_pixel/4; + else if (mode == PACKED_444_FP16) /* packed mode 64 bpp */ + time_per_byte = time_per_pixel/8; + + if (time_per_byte == 0) + time_per_byte = 1; + + buf_lh_capability = (total_free_entry*time_per_byte*32) >> 6; /* time_per_byte is in u6.6*/ + max_scaled_time = buf_lh_capability - urgent_watermark; + return max_scaled_time; +} + +void dml2_extract_writeback_wm(struct dc_state *context, struct display_mode_lib_st *dml_core_ctx) +{ + int i, j = 0;; + struct mcif_arb_params *wb_arb_params = NULL; + struct dcn_bw_writeback *bw_writeback = NULL; + enum mmhubbub_wbif_mode wbif_mode = PACKED_444_FP16; /*for now*/ + + if (context->stream_count != 0) { + for (i = 0; i < context->stream_count; i++) { + if (context->streams[i]->num_wb_info != 0) + j++; + } + } + if (j == 0) /*no dwb */ + return; + for (i = 0; i < __DML_NUM_DMB__; i++) { + bw_writeback = &context->bw_ctx.bw.dcn.bw_writeback; + wb_arb_params = &context->bw_ctx.bw.dcn.bw_writeback.mcif_wb_arb[i]; + + for (j = 0 ; j < 4; j++) { + /*current dml only has one set of watermark, need to follow up*/ + bw_writeback->mcif_wb_arb[i].cli_watermark[j] = + dml_get_wm_writeback_urgent(dml_core_ctx) * 1000; + bw_writeback->mcif_wb_arb[i].pstate_watermark[j] = + dml_get_wm_writeback_dram_clock_change(dml_core_ctx) * 1000; + } + if (context->res_ctx.pipe_ctx[i].stream->phy_pix_clk != 0) { + /* time_per_pixel should be in u6.6 format */ + bw_writeback->mcif_wb_arb[i].time_per_pixel = + (1000000 << 6) / context->res_ctx.pipe_ctx[i].stream->phy_pix_clk; + } + bw_writeback->mcif_wb_arb[i].slice_lines = 32; + bw_writeback->mcif_wb_arb[i].arbitration_slice = 2; + bw_writeback->mcif_wb_arb[i].max_scaled_time = + dml2_calc_max_scaled_time(wb_arb_params->time_per_pixel, + wbif_mode, wb_arb_params->cli_watermark[0]); + /*not required any more*/ + bw_writeback->mcif_wb_arb[i].dram_speed_change_duration = + dml_get_wm_writeback_dram_clock_change(dml_core_ctx) * 1000; + + } +} void dml2_initialize_det_scratch(struct dml2_context *in_ctx) { int i; @@ -468,6 +541,9 @@ bool dml2_verify_det_buffer_configuration(struct dml2_context *in_ctx, struct dc dml_pipe_idx = find_dml_pipe_idx_by_plane_id(in_ctx, plane_id); else dml_pipe_idx = dml2_helper_find_dml_pipe_idx_by_stream_id(in_ctx, display_state->res_ctx.pipe_ctx[i].stream->stream_id); + + if (dml_pipe_idx == 0xFFFFFFFF) + continue; total_det_allocated += dml_get_det_buffer_size_kbytes(&in_ctx->v20.dml_core_ctx, dml_pipe_idx); if (total_det_allocated > max_det_size) { need_recalculation = true; diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml2_utils.h b/drivers/gpu/drm/amd/display/dc/dml2/dml2_utils.h index 5842d6d3c4b6..04fcfe637119 100644 --- a/drivers/gpu/drm/amd/display/dc/dml2/dml2_utils.h +++ b/drivers/gpu/drm/amd/display/dc/dml2/dml2_utils.h @@ -40,9 +40,14 @@ void dml2_util_copy_dml_output(struct dml_output_cfg_st *dml_output_array, unsig unsigned int dml2_util_get_maximum_odm_combine_for_output(bool force_odm_4to1, enum dml_output_encoder_class encoder, bool dsc_enabled); void dml2_copy_clocks_to_dc_state(struct dml2_dcn_clocks *out_clks, struct dc_state *context); void dml2_extract_watermark_set(struct dcn_watermarks *watermark, struct display_mode_lib_st *dml_core_ctx); +void dml2_extract_writeback_wm(struct dc_state *context, struct display_mode_lib_st *dml_core_ctx); int dml2_helper_find_dml_pipe_idx_by_stream_id(struct dml2_context *ctx, unsigned int stream_id); bool is_dtbclk_required(const struct dc *dc, struct dc_state *context); bool dml2_is_stereo_timing(const struct dc_stream_state *stream); +unsigned int dml2_calc_max_scaled_time( + unsigned int time_per_pixel, + enum mmhubbub_wbif_mode mode, + unsigned int urgent_watermark); /* * dml2_dc_construct_pipes - This function will determine if we need additional pipes based diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml2_wrapper.c b/drivers/gpu/drm/amd/display/dc/dml2/dml2_wrapper.c index 72cca367062e..9412d5384a41 100644 --- a/drivers/gpu/drm/amd/display/dc/dml2/dml2_wrapper.c +++ b/drivers/gpu/drm/amd/display/dc/dml2/dml2_wrapper.c @@ -570,6 +570,7 @@ static bool dml2_validate_and_build_resource(const struct dc *in_dc, struct dc_s struct dml2_dcn_clocks out_clks; unsigned int result = 0; bool need_recalculation = false; + uint32_t cstate_enter_plus_exit_z8_ns; if (!context || context->stream_count == 0) return true; @@ -639,8 +640,17 @@ static bool dml2_validate_and_build_resource(const struct dc *in_dc, struct dc_s dml2_extract_watermark_set(&context->bw_ctx.bw.dcn.watermarks.b, &dml2->v20.dml_core_ctx); memcpy(&context->bw_ctx.bw.dcn.watermarks.c, &dml2->v20.g6_temp_read_watermark_set, sizeof(context->bw_ctx.bw.dcn.watermarks.c)); dml2_extract_watermark_set(&context->bw_ctx.bw.dcn.watermarks.d, &dml2->v20.dml_core_ctx); + dml2_extract_writeback_wm(context, &dml2->v20.dml_core_ctx); //copy for deciding zstate use context->bw_ctx.dml.vba.StutterPeriod = context->bw_ctx.dml2->v20.dml_core_ctx.mp.StutterPeriod; + + cstate_enter_plus_exit_z8_ns = context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.cstate_enter_plus_exit_z8_ns; + + if (context->bw_ctx.dml.vba.StutterPeriod < in_dc->debug.minimum_z8_residency_time && + cstate_enter_plus_exit_z8_ns < in_dc->debug.minimum_z8_residency_time * 1000) + cstate_enter_plus_exit_z8_ns = in_dc->debug.minimum_z8_residency_time * 1000; + + context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.cstate_enter_plus_exit_z8_ns = cstate_enter_plus_exit_z8_ns; } return result; @@ -681,13 +691,13 @@ static void dml2_apply_debug_options(const struct dc *dc, struct dml2_context *d } } -bool dml2_validate(const struct dc *in_dc, struct dc_state *context, bool fast_validate) +bool dml2_validate(const struct dc *in_dc, struct dc_state *context, struct dml2_context *dml2, bool fast_validate) { bool out = false; - if (!(context->bw_ctx.dml2)) + if (!dml2) return false; - dml2_apply_debug_options(in_dc, context->bw_ctx.dml2); + dml2_apply_debug_options(in_dc, dml2); /* Use dml_validate_only for fast_validate path */ diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml2_wrapper.h b/drivers/gpu/drm/amd/display/dc/dml2/dml2_wrapper.h index cc662d682fd4..4a8bd2f4195e 100644 --- a/drivers/gpu/drm/amd/display/dc/dml2/dml2_wrapper.h +++ b/drivers/gpu/drm/amd/display/dc/dml2/dml2_wrapper.h @@ -71,6 +71,7 @@ struct dml2_dcn_clocks { struct dml2_dc_callbacks { struct dc *dc; bool (*build_scaling_params)(struct pipe_ctx *pipe_ctx); + void (*build_test_pattern_params)(struct resource_context *res_ctx, struct pipe_ctx *otg_master); bool (*can_support_mclk_switch_using_fw_based_vblank_stretch)(struct dc *dc, struct dc_state *context); bool (*acquire_secondary_pipe_for_mpc_odm)(const struct dc *dc, struct dc_state *state, struct pipe_ctx *pri_pipe, struct pipe_ctx *sec_pipe, bool odm); bool (*update_pipes_for_stream_with_slice_count)( @@ -86,8 +87,23 @@ struct dml2_dc_callbacks { const struct dc_plane_state *plane, int slice_count); int (*get_odm_slice_index)(const struct pipe_ctx *opp_head); + int (*get_odm_slice_count)(const struct pipe_ctx *opp_head); int (*get_mpc_slice_index)(const struct pipe_ctx *dpp_pipe); + int (*get_mpc_slice_count)(const struct pipe_ctx *dpp_pipe); struct pipe_ctx *(*get_opp_head)(const struct pipe_ctx *pipe_ctx); + struct pipe_ctx *(*get_otg_master_for_stream)( + struct resource_context *res_ctx, + const struct dc_stream_state *stream); + int (*get_opp_heads_for_otg_master)(const struct pipe_ctx *otg_master, + struct resource_context *res_ctx, + struct pipe_ctx *opp_heads[MAX_PIPES]); + int (*get_dpp_pipes_for_plane)(const struct dc_plane_state *plane, + struct resource_context *res_ctx, + struct pipe_ctx *dpp_pipes[MAX_PIPES]); + struct dc_stream_status *(*get_stream_status)( + struct dc_state *state, + const struct dc_stream_state *stream); + struct dc_stream_state *(*get_stream_from_id)(const struct dc_state *state, unsigned int id); }; struct dml2_dc_svp_callbacks { @@ -96,10 +112,10 @@ struct dml2_dc_svp_callbacks { struct dc_stream_state* (*create_phantom_stream)(const struct dc *dc, struct dc_state *state, struct dc_stream_state *main_stream); - struct dc_plane_state* (*create_phantom_plane)(struct dc *dc, + struct dc_plane_state* (*create_phantom_plane)(const struct dc *dc, struct dc_state *state, struct dc_plane_state *main_plane); - enum dc_status (*add_phantom_stream)(struct dc *dc, + enum dc_status (*add_phantom_stream)(const struct dc *dc, struct dc_state *state, struct dc_stream_state *phantom_stream, struct dc_stream_state *main_stream); @@ -108,7 +124,7 @@ struct dml2_dc_svp_callbacks { struct dc_stream_state *stream, struct dc_plane_state *plane_state, struct dc_state *context); - enum dc_status (*remove_phantom_stream)(struct dc *dc, + enum dc_status (*remove_phantom_stream)(const struct dc *dc, struct dc_state *state, struct dc_stream_state *stream); void (*release_phantom_plane)(const struct dc *dc, @@ -121,6 +137,15 @@ struct dml2_dc_svp_callbacks { enum mall_stream_type (*get_pipe_subvp_type)(const struct dc_state *state, const struct pipe_ctx *pipe_ctx); enum mall_stream_type (*get_stream_subvp_type)(const struct dc_state *state, const struct dc_stream_state *stream); struct dc_stream_state *(*get_paired_subvp_stream)(const struct dc_state *state, const struct dc_stream_state *stream); + bool (*remove_phantom_streams_and_planes)( + const struct dc *dc, + struct dc_state *state); + void (*release_phantom_streams_and_planes)( + const struct dc *dc, + struct dc_state *state); + unsigned int (*calculate_mall_ways_from_bytes)( + const struct dc *dc, + unsigned int total_size_in_mall_bytes); }; struct dml2_clks_table_entry { @@ -191,6 +216,8 @@ struct dml2_configuration_options { unsigned int max_segments_per_hubp; unsigned int det_segment_size; bool map_dc_pipes_with_callbacks; + + bool use_clock_dc_limits; }; /* @@ -244,6 +271,7 @@ void dml2_reinit(const struct dc *in_dc, */ bool dml2_validate(const struct dc *in_dc, struct dc_state *context, + struct dml2_context *dml2, bool fast_validate); /* diff --git a/drivers/gpu/drm/amd/display/dc/dpp/Makefile b/drivers/gpu/drm/amd/display/dc/dpp/Makefile new file mode 100644 index 000000000000..99bd36073561 --- /dev/null +++ b/drivers/gpu/drm/amd/display/dc/dpp/Makefile @@ -0,0 +1,77 @@ + +# Copyright 2022 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. +# +# Makefile for the 'dpp' sub-component of DAL. +# +ifdef CONFIG_DRM_AMD_DC_FP +############################################################################### +# DCN +############################################################################### + +DPP_DCN10 = dcn10_dpp.o dcn10_dpp_dscl.o dcn10_dpp_cm.o + +AMD_DAL_DPP_DCN10 = $(addprefix $(AMDDALPATH)/dc/dpp/dcn10/,$(DPP_DCN10)) + +AMD_DISPLAY_FILES += $(AMD_DAL_DPP_DCN10) + +############################################################################### + +DPP_DCN20 = dcn20_dpp.o dcn20_dpp_cm.o + +AMD_DAL_DPP_DCN20 = $(addprefix $(AMDDALPATH)/dc/dpp/dcn20/,$(DPP_DCN20)) + +AMD_DISPLAY_FILES += $(AMD_DAL_DPP_DCN20) + +############################################################################### + +DPP_DCN201 = dcn201_dpp.o + +AMD_DAL_DPP_DCN201 = $(addprefix $(AMDDALPATH)/dc/dpp/dcn201/,$(DPP_DCN201)) + +AMD_DISPLAY_FILES += $(AMD_DAL_DPP_DCN201) + +############################################################################### + +DPP_DCN30 = dcn30_dpp.o dcn30_dpp_cm.o + +AMD_DAL_DPP_DCN30 = $(addprefix $(AMDDALPATH)/dc/dpp/dcn30/,$(DPP_DCN30)) + +AMD_DISPLAY_FILES += $(AMD_DAL_DPP_DCN30) + +############################################################################### + +DPP_DCN32 = dcn32_dpp.o + +AMD_DAL_DPP_DCN32 = $(addprefix $(AMDDALPATH)/dc/dpp/dcn32/,$(DPP_DCN32)) + +AMD_DISPLAY_FILES += $(AMD_DAL_DPP_DCN32) + +############################################################################### + +DPP_DCN35 = dcn35_dpp.o + +AMD_DAL_DPP_DCN35 = $(addprefix $(AMDDALPATH)/dc/dpp/dcn35/,$(DPP_DCN35)) + +AMD_DISPLAY_FILES += $(AMD_DAL_DPP_DCN35) + +############################################################################### + +endif
\ No newline at end of file diff --git a/drivers/gpu/drm/amd/display/dc/dpp/dcn10/CMakeLists.txt b/drivers/gpu/drm/amd/display/dc/dpp/dcn10/CMakeLists.txt new file mode 100644 index 000000000000..1318c6fba3e7 --- /dev/null +++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn10/CMakeLists.txt @@ -0,0 +1,6 @@ +dal3_subdirectory_sources( + dcn10_dpp.c + dcn10_dpp_cm.c + dcn10_dpp_dscl.c + dcn10_dpp.h +) diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c b/drivers/gpu/drm/amd/display/dc/dpp/dcn10/dcn10_dpp.c index 4e391fd1d71c..e1da48b05d00 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c +++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn10/dcn10_dpp.c @@ -28,7 +28,7 @@ #include "core_types.h" #include "reg_helper.h" -#include "dcn10_dpp.h" +#include "dcn10/dcn10_dpp.h" #include "basics/conversion.h" #define NUM_PHASES 64 diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h b/drivers/gpu/drm/amd/display/dc/dpp/dcn10/dcn10_dpp.h index a039eedc7c24..c48139bed11f 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h +++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn10/dcn10_dpp.h @@ -1090,7 +1090,8 @@ type DPP_CLOCK_ENABLE; \ type CM_HDR_MULT_COEF; \ type CUR0_FP_BIAS; \ - type CUR0_FP_SCALE; + type CUR0_FP_SCALE;\ + type DISPCLK_R_GATE_DISABLE; struct dcn_dpp_shift { TF_REG_FIELD_LIST(uint8_t) diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_cm.c b/drivers/gpu/drm/amd/display/dc/dpp/dcn10/dcn10_dpp_cm.c index 2f994a3a0b9c..006e23842016 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_cm.c +++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn10/dcn10_dpp_cm.c @@ -28,9 +28,9 @@ #include "core_types.h" #include "reg_helper.h" -#include "dcn10_dpp.h" +#include "dcn10/dcn10_dpp.h" #include "basics/conversion.h" -#include "dcn10_cm_common.h" +#include "dcn10/dcn10_cm_common.h" #define NUM_PHASES 64 #define HORZ_MAX_TAPS 8 @@ -234,7 +234,7 @@ void dpp1_cm_get_gamut_remap(struct dpp *dpp_base, struct dpp_grph_csc_adjustment *adjust) { struct dcn10_dpp *dpp = TO_DCN10_DPP(dpp_base); - uint16_t arr_reg_val[12]; + uint16_t arr_reg_val[12] = {0}; enum gamut_remap_select select; read_gamut_remap(dpp, arr_reg_val, &select); diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_dscl.c b/drivers/gpu/drm/amd/display/dc/dpp/dcn10/dcn10_dpp_dscl.c index 5ca9ab8a76e8..808bca9fb804 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_dscl.c +++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn10/dcn10_dpp_dscl.c @@ -28,7 +28,7 @@ #include "core_types.h" #include "reg_helper.h" -#include "dcn10_dpp.h" +#include "dcn10/dcn10_dpp.h" #include "basics/conversion.h" diff --git a/drivers/gpu/drm/amd/display/dc/dpp/dcn20/CMakeLists.txt b/drivers/gpu/drm/amd/display/dc/dpp/dcn20/CMakeLists.txt new file mode 100644 index 000000000000..9c2d7096348e --- /dev/null +++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn20/CMakeLists.txt @@ -0,0 +1,5 @@ +dal3_subdirectory_sources( + dcn20_dpp.c + dcn20_dpp_cm.c + dcn20_dpp.h +) diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dpp.c b/drivers/gpu/drm/amd/display/dc/dpp/dcn20/dcn20_dpp.c index 1516c0a48726..56ebd7164dd7 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dpp.c +++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn20/dcn20_dpp.c @@ -28,7 +28,7 @@ #include "core_types.h" #include "reg_helper.h" -#include "dcn20_dpp.h" +#include "dcn20/dcn20_dpp.h" #include "basics/conversion.h" #define NUM_PHASES 64 diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dpp.h b/drivers/gpu/drm/amd/display/dc/dpp/dcn20/dcn20_dpp.h index 672cde46c4b9..49cb25c9cb36 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dpp.h +++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn20/dcn20_dpp.h @@ -736,7 +736,7 @@ bool dpp20_program_shaper( bool dpp20_program_3dlut( struct dpp *dpp_base, - struct tetrahedral_params *params); + const struct tetrahedral_params *params); void dpp2_cnv_set_alpha_keyer( struct dpp *dpp_base, diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dpp_cm.c b/drivers/gpu/drm/amd/display/dc/dpp/dcn20/dcn20_dpp_cm.c index 58dc69926e8a..31613372e214 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dpp_cm.c +++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn20/dcn20_dpp_cm.c @@ -28,7 +28,7 @@ #include "core_types.h" #include "reg_helper.h" -#include "dcn20_dpp.h" +#include "dcn20/dcn20_dpp.h" #include "basics/conversion.h" #include "dcn10/dcn10_cm_common.h" @@ -274,7 +274,7 @@ void dpp2_cm_get_gamut_remap(struct dpp *dpp_base, struct dpp_grph_csc_adjustment *adjust) { struct dcn20_dpp *dpp = TO_DCN20_DPP(dpp_base); - uint16_t arr_reg_val[12]; + uint16_t arr_reg_val[12] = {0}; enum dcn20_gamut_remap_select select; read_gamut_remap(dpp, arr_reg_val, &select); @@ -1114,15 +1114,15 @@ static void dpp20_select_3dlut_ram_mask( bool dpp20_program_3dlut( struct dpp *dpp_base, - struct tetrahedral_params *params) + const struct tetrahedral_params *params) { enum dc_lut_mode mode; bool is_17x17x17; bool is_12bits_color_channel; - struct dc_rgb *lut0; - struct dc_rgb *lut1; - struct dc_rgb *lut2; - struct dc_rgb *lut3; + const struct dc_rgb *lut0; + const struct dc_rgb *lut1; + const struct dc_rgb *lut2; + const struct dc_rgb *lut3; int lut_size0; int lut_size; diff --git a/drivers/gpu/drm/amd/display/dc/dpp/dcn201/CMakeLists.txt b/drivers/gpu/drm/amd/display/dc/dpp/dcn201/CMakeLists.txt new file mode 100644 index 000000000000..7711cd3c47a7 --- /dev/null +++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn201/CMakeLists.txt @@ -0,0 +1,4 @@ +dal3_subdirectory_sources( + dcn201_dpp.c + dcn201_dpp.h +) diff --git a/drivers/gpu/drm/amd/display/dc/dcn201/dcn201_dpp.c b/drivers/gpu/drm/amd/display/dc/dpp/dcn201/dcn201_dpp.c index f809a7d21033..345202fee40f 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn201/dcn201_dpp.c +++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn201/dcn201_dpp.c @@ -28,7 +28,7 @@ #include "core_types.h" #include "reg_helper.h" -#include "dcn201_dpp.h" +#include "dcn201/dcn201_dpp.h" #include "basics/conversion.h" #define REG(reg)\ diff --git a/drivers/gpu/drm/amd/display/dc/dcn201/dcn201_dpp.h b/drivers/gpu/drm/amd/display/dc/dpp/dcn201/dcn201_dpp.h index cbd5b47b4acf..cbd5b47b4acf 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn201/dcn201_dpp.h +++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn201/dcn201_dpp.h diff --git a/drivers/gpu/drm/amd/display/dc/dpp/dcn30/CMakeLists.txt b/drivers/gpu/drm/amd/display/dc/dpp/dcn30/CMakeLists.txt new file mode 100644 index 000000000000..0faee2a1e32b --- /dev/null +++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn30/CMakeLists.txt @@ -0,0 +1,5 @@ +dal3_subdirectory_sources( + dcn30_dpp.c + dcn30_dpp_cm.c + dcn30_dpp.h +) diff --git a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dpp.c b/drivers/gpu/drm/amd/display/dc/dpp/dcn30/dcn30_dpp.c index a3a769aad042..f8c0cee34080 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dpp.c +++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn30/dcn30_dpp.c @@ -26,9 +26,9 @@ #include "dm_services.h" #include "core_types.h" #include "reg_helper.h" -#include "dcn30_dpp.h" +#include "dcn30/dcn30_dpp.h" #include "basics/conversion.h" -#include "dcn30_cm_common.h" +#include "dcn30/dcn30_cm_common.h" #define REG(reg)\ dpp->tf_regs->reg @@ -293,9 +293,11 @@ void dpp3_cnv_setup ( break; case SURFACE_PIXEL_FORMAT_GRPH_RGB111110_FIX: pixel_format = 112; + alpha_en = 0; break; case SURFACE_PIXEL_FORMAT_GRPH_BGR101111_FIX: pixel_format = 113; + alpha_en = 0; break; case SURFACE_PIXEL_FORMAT_VIDEO_ACrYCb2101010: pixel_format = 114; @@ -319,9 +321,11 @@ void dpp3_cnv_setup ( break; case SURFACE_PIXEL_FORMAT_GRPH_RGB111110_FLOAT: pixel_format = 118; + alpha_en = 0; break; case SURFACE_PIXEL_FORMAT_GRPH_BGR101111_FLOAT: pixel_format = 119; + alpha_en = 0; break; default: break; @@ -1384,15 +1388,15 @@ static void dpp3_select_3dlut_ram_mask( } static bool dpp3_program_3dlut(struct dpp *dpp_base, - struct tetrahedral_params *params) + const struct tetrahedral_params *params) { enum dc_lut_mode mode; bool is_17x17x17; bool is_12bits_color_channel; - struct dc_rgb *lut0; - struct dc_rgb *lut1; - struct dc_rgb *lut2; - struct dc_rgb *lut3; + const struct dc_rgb *lut0; + const struct dc_rgb *lut1; + const struct dc_rgb *lut2; + const struct dc_rgb *lut3; int lut_size0; int lut_size; diff --git a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dpp.h b/drivers/gpu/drm/amd/display/dc/dpp/dcn30/dcn30_dpp.h index 2ac8045a87a1..269f437c1633 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dpp.h +++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn30/dcn30_dpp.h @@ -132,6 +132,8 @@ SRI(CM_POST_CSC_B_C33_C34, CM, id), \ SRI(CM_MEM_PWR_CTRL, CM, id), \ SRI(CM_CONTROL, CM, id), \ + SRI(CM_TEST_DEBUG_INDEX, CM, id), \ + SRI(CM_TEST_DEBUG_DATA, CM, id), \ SRI(FORMAT_CONTROL, CNVC_CFG, id), \ SRI(CNVC_SURFACE_PIXEL_FORMAT, CNVC_CFG, id), \ SRI(CURSOR0_CONTROL, CNVC_CUR, id), \ @@ -294,6 +296,7 @@ TF_SF(CM0_CM_POST_CSC_C11_C12, CM_POST_CSC_C12, mask_sh), \ TF_SF(CM0_CM_POST_CSC_C33_C34, CM_POST_CSC_C33, mask_sh), \ TF_SF(CM0_CM_POST_CSC_C33_C34, CM_POST_CSC_C34, mask_sh), \ + TF_SF(CM0_CM_TEST_DEBUG_INDEX, CM_TEST_DEBUG_INDEX, mask_sh), \ TF_SF(CNVC_CFG0_FORMAT_CONTROL, CNVC_BYPASS, mask_sh), \ TF2_SF(CNVC_CFG0, FORMAT_CONTROL__ALPHA_EN, mask_sh), \ TF_SF(CNVC_CFG0_FORMAT_CONTROL, FORMAT_EXPANSION_MODE, mask_sh), \ @@ -426,6 +429,7 @@ type CM_GAMCOR_LUT_DATA; \ type CM_GAMCOR_LUT_WRITE_COLOR_MASK; \ type CM_GAMCOR_LUT_READ_COLOR_SEL; \ + type CM_GAMCOR_LUT_READ_DBG; \ type CM_GAMCOR_LUT_HOST_SEL; \ type CM_GAMCOR_LUT_CONFIG_MODE; \ type CM_GAMCOR_LUT_STATUS; \ diff --git a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dpp_cm.c b/drivers/gpu/drm/amd/display/dc/dpp/dcn30/dcn30_dpp_cm.c index 2f5b3fbd3507..82eca0e7b7d0 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dpp_cm.c +++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn30/dcn30_dpp_cm.c @@ -26,9 +26,9 @@ #include "dm_services.h" #include "core_types.h" #include "reg_helper.h" -#include "dcn30_dpp.h" +#include "dcn30/dcn30_dpp.h" #include "basics/conversion.h" -#include "dcn30_cm_common.h" +#include "dcn30/dcn30_cm_common.h" #define REG(reg)\ dpp->tf_regs->reg @@ -445,7 +445,7 @@ void dpp3_cm_get_gamut_remap(struct dpp *dpp_base, struct dpp_grph_csc_adjustment *adjust) { struct dcn3_dpp *dpp = TO_DCN30_DPP(dpp_base); - uint16_t arr_reg_val[12]; + uint16_t arr_reg_val[12] = {0}; int select; read_gamut_remap(dpp, arr_reg_val, &select); diff --git a/drivers/gpu/drm/amd/display/dc/dpp/dcn32/CMakeLists.txt b/drivers/gpu/drm/amd/display/dc/dpp/dcn32/CMakeLists.txt new file mode 100644 index 000000000000..7743edc4599f --- /dev/null +++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn32/CMakeLists.txt @@ -0,0 +1,4 @@ +dal3_subdirectory_sources( + dcn32_dpp.c + dcn32_dpp.h +) diff --git a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dpp.c b/drivers/gpu/drm/amd/display/dc/dpp/dcn32/dcn32_dpp.c index 681e75c6dbaf..41679997b44d 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dpp.c +++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn32/dcn32_dpp.c @@ -26,7 +26,7 @@ #include "dm_services.h" #include "core_types.h" #include "reg_helper.h" -#include "dcn32_dpp.h" +#include "dcn32/dcn32_dpp.h" #include "basics/conversion.h" #include "dcn30/dcn30_cm_common.h" diff --git a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dpp.h b/drivers/gpu/drm/amd/display/dc/dpp/dcn32/dcn32_dpp.h index 572958d287eb..572958d287eb 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dpp.h +++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn32/dcn32_dpp.h diff --git a/drivers/gpu/drm/amd/display/dc/dpp/dcn35/CMakeLists.txt b/drivers/gpu/drm/amd/display/dc/dpp/dcn35/CMakeLists.txt new file mode 100644 index 000000000000..91df5db26435 --- /dev/null +++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn35/CMakeLists.txt @@ -0,0 +1,4 @@ +dal3_subdirectory_sources( + dcn35_dpp.c + dcn35_dpp.h +) diff --git a/drivers/gpu/drm/amd/display/dc/dpp/dcn35/dcn35_dpp.c b/drivers/gpu/drm/amd/display/dc/dpp/dcn35/dcn35_dpp.c new file mode 100644 index 000000000000..e16274fee31d --- /dev/null +++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn35/dcn35_dpp.c @@ -0,0 +1,112 @@ +/* SPDX-License-Identifier: MIT */ +/* + * Copyright 2023 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 "core_types.h" +#include "dcn35/dcn35_dpp.h" +#include "reg_helper.h" + +#define REG(reg) dpp->tf_regs->reg + +#define CTX dpp->base.ctx + +#undef FN +#define FN(reg_name, field_name) \ + ((const struct dcn35_dpp_shift *)(dpp->tf_shift))->field_name, \ + ((const struct dcn35_dpp_mask *)(dpp->tf_mask))->field_name + +void dpp35_dppclk_control( + struct dpp *dpp_base, + bool dppclk_div, + bool enable) +{ + struct dcn20_dpp *dpp = TO_DCN20_DPP(dpp_base); + + if (enable) { + if (dpp->tf_mask->DPPCLK_RATE_CONTROL) + REG_UPDATE_2(DPP_CONTROL, + DPPCLK_RATE_CONTROL, dppclk_div, + DPP_CLOCK_ENABLE, 1); + else + REG_UPDATE_2(DPP_CONTROL, + DPP_CLOCK_ENABLE, 1, + DISPCLK_R_GATE_DISABLE, 1); + } else + REG_UPDATE_2(DPP_CONTROL, + DPP_CLOCK_ENABLE, 0, + DISPCLK_R_GATE_DISABLE, 0); +} + +static struct dpp_funcs dcn35_dpp_funcs = { + .dpp_program_gamcor_lut = dpp3_program_gamcor_lut, + .dpp_read_state = dpp30_read_state, + .dpp_reset = dpp_reset, + .dpp_set_scaler = dpp1_dscl_set_scaler_manual_scale, + .dpp_get_optimal_number_of_taps = dpp3_get_optimal_number_of_taps, + .dpp_set_gamut_remap = dpp3_cm_set_gamut_remap, + .dpp_set_csc_adjustment = NULL, + .dpp_set_csc_default = NULL, + .dpp_program_regamma_pwl = NULL, + .dpp_set_pre_degam = dpp3_set_pre_degam, + .dpp_program_input_lut = NULL, + .dpp_full_bypass = dpp1_full_bypass, + .dpp_setup = dpp3_cnv_setup, + .dpp_program_degamma_pwl = NULL, + .dpp_program_cm_dealpha = dpp3_program_cm_dealpha, + .dpp_program_cm_bias = dpp3_program_cm_bias, + + .dpp_program_blnd_lut = NULL, // BLNDGAM is removed completely in DCN3.2 DPP + .dpp_program_shaper_lut = NULL, // CM SHAPER block is removed in DCN3.2 DPP, (it is in MPCC, programmable before or after BLND) + .dpp_program_3dlut = NULL, // CM 3DLUT block is removed in DCN3.2 DPP, (it is in MPCC, programmable before or after BLND) + + .dpp_program_bias_and_scale = NULL, + .dpp_cnv_set_alpha_keyer = dpp2_cnv_set_alpha_keyer, + .set_cursor_attributes = dpp3_set_cursor_attributes, + .set_cursor_position = dpp1_set_cursor_position, + .set_optional_cursor_attributes = dpp1_cnv_set_optional_cursor_attributes, + .dpp_dppclk_control = dpp35_dppclk_control, + .dpp_set_hdr_multiplier = dpp3_set_hdr_multiplier, + .dpp_get_gamut_remap = dpp3_cm_get_gamut_remap, +}; + + +bool dpp35_construct( + struct dcn3_dpp *dpp, struct dc_context *ctx, + uint32_t inst, const struct dcn3_dpp_registers *tf_regs, + const struct dcn35_dpp_shift *tf_shift, + const struct dcn35_dpp_mask *tf_mask) +{ + bool ret = dpp32_construct(dpp, ctx, inst, tf_regs, + (const struct dcn3_dpp_shift *)(tf_shift), + (const struct dcn3_dpp_mask *)(tf_mask)); + + dpp->base.funcs = &dcn35_dpp_funcs; + return ret; +} + +void dpp35_set_fgcg(struct dcn3_dpp *dpp, bool enable) +{ + REG_UPDATE(DPP_CONTROL, DPP_FGCG_REP_DIS, !enable); +} diff --git a/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_dpp.h b/drivers/gpu/drm/amd/display/dc/dpp/dcn35/dcn35_dpp.h index 09b84307cd9e..135872d88219 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_dpp.h +++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn35/dcn35_dpp.h @@ -31,7 +31,9 @@ #define DPP_REG_LIST_SH_MASK_DCN35(mask_sh) \ DPP_REG_LIST_SH_MASK_DCN30_COMMON(mask_sh), \ - TF_SF(DPP_TOP0_DPP_CONTROL, DPP_FGCG_REP_DIS, mask_sh) + TF_SF(DPP_TOP0_DPP_CONTROL, DPP_FGCG_REP_DIS, mask_sh), \ + TF_SF(DPP_TOP0_DPP_CONTROL, DPP_FGCG_REP_DIS, mask_sh), \ + TF_SF(DPP_TOP0_DPP_CONTROL, DISPCLK_R_GATE_DISABLE, mask_sh) #define DPP_REG_FIELD_LIST_DCN35(type) \ struct { \ @@ -47,6 +49,11 @@ struct dcn35_dpp_mask { DPP_REG_FIELD_LIST_DCN35(uint32_t); }; +void dpp35_dppclk_control( + struct dpp *dpp_base, + bool dppclk_div, + bool enable); + bool dpp35_construct(struct dcn3_dpp *dpp3, struct dc_context *ctx, uint32_t inst, const struct dcn3_dpp_registers *tf_regs, const struct dcn35_dpp_shift *tf_shift, diff --git a/drivers/gpu/drm/amd/display/dc/dsc/dc_dsc.c b/drivers/gpu/drm/amd/display/dc/dsc/dc_dsc.c index ac41f9c0a283..150ef23440a2 100644 --- a/drivers/gpu/drm/amd/display/dc/dsc/dc_dsc.c +++ b/drivers/gpu/drm/amd/display/dc/dsc/dc_dsc.c @@ -458,7 +458,7 @@ bool dc_dsc_compute_bandwidth_range( bool is_dsc_possible = false; struct dsc_enc_caps dsc_enc_caps; struct dsc_enc_caps dsc_common_caps; - struct dc_dsc_config config; + struct dc_dsc_config config = {0}; struct dc_dsc_config_options options = {0}; options.dsc_min_slice_height_override = dsc_min_slice_height_override; @@ -868,9 +868,9 @@ static bool setup_dsc_config( struct dc_dsc_config *dsc_cfg) { struct dsc_enc_caps dsc_common_caps; - int max_slices_h; - int min_slices_h; - int num_slices_h; + int max_slices_h = 0; + int min_slices_h = 0; + int num_slices_h = 0; int pic_width; int slice_width; int target_bpp; @@ -1055,7 +1055,12 @@ static bool setup_dsc_config( if (!is_dsc_possible) goto done; - dsc_cfg->num_slices_v = pic_height/slice_height; + if (slice_height > 0) { + dsc_cfg->num_slices_v = pic_height / slice_height; + } else { + is_dsc_possible = false; + goto done; + } if (target_bandwidth_kbps > 0) { is_dsc_possible = decide_dsc_target_bpp_x16( diff --git a/drivers/gpu/drm/amd/display/dc/dsc/rc_calc_dpi.c b/drivers/gpu/drm/amd/display/dc/dsc/rc_calc_dpi.c index 36d6c1646a51..59864130cf83 100644 --- a/drivers/gpu/drm/amd/display/dc/dsc/rc_calc_dpi.c +++ b/drivers/gpu/drm/amd/display/dc/dsc/rc_calc_dpi.c @@ -101,7 +101,6 @@ int dscc_compute_dsc_parameters(const struct drm_dsc_config *pps, { int ret; struct drm_dsc_config dsc_cfg; - unsigned long long tmp; dsc_params->pps = *pps; dsc_params->pps.initial_scale_value = 8 * rc->rc_model_size / (rc->rc_model_size - rc->initial_fullness_offset); @@ -112,9 +111,9 @@ int dscc_compute_dsc_parameters(const struct drm_dsc_config *pps, dsc_cfg.mux_word_size = dsc_params->pps.bits_per_component <= 10 ? 48 : 64; ret = drm_dsc_compute_rc_parameters(&dsc_cfg); - tmp = (unsigned long long)dsc_cfg.slice_chunk_size * 0x10000000 + (dsc_cfg.slice_width - 1); - do_div(tmp, (uint32_t)dsc_cfg.slice_width); //ROUND-UP - dsc_params->bytes_per_pixel = (uint32_t)tmp; + dsc_params->bytes_per_pixel = + (uint32_t)(div_u64(((uint64_t)dsc_cfg.slice_chunk_size * 0x10000000 + (dsc_cfg.slice_width - 1)), + (uint32_t)dsc_cfg.slice_width)); /* Round-up */ copy_pps_fields(&dsc_params->pps, &dsc_cfg); dsc_params->rc_buffer_model_size = dsc_cfg.rc_bits; diff --git a/drivers/gpu/drm/amd/display/dc/gpio/dcn21/hw_translate_dcn21.c b/drivers/gpu/drm/amd/display/dc/gpio/dcn21/hw_translate_dcn21.c index d734e3a134d1..2840ed5c57d8 100644 --- a/drivers/gpu/drm/amd/display/dc/gpio/dcn21/hw_translate_dcn21.c +++ b/drivers/gpu/drm/amd/display/dc/gpio/dcn21/hw_translate_dcn21.c @@ -95,10 +95,6 @@ static bool offset_to_id( return true; default: ASSERT_CRITICAL(false); -#ifdef PALLADIUM_SUPPORTED - *en = GPIO_DDC_LINE_DDC1; - return true; -#endif return false; } break; @@ -184,11 +180,6 @@ static bool offset_to_id( /* UNEXPECTED */ default: /* case REG(DC_GPIO_SYNCA_A): not exista */ -#ifdef PALLADIUM_SUPPORTED - *id = GPIO_ID_HPD; - *en = GPIO_DDC_LINE_DDC1; - return true; -#endif ASSERT_CRITICAL(false); return false; } @@ -308,10 +299,6 @@ static bool id_to_offset( break; default: ASSERT_CRITICAL(false); -#ifdef PALLADIUM_SUPPORTED - info->mask = DC_GPIO_HPD_A__DC_GPIO_HPD1_A_MASK; - result = true; -#endif result = false; } break; diff --git a/drivers/gpu/drm/amd/display/dc/gpio/gpio_service.c b/drivers/gpu/drm/amd/display/dc/gpio/gpio_service.c index 3ede6e02c3a7..663c17f52779 100644 --- a/drivers/gpu/drm/amd/display/dc/gpio/gpio_service.c +++ b/drivers/gpu/drm/amd/display/dc/gpio/gpio_service.c @@ -128,7 +128,7 @@ struct gpio *dal_gpio_service_create_irq( uint32_t offset, uint32_t mask) { - enum gpio_id id; + enum gpio_id id = 0; uint32_t en; if (!service->translate.funcs->offset_to_id(offset, mask, &id, &en)) { @@ -144,7 +144,7 @@ struct gpio *dal_gpio_service_create_generic_mux( uint32_t offset, uint32_t mask) { - enum gpio_id id; + enum gpio_id id = 0; uint32_t en; struct gpio *generic; @@ -178,7 +178,7 @@ struct gpio_pin_info dal_gpio_get_generic_pin_info( enum gpio_id id, uint32_t en) { - struct gpio_pin_info pin; + struct gpio_pin_info pin = {0}; if (service->translate.funcs->id_to_offset) { service->translate.funcs->id_to_offset(id, en, &pin); diff --git a/drivers/gpu/drm/amd/display/dc/hwss/Makefile b/drivers/gpu/drm/amd/display/dc/hwss/Makefile index 9e8e9de51a92..cf8aa23b4415 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/Makefile +++ b/drivers/gpu/drm/amd/display/dc/hwss/Makefile @@ -180,7 +180,7 @@ AMD_DISPLAY_FILES += $(AMD_DAL_HWSS_DCN35) ############################################################################### -HWSS_DCN351 = dcn351_init.o +HWSS_DCN351 = dcn351_hwseq.o dcn351_init.o AMD_DAL_HWSS_DCN351 = $(addprefix $(AMDDALPATH)/dc/hwss/dcn351/,$(HWSS_DCN351)) diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dce110/dce110_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dce110/dce110_hwseq.c index 9d5df4c0da59..0d3ea291eeee 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dce110/dce110_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dce110/dce110_hwseq.c @@ -249,7 +249,7 @@ static bool dce110_enable_display_power_gating( return false; } -static void build_prescale_params(struct ipp_prescale_params *prescale_params, +static void dce110_prescale_params(struct ipp_prescale_params *prescale_params, const struct dc_plane_state *plane_state) { prescale_params->mode = IPP_PRESCALE_MODE_FIXED_UNSIGNED; @@ -289,16 +289,14 @@ dce110_set_input_transfer_func(struct dc *dc, struct pipe_ctx *pipe_ctx, if (ipp == NULL) return false; - if (plane_state->in_transfer_func) - tf = plane_state->in_transfer_func; + tf = &plane_state->in_transfer_func; - build_prescale_params(&prescale_params, plane_state); + dce110_prescale_params(&prescale_params, plane_state); ipp->funcs->ipp_program_prescale(ipp, &prescale_params); - if (plane_state->gamma_correction && - !plane_state->gamma_correction->is_identity && + if (!plane_state->gamma_correction.is_identity && dce_use_lut(plane_state->format)) - ipp->funcs->ipp_program_input_lut(ipp, plane_state->gamma_correction); + ipp->funcs->ipp_program_input_lut(ipp, &plane_state->gamma_correction); if (tf == NULL) { /* Default case if no input transfer function specified */ @@ -614,11 +612,10 @@ dce110_set_output_transfer_func(struct dc *dc, struct pipe_ctx *pipe_ctx, xfm->funcs->opp_power_on_regamma_lut(xfm, true); xfm->regamma_params.hw_points_num = GAMMA_HW_POINTS_NUM; - if (stream->out_transfer_func && - stream->out_transfer_func->type == TF_TYPE_PREDEFINED && - stream->out_transfer_func->tf == TRANSFER_FUNCTION_SRGB) { + if (stream->out_transfer_func.type == TF_TYPE_PREDEFINED && + stream->out_transfer_func.tf == TRANSFER_FUNCTION_SRGB) { xfm->funcs->opp_set_regamma_mode(xfm, OPP_REGAMMA_SRGB); - } else if (dce110_translate_regamma_to_hw_format(stream->out_transfer_func, + } else if (dce110_translate_regamma_to_hw_format(&stream->out_transfer_func, &xfm->regamma_params)) { xfm->funcs->opp_program_regamma_pwl(xfm, &xfm->regamma_params); xfm->funcs->opp_set_regamma_mode(xfm, OPP_REGAMMA_USER); @@ -1185,22 +1182,13 @@ void dce110_disable_stream(struct pipe_ctx *pipe_ctx) if (dccg) { dccg->funcs->disable_symclk32_se(dccg, dp_hpo_inst); dccg->funcs->set_dpstreamclk(dccg, REFCLK, tg->inst, dp_hpo_inst); - dccg->funcs->set_dtbclk_dto(dccg, &dto_params); + if (dccg && dccg->funcs->set_dtbclk_dto) + dccg->funcs->set_dtbclk_dto(dccg, &dto_params); } } else if (dccg && dccg->funcs->disable_symclk_se) { dccg->funcs->disable_symclk_se(dccg, stream_enc->stream_enc_inst, link_enc->transmitter - TRANSMITTER_UNIPHY_A); } - - if (dc->link_srv->dp_is_128b_132b_signal(pipe_ctx)) { - /* TODO: This looks like a bug to me as we are disabling HPO IO when - * we are just disabling a single HPO stream. Shouldn't we disable HPO - * HW control only when HPOs for all streams are disabled? - */ - if (pipe_ctx->stream->ctx->dc->hwseq->funcs.setup_hpo_hw_control) - pipe_ctx->stream->ctx->dc->hwseq->funcs.setup_hpo_hw_control( - pipe_ctx->stream->ctx->dc->hwseq, false); - } } void dce110_unblank_stream(struct pipe_ctx *pipe_ctx, @@ -1549,7 +1537,7 @@ enum dc_status dce110_apply_single_controller_ctx_to_hw( } if (pipe_ctx->stream_res.audio != NULL) { - struct audio_output audio_output; + struct audio_output audio_output = {0}; build_audio_output(context, pipe_ctx, &audio_output); @@ -2200,7 +2188,7 @@ static void dce110_setup_audio_dto( struct dc *dc, struct dc_state *context) { - int i; + unsigned int i; /* program audio wall clock. use HDMI as clock source if HDMI * audio active. Otherwise, use DP as clock source @@ -2272,7 +2260,7 @@ static void dce110_setup_audio_dto( continue; if (pipe_ctx->stream_res.audio != NULL) { - struct audio_output audio_output; + struct audio_output audio_output = {0}; build_audio_output(context, pipe_ctx, &audio_output); @@ -2287,6 +2275,19 @@ static void dce110_setup_audio_dto( } } +static bool dce110_is_hpo_enabled(struct dc_state *context) +{ + int i; + + for (i = 0; i < MAX_HPO_DP2_ENCODERS; i++) { + if (context->res_ctx.is_hpo_dp_stream_enc_acquired[i]) { + return true; + } + } + + return false; +} + enum dc_status dce110_apply_ctx_to_hw( struct dc *dc, struct dc_state *context) @@ -2295,6 +2296,8 @@ enum dc_status dce110_apply_ctx_to_hw( struct dc_bios *dcb = dc->ctx->dc_bios; enum dc_status status; int i; + bool was_hpo_enabled = dce110_is_hpo_enabled(dc->current_state); + bool is_hpo_enabled = dce110_is_hpo_enabled(context); /* reset syncd pipes from disabled pipes */ if (dc->config.use_pipe_ctx_sync_logic) @@ -2337,6 +2340,10 @@ enum dc_status dce110_apply_ctx_to_hw( dce110_setup_audio_dto(dc, context); + if (dc->hwseq->funcs.setup_hpo_hw_control && was_hpo_enabled != is_hpo_enabled) { + dc->hwseq->funcs.setup_hpo_hw_control(dc->hwseq, is_hpo_enabled); + } + for (i = 0; i < dc->res_pool->pipe_count; i++) { struct pipe_ctx *pipe_ctx_old = &dc->current_state->res_ctx.pipe_ctx[i]; diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn10/dcn10_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn10/dcn10_hwseq.c index 314798400b16..0c4aef8ffe2c 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn10/dcn10_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn10/dcn10_hwseq.c @@ -367,7 +367,7 @@ static void dcn10_log_color_state(struct dc *dc, dc->caps.color.dpp.ocsc); DTN_INFO("MPCC: OPP DPP MPCCBOT MODE ALPHA_MODE PREMULT OVERLAP_ONLY IDLE\n"); - for (i = 0; i < pool->pipe_count; i++) { + for (i = 0; i < pool->mpcc_count; i++) { struct mpcc_state s = {0}; pool->mpc->funcs->read_mpcc_state(pool->mpc, i, &s); @@ -1366,6 +1366,7 @@ void dcn10_init_pipes(struct dc *dc, struct dc_state *context) struct dce_hwseq *hws = dc->hwseq; struct hubbub *hubbub = dc->res_pool->hubbub; bool can_apply_seamless_boot = false; + bool tg_enabled[MAX_PIPES] = {false}; for (i = 0; i < context->stream_count; i++) { if (context->streams[i]->apply_seamless_boot_optimization) { @@ -1447,6 +1448,7 @@ void dcn10_init_pipes(struct dc *dc, struct dc_state *context) // requesting data while in PSR. tg->funcs->tg_init(tg); hubp->power_gated = true; + tg_enabled[i] = true; continue; } @@ -1488,6 +1490,20 @@ void dcn10_init_pipes(struct dc *dc, struct dc_state *context) tg->funcs->tg_init(tg); } + /* Clean up MPC tree */ + for (i = 0; i < dc->res_pool->pipe_count; i++) { + if (tg_enabled[i]) { + if (dc->res_pool->opps[i]->mpc_tree_params.opp_list) { + if (dc->res_pool->opps[i]->mpc_tree_params.opp_list->mpcc_bot) { + int bot_id = dc->res_pool->opps[i]->mpc_tree_params.opp_list->mpcc_bot->mpcc_id; + + if ((bot_id < MAX_MPCC) && (bot_id < MAX_PIPES) && (!tg_enabled[bot_id])) + dc->res_pool->opps[i]->mpc_tree_params.opp_list = NULL; + } + } + } + } + /* Power gate DSCs */ if (hws->funcs.dsc_pg_control != NULL) { uint32_t num_opps = 0; @@ -1813,14 +1829,12 @@ bool dcn10_set_input_transfer_func(struct dc *dc, struct pipe_ctx *pipe_ctx, if (dpp_base == NULL) return false; - if (plane_state->in_transfer_func) - tf = plane_state->in_transfer_func; + tf = &plane_state->in_transfer_func; - if (plane_state->gamma_correction && - !dpp_base->ctx->dc->debug.always_use_regamma - && !plane_state->gamma_correction->is_identity + if (!dpp_base->ctx->dc->debug.always_use_regamma + && !plane_state->gamma_correction.is_identity && dce_use_lut(plane_state->format)) - dpp_base->funcs->dpp_program_input_lut(dpp_base, plane_state->gamma_correction); + dpp_base->funcs->dpp_program_input_lut(dpp_base, &plane_state->gamma_correction); if (tf == NULL) dpp_base->funcs->dpp_set_degamma(dpp_base, IPP_DEGAMMA_MODE_BYPASS); @@ -1861,7 +1875,7 @@ bool dcn10_set_input_transfer_func(struct dc *dc, struct pipe_ctx *pipe_ctx, #define MAX_NUM_HW_POINTS 0x200 static void log_tf(struct dc_context *ctx, - struct dc_transfer_func *tf, uint32_t hw_points_num) + const struct dc_transfer_func *tf, uint32_t hw_points_num) { // DC_LOG_GAMMA is default logging of all hw points // DC_LOG_ALL_GAMMA logs all points, not only hw points @@ -1898,16 +1912,15 @@ bool dcn10_set_output_transfer_func(struct dc *dc, struct pipe_ctx *pipe_ctx, dpp->regamma_params.hw_points_num = GAMMA_HW_POINTS_NUM; - if (stream->out_transfer_func && - stream->out_transfer_func->type == TF_TYPE_PREDEFINED && - stream->out_transfer_func->tf == TRANSFER_FUNCTION_SRGB) + if (stream->out_transfer_func.type == TF_TYPE_PREDEFINED && + stream->out_transfer_func.tf == TRANSFER_FUNCTION_SRGB) dpp->funcs->dpp_program_regamma_pwl(dpp, NULL, OPP_REGAMMA_SRGB); /* dcn10_translate_regamma_to_hw_format takes 750us, only do it when full * update. */ else if (cm_helper_translate_curve_to_hw_format(dc->ctx, - stream->out_transfer_func, + &stream->out_transfer_func, &dpp->regamma_params, false)) { dpp->funcs->dpp_program_regamma_pwl( dpp, @@ -1915,10 +1928,9 @@ bool dcn10_set_output_transfer_func(struct dc *dc, struct pipe_ctx *pipe_ctx, } else dpp->funcs->dpp_program_regamma_pwl(dpp, NULL, OPP_REGAMMA_BYPASS); - if (stream->ctx && - stream->out_transfer_func) { + if (stream->ctx) { log_tf(stream->ctx, - stream->out_transfer_func, + &stream->out_transfer_func, dpp->regamma_params.hw_points_num); } @@ -2173,7 +2185,7 @@ static int dcn10_align_pixel_clocks(struct dc *dc, int group_size, struct dc_crtc_timing *hw_crtc_timing; uint64_t phase[MAX_PIPES]; uint64_t modulo[MAX_PIPES]; - unsigned int pclk; + unsigned int pclk = 0; uint32_t embedded_pix_clk_100hz; uint16_t embedded_h_total; @@ -2264,7 +2276,7 @@ void dcn10_enable_vblanks_synchronization( struct dc_context *dc_ctx = dc->ctx; struct output_pixel_processor *opp; struct timing_generator *tg; - int i, width, height, master; + int i, width = 0, height = 0, master; DC_LOGGER_INIT(dc_ctx->logger); @@ -2330,7 +2342,7 @@ void dcn10_enable_timing_synchronization( struct dc_context *dc_ctx = dc->ctx; struct output_pixel_processor *opp; struct timing_generator *tg; - int i, width, height; + int i, width = 0, height = 0; DC_LOGGER_INIT(dc_ctx->logger); diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn20/dcn20_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn20/dcn20_hwseq.c index 8b3536c380b8..7d833fa6dd77 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn20/dcn20_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn20/dcn20_hwseq.c @@ -155,7 +155,7 @@ void dcn20_log_color_state(struct dc *dc, DTN_INFO("MPCC: OPP DPP MPCCBOT MODE ALPHA_MODE PREMULT OVERLAP_ONLY IDLE" " OGAM mode\n"); - for (i = 0; i < pool->pipe_count; i++) { + for (i = 0; i < pool->mpcc_count; i++) { struct mpcc_state s = {0}; pool->mpc->funcs->read_mpcc_state(pool->mpc, i, &s); @@ -403,7 +403,7 @@ void dcn20_init_blank( struct output_pixel_processor *opp = NULL; struct output_pixel_processor *bottom_opp = NULL; uint32_t num_opps, opp_id_src0, opp_id_src1; - uint32_t otg_active_width, otg_active_height; + uint32_t otg_active_width = 0, otg_active_height = 0; /* program opp dpg blank color */ color_space = COLOR_SPACE_SRGB; @@ -873,6 +873,22 @@ enum dc_status dcn20_enable_stream_timing( return DC_ERROR_UNEXPECTED; } + if (dc->link_srv->dp_is_128b_132b_signal(pipe_ctx)) { + struct dccg *dccg = dc->res_pool->dccg; + struct timing_generator *tg = pipe_ctx->stream_res.tg; + struct dtbclk_dto_params dto_params = {0}; + + if (dccg->funcs->set_dtbclk_p_src) + dccg->funcs->set_dtbclk_p_src(dccg, DTBCLK0, tg->inst); + + dto_params.otg_inst = tg->inst; + dto_params.pixclk_khz = pipe_ctx->stream->timing.pix_clk_100hz / 10; + dto_params.num_odm_segments = get_odm_segment_count(pipe_ctx); + dto_params.timing = &pipe_ctx->stream->timing; + dto_params.ref_dtbclk_khz = dc->clk_mgr->funcs->get_dtb_ref_clk_frequency(dc->clk_mgr); + dccg->funcs->set_dtbclk_dto(dccg, &dto_params); + } + if (dc_is_hdmi_tmds_signal(stream->signal)) { stream->link->phy_state.symclk_ref_cnts.otg = 1; if (stream->link->phy_state.symclk_state == SYMCLK_OFF_TX_OFF) @@ -959,22 +975,6 @@ enum dc_status dcn20_enable_stream_timing( pipe_ctx->stream_res.tg->funcs->phantom_crtc_post_enable(pipe_ctx->stream_res.tg); } - if (dc->link_srv->dp_is_128b_132b_signal(pipe_ctx)) { - struct dccg *dccg = dc->res_pool->dccg; - struct timing_generator *tg = pipe_ctx->stream_res.tg; - struct dtbclk_dto_params dto_params = {0}; - - if (dccg->funcs->set_dtbclk_p_src) - dccg->funcs->set_dtbclk_p_src(dccg, DTBCLK0, tg->inst); - - dto_params.otg_inst = tg->inst; - dto_params.pixclk_khz = pipe_ctx->stream->timing.pix_clk_100hz / 10; - dto_params.num_odm_segments = get_odm_segment_count(pipe_ctx); - dto_params.timing = &pipe_ctx->stream->timing; - dto_params.ref_dtbclk_khz = dc->clk_mgr->funcs->get_dtb_ref_clk_frequency(dc->clk_mgr); - dccg->funcs->set_dtbclk_dto(dccg, &dto_params); - } - return DC_OK; } @@ -1011,7 +1011,7 @@ bool dcn20_set_output_transfer_func(struct dc *dc, struct pipe_ctx *pipe_ctx, { int mpcc_id = pipe_ctx->plane_res.hubp->inst; struct mpc *mpc = pipe_ctx->stream_res.opp->ctx->dc->res_pool->mpc; - struct pwl_params *params = NULL; + const struct pwl_params *params = NULL; /* * program OGAM only for the top pipe * if there is a pipe split then fix diagnostic is required: @@ -1022,19 +1022,19 @@ bool dcn20_set_output_transfer_func(struct dc *dc, struct pipe_ctx *pipe_ctx, if (mpc->funcs->power_on_mpc_mem_pwr) mpc->funcs->power_on_mpc_mem_pwr(mpc, mpcc_id, true); if (pipe_ctx->top_pipe == NULL - && mpc->funcs->set_output_gamma && stream->out_transfer_func) { - if (stream->out_transfer_func->type == TF_TYPE_HWPWL) - params = &stream->out_transfer_func->pwl; - else if (pipe_ctx->stream->out_transfer_func->type == + && mpc->funcs->set_output_gamma) { + if (stream->out_transfer_func.type == TF_TYPE_HWPWL) + params = &stream->out_transfer_func.pwl; + else if (pipe_ctx->stream->out_transfer_func.type == TF_TYPE_DISTRIBUTED_POINTS && cm_helper_translate_curve_to_hw_format(dc->ctx, - stream->out_transfer_func, + &stream->out_transfer_func, &mpc->blender_params, false)) params = &mpc->blender_params; /* * there is no ROM */ - if (stream->out_transfer_func->type == TF_TYPE_PREDEFINED) + if (stream->out_transfer_func.type == TF_TYPE_PREDEFINED) BREAK_TO_DEBUGGER(); } /* @@ -1050,17 +1050,15 @@ bool dcn20_set_blend_lut( { struct dpp *dpp_base = pipe_ctx->plane_res.dpp; bool result = true; - struct pwl_params *blend_lut = NULL; - - if (plane_state->blend_tf) { - if (plane_state->blend_tf->type == TF_TYPE_HWPWL) - blend_lut = &plane_state->blend_tf->pwl; - else if (plane_state->blend_tf->type == TF_TYPE_DISTRIBUTED_POINTS) { - cm_helper_translate_curve_to_hw_format(plane_state->ctx, - plane_state->blend_tf, - &dpp_base->regamma_params, false); - blend_lut = &dpp_base->regamma_params; - } + const struct pwl_params *blend_lut = NULL; + + if (plane_state->blend_tf.type == TF_TYPE_HWPWL) + blend_lut = &plane_state->blend_tf.pwl; + else if (plane_state->blend_tf.type == TF_TYPE_DISTRIBUTED_POINTS) { + cm_helper_translate_curve_to_hw_format(plane_state->ctx, + &plane_state->blend_tf, + &dpp_base->regamma_params, false); + blend_lut = &dpp_base->regamma_params; } result = dpp_base->funcs->dpp_program_blnd_lut(dpp_base, blend_lut); @@ -1072,24 +1070,21 @@ bool dcn20_set_shaper_3dlut( { struct dpp *dpp_base = pipe_ctx->plane_res.dpp; bool result = true; - struct pwl_params *shaper_lut = NULL; - - if (plane_state->in_shaper_func) { - if (plane_state->in_shaper_func->type == TF_TYPE_HWPWL) - shaper_lut = &plane_state->in_shaper_func->pwl; - else if (plane_state->in_shaper_func->type == TF_TYPE_DISTRIBUTED_POINTS) { - cm_helper_translate_curve_to_hw_format(plane_state->ctx, - plane_state->in_shaper_func, - &dpp_base->shaper_params, true); - shaper_lut = &dpp_base->shaper_params; - } + const struct pwl_params *shaper_lut = NULL; + + if (plane_state->in_shaper_func.type == TF_TYPE_HWPWL) + shaper_lut = &plane_state->in_shaper_func.pwl; + else if (plane_state->in_shaper_func.type == TF_TYPE_DISTRIBUTED_POINTS) { + cm_helper_translate_curve_to_hw_format(plane_state->ctx, + &plane_state->in_shaper_func, + &dpp_base->shaper_params, true); + shaper_lut = &dpp_base->shaper_params; } result = dpp_base->funcs->dpp_program_shaper_lut(dpp_base, shaper_lut); - if (plane_state->lut3d_func && - plane_state->lut3d_func->state.bits.initialized == 1) + if (plane_state->lut3d_func.state.bits.initialized == 1) result = dpp_base->funcs->dpp_program_3dlut(dpp_base, - &plane_state->lut3d_func->lut_3d); + &plane_state->lut3d_func.lut_3d); else result = dpp_base->funcs->dpp_program_3dlut(dpp_base, NULL); @@ -1112,15 +1107,7 @@ bool dcn20_set_input_transfer_func(struct dc *dc, hws->funcs.set_shaper_3dlut(pipe_ctx, plane_state); hws->funcs.set_blend_lut(pipe_ctx, plane_state); - if (plane_state->in_transfer_func) - tf = plane_state->in_transfer_func; - - - if (tf == NULL) { - dpp_base->funcs->dpp_set_degamma(dpp_base, - IPP_DEGAMMA_MODE_BYPASS); - return true; - } + tf = &plane_state->in_transfer_func; if (tf->type == TF_TYPE_HWPWL || tf->type == TF_TYPE_DISTRIBUTED_POINTS) use_degamma_ram = true; @@ -1917,9 +1904,11 @@ static void dcn20_program_pipe( dc->res_pool->hubbub->funcs->force_wm_propagate_to_pipes(dc->res_pool->hubbub); } - if (dc->res_pool->hubbub->funcs->program_det_size && pipe_ctx->update_flags.bits.det_size) - dc->res_pool->hubbub->funcs->program_det_size( - dc->res_pool->hubbub, pipe_ctx->plane_res.hubp->inst, pipe_ctx->det_buffer_size_kb); + if (pipe_ctx->update_flags.bits.det_size) { + if (dc->res_pool->hubbub->funcs->program_det_size) + dc->res_pool->hubbub->funcs->program_det_size( + dc->res_pool->hubbub, pipe_ctx->plane_res.hubp->inst, pipe_ctx->det_buffer_size_kb); + } if (pipe_ctx->update_flags.raw || pipe_ctx->plane_state->update_flags.raw || pipe_ctx->stream->update_flags.raw) dcn20_update_dchubp_dpp(dc, pipe_ctx, context); @@ -2080,9 +2069,11 @@ void dcn20_program_front_end_for_ctx( * turned on (i.e. in an MCLK switch) which can come in too late and cause issues with * DET allocation. */ - if (hubbub->funcs->program_det_size && (context->res_ctx.pipe_ctx[i].update_flags.bits.disable || - (context->res_ctx.pipe_ctx[i].plane_state && dc_state_get_pipe_subvp_type(context, &context->res_ctx.pipe_ctx[i]) == SUBVP_PHANTOM))) - hubbub->funcs->program_det_size(hubbub, dc->current_state->res_ctx.pipe_ctx[i].plane_res.hubp->inst, 0); + if ((context->res_ctx.pipe_ctx[i].update_flags.bits.disable || + (context->res_ctx.pipe_ctx[i].plane_state && dc_state_get_pipe_subvp_type(context, &context->res_ctx.pipe_ctx[i]) == SUBVP_PHANTOM))) { + if (hubbub->funcs->program_det_size) + hubbub->funcs->program_det_size(hubbub, dc->current_state->res_ctx.pipe_ctx[i].plane_res.hubp->inst, 0); + } hws->funcs.plane_atomic_disconnect(dc, dc->current_state, &dc->current_state->res_ctx.pipe_ctx[i]); DC_LOG_DC("Reset mpcc for pipe %d\n", dc->current_state->res_ctx.pipe_ctx[i].pipe_idx); } @@ -2893,11 +2884,6 @@ void dcn20_enable_stream(struct pipe_ctx *pipe_ctx) struct stream_encoder *stream_enc = pipe_ctx->stream_res.stream_enc; if (dc->link_srv->dp_is_128b_132b_signal(pipe_ctx)) { - if (dc->hwseq->funcs.setup_hpo_hw_control) - dc->hwseq->funcs.setup_hpo_hw_control(dc->hwseq, true); - } - - if (dc->link_srv->dp_is_128b_132b_signal(pipe_ctx)) { dto_params.otg_inst = tg->inst; dto_params.pixclk_khz = pipe_ctx->stream->timing.pix_clk_100hz / 10; dto_params.num_odm_segments = get_odm_segment_count(pipe_ctx); diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn20/dcn20_init.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn20/dcn20_init.c index 884e3e323338..ef6488165b8f 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn20/dcn20_init.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn20/dcn20_init.c @@ -67,6 +67,7 @@ static const struct hw_sequencer_funcs dcn20_funcs = { .setup_stereo = dcn10_setup_stereo, .set_avmute = dce110_set_avmute, .log_hw_state = dcn10_log_hw_state, + .log_color_state = dcn20_log_color_state, .get_hw_state = dcn10_get_hw_state, .clear_status_bits = dcn10_clear_status_bits, .wait_for_mpcc_disconnect = dcn10_wait_for_mpcc_disconnect, diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn201/dcn201_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn201/dcn201_hwseq.c index d5769f38874f..6be846635a79 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn201/dcn201_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn201/dcn201_hwseq.c @@ -167,7 +167,7 @@ void dcn201_init_blank( struct tg_color black_color = {0}; struct output_pixel_processor *opp = NULL; uint32_t num_opps, opp_id_src0, opp_id_src1; - uint32_t otg_active_width, otg_active_height; + uint32_t otg_active_width = 0, otg_active_height = 0; /* program opp dpg blank color */ color_space = COLOR_SPACE_SRGB; diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn21/dcn21_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn21/dcn21_hwseq.c index 7252f5f781f0..804be977ea47 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn21/dcn21_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn21/dcn21_hwseq.c @@ -66,7 +66,7 @@ static void mmhub_update_page_table_config(struct dcn_hubbub_phys_addr_config *c int dcn21_init_sys_ctx(struct dce_hwseq *hws, struct dc *dc, struct dc_phy_addr_space_config *pa_config) { - struct dcn_hubbub_phys_addr_config config; + struct dcn_hubbub_phys_addr_config config = {0}; config.system_aperture.fb_top = pa_config->system_aperture.fb_top; config.system_aperture.fb_offset = pa_config->system_aperture.fb_offset; diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_hwseq.c index 8bc3d01537bb..ed9141a67db3 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_hwseq.c @@ -166,7 +166,7 @@ void dcn30_log_color_state(struct dc *dc, "C21 C22 C23 C24 " "C31 C32 C33 C34 \n"); - for (i = 0; i < pool->pipe_count; i++) { + for (i = 0; i < pool->mpcc_count; i++) { struct mpcc_state s = {0}; pool->mpc->funcs->read_mpcc_state(pool->mpc, i, &s); @@ -223,16 +223,14 @@ bool dcn30_set_blend_lut( { struct dpp *dpp_base = pipe_ctx->plane_res.dpp; bool result = true; - struct pwl_params *blend_lut = NULL; - - if (plane_state->blend_tf) { - if (plane_state->blend_tf->type == TF_TYPE_HWPWL) - blend_lut = &plane_state->blend_tf->pwl; - else if (plane_state->blend_tf->type == TF_TYPE_DISTRIBUTED_POINTS) { - cm3_helper_translate_curve_to_hw_format( - plane_state->blend_tf, &dpp_base->regamma_params, false); - blend_lut = &dpp_base->regamma_params; - } + const struct pwl_params *blend_lut = NULL; + + if (plane_state->blend_tf.type == TF_TYPE_HWPWL) + blend_lut = &plane_state->blend_tf.pwl; + else if (plane_state->blend_tf.type == TF_TYPE_DISTRIBUTED_POINTS) { + cm3_helper_translate_curve_to_hw_format( + &plane_state->blend_tf, &dpp_base->regamma_params, false); + blend_lut = &dpp_base->regamma_params; } result = dpp_base->funcs->dpp_program_blnd_lut(dpp_base, blend_lut); @@ -300,27 +298,24 @@ bool dcn30_set_input_transfer_func(struct dc *dc, struct dpp *dpp_base = pipe_ctx->plane_res.dpp; enum dc_transfer_func_predefined tf; bool result = true; - struct pwl_params *params = NULL; + const struct pwl_params *params = NULL; if (dpp_base == NULL || plane_state == NULL) return false; tf = TRANSFER_FUNCTION_UNITY; - if (plane_state->in_transfer_func && - plane_state->in_transfer_func->type == TF_TYPE_PREDEFINED) - tf = plane_state->in_transfer_func->tf; + if (plane_state->in_transfer_func.type == TF_TYPE_PREDEFINED) + tf = plane_state->in_transfer_func.tf; dpp_base->funcs->dpp_set_pre_degam(dpp_base, tf); - if (plane_state->in_transfer_func) { - if (plane_state->in_transfer_func->type == TF_TYPE_HWPWL) - params = &plane_state->in_transfer_func->pwl; - else if (plane_state->in_transfer_func->type == TF_TYPE_DISTRIBUTED_POINTS && - cm3_helper_translate_curve_to_hw_format(plane_state->in_transfer_func, - &dpp_base->degamma_params, false)) - params = &dpp_base->degamma_params; - } + if (plane_state->in_transfer_func.type == TF_TYPE_HWPWL) + params = &plane_state->in_transfer_func.pwl; + else if (plane_state->in_transfer_func.type == TF_TYPE_DISTRIBUTED_POINTS && + cm3_helper_translate_curve_to_hw_format(&plane_state->in_transfer_func, + &dpp_base->degamma_params, false)) + params = &dpp_base->degamma_params; result = dpp_base->funcs->dpp_program_gamcor_lut(dpp_base, params); @@ -378,24 +373,24 @@ bool dcn30_set_output_transfer_func(struct dc *dc, { int mpcc_id = pipe_ctx->plane_res.hubp->inst; struct mpc *mpc = pipe_ctx->stream_res.opp->ctx->dc->res_pool->mpc; - struct pwl_params *params = NULL; + const struct pwl_params *params = NULL; bool ret = false; /* program OGAM or 3DLUT only for the top pipe*/ if (pipe_ctx->top_pipe == NULL) { /*program rmu shaper and 3dlut in MPC*/ ret = dcn30_set_mpc_shaper_3dlut(pipe_ctx, stream); - if (ret == false && mpc->funcs->set_output_gamma && stream->out_transfer_func) { - if (stream->out_transfer_func->type == TF_TYPE_HWPWL) - params = &stream->out_transfer_func->pwl; - else if (pipe_ctx->stream->out_transfer_func->type == + if (ret == false && mpc->funcs->set_output_gamma) { + if (stream->out_transfer_func.type == TF_TYPE_HWPWL) + params = &stream->out_transfer_func.pwl; + else if (pipe_ctx->stream->out_transfer_func.type == TF_TYPE_DISTRIBUTED_POINTS && cm3_helper_translate_curve_to_hw_format( - stream->out_transfer_func, + &stream->out_transfer_func, &mpc->blender_params, false)) params = &mpc->blender_params; /* there are no ROM LUTs in OUTGAM */ - if (stream->out_transfer_func->type == TF_TYPE_PREDEFINED) + if (stream->out_transfer_func.type == TF_TYPE_PREDEFINED) BREAK_TO_DEBUGGER(); } } @@ -804,7 +799,7 @@ void dcn30_init_hw(struct dc *dc) // Get DMCUB capabilities dc_dmub_srv_query_caps_cmd(dc->ctx->dmub_srv); dc->caps.dmub_caps.psr = dc->ctx->dmub_srv->dmub->feature_caps.psr; - dc->caps.dmub_caps.mclk_sw = dc->ctx->dmub_srv->dmub->feature_caps.fw_assisted_mclk_switch; + dc->caps.dmub_caps.mclk_sw = dc->ctx->dmub_srv->dmub->feature_caps.fw_assisted_mclk_switch_ver; } void dcn30_set_avmute(struct pipe_ctx *pipe_ctx, bool enable) @@ -818,7 +813,7 @@ void dcn30_set_avmute(struct pipe_ctx *pipe_ctx, bool enable) enable); /* Wait for two frame to make sure AV mute is sent out */ - if (enable) { + if (enable && pipe_ctx->stream_res.tg->funcs->is_tg_enabled(pipe_ctx->stream_res.tg)) { pipe_ctx->stream_res.tg->funcs->wait_for_state(pipe_ctx->stream_res.tg, CRTC_STATE_VACTIVE); pipe_ctx->stream_res.tg->funcs->wait_for_state(pipe_ctx->stream_res.tg, CRTC_STATE_VBLANK); pipe_ctx->stream_res.tg->funcs->wait_for_state(pipe_ctx->stream_res.tg, CRTC_STATE_VACTIVE); @@ -890,7 +885,7 @@ bool dcn30_apply_idle_power_optimizations(struct dc *dc, bool enable) { union dmub_rb_cmd cmd; uint32_t tmr_delay = 0, tmr_scale = 0; - struct dc_cursor_attributes cursor_attr; + struct dc_cursor_attributes cursor_attr = {0}; bool cursor_cache_enable = false; struct dc_stream_state *stream = NULL; struct dc_plane_state *plane = NULL; @@ -946,7 +941,8 @@ bool dcn30_apply_idle_power_optimizations(struct dc *dc, bool enable) plane->format >= SURFACE_PIXEL_FORMAT_GRPH_ARGB8888 && plane->address.page_table_base.quad_part == 0 && dc->hwss.does_plane_fit_in_mall && - dc->hwss.does_plane_fit_in_mall(dc, plane, + dc->hwss.does_plane_fit_in_mall(dc, plane->plane_size.surface_pitch, + plane->plane_size.surface_size.height, plane->format, cursor_cache_enable ? &cursor_attr : NULL)) { unsigned int v_total = stream->adjust.v_total_max ? stream->adjust.v_total_max : stream->timing.v_total; @@ -1076,11 +1072,15 @@ bool dcn30_apply_idle_power_optimizations(struct dc *dc, bool enable) return true; } -bool dcn30_does_plane_fit_in_mall(struct dc *dc, struct dc_plane_state *plane, struct dc_cursor_attributes *cursor_attr) +bool dcn30_does_plane_fit_in_mall(struct dc *dc, + unsigned int pitch, + unsigned int height, + enum surface_pixel_format format, + struct dc_cursor_attributes *cursor_attr) { // add meta size? - unsigned int surface_size = plane->plane_size.surface_pitch * plane->plane_size.surface_size.height * - (plane->format >= SURFACE_PIXEL_FORMAT_GRPH_ARGB16161616 ? 8 : 4); + unsigned int surface_size = pitch * height * + (format >= SURFACE_PIXEL_FORMAT_GRPH_ARGB16161616 ? 8 : 4); unsigned int mall_size = dc->caps.mall_size_total; unsigned int cursor_size = 0; diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_hwseq.h b/drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_hwseq.h index 638f018a3cb5..76b16839486a 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_hwseq.h +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_hwseq.h @@ -71,7 +71,10 @@ void dcn30_set_avmute(struct pipe_ctx *pipe_ctx, bool enable); void dcn30_update_info_frame(struct pipe_ctx *pipe_ctx); void dcn30_program_dmdata_engine(struct pipe_ctx *pipe_ctx); -bool dcn30_does_plane_fit_in_mall(struct dc *dc, struct dc_plane_state *plane, +bool dcn30_does_plane_fit_in_mall(struct dc *dc, + unsigned int pitch, + unsigned int height, + enum surface_pixel_format format, struct dc_cursor_attributes *cursor_attr); bool dcn30_apply_idle_power_optimizations(struct dc *dc, bool enable); diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn31/dcn31_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn31/dcn31_hwseq.c index a760f0c6fe98..1c8abb417b6e 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn31/dcn31_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn31/dcn31_hwseq.c @@ -273,7 +273,7 @@ void dcn31_init_hw(struct dc *dc) // Get DMCUB capabilities dc_dmub_srv_query_caps_cmd(dc->ctx->dmub_srv); dc->caps.dmub_caps.psr = dc->ctx->dmub_srv->dmub->feature_caps.psr; - dc->caps.dmub_caps.mclk_sw = dc->ctx->dmub_srv->dmub->feature_caps.fw_assisted_mclk_switch; + dc->caps.dmub_caps.mclk_sw = dc->ctx->dmub_srv->dmub->feature_caps.fw_assisted_mclk_switch_ver; } void dcn31_dsc_pg_control( @@ -479,7 +479,7 @@ void dcn31_hubp_pg_control(struct dce_hwseq *hws, unsigned int hubp_inst, bool p int dcn31_init_sys_ctx(struct dce_hwseq *hws, struct dc *dc, struct dc_phy_addr_space_config *pa_config) { - struct dcn_hubbub_phys_addr_config config; + struct dcn_hubbub_phys_addr_config config = {0}; config.system_aperture.fb_top = pa_config->system_aperture.fb_top; config.system_aperture.fb_offset = pa_config->system_aperture.fb_offset; diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn314/dcn314_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn314/dcn314_hwseq.c index 3a9cc8ac0c07..0d8a05cf8b1a 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn314/dcn314_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn314/dcn314_hwseq.c @@ -69,29 +69,6 @@ #define FN(reg_name, field_name) \ hws->shifts->field_name, hws->masks->field_name -static int calc_mpc_flow_ctrl_cnt(const struct dc_stream_state *stream, - int opp_cnt) -{ - bool hblank_halved = optc2_is_two_pixels_per_containter(&stream->timing); - int flow_ctrl_cnt; - - if (opp_cnt >= 2) - hblank_halved = true; - - flow_ctrl_cnt = stream->timing.h_total - stream->timing.h_addressable - - stream->timing.h_border_left - - stream->timing.h_border_right; - - if (hblank_halved) - flow_ctrl_cnt /= 2; - - /* ODM combine 4:1 case */ - if (opp_cnt == 4) - flow_ctrl_cnt /= 2; - - return flow_ctrl_cnt; -} - static void update_dsc_on_stream(struct pipe_ctx *pipe_ctx, bool enable) { struct display_stream_compressor *dsc = pipe_ctx->stream_res.dsc; @@ -105,7 +82,7 @@ static void update_dsc_on_stream(struct pipe_ctx *pipe_ctx, bool enable) if (enable) { struct dsc_config dsc_cfg; - struct dsc_optc_config dsc_optc_cfg; + struct dsc_optc_config dsc_optc_cfg = {0}; enum optc_dsc_mode optc_dsc_mode; /* Enable DSC hw block */ @@ -183,10 +160,6 @@ void dcn314_update_odm(struct dc *dc, struct dc_state *context, struct pipe_ctx struct pipe_ctx *odm_pipe; int opp_cnt = 0; int opp_inst[MAX_PIPES] = {0}; - bool rate_control_2x_pclk = (pipe_ctx->stream->timing.flags.INTERLACE || optc2_is_two_pixels_per_containter(&pipe_ctx->stream->timing)); - struct mpc_dwb_flow_control flow_control; - struct mpc *mpc = dc->res_pool->mpc; - int i; opp_cnt = get_odm_config(pipe_ctx, opp_inst); @@ -199,20 +172,6 @@ void dcn314_update_odm(struct dc *dc, struct dc_state *context, struct pipe_ctx pipe_ctx->stream_res.tg->funcs->set_odm_bypass( pipe_ctx->stream_res.tg, &pipe_ctx->stream->timing); - rate_control_2x_pclk = rate_control_2x_pclk || opp_cnt > 1; - flow_control.flow_ctrl_mode = 0; - flow_control.flow_ctrl_cnt0 = 0x80; - flow_control.flow_ctrl_cnt1 = calc_mpc_flow_ctrl_cnt(pipe_ctx->stream, opp_cnt); - if (mpc->funcs->set_out_rate_control) { - for (i = 0; i < opp_cnt; ++i) { - mpc->funcs->set_out_rate_control( - mpc, opp_inst[i], - true, - rate_control_2x_pclk, - &flow_control); - } - } - for (odm_pipe = pipe_ctx->next_odm_pipe; odm_pipe; odm_pipe = odm_pipe->next_odm_pipe) { odm_pipe->stream_res.opp->funcs->opp_pipe_clock_control( odm_pipe->stream_res.opp, diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_hwseq.c index c0b526cf1786..b8e884368dc6 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_hwseq.c @@ -239,8 +239,10 @@ static uint32_t dcn32_calculate_cab_allocation(struct dc *dc, struct dc_state *c // Convert number of cache lines required to number of ways if (dc->debug.force_mall_ss_num_ways > 0) { num_ways = dc->debug.force_mall_ss_num_ways; + } else if (dc->res_pool->funcs->calculate_mall_ways_from_bytes) { + num_ways = dc->res_pool->funcs->calculate_mall_ways_from_bytes(dc, mall_ss_size_bytes); } else { - num_ways = dcn32_helper_mall_bytes_to_ways(dc, mall_ss_size_bytes); + num_ways = 0; } return num_ways; @@ -261,7 +263,9 @@ bool dcn32_apply_idle_power_optimizations(struct dc *dc, bool enable) for (i = 0; i < dc->current_state->stream_count; i++) { /* MALL SS messaging is not supported with PSR at this time */ if (dc->current_state->streams[i] != NULL && - dc->current_state->streams[i]->link->psr_settings.psr_version != DC_PSR_VERSION_UNSUPPORTED) + dc->current_state->streams[i]->link->psr_settings.psr_version != DC_PSR_VERSION_UNSUPPORTED && + (dc->current_state->stream_count > 1 || (!dc->current_state->streams[i]->dpms_off && + dc->current_state->stream_status[i].plane_count > 0))) return false; } @@ -475,39 +479,35 @@ bool dcn32_set_mcm_luts( int mpcc_id = pipe_ctx->plane_res.hubp->inst; struct mpc *mpc = pipe_ctx->stream_res.opp->ctx->dc->res_pool->mpc; bool result = true; - struct pwl_params *lut_params = NULL; + const struct pwl_params *lut_params = NULL; // 1D LUT - if (plane_state->blend_tf) { - if (plane_state->blend_tf->type == TF_TYPE_HWPWL) - lut_params = &plane_state->blend_tf->pwl; - else if (plane_state->blend_tf->type == TF_TYPE_DISTRIBUTED_POINTS) { - cm3_helper_translate_curve_to_hw_format(plane_state->blend_tf, - &dpp_base->regamma_params, false); - lut_params = &dpp_base->regamma_params; - } + if (plane_state->blend_tf.type == TF_TYPE_HWPWL) + lut_params = &plane_state->blend_tf.pwl; + else if (plane_state->blend_tf.type == TF_TYPE_DISTRIBUTED_POINTS) { + cm3_helper_translate_curve_to_hw_format(&plane_state->blend_tf, + &dpp_base->regamma_params, false); + lut_params = &dpp_base->regamma_params; } result = mpc->funcs->program_1dlut(mpc, lut_params, mpcc_id); lut_params = NULL; // Shaper - if (plane_state->in_shaper_func) { - if (plane_state->in_shaper_func->type == TF_TYPE_HWPWL) - lut_params = &plane_state->in_shaper_func->pwl; - else if (plane_state->in_shaper_func->type == TF_TYPE_DISTRIBUTED_POINTS) { - // TODO: dpp_base replace - ASSERT(false); - cm3_helper_translate_curve_to_hw_format(plane_state->in_shaper_func, - &dpp_base->shaper_params, true); - lut_params = &dpp_base->shaper_params; - } + if (plane_state->in_shaper_func.type == TF_TYPE_HWPWL) + lut_params = &plane_state->in_shaper_func.pwl; + else if (plane_state->in_shaper_func.type == TF_TYPE_DISTRIBUTED_POINTS) { + // TODO: dpp_base replace + ASSERT(false); + cm3_helper_translate_curve_to_hw_format(&plane_state->in_shaper_func, + &dpp_base->shaper_params, true); + lut_params = &dpp_base->shaper_params; } result = mpc->funcs->program_shaper(mpc, lut_params, mpcc_id); // 3D - if (plane_state->lut3d_func && plane_state->lut3d_func->state.bits.initialized == 1) - result = mpc->funcs->program_3dlut(mpc, &plane_state->lut3d_func->lut_3d, mpcc_id); + if (plane_state->lut3d_func.state.bits.initialized == 1) + result = mpc->funcs->program_3dlut(mpc, &plane_state->lut3d_func.lut_3d, mpcc_id); else result = mpc->funcs->program_3dlut(mpc, NULL, mpcc_id); @@ -524,27 +524,24 @@ bool dcn32_set_input_transfer_func(struct dc *dc, enum dc_transfer_func_predefined tf; bool result = true; - struct pwl_params *params = NULL; + const struct pwl_params *params = NULL; if (mpc == NULL || plane_state == NULL) return false; tf = TRANSFER_FUNCTION_UNITY; - if (plane_state->in_transfer_func && - plane_state->in_transfer_func->type == TF_TYPE_PREDEFINED) - tf = plane_state->in_transfer_func->tf; + if (plane_state->in_transfer_func.type == TF_TYPE_PREDEFINED) + tf = plane_state->in_transfer_func.tf; dpp_base->funcs->dpp_set_pre_degam(dpp_base, tf); - if (plane_state->in_transfer_func) { - if (plane_state->in_transfer_func->type == TF_TYPE_HWPWL) - params = &plane_state->in_transfer_func->pwl; - else if (plane_state->in_transfer_func->type == TF_TYPE_DISTRIBUTED_POINTS && - cm3_helper_translate_curve_to_hw_format(plane_state->in_transfer_func, - &dpp_base->degamma_params, false)) - params = &dpp_base->degamma_params; - } + if (plane_state->in_transfer_func.type == TF_TYPE_HWPWL) + params = &plane_state->in_transfer_func.pwl; + else if (plane_state->in_transfer_func.type == TF_TYPE_DISTRIBUTED_POINTS && + cm3_helper_translate_curve_to_hw_format(&plane_state->in_transfer_func, + &dpp_base->degamma_params, false)) + params = &dpp_base->degamma_params; dpp_base->funcs->dpp_program_gamcor_lut(dpp_base, params); @@ -562,24 +559,24 @@ bool dcn32_set_output_transfer_func(struct dc *dc, { int mpcc_id = pipe_ctx->plane_res.hubp->inst; struct mpc *mpc = pipe_ctx->stream_res.opp->ctx->dc->res_pool->mpc; - struct pwl_params *params = NULL; + const struct pwl_params *params = NULL; bool ret = false; /* program OGAM or 3DLUT only for the top pipe*/ if (resource_is_pipe_type(pipe_ctx, OPP_HEAD)) { /*program shaper and 3dlut in MPC*/ ret = dcn32_set_mpc_shaper_3dlut(pipe_ctx, stream); - if (ret == false && mpc->funcs->set_output_gamma && stream->out_transfer_func) { - if (stream->out_transfer_func->type == TF_TYPE_HWPWL) - params = &stream->out_transfer_func->pwl; - else if (pipe_ctx->stream->out_transfer_func->type == + if (ret == false && mpc->funcs->set_output_gamma) { + if (stream->out_transfer_func.type == TF_TYPE_HWPWL) + params = &stream->out_transfer_func.pwl; + else if (pipe_ctx->stream->out_transfer_func.type == TF_TYPE_DISTRIBUTED_POINTS && cm3_helper_translate_curve_to_hw_format( - stream->out_transfer_func, + &stream->out_transfer_func, &mpc->blender_params, false)) params = &mpc->blender_params; /* there are no ROM LUTs in OUTGAM */ - if (stream->out_transfer_func->type == TF_TYPE_PREDEFINED) + if (stream->out_transfer_func.type == TF_TYPE_PREDEFINED) BREAK_TO_DEBUGGER(); } } @@ -956,39 +953,16 @@ void dcn32_init_hw(struct dc *dc) dc->caps.dmub_caps.psr = dc->ctx->dmub_srv->dmub->feature_caps.psr; dc->caps.dmub_caps.subvp_psr = dc->ctx->dmub_srv->dmub->feature_caps.subvp_psr_support; dc->caps.dmub_caps.gecc_enable = dc->ctx->dmub_srv->dmub->feature_caps.gecc_enable; - dc->caps.dmub_caps.mclk_sw = dc->ctx->dmub_srv->dmub->feature_caps.fw_assisted_mclk_switch; + dc->caps.dmub_caps.mclk_sw = dc->ctx->dmub_srv->dmub->feature_caps.fw_assisted_mclk_switch_ver; if (dc->ctx->dmub_srv->dmub->fw_version < - DMUB_FW_VERSION(7, 0, 35)) { + DMUB_FW_VERSION(7, 0, 35)) { dc->debug.force_disable_subvp = true; dc->debug.disable_fpo_optimizations = true; } } } -static int calc_mpc_flow_ctrl_cnt(const struct dc_stream_state *stream, - int opp_cnt) -{ - bool hblank_halved = optc2_is_two_pixels_per_containter(&stream->timing); - int flow_ctrl_cnt; - - if (opp_cnt >= 2) - hblank_halved = true; - - flow_ctrl_cnt = stream->timing.h_total - stream->timing.h_addressable - - stream->timing.h_border_left - - stream->timing.h_border_right; - - if (hblank_halved) - flow_ctrl_cnt /= 2; - - /* ODM combine 4:1 case */ - if (opp_cnt == 4) - flow_ctrl_cnt /= 2; - - return flow_ctrl_cnt; -} - static void update_dsc_on_stream(struct pipe_ctx *pipe_ctx, bool enable) { struct display_stream_compressor *dsc = pipe_ctx->stream_res.dsc; @@ -1015,7 +989,7 @@ static void update_dsc_on_stream(struct pipe_ctx *pipe_ctx, bool enable) if (enable) { struct dsc_config dsc_cfg; - struct dsc_optc_config dsc_optc_cfg; + struct dsc_optc_config dsc_optc_cfg = {0}; enum optc_dsc_mode optc_dsc_mode; /* Enable DSC hw block */ @@ -1103,10 +1077,6 @@ void dcn32_update_odm(struct dc *dc, struct dc_state *context, struct pipe_ctx * struct pipe_ctx *odm_pipe; int opp_cnt = 0; int opp_inst[MAX_PIPES] = {0}; - bool rate_control_2x_pclk = (pipe_ctx->stream->timing.flags.INTERLACE || optc2_is_two_pixels_per_containter(&pipe_ctx->stream->timing)); - struct mpc_dwb_flow_control flow_control; - struct mpc *mpc = dc->res_pool->mpc; - int i; opp_cnt = get_odm_config(pipe_ctx, opp_inst); @@ -1119,20 +1089,6 @@ void dcn32_update_odm(struct dc *dc, struct dc_state *context, struct pipe_ctx * pipe_ctx->stream_res.tg->funcs->set_odm_bypass( pipe_ctx->stream_res.tg, &pipe_ctx->stream->timing); - rate_control_2x_pclk = rate_control_2x_pclk || opp_cnt > 1; - flow_control.flow_ctrl_mode = 0; - flow_control.flow_ctrl_cnt0 = 0x80; - flow_control.flow_ctrl_cnt1 = calc_mpc_flow_ctrl_cnt(pipe_ctx->stream, opp_cnt); - if (mpc->funcs->set_out_rate_control) { - for (i = 0; i < opp_cnt; ++i) { - mpc->funcs->set_out_rate_control( - mpc, opp_inst[i], - true, - rate_control_2x_pclk, - &flow_control); - } - } - for (odm_pipe = pipe_ctx->next_odm_pipe; odm_pipe; odm_pipe = odm_pipe->next_odm_pipe) { odm_pipe->stream_res.opp->funcs->opp_pipe_clock_control( odm_pipe->stream_res.opp, @@ -1586,7 +1542,7 @@ void dcn32_init_blank( struct output_pixel_processor *opp = NULL; struct output_pixel_processor *bottom_opp = NULL; uint32_t num_opps, opp_id_src0, opp_id_src1; - uint32_t otg_active_width, otg_active_height; + uint32_t otg_active_width = 0, otg_active_height = 0; uint32_t i; /* program opp dpg blank color */ diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.c index 4b92df23ff0d..5295f52e4fc8 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.c @@ -349,7 +349,7 @@ void dcn35_init_hw(struct dc *dc) if (dc->ctx->dmub_srv) { dc_dmub_srv_query_caps_cmd(dc->ctx->dmub_srv); dc->caps.dmub_caps.psr = dc->ctx->dmub_srv->dmub->feature_caps.psr; - dc->caps.dmub_caps.mclk_sw = dc->ctx->dmub_srv->dmub->feature_caps.fw_assisted_mclk_switch; + dc->caps.dmub_caps.mclk_sw = dc->ctx->dmub_srv->dmub->feature_caps.fw_assisted_mclk_switch_ver; } if (dc->res_pool->pg_cntl) { @@ -358,29 +358,6 @@ void dcn35_init_hw(struct dc *dc) } } -static int calc_mpc_flow_ctrl_cnt(const struct dc_stream_state *stream, - int opp_cnt) -{ - bool hblank_halved = optc2_is_two_pixels_per_containter(&stream->timing); - int flow_ctrl_cnt; - - if (opp_cnt >= 2) - hblank_halved = true; - - flow_ctrl_cnt = stream->timing.h_total - stream->timing.h_addressable - - stream->timing.h_border_left - - stream->timing.h_border_right; - - if (hblank_halved) - flow_ctrl_cnt /= 2; - - /* ODM combine 4:1 case */ - if (opp_cnt == 4) - flow_ctrl_cnt /= 2; - - return flow_ctrl_cnt; -} - static void update_dsc_on_stream(struct pipe_ctx *pipe_ctx, bool enable) { struct display_stream_compressor *dsc = pipe_ctx->stream_res.dsc; @@ -396,7 +373,7 @@ static void update_dsc_on_stream(struct pipe_ctx *pipe_ctx, bool enable) if (enable) { struct dsc_config dsc_cfg; - struct dsc_optc_config dsc_optc_cfg; + struct dsc_optc_config dsc_optc_cfg = {0}; enum optc_dsc_mode optc_dsc_mode; /* Enable DSC hw block */ @@ -474,10 +451,6 @@ void dcn35_update_odm(struct dc *dc, struct dc_state *context, struct pipe_ctx * struct pipe_ctx *odm_pipe; int opp_cnt = 0; int opp_inst[MAX_PIPES] = {0}; - bool rate_control_2x_pclk = (pipe_ctx->stream->timing.flags.INTERLACE || optc2_is_two_pixels_per_containter(&pipe_ctx->stream->timing)); - struct mpc_dwb_flow_control flow_control; - struct mpc *mpc = dc->res_pool->mpc; - int i; opp_cnt = get_odm_config(pipe_ctx, opp_inst); @@ -490,20 +463,6 @@ void dcn35_update_odm(struct dc *dc, struct dc_state *context, struct pipe_ctx * pipe_ctx->stream_res.tg->funcs->set_odm_bypass( pipe_ctx->stream_res.tg, &pipe_ctx->stream->timing); - rate_control_2x_pclk = rate_control_2x_pclk || opp_cnt > 1; - flow_control.flow_ctrl_mode = 0; - flow_control.flow_ctrl_cnt0 = 0x80; - flow_control.flow_ctrl_cnt1 = calc_mpc_flow_ctrl_cnt(pipe_ctx->stream, opp_cnt); - if (mpc->funcs->set_out_rate_control) { - for (i = 0; i < opp_cnt; ++i) { - mpc->funcs->set_out_rate_control( - mpc, opp_inst[i], - true, - rate_control_2x_pclk, - &flow_control); - } - } - for (odm_pipe = pipe_ctx->next_odm_pipe; odm_pipe; odm_pipe = odm_pipe->next_odm_pipe) { odm_pipe->stream_res.opp->funcs->opp_pipe_clock_control( odm_pipe->stream_res.opp, @@ -536,6 +495,17 @@ void dcn35_dpp_root_clock_control(struct dce_hwseq *hws, unsigned int dpp_inst, } } +void dcn35_dpstream_root_clock_control(struct dce_hwseq *hws, unsigned int dp_hpo_inst, bool clock_on) +{ + if (!hws->ctx->dc->debug.root_clock_optimization.bits.dpstream) + return; + + if (hws->ctx->dc->res_pool->dccg->funcs->set_dpstreamclk_root_clock_gating) { + hws->ctx->dc->res_pool->dccg->funcs->set_dpstreamclk_root_clock_gating( + hws->ctx->dc->res_pool->dccg, dp_hpo_inst, clock_on); + } +} + void dcn35_dsc_pg_control( struct dce_hwseq *hws, unsigned int dsc_inst, @@ -679,22 +649,43 @@ void dcn35_power_down_on_boot(struct dc *dc) bool dcn35_apply_idle_power_optimizations(struct dc *dc, bool enable) { - struct dc_link *edp_links[MAX_NUM_EDP]; - int i, edp_num; if (dc->debug.dmcub_emulation) return true; if (enable) { - dc_get_edp_links(dc, edp_links, &edp_num); - if (edp_num == 0 || edp_num > 1) - return false; + uint32_t num_active_edp = 0; + int i; for (i = 0; i < dc->current_state->stream_count; ++i) { struct dc_stream_state *stream = dc->current_state->streams[i]; + struct dc_link *link = stream->link; + bool is_psr = link && !link->panel_config.psr.disable_psr && + (link->psr_settings.psr_version == DC_PSR_VERSION_1 || + link->psr_settings.psr_version == DC_PSR_VERSION_SU_1); + bool is_replay = link && link->replay_settings.replay_feature_enabled; + + /* Ignore streams that disabled. */ + if (stream->dpms_off) + continue; + + /* Active external displays block idle optimizations. */ + if (!dc_is_embedded_signal(stream->signal)) + return false; + + /* If not PWRSEQ0 can't enter idle optimizations */ + if (link && link->link_index != 0) + return false; - if (!stream->dpms_off && !dc_is_embedded_signal(stream->signal)) + /* Check for panel power features required for idle optimizations. */ + if (!is_psr && !is_replay) return false; + + num_active_edp += 1; } + + /* If more than one active eDP then disallow. */ + if (num_active_edp > 1) + return false; } // TODO: review other cases when idle optimization is allowed @@ -720,6 +711,7 @@ void dcn35_init_pipes(struct dc *dc, struct dc_state *context) struct hubbub *hubbub = dc->res_pool->hubbub; struct pg_cntl *pg_cntl = dc->res_pool->pg_cntl; bool can_apply_seamless_boot = false; + bool tg_enabled[MAX_PIPES] = {false}; for (i = 0; i < context->stream_count; i++) { if (context->streams[i]->apply_seamless_boot_optimization) { @@ -801,6 +793,7 @@ void dcn35_init_pipes(struct dc *dc, struct dc_state *context) // requesting data while in PSR. tg->funcs->tg_init(tg); hubp->power_gated = true; + tg_enabled[i] = true; continue; } @@ -842,6 +835,20 @@ void dcn35_init_pipes(struct dc *dc, struct dc_state *context) tg->funcs->tg_init(tg); } + /* Clean up MPC tree */ + for (i = 0; i < dc->res_pool->pipe_count; i++) { + if (tg_enabled[i]) { + if (dc->res_pool->opps[i]->mpc_tree_params.opp_list) { + if (dc->res_pool->opps[i]->mpc_tree_params.opp_list->mpcc_bot) { + int bot_id = dc->res_pool->opps[i]->mpc_tree_params.opp_list->mpcc_bot->mpcc_id; + + if ((bot_id < MAX_MPCC) && (bot_id < MAX_PIPES) && (!tg_enabled[bot_id])) + dc->res_pool->opps[i]->mpc_tree_params.opp_list = NULL; + } + } + } + } + if (pg_cntl != NULL) { if (pg_cntl->funcs->dsc_pg_control != NULL) { uint32_t num_opps = 0; @@ -1002,6 +1009,9 @@ void dcn35_calc_blocks_to_gate(struct dc *dc, struct dc_state *context, if (!hpo_frl_stream_enc_acquired && !hpo_dp_stream_enc_acquired) update_state->pg_res_update[PG_HPO] = true; + if (hpo_frl_stream_enc_acquired) + update_state->pg_pipe_res_update[PG_HDMISTREAM][0] = true; + update_state->pg_res_update[PG_DWB] = true; for (i = 0; i < dc->res_pool->pipe_count; i++) { @@ -1019,8 +1029,7 @@ void dcn35_calc_blocks_to_gate(struct dc *dc, struct dc_state *context, if (pipe_ctx->plane_res.dpp) update_state->pg_pipe_res_update[PG_DPP][pipe_ctx->plane_res.hubp->inst] = false; - if ((pipe_ctx->plane_res.dpp || pipe_ctx->stream_res.opp) && - pipe_ctx->plane_res.mpcc_inst >= 0) + if (pipe_ctx->plane_res.dpp || pipe_ctx->stream_res.opp) update_state->pg_pipe_res_update[PG_MPCC][pipe_ctx->plane_res.mpcc_inst] = false; if (pipe_ctx->stream_res.dsc) @@ -1028,6 +1037,9 @@ void dcn35_calc_blocks_to_gate(struct dc *dc, struct dc_state *context, if (pipe_ctx->stream_res.opp) update_state->pg_pipe_res_update[PG_OPP][pipe_ctx->stream_res.opp->inst] = false; + + if (pipe_ctx->stream_res.hpo_dp_stream_enc) + update_state->pg_pipe_res_update[PG_DPSTREAM][pipe_ctx->stream_res.hpo_dp_stream_enc->inst] = false; } /*domain24 controls all the otg, mpc, opp, as long as one otg is still up, avoid enabling OTG PG*/ for (i = 0; i < dc->res_pool->timing_generator_count; i++) { @@ -1085,6 +1097,9 @@ void dcn35_calc_blocks_to_ungate(struct dc *dc, struct dc_state *context, if (j == PG_OPTC && new_pipe->stream_res.tg) update_state->pg_pipe_res_update[j][new_pipe->stream_res.tg->inst] = true; + + if (j == PG_DPSTREAM && new_pipe->stream_res.hpo_dp_stream_enc) + update_state->pg_pipe_res_update[j][new_pipe->stream_res.hpo_dp_stream_enc->inst] = true; } } else if (cur_pipe->plane_state == new_pipe->plane_state || cur_pipe == new_pipe) { @@ -1114,6 +1129,11 @@ void dcn35_calc_blocks_to_ungate(struct dc *dc, struct dc_state *context, cur_pipe->stream_res.tg != new_pipe->stream_res.tg && new_pipe->stream_res.tg) update_state->pg_pipe_res_update[j][new_pipe->stream_res.tg->inst] = true; + + if (j == PG_DPSTREAM && + cur_pipe->stream_res.hpo_dp_stream_enc != new_pipe->stream_res.hpo_dp_stream_enc && + new_pipe->stream_res.hpo_dp_stream_enc) + update_state->pg_pipe_res_update[j][new_pipe->stream_res.hpo_dp_stream_enc->inst] = true; } } } @@ -1129,6 +1149,9 @@ void dcn35_calc_blocks_to_ungate(struct dc *dc, struct dc_state *context, if (hpo_frl_stream_enc_acquired || hpo_dp_stream_enc_acquired) update_state->pg_res_update[PG_HPO] = true; + if (hpo_frl_stream_enc_acquired) + update_state->pg_pipe_res_update[PG_HDMISTREAM][0] = true; + } /** @@ -1253,14 +1276,19 @@ void dcn35_root_clock_control(struct dc *dc, if (!pg_cntl) return; /*enable root clock first when power up*/ - if (power_on) + if (power_on) { for (i = 0; i < dc->res_pool->pipe_count; i++) { if (update_state->pg_pipe_res_update[PG_HUBP][i] && update_state->pg_pipe_res_update[PG_DPP][i]) { if (dc->hwseq->funcs.dpp_root_clock_control) dc->hwseq->funcs.dpp_root_clock_control(dc->hwseq, i, power_on); } + if (update_state->pg_pipe_res_update[PG_DPSTREAM][i]) + if (dc->hwseq->funcs.dpstream_root_clock_control) + dc->hwseq->funcs.dpstream_root_clock_control(dc->hwseq, i, power_on); } + + } for (i = 0; i < dc->res_pool->res_cap->num_dsc; i++) { if (update_state->pg_pipe_res_update[PG_DSC][i]) { if (power_on) { @@ -1273,14 +1301,19 @@ void dcn35_root_clock_control(struct dc *dc, } } /*disable root clock first when power down*/ - if (!power_on) + if (!power_on) { for (i = 0; i < dc->res_pool->pipe_count; i++) { if (update_state->pg_pipe_res_update[PG_HUBP][i] && update_state->pg_pipe_res_update[PG_DPP][i]) { if (dc->hwseq->funcs.dpp_root_clock_control) dc->hwseq->funcs.dpp_root_clock_control(dc->hwseq, i, power_on); } + if (update_state->pg_pipe_res_update[PG_DPSTREAM][i]) + if (dc->hwseq->funcs.dpstream_root_clock_control) + dc->hwseq->funcs.dpstream_root_clock_control(dc->hwseq, i, power_on); } + + } } void dcn35_prepare_bandwidth( @@ -1321,22 +1354,6 @@ void dcn35_optimize_bandwidth( } } -void dcn35_set_idle_state(const struct dc *dc, bool allow_idle) -{ - // TODO: Find a more suitable communcation - if (dc->clk_mgr->funcs->set_idle_state) - dc->clk_mgr->funcs->set_idle_state(dc->clk_mgr, allow_idle); -} - -uint32_t dcn35_get_idle_state(const struct dc *dc) -{ - // TODO: Find a more suitable communcation - if (dc->clk_mgr->funcs->get_idle_state) - return dc->clk_mgr->funcs->get_idle_state(dc->clk_mgr); - - return 0; -} - void dcn35_set_drr(struct pipe_ctx **pipe_ctx, int num_pipes, struct dc_crtc_timing_adjust adjust) { @@ -1394,3 +1411,31 @@ void dcn35_set_static_screen_control(struct pipe_ctx **pipe_ctx, set_static_screen_control(pipe_ctx[i]->stream_res.tg, triggers, params->num_frames); } + +void dcn35_set_long_vblank(struct pipe_ctx **pipe_ctx, + int num_pipes, uint32_t v_total_min, uint32_t v_total_max) +{ + int i = 0; + struct long_vtotal_params params = {0}; + + params.vertical_total_max = v_total_max; + params.vertical_total_min = v_total_min; + + for (i = 0; i < num_pipes; i++) { + if (!pipe_ctx[i]) + continue; + + if (pipe_ctx[i]->stream) { + struct dc_crtc_timing *timing = &pipe_ctx[i]->stream->timing; + + if (timing) + params.vertical_blank_start = timing->v_total - timing->v_front_porch; + else + params.vertical_blank_start = 0; + + if ((pipe_ctx[i]->stream_res.tg != NULL) && pipe_ctx[i]->stream_res.tg->funcs && + pipe_ctx[i]->stream_res.tg->funcs->set_long_vtotal) + pipe_ctx[i]->stream_res.tg->funcs->set_long_vtotal(pipe_ctx[i]->stream_res.tg, ¶ms); + } + } +} diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.h b/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.h index c354efa6c1b2..a731c8880d60 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.h +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.h @@ -37,6 +37,8 @@ void dcn35_dsc_pg_control(struct dce_hwseq *hws, unsigned int dsc_inst, bool pow void dcn35_dpp_root_clock_control(struct dce_hwseq *hws, unsigned int dpp_inst, bool clock_on); +void dcn35_dpstream_root_clock_control(struct dce_hwseq *hws, unsigned int dp_hpo_inst, bool clock_on); + void dcn35_enable_power_gating_plane(struct dce_hwseq *hws, bool enable); void dcn35_set_dmu_fgcg(struct dce_hwseq *hws, bool enable); @@ -84,13 +86,13 @@ void dcn35_dsc_pg_control( unsigned int dsc_inst, bool power_on); -void dcn35_set_idle_state(const struct dc *dc, bool allow_idle); -uint32_t dcn35_get_idle_state(const struct dc *dc); - void dcn35_set_drr(struct pipe_ctx **pipe_ctx, int num_pipes, struct dc_crtc_timing_adjust adjust); void dcn35_set_static_screen_control(struct pipe_ctx **pipe_ctx, int num_pipes, const struct dc_static_screen_params *params); +void dcn35_set_long_vblank(struct pipe_ctx **pipe_ctx, + int num_pipes, uint32_t v_total_min, uint32_t v_total_max); + #endif /* __DC_HWSS_DCN35_H__ */ diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_init.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_init.c index a93073055e7b..df3bf77f3fb4 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_init.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_init.c @@ -121,8 +121,7 @@ static const struct hw_sequencer_funcs dcn35_funcs = { .hw_block_power_up = dcn35_hw_block_power_up, .hw_block_power_down = dcn35_hw_block_power_down, .root_clock_control = dcn35_root_clock_control, - .set_idle_state = dcn35_set_idle_state, - .get_idle_state = dcn35_get_idle_state + .set_long_vtotal = dcn35_set_long_vblank, }; static const struct hwseq_private_funcs dcn35_private_funcs = { @@ -148,6 +147,7 @@ static const struct hwseq_private_funcs dcn35_private_funcs = { //.hubp_pg_control = dcn35_hubp_pg_control, .enable_power_gating_plane = dcn35_enable_power_gating_plane, .dpp_root_clock_control = dcn35_dpp_root_clock_control, + .dpstream_root_clock_control = dcn35_dpstream_root_clock_control, .program_all_writeback_pipes_in_tree = dcn30_program_all_writeback_pipes_in_tree, .update_odm = dcn35_update_odm, .set_hdr_multiplier = dcn10_set_hdr_multiplier, diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn351/Makefile b/drivers/gpu/drm/amd/display/dc/hwss/dcn351/Makefile index b24ad27fe6ef..a4b3c1e99ec6 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn351/Makefile +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn351/Makefile @@ -1,16 +1,27 @@ # -# (c) Copyright 2022 Advanced Micro Devices, Inc. All the rights reserved +# Copyright (c) 2022-2024 Advanced Micro Devices, Inc. # -# All rights reserved. This notice is intended as a precaution against -# inadvertent publication and does not imply publication or any waiver -# of confidentiality. The year included in the foregoing notice is the -# year of creation of the work. +# 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: # -# Authors: AMD +# 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. # # Makefile for DCN351. -DCN351 = dcn351_init.o +DCN351 = dcn351_hwseq.o dcn351_init.o AMD_DAL_DCN351 = $(addprefix $(AMDDALPATH)/dc/dcn351/,$(DCN351)) diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn351/dcn351_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn351/dcn351_hwseq.c new file mode 100644 index 000000000000..93fe5b262a3d --- /dev/null +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn351/dcn351_hwseq.c @@ -0,0 +1,182 @@ +/* SPDX-License-Identifier: MIT */ +/* + * Copyright 2024 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 "core_types.h" +#include "resource.h" +#include "dcn351_hwseq.h" +#include "dcn35/dcn35_hwseq.h" + +#define DC_LOGGER_INIT(logger) \ + struct dal_logger *dc_logger = logger + +#define DC_LOGGER \ + dc_logger + +void dcn351_calc_blocks_to_gate(struct dc *dc, struct dc_state *context, + struct pg_block_update *update_state) +{ + int i, j; + + dcn35_calc_blocks_to_gate(dc, context, update_state); + + for (i = dc->res_pool->pipe_count - 1; i >= 0; i--) { + if (!update_state->pg_pipe_res_update[PG_HUBP][i] && + !update_state->pg_pipe_res_update[PG_DPP][i]) { + for (j = i - 1; j >= 0; j--) { + update_state->pg_pipe_res_update[PG_HUBP][j] = false; + update_state->pg_pipe_res_update[PG_DPP][j] = false; + } + + break; + } + } +} + +void dcn351_calc_blocks_to_ungate(struct dc *dc, struct dc_state *context, + struct pg_block_update *update_state) +{ + int i, j; + + dcn35_calc_blocks_to_ungate(dc, context, update_state); + + for (i = dc->res_pool->pipe_count - 1; i >= 0; i--) { + if (update_state->pg_pipe_res_update[PG_HUBP][i] && + update_state->pg_pipe_res_update[PG_DPP][i]) { + for (j = i - 1; j >= 0; j--) { + update_state->pg_pipe_res_update[PG_HUBP][j] = true; + update_state->pg_pipe_res_update[PG_DPP][j] = true; + } + + break; + } + } +} + +/** + * dcn351_hw_block_power_down() - power down sequence + * + * The following sequence describes the ON-OFF (ONO) for power down: + * + * ONO Region 11, DCPG 19: dsc3 + * ONO Region 10, DCPG 3: dchubp3, dpp3 + * ONO Region 9, DCPG 18: dsc2 + * ONO Region 8, DCPG 2: dchubp2, dpp2 + * ONO Region 7, DCPG 17: dsc1 + * ONO Region 6, DCPG 1: dchubp1, dpp1 + * ONO Region 5, DCPG 16: dsc0 + * ONO Region 4, DCPG 0: dchubp0, dpp0 + * ONO Region 3, DCPG 25: hpo - SKIPPED. Should be kept on + * ONO Region 2, DCPG 24: mpc opp optc dwb + * ONO Region 1, DCPG 23: dchubbub dchvm dchubbubmem - SKIPPED. PMFW will pwr dwn at IPS2 entry + * ONO Region 0, DCPG 22: dccg dio dcio - SKIPPED. will be pwr dwn after lono timer is armed + * + * @dc: Current DC state + * @update_state: update PG sequence states for HW block + */ +void dcn351_hw_block_power_down(struct dc *dc, + struct pg_block_update *update_state) +{ + int i = 0; + struct pg_cntl *pg_cntl = dc->res_pool->pg_cntl; + + if (!pg_cntl || dc->debug.ignore_pg) + return; + + for (i = dc->res_pool->pipe_count - 1; i >= 0; i--) { + if (update_state->pg_pipe_res_update[PG_DSC][i]) { + if (pg_cntl->funcs->dsc_pg_control) + pg_cntl->funcs->dsc_pg_control(pg_cntl, i, false); + } + + if (update_state->pg_pipe_res_update[PG_HUBP][i] && + update_state->pg_pipe_res_update[PG_DPP][i]) { + if (pg_cntl->funcs->hubp_dpp_pg_control) + pg_cntl->funcs->hubp_dpp_pg_control(pg_cntl, i, false); + } + } + + // domain25 currently always on. + + /* this will need all the clients to unregister optc interrupts, let dmubfw handle this */ + if (pg_cntl->funcs->plane_otg_pg_control) + pg_cntl->funcs->plane_otg_pg_control(pg_cntl, false); + + // domain23 currently always on. + // domain22 currently always on. +} + +/** + * dcn351_hw_block_power_up() - power up sequence + * + * The following sequence describes the ON-OFF (ONO) for power up: + * + * ONO Region 0, DCPG 22: dccg dio dcio - SKIPPED + * ONO Region 1, DCPG 23: dchubbub dchvm dchubbubmem - SKIPPED. PMFW will power up at IPS2 exit + * ONO Region 2, DCPG 24: mpc opp optc dwb + * ONO Region 3, DCPG 25: hpo - SKIPPED + * ONO Region 4, DCPG 0: dchubp0, dpp0 + * ONO Region 5, DCPG 16: dsc0 + * ONO Region 6, DCPG 1: dchubp1, dpp1 + * ONO Region 7, DCPG 17: dsc1 + * ONO Region 8, DCPG 2: dchubp2, dpp2 + * ONO Region 9, DCPG 18: dsc2 + * ONO Region 10, DCPG 3: dchubp3, dpp3 + * ONO Region 11, DCPG 19: dsc3 + * + * @dc: Current DC state + * @update_state: update PG sequence states for HW block + */ +void dcn351_hw_block_power_up(struct dc *dc, + struct pg_block_update *update_state) +{ + int i = 0; + struct pg_cntl *pg_cntl = dc->res_pool->pg_cntl; + + if (!pg_cntl || dc->debug.ignore_pg) + return; + + // domain22 currently always on. + // domain23 currently always on. + + /* this will need all the clients to unregister optc interrupts, let dmubfw handle this */ + if (pg_cntl->funcs->plane_otg_pg_control) + pg_cntl->funcs->plane_otg_pg_control(pg_cntl, true); + + // domain25 currently always on. + + for (i = 0; i < dc->res_pool->pipe_count; i++) { + if (update_state->pg_pipe_res_update[PG_HUBP][i] && + update_state->pg_pipe_res_update[PG_DPP][i]) { + if (pg_cntl->funcs->hubp_dpp_pg_control) + pg_cntl->funcs->hubp_dpp_pg_control(pg_cntl, i, true); + } + + if (update_state->pg_pipe_res_update[PG_DSC][i]) { + if (pg_cntl->funcs->dsc_pg_control) + pg_cntl->funcs->dsc_pg_control(pg_cntl, i, true); + } + } +} diff --git a/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_dpp.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn351/dcn351_hwseq.h index 3341ef71009b..6d8f3bfb668e 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_dpp.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn351/dcn351_hwseq.h @@ -1,6 +1,6 @@ /* SPDX-License-Identifier: MIT */ /* - * Copyright 2023 Advanced Micro Devices, Inc. + * Copyright 2024 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"), @@ -24,30 +24,18 @@ * */ -#include "core_types.h" -#include "dcn35_dpp.h" -#include "reg_helper.h" +#ifndef __DC_HWSS_DCN351_H__ +#define __DC_HWSS_DCN351_H__ -#define REG(reg) dpp->tf_regs->reg +#include "hw_sequencer_private.h" -#define CTX dpp->base.ctx +void dcn351_calc_blocks_to_gate(struct dc *dc, struct dc_state *context, + struct pg_block_update *update_state); +void dcn351_calc_blocks_to_ungate(struct dc *dc, struct dc_state *context, + struct pg_block_update *update_state); +void dcn351_hw_block_power_up(struct dc *dc, + struct pg_block_update *update_state); +void dcn351_hw_block_power_down(struct dc *dc, + struct pg_block_update *update_state); -#undef FN -#define FN(reg_name, field_name) \ - ((const struct dcn35_dpp_shift *)(dpp->tf_shift))->field_name, \ - ((const struct dcn35_dpp_mask *)(dpp->tf_mask))->field_name - -bool dpp35_construct(struct dcn3_dpp *dpp, struct dc_context *ctx, - uint32_t inst, const struct dcn3_dpp_registers *tf_regs, - const struct dcn35_dpp_shift *tf_shift, - const struct dcn35_dpp_mask *tf_mask) -{ - return dpp32_construct(dpp, ctx, inst, tf_regs, - (const struct dcn3_dpp_shift *)(tf_shift), - (const struct dcn3_dpp_mask *)(tf_mask)); -} - -void dpp35_set_fgcg(struct dcn3_dpp *dpp, bool enable) -{ - REG_UPDATE(DPP_CONTROL, DPP_FGCG_REP_DIS, !enable); -} +#endif /* __DC_HWSS_DCN351_H__ */ diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn351/dcn351_init.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn351/dcn351_init.c index ab17fa1c64e8..a53092cd619b 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn351/dcn351_init.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn351/dcn351_init.c @@ -32,6 +32,7 @@ #include "dcn31/dcn31_hwseq.h" #include "dcn32/dcn32_hwseq.h" #include "dcn35/dcn35_hwseq.h" +#include "dcn351/dcn351_hwseq.h" #include "dcn351_init.h" @@ -67,7 +68,7 @@ static const struct hw_sequencer_funcs dcn351_funcs = { .prepare_bandwidth = dcn35_prepare_bandwidth, .optimize_bandwidth = dcn35_optimize_bandwidth, .update_bandwidth = dcn20_update_bandwidth, - .set_drr = dcn10_set_drr, + .set_drr = dcn35_set_drr, .get_position = dcn10_get_position, .set_static_screen_control = dcn35_set_static_screen_control, .setup_stereo = dcn10_setup_stereo, @@ -120,8 +121,6 @@ static const struct hw_sequencer_funcs dcn351_funcs = { .hw_block_power_up = dcn35_hw_block_power_up, .hw_block_power_down = dcn35_hw_block_power_down, .root_clock_control = dcn35_root_clock_control, - .set_idle_state = dcn35_set_idle_state, - .get_idle_state = dcn35_get_idle_state }; static const struct hwseq_private_funcs dcn351_private_funcs = { @@ -147,6 +146,7 @@ static const struct hwseq_private_funcs dcn351_private_funcs = { //.hubp_pg_control = dcn35_hubp_pg_control, .enable_power_gating_plane = dcn35_enable_power_gating_plane, .dpp_root_clock_control = dcn35_dpp_root_clock_control, + .dpstream_root_clock_control = dcn35_dpstream_root_clock_control, .program_all_writeback_pipes_in_tree = dcn30_program_all_writeback_pipes_in_tree, .update_odm = dcn35_update_odm, .set_hdr_multiplier = dcn10_set_hdr_multiplier, diff --git a/drivers/gpu/drm/amd/display/dc/hwss/hw_sequencer.h b/drivers/gpu/drm/amd/display/dc/hwss/hw_sequencer.h index f89f205e42a1..7c339e7e7117 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/hw_sequencer.h +++ b/drivers/gpu/drm/amd/display/dc/hwss/hw_sequencer.h @@ -377,7 +377,10 @@ struct hw_sequencer_funcs { /* Idle Optimization Related */ bool (*apply_idle_power_optimizations)(struct dc *dc, bool enable); - bool (*does_plane_fit_in_mall)(struct dc *dc, struct dc_plane_state *plane, + bool (*does_plane_fit_in_mall)(struct dc *dc, + unsigned int pitch, + unsigned int height, + enum surface_pixel_format format, struct dc_cursor_attributes *cursor_attr); void (*commit_subvp_config)(struct dc *dc, struct dc_state *context); void (*enable_phantom_streams)(struct dc *dc, struct dc_state *context); @@ -424,11 +427,10 @@ struct hw_sequencer_funcs { struct pg_block_update *update_state); void (*root_clock_control)(struct dc *dc, struct pg_block_update *update_state, bool power_on); - void (*set_idle_state)(const struct dc *dc, bool allow_idle); - uint32_t (*get_idle_state)(const struct dc *dc); bool (*is_pipe_topology_transition_seamless)(struct dc *dc, const struct dc_state *cur_ctx, const struct dc_state *new_ctx); + void (*set_long_vtotal)(struct pipe_ctx **pipe_ctx, int num_pipes, uint32_t v_total_min, uint32_t v_total_max); }; void color_space_to_black_color( @@ -478,9 +480,10 @@ void hwss_build_fast_sequence(struct dc *dc, struct dc_dmub_cmd *dc_dmub_cmd, unsigned int dmub_cmd_count, struct block_sequence block_sequence[], - int *num_steps, + unsigned int *num_steps, struct pipe_ctx *pipe_ctx, - struct dc_stream_status *stream_status); + struct dc_stream_status *stream_status, + struct dc_state *context); void hwss_send_dmcub_cmd(union block_sequence_params *params); diff --git a/drivers/gpu/drm/amd/display/dc/hwss/hw_sequencer_private.h b/drivers/gpu/drm/amd/display/dc/hwss/hw_sequencer_private.h index 554cfab5ab24..341219cf4144 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/hw_sequencer_private.h +++ b/drivers/gpu/drm/amd/display/dc/hwss/hw_sequencer_private.h @@ -120,6 +120,10 @@ struct hwseq_private_funcs { struct dce_hwseq *hws, unsigned int dpp_inst, bool clock_on); + void (*dpstream_root_clock_control)( + struct dce_hwseq *hws, + unsigned int dpp_inst, + bool clock_on); void (*dpp_pg_control)(struct dce_hwseq *hws, unsigned int dpp_inst, bool power_on); diff --git a/drivers/gpu/drm/amd/display/dc/inc/core_types.h b/drivers/gpu/drm/amd/display/dc/inc/core_types.h index b1b72e688f74..028b2f971e36 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/core_types.h +++ b/drivers/gpu/drm/amd/display/dc/inc/core_types.h @@ -90,6 +90,9 @@ struct resource_funcs { void (*update_soc_for_wm_a)( struct dc *dc, struct dc_state *context); + unsigned int (*calculate_mall_ways_from_bytes)( + const struct dc *dc, + unsigned int total_size_in_mall_bytes); /** * @populate_dml_pipes - Populate pipe data struct * @@ -336,7 +339,9 @@ struct stream_resource { }; struct plane_resource { + /* scl_data is scratch space required to program a plane */ struct scaler_data scl_data; + /* Below pointers to hw objects are required to enable the plane */ struct hubp *hubp; struct mem_input *mi; struct input_pixel_processor *ipp; @@ -496,7 +501,7 @@ struct dcn_bw_writeback { struct dcn_bw_output { struct dc_clocks clk; - struct dcn_watermark_set watermarks; + union dcn_watermark_set watermarks; struct dcn_bw_writeback bw_writeback; int compbuf_size_kb; unsigned int mall_ss_size_bytes; @@ -515,6 +520,7 @@ struct bw_context { union bw_output bw; struct display_mode_lib dml; struct dml2_context *dml2; + struct dml2_context *dml2_dc_power_source; }; struct dc_dmub_cmd { @@ -522,25 +528,6 @@ struct dc_dmub_cmd { enum dm_dmub_wait_type wait_type; }; -struct dc_scratch_space { - /* used to temporarily backup plane states of a stream during - * dc update. The reason is that plane states are overwritten - * with surface updates in dc update. Once they are overwritten - * current state is no longer valid. We want to temporarily - * store current value in plane states so we can still recover - * a valid current state during dc update. - */ - struct dc_plane_state plane_states[MAX_SURFACE_NUM]; - struct dc_gamma gamma_correction[MAX_SURFACE_NUM]; - struct dc_transfer_func in_transfer_func[MAX_SURFACE_NUM]; - struct dc_3dlut lut3d_func[MAX_SURFACE_NUM]; - struct dc_transfer_func in_shaper_func[MAX_SURFACE_NUM]; - struct dc_transfer_func blend_tf[MAX_SURFACE_NUM]; - - struct dc_stream_state stream_state; - struct dc_transfer_func out_transfer_func; -}; - /** * struct dc_state - The full description of a state requested by users */ @@ -623,8 +610,7 @@ struct dc_state { unsigned int stutter_period_us; } perf_params; - - struct dc_scratch_space scratch; + enum dc_power_source_type power_source; }; struct replay_context { diff --git a/drivers/gpu/drm/amd/display/dc/inc/dcn_calcs.h b/drivers/gpu/drm/amd/display/dc/inc/dcn_calcs.h index 9e4ddc985240..55529c5f471c 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/dcn_calcs.h +++ b/drivers/gpu/drm/amd/display/dc/inc/dcn_calcs.h @@ -31,7 +31,7 @@ #define __DCN_CALCS_H__ #include "bw_fixed.h" -#include "../dml/display_mode_lib.h" +#include "dml/display_mode_lib.h" struct dc; diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr.h b/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr.h index 17e014d3bdc8..4f7480f60c85 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr.h @@ -281,8 +281,6 @@ struct clk_mgr_funcs { void (*set_low_power_state)(struct clk_mgr *clk_mgr); void (*exit_low_power_state)(struct clk_mgr *clk_mgr); bool (*is_ips_supported)(struct clk_mgr *clk_mgr); - void (*set_idle_state)(struct clk_mgr *clk_mgr, bool allow_idle); - uint32_t (*get_idle_state)(struct clk_mgr *clk_mgr); void (*init_clocks)(struct clk_mgr *clk_mgr); diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr_internal.h b/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr_internal.h index f4d4a68c91dc..4ba18ea57aad 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr_internal.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr_internal.h @@ -349,7 +349,7 @@ struct clk_mgr_internal { enum dm_pp_clocks_state cur_min_clks_state; bool periodic_retraining_disabled; - unsigned int cur_phyclk_req_table[MAX_PIPES * 2]; + unsigned int cur_phyclk_req_table[MAX_LINKS]; bool smu_present; void *wm_range_table; diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/dccg.h b/drivers/gpu/drm/amd/display/dc/inc/hw/dccg.h index b9a06bf84cc9..d4c7885fc916 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/dccg.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/dccg.h @@ -59,6 +59,7 @@ enum dentist_dispclk_change_mode { struct dp_dto_params { int otg_inst; enum signal_type signal; + enum streamclk_source clk_src; uint64_t pixclk_hz; uint64_t refclk_hz; }; @@ -105,6 +106,10 @@ struct dccg_funcs { void (*otg_drop_pixel)(struct dccg *dccg, uint32_t otg_inst); void (*dccg_init)(struct dccg *dccg); + void (*set_dpstreamclk_root_clock_gating)( + struct dccg *dccg, + int dp_hpo_inst, + bool enable); void (*set_dpstreamclk)( struct dccg *dccg, diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/dchubbub.h b/drivers/gpu/drm/amd/display/dc/inc/hw/dchubbub.h index 2ae7484d18af..305fdc127bfc 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/dchubbub.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/dchubbub.h @@ -160,7 +160,7 @@ struct hubbub_funcs { bool (*program_watermarks)( struct hubbub *hubbub, - struct dcn_watermark_set *watermarks, + union dcn_watermark_set *watermarks, unsigned int refclk_mhz, bool safe_to_lower); diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h b/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h index 0f24afbf4388..ca8de345d039 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h @@ -322,7 +322,7 @@ struct dpp_funcs { const struct pwl_params *params); bool (*dpp_program_3dlut)( struct dpp *dpp, - struct tetrahedral_params *params); + const struct tetrahedral_params *params); void (*dpp_cnv_set_alpha_keyer)( struct dpp *dpp_base, struct cnv_color_keyer_params *color_keyer); diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/dwb.h b/drivers/gpu/drm/amd/display/dc/inc/hw/dwb.h index 729ca0064e94..063efc8128a7 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/dwb.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/dwb.h @@ -147,9 +147,10 @@ struct dwb_caps { unsigned int support_ogam :1; unsigned int support_wbscl :1; unsigned int support_ocsc :1; - unsigned int support_stereo :1; + unsigned int support_stereo :1; + unsigned int support_4k_120p :1; } caps; - unsigned int reserved2[9]; /* Reserved for future use, MUST BE 0. */ + unsigned int reserved2[10]; /* Reserved for future use, MUST BE 0. */ }; struct dwbc { @@ -166,8 +167,9 @@ struct dwbc { bool dwb_is_drc; int wb_src_plane_inst;/*hubp, mpcc, inst*/ uint32_t mask_id; - int otg_inst; - bool mvc_cfg; + int otg_inst; + bool mvc_cfg; + struct dc_dwb_params params; }; struct dwbc_funcs { @@ -192,6 +194,10 @@ struct dwbc_funcs { struct dwbc *dwbc, enum dwb_frame_capture_enable enable); + void (*dwb_set_scaler)( + struct dwbc *dwbc, + struct dc_dwb_params *params); + void (*set_stereo)( struct dwbc *dwbc, struct dwb_stereo_params *stereo_params); @@ -205,9 +211,11 @@ struct dwbc_funcs { struct dwbc *dwbc, struct dwb_warmup_params *warmup_params); - + bool (*dwb_get_mcifbuf_line)( + struct dwbc *dwbc, unsigned int *buf_idx, + unsigned int *cur_line, + unsigned int *over_run); #if defined(CONFIG_DRM_AMD_DC_FP) - void (*dwb_program_output_csc)( struct dwbc *dwbc, enum dc_color_space color_space, @@ -216,17 +224,17 @@ struct dwbc_funcs { bool (*dwb_ogam_set_output_transfer_func)( struct dwbc *dwbc, const struct dc_transfer_func *in_transfer_func_dwb_ogam); - +#endif //TODO: merge with output_transfer_func? bool (*dwb_ogam_set_input_transfer_func)( struct dwbc *dwbc, const struct dc_transfer_func *in_transfer_func_dwb_ogam); -#endif + + void (*get_drr_time_stamp)( + struct dwbc *dwbc, uint32_t *time_stamp); + bool (*get_dwb_status)( struct dwbc *dwbc); - void (*dwb_set_scaler)( - struct dwbc *dwbc, - struct dc_dwb_params *params); }; #endif diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h b/drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h index dcae23faeee3..c80ebb407add 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h @@ -44,10 +44,11 @@ */ #define MAX_PIPES 6 #define MAX_PHANTOM_PIPES (MAX_PIPES / 2) +#define MAX_LINKS (MAX_PIPES * 2) #define MAX_DIG_LINK_ENCODERS 7 #define MAX_DWB_PIPES 1 #define MAX_HPO_DP2_ENCODERS 4 -#define MAX_HPO_DP2_LINK_ENCODERS 2 +#define MAX_HPO_DP2_LINK_ENCODERS 4 struct gamma_curve { uint32_t offset; diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/link_encoder.h b/drivers/gpu/drm/amd/display/dc/inc/hw/link_encoder.h index dbe7afa9d3a2..af9183f5d69b 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/link_encoder.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/link_encoder.h @@ -163,12 +163,11 @@ struct link_encoder_funcs { enum signal_type (*get_dig_mode)( struct link_encoder *enc); + void (*set_dio_phy_mux)( struct link_encoder *enc, enum encoder_type_select sel, uint32_t hpo_inst); - void (*set_dig_output_mode)( - struct link_encoder *enc, uint8_t pix_per_container); }; /* diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/mem_input.h b/drivers/gpu/drm/amd/display/dc/inc/hw/mem_input.h index b72fb314d804..86c12cd6f47d 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/mem_input.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/mem_input.h @@ -50,11 +50,13 @@ struct dcn_watermarks { uint32_t usr_retraining_ns; }; -struct dcn_watermark_set { - struct dcn_watermarks a; - struct dcn_watermarks b; - struct dcn_watermarks c; - struct dcn_watermarks d; +union dcn_watermark_set { + struct { + struct dcn_watermarks a; + struct dcn_watermarks b; + struct dcn_watermarks c; + struct dcn_watermarks d; + }; // legacy }; struct dce_watermarks { diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/optc.h b/drivers/gpu/drm/amd/display/dc/inc/hw/optc.h index 9a8bf6ec70ea..8d32e525f05a 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/optc.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/optc.h @@ -93,6 +93,8 @@ struct dcn_otg_state { uint32_t vertical_interrupt1_line; uint32_t vertical_interrupt2_en; uint32_t vertical_interrupt2_line; + uint32_t otg_master_update_lock; + uint32_t otg_double_buffer_control; }; void optc1_read_otg_state(struct optc *optc1, struct dcn_otg_state *s); diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/stream_encoder.h b/drivers/gpu/drm/amd/display/dc/inc/hw/stream_encoder.h index a15efadb9183..75b9ec21f297 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/stream_encoder.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/stream_encoder.h @@ -178,10 +178,6 @@ struct stream_encoder_funcs { void (*stop_dp_info_packets)( struct stream_encoder *enc); - void (*reset_fifo)( - struct stream_encoder *enc - ); - void (*dp_blank)( struct dc_link *link, struct stream_encoder *enc); diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h b/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h index ffad8fe16c54..cd68ecc242c1 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h @@ -64,6 +64,12 @@ struct drr_params { bool immediate_flip; }; +struct long_vtotal_params { + uint32_t vertical_total_min; + uint32_t vertical_total_max; + uint32_t vertical_blank_start; +}; + #define LEFT_EYE_3D_PRIMARY_SURFACE 1 #define RIGHT_EYE_3D_PRIMARY_SURFACE 0 @@ -331,6 +337,7 @@ struct timing_generator_funcs { void (*init_odm)(struct timing_generator *tg); void (*wait_drr_doublebuffer_pending_clear)(struct timing_generator *tg); + void (*set_long_vtotal)(struct timing_generator *optc, const struct long_vtotal_params *params); void (*wait_odm_doublebuffer_pending_clear)(struct timing_generator *tg); }; diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/vpg.h b/drivers/gpu/drm/amd/display/dc/inc/hw/vpg.h new file mode 100644 index 000000000000..51da368f5c3e --- /dev/null +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/vpg.h @@ -0,0 +1,53 @@ +/* + * Copyright 2024 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: AMD + */ + +#ifndef __DC_VPG_H__ +#define __DC_VPG_H__ + +struct dc_context; +struct dc_info_packet; + +struct vpg; + +struct vpg_funcs { + void (*update_generic_info_packet)( + struct vpg *vpg, + uint32_t packet_index, + const struct dc_info_packet *info_packet, + bool immediate_update); + + void (*vpg_poweron)( + struct vpg *vpg); + + void (*vpg_powerdown)( + struct vpg *vpg); +}; + +struct vpg { + const struct vpg_funcs *funcs; + struct dc_context *ctx; + int inst; +}; + +#endif /* DC_INC_VPG_H_ */
\ No newline at end of file diff --git a/drivers/gpu/drm/amd/display/dc/inc/link.h b/drivers/gpu/drm/amd/display/dc/inc/link.h index bf29fc58ea6a..7ab8ba5e23ed 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/link.h +++ b/drivers/gpu/drm/amd/display/dc/inc/link.h @@ -288,7 +288,7 @@ struct link_service { struct dc_link *link, uint32_t coasting_vtotal); bool (*edp_replay_residency)(const struct dc_link *link, unsigned int *residency, const bool is_start, - const bool is_alpm); + const enum pr_residency_mode mode); bool (*edp_set_replay_power_opt_and_coasting_vtotal)(struct dc_link *link, const unsigned int *power_opts, uint32_t coasting_vtotal); diff --git a/drivers/gpu/drm/amd/display/dc/inc/resource.h b/drivers/gpu/drm/amd/display/dc/inc/resource.h index 77a60aa9f27b..361ad6b16b96 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/resource.h +++ b/drivers/gpu/drm/amd/display/dc/inc/resource.h @@ -510,6 +510,17 @@ int recource_find_free_pipe_used_as_otg_master_in_cur_res_ctx( /* * Look for a free pipe in new resource context that is used as a secondary DPP + * pipe in current resource context. + * return - FREE_PIPE_INDEX_NOT_FOUND if free pipe is not found, otherwise + * pipe idx of the free pipe + */ +int resource_find_free_pipe_used_as_cur_sec_dpp( + const struct resource_context *cur_res_ctx, + struct resource_context *new_res_ctx, + const struct resource_pool *pool); + +/* + * Look for a free pipe in new resource context that is used as a secondary DPP * pipe in any MPCC combine in current resource context. * return - FREE_PIPE_INDEX_NOT_FOUND if free pipe is not found, otherwise * pipe idx of the free pipe @@ -573,13 +584,6 @@ bool get_temp_dp_link_res(struct dc_link *link, struct link_resource *link_res, struct dc_link_settings *link_settings); -#if defined(CONFIG_DRM_AMD_DC_FP) -struct hpo_dp_link_encoder *resource_get_hpo_dp_link_enc_for_det_lt( - const struct resource_context *res_ctx, - const struct resource_pool *pool, - const struct dc_link *link); -#endif - void reset_syncd_pipes_from_disabled_pipes(struct dc *dc, struct dc_state *context); @@ -615,4 +619,10 @@ enum dc_status update_dp_encoder_resources_for_test_harness(const struct dc *dc, struct pipe_ctx *pipe_ctx); bool check_subvp_sw_cursor_fallback_req(const struct dc *dc, struct dc_stream_state *stream); + +/* Setup dc callbacks for dml2 + * @dc: the display core structure + * @dml2_options: struct to hold callbacks + */ +void resource_init_common_dml2_callbacks(struct dc *dc, struct dml2_configuration_options *dml2_options); #endif /* DRIVERS_GPU_DRM_AMD_DC_DEV_DC_INC_RESOURCE_H_ */ diff --git a/drivers/gpu/drm/amd/display/dc/irq/dce110/irq_service_dce110.c b/drivers/gpu/drm/amd/display/dc/irq/dce110/irq_service_dce110.c index 1c0d89e675da..bb576a9c5fdb 100644 --- a/drivers/gpu/drm/amd/display/dc/irq/dce110/irq_service_dce110.c +++ b/drivers/gpu/drm/amd/display/dc/irq/dce110/irq_service_dce110.c @@ -211,8 +211,12 @@ bool dce110_vblank_set(struct irq_service *irq_service, info->ext_id); uint8_t pipe_offset = dal_irq_src - IRQ_TYPE_VBLANK; - struct timing_generator *tg = - dc->current_state->res_ctx.pipe_ctx[pipe_offset].stream_res.tg; + struct timing_generator *tg; + + if (pipe_offset >= MAX_PIPES) + return false; + + tg = dc->current_state->res_ctx.pipe_ctx[pipe_offset].stream_res.tg; if (enable) { if (!tg || !tg->funcs->arm_vert_intr(tg, 2)) { diff --git a/drivers/gpu/drm/amd/display/dc/link/accessories/link_dp_cts.c b/drivers/gpu/drm/amd/display/dc/link/accessories/link_dp_cts.c index 22b24749c9d2..8d1a1cc94a8b 100644 --- a/drivers/gpu/drm/amd/display/dc/link/accessories/link_dp_cts.c +++ b/drivers/gpu/drm/amd/display/dc/link/accessories/link_dp_cts.c @@ -884,7 +884,7 @@ void dp_set_preferred_link_settings(struct dc *dc, { int i; struct pipe_ctx *pipe; - struct dc_stream_state *link_stream; + struct dc_stream_state *link_stream = 0; struct dc_link_settings store_settings = *link_setting; link->preferred_link_setting = store_settings; diff --git a/drivers/gpu/drm/amd/display/dc/link/accessories/link_dp_trace.c b/drivers/gpu/drm/amd/display/dc/link/accessories/link_dp_trace.c index fbcd8fb58ea8..c8c55f196f8d 100644 --- a/drivers/gpu/drm/amd/display/dc/link/accessories/link_dp_trace.c +++ b/drivers/gpu/drm/amd/display/dc/link/accessories/link_dp_trace.c @@ -24,7 +24,6 @@ */ #include "link_dp_trace.h" #include "link/protocols/link_dpcd.h" -#include "link.h" void dp_trace_init(struct dc_link *link) { diff --git a/drivers/gpu/drm/amd/display/dc/link/link_detection.c b/drivers/gpu/drm/amd/display/dc/link/link_detection.c index b8c4a04dd175..0d523dc43d02 100644 --- a/drivers/gpu/drm/amd/display/dc/link/link_detection.c +++ b/drivers/gpu/drm/amd/display/dc/link/link_detection.c @@ -516,8 +516,8 @@ static void query_hdcp_capability(enum signal_type signal, struct dc_link *link) static void read_current_link_settings_on_detect(struct dc_link *link) { union lane_count_set lane_count_set = {0}; - uint8_t link_bw_set; - uint8_t link_rate_set; + uint8_t link_bw_set = 0; + uint8_t link_rate_set = 0; uint32_t read_dpcd_retry_cnt = 10; enum dc_status status = DC_ERROR_UNEXPECTED; int i; diff --git a/drivers/gpu/drm/amd/display/dc/link/link_dpms.c b/drivers/gpu/drm/amd/display/dc/link/link_dpms.c index a72de44a5747..b53ad18dbfbc 100644 --- a/drivers/gpu/drm/amd/display/dc/link/link_dpms.c +++ b/drivers/gpu/drm/amd/display/dc/link/link_dpms.c @@ -55,6 +55,8 @@ #include "dccg.h" #include "clk_mgr.h" #include "atomfirmware.h" +#include "vpg.h" + #define DC_LOGGER \ dc_logger #define DC_LOGGER_INIT(logger) \ @@ -67,7 +69,6 @@ #define RETIMER_REDRIVER_INFO(...) \ DC_LOG_RETIMER_REDRIVER( \ __VA_ARGS__) -#include "dc/dcn30/dcn30_vpg.h" #define MAX_MTP_SLOT_COUNT 64 #define LINK_TRAINING_ATTEMPTS 4 @@ -127,7 +128,7 @@ void link_blank_dp_stream(struct dc_link *link, bool hw_init) if (link->ep_type == DISPLAY_ENDPOINT_PHY && link->link_enc->funcs->get_dig_frontend && link->link_enc->funcs->is_dig_enabled(link->link_enc)) { - unsigned int fe = link->link_enc->funcs->get_dig_frontend(link->link_enc); + int fe = link->link_enc->funcs->get_dig_frontend(link->link_enc); if (fe != ENGINE_ID_UNKNOWN) for (j = 0; j < dc->res_pool->stream_enc_count; j++) { @@ -725,7 +726,7 @@ static void set_avmute(struct pipe_ctx *pipe_ctx, bool enable) static void enable_mst_on_sink(struct dc_link *link, bool enable) { - unsigned char mstmCntl; + unsigned char mstmCntl = 0; core_link_read_dpcd(link, DP_MSTM_CTRL, &mstmCntl, 1); if (enable) @@ -803,7 +804,7 @@ void link_set_dsc_on_stream(struct pipe_ctx *pipe_ctx, bool enable) if (enable) { struct dsc_config dsc_cfg; - struct dsc_optc_config dsc_optc_cfg; + struct dsc_optc_config dsc_optc_cfg = {0}; enum optc_dsc_mode optc_dsc_mode; /* Enable DSC hw block */ @@ -1575,7 +1576,7 @@ static bool write_128b_132b_sst_payload_allocation_table( break; } } else { - union dpcd_rev dpcdRev; + union dpcd_rev dpcdRev = {0}; if (core_link_read_dpcd( link, @@ -2119,7 +2120,7 @@ static enum dc_status enable_link_dp_mst( struct pipe_ctx *pipe_ctx) { struct dc_link *link = pipe_ctx->stream->link; - unsigned char mstm_cntl; + unsigned char mstm_cntl = 0; /* sink signal type after MST branch is MST. Multiple MST sinks * share one link. Link DP PHY is enable or training only once. @@ -2285,6 +2286,7 @@ void link_set_dpms_off(struct pipe_ctx *pipe_ctx) struct dc_stream_state *stream = pipe_ctx->stream; struct dc_link *link = stream->sink->link; struct vpg *vpg = pipe_ctx->stream_res.stream_enc->vpg; + enum dp_panel_mode panel_mode_dp = dp_get_panel_mode(link); DC_LOGGER_INIT(pipe_ctx->stream->ctx->logger); @@ -2311,6 +2313,8 @@ void link_set_dpms_off(struct pipe_ctx *pipe_ctx) dc->hwss.disable_audio_stream(pipe_ctx); + edp_set_panel_assr(link, pipe_ctx, &panel_mode_dp, false); + update_psp_stream_config(pipe_ctx, true); dc->hwss.blank_stream(pipe_ctx); diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_capability.c b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_capability.c index 289f5d133342..a01d0842bf8e 100644 --- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_capability.c +++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_capability.c @@ -992,7 +992,7 @@ enum dp_link_encoding mst_decide_link_encoding_format(const struct dc_link *link static void read_dp_device_vendor_id(struct dc_link *link) { - struct dp_device_vendor_id dp_id; + struct dp_device_vendor_id dp_id = {0}; /* read IEEE branch device id */ core_link_read_dpcd( @@ -1087,7 +1087,7 @@ static void get_active_converter_info( } if (link->dpcd_caps.dpcd_rev.raw >= DPCD_REV_11) { - uint8_t det_caps[16]; /* CTS 4.2.2.7 expects source to read Detailed Capabilities Info : 00080h-0008F.*/ + uint8_t det_caps[16] = {0}; /* CTS 4.2.2.7 expects source to read Detailed Capabilities Info : 00080h-0008F.*/ union dwnstream_port_caps_byte0 *port_caps = (union dwnstream_port_caps_byte0 *)det_caps; if (core_link_read_dpcd(link, DP_DOWNSTREAM_PORT_0, @@ -1172,7 +1172,7 @@ static void get_active_converter_info( set_dongle_type(link->ddc, link->dpcd_caps.dongle_type); { - struct dp_sink_hw_fw_revision dp_hw_fw_revision; + struct dp_sink_hw_fw_revision dp_hw_fw_revision = {0}; core_link_read_dpcd( link, @@ -1242,7 +1242,7 @@ static void apply_usbc_combo_phy_reset_wa(struct dc_link *link, bool dp_overwrite_extended_receiver_cap(struct dc_link *link) { - uint8_t dpcd_data[16]; + uint8_t dpcd_data[16] = {0}; uint32_t read_dpcd_retry_cnt = 3; enum dc_status status = DC_ERROR_UNEXPECTED; union dp_downstream_port_present ds_port = { 0 }; @@ -1408,7 +1408,7 @@ static bool get_usbc_cable_id(struct dc_link *link, union dp_cable_id *cable_id) static void retrieve_cable_id(struct dc_link *link) { - union dp_cable_id usbc_cable_id; + union dp_cable_id usbc_cable_id = {0}; link->dpcd_caps.cable_id.raw = 0; core_link_read_dpcd(link, DP_CABLE_ATTRIBUTES_UPDATED_BY_DPRX, @@ -1475,7 +1475,7 @@ static bool dpcd_read_sink_ext_caps(struct dc_link *link) enum dc_status dp_retrieve_lttpr_cap(struct dc_link *link) { - uint8_t lttpr_dpcd_data[8]; + uint8_t lttpr_dpcd_data[8] = {0}; enum dc_status status; bool is_lttpr_present; @@ -1931,8 +1931,8 @@ void detect_edp_sink_caps(struct dc_link *link) uint32_t entry; uint32_t link_rate_in_khz; enum dc_link_rate link_rate = LINK_RATE_UNKNOWN; - uint8_t backlight_adj_cap; - uint8_t general_edp_cap; + uint8_t backlight_adj_cap = 0; + uint8_t general_edp_cap = 0; retrieve_link_cap(link); link->dpcd_caps.edp_supported_link_rates_count = 0; diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_dpia_bw.c b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_dpia_bw.c index 5491b707cec8..0f1c411523a2 100644 --- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_dpia_bw.c +++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_dpia_bw.c @@ -166,7 +166,7 @@ static uint8_t get_lowest_dpia_index(struct dc_link *link) uint8_t idx = 0xFF; int i; - for (i = 0; i < MAX_PIPES * 2; ++i) { + for (i = 0; i < MAX_LINKS; ++i) { if (!dc_struct->links[i] || dc_struct->links[i]->ep_type != DISPLAY_ENDPOINT_USB4_DPIA) @@ -196,7 +196,7 @@ static int get_host_router_total_dp_tunnel_bw(const struct dc *dc, uint8_t hr_in struct dc_link *link_dpia_primary, *link_dpia_secondary; int total_bw = 0; - for (uint8_t i = 0; i < (MAX_PIPES * 2) - 1; ++i) { + for (uint8_t i = 0; i < MAX_LINKS - 1; ++i) { if (!dc->links[i] || dc->links[i]->ep_type != DISPLAY_ENDPOINT_USB4_DPIA) continue; @@ -270,7 +270,7 @@ static void set_usb4_req_bw_req(struct dc_link *link, int req_bw) /* Error check whether requested and allocated are equal */ req_bw = requested_bw * (Kbps_TO_Gbps / link->dpia_bw_alloc_config.bw_granularity); - if (req_bw == link->dpia_bw_alloc_config.allocated_bw) { + if (req_bw && (req_bw == link->dpia_bw_alloc_config.allocated_bw)) { DC_LOG_ERROR("%s: Request bw equals to allocated bw for link(%d)\n", __func__, link->link_index); } @@ -341,6 +341,14 @@ bool link_dp_dpia_set_dptx_usb4_bw_alloc_support(struct dc_link *link) ret = true; init_usb4_bw_struct(link); link->dpia_bw_alloc_config.bw_alloc_enabled = true; + + /* + * During DP tunnel creation, CM preallocates BW and reduces estimated BW of other + * DPIA. CM release preallocation only when allocation is complete. Do zero alloc + * to make the CM to release preallocation and update estimated BW correctly for + * all DPIAs per host router + */ + link_dp_dpia_allocate_usb4_bandwidth_for_stream(link, 0); } } diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_irq_handler.c b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_irq_handler.c index ba69874be5a4..0fcf0b8530ac 100644 --- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_irq_handler.c +++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_irq_handler.c @@ -120,7 +120,7 @@ bool dp_parse_link_loss_status( static bool handle_hpd_irq_psr_sink(struct dc_link *link) { - union dpcd_psr_configuration psr_configuration; + union dpcd_psr_configuration psr_configuration = {0}; if (!link->psr_settings.psr_feature_enabled) return false; @@ -186,9 +186,9 @@ static bool handle_hpd_irq_psr_sink(struct dc_link *link) static void handle_hpd_irq_replay_sink(struct dc_link *link) { - union dpcd_replay_configuration replay_configuration; + union dpcd_replay_configuration replay_configuration = {0}; /*AMD Replay version reuse DP_PSR_ERROR_STATUS for REPLAY_ERROR status.*/ - union psr_error_status replay_error_status; + union psr_error_status replay_error_status = {0}; if (!link->replay_settings.replay_feature_enabled) return; @@ -280,7 +280,7 @@ void dp_handle_link_loss(struct dc_link *link) static void read_dpcd204h_on_irq_hpd(struct dc_link *link, union hpd_irq_data *irq_data) { enum dc_status retval; - union lane_align_status_updated dpcd_lane_status_updated; + union lane_align_status_updated dpcd_lane_status_updated = {0}; retval = core_link_read_dpcd( link, @@ -320,7 +320,7 @@ enum dc_status dp_read_hpd_rx_irq_data( /* Read 14 bytes in a single read and then copy only the required fields. * This is more efficient than doing it in two separate AUX reads. */ - uint8_t tmp[DP_SINK_STATUS_ESI - DP_SINK_COUNT_ESI + 1]; + uint8_t tmp[DP_SINK_STATUS_ESI - DP_SINK_COUNT_ESI + 1] = {0}; retval = core_link_read_dpcd( link, diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training.c b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training.c index e538c67d3ed9..1818970b8eaf 100644 --- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training.c +++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training.c @@ -1071,7 +1071,7 @@ enum dc_status dpcd_set_link_settings( * MUX chip gets link rate set back before link training. */ if (link->connector_signal == SIGNAL_TYPE_EDP) { - uint8_t supported_link_rates[16]; + uint8_t supported_link_rates[16] = {0}; core_link_read_dpcd(link, DP_SUPPORTED_LINK_RATES, supported_link_rates, sizeof(supported_link_rates)); @@ -1587,21 +1587,7 @@ bool perform_link_training_with_retries( msleep(delay_dp_power_up_in_ms); } - if (panel_mode == DP_PANEL_MODE_EDP) { - struct cp_psp *cp_psp = &stream->ctx->cp_psp; - - if (cp_psp && cp_psp->funcs.enable_assr) { - /* ASSR is bound to fail with unsigned PSP - * verstage used during devlopment phase. - * Report and continue with eDP panel mode to - * perform eDP link training with right settings - */ - bool result; - result = cp_psp->funcs.enable_assr(cp_psp->handle, link); - if (!result && link->panel_mode != DP_PANEL_MODE_EDP) - panel_mode = DP_PANEL_MODE_DEFAULT; - } - } + edp_set_panel_assr(link, pipe_ctx, &panel_mode, true); dp_set_panel_mode(link, panel_mode); diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_dpia.c b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_dpia.c index 5d36bab0029c..edb21d21952a 100644 --- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_dpia.c +++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_dpia.c @@ -291,7 +291,7 @@ static enum link_training_result dpia_training_cr_non_transparent( { enum link_training_result result = LINK_TRAINING_CR_FAIL_LANE0; uint8_t repeater_cnt = 0; /* Number of hops/repeaters in display path. */ - enum dc_status status; + enum dc_status status = DC_ERROR_UNEXPECTED; uint32_t retries_cr = 0; /* Number of consecutive attempts with same VS or PE. */ uint32_t retry_count = 0; uint32_t wait_time_microsec = TRAINING_AUX_RD_INTERVAL; /* From DP spec, CR read interval is always 100us. */ @@ -617,7 +617,7 @@ static enum link_training_result dpia_training_eq_non_transparent( enum link_training_result result = LINK_TRAINING_EQ_FAIL_EQ; uint8_t repeater_cnt = 0; /* Number of hops/repeaters in display path. */ uint32_t retries_eq = 0; - enum dc_status status; + enum dc_status status = DC_ERROR_UNEXPECTED; enum dc_dp_training_pattern tr_pattern; uint32_t wait_time_microsec = 0; enum dc_lane_count lane_count = lt_settings->link_settings.lane_count; diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dpcd.c b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dpcd.c index c5de6ed5bf58..a72c898b64fa 100644 --- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dpcd.c +++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dpcd.c @@ -130,7 +130,7 @@ static uint32_t dpcd_get_next_partition_size(const uint32_t address, const uint3 * XXX: Do not allow any two address ranges in this array to overlap */ static const struct dpcd_address_range mandatory_dpcd_blocks[] = { - { DP_LT_TUNABLE_PHY_REPEATER_FIELD_DATA_STRUCTURE_REV, DP_PHY_REPEATER_EXTENDED_WAIT_TIMEOUT }}; + { DP_LT_TUNABLE_PHY_REPEATER_FIELD_DATA_STRUCTURE_REV, DP_PHY_REPEATER_128B132B_RATES }}; /* * extend addresses to read all mandatory blocks together diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.c b/drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.c index 3baa2bdd6dd6..ad9aca790dd7 100644 --- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.c +++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.c @@ -38,6 +38,7 @@ #include "dc/dc_dmub_srv.h" #include "dce/dmub_replay.h" #include "abm.h" +#include "resource.h" #define DC_LOGGER \ link->ctx->logger #define DC_LOGGER_INIT(logger) @@ -320,8 +321,8 @@ bool edp_is_ilr_optimization_required(struct dc_link *link, struct dc_crtc_timing *crtc_timing) { struct dc_link_settings link_setting; - uint8_t link_bw_set; - uint8_t link_rate_set; + uint8_t link_bw_set = 0; + uint8_t link_rate_set = 0; uint32_t req_bw; union lane_count_set lane_count_set = {0}; @@ -1055,7 +1056,7 @@ bool edp_set_coasting_vtotal(struct dc_link *link, uint32_t coasting_vtotal) } bool edp_replay_residency(const struct dc_link *link, - unsigned int *residency, const bool is_start, const bool is_alpm) + unsigned int *residency, const bool is_start, const enum pr_residency_mode mode) { struct dc *dc = link->ctx->dc; struct dmub_replay *replay = dc->res_pool->replay; @@ -1064,8 +1065,11 @@ bool edp_replay_residency(const struct dc_link *link, if (!dc_get_edp_link_panel_inst(dc, link, &panel_inst)) return false; + if (!residency) + return false; + if (replay != NULL && link->replay_settings.replay_feature_enabled) - replay->funcs->replay_residency(replay, panel_inst, residency, is_start, is_alpm); + replay->funcs->replay_residency(replay, panel_inst, residency, is_start, mode); else *residency = 0; @@ -1145,3 +1149,66 @@ int edp_get_target_backlight_pwm(const struct dc_link *link) return (int) abm->funcs->get_target_backlight(abm); } + +static void edp_set_assr_enable(const struct dc *pDC, struct dc_link *link, + struct link_resource *link_res, bool enable) +{ + union dmub_rb_cmd cmd; + bool use_hpo_dp_link_enc = false; + uint8_t link_enc_index = 0; + uint8_t phy_type = 0; + uint8_t phy_id = 0; + + if (!pDC->config.use_assr_psp_message) + return; + + memset(&cmd, 0, sizeof(cmd)); + + link_enc_index = link->link_enc->transmitter - TRANSMITTER_UNIPHY_A; + + if (link_res->hpo_dp_link_enc) { + link_enc_index = link_res->hpo_dp_link_enc->inst; + use_hpo_dp_link_enc = true; + } + + if (enable) + phy_type = ((dp_get_panel_mode(link) == DP_PANEL_MODE_EDP) ? 1 : 0); + + phy_id = resource_transmitter_to_phy_idx(pDC, link->link_enc->transmitter); + + cmd.assr_enable.header.type = DMUB_CMD__PSP; + cmd.assr_enable.header.sub_type = DMUB_CMD__PSP_ASSR_ENABLE; + cmd.assr_enable.assr_data.enable = enable; + cmd.assr_enable.assr_data.phy_port_type = phy_type; + cmd.assr_enable.assr_data.phy_port_id = phy_id; + cmd.assr_enable.assr_data.link_enc_index = link_enc_index; + cmd.assr_enable.assr_data.hpo_mode = use_hpo_dp_link_enc; + + dc_wake_and_execute_dmub_cmd(pDC->ctx, &cmd, DM_DMUB_WAIT_TYPE_WAIT); +} + +void edp_set_panel_assr(struct dc_link *link, struct pipe_ctx *pipe_ctx, + enum dp_panel_mode *panel_mode, bool enable) +{ + struct link_resource *link_res = &pipe_ctx->link_res; + struct cp_psp *cp_psp = &pipe_ctx->stream->ctx->cp_psp; + + if (*panel_mode != DP_PANEL_MODE_EDP) + return; + + if (link->dc->config.use_assr_psp_message) { + edp_set_assr_enable(link->dc, link, link_res, enable); + } else if (cp_psp && cp_psp->funcs.enable_assr && enable) { + /* ASSR is bound to fail with unsigned PSP + * verstage used during devlopment phase. + * Report and continue with eDP panel mode to + * perform eDP link training with right settings + */ + bool result; + + result = cp_psp->funcs.enable_assr(cp_psp->handle, link); + + if (!result && link->panel_mode != DP_PANEL_MODE_EDP) + *panel_mode = DP_PANEL_MODE_DEFAULT; + } +} diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.h b/drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.h index a158c6234d42..cb6d95cc36e4 100644 --- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.h +++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.h @@ -61,7 +61,7 @@ bool edp_send_replay_cmd(struct dc_link *link, union dmub_replay_cmd_set *cmd_data); bool edp_set_coasting_vtotal(struct dc_link *link, uint32_t coasting_vtotal); bool edp_replay_residency(const struct dc_link *link, - unsigned int *residency, const bool is_start, const bool is_alpm); + unsigned int *residency, const bool is_start, const enum pr_residency_mode mode); bool edp_get_replay_state(const struct dc_link *link, uint64_t *state); bool edp_set_replay_power_opt_and_coasting_vtotal(struct dc_link *link, const unsigned int *power_opts, uint32_t coasting_vtotal); @@ -76,4 +76,6 @@ bool edp_receiver_ready_T9(struct dc_link *link); bool edp_receiver_ready_T7(struct dc_link *link); bool edp_power_alpm_dpcd_enable(struct dc_link *link, bool enable); void edp_set_panel_power(struct dc_link *link, bool powerOn); +void edp_set_panel_assr(struct dc_link *link, struct pipe_ctx *pipe_ctx, + enum dp_panel_mode *panel_mode, bool enable); #endif /* __DC_LINK_EDP_POWER_CONTROL_H__ */ diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_hpd.c b/drivers/gpu/drm/amd/display/dc/link/protocols/link_hpd.c index e3d729ab5b9f..caa617883f62 100644 --- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_hpd.c +++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_hpd.c @@ -35,7 +35,7 @@ bool link_get_hpd_state(struct dc_link *link) { - uint32_t state; + uint32_t state = 0; dal_gpio_lock_pin(link->hpd_gpio); dal_gpio_get_value(link->hpd_gpio, &state); diff --git a/drivers/gpu/drm/amd/display/dc/optc/dcn10/dcn10_optc.c b/drivers/gpu/drm/amd/display/dc/optc/dcn10/dcn10_optc.c index 0e8f4f36c87c..5574bc628053 100644 --- a/drivers/gpu/drm/amd/display/dc/optc/dcn10/dcn10_optc.c +++ b/drivers/gpu/drm/amd/display/dc/optc/dcn10/dcn10_optc.c @@ -945,10 +945,19 @@ void optc1_set_drr( OTG_FORCE_LOCK_ON_EVENT, 0, OTG_SET_V_TOTAL_MIN_MASK_EN, 0, OTG_SET_V_TOTAL_MIN_MASK, 0); - } - // Setup manual flow control for EOF via TRIG_A - optc->funcs->setup_manual_trigger(optc); + // Setup manual flow control for EOF via TRIG_A + optc->funcs->setup_manual_trigger(optc); + + } else { + REG_UPDATE_4(OTG_V_TOTAL_CONTROL, + OTG_SET_V_TOTAL_MIN_MASK, 0, + OTG_V_TOTAL_MIN_SEL, 0, + OTG_V_TOTAL_MAX_SEL, 0, + OTG_FORCE_LOCK_ON_EVENT, 0); + + optc->funcs->set_vtotal_min_max(optc, 0, 0); + } } void optc1_set_vtotal_min_max(struct timing_generator *optc, int vtotal_min, int vtotal_max) @@ -1383,6 +1392,9 @@ void optc1_read_otg_state(struct optc *optc1, REG_GET(OTG_VERTICAL_INTERRUPT2_POSITION, OTG_VERTICAL_INTERRUPT2_LINE_START, &s->vertical_interrupt2_line); + + s->otg_master_update_lock = REG_READ(OTG_MASTER_UPDATE_LOCK); + s->otg_double_buffer_control = REG_READ(OTG_DOUBLE_BUFFER_CONTROL); } bool optc1_get_otg_active_size(struct timing_generator *optc, diff --git a/drivers/gpu/drm/amd/display/dc/optc/dcn10/dcn10_optc.h b/drivers/gpu/drm/amd/display/dc/optc/dcn10/dcn10_optc.h index 6c2e84d3967f..2f3bd7648ba7 100644 --- a/drivers/gpu/drm/amd/display/dc/optc/dcn10/dcn10_optc.h +++ b/drivers/gpu/drm/amd/display/dc/optc/dcn10/dcn10_optc.h @@ -129,6 +129,8 @@ struct dcn_optc_registers { uint32_t OTG_V_TOTAL_MID; uint32_t OTG_V_TOTAL_MIN; uint32_t OTG_V_TOTAL_CONTROL; + uint32_t OTG_V_COUNT_STOP_CONTROL; + uint32_t OTG_V_COUNT_STOP_CONTROL2; uint32_t OTG_TRIGA_CNTL; uint32_t OTG_TRIGA_MANUAL_TRIG; uint32_t OTG_MANUAL_FLOW_CONTROL; @@ -515,12 +517,15 @@ struct dcn_optc_registers { type MANUAL_FLOW_CONTROL;\ type MANUAL_FLOW_CONTROL_SEL; +#define V_TOTAL_REGS(type) + #define TG_REG_FIELD_LIST(type) \ TG_REG_FIELD_LIST_DCN1_0(type)\ type OTG_V_SYNC_MODE;\ type OTG_DRR_TRIGGER_WINDOW_START_X;\ type OTG_DRR_TRIGGER_WINDOW_END_X;\ type OTG_DRR_V_TOTAL_CHANGE_LIMIT;\ + V_TOTAL_REGS(type)\ type OTG_OUT_MUX;\ type OTG_M_CONST_DTO_PHASE;\ type OTG_M_CONST_DTO_MODULO;\ @@ -581,7 +586,9 @@ struct dcn_optc_registers { type OTG_CRC1_WINDOWB_X_END_READBACK;\ type OTG_CRC1_WINDOWB_Y_START_READBACK;\ type OTG_CRC1_WINDOWB_Y_END_READBACK;\ - type OPTC_FGCG_REP_DIS; + type OPTC_FGCG_REP_DIS;\ + type OTG_V_COUNT_STOP;\ + type OTG_V_COUNT_STOP_TIMER; struct dcn_optc_shift { TG_REG_FIELD_LIST(uint8_t) diff --git a/drivers/gpu/drm/amd/display/dc/optc/dcn20/dcn20_optc.c b/drivers/gpu/drm/amd/display/dc/optc/dcn20/dcn20_optc.c index 58bdbd859bf9..d6f095b4555d 100644 --- a/drivers/gpu/drm/amd/display/dc/optc/dcn20/dcn20_optc.c +++ b/drivers/gpu/drm/amd/display/dc/optc/dcn20/dcn20_optc.c @@ -462,16 +462,6 @@ void optc2_setup_manual_trigger(struct timing_generator *optc) { struct optc *optc1 = DCN10TG_FROM_TG(optc); - /* Set the min/max selectors unconditionally so that - * DMCUB fw may change OTG timings when necessary - * TODO: Remove the w/a after fixing the issue in DMCUB firmware - */ - REG_UPDATE_4(OTG_V_TOTAL_CONTROL, - OTG_V_TOTAL_MIN_SEL, 1, - OTG_V_TOTAL_MAX_SEL, 1, - OTG_FORCE_LOCK_ON_EVENT, 0, - OTG_SET_V_TOTAL_MIN_MASK, (1 << 1)); /* TRIGA */ - REG_SET_8(OTG_TRIGA_CNTL, 0, OTG_TRIGA_SOURCE_SELECT, 21, OTG_TRIGA_SOURCE_PIPE_SELECT, optc->inst, diff --git a/drivers/gpu/drm/amd/display/dc/optc/dcn32/dcn32_optc.c b/drivers/gpu/drm/amd/display/dc/optc/dcn32/dcn32_optc.c index f07a4c7e48bc..52eab8fccb7f 100644 --- a/drivers/gpu/drm/amd/display/dc/optc/dcn32/dcn32_optc.c +++ b/drivers/gpu/drm/amd/display/dc/optc/dcn32/dcn32_optc.c @@ -267,9 +267,6 @@ static void optc32_setup_manual_trigger(struct timing_generator *optc) OTG_V_TOTAL_MAX_SEL, 1, OTG_FORCE_LOCK_ON_EVENT, 0, OTG_SET_V_TOTAL_MIN_MASK, (1 << 1)); /* TRIGA */ - - // Setup manual flow control for EOF via TRIG_A - optc->funcs->setup_manual_trigger(optc); } } diff --git a/drivers/gpu/drm/amd/display/dc/optc/dcn35/dcn35_optc.c b/drivers/gpu/drm/amd/display/dc/optc/dcn35/dcn35_optc.c index 5b1547508850..d393be30dff8 100644 --- a/drivers/gpu/drm/amd/display/dc/optc/dcn35/dcn35_optc.c +++ b/drivers/gpu/drm/amd/display/dc/optc/dcn35/dcn35_optc.c @@ -32,6 +32,7 @@ #include "reg_helper.h" #include "dc.h" #include "dcn_calc_math.h" +#include "dc_dmub_srv.h" #define REG(reg)\ optc1->tg_regs->reg @@ -213,6 +214,167 @@ static bool optc35_configure_crc(struct timing_generator *optc, return true; } +static void optc35_setup_manual_trigger(struct timing_generator *optc) +{ + if (!optc || !optc->ctx) + return; + + struct optc *optc1 = DCN10TG_FROM_TG(optc); + struct dc *dc = optc->ctx->dc; + + if (dc->caps.dmub_caps.mclk_sw && !dc->debug.disable_fams) + dc_dmub_srv_set_drr_manual_trigger_cmd(dc, optc->inst); + else { + /* + * MIN_MASK_EN is gone and MASK is now always enabled. + * + * To get it to it work with manual trigger we need to make sure + * we program the correct bit. + */ + REG_UPDATE_4(OTG_V_TOTAL_CONTROL, + OTG_V_TOTAL_MIN_SEL, 1, + OTG_V_TOTAL_MAX_SEL, 1, + OTG_FORCE_LOCK_ON_EVENT, 0, + OTG_SET_V_TOTAL_MIN_MASK, (1 << 1)); /* TRIGA */ + + // Setup manual flow control for EOF via TRIG_A + if (optc->funcs && optc->funcs->setup_manual_trigger) + optc->funcs->setup_manual_trigger(optc); + } +} + +void optc35_set_drr( + struct timing_generator *optc, + const struct drr_params *params) +{ + if (!optc || !params) + return; + + struct optc *optc1 = DCN10TG_FROM_TG(optc); + uint32_t max_otg_v_total = optc1->max_v_total - 1; + + if (params != NULL && + params->vertical_total_max > 0 && + params->vertical_total_min > 0) { + + if (params->vertical_total_mid != 0) { + + REG_SET(OTG_V_TOTAL_MID, 0, + OTG_V_TOTAL_MID, params->vertical_total_mid - 1); + + REG_UPDATE_2(OTG_V_TOTAL_CONTROL, + OTG_VTOTAL_MID_REPLACING_MAX_EN, 1, + OTG_VTOTAL_MID_FRAME_NUM, + (uint8_t)params->vertical_total_mid_frame_num); + + } + + if (optc->funcs && optc->funcs->set_vtotal_min_max) + optc->funcs->set_vtotal_min_max(optc, + params->vertical_total_min - 1, params->vertical_total_max - 1); + optc35_setup_manual_trigger(optc); + } else { + REG_UPDATE_4(OTG_V_TOTAL_CONTROL, + OTG_SET_V_TOTAL_MIN_MASK, 0, + OTG_V_TOTAL_MIN_SEL, 0, + OTG_V_TOTAL_MAX_SEL, 0, + OTG_FORCE_LOCK_ON_EVENT, 0); + + if (optc->funcs && optc->funcs->set_vtotal_min_max) + optc->funcs->set_vtotal_min_max(optc, 0, 0); + } + + REG_WRITE(OTG_V_COUNT_STOP_CONTROL, max_otg_v_total); + REG_WRITE(OTG_V_COUNT_STOP_CONTROL2, 0); +} + +static void optc35_set_long_vtotal( + struct timing_generator *optc, + const struct long_vtotal_params *params) +{ + if (!optc || !params) + return; + + struct optc *optc1 = DCN10TG_FROM_TG(optc); + uint32_t vcount_stop_timer = 0, vcount_stop = 0; + uint32_t max_otg_v_total = optc1->max_v_total - 1; + + if (params->vertical_total_min <= max_otg_v_total && params->vertical_total_max <= max_otg_v_total) + return; + + if (params->vertical_total_max == 0 || params->vertical_total_min == 0) { + REG_UPDATE_4(OTG_V_TOTAL_CONTROL, + OTG_SET_V_TOTAL_MIN_MASK, 0, + OTG_V_TOTAL_MIN_SEL, 0, + OTG_V_TOTAL_MAX_SEL, 0, + OTG_FORCE_LOCK_ON_EVENT, 0); + + if (optc->funcs && optc->funcs->set_vtotal_min_max) + optc->funcs->set_vtotal_min_max(optc, 0, 0); + } else if (params->vertical_total_max == params->vertical_total_min) { + vcount_stop = params->vertical_blank_start; + vcount_stop_timer = params->vertical_total_max - max_otg_v_total; + + REG_UPDATE_4(OTG_V_TOTAL_CONTROL, + OTG_V_TOTAL_MIN_SEL, 1, + OTG_V_TOTAL_MAX_SEL, 1, + OTG_FORCE_LOCK_ON_EVENT, 0, + OTG_SET_V_TOTAL_MIN_MASK, 0); + + if (optc->funcs && optc->funcs->set_vtotal_min_max) + optc->funcs->set_vtotal_min_max(optc, max_otg_v_total, max_otg_v_total); + + REG_WRITE(OTG_V_COUNT_STOP_CONTROL, vcount_stop); + REG_WRITE(OTG_V_COUNT_STOP_CONTROL2, vcount_stop_timer); + } else { + // Variable rate, keep DRR trigger mask + if (params->vertical_total_min > max_otg_v_total) { + // cannot be supported + // If MAX_OTG_V_COUNT < DRR trigger < v_total_min < v_total_max, + // DRR trigger will drop the vtotal counting directly to a new frame. + // But it should trigger between v_total_min and v_total_max. + ASSERT(0); + + REG_UPDATE_4(OTG_V_TOTAL_CONTROL, + OTG_SET_V_TOTAL_MIN_MASK, 0, + OTG_V_TOTAL_MIN_SEL, 0, + OTG_V_TOTAL_MAX_SEL, 0, + OTG_FORCE_LOCK_ON_EVENT, 0); + + if (optc->funcs && optc->funcs->set_vtotal_min_max) + optc->funcs->set_vtotal_min_max(optc, 0, 0); + + REG_WRITE(OTG_V_COUNT_STOP_CONTROL, max_otg_v_total); + REG_WRITE(OTG_V_COUNT_STOP_CONTROL2, 0); + } else { + // For total_min <= MAX_OTG_V_COUNT and total_max > MAX_OTG_V_COUNT + vcount_stop = params->vertical_total_min; + vcount_stop_timer = params->vertical_total_max - max_otg_v_total; + + // Example: + // params->vertical_total_min 1000 + // params->vertical_total_max 2000 + // MAX_OTG_V_COUNT_STOP = 1500 + // + // If DRR event not happened, + // time 0,1,2,3,4,...1000,1001,........,1500,1501,1502, ...1999 + // vcount 0,1,2,3,4....1000...................,1001,1002,1003,...1399 + // vcount2 0,1,2,3,4,..499, + // else (DRR event happened, ex : at line 1004) + // time 0,1,2,3,4,...1000,1001.....1004, 0 + // vcount 0,1,2,3,4....1000,.............. 0 (new frame) + // vcount2 0,1,2, 3, - + if (optc->funcs && optc->funcs->set_vtotal_min_max) + optc->funcs->set_vtotal_min_max(optc, + params->vertical_total_min - 1, max_otg_v_total); + optc35_setup_manual_trigger(optc); + + REG_WRITE(OTG_V_COUNT_STOP_CONTROL, vcount_stop); + REG_WRITE(OTG_V_COUNT_STOP_CONTROL2, vcount_stop_timer); + } + } +} + static struct timing_generator_funcs dcn35_tg_funcs = { .validate_timing = optc1_validate_timing, .program_timing = optc1_program_timing, @@ -245,7 +407,7 @@ static struct timing_generator_funcs dcn35_tg_funcs = { .lock_doublebuffer_enable = optc3_lock_doublebuffer_enable, .lock_doublebuffer_disable = optc3_lock_doublebuffer_disable, .enable_optc_clock = optc1_enable_optc_clock, - .set_drr = optc31_set_drr, + .set_drr = optc35_set_drr, .get_last_used_drr_vtotal = optc2_get_last_used_drr_vtotal, .set_vtotal_min_max = optc1_set_vtotal_min_max, .set_static_screen_control = optc1_set_static_screen_control, @@ -275,6 +437,7 @@ static struct timing_generator_funcs dcn35_tg_funcs = { .setup_manual_trigger = optc2_setup_manual_trigger, .get_hw_timing = optc1_get_hw_timing, .init_odm = optc3_init_odm, + .set_long_vtotal = optc35_set_long_vtotal, }; void dcn35_timing_generator_init(struct optc *optc1) diff --git a/drivers/gpu/drm/amd/display/dc/optc/dcn35/dcn35_optc.h b/drivers/gpu/drm/amd/display/dc/optc/dcn35/dcn35_optc.h index 1f422e4c468f..d077e2392379 100644 --- a/drivers/gpu/drm/amd/display/dc/optc/dcn35/dcn35_optc.h +++ b/drivers/gpu/drm/amd/display/dc/optc/dcn35/dcn35_optc.h @@ -65,10 +65,14 @@ SF(OTG0_OTG_CRC1_WINDOWB_X_CONTROL_READBACK, OTG_CRC1_WINDOWB_X_END_READBACK, mask_sh),\ SF(OTG0_OTG_CRC1_WINDOWB_Y_CONTROL_READBACK, OTG_CRC1_WINDOWB_Y_START_READBACK, mask_sh),\ SF(OTG0_OTG_CRC1_WINDOWB_Y_CONTROL_READBACK, OTG_CRC1_WINDOWB_Y_END_READBACK, mask_sh),\ - SF(OPTC_CLOCK_CONTROL, OPTC_FGCG_REP_DIS, mask_sh) + SF(OPTC_CLOCK_CONTROL, OPTC_FGCG_REP_DIS, mask_sh),\ + SF(OTG0_OTG_V_COUNT_STOP_CONTROL, OTG_V_COUNT_STOP, mask_sh),\ + SF(OTG0_OTG_V_COUNT_STOP_CONTROL2, OTG_V_COUNT_STOP_TIMER, mask_sh) void dcn35_timing_generator_init(struct optc *optc1); void dcn35_timing_generator_set_fgcg(struct optc *optc1, bool enable); +void optc35_set_drr(struct timing_generator *optc, const struct drr_params *params); + #endif /* __DC_OPTC_DCN35_H__ */ diff --git a/drivers/gpu/drm/amd/display/dc/resource/Makefile b/drivers/gpu/drm/amd/display/dc/resource/Makefile index 184b1f23aa77..db9048974d74 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/Makefile +++ b/drivers/gpu/drm/amd/display/dc/resource/Makefile @@ -102,10 +102,6 @@ AMD_DISPLAY_FILES += $(AMD_DAL_RESOURCE_DCN21) ############################################################################### -############################################################################### - -############################################################################### - RESOURCE_DCN30 = dcn30_resource.o AMD_DAL_RESOURCE_DCN30 = $(addprefix $(AMDDALPATH)/dc/resource/dcn30/,$(RESOURCE_DCN30)) @@ -202,6 +198,4 @@ AMD_DISPLAY_FILES += $(AMD_DAL_RESOURCE_DCN351) ############################################################################### -############################################################################### - endif diff --git a/drivers/gpu/drm/amd/display/dc/resource/dce112/dce112_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dce112/dce112_resource.c index d1edac46c9a0..88afb2a30eef 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dce112/dce112_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dce112/dce112_resource.c @@ -864,8 +864,6 @@ static struct clock_source *find_matching_pll( default: return NULL; } - - return NULL; } static enum dc_status build_mapped_resource( diff --git a/drivers/gpu/drm/amd/display/dc/resource/dce120/dce120_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dce120/dce120_resource.c index 20662edd0ae4..621825a51f46 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dce120/dce120_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dce120/dce120_resource.c @@ -1060,7 +1060,7 @@ static bool dce120_resource_construct( struct irq_service_init_data irq_init_data; static const struct resource_create_funcs *res_funcs; bool is_vg20 = ASICREV_IS_VEGA20_P(ctx->asic_id.hw_internal_rev); - uint32_t pipe_fuses; + uint32_t pipe_fuses = 0; ctx->dc_bios->regs = &bios_regs; diff --git a/drivers/gpu/drm/amd/display/dc/resource/dce80/dce80_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dce80/dce80_resource.c index 35a2cce0c2b8..56ee45e12b46 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dce80/dce80_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dce80/dce80_resource.c @@ -56,7 +56,6 @@ #include "dce/dce_aux.h" #include "dce/dce_abm.h" #include "dce/dce_i2c.h" -/* TODO remove this include */ #ifndef mmMC_HUB_RDREQ_DMIF_LIMIT #include "gmc/gmc_7_1_d.h" diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn10/dcn10_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn10/dcn10_resource.c index d08d10969251..563c5eec83ff 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn10/dcn10_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn10/dcn10_resource.c @@ -513,7 +513,7 @@ static const struct dc_plane_cap plane_cap = { .argb8888 = true, .nv12 = true, .fp16 = true, - .p010 = true + .p010 = false }, .max_upscale_factor = { @@ -569,6 +569,7 @@ static const struct dc_debug_options debug_defaults_diags = { .disable_pplib_clock_request = true, .disable_pplib_wm_range = true, .underflow_assert_delay_us = 0xFFFFFFFF, + .enable_legacy_fast_update = true, }; static void dcn10_dpp_destroy(struct dpp **dpp) @@ -1631,6 +1632,7 @@ static bool dcn10_resource_construct( /* valid pipe num */ pool->base.pipe_count = j; pool->base.timing_generator_count = j; + pool->base.mpcc_count = j; /* within dml lib, it is hard code to 4. If ASIC pipe is fused, * the value may be changed diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn20/dcn20_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn20/dcn20_resource.c index a2387cea1af9..0a939437e19f 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn20/dcn20_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn20/dcn20_resource.c @@ -62,6 +62,9 @@ #include "dcn20/dcn20_vmid.h" #include "dce/dce_panel_cntl.h" +#include "dcn20/dcn20_dwb.h" +#include "dcn20/dcn20_mmhubbub.h" + #include "navi10_ip_offset.h" #include "dcn/dcn_2_0_0_offset.h" @@ -71,9 +74,6 @@ #include "nbio/nbio_2_3_offset.h" -#include "dcn20/dcn20_dwb.h" -#include "dcn20/dcn20_mmhubbub.h" - #include "mmhub/mmhub_2_0_0_offset.h" #include "mmhub/mmhub_2_0_0_sh_mask.h" @@ -83,11 +83,10 @@ #include "dce/dce_aux.h" #include "dce/dce_i2c.h" #include "vm_helper.h" -#include "link_enc_cfg.h" - -#include "amdgpu_socbb.h" +#include "link_enc_cfg.h" #include "link.h" + #define DC_LOGGER_INIT(logger) #ifndef mmDP0_DP_DPHY_INTERNAL_CTRL @@ -1282,8 +1281,13 @@ void dcn20_build_pipe_pix_clk_params(struct pipe_ctx *pipe_ctx) static enum dc_status build_pipe_hw_param(struct pipe_ctx *pipe_ctx) { + struct resource_pool *pool = pipe_ctx->stream->ctx->dc->res_pool; - dcn20_build_pipe_pix_clk_params(pipe_ctx); + if (pool->funcs->build_pipe_pix_clk_params) { + pool->funcs->build_pipe_pix_clk_params(pipe_ctx); + } else { + dcn20_build_pipe_pix_clk_params(pipe_ctx); + } pipe_ctx->stream->clamping.pixel_encoding = pipe_ctx->stream->timing.pixel_encoding; @@ -2449,6 +2453,7 @@ static bool dcn20_resource_construct( dc->caps.post_blend_color_processing = true; dc->caps.force_dp_tps4_for_cp2520 = true; dc->caps.extended_aux_timeout_support = true; + dc->caps.dmcub_support = true; /* Color pipeline capabilities */ dc->caps.color.dpp.dcn_arch = 1; diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn201/dcn201_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn201/dcn201_resource.c index 914b234d7f6b..070a4efb308b 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn201/dcn201_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn201/dcn201_resource.c @@ -55,7 +55,6 @@ #include "dce110/dce110_resource.h" #include "dce/dce_aux.h" #include "dce/dce_i2c.h" -#include "dcn201/dcn201_hubbub.h" #include "dcn10/dcn10_resource.h" #include "cyan_skillfish_ip_offset.h" @@ -182,6 +181,7 @@ static struct _vcs_dpi_soc_bounding_box_st dcn201_soc = { .socclk_mhz = 1254.0, .dram_speed_mts = 14000.0, }, + /* state4 is not an actual state, just defines unsupported for dml*/ { .state = 4, .dscclk_mhz = 400.0, @@ -566,6 +566,8 @@ static const struct resource_caps res_cap_dnc201 = { .num_audio = 2, .num_stream_encoder = 2, .num_pll = 2, + .num_dwb = 0, + .num_dsc = 0, .num_ddc = 2, }; @@ -612,7 +614,7 @@ static const struct dc_debug_options debug_defaults_drv = { .scl_reset_length10 = true, .sanity_checks = false, .underflow_assert_delay_us = 0xFFFFFFFF, - .enable_tri_buf = false, + .enable_tri_buf = true, .enable_legacy_fast_update = true, .using_dml2 = false, }; diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn21/dcn21_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn21/dcn21_resource.c index 65d337731f56..8663cbc3d1cf 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn21/dcn21_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn21/dcn21_resource.c @@ -581,32 +581,6 @@ static const struct resource_caps res_cap_rn = { .num_dsc = 3, }; -#ifdef DIAGS_BUILD -static const struct resource_caps res_cap_rn_FPGA_4pipe = { - .num_timing_generator = 4, - .num_opp = 4, - .num_video_plane = 4, - .num_audio = 7, - .num_stream_encoder = 4, - .num_pll = 4, - .num_dwb = 1, - .num_ddc = 4, - .num_dsc = 0, -}; - -static const struct resource_caps res_cap_rn_FPGA_2pipe_dsc = { - .num_timing_generator = 2, - .num_opp = 2, - .num_video_plane = 2, - .num_audio = 7, - .num_stream_encoder = 2, - .num_pll = 4, - .num_dwb = 1, - .num_ddc = 4, - .num_dsc = 2, -}; -#endif - static const struct dc_plane_cap plane_cap = { .type = DC_PLANE_TYPE_DCN_UNIVERSAL, .per_pixel_alpha = true, @@ -1415,16 +1389,11 @@ static bool dcn21_resource_construct( struct dc_context *ctx = dc->ctx; struct irq_service_init_data init_data; uint32_t pipe_fuses = read_pipe_fuses(ctx); - uint32_t num_pipes; + uint32_t num_pipes = 0; ctx->dc_bios->regs = &bios_regs; pool->base.res_cap = &res_cap_rn; -#ifdef DIAGS_BUILD - if (IS_FPGA_MAXIMUS_DC(dc->ctx->dce_environment)) - //pool->base.res_cap = &res_cap_nv10_FPGA_2pipe_dsc; - pool->base.res_cap = &res_cap_rn_FPGA_4pipe; -#endif pool->base.funcs = &dcn21_res_pool_funcs; diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn30/dcn30_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn30/dcn30_resource.c index ecc477ef8e3b..f35cc307830b 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn30/dcn30_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn30/dcn30_resource.c @@ -1639,7 +1639,7 @@ noinline bool dcn30_internal_validate_bw( int split[MAX_PIPES] = { 0 }; bool merge[MAX_PIPES] = { false }; bool newly_split[MAX_PIPES] = { false }; - int pipe_cnt, i, pipe_idx, vlevel; + int pipe_cnt, i, pipe_idx, vlevel = 0; struct vba_vars_st *vba = &context->bw_ctx.dml.vba; ASSERT(pipes); @@ -2050,6 +2050,9 @@ bool dcn30_validate_bandwidth(struct dc *dc, BW_VAL_TRACE_COUNT(); + if (!pipes) + goto validate_fail; + DC_FP_START(); out = dcn30_internal_validate_bw(dc, context, pipes, &pipe_cnt, &vlevel, fast_validate, true); DC_FP_END(); diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn303/dcn303_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn303/dcn303_resource.c index 25cd6236b054..8bc1bcaeaa47 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn303/dcn303_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn303/dcn303_resource.c @@ -1143,7 +1143,7 @@ static bool dcn303_resource_construct( int i; struct dc_context *ctx = dc->ctx; struct irq_service_init_data init_data; - struct ddc_service_init_data ddc_init_data; + struct ddc_service_init_data ddc_init_data = {0}; ctx->dc_bios->regs = &bios_regs; diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn31/dcn31_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn31/dcn31_resource.c index 04d142f97474..d4c3e2754f51 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn31/dcn31_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn31/dcn31_resource.c @@ -75,7 +75,6 @@ #include "dcn30/dcn30_dwb.h" #include "dcn30/dcn30_mmhubbub.h" -// TODO: change include headers /amd/include/asic_reg after upstream #include "yellow_carp_offset.h" #include "dcn/dcn_3_1_2_offset.h" #include "dcn/dcn_3_1_2_sh_mask.h" @@ -892,7 +891,7 @@ static const struct dc_debug_options debug_defaults_drv = { .disable_z10 = true, .enable_legacy_fast_update = true, .enable_z9_disable_interface = true, /* Allow support for the PMFW interface for disable Z9*/ - .dml_hostvm_override = DML_HOSTVM_OVERRIDE_FALSE, + .dml_hostvm_override = DML_HOSTVM_NO_OVERRIDE, .using_dml2 = false, }; @@ -1311,6 +1310,8 @@ static struct hpo_dp_link_encoder *dcn31_hpo_dp_link_encoder_create( /* allocate HPO link encoder */ hpo_dp_enc31 = kzalloc(sizeof(struct dcn31_hpo_dp_link_encoder), GFP_KERNEL); + if (!hpo_dp_enc31) + return NULL; /* out of memory */ hpo_dp_link_encoder31_construct(hpo_dp_enc31, ctx, inst, &hpo_dp_link_enc_regs[inst], @@ -1645,7 +1646,7 @@ int dcn31_populate_dml_pipes_from_context( { int i, pipe_cnt; struct resource_context *res_ctx = &context->res_ctx; - struct pipe_ctx *pipe; + struct pipe_ctx *pipe = 0; bool upscaled = false; DC_FP_START(); @@ -1767,11 +1768,14 @@ bool dcn31_validate_bandwidth(struct dc *dc, BW_VAL_TRACE_COUNT(); + if (!pipes) + goto validate_fail; + DC_FP_START(); out = dcn30_internal_validate_bw(dc, context, pipes, &pipe_cnt, &vlevel, fast_validate, true); DC_FP_END(); - // Disable fast_validate to set min dcfclk in alculate_wm_and_dlg + // Disable fast_validate to set min dcfclk in calculate_wm_and_dlg if (pipe_cnt == 0) fast_validate = false; diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn314/dcn314_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn314/dcn314_resource.c index c97391edb5ff..ff50f43e4c00 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn314/dcn314_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn314/dcn314_resource.c @@ -925,27 +925,10 @@ static const struct dc_debug_options debug_defaults_drv = { }, .seamless_boot_odm_combine = true, + .enable_legacy_fast_update = true, .using_dml2 = false, }; -static const struct dc_debug_options debug_defaults_diags = { - .disable_dmcu = true, - .force_abm_enable = false, - .timing_trace = true, - .clock_trace = true, - .disable_dpp_power_gate = true, - .disable_hubp_power_gate = true, - .disable_clock_gate = true, - .disable_pplib_clock_request = true, - .disable_pplib_wm_range = true, - .disable_stutter = false, - .scl_reset_length10 = true, - .dwb_fi_phase = -1, // -1 = disable - .dmub_command_table = true, - .enable_tri_buf = true, - .use_max_lb = true -}; - static const struct dc_panel_config panel_config_defaults = { .psr = { .disable_psr = false, @@ -1384,6 +1367,8 @@ static struct hpo_dp_link_encoder *dcn31_hpo_dp_link_encoder_create( /* allocate HPO link encoder */ hpo_dp_enc31 = kzalloc(sizeof(struct dcn31_hpo_dp_link_encoder), GFP_KERNEL); + if (!hpo_dp_enc31) + return NULL; /* out of memory */ hpo_dp_link_encoder31_construct(hpo_dp_enc31, ctx, inst, &hpo_dp_link_enc_regs[inst], @@ -1744,6 +1729,9 @@ bool dcn314_validate_bandwidth(struct dc *dc, BW_VAL_TRACE_COUNT(); + if (!pipes) + goto validate_fail; + if (filter_modes_for_single_channel_workaround(dc, context)) goto validate_fail; @@ -1938,8 +1926,6 @@ static bool dcn314_resource_construct( if (dc->ctx->dce_environment == DCE_ENV_PRODUCTION_DRV) dc->debug = debug_defaults_drv; - else - dc->debug = debug_defaults_diags; /* Disable pipe power gating */ dc->debug.disable_dpp_power_gate = true; diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn315/dcn315_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn315/dcn315_resource.c index 515ba435f759..4ce0f4bf1d9b 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn315/dcn315_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn315/dcn315_resource.c @@ -1309,6 +1309,8 @@ static struct hpo_dp_link_encoder *dcn31_hpo_dp_link_encoder_create( /* allocate HPO link encoder */ hpo_dp_enc31 = kzalloc(sizeof(struct dcn31_hpo_dp_link_encoder), GFP_KERNEL); + if (!hpo_dp_enc31) + return NULL; /* out of memory */ hpo_dp_link_encoder31_construct(hpo_dp_enc31, ctx, inst, &hpo_dp_link_enc_regs[inst], diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn316/dcn316_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn316/dcn316_resource.c index b9753d4606f8..5fd52c5fcee4 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn316/dcn316_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn316/dcn316_resource.c @@ -125,7 +125,6 @@ #include "link_enc_cfg.h" #define DCN3_16_MAX_DET_SIZE 384 -#define DCN3_16_MIN_COMPBUF_SIZE_KB 128 #define DCN3_16_CRB_SEGMENT_SIZE_KB 64 enum dcn31_clk_src_array_id { @@ -1306,6 +1305,8 @@ static struct hpo_dp_link_encoder *dcn31_hpo_dp_link_encoder_create( /* allocate HPO link encoder */ hpo_dp_enc31 = kzalloc(sizeof(struct dcn31_hpo_dp_link_encoder), GFP_KERNEL); + if (!hpo_dp_enc31) + return NULL; /* out of memory */ hpo_dp_link_encoder31_construct(hpo_dp_enc31, ctx, inst, &hpo_dp_link_enc_regs[inst], @@ -1614,7 +1615,7 @@ static int dcn316_populate_dml_pipes_from_context( { int i, pipe_cnt; struct resource_context *res_ctx = &context->res_ctx; - struct pipe_ctx *pipe; + struct pipe_ctx *pipe = 0; const int max_usable_det = context->bw_ctx.dml.ip.config_return_buffer_size_in_kbytes - DCN3_16_MIN_COMPBUF_SIZE_KB; DC_FP_START(); diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn32/dcn32_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn32/dcn32_resource.c index ce1754cc1f46..abd76345d1e4 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn32/dcn32_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn32/dcn32_resource.c @@ -1304,6 +1304,8 @@ static struct hpo_dp_link_encoder *dcn32_hpo_dp_link_encoder_create( /* allocate HPO link encoder */ hpo_dp_enc31 = kzalloc(sizeof(struct dcn31_hpo_dp_link_encoder), GFP_KERNEL); + if (!hpo_dp_enc31) + return NULL; /* out of memory */ #undef REG_STRUCT #define REG_STRUCT hpo_dp_link_enc_regs @@ -1751,6 +1753,9 @@ static bool dml1_validate(struct dc *dc, struct dc_state *context, bool fast_val BW_VAL_TRACE_COUNT(); + if (!pipes) + goto validate_fail; + DC_FP_START(); out = dcn32_internal_validate_bw(dc, context, pipes, &pipe_cnt, &vlevel, fast_validate); DC_FP_END(); @@ -1799,7 +1804,9 @@ bool dcn32_validate_bandwidth(struct dc *dc, bool out = false; if (dc->debug.using_dml2) - out = dml2_validate(dc, context, fast_validate); + out = dml2_validate(dc, context, + context->power_source == DC_POWER_SOURCE_DC ? context->bw_ctx.dml2_dc_power_source : context->bw_ctx.dml2, + fast_validate); else out = dml1_validate(dc, context, fast_validate); return out; @@ -1815,9 +1822,48 @@ int dcn32_populate_dml_pipes_from_context( struct pipe_ctx *pipe = NULL; bool subvp_in_use = false; struct dc_crtc_timing *timing; + int subvp_main_pipe_index = -1; + enum mall_stream_type mall_type; + bool single_display_subvp = false; + struct dc_stream_state *stream = NULL; + int num_subvp_main = 0; + int num_subvp_phantom = 0; + int num_subvp_none = 0; + int odm_slice_count; dcn20_populate_dml_pipes_from_context(dc, context, pipes, fast_validate); + /* For single display subvp, look for subvp main so if we have phantom + * pipe, we can set odm policy to match main pipe + */ + for (i = 0; i < context->stream_count; i++) { + stream = context->streams[i]; + mall_type = dc_state_get_stream_subvp_type(context, stream); + if (mall_type == SUBVP_MAIN) + num_subvp_main++; + else if (mall_type == SUBVP_PHANTOM) + num_subvp_phantom++; + else + num_subvp_none++; + } + if (num_subvp_main == 1 && num_subvp_phantom == 1 && num_subvp_none == 0) + single_display_subvp = true; + + if (single_display_subvp) { + for (i = 0, pipe_cnt = 0; i < dc->res_pool->pipe_count; i++) { + pipe = &res_ctx->pipe_ctx[i]; + if (!res_ctx->pipe_ctx[i].stream) + continue; + + mall_type = dc_state_get_pipe_subvp_type(context, pipe); + if (mall_type == SUBVP_MAIN) { + if (resource_is_pipe_type(pipe, OTG_MASTER)) + subvp_main_pipe_index = i; + } + pipe_cnt++; + } + } + for (i = 0, pipe_cnt = 0; i < dc->res_pool->pipe_count; i++) { if (!res_ctx->pipe_ctx[i].stream) @@ -1832,7 +1878,21 @@ int dcn32_populate_dml_pipes_from_context( pipes[pipe_cnt].pipe.dest.vfront_porch = timing->v_front_porch; if (dc->config.enable_windowed_mpo_odm && dc->debug.enable_single_display_2to1_odm_policy) { - switch (resource_get_odm_slice_count(pipe)) { + /* For single display subvp, if pipe is phantom pipe, + * then copy odm policy from subvp main pipe + */ + mall_type = dc_state_get_pipe_subvp_type(context, pipe); + if (single_display_subvp && (mall_type == SUBVP_PHANTOM)) { + if (subvp_main_pipe_index < 0) { + odm_slice_count = -1; + ASSERT(0); + } else { + odm_slice_count = resource_get_odm_slice_count(&res_ctx->pipe_ctx[subvp_main_pipe_index]); + } + } else { + odm_slice_count = resource_get_odm_slice_count(pipe); + } + switch (odm_slice_count) { case 2: pipes[pipe_cnt].pipe.dest.odm_combine_policy = dm_odm_combine_policy_2to1; break; @@ -1845,6 +1905,7 @@ int dcn32_populate_dml_pipes_from_context( } else { pipes[pipe_cnt].pipe.dest.odm_combine_policy = dm_odm_combine_policy_dal; } + pipes[pipe_cnt].pipe.src.gpuvm_min_page_size_kbytes = 256; // according to spreadsheet pipes[pipe_cnt].pipe.src.unbounded_req_mode = false; pipes[pipe_cnt].pipe.scale_ratio_depth.lb_depth = dm_lb_19; @@ -1912,6 +1973,22 @@ int dcn32_populate_dml_pipes_from_context( return pipe_cnt; } +unsigned int dcn32_calculate_mall_ways_from_bytes(const struct dc *dc, unsigned int total_size_in_mall_bytes) +{ + uint32_t cache_lines_used, lines_per_way, total_cache_lines, num_ways; + + /* add 2 lines for worst case alignment */ + cache_lines_used = total_size_in_mall_bytes / dc->caps.cache_line_size + 2; + + total_cache_lines = dc->caps.max_cab_allocation_bytes / dc->caps.cache_line_size; + lines_per_way = total_cache_lines / dc->caps.cache_num_ways; + num_ways = cache_lines_used / lines_per_way; + if (cache_lines_used % lines_per_way > 0) + num_ways++; + + return num_ways; +} + static struct dc_cap_funcs cap_funcs = { .get_dcc_compression_cap = dcn20_get_dcc_compression_cap, .get_subvp_en = dcn32_subvp_in_use, @@ -1929,10 +2006,20 @@ void dcn32_calculate_wm_and_dlg(struct dc *dc, struct dc_state *context, static void dcn32_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_params) { + struct dml2_configuration_options dml2_opt = dc->dml2_options; + DC_FP_START(); + dcn32_update_bw_bounding_box_fpu(dc, bw_params); + + dml2_opt.use_clock_dc_limits = false; if (dc->debug.using_dml2 && dc->current_state && dc->current_state->bw_ctx.dml2) - dml2_reinit(dc, &dc->dml2_options, &dc->current_state->bw_ctx.dml2); + dml2_reinit(dc, &dml2_opt, &dc->current_state->bw_ctx.dml2); + + dml2_opt.use_clock_dc_limits = true; + if (dc->debug.using_dml2 && dc->current_state && dc->current_state->bw_ctx.dml2_dc_power_source) + dml2_reinit(dc, &dml2_opt, &dc->current_state->bw_ctx.dml2_dc_power_source); + DC_FP_END(); } @@ -1960,6 +2047,7 @@ static struct resource_funcs dcn32_res_pool_funcs = { .update_soc_for_wm_a = dcn30_update_soc_for_wm_a, .add_phantom_pipes = dcn32_add_phantom_pipes, .build_pipe_pix_clk_params = dcn20_build_pipe_pix_clk_params, + .calculate_mall_ways_from_bytes = dcn32_calculate_mall_ways_from_bytes, }; static uint32_t read_pipe_fuses(struct dc_context *ctx) @@ -2048,7 +2136,8 @@ static bool dcn32_resource_construct( dc->caps.min_horizontal_blanking_period = 80; dc->caps.dmdata_alloc_size = 2048; dc->caps.mall_size_per_mem_channel = 4; - dc->caps.mall_size_total = 0; + /* total size = mall per channel * num channels * 1024 * 1024 */ + dc->caps.mall_size_total = dc->caps.mall_size_per_mem_channel * dc->ctx->dc_bios->vram_info.num_chans * 1048576; dc->caps.cursor_cache_size = dc->caps.max_cursor_size * dc->caps.max_cursor_size * 8; dc->caps.cache_line_size = 64; @@ -2362,30 +2451,10 @@ static bool dcn32_resource_construct( dc->dml2_options.use_native_soc_bb_construction = true; dc->dml2_options.minimize_dispclk_using_odm = true; - dc->dml2_options.callbacks.dc = dc; - dc->dml2_options.callbacks.build_scaling_params = &resource_build_scaling_params; + resource_init_common_dml2_callbacks(dc, &dc->dml2_options); dc->dml2_options.callbacks.can_support_mclk_switch_using_fw_based_vblank_stretch = &dcn30_can_support_mclk_switch_using_fw_based_vblank_stretch; - dc->dml2_options.callbacks.acquire_secondary_pipe_for_mpc_odm = &dc_resource_acquire_secondary_pipe_for_mpc_odm_legacy; - dc->dml2_options.callbacks.update_pipes_for_stream_with_slice_count = &resource_update_pipes_for_stream_with_slice_count; - dc->dml2_options.callbacks.update_pipes_for_plane_with_slice_count = &resource_update_pipes_for_plane_with_slice_count; - dc->dml2_options.callbacks.get_mpc_slice_index = &resource_get_mpc_slice_index; - dc->dml2_options.callbacks.get_odm_slice_index = &resource_get_odm_slice_index; - dc->dml2_options.callbacks.get_opp_head = &resource_get_opp_head; - - dc->dml2_options.svp_pstate.callbacks.dc = dc; - dc->dml2_options.svp_pstate.callbacks.add_phantom_plane = &dc_state_add_phantom_plane; - dc->dml2_options.svp_pstate.callbacks.add_phantom_stream = &dc_state_add_phantom_stream; - dc->dml2_options.svp_pstate.callbacks.build_scaling_params = &resource_build_scaling_params; - dc->dml2_options.svp_pstate.callbacks.create_phantom_plane = &dc_state_create_phantom_plane; - dc->dml2_options.svp_pstate.callbacks.remove_phantom_plane = &dc_state_remove_phantom_plane; - dc->dml2_options.svp_pstate.callbacks.remove_phantom_stream = &dc_state_remove_phantom_stream; - dc->dml2_options.svp_pstate.callbacks.create_phantom_stream = &dc_state_create_phantom_stream; - dc->dml2_options.svp_pstate.callbacks.release_phantom_plane = &dc_state_release_phantom_plane; - dc->dml2_options.svp_pstate.callbacks.release_phantom_stream = &dc_state_release_phantom_stream; dc->dml2_options.svp_pstate.callbacks.release_dsc = &dcn20_release_dsc; - dc->dml2_options.svp_pstate.callbacks.get_pipe_subvp_type = &dc_state_get_pipe_subvp_type; - dc->dml2_options.svp_pstate.callbacks.get_stream_subvp_type = &dc_state_get_stream_subvp_type; - dc->dml2_options.svp_pstate.callbacks.get_paired_subvp_stream = &dc_state_get_paired_subvp_stream; + dc->dml2_options.svp_pstate.callbacks.calculate_mall_ways_from_bytes = pool->base.funcs->calculate_mall_ways_from_bytes; dc->dml2_options.svp_pstate.subvp_fw_processing_delay_us = dc->caps.subvp_fw_processing_delay_us; dc->dml2_options.svp_pstate.subvp_prefetch_end_to_mall_start_us = dc->caps.subvp_prefetch_end_to_mall_start_us; @@ -2483,7 +2552,7 @@ struct resource_pool *dcn32_create_resource_pool( * full update which delays the flip for 1 frame. If we use the original pipe * we don't have to toggle its power. So we can flip faster. */ -static int find_optimal_free_pipe_as_secondary_dpp_pipe( +int dcn32_find_optimal_free_pipe_as_secondary_dpp_pipe( const struct resource_context *cur_res_ctx, struct resource_context *new_res_ctx, const struct resource_pool *pool, @@ -2666,7 +2735,7 @@ struct pipe_ctx *dcn32_acquire_free_pipe_as_secondary_dpp_pipe( return dcn32_acquire_idle_pipe_for_head_pipe_in_layer( new_ctx, pool, opp_head_pipe->stream, opp_head_pipe); - free_pipe_idx = find_optimal_free_pipe_as_secondary_dpp_pipe( + free_pipe_idx = dcn32_find_optimal_free_pipe_as_secondary_dpp_pipe( &cur_ctx->res_ctx, &new_ctx->res_ctx, pool, opp_head_pipe); if (free_pipe_idx >= 0) { diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn32/dcn32_resource.h b/drivers/gpu/drm/amd/display/dc/resource/dcn32/dcn32_resource.h index 2258c5c7212d..fee67fbab8e2 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn32/dcn32_resource.h +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn32/dcn32_resource.h @@ -113,10 +113,6 @@ void dcn32_calculate_wm_and_dlg( int pipe_cnt, int vlevel); -uint32_t dcn32_helper_mall_bytes_to_ways( - struct dc *dc, - uint32_t total_size_in_mall_bytes); - uint32_t dcn32_helper_calculate_mall_bytes_for_cursor( struct dc *dc, struct pipe_ctx *pipe_ctx, @@ -141,6 +137,12 @@ bool dcn32_any_surfaces_rotated(struct dc *dc, struct dc_state *context); bool dcn32_is_center_timing(struct pipe_ctx *pipe); bool dcn32_is_psr_capable(struct pipe_ctx *pipe); +int dcn32_find_optimal_free_pipe_as_secondary_dpp_pipe( + const struct resource_context *cur_res_ctx, + struct resource_context *new_res_ctx, + const struct resource_pool *pool, + const struct pipe_ctx *new_opp_head); + struct pipe_ctx *dcn32_acquire_free_pipe_as_secondary_dpp_pipe( const struct dc_state *cur_ctx, struct dc_state *new_ctx, @@ -184,6 +186,8 @@ void dcn32_update_dml_pipes_odm_policy_based_on_context(struct dc *dc, struct dc void dcn32_override_min_req_dcfclk(struct dc *dc, struct dc_state *context); +unsigned int dcn32_calculate_mall_ways_from_bytes(const struct dc *dc, unsigned int total_size_in_mall_bytes); + /* definitions for run time init of reg offsets */ /* CLK SRC */ diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn321/dcn321_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn321/dcn321_resource.c index 296a0a8e7145..e4b360d89b3b 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn321/dcn321_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn321/dcn321_resource.c @@ -1288,6 +1288,8 @@ static struct hpo_dp_link_encoder *dcn321_hpo_dp_link_encoder_create( /* allocate HPO link encoder */ hpo_dp_enc31 = kzalloc(sizeof(struct dcn31_hpo_dp_link_encoder), GFP_KERNEL); + if (!hpo_dp_enc31) + return NULL; /* out of memory */ #undef REG_STRUCT #define REG_STRUCT hpo_dp_link_enc_regs @@ -1579,10 +1581,20 @@ static struct dc_cap_funcs cap_funcs = { static void dcn321_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_params) { + struct dml2_configuration_options dml2_opt = dc->dml2_options; + DC_FP_START(); + dcn321_update_bw_bounding_box_fpu(dc, bw_params); + + dml2_opt.use_clock_dc_limits = false; if (dc->debug.using_dml2 && dc->current_state && dc->current_state->bw_ctx.dml2) - dml2_reinit(dc, &dc->dml2_options, &dc->current_state->bw_ctx.dml2); + dml2_reinit(dc, &dml2_opt, &dc->current_state->bw_ctx.dml2); + + dml2_opt.use_clock_dc_limits = true; + if (dc->debug.using_dml2 && dc->current_state && dc->current_state->bw_ctx.dml2_dc_power_source) + dml2_reinit(dc, &dml2_opt, &dc->current_state->bw_ctx.dml2_dc_power_source); + DC_FP_END(); } @@ -1610,6 +1622,7 @@ static struct resource_funcs dcn321_res_pool_funcs = { .update_soc_for_wm_a = dcn30_update_soc_for_wm_a, .add_phantom_pipes = dcn32_add_phantom_pipes, .build_pipe_pix_clk_params = dcn20_build_pipe_pix_clk_params, + .calculate_mall_ways_from_bytes = dcn32_calculate_mall_ways_from_bytes, }; static uint32_t read_pipe_fuses(struct dc_context *ctx) @@ -1697,7 +1710,9 @@ static bool dcn321_resource_construct( dc->caps.min_horizontal_blanking_period = 80; dc->caps.dmdata_alloc_size = 2048; dc->caps.mall_size_per_mem_channel = 4; - dc->caps.mall_size_total = 0; + /* total size = mall per channel * num channels * 1024 * 1024 */ + dc->caps.mall_size_total = dc->caps.mall_size_per_mem_channel * dc->ctx->dc_bios->vram_info.num_chans * 1048576; + dc->caps.cursor_cache_size = dc->caps.max_cursor_size * dc->caps.max_cursor_size * 8; dc->caps.cache_line_size = 64; dc->caps.cache_num_ways = 16; @@ -1998,30 +2013,10 @@ static bool dcn321_resource_construct( dc->dml2_options.use_native_soc_bb_construction = true; dc->dml2_options.minimize_dispclk_using_odm = true; - dc->dml2_options.callbacks.dc = dc; - dc->dml2_options.callbacks.build_scaling_params = &resource_build_scaling_params; + resource_init_common_dml2_callbacks(dc, &dc->dml2_options); dc->dml2_options.callbacks.can_support_mclk_switch_using_fw_based_vblank_stretch = &dcn30_can_support_mclk_switch_using_fw_based_vblank_stretch; - dc->dml2_options.callbacks.acquire_secondary_pipe_for_mpc_odm = &dc_resource_acquire_secondary_pipe_for_mpc_odm_legacy; - dc->dml2_options.callbacks.update_pipes_for_stream_with_slice_count = &resource_update_pipes_for_stream_with_slice_count; - dc->dml2_options.callbacks.update_pipes_for_plane_with_slice_count = &resource_update_pipes_for_plane_with_slice_count; - dc->dml2_options.callbacks.get_mpc_slice_index = &resource_get_mpc_slice_index; - dc->dml2_options.callbacks.get_odm_slice_index = &resource_get_odm_slice_index; - dc->dml2_options.callbacks.get_opp_head = &resource_get_opp_head; - - dc->dml2_options.svp_pstate.callbacks.dc = dc; - dc->dml2_options.svp_pstate.callbacks.add_phantom_plane = &dc_state_add_phantom_plane; - dc->dml2_options.svp_pstate.callbacks.add_phantom_stream = &dc_state_add_phantom_stream; - dc->dml2_options.svp_pstate.callbacks.build_scaling_params = &resource_build_scaling_params; - dc->dml2_options.svp_pstate.callbacks.create_phantom_plane = &dc_state_create_phantom_plane; - dc->dml2_options.svp_pstate.callbacks.remove_phantom_plane = &dc_state_remove_phantom_plane; - dc->dml2_options.svp_pstate.callbacks.remove_phantom_stream = &dc_state_remove_phantom_stream; - dc->dml2_options.svp_pstate.callbacks.create_phantom_stream = &dc_state_create_phantom_stream; - dc->dml2_options.svp_pstate.callbacks.release_phantom_plane = &dc_state_release_phantom_plane; - dc->dml2_options.svp_pstate.callbacks.release_phantom_stream = &dc_state_release_phantom_stream; dc->dml2_options.svp_pstate.callbacks.release_dsc = &dcn20_release_dsc; - dc->dml2_options.svp_pstate.callbacks.get_pipe_subvp_type = &dc_state_get_pipe_subvp_type; - dc->dml2_options.svp_pstate.callbacks.get_stream_subvp_type = &dc_state_get_stream_subvp_type; - dc->dml2_options.svp_pstate.callbacks.get_paired_subvp_stream = &dc_state_get_paired_subvp_stream; + dc->dml2_options.svp_pstate.callbacks.calculate_mall_ways_from_bytes = pool->base.funcs->calculate_mall_ways_from_bytes; dc->dml2_options.svp_pstate.subvp_fw_processing_delay_us = dc->caps.subvp_fw_processing_delay_us; dc->dml2_options.svp_pstate.subvp_prefetch_end_to_mall_start_us = dc->caps.subvp_prefetch_end_to_mall_start_us; diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn35/dcn35_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn35/dcn35_resource.c index 5d52853cac96..2df8a742516c 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn35/dcn35_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn35/dcn35_resource.c @@ -721,7 +721,7 @@ static const struct dc_debug_options debug_defaults_drv = { .disable_dpp_power_gate = true, .disable_hubp_power_gate = true, .disable_optc_power_gate = true, /*should the same as above two*/ - .disable_hpo_power_gate = true, /*dmubfw force domain25 on*/ + .disable_hpo_power_gate = false, /*dmubfw force domain25 on*/ .disable_clock_gate = false, .disable_dsc_power_gate = true, .vsr_support = true, @@ -764,12 +764,12 @@ static const struct dc_debug_options debug_defaults_drv = { }, .seamless_boot_odm_combine = DML_FAIL_SOURCE_PIXEL_FORMAT, .enable_z9_disable_interface = true, /* Allow support for the PMFW interface for disable Z9*/ - .minimum_z8_residency_time = 2100, + .minimum_z8_residency_time = 1, /* Always allow when other conditions are met */ .using_dml2 = true, .support_eDP1_5 = true, .enable_hpo_pg_support = false, .enable_legacy_fast_update = true, - .enable_single_display_2to1_odm_policy = false, + .enable_single_display_2to1_odm_policy = true, .disable_idle_power_optimizations = false, .dmcub_emulation = false, .disable_boot_optimizations = false, @@ -783,7 +783,7 @@ static const struct dc_debug_options debug_defaults_drv = { .psp_disabled_wa = true, .ips2_eval_delay_us = 2000, .ips2_entry_delay_us = 800, - .disable_dmub_reallow_idle = true, + .disable_dmub_reallow_idle = false, .static_screen_wait_frames = 2, }; @@ -1368,6 +1368,8 @@ static struct hpo_dp_link_encoder *dcn31_hpo_dp_link_encoder_create( /* allocate HPO link encoder */ hpo_dp_enc31 = kzalloc(sizeof(struct dcn31_hpo_dp_link_encoder), GFP_KERNEL); + if (!hpo_dp_enc31) + return NULL; /* out of memory */ #undef REG_STRUCT #define REG_STRUCT hpo_dp_link_enc_regs @@ -1734,7 +1736,9 @@ static bool dcn35_validate_bandwidth(struct dc *dc, { bool out = false; - out = dml2_validate(dc, context, fast_validate); + out = dml2_validate(dc, context, + context->power_source == DC_POWER_SOURCE_DC ? context->bw_ctx.dml2_dc_power_source : context->bw_ctx.dml2, + fast_validate); if (fast_validate) return out; @@ -2138,15 +2142,9 @@ static bool dcn35_resource_construct( dc->dml2_options.minimize_dispclk_using_odm = true; dc->dml2_options.enable_windowed_mpo_odm = dc->config.enable_windowed_mpo_odm; - dc->dml2_options.callbacks.dc = dc; - dc->dml2_options.callbacks.build_scaling_params = &resource_build_scaling_params; + resource_init_common_dml2_callbacks(dc, &dc->dml2_options); dc->dml2_options.callbacks.can_support_mclk_switch_using_fw_based_vblank_stretch = &dcn30_can_support_mclk_switch_using_fw_based_vblank_stretch; - dc->dml2_options.callbacks.acquire_secondary_pipe_for_mpc_odm = &dc_resource_acquire_secondary_pipe_for_mpc_odm_legacy; - dc->dml2_options.callbacks.update_pipes_for_stream_with_slice_count = &resource_update_pipes_for_stream_with_slice_count; - dc->dml2_options.callbacks.update_pipes_for_plane_with_slice_count = &resource_update_pipes_for_plane_with_slice_count; - dc->dml2_options.callbacks.get_mpc_slice_index = &resource_get_mpc_slice_index; - dc->dml2_options.callbacks.get_odm_slice_index = &resource_get_odm_slice_index; - dc->dml2_options.callbacks.get_opp_head = &resource_get_opp_head; + dc->dml2_options.max_segments_per_hubp = 24; dc->dml2_options.det_segment_size = DCN3_2_DET_SEG_SIZE;/*todo*/ diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn35/dcn35_resource.h b/drivers/gpu/drm/amd/display/dc/resource/dcn35/dcn35_resource.h index a51c4a9eaafe..f97bb4cb3761 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn35/dcn35_resource.h +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn35/dcn35_resource.h @@ -240,6 +240,8 @@ struct resource_pool *dcn35_create_resource_pool( SRI_ARR(OTG_V_TOTAL_MAX, OTG, inst),\ SRI_ARR(OTG_V_TOTAL_MIN, OTG, inst),\ SRI_ARR(OTG_V_TOTAL_CONTROL, OTG, inst),\ + SRI_ARR(OTG_V_COUNT_STOP_CONTROL, OTG, inst),\ + SRI_ARR(OTG_V_COUNT_STOP_CONTROL2, OTG, inst),\ SRI_ARR(OTG_TRIGA_CNTL, OTG, inst),\ SRI_ARR(OTG_FORCE_COUNT_NOW_CNTL, OTG, inst),\ SRI_ARR(OTG_STATIC_SCREEN_CONTROL, OTG, inst),\ diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn351/dcn351_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn351/dcn351_resource.c index 5b486400dfdb..ddf9560ab772 100644 --- a/drivers/gpu/drm/amd/display/dc/resource/dcn351/dcn351_resource.c +++ b/drivers/gpu/drm/amd/display/dc/resource/dcn351/dcn351_resource.c @@ -700,6 +700,8 @@ static const struct dc_debug_options debug_defaults_drv = { .disable_dcc = DCC_ENABLE, .disable_dpp_power_gate = true, .disable_hubp_power_gate = true, + .disable_optc_power_gate = true, /*should the same as above two*/ + .disable_hpo_power_gate = true, /*dmubfw force domain25 on*/ .disable_clock_gate = false, .disable_dsc_power_gate = true, .vsr_support = true, @@ -742,12 +744,13 @@ static const struct dc_debug_options debug_defaults_drv = { }, .seamless_boot_odm_combine = DML_FAIL_SOURCE_PIXEL_FORMAT, .enable_z9_disable_interface = true, /* Allow support for the PMFW interface for disable Z9*/ + .minimum_z8_residency_time = 2100, .using_dml2 = true, .support_eDP1_5 = true, .enable_hpo_pg_support = false, .enable_legacy_fast_update = true, .enable_single_display_2to1_odm_policy = true, - .disable_idle_power_optimizations = true, + .disable_idle_power_optimizations = false, .dmcub_emulation = false, .disable_boot_optimizations = false, .disable_unbounded_requesting = false, @@ -755,11 +758,13 @@ static const struct dc_debug_options debug_defaults_drv = { //must match enable_single_display_2to1_odm_policy to support dynamic ODM transitions .enable_double_buffered_dsc_pg_support = true, .enable_dp_dig_pixel_rate_div_policy = 1, - .disable_z10 = true, + .disable_z10 = false, .ignore_pg = true, .psp_disabled_wa = true, - .ips2_eval_delay_us = 200, - .ips2_entry_delay_us = 400 + .ips2_eval_delay_us = 2000, + .ips2_entry_delay_us = 800, + .disable_dmub_reallow_idle = true, + .static_screen_wait_frames = 2, }; static const struct dc_panel_config panel_config_defaults = { @@ -1343,6 +1348,8 @@ static struct hpo_dp_link_encoder *dcn31_hpo_dp_link_encoder_create( /* allocate HPO link encoder */ hpo_dp_enc31 = kzalloc(sizeof(struct dcn31_hpo_dp_link_encoder), GFP_KERNEL); + if (!hpo_dp_enc31) + return NULL; /* out of memory */ #undef REG_STRUCT #define REG_STRUCT hpo_dp_link_enc_regs @@ -1709,19 +1716,20 @@ static bool dcn351_validate_bandwidth(struct dc *dc, { bool out = false; - out = dml2_validate(dc, context, fast_validate); + out = dml2_validate(dc, context, + context->power_source == DC_POWER_SOURCE_DC ? context->bw_ctx.dml2_dc_power_source : context->bw_ctx.dml2, + fast_validate); if (fast_validate) return out; DC_FP_START(); - dcn351_decide_zstate_support(dc, context); + dcn35_decide_zstate_support(dc, context); DC_FP_END(); return out; } - static struct resource_funcs dcn351_res_pool_funcs = { .destroy = dcn351_destroy_resource_pool, .link_enc_create = dcn35_link_encoder_create, @@ -1864,6 +1872,9 @@ static bool dcn351_resource_construct( /* Use pipe context based otg sync logic */ dc->config.use_pipe_ctx_sync_logic = true; + /* Use psp mailbox to enable assr */ + dc->config.use_assr_psp_message = true; + /* read VBIOS LTTPR caps */ { if (ctx->dc_bios->funcs->get_lttpr_caps) { @@ -1883,6 +1894,8 @@ static bool dcn351_resource_construct( if (dc->ctx->dce_environment == DCE_ENV_PRODUCTION_DRV) dc->debug = debug_defaults_drv; + /*HW default is to have all the FGCG enabled, SW no need to program them*/ + dc->debug.enable_fine_grain_clock_gating.u32All = 0xFFFF; // Init the vm_helper if (dc->vm_helper) vm_helper_init(dc->vm_helper, 16); @@ -2113,15 +2126,9 @@ static bool dcn351_resource_construct( dc->dml2_options.minimize_dispclk_using_odm = true; dc->dml2_options.enable_windowed_mpo_odm = dc->config.enable_windowed_mpo_odm; - dc->dml2_options.callbacks.dc = dc; - dc->dml2_options.callbacks.build_scaling_params = &resource_build_scaling_params; + resource_init_common_dml2_callbacks(dc, &dc->dml2_options); dc->dml2_options.callbacks.can_support_mclk_switch_using_fw_based_vblank_stretch = &dcn30_can_support_mclk_switch_using_fw_based_vblank_stretch; - dc->dml2_options.callbacks.acquire_secondary_pipe_for_mpc_odm = &dc_resource_acquire_secondary_pipe_for_mpc_odm_legacy; - dc->dml2_options.callbacks.update_pipes_for_stream_with_slice_count = &resource_update_pipes_for_stream_with_slice_count; - dc->dml2_options.callbacks.update_pipes_for_plane_with_slice_count = &resource_update_pipes_for_plane_with_slice_count; - dc->dml2_options.callbacks.get_mpc_slice_index = &resource_get_mpc_slice_index; - dc->dml2_options.callbacks.get_odm_slice_index = &resource_get_odm_slice_index; - dc->dml2_options.callbacks.get_opp_head = &resource_get_opp_head; + dc->dml2_options.max_segments_per_hubp = 24; dc->dml2_options.det_segment_size = DCN3_2_DET_SEG_SIZE;/*todo*/ diff --git a/drivers/gpu/drm/amd/display/dmub/dmub_srv.h b/drivers/gpu/drm/amd/display/dmub/dmub_srv.h index 7785908a6676..2fde1f043d50 100644 --- a/drivers/gpu/drm/amd/display/dmub/dmub_srv.h +++ b/drivers/gpu/drm/amd/display/dmub/dmub_srv.h @@ -71,6 +71,8 @@ extern "C" { #endif +#define DMUB_PC_SNAPSHOT_COUNT 10 + /* Forward declarations */ struct dmub_srv; struct dmub_srv_common_regs; @@ -295,18 +297,30 @@ struct dmub_srv_hw_params { bool dpia_hpd_int_enable_supported; bool disable_clock_gate; bool disallow_dispclk_dppclk_ds; + bool ips_sequential_ono; enum dmub_memory_access_type mem_access_type; enum dmub_ips_disable_type disable_ips; }; /** + * struct dmub_srv_debug - Debug info for dmub_srv + * @timeout_occured: Indicates a timeout occured on any message from driver to dmub + * @timeout_cmd: first cmd sent from driver that timed out - subsequent timeouts are not stored + */ +struct dmub_srv_debug { + bool timeout_occured; + union dmub_rb_cmd timeout_cmd; + unsigned long long timestamp; +}; + +/** * struct dmub_diagnostic_data - Diagnostic data retrieved from DMCUB for * debugging purposes, including logging, crash analysis, etc. */ struct dmub_diagnostic_data { uint32_t dmcub_version; uint32_t scratch[17]; - uint32_t pc; + uint32_t pc[DMUB_PC_SNAPSHOT_COUNT]; uint32_t undefined_address_fault_addr; uint32_t inst_fetch_fault_addr; uint32_t data_write_fault_addr; @@ -317,6 +331,7 @@ struct dmub_diagnostic_data { uint32_t inbox0_wptr; uint32_t inbox0_size; uint32_t gpint_datain0; + struct dmub_srv_debug timeout_info; uint8_t is_dmcub_enabled : 1; uint8_t is_dmcub_soft_reset : 1; uint8_t is_dmcub_secure_reset : 1; @@ -506,6 +521,7 @@ struct dmub_srv { struct dmub_visual_confirm_color visual_confirm_color; enum dmub_srv_power_state_type power_state; + struct dmub_srv_debug debug; }; /** 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 af3fe8bb0728..e85fd3ac52c7 100644 --- a/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h +++ b/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h @@ -97,6 +97,9 @@ /* Maximum number of planes on any ASIC. */ #define DMUB_MAX_PLANES 6 +/* Maximum number of phantom planes on any ASIC */ +#define DMUB_MAX_PHANTOM_PLANES ((DMUB_MAX_PLANES) / 2) + /* Trace buffer offset for entry */ #define TRACE_BUFFER_ENTRY_OFFSET 16 @@ -194,6 +197,11 @@ union abm_flags { * of user backlight level. */ unsigned int abm_gradual_bl_change : 1; + + /** + * @abm_new_frame: Indicates if a new frame update needed for ABM to ramp up into steady + */ + unsigned int abm_new_frame : 1; } bitfields; unsigned int u32All; @@ -461,7 +469,7 @@ struct dmub_feature_caps { * Max PSR version supported by FW. */ uint8_t psr; - uint8_t fw_assisted_mclk_switch; + uint8_t fw_assisted_mclk_switch_ver; uint8_t reserved[4]; uint8_t subvp_psr_support; uint8_t gecc_enable; @@ -619,6 +627,7 @@ enum dmub_ips_disable_type { DMUB_IPS_DISABLE_IPS2 = 3, DMUB_IPS_DISABLE_IPS2_Z10 = 4, DMUB_IPS_DISABLE_DYNAMIC = 5, + DMUB_IPS_RCG_IN_ACTIVE_IPS2_IN_OFF = 6, }; #define DMUB_IPS1_ALLOW_MASK 0x00000001 @@ -653,6 +662,7 @@ union dmub_fw_boot_options { uint32_t disable_timeout_recovery : 1; /* 1 if timeout recovery should be disabled */ uint32_t ips_pg_disable: 1; /* 1 to disable ONO domains power gating*/ uint32_t ips_disable: 3; /* options to disable ips support*/ + uint32_t ips_sequential_ono: 1; /**< 1 to enable sequential ONO IPS sequence */ uint32_t reserved : 9; /**< reserved */ } bits; /**< boot bits */ uint32_t all; /**< 32-bit access to bits */ @@ -695,7 +705,8 @@ union dmub_shared_state_ips_fw_signals { struct { uint32_t ips1_commit : 1; /**< 1 if in IPS1 */ uint32_t ips2_commit : 1; /**< 1 if in IPS2 */ - uint32_t reserved_bits : 30; /**< Reversed */ + uint32_t in_idle : 1; /**< 1 if DMCUB is in idle */ + uint32_t reserved_bits : 29; /**< Reversed */ } bits; uint32_t all; }; @@ -724,7 +735,13 @@ union dmub_shared_state_ips_driver_signals { */ struct dmub_shared_state_ips_fw { union dmub_shared_state_ips_fw_signals signals; /**< 4 bytes, IPS signal bits */ - uint32_t reserved[61]; /**< Reversed, to be updated when adding new fields. */ + uint32_t rcg_entry_count; /**< Entry counter for RCG */ + uint32_t rcg_exit_count; /**< Exit counter for RCG */ + uint32_t ips1_entry_count; /**< Entry counter for IPS1 */ + uint32_t ips1_exit_count; /**< Exit counter for IPS1 */ + uint32_t ips2_entry_count; /**< Entry counter for IPS2 */ + uint32_t ips2_exit_count; /**< Exit counter for IPS2 */ + uint32_t reserved[55]; /**< Reversed, to be updated when adding new fields. */ }; /* 248-bytes, fixed */ /** @@ -812,6 +829,10 @@ enum dmub_cmd_vbios_type { */ DMUB_CMD__VBIOS_TRANSMITTER_QUERY_DP_ALT = 26, /** + * Control PHY FSM + */ + DMUB_CMD__VBIOS_TRANSMITTER_SET_PHY_FSM = 29, + /** * Controls domain power gating */ DMUB_CMD__VBIOS_DOMAIN_CONTROL = 28, @@ -1186,6 +1207,11 @@ enum dmub_cmd_type { */ DMUB_CMD__DPIA_HPD_INT_ENABLE = 86, + /** + * Command type used for all PSP commands. + */ + DMUB_CMD__PSP = 88, + DMUB_CMD__VBIOS = 128, }; @@ -1588,7 +1614,7 @@ struct dmub_rb_cmd_idle_opt_dcn_restore { */ struct dmub_dcn_notify_idle_cntl_data { uint8_t driver_idle; - uint8_t pad[1]; + uint8_t reserved[59]; }; /** @@ -2309,6 +2335,11 @@ enum phy_link_rate { * UHBR10 - 20.0 Gbps/Lane */ PHY_RATE_2000 = 11, + + PHY_RATE_675 = 12, + /** + * Rate 12 - 6.75 Gbps/Lane + */ }; /** @@ -2327,6 +2358,7 @@ enum dmub_phy_fsm_state { DMUB_PHY_FSM_POWER_DOWN, DMUB_PHY_FSM_PLL_EN, DMUB_PHY_FSM_TX_EN, + DMUB_PHY_FSM_TX_EN_TEST_MODE, DMUB_PHY_FSM_FAST_LP, DMUB_PHY_FSM_P2_PLL_OFF_CPM, DMUB_PHY_FSM_P2_PLL_OFF_PG, @@ -2931,18 +2963,49 @@ struct dmub_rb_cmd_psr_set_power_opt { struct dmub_cmd_psr_set_power_opt_data psr_set_power_opt_data; }; +/** + * Definition of Replay Residency GPINT command. + * Bit[0] - Residency mode for Revision 0 + * Bit[1] - Enable/Disable state + * Bit[2-3] - Revision number + * Bit[4-7] - Residency mode for Revision 1 + * Bit[8] - Panel instance + * Bit[9-15] - Reserved + */ + +enum pr_residency_mode { + PR_RESIDENCY_MODE_PHY = 0x0, + PR_RESIDENCY_MODE_ALPM, + PR_RESIDENCY_MODE_IPS2, + PR_RESIDENCY_MODE_FRAME_CNT, + PR_RESIDENCY_MODE_ENABLEMENT_PERIOD, +}; + #define REPLAY_RESIDENCY_MODE_SHIFT (0) #define REPLAY_RESIDENCY_ENABLE_SHIFT (1) +#define REPLAY_RESIDENCY_REVISION_SHIFT (2) +#define REPLAY_RESIDENCY_MODE2_SHIFT (4) #define REPLAY_RESIDENCY_MODE_MASK (0x1 << REPLAY_RESIDENCY_MODE_SHIFT) -# define REPLAY_RESIDENCY_MODE_PHY (0x0 << REPLAY_RESIDENCY_MODE_SHIFT) -# define REPLAY_RESIDENCY_MODE_ALPM (0x1 << REPLAY_RESIDENCY_MODE_SHIFT) -# define REPLAY_RESIDENCY_MODE_IPS 0x10 +# define REPLAY_RESIDENCY_FIELD_MODE_PHY (0x0 << REPLAY_RESIDENCY_MODE_SHIFT) +# define REPLAY_RESIDENCY_FIELD_MODE_ALPM (0x1 << REPLAY_RESIDENCY_MODE_SHIFT) + +#define REPLAY_RESIDENCY_MODE2_MASK (0xF << REPLAY_RESIDENCY_MODE2_SHIFT) +# define REPLAY_RESIDENCY_FIELD_MODE2_IPS (0x1 << REPLAY_RESIDENCY_MODE2_SHIFT) +# define REPLAY_RESIDENCY_FIELD_MODE2_FRAME_CNT (0x2 << REPLAY_RESIDENCY_MODE2_SHIFT) +# define REPLAY_RESIDENCY_FIELD_MODE2_EN_PERIOD (0x3 << REPLAY_RESIDENCY_MODE2_SHIFT) #define REPLAY_RESIDENCY_ENABLE_MASK (0x1 << REPLAY_RESIDENCY_ENABLE_SHIFT) # define REPLAY_RESIDENCY_DISABLE (0x0 << REPLAY_RESIDENCY_ENABLE_SHIFT) # define REPLAY_RESIDENCY_ENABLE (0x1 << REPLAY_RESIDENCY_ENABLE_SHIFT) +#define REPLAY_RESIDENCY_REVISION_MASK (0x3 << REPLAY_RESIDENCY_REVISION_SHIFT) +# define REPLAY_RESIDENCY_REVISION_0 (0x0 << REPLAY_RESIDENCY_REVISION_SHIFT) +# define REPLAY_RESIDENCY_REVISION_1 (0x1 << REPLAY_RESIDENCY_REVISION_SHIFT) + +/** + * Definition of a replay_state. + */ enum replay_state { REPLAY_STATE_0 = 0x0, REPLAY_STATE_1 = 0x10, @@ -3004,6 +3067,11 @@ enum dmub_cmd_replay_type { * Set pseudo vtotal */ DMUB_CMD__REPLAY_SET_PSEUDO_VTOTAL = 7, + /** + * Set adaptive sync sdp enabled + */ + DMUB_CMD__REPLAY_DISABLED_ADAPTIVE_SYNC_SDP = 8, + }; /** @@ -3205,6 +3273,20 @@ struct dmub_cmd_replay_set_pseudo_vtotal { */ uint8_t pad; }; +struct dmub_cmd_replay_disabled_adaptive_sync_sdp_data { + /** + * Panel Instance. + * Panel isntance to identify which replay_state to use + * Currently the support is only for 0 or 1 + */ + uint8_t panel_inst; + /** + * enabled: set adaptive sync sdp enabled + */ + uint8_t force_disabled; + + uint8_t pad[2]; +}; /** * Definition of a DMUB_CMD__SET_REPLAY_POWER_OPT command. @@ -3309,6 +3391,20 @@ struct dmub_rb_cmd_replay_set_pseudo_vtotal { }; /** + * Definition of a DMUB_CMD__REPLAY_DISABLED_ADAPTIVE_SYNC_SDP command. + */ +struct dmub_rb_cmd_replay_disabled_adaptive_sync_sdp { + /** + * Command header. + */ + struct dmub_cmd_header header; + /** + * Definition of DMUB_CMD__REPLAY_DISABLED_ADAPTIVE_SYNC_SDP command. + */ + struct dmub_cmd_replay_disabled_adaptive_sync_sdp_data data; +}; + +/** * Data passed from driver to FW in DMUB_CMD__REPLAY_SET_RESIDENCY_FRAMEUPDATE_TIMER command. */ struct dmub_cmd_replay_frameupdate_timer_data { @@ -3363,6 +3459,11 @@ union dmub_replay_cmd_set { * Definition of DMUB_CMD__REPLAY_SET_PSEUDO_VTOTAL command data. */ struct dmub_cmd_replay_set_pseudo_vtotal pseudo_vtotal_data; + /** + * Definition of DMUB_CMD__REPLAY_DISABLED_ADAPTIVE_SYNC_SDP command data. + */ + struct dmub_cmd_replay_disabled_adaptive_sync_sdp_data disabled_adaptive_sync_sdp_data; + }; /** @@ -3445,7 +3546,7 @@ enum hw_lock_client { /** * Replay is the client of HW Lock Manager. */ - HW_LOCK_CLIENT_REPLAY = 4, + HW_LOCK_CLIENT_REPLAY = 4, /** * Invalid client. */ @@ -4038,6 +4139,10 @@ enum dmub_cmd_panel_cntl_type { * Queries backlight info for the embedded panel. */ DMUB_CMD__PANEL_CNTL_QUERY_BACKLIGHT_INFO = 1, + /** + * Sets the PWM Freq as per user's requirement. + */ + DMUB_CMD__PANEL_DEBUG_PWM_FREQ = 2, }; /** @@ -4139,6 +4244,34 @@ struct dmub_rb_cmd_transmitter_query_dp_alt { struct dmub_rb_cmd_transmitter_query_dp_alt_data data; /**< payload */ }; +struct phy_test_mode { + uint8_t mode; + uint8_t pat0; + uint8_t pad[2]; +}; + +/** + * Data passed in/out in a DMUB_CMD__VBIOS_TRANSMITTER_SET_PHY_FSM command. + */ +struct dmub_rb_cmd_transmitter_set_phy_fsm_data { + uint8_t phy_id; /**< 0=UNIPHYA, 1=UNIPHYB, 2=UNIPHYC, 3=UNIPHYD, 4=UNIPHYE, 5=UNIPHYF */ + uint8_t mode; /**< HDMI/DP/DP2 etc */ + uint8_t lane_num; /**< Number of lanes */ + uint32_t symclk_100Hz; /**< PLL symclock in 100hz */ + struct phy_test_mode test_mode; + enum dmub_phy_fsm_state state; + uint32_t status; + uint8_t pad; +}; + +/** + * Definition of a DMUB_CMD__VBIOS_TRANSMITTER_SET_PHY_FSM command. + */ +struct dmub_rb_cmd_transmitter_set_phy_fsm { + struct dmub_cmd_header header; /**< header */ + struct dmub_rb_cmd_transmitter_set_phy_fsm_data data; /**< payload */ +}; + /** * Maximum number of bytes a chunk sent to DMUB for parsing */ @@ -4261,6 +4394,65 @@ struct dmub_rb_cmd_secure_display { }; /** + * Command type of a DMUB_CMD__PSP command + */ +enum dmub_cmd_psp_type { + DMUB_CMD__PSP_ASSR_ENABLE = 0 +}; + +/** + * Data passed from driver to FW in a DMUB_CMD__PSP_ASSR_ENABLE command. + */ +struct dmub_cmd_assr_enable_data { + /** + * ASSR enable or disable. + */ + uint8_t enable; + /** + * PHY port type. + * Indicates eDP / non-eDP port type + */ + uint8_t phy_port_type; + /** + * PHY port ID. + */ + uint8_t phy_port_id; + /** + * Link encoder index. + */ + uint8_t link_enc_index; + /** + * HPO mode. + */ + uint8_t hpo_mode; + + /** + * Reserved field. + */ + uint8_t reserved[7]; +}; + +/** + * Definition of a DMUB_CMD__PSP_ASSR_ENABLE command. + */ +struct dmub_rb_cmd_assr_enable { + /** + * Command header. + */ + struct dmub_cmd_header header; + + /** + * Assr data. + */ + struct dmub_cmd_assr_enable_data assr_data; + + /** + * Reserved field. + */ + uint32_t reserved[3]; +}; + +/** * union dmub_rb_cmd - DMUB inbox command. */ union dmub_rb_cmd { @@ -4451,6 +4643,10 @@ union dmub_rb_cmd { */ struct dmub_rb_cmd_transmitter_query_dp_alt query_dp_alt; /** + * Definition of a DMUB_CMD__VBIOS_TRANSMITTER_SET_PHY_FSM command. + */ + struct dmub_rb_cmd_transmitter_set_phy_fsm set_phy_fsm; + /** * Definition of a DMUB_CMD__DPIA_DIG1_CONTROL command. */ struct dmub_rb_cmd_dig1_dpia_control dig1_dpia_control; @@ -4518,6 +4714,15 @@ union dmub_rb_cmd { * Definition of a DMUB_CMD__REPLAY_SET_PSEUDO_VTOTAL command. */ struct dmub_rb_cmd_replay_set_pseudo_vtotal replay_set_pseudo_vtotal; + /** + * Definition of a DMUB_CMD__REPLAY_DISABLED_ADAPTIVE_SYNC_SDP command. + */ + struct dmub_rb_cmd_replay_disabled_adaptive_sync_sdp replay_disabled_adaptive_sync_sdp; + /** + * Definition of a DMUB_CMD__PSP_ASSR_ENABLE command. + */ + struct dmub_rb_cmd_assr_enable assr_enable; + }; /** 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 cae96fba6349..e500ca9ae09c 100644 --- a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn20.c +++ b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn20.c @@ -472,4 +472,5 @@ void dmub_dcn20_get_diagnostic_data(struct dmub_srv *dmub, struct dmub_diagnosti REG_GET(DMCUB_REGION3_CW6_TOP_ADDRESS, DMCUB_REGION3_CW6_ENABLE, &is_cw6_enabled); diag_data->is_cw6_enabled = is_cw6_enabled; + diag_data->timeout_info = dmub->debug; } diff --git a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn31.c b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn31.c index 2bcf5fb87dd9..662c34e9495c 100644 --- a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn31.c +++ b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn31.c @@ -466,6 +466,7 @@ void dmub_dcn31_get_diagnostic_data(struct dmub_srv *dmub, struct dmub_diagnosti REG_GET(DMCUB_REGION3_CW6_TOP_ADDRESS, DMCUB_REGION3_CW6_ENABLE, &is_cw6_enabled); diag_data->is_cw6_enabled = is_cw6_enabled; + diag_data->timeout_info = dmub->debug; } bool dmub_dcn31_should_detect(struct dmub_srv *dmub) diff --git a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn32.c b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn32.c index 0d521eeda050..e1da270502cc 100644 --- a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn32.c +++ b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn32.c @@ -478,6 +478,8 @@ void dmub_dcn32_get_diagnostic_data(struct dmub_srv *dmub, struct dmub_diagnosti diag_data->is_cw6_enabled = is_cw6_enabled; diag_data->gpint_datain0 = REG_READ(DMCUB_GPINT_DATAIN0); + + diag_data->timeout_info = dmub->debug; } void dmub_dcn32_configure_dmub_in_system_memory(struct dmub_srv *dmub) { diff --git a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn35.c b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn35.c index 53f359f3fae2..70e63aeb8f89 100644 --- a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn35.c +++ b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn35.c @@ -420,6 +420,7 @@ void dmub_dcn35_enable_dmub_boot_options(struct dmub_srv *dmub, const struct dmu boot_options.bits.disable_clk_ds = params->disallow_dispclk_dppclk_ds; boot_options.bits.disable_clk_gate = params->disable_clock_gate; boot_options.bits.ips_disable = params->disable_ips; + boot_options.bits.ips_sequential_ono = params->ips_sequential_ono; REG_WRITE(DMCUB_SCRATCH14, boot_options.all); } @@ -516,6 +517,7 @@ void dmub_dcn35_get_diagnostic_data(struct dmub_srv *dmub, struct dmub_diagnosti diag_data->is_cw6_enabled = is_cw6_enabled; diag_data->gpint_datain0 = REG_READ(DMCUB_GPINT_DATAIN0); + diag_data->timeout_info = dmub->debug; } void dmub_dcn35_configure_dmub_in_system_memory(struct dmub_srv *dmub) { diff --git a/drivers/gpu/drm/amd/display/include/dal_types.h b/drivers/gpu/drm/amd/display/include/dal_types.h index 1c6f24cb1d2f..447768dec887 100644 --- a/drivers/gpu/drm/amd/display/include/dal_types.h +++ b/drivers/gpu/drm/amd/display/include/dal_types.h @@ -27,7 +27,6 @@ #define __DAL_TYPES_H__ #include "signal_types.h" -#include "dc_types.h" struct dal_logger; struct dc_bios; diff --git a/drivers/gpu/drm/amd/display/include/grph_object_id.h b/drivers/gpu/drm/amd/display/include/grph_object_id.h index c6bbd262f1ac..08ee0350b31f 100644 --- a/drivers/gpu/drm/amd/display/include/grph_object_id.h +++ b/drivers/gpu/drm/amd/display/include/grph_object_id.h @@ -226,8 +226,8 @@ enum dp_alt_mode { struct graphics_object_id { uint32_t id:8; - uint32_t enum_id:4; - uint32_t type:4; + enum object_enum_id enum_id; + enum object_type type; uint32_t reserved:16; /* for padding. total size should be u32 */ }; diff --git a/drivers/gpu/drm/amd/display/include/link_service_types.h b/drivers/gpu/drm/amd/display/include/link_service_types.h index 92dbff22a7c6..1867aac57cf2 100644 --- a/drivers/gpu/drm/amd/display/include/link_service_types.h +++ b/drivers/gpu/drm/amd/display/include/link_service_types.h @@ -73,7 +73,6 @@ struct link_training_settings { enum dc_pre_emphasis *pre_emphasis; enum dc_post_cursor2 *post_cursor2; bool should_set_fec_ready; - /* TODO - factor lane_settings out because it changes during LT */ union dc_dp_ffe_preset *ffe_preset; uint16_t cr_pattern_time; diff --git a/drivers/gpu/drm/amd/display/include/logger_types.h b/drivers/gpu/drm/amd/display/include/logger_types.h index f39e2785e618..83479951732a 100644 --- a/drivers/gpu/drm/amd/display/include/logger_types.h +++ b/drivers/gpu/drm/amd/display/include/logger_types.h @@ -64,6 +64,7 @@ #define DC_LOG_DWB(...) drm_dbg((DC_LOGGER)->dev, __VA_ARGS__) #define DC_LOG_DP2(...) drm_dbg_dp((DC_LOGGER)->dev, __VA_ARGS__) #define DC_LOG_AUTO_DPM_TEST(...) pr_debug("[AutoDPMTest]: "__VA_ARGS__) +#define DC_LOG_IPS(...) pr_debug("[IPS]: "__VA_ARGS__) struct dc_log_buffer_ctx { char *buf; diff --git a/drivers/gpu/drm/amd/display/include/signal_types.h b/drivers/gpu/drm/amd/display/include/signal_types.h index 1b14b17a79c7..a10d6b988aab 100644 --- a/drivers/gpu/drm/amd/display/include/signal_types.h +++ b/drivers/gpu/drm/amd/display/include/signal_types.h @@ -118,6 +118,19 @@ static inline bool dc_is_dvi_signal(enum signal_type signal) } } +static inline bool dc_is_tmds_signal(enum signal_type signal) +{ + switch (signal) { + case SIGNAL_TYPE_DVI_SINGLE_LINK: + case SIGNAL_TYPE_DVI_DUAL_LINK: + case SIGNAL_TYPE_HDMI_TYPE_A: + return true; + break; + default: + return false; + } +} + static inline bool dc_is_dvi_single_link_signal(enum signal_type signal) { return (signal == SIGNAL_TYPE_DVI_SINGLE_LINK); diff --git a/drivers/gpu/drm/amd/display/modules/color/color_gamma.c b/drivers/gpu/drm/amd/display/modules/color/color_gamma.c index 8b5c27857671..3699e633801d 100644 --- a/drivers/gpu/drm/amd/display/modules/color/color_gamma.c +++ b/drivers/gpu/drm/amd/display/modules/color/color_gamma.c @@ -1059,7 +1059,7 @@ static bool build_freesync_hdr(struct pwl_float_data_ex *rgb_regamma, struct fixed31_32 min_display; struct fixed31_32 max_content; struct fixed31_32 clip = dc_fixpt_one; - struct fixed31_32 output; + struct fixed31_32 output = dc_fixpt_zero; bool use_eetf = false; bool is_clipped = false; struct fixed31_32 sdr_white_level; diff --git a/drivers/gpu/drm/amd/display/modules/freesync/freesync.c b/drivers/gpu/drm/amd/display/modules/freesync/freesync.c index 3955b7e4b2e2..d09627c15b9c 100644 --- a/drivers/gpu/drm/amd/display/modules/freesync/freesync.c +++ b/drivers/gpu/drm/amd/display/modules/freesync/freesync.c @@ -158,13 +158,13 @@ static unsigned int calc_v_total_from_duration( if (duration_in_us > vrr->max_duration_in_us) duration_in_us = vrr->max_duration_in_us; - if (dc_is_hdmi_signal(stream->signal)) { + if (dc_is_hdmi_signal(stream->signal)) { // change for HDMI to comply with spec uint32_t h_total_up_scaled; h_total_up_scaled = stream->timing.h_total * 10000; v_total = div_u64((unsigned long long)duration_in_us * stream->timing.pix_clk_100hz + (h_total_up_scaled - 1), - h_total_up_scaled); + h_total_up_scaled); //ceiling for MMax and MMin for MVRR } else { v_total = div64_u64(div64_u64(((unsigned long long)( duration_in_us) * (stream->timing.pix_clk_100hz / 10)), @@ -1057,7 +1057,7 @@ void mod_freesync_build_vrr_params(struct mod_freesync *mod_freesync, in_out_vrr->fixed_refresh_in_uhz = 0; refresh_range = div_u64(in_out_vrr->max_refresh_in_uhz + 500000, 1000000) - -+ div_u64(in_out_vrr->min_refresh_in_uhz + 500000, 1000000); + div_u64(in_out_vrr->min_refresh_in_uhz + 500000, 1000000); in_out_vrr->supported = true; } @@ -1126,6 +1126,8 @@ void mod_freesync_build_vrr_params(struct mod_freesync *mod_freesync, in_out_vrr->adjust.v_total_min = stream->timing.v_total; in_out_vrr->adjust.v_total_max = stream->timing.v_total; } + + in_out_vrr->adjust.allow_otg_v_count_halt = (in_config->state == VRR_STATE_ACTIVE_FIXED) ? true : false; } void mod_freesync_handle_preflip(struct mod_freesync *mod_freesync, diff --git a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp2_execution.c b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp2_execution.c index 733f22bed021..c996365e84b0 100644 --- a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp2_execution.c +++ b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp2_execution.c @@ -151,7 +151,7 @@ out: static enum mod_hdcp_status poll_l_prime_available(struct mod_hdcp *hdcp) { - enum mod_hdcp_status status; + enum mod_hdcp_status status = MOD_HDCP_STATUS_FAILURE; uint8_t size; uint16_t max_wait = 20; // units of ms uint16_t num_polls = 5; diff --git a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_ddc.c b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_ddc.c index f7b5583ee609..8e9caae7c955 100644 --- a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_ddc.c +++ b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_ddc.c @@ -156,6 +156,10 @@ static enum mod_hdcp_status read(struct mod_hdcp *hdcp, uint32_t cur_size = 0; uint32_t data_offset = 0; + if (msg_id == MOD_HDCP_MESSAGE_ID_INVALID) { + return MOD_HDCP_STATUS_DDC_FAILURE; + } + if (is_dp_hdcp(hdcp)) { while (buf_len > 0) { cur_size = MIN(buf_len, HDCP_MAX_AUX_TRANSACTION_SIZE); @@ -215,6 +219,10 @@ static enum mod_hdcp_status write(struct mod_hdcp *hdcp, uint32_t cur_size = 0; uint32_t data_offset = 0; + if (msg_id == MOD_HDCP_MESSAGE_ID_INVALID) { + return MOD_HDCP_STATUS_DDC_FAILURE; + } + if (is_dp_hdcp(hdcp)) { while (buf_len > 0) { cur_size = MIN(buf_len, HDCP_MAX_AUX_TRANSACTION_SIZE); 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 738ee763f24a..a344e2e49b0e 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 @@ -147,15 +147,12 @@ void mod_build_vsc_infopacket(const struct dc_stream_state *stream, } /* VSC packet set to 4 for PSR-SU, or 2 for PSR1 */ - if (stream->link->psr_settings.psr_feature_enabled) { - if (stream->link->psr_settings.psr_version == DC_PSR_VERSION_SU_1) - vsc_packet_revision = vsc_packet_rev4; - else if (stream->link->psr_settings.psr_version == DC_PSR_VERSION_1) - vsc_packet_revision = vsc_packet_rev2; - } - - if (stream->link->replay_settings.config.replay_supported) + if (stream->link->psr_settings.psr_version == DC_PSR_VERSION_SU_1) + vsc_packet_revision = vsc_packet_rev4; + else if (stream->link->replay_settings.config.replay_supported) vsc_packet_revision = vsc_packet_rev4; + else if (stream->link->psr_settings.psr_version == DC_PSR_VERSION_1) + vsc_packet_revision = vsc_packet_rev2; /* Update to revision 5 for extended colorimetry support */ if (stream->use_vsc_sdp_for_colorimetry) @@ -539,8 +536,6 @@ void mod_build_adaptive_sync_infopacket(const struct dc_stream_state *stream, mod_build_adaptive_sync_infopacket_v2(stream, param, info_packet); break; case FREESYNC_TYPE_PCON_IN_WHITELIST: - mod_build_adaptive_sync_infopacket_v1(info_packet); - break; case ADAPTIVE_SYNC_TYPE_EDP: mod_build_adaptive_sync_infopacket_v1(info_packet); break; diff --git a/drivers/gpu/drm/amd/include/amd_shared.h b/drivers/gpu/drm/amd/include/amd_shared.h index b0a6256e89f4..7536c173a546 100644 --- a/drivers/gpu/drm/amd/include/amd_shared.h +++ b/drivers/gpu/drm/amd/include/amd_shared.h @@ -24,6 +24,7 @@ #define __AMD_SHARED_H__ #include <drm/amd_asic_type.h> +#include <drm/drm_print.h> #define AMD_MAX_USEC_TIMEOUT 1000000 /* 1000 ms */ @@ -321,6 +322,8 @@ struct amd_ip_funcs { int (*set_powergating_state)(void *handle, enum amd_powergating_state state); void (*get_clockgating_state)(void *handle, u64 *flags); + void (*dump_ip_state)(void *handle); + void (*print_ip_state)(void *handle, struct drm_printer *p); }; diff --git a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_0_0_offset.h b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_0_0_offset.h index f2f8f9b39c6b..fc72c2267060 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_0_0_offset.h +++ b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_0_0_offset.h @@ -311,6 +311,10 @@ #define mmPHYESYMCLK_CLOCK_CNTL_BASE_IDX 2 #define mmPHYFSYMCLK_CLOCK_CNTL 0x0057 #define mmPHYFSYMCLK_CLOCK_CNTL_BASE_IDX 2 +#define regHDMICHARCLK0_CLOCK_CNTL 0x004a +#define regHDMICHARCLK0_CLOCK_CNTL_BASE_IDX 2 +#define mmHDMICHARCLK0_CLOCK_CNTL 0x004a +#define mmHDMICHARCLK0_CLOCK_CNTL_BASE_IDX 2 // addressBlock: dce_dc_dccg_dccg_dfs_dispdec @@ -4513,6 +4517,10 @@ #define mmCM0_CM_3DLUT_OUT_OFFSET_G_BASE_IDX 2 #define mmCM0_CM_3DLUT_OUT_OFFSET_B 0x0e18 #define mmCM0_CM_3DLUT_OUT_OFFSET_B_BASE_IDX 2 +#define mmCM0_CM_TEST_DEBUG_INDEX 0x0e19 +#define mmCM0_CM_TEST_DEBUG_INDEX_BASE_IDX 2 +#define mmCM0_CM_TEST_DEBUG_DATA 0x0e1a +#define mmCM0_CM_TEST_DEBUG_DATA_BASE_IDX 2 // addressBlock: dce_dc_dpp0_dispdec_dpp_dcperfmon_dc_perfmon_dispdec @@ -5201,6 +5209,10 @@ #define mmCM1_CM_3DLUT_OUT_OFFSET_G_BASE_IDX 2 #define mmCM1_CM_3DLUT_OUT_OFFSET_B 0x0f83 #define mmCM1_CM_3DLUT_OUT_OFFSET_B_BASE_IDX 2 +#define mmCM1_CM_TEST_DEBUG_INDEX 0x0f84 +#define mmCM1_CM_TEST_DEBUG_INDEX_BASE_IDX 2 +#define mmCM1_CM_TEST_DEBUG_DATA 0x0f85 +#define mmCM1_CM_TEST_DEBUG_DATA_BASE_IDX 2 // addressBlock: dce_dc_dpp1_dispdec_dpp_dcperfmon_dc_perfmon_dispdec @@ -5888,6 +5900,10 @@ #define mmCM2_CM_3DLUT_OUT_OFFSET_G_BASE_IDX 2 #define mmCM2_CM_3DLUT_OUT_OFFSET_B 0x10ee #define mmCM2_CM_3DLUT_OUT_OFFSET_B_BASE_IDX 2 +#define mmCM2_CM_TEST_DEBUG_INDEX 0x10ef +#define mmCM2_CM_TEST_DEBUG_INDEX_BASE_IDX 2 +#define mmCM2_CM_TEST_DEBUG_DATA 0x10f0 +#define mmCM2_CM_TEST_DEBUG_DATA_BASE_IDX 2 // addressBlock: dce_dc_dpp2_dispdec_dpp_dcperfmon_dc_perfmon_dispdec @@ -6576,6 +6592,10 @@ #define mmCM3_CM_3DLUT_OUT_OFFSET_G_BASE_IDX 2 #define mmCM3_CM_3DLUT_OUT_OFFSET_B 0x1259 #define mmCM3_CM_3DLUT_OUT_OFFSET_B_BASE_IDX 2 +#define mmCM3_CM_TEST_DEBUG_INDEX 0x125a +#define mmCM3_CM_TEST_DEBUG_INDEX_BASE_IDX 2 +#define mmCM3_CM_TEST_DEBUG_DATA 0x125b +#define mmCM3_CM_TEST_DEBUG_DATA_BASE_IDX 2 // addressBlock: dce_dc_dpp3_dispdec_dpp_dcperfmon_dc_perfmon_dispdec @@ -7264,6 +7284,10 @@ #define mmCM4_CM_3DLUT_OUT_OFFSET_G_BASE_IDX 2 #define mmCM4_CM_3DLUT_OUT_OFFSET_B 0x13c4 #define mmCM4_CM_3DLUT_OUT_OFFSET_B_BASE_IDX 2 +#define mmCM4_CM_TEST_DEBUG_INDEX 0x13c5 +#define mmCM4_CM_TEST_DEBUG_INDEX_BASE_IDX 2 +#define mmCM4_CM_TEST_DEBUG_DATA 0x13c6 +#define mmCM4_CM_TEST_DEBUG_DATA_BASE_IDX 2 // addressBlock: dce_dc_dpp4_dispdec_dpp_dcperfmon_dc_perfmon_dispdec @@ -7952,6 +7976,10 @@ #define mmCM5_CM_3DLUT_OUT_OFFSET_G_BASE_IDX 2 #define mmCM5_CM_3DLUT_OUT_OFFSET_B 0x152f #define mmCM5_CM_3DLUT_OUT_OFFSET_B_BASE_IDX 2 +#define mmCM5_CM_TEST_DEBUG_INDEX 0x1530 +#define mmCM5_CM_TEST_DEBUG_INDEX_BASE_IDX 2 +#define mmCM5_CM_TEST_DEBUG_DATA 0x1531 +#define mmCM5_CM_TEST_DEBUG_DATA_BASE_IDX 2 // addressBlock: dce_dc_dpp5_dispdec_dpp_dcperfmon_dc_perfmon_dispdec diff --git a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_0_0_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_0_0_sh_mask.h index e0a447351623..daf71e82f0ba 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_0_0_sh_mask.h +++ b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_0_0_sh_mask.h @@ -1189,6 +1189,11 @@ #define PHYFSYMCLK_CLOCK_CNTL__PHYFSYMCLK_FORCE_SRC_SEL__SHIFT 0x4 #define PHYFSYMCLK_CLOCK_CNTL__PHYFSYMCLK_FORCE_EN_MASK 0x00000001L #define PHYFSYMCLK_CLOCK_CNTL__PHYFSYMCLK_FORCE_SRC_SEL_MASK 0x00000010L +//HDMICHARCLK0_CLOCK_CNTL +#define HDMICHARCLK0_CLOCK_CNTL__HDMICHARCLK0_EN__SHIFT 0x0 +#define HDMICHARCLK0_CLOCK_CNTL__HDMICHARCLK0_SRC_SEL__SHIFT 0x4 +#define HDMICHARCLK0_CLOCK_CNTL__HDMICHARCLK0_EN_MASK 0x00000001L +#define HDMICHARCLK0_CLOCK_CNTL__HDMICHARCLK0_SRC_SEL_MASK 0x00000070L // addressBlock: dce_dc_dccg_dccg_dfs_dispdec @@ -16739,6 +16744,15 @@ #define CM0_CM_3DLUT_OUT_OFFSET_B__CM_3DLUT_OUT_SCALE_B__SHIFT 0x10 #define CM0_CM_3DLUT_OUT_OFFSET_B__CM_3DLUT_OUT_OFFSET_B_MASK 0x0000FFFFL #define CM0_CM_3DLUT_OUT_OFFSET_B__CM_3DLUT_OUT_SCALE_B_MASK 0xFFFF0000L +//CM0_CM_TEST_DEBUG_INDEX +#define CM0_CM_TEST_DEBUG_INDEX__CM_TEST_DEBUG_INDEX__SHIFT 0x0 +#define CM0_CM_TEST_DEBUG_INDEX__CM_TEST_DEBUG_WRITE_EN__SHIFT 0x8 +#define CM0_CM_TEST_DEBUG_INDEX__CM_TEST_DEBUG_INDEX_MASK 0x000000FFL +#define CM0_CM_TEST_DEBUG_INDEX__CM_TEST_DEBUG_WRITE_EN_MASK 0x00000100L +//CM0_CM_TEST_DEBUG_DATA +#define CM0_CM_TEST_DEBUG_DATA__CM_TEST_DEBUG_DATA__SHIFT 0x0 +#define CM0_CM_TEST_DEBUG_DATA__CM_TEST_DEBUG_DATA_MASK 0xFFFFFFFFL + // addressBlock: dce_dc_dpp0_dispdec_dpp_dcperfmon_dc_perfmon_dispdec //DC_PERFMON12_PERFCOUNTER_CNTL diff --git a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_0_2_offset.h b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_0_2_offset.h index b45a35aae241..bf84f97d9162 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_0_2_offset.h +++ b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_0_2_offset.h @@ -4466,6 +4466,10 @@ #define mmCM0_CM_3DLUT_OUT_OFFSET_G_BASE_IDX 2 #define mmCM0_CM_3DLUT_OUT_OFFSET_B 0x0e18 #define mmCM0_CM_3DLUT_OUT_OFFSET_B_BASE_IDX 2 +#define mmCM0_CM_TEST_DEBUG_INDEX 0x0e19 +#define mmCM0_CM_TEST_DEBUG_INDEX_BASE_IDX 2 +#define mmCM0_CM_TEST_DEBUG_DATA 0x0e1a +#define mmCM0_CM_TEST_DEBUG_DATA_BASE_IDX 2 // addressBlock: dce_dc_dpp0_dispdec_dpp_dcperfmon_dc_perfmon_dispdec @@ -5154,6 +5158,10 @@ #define mmCM1_CM_3DLUT_OUT_OFFSET_G_BASE_IDX 2 #define mmCM1_CM_3DLUT_OUT_OFFSET_B 0x0f83 #define mmCM1_CM_3DLUT_OUT_OFFSET_B_BASE_IDX 2 +#define mmCM1_CM_TEST_DEBUG_INDEX 0x0f84 +#define mmCM1_CM_TEST_DEBUG_INDEX_BASE_IDX 2 +#define mmCM1_CM_TEST_DEBUG_DATA 0x0f85 +#define mmCM1_CM_TEST_DEBUG_DATA_BASE_IDX 2 // addressBlock: dce_dc_dpp1_dispdec_dpp_dcperfmon_dc_perfmon_dispdec @@ -5841,6 +5849,10 @@ #define mmCM2_CM_3DLUT_OUT_OFFSET_G_BASE_IDX 2 #define mmCM2_CM_3DLUT_OUT_OFFSET_B 0x10ee #define mmCM2_CM_3DLUT_OUT_OFFSET_B_BASE_IDX 2 +#define mmCM2_CM_TEST_DEBUG_INDEX 0x10ef +#define mmCM2_CM_TEST_DEBUG_INDEX_BASE_IDX 2 +#define mmCM2_CM_TEST_DEBUG_DATA 0x10f0 +#define mmCM2_CM_TEST_DEBUG_DATA_BASE_IDX 2 // addressBlock: dce_dc_dpp2_dispdec_dpp_dcperfmon_dc_perfmon_dispdec @@ -6529,6 +6541,10 @@ #define mmCM3_CM_3DLUT_OUT_OFFSET_G_BASE_IDX 2 #define mmCM3_CM_3DLUT_OUT_OFFSET_B 0x1259 #define mmCM3_CM_3DLUT_OUT_OFFSET_B_BASE_IDX 2 +#define mmCM3_CM_TEST_DEBUG_INDEX 0x125a +#define mmCM3_CM_TEST_DEBUG_INDEX_BASE_IDX 2 +#define mmCM3_CM_TEST_DEBUG_DATA 0x125b +#define mmCM3_CM_TEST_DEBUG_DATA_BASE_IDX 2 // addressBlock: dce_dc_dpp3_dispdec_dpp_dcperfmon_dc_perfmon_dispdec @@ -7217,6 +7233,10 @@ #define mmCM4_CM_3DLUT_OUT_OFFSET_G_BASE_IDX 2 #define mmCM4_CM_3DLUT_OUT_OFFSET_B 0x13c4 #define mmCM4_CM_3DLUT_OUT_OFFSET_B_BASE_IDX 2 +#define mmCM4_CM_TEST_DEBUG_INDEX 0x13c5 +#define mmCM4_CM_TEST_DEBUG_INDEX_BASE_IDX 2 +#define mmCM4_CM_TEST_DEBUG_DATA 0x13c6 +#define mmCM4_CM_TEST_DEBUG_DATA_BASE_IDX 2 // addressBlock: dce_dc_dpp4_dispdec_dpp_dcperfmon_dc_perfmon_dispdec diff --git a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_0_2_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_0_2_sh_mask.h index 3dae29f9581e..56cdb219874a 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_0_2_sh_mask.h +++ b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_0_2_sh_mask.h @@ -15676,6 +15676,14 @@ #define CM0_CM_3DLUT_OUT_OFFSET_B__CM_3DLUT_OUT_SCALE_B__SHIFT 0x10 #define CM0_CM_3DLUT_OUT_OFFSET_B__CM_3DLUT_OUT_OFFSET_B_MASK 0x0000FFFFL #define CM0_CM_3DLUT_OUT_OFFSET_B__CM_3DLUT_OUT_SCALE_B_MASK 0xFFFF0000L +//CM0_CM_TEST_DEBUG_INDEX +#define CM0_CM_TEST_DEBUG_INDEX__CM_TEST_DEBUG_INDEX__SHIFT 0x0 +#define CM0_CM_TEST_DEBUG_INDEX__CM_TEST_DEBUG_WRITE_EN__SHIFT 0x8 +#define CM0_CM_TEST_DEBUG_INDEX__CM_TEST_DEBUG_INDEX_MASK 0x000000FFL +#define CM0_CM_TEST_DEBUG_INDEX__CM_TEST_DEBUG_WRITE_EN_MASK 0x00000100L +//CM0_CM_TEST_DEBUG_DATA +#define CM0_CM_TEST_DEBUG_DATA__CM_TEST_DEBUG_DATA__SHIFT 0x0 +#define CM0_CM_TEST_DEBUG_DATA__CM_TEST_DEBUG_DATA_MASK 0xFFFFFFFFL // addressBlock: dce_dc_dpp0_dispdec_dpp_dcperfmon_dc_perfmon_dispdec diff --git a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_0_3_offset.h b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_0_3_offset.h index daa8130636f0..8b0d2638a6b0 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_0_3_offset.h +++ b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_0_3_offset.h @@ -3110,6 +3110,10 @@ #define mmCM0_CM_3DLUT_OUT_OFFSET_G_BASE_IDX 2 #define mmCM0_CM_3DLUT_OUT_OFFSET_B 0x0e18 #define mmCM0_CM_3DLUT_OUT_OFFSET_B_BASE_IDX 2 +#define mmCM0_CM_TEST_DEBUG_INDEX 0x0e19 +#define mmCM0_CM_TEST_DEBUG_INDEX_BASE_IDX 2 +#define mmCM0_CM_TEST_DEBUG_DATA 0x0e1a +#define mmCM0_CM_TEST_DEBUG_DATA_BASE_IDX 2 // addressBlock: dce_dc_dpp0_dispdec_dpp_dcperfmon_dc_perfmon_dispdec @@ -3798,6 +3802,10 @@ #define mmCM1_CM_3DLUT_OUT_OFFSET_G_BASE_IDX 2 #define mmCM1_CM_3DLUT_OUT_OFFSET_B 0x0f83 #define mmCM1_CM_3DLUT_OUT_OFFSET_B_BASE_IDX 2 +#define mmCM1_CM_TEST_DEBUG_INDEX 0x0f84 +#define mmCM1_CM_TEST_DEBUG_INDEX_BASE_IDX 2 +#define mmCM1_CM_TEST_DEBUG_DATA 0x0f85 +#define mmCM1_CM_TEST_DEBUG_DATA_BASE_IDX 2 // addressBlock: dce_dc_dpp1_dispdec_dpp_dcperfmon_dc_perfmon_dispdec @@ -5687,6 +5695,16 @@ #define mmDSCC0_DSCC_RATE_CONTROL_BUFFER2_MAX_FULLNESS_LEVEL_BASE_IDX 2 #define mmDSCC0_DSCC_RATE_CONTROL_BUFFER3_MAX_FULLNESS_LEVEL 0x3035 #define mmDSCC0_DSCC_RATE_CONTROL_BUFFER3_MAX_FULLNESS_LEVEL_BASE_IDX 2 +#define mmDSCC0_DSCC_TEST_DEBUG_BUS_ROTATE 0x303a +#define mmDSCC0_DSCC_TEST_DEBUG_BUS_ROTATE_BASE_IDX 2 +#define mmDSCC0_DSCC_TEST_DEBUG_DATA0 0x303b +#define mmDSCC0_DSCC_TEST_DEBUG_DATA0_BASE_IDX 2 +#define mmDSCC0_DSCC_TEST_DEBUG_DATA1 0x303c +#define mmDSCC0_DSCC_TEST_DEBUG_DATA1_BASE_IDX 2 +#define mmDSCC0_DSCC_TEST_DEBUG_DATA2 0x303d +#define mmDSCC0_DSCC_TEST_DEBUG_DATA2_BASE_IDX 2 +#define mmDSCC0_DSCC_TEST_DEBUG_DATA3 0x303e +#define mmDSCC0_DSCC_TEST_DEBUG_DATA3_BASE_IDX 2 // addressBlock: dce_dc_dsc0_dispdec_dsc_dcperfmon_dc_perfmon_dispdec @@ -5817,6 +5835,16 @@ #define mmDSCC1_DSCC_RATE_CONTROL_BUFFER2_MAX_FULLNESS_LEVEL_BASE_IDX 2 #define mmDSCC1_DSCC_RATE_CONTROL_BUFFER3_MAX_FULLNESS_LEVEL 0x3091 #define mmDSCC1_DSCC_RATE_CONTROL_BUFFER3_MAX_FULLNESS_LEVEL_BASE_IDX 2 +#define mmDSCC1_DSCC_TEST_DEBUG_BUS_ROTATE 0x3096 +#define mmDSCC1_DSCC_TEST_DEBUG_BUS_ROTATE_BASE_IDX 2 +#define mmDSCC1_DSCC_TEST_DEBUG_DATA0 0x3097 +#define mmDSCC1_DSCC_TEST_DEBUG_DATA0_BASE_IDX 2 +#define mmDSCC1_DSCC_TEST_DEBUG_DATA1 0x3098 +#define mmDSCC1_DSCC_TEST_DEBUG_DATA1_BASE_IDX 2 +#define mmDSCC1_DSCC_TEST_DEBUG_DATA2 0x3099 +#define mmDSCC1_DSCC_TEST_DEBUG_DATA2_BASE_IDX 2 +#define mmDSCC1_DSCC_TEST_DEBUG_DATA3 0x309a +#define mmDSCC1_DSCC_TEST_DEBUG_DATA3_BASE_IDX 2 // addressBlock: dce_dc_dsc1_dispdec_dsc_dcperfmon_dc_perfmon_dispdec diff --git a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_0_3_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_0_3_sh_mask.h index 5c469cf635e5..53f1705f8d99 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_0_3_sh_mask.h +++ b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_0_3_sh_mask.h @@ -10701,6 +10701,13 @@ #define CM0_CM_COEF_FORMAT__CM_BIAS_FORMAT_MASK 0x00000001L #define CM0_CM_COEF_FORMAT__CM_POST_CSC_COEF_FORMAT_MASK 0x00000010L #define CM0_CM_COEF_FORMAT__CM_GAMUT_REMAP_COEF_FORMAT_MASK 0x00000100L + +//CM0_CM_TEST_DEBUG_INDEX +#define CM0_CM_TEST_DEBUG_INDEX__CM_TEST_DEBUG_INDEX__SHIFT 0x0 +#define CM0_CM_TEST_DEBUG_INDEX__CM_TEST_DEBUG_WRITE_EN__SHIFT 0x8 +#define CM0_CM_TEST_DEBUG_INDEX__CM_TEST_DEBUG_INDEX_MASK 0x000000FFL +#define CM0_CM_TEST_DEBUG_INDEX__CM_TEST_DEBUG_WRITE_EN_MASK 0x00000100L + //CM0_CM_SHAPER_CONTROL #define CM0_CM_SHAPER_CONTROL__CM_SHAPER_LUT_MODE__SHIFT 0x0 #define CM0_CM_SHAPER_CONTROL__CM_SHAPER_MODE_CURRENT__SHIFT 0x2 @@ -22258,7 +22265,9 @@ #define DSC_TOP0_DSC_TOP_CONTROL__DSC_DSCCLK_R_GATE_DIS_MASK 0x00000100L //DSC_TOP0_DSC_DEBUG_CONTROL #define DSC_TOP0_DSC_DEBUG_CONTROL__DSC_DBG_EN__SHIFT 0x0 +#define DSC_TOP0_DSC_DEBUG_CONTROL__DSC_TEST_CLOCK_MUX_SEL__SHIFT 0x4 #define DSC_TOP0_DSC_DEBUG_CONTROL__DSC_DBG_EN_MASK 0x00000001L +#define DSC_TOP0_DSC_DEBUG_CONTROL__DSC_TEST_CLOCK_MUX_SEL_MASK 0x00000070L // addressBlock: dce_dc_dsc0_dispdec_dsccif_dispdec @@ -22631,6 +22640,15 @@ //DSCC0_DSCC_RATE_CONTROL_BUFFER3_MAX_FULLNESS_LEVEL #define DSCC0_DSCC_RATE_CONTROL_BUFFER3_MAX_FULLNESS_LEVEL__DSCC_RATE_CONTROL_BUFFER3_MAX_FULLNESS_LEVEL__SHIFT 0x0 #define DSCC0_DSCC_RATE_CONTROL_BUFFER3_MAX_FULLNESS_LEVEL__DSCC_RATE_CONTROL_BUFFER3_MAX_FULLNESS_LEVEL_MASK 0x0003FFFFL +//DSCC0_DSCC_TEST_DEBUG_BUS_ROTATE +#define DSCC0_DSCC_TEST_DEBUG_BUS_ROTATE__DSCC_TEST_DEBUG_BUS0_ROTATE__SHIFT 0x0 +#define DSCC0_DSCC_TEST_DEBUG_BUS_ROTATE__DSCC_TEST_DEBUG_BUS1_ROTATE__SHIFT 0x8 +#define DSCC0_DSCC_TEST_DEBUG_BUS_ROTATE__DSCC_TEST_DEBUG_BUS2_ROTATE__SHIFT 0x10 +#define DSCC0_DSCC_TEST_DEBUG_BUS_ROTATE__DSCC_TEST_DEBUG_BUS3_ROTATE__SHIFT 0x18 +#define DSCC0_DSCC_TEST_DEBUG_BUS_ROTATE__DSCC_TEST_DEBUG_BUS0_ROTATE_MASK 0x0000001FL +#define DSCC0_DSCC_TEST_DEBUG_BUS_ROTATE__DSCC_TEST_DEBUG_BUS1_ROTATE_MASK 0x00001F00L +#define DSCC0_DSCC_TEST_DEBUG_BUS_ROTATE__DSCC_TEST_DEBUG_BUS2_ROTATE_MASK 0x001F0000L +#define DSCC0_DSCC_TEST_DEBUG_BUS_ROTATE__DSCC_TEST_DEBUG_BUS3_ROTATE_MASK 0x1F000000L // addressBlock: dce_dc_dsc0_dispdec_dsc_dcperfmon_dc_perfmon_dispdec diff --git a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_1_2_offset.h b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_1_2_offset.h index f268d33c4744..7fd906f10803 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_1_2_offset.h +++ b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_1_2_offset.h @@ -424,6 +424,8 @@ #define regDTBCLK_DTO2_MODULO_BASE_IDX 2 #define regDTBCLK_DTO3_MODULO 0x0022 #define regDTBCLK_DTO3_MODULO_BASE_IDX 2 +#define regHDMICHARCLK0_CLOCK_CNTL 0x004a +#define regHDMICHARCLK0_CLOCK_CNTL_BASE_IDX 2 #define regPHYASYMCLK_CLOCK_CNTL 0x0052 #define regPHYASYMCLK_CLOCK_CNTL_BASE_IDX 2 #define regPHYBSYMCLK_CLOCK_CNTL 0x0053 @@ -434,6 +436,8 @@ #define regPHYDSYMCLK_CLOCK_CNTL_BASE_IDX 2 #define regPHYESYMCLK_CLOCK_CNTL 0x0056 #define regPHYESYMCLK_CLOCK_CNTL_BASE_IDX 2 +#define regHDMISTREAMCLK_CNTL 0x0059 +#define regHDMISTREAMCLK_CNTL_BASE_IDX 2 #define regDCCG_GATE_DISABLE_CNTL3 0x005a #define regDCCG_GATE_DISABLE_CNTL3_BASE_IDX 2 #define regHDMISTREAMCLK0_DTO_PARAM 0x005b diff --git a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_1_2_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_1_2_sh_mask.h index cf3398f15666..07fbfafe6056 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_1_2_sh_mask.h +++ b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_1_2_sh_mask.h @@ -1372,6 +1372,11 @@ //DTBCLK_DTO3_MODULO #define DTBCLK_DTO3_MODULO__DTBCLK_DTO3_MODULO__SHIFT 0x0 #define DTBCLK_DTO3_MODULO__DTBCLK_DTO3_MODULO_MASK 0xFFFFFFFFL +//HDMICHARCLK0_CLOCK_CNTL +#define HDMICHARCLK0_CLOCK_CNTL__HDMICHARCLK0_EN__SHIFT 0x0 +#define HDMICHARCLK0_CLOCK_CNTL__HDMICHARCLK0_SRC_SEL__SHIFT 0x4 +#define HDMICHARCLK0_CLOCK_CNTL__HDMICHARCLK0_EN_MASK 0x00000001L +#define HDMICHARCLK0_CLOCK_CNTL__HDMICHARCLK0_SRC_SEL_MASK 0x00000070L //PHYASYMCLK_CLOCK_CNTL #define PHYASYMCLK_CLOCK_CNTL__PHYASYMCLK_FORCE_EN__SHIFT 0x0 #define PHYASYMCLK_CLOCK_CNTL__PHYASYMCLK_FORCE_SRC_SEL__SHIFT 0x4 @@ -1397,6 +1402,13 @@ #define PHYESYMCLK_CLOCK_CNTL__PHYESYMCLK_FORCE_SRC_SEL__SHIFT 0x4 #define PHYESYMCLK_CLOCK_CNTL__PHYESYMCLK_FORCE_EN_MASK 0x00000001L #define PHYESYMCLK_CLOCK_CNTL__PHYESYMCLK_FORCE_SRC_SEL_MASK 0x00000030L +//HDMISTREAMCLK_CNTL +#define HDMISTREAMCLK_CNTL__HDMISTREAMCLK0_SRC_SEL__SHIFT 0x0 +#define HDMISTREAMCLK_CNTL__HDMISTREAMCLK0_EN__SHIFT 0x3 +#define HDMISTREAMCLK_CNTL__HDMISTREAMCLK0_DTO_FORCE_DIS__SHIFT 0x4 +#define HDMISTREAMCLK_CNTL__HDMISTREAMCLK0_SRC_SEL_MASK 0x00000007L +#define HDMISTREAMCLK_CNTL__HDMISTREAMCLK0_EN_MASK 0x00000008L +#define HDMISTREAMCLK_CNTL__HDMISTREAMCLK0_DTO_FORCE_DIS_MASK 0x00000010L //DCCG_GATE_DISABLE_CNTL3 #define DCCG_GATE_DISABLE_CNTL3__HDMISTREAMCLK0_GATE_DISABLE__SHIFT 0x0 #define DCCG_GATE_DISABLE_CNTL3__HDMISTREAMCLK1_GATE_DISABLE__SHIFT 0x1 @@ -46978,6 +46990,13 @@ #define DSC_TOP0_DSC_TOP_CONTROL__DSC_CLOCK_EN_MASK 0x00000001L #define DSC_TOP0_DSC_TOP_CONTROL__DSC_DISPCLK_R_GATE_DIS_MASK 0x00000010L #define DSC_TOP0_DSC_TOP_CONTROL__DSC_DSCCLK_R_GATE_DIS_MASK 0x00000100L + + +//DSC_TOP0_DSC_DEBUG_CONTROL +#define DSC_TOP0_DSC_DEBUG_CONTROL__DSC_DBG_EN__SHIFT 0x0 +#define DSC_TOP0_DSC_DEBUG_CONTROL__DSC_TEST_CLOCK_MUX_SEL__SHIFT 0x4 +#define DSC_TOP0_DSC_DEBUG_CONTROL__DSC_DBG_EN_MASK 0x00000001L +#define DSC_TOP0_DSC_DEBUG_CONTROL__DSC_TEST_CLOCK_MUX_SEL_MASK 0x00000070L //DSC_TOP0_DSC_DEBUG_CONTROL #define DSC_TOP0_DSC_DEBUG_CONTROL__DSC_DBG_EN__SHIFT 0x0 #define DSC_TOP0_DSC_DEBUG_CONTROL__DSC_TEST_CLOCK_MUX_SEL__SHIFT 0x4 diff --git a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_1_5_offset.h b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_1_5_offset.h index 50c34d88c17c..16a69d17bb1e 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_1_5_offset.h +++ b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_1_5_offset.h @@ -213,6 +213,8 @@ #define regDTBCLK_DTO2_MODULO_BASE_IDX 2 #define regDTBCLK_DTO3_MODULO 0x0022 #define regDTBCLK_DTO3_MODULO_BASE_IDX 2 +#define regHDMICHARCLK0_CLOCK_CNTL 0x004a +#define regHDMICHARCLK0_CLOCK_CNTL_BASE_IDX 2 #define regPHYASYMCLK_CLOCK_CNTL 0x0052 #define regPHYASYMCLK_CLOCK_CNTL_BASE_IDX 2 #define regPHYBSYMCLK_CLOCK_CNTL 0x0053 @@ -233,6 +235,8 @@ #define regDCCG_AUDIO_DTBCLK_DTO_MODULO_BASE_IDX 2 #define regDTBCLK_DTO_DBUF_EN 0x0063 #define regDTBCLK_DTO_DBUF_EN_BASE_IDX 2 +#define regHDMISTREAMCLK_CNTL 0x0059 +#define regHDMISTREAMCLK_CNTL_BASE_IDX 2 // addressBlock: dce_dc_dccg_dccg_dcperfmon0_dc_perfmon_dispdec diff --git a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_1_5_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_1_5_sh_mask.h index 295e0dac9ffa..6473362e39a8 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_1_5_sh_mask.h +++ b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_1_5_sh_mask.h @@ -886,6 +886,11 @@ //DTBCLK_DTO3_MODULO #define DTBCLK_DTO3_MODULO__DTBCLK_DTO3_MODULO__SHIFT 0x0 #define DTBCLK_DTO3_MODULO__DTBCLK_DTO3_MODULO_MASK 0xFFFFFFFFL +//HDMICHARCLK0_CLOCK_CNTL +#define HDMICHARCLK0_CLOCK_CNTL__HDMICHARCLK0_EN__SHIFT 0x0 +#define HDMICHARCLK0_CLOCK_CNTL__HDMICHARCLK0_SRC_SEL__SHIFT 0x4 +#define HDMICHARCLK0_CLOCK_CNTL__HDMICHARCLK0_EN_MASK 0x00000001L +#define HDMICHARCLK0_CLOCK_CNTL__HDMICHARCLK0_SRC_SEL_MASK 0x00000070L //PHYASYMCLK_CLOCK_CNTL #define PHYASYMCLK_CLOCK_CNTL__PHYASYMCLK_FORCE_EN__SHIFT 0x0 #define PHYASYMCLK_CLOCK_CNTL__PHYASYMCLK_FORCE_SRC_SEL__SHIFT 0x4 @@ -911,6 +916,11 @@ #define PHYESYMCLK_CLOCK_CNTL__PHYESYMCLK_FORCE_SRC_SEL__SHIFT 0x4 #define PHYESYMCLK_CLOCK_CNTL__PHYESYMCLK_FORCE_EN_MASK 0x00000001L #define PHYESYMCLK_CLOCK_CNTL__PHYESYMCLK_FORCE_SRC_SEL_MASK 0x00000030L +//HDMISTREAMCLK_CNTL +#define HDMISTREAMCLK_CNTL__HDMISTREAMCLK0_SRC_SEL__SHIFT 0x0 +#define HDMISTREAMCLK_CNTL__HDMISTREAMCLK0_DTO_FORCE_DIS__SHIFT 0x10 +#define HDMISTREAMCLK_CNTL__HDMISTREAMCLK0_SRC_SEL_MASK 0x00000003L +#define HDMISTREAMCLK_CNTL__HDMISTREAMCLK0_DTO_FORCE_DIS_MASK 0x00010000L //DCCG_GATE_DISABLE_CNTL3 #define DCCG_GATE_DISABLE_CNTL3__HDMISTREAMCLK0_GATE_DISABLE__SHIFT 0x0 #define DCCG_GATE_DISABLE_CNTL3__HDMISTREAMCLK1_GATE_DISABLE__SHIFT 0x1 diff --git a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_2_0_offset.h b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_2_0_offset.h index 14c29ce4c7b3..78cb61d5800a 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_2_0_offset.h +++ b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_2_0_offset.h @@ -1719,6 +1719,10 @@ #define regDCHUBBUB_TIMEOUT_INTERRUPT_STATUS_BASE_IDX 2 #define regFMON_CTRL 0x0541 #define regFMON_CTRL_BASE_IDX 2 +#define regDCHUBBUB_TEST_DEBUG_INDEX 0x0542 +#define regDCHUBBUB_TEST_DEBUG_INDEX_BASE_IDX 2 +#define regDCHUBBUB_TEST_DEBUG_DATA 0x0543 +#define regDCHUBBUB_TEST_DEBUG_DATA_BASE_IDX 2 // addressBlock: dcn_dc_dchubbubl_hubbub_sdpif_dispdec @@ -3574,6 +3578,10 @@ #define regCM0_CM_DEALPHA_BASE_IDX 2 #define regCM0_CM_COEF_FORMAT 0x0d8c #define regCM0_CM_COEF_FORMAT_BASE_IDX 2 +#define regCM0_CM_TEST_DEBUG_INDEX 0x0d8d +#define regCM0_CM_TEST_DEBUG_INDEX_BASE_IDX 2 +#define regCM0_CM_TEST_DEBUG_DATA 0x0d8e +#define regCM0_CM_TEST_DEBUG_DATA_BASE_IDX 2 // addressBlock: dcn_dc_dpp0_dispdec_dpp_top_dispdec @@ -3960,6 +3968,10 @@ #define regCM1_CM_DEALPHA_BASE_IDX 2 #define regCM1_CM_COEF_FORMAT 0x0ef7 #define regCM1_CM_COEF_FORMAT_BASE_IDX 2 +#define regCM1_CM_TEST_DEBUG_INDEX 0x0ef8 +#define regCM1_CM_TEST_DEBUG_INDEX_BASE_IDX 2 +#define regCM1_CM_TEST_DEBUG_DATA 0x0ef9 +#define regCM1_CM_TEST_DEBUG_DATA_BASE_IDX 2 // addressBlock: dcn_dc_dpp1_dispdec_dpp_top_dispdec @@ -4346,6 +4358,10 @@ #define regCM2_CM_DEALPHA_BASE_IDX 2 #define regCM2_CM_COEF_FORMAT 0x1062 #define regCM2_CM_COEF_FORMAT_BASE_IDX 2 +#define regCM2_CM_TEST_DEBUG_INDEX 0x1063 +#define regCM2_CM_TEST_DEBUG_INDEX_BASE_IDX 2 +#define regCM2_CM_TEST_DEBUG_DATA 0x1064 +#define regCM2_CM_TEST_DEBUG_DATA_BASE_IDX 2 // addressBlock: dcn_dc_dpp2_dispdec_dpp_top_dispdec @@ -4732,6 +4748,10 @@ #define regCM3_CM_DEALPHA_BASE_IDX 2 #define regCM3_CM_COEF_FORMAT 0x11cd #define regCM3_CM_COEF_FORMAT_BASE_IDX 2 +#define regCM3_CM_TEST_DEBUG_INDEX 0x11ce +#define regCM3_CM_TEST_DEBUG_INDEX_BASE_IDX 2 +#define regCM3_CM_TEST_DEBUG_DATA 0x11cf +#define regCM3_CM_TEST_DEBUG_DATA_BASE_IDX 2 // addressBlock: dcn_dc_dpp3_dispdec_dpp_top_dispdec @@ -11780,6 +11800,16 @@ #define regDSCC0_DSCC_RATE_CONTROL_BUFFER2_MAX_FULLNESS_LEVEL_BASE_IDX 2 #define regDSCC0_DSCC_RATE_CONTROL_BUFFER3_MAX_FULLNESS_LEVEL 0x3035 #define regDSCC0_DSCC_RATE_CONTROL_BUFFER3_MAX_FULLNESS_LEVEL_BASE_IDX 2 +#define regDSCC0_DSCC_TEST_DEBUG_BUS_ROTATE 0x303a +#define regDSCC0_DSCC_TEST_DEBUG_BUS_ROTATE_BASE_IDX 2 +#define regDSCC0_DSCC_TEST_DEBUG_DATA0 0x303b +#define regDSCC0_DSCC_TEST_DEBUG_DATA0_BASE_IDX 2 +#define regDSCC0_DSCC_TEST_DEBUG_DATA1 0x303c +#define regDSCC0_DSCC_TEST_DEBUG_DATA1_BASE_IDX 2 +#define regDSCC0_DSCC_TEST_DEBUG_DATA2 0x303d +#define regDSCC0_DSCC_TEST_DEBUG_DATA2_BASE_IDX 2 +#define regDSCC0_DSCC_TEST_DEBUG_DATA3 0x303e +#define regDSCC0_DSCC_TEST_DEBUG_DATA3_BASE_IDX 2 // addressBlock: dcn_dc_dsc0_dispdec_dsccif_dispdec @@ -11888,6 +11918,16 @@ #define regDSCC1_DSCC_RATE_CONTROL_BUFFER2_MAX_FULLNESS_LEVEL_BASE_IDX 2 #define regDSCC1_DSCC_RATE_CONTROL_BUFFER3_MAX_FULLNESS_LEVEL 0x3091 #define regDSCC1_DSCC_RATE_CONTROL_BUFFER3_MAX_FULLNESS_LEVEL_BASE_IDX 2 +#define regDSCC1_DSCC_TEST_DEBUG_BUS_ROTATE 0x3096 +#define regDSCC1_DSCC_TEST_DEBUG_BUS_ROTATE_BASE_IDX 2 +#define regDSCC1_DSCC_TEST_DEBUG_DATA0 0x3097 +#define regDSCC1_DSCC_TEST_DEBUG_DATA0_BASE_IDX 2 +#define regDSCC1_DSCC_TEST_DEBUG_DATA1 0x3098 +#define regDSCC1_DSCC_TEST_DEBUG_DATA1_BASE_IDX 2 +#define regDSCC1_DSCC_TEST_DEBUG_DATA2 0x3099 +#define regDSCC1_DSCC_TEST_DEBUG_DATA2_BASE_IDX 2 +#define regDSCC1_DSCC_TEST_DEBUG_DATA3 0x309a +#define regDSCC1_DSCC_TEST_DEBUG_DATA3_BASE_IDX 2 // addressBlock: dcn_dc_dsc1_dispdec_dsccif_dispdec @@ -11996,6 +12036,16 @@ #define regDSCC2_DSCC_RATE_CONTROL_BUFFER2_MAX_FULLNESS_LEVEL_BASE_IDX 2 #define regDSCC2_DSCC_RATE_CONTROL_BUFFER3_MAX_FULLNESS_LEVEL 0x30ed #define regDSCC2_DSCC_RATE_CONTROL_BUFFER3_MAX_FULLNESS_LEVEL_BASE_IDX 2 +#define regDSCC2_DSCC_TEST_DEBUG_BUS_ROTATE 0x30f2 +#define regDSCC2_DSCC_TEST_DEBUG_BUS_ROTATE_BASE_IDX 2 +#define regDSCC2_DSCC_TEST_DEBUG_DATA0 0x30f3 +#define regDSCC2_DSCC_TEST_DEBUG_DATA0_BASE_IDX 2 +#define regDSCC2_DSCC_TEST_DEBUG_DATA1 0x30f4 +#define regDSCC2_DSCC_TEST_DEBUG_DATA1_BASE_IDX 2 +#define regDSCC2_DSCC_TEST_DEBUG_DATA2 0x30f5 +#define regDSCC2_DSCC_TEST_DEBUG_DATA2_BASE_IDX 2 +#define regDSCC2_DSCC_TEST_DEBUG_DATA3 0x30f6 +#define regDSCC2_DSCC_TEST_DEBUG_DATA3_BASE_IDX 2 // addressBlock: dcn_dc_dsc2_dispdec_dsccif_dispdec @@ -12104,6 +12154,16 @@ #define regDSCC3_DSCC_RATE_CONTROL_BUFFER2_MAX_FULLNESS_LEVEL_BASE_IDX 2 #define regDSCC3_DSCC_RATE_CONTROL_BUFFER3_MAX_FULLNESS_LEVEL 0x3149 #define regDSCC3_DSCC_RATE_CONTROL_BUFFER3_MAX_FULLNESS_LEVEL_BASE_IDX 2 +#define regDSCC3_DSCC_TEST_DEBUG_BUS_ROTATE 0x314e +#define regDSCC3_DSCC_TEST_DEBUG_BUS_ROTATE_BASE_IDX 2 +#define regDSCC3_DSCC_TEST_DEBUG_DATA0 0x314f +#define regDSCC3_DSCC_TEST_DEBUG_DATA0_BASE_IDX 2 +#define regDSCC3_DSCC_TEST_DEBUG_DATA1 0x3150 +#define regDSCC3_DSCC_TEST_DEBUG_DATA1_BASE_IDX 2 +#define regDSCC3_DSCC_TEST_DEBUG_DATA2 0x3151 +#define regDSCC3_DSCC_TEST_DEBUG_DATA2_BASE_IDX 2 +#define regDSCC3_DSCC_TEST_DEBUG_DATA3 0x3152 +#define regDSCC3_DSCC_TEST_DEBUG_DATA3_BASE_IDX 2 // addressBlock: dcn_dc_dsc3_dispdec_dsccif_dispdec diff --git a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_2_0_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_2_0_sh_mask.h index 0691e328d0f0..1093105ca35b 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_2_0_sh_mask.h +++ b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_2_0_sh_mask.h @@ -11544,6 +11544,11 @@ #define CM0_CM_COEF_FORMAT__CM_BIAS_FORMAT_MASK 0x00000001L #define CM0_CM_COEF_FORMAT__CM_POST_CSC_COEF_FORMAT_MASK 0x00000010L #define CM0_CM_COEF_FORMAT__CM_GAMUT_REMAP_COEF_FORMAT_MASK 0x00000100L +//CM0_CM_TEST_DEBUG_INDEX +#define CM0_CM_TEST_DEBUG_INDEX__CM_TEST_DEBUG_INDEX__SHIFT 0x0 +#define CM0_CM_TEST_DEBUG_INDEX__CM_TEST_DEBUG_WRITE_EN__SHIFT 0x8 +#define CM0_CM_TEST_DEBUG_INDEX__CM_TEST_DEBUG_INDEX_MASK 0x000000FFL +#define CM0_CM_TEST_DEBUG_INDEX__CM_TEST_DEBUG_WRITE_EN_MASK 0x00000100L // addressBlock: dcn_dc_dpp0_dispdec_dpp_top_dispdec @@ -42267,6 +42272,18 @@ //DSCC0_DSCC_RATE_CONTROL_BUFFER3_MAX_FULLNESS_LEVEL #define DSCC0_DSCC_RATE_CONTROL_BUFFER3_MAX_FULLNESS_LEVEL__DSCC_RATE_CONTROL_BUFFER3_MAX_FULLNESS_LEVEL__SHIFT 0x0 #define DSCC0_DSCC_RATE_CONTROL_BUFFER3_MAX_FULLNESS_LEVEL__DSCC_RATE_CONTROL_BUFFER3_MAX_FULLNESS_LEVEL_MASK 0x0003FFFFL +//DSCC0_DSCC_TEST_DEBUG_INDEX2 +#define DSCC0_DSCC_TEST_DEBUG_INDEX2__DSCC_TEST_DEBUG_INDEX2__SHIFT 0x0 +#define DSCC0_DSCC_TEST_DEBUG_INDEX2__DSCC_TEST_DEBUG_INDEX2_MASK 0x000000FFL +//DSCC0_DSCC_TEST_DEBUG_BUS_ROTATE +#define DSCC0_DSCC_TEST_DEBUG_BUS_ROTATE__DSCC_TEST_DEBUG_BUS0_ROTATE__SHIFT 0x0 +#define DSCC0_DSCC_TEST_DEBUG_BUS_ROTATE__DSCC_TEST_DEBUG_BUS1_ROTATE__SHIFT 0x8 +#define DSCC0_DSCC_TEST_DEBUG_BUS_ROTATE__DSCC_TEST_DEBUG_BUS2_ROTATE__SHIFT 0x10 +#define DSCC0_DSCC_TEST_DEBUG_BUS_ROTATE__DSCC_TEST_DEBUG_BUS3_ROTATE__SHIFT 0x18 +#define DSCC0_DSCC_TEST_DEBUG_BUS_ROTATE__DSCC_TEST_DEBUG_BUS0_ROTATE_MASK 0x0000001FL +#define DSCC0_DSCC_TEST_DEBUG_BUS_ROTATE__DSCC_TEST_DEBUG_BUS1_ROTATE_MASK 0x00001F00L +#define DSCC0_DSCC_TEST_DEBUG_BUS_ROTATE__DSCC_TEST_DEBUG_BUS2_ROTATE_MASK 0x001F0000L +#define DSCC0_DSCC_TEST_DEBUG_BUS_ROTATE__DSCC_TEST_DEBUG_BUS3_ROTATE_MASK 0x1F000000L // addressBlock: dcn_dc_dsc0_dispdec_dsccif_dispdec @@ -42300,6 +42317,16 @@ #define DSC_TOP0_DSC_TOP_CONTROL__DSC_DSCCLK_R_GATE_DIS_MASK 0x00000100L //DSC_TOP0_DSC_DEBUG_CONTROL #define DSC_TOP0_DSC_DEBUG_CONTROL__DSC_DBG_EN__SHIFT 0x0 +#define DSC_TOP0_DSC_DEBUG_CONTROL__DSC_TEST_CLOCK_MUX_SEL__SHIFT 0x4 +#define DSC_TOP0_DSC_DEBUG_CONTROL__DSC_DBG_EN_MASK 0x00000001L +#define DSC_TOP0_DSC_DEBUG_CONTROL__DSC_TEST_CLOCK_MUX_SEL_MASK 0x00000070L + + +//DSC_TOP0_DSC_DEBUG_CONTROL +#define DSC_TOP0_DSC_DEBUG_CONTROL__DSC_DBG_EN__SHIFT 0x0 +#define DSC_TOP0_DSC_DEBUG_CONTROL__DSC_TEST_CLOCK_MUX_SEL__SHIFT 0x4 +//DSC_TOP0_DSC_DEBUG_CONTROL +#define DSC_TOP0_DSC_DEBUG_CONTROL__DSC_DBG_EN__SHIFT 0x0 #define DSC_TOP0_DSC_DEBUG_CONTROL__DSC_DBG_EN_MASK 0x00000001L diff --git a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_2_1_offset.h b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_2_1_offset.h index 3bd8792fd7b3..a04b8c32c564 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_2_1_offset.h +++ b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_2_1_offset.h @@ -1719,6 +1719,10 @@ #define regDCHUBBUB_TIMEOUT_INTERRUPT_STATUS_BASE_IDX 2 #define regFMON_CTRL 0x0541 #define regFMON_CTRL_BASE_IDX 2 +#define regDCHUBBUB_TEST_DEBUG_INDEX 0x0542 +#define regDCHUBBUB_TEST_DEBUG_INDEX_BASE_IDX 2 +#define regDCHUBBUB_TEST_DEBUG_DATA 0x0543 +#define regDCHUBBUB_TEST_DEBUG_DATA_BASE_IDX 2 // addressBlock: dce_dc_dchubbubl_hubbub_sdpif_dispdec @@ -3573,6 +3577,10 @@ #define regCM0_CM_DEALPHA_BASE_IDX 2 #define regCM0_CM_COEF_FORMAT 0x0d8c #define regCM0_CM_COEF_FORMAT_BASE_IDX 2 +#define regCM0_CM_TEST_DEBUG_INDEX 0x0d8d +#define regCM0_CM_TEST_DEBUG_INDEX_BASE_IDX 2 +#define regCM0_CM_TEST_DEBUG_DATA 0x0d8e +#define regCM0_CM_TEST_DEBUG_DATA_BASE_IDX 2 // addressBlock: dce_dc_dpp0_dispdec_dpp_top_dispdec @@ -3959,6 +3967,10 @@ #define regCM1_CM_DEALPHA_BASE_IDX 2 #define regCM1_CM_COEF_FORMAT 0x0ef7 #define regCM1_CM_COEF_FORMAT_BASE_IDX 2 +#define regCM1_CM_TEST_DEBUG_INDEX 0x0ef8 +#define regCM1_CM_TEST_DEBUG_INDEX_BASE_IDX 2 +#define regCM1_CM_TEST_DEBUG_DATA 0x0ef9 +#define regCM1_CM_TEST_DEBUG_DATA_BASE_IDX 2 // addressBlock: dce_dc_dpp1_dispdec_dpp_top_dispdec @@ -4345,6 +4357,10 @@ #define regCM2_CM_DEALPHA_BASE_IDX 2 #define regCM2_CM_COEF_FORMAT 0x1062 #define regCM2_CM_COEF_FORMAT_BASE_IDX 2 +#define regCM2_CM_TEST_DEBUG_INDEX 0x1063 +#define regCM2_CM_TEST_DEBUG_INDEX_BASE_IDX 2 +#define regCM2_CM_TEST_DEBUG_DATA 0x1064 +#define regCM2_CM_TEST_DEBUG_DATA_BASE_IDX 2 // addressBlock: dce_dc_dpp2_dispdec_dpp_top_dispdec @@ -4731,6 +4747,10 @@ #define regCM3_CM_DEALPHA_BASE_IDX 2 #define regCM3_CM_COEF_FORMAT 0x11cd #define regCM3_CM_COEF_FORMAT_BASE_IDX 2 +#define regCM3_CM_TEST_DEBUG_INDEX 0x11ce +#define regCM3_CM_TEST_DEBUG_INDEX_BASE_IDX 2 +#define regCM3_CM_TEST_DEBUG_DATA 0x11cf +#define regCM3_CM_TEST_DEBUG_DATA_BASE_IDX 2 // addressBlock: dce_dc_dpp3_dispdec_dpp_top_dispdec @@ -11789,6 +11809,10 @@ #define regDSCC0_DSCC_RATE_CONTROL_BUFFER2_MAX_FULLNESS_LEVEL_BASE_IDX 2 #define regDSCC0_DSCC_RATE_CONTROL_BUFFER3_MAX_FULLNESS_LEVEL 0x3035 #define regDSCC0_DSCC_RATE_CONTROL_BUFFER3_MAX_FULLNESS_LEVEL_BASE_IDX 2 +#define regDSCC0_DSCC_TEST_DEBUG_BUS_ROTATE 0x303a +#define regDSCC0_DSCC_TEST_DEBUG_BUS_ROTATE_BASE_IDX 2 +#define regDSCC0_DSCC_TEST_DEBUG_DATA0 0x303b +#define regDSCC0_DSCC_TEST_DEBUG_DATA0_BASE_IDX 2 // addressBlock: dce_dc_dsc0_dispdec_dsccif_dispdec @@ -11897,6 +11921,10 @@ #define regDSCC1_DSCC_RATE_CONTROL_BUFFER2_MAX_FULLNESS_LEVEL_BASE_IDX 2 #define regDSCC1_DSCC_RATE_CONTROL_BUFFER3_MAX_FULLNESS_LEVEL 0x3091 #define regDSCC1_DSCC_RATE_CONTROL_BUFFER3_MAX_FULLNESS_LEVEL_BASE_IDX 2 +#define regDSCC1_DSCC_TEST_DEBUG_BUS_ROTATE 0x3096 +#define regDSCC1_DSCC_TEST_DEBUG_BUS_ROTATE_BASE_IDX 2 +#define regDSCC1_DSCC_TEST_DEBUG_DATA0 0x3097 +#define regDSCC1_DSCC_TEST_DEBUG_DATA0_BASE_IDX 2 // addressBlock: dce_dc_dsc1_dispdec_dsccif_dispdec @@ -12005,7 +12033,10 @@ #define regDSCC2_DSCC_RATE_CONTROL_BUFFER2_MAX_FULLNESS_LEVEL_BASE_IDX 2 #define regDSCC2_DSCC_RATE_CONTROL_BUFFER3_MAX_FULLNESS_LEVEL 0x30ed #define regDSCC2_DSCC_RATE_CONTROL_BUFFER3_MAX_FULLNESS_LEVEL_BASE_IDX 2 - +#define regDSCC2_DSCC_TEST_DEBUG_BUS_ROTATE 0x30f2 +#define regDSCC2_DSCC_TEST_DEBUG_BUS_ROTATE_BASE_IDX 2 +#define regDSCC2_DSCC_TEST_DEBUG_DATA0 0x30f3 +#define regDSCC2_DSCC_TEST_DEBUG_DATA0_BASE_IDX 2 // addressBlock: dce_dc_dsc2_dispdec_dsccif_dispdec // base address: 0x2e0 @@ -12113,6 +12144,10 @@ #define regDSCC3_DSCC_RATE_CONTROL_BUFFER2_MAX_FULLNESS_LEVEL_BASE_IDX 2 #define regDSCC3_DSCC_RATE_CONTROL_BUFFER3_MAX_FULLNESS_LEVEL 0x3149 #define regDSCC3_DSCC_RATE_CONTROL_BUFFER3_MAX_FULLNESS_LEVEL_BASE_IDX 2 +#define regDSCC3_DSCC_TEST_DEBUG_BUS_ROTATE 0x314e +#define regDSCC3_DSCC_TEST_DEBUG_BUS_ROTATE_BASE_IDX 2 +#define regDSCC3_DSCC_TEST_DEBUG_DATA0 0x314f +#define regDSCC3_DSCC_TEST_DEBUG_DATA0_BASE_IDX 2 // addressBlock: dce_dc_dsc3_dispdec_dsccif_dispdec diff --git a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_2_1_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_2_1_sh_mask.h index e82dffc2b9b0..ce773fca621f 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_2_1_sh_mask.h +++ b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_2_1_sh_mask.h @@ -11547,6 +11547,11 @@ #define CM0_CM_COEF_FORMAT__CM_BIAS_FORMAT_MASK 0x00000001L #define CM0_CM_COEF_FORMAT__CM_POST_CSC_COEF_FORMAT_MASK 0x00000010L #define CM0_CM_COEF_FORMAT__CM_GAMUT_REMAP_COEF_FORMAT_MASK 0x00000100L +//CM0_CM_TEST_DEBUG_INDEX +#define CM0_CM_TEST_DEBUG_INDEX__CM_TEST_DEBUG_INDEX__SHIFT 0x0 +#define CM0_CM_TEST_DEBUG_INDEX__CM_TEST_DEBUG_WRITE_EN__SHIFT 0x8 +#define CM0_CM_TEST_DEBUG_INDEX__CM_TEST_DEBUG_INDEX_MASK 0x000000FFL +#define CM0_CM_TEST_DEBUG_INDEX__CM_TEST_DEBUG_WRITE_EN_MASK 0x00000100L // addressBlock: dce_dc_dpp0_dispdec_dpp_top_dispdec @@ -42315,6 +42320,15 @@ //DSCC0_DSCC_RATE_CONTROL_BUFFER3_MAX_FULLNESS_LEVEL #define DSCC0_DSCC_RATE_CONTROL_BUFFER3_MAX_FULLNESS_LEVEL__DSCC_RATE_CONTROL_BUFFER3_MAX_FULLNESS_LEVEL__SHIFT 0x0 #define DSCC0_DSCC_RATE_CONTROL_BUFFER3_MAX_FULLNESS_LEVEL__DSCC_RATE_CONTROL_BUFFER3_MAX_FULLNESS_LEVEL_MASK 0x0003FFFFL +//DSCC0_DSCC_TEST_DEBUG_BUS_ROTATE +#define DSCC0_DSCC_TEST_DEBUG_BUS_ROTATE__DSCC_TEST_DEBUG_BUS0_ROTATE__SHIFT 0x0 +#define DSCC0_DSCC_TEST_DEBUG_BUS_ROTATE__DSCC_TEST_DEBUG_BUS1_ROTATE__SHIFT 0x8 +#define DSCC0_DSCC_TEST_DEBUG_BUS_ROTATE__DSCC_TEST_DEBUG_BUS2_ROTATE__SHIFT 0x10 +#define DSCC0_DSCC_TEST_DEBUG_BUS_ROTATE__DSCC_TEST_DEBUG_BUS3_ROTATE__SHIFT 0x18 +#define DSCC0_DSCC_TEST_DEBUG_BUS_ROTATE__DSCC_TEST_DEBUG_BUS0_ROTATE_MASK 0x0000001FL +#define DSCC0_DSCC_TEST_DEBUG_BUS_ROTATE__DSCC_TEST_DEBUG_BUS1_ROTATE_MASK 0x00001F00L +#define DSCC0_DSCC_TEST_DEBUG_BUS_ROTATE__DSCC_TEST_DEBUG_BUS2_ROTATE_MASK 0x001F0000L +#define DSCC0_DSCC_TEST_DEBUG_BUS_ROTATE__DSCC_TEST_DEBUG_BUS3_ROTATE_MASK 0x1F000000L // addressBlock: dce_dc_dsc0_dispdec_dsccif_dispdec @@ -42348,7 +42362,9 @@ #define DSC_TOP0_DSC_TOP_CONTROL__DSC_DSCCLK_R_GATE_DIS_MASK 0x00000100L //DSC_TOP0_DSC_DEBUG_CONTROL #define DSC_TOP0_DSC_DEBUG_CONTROL__DSC_DBG_EN__SHIFT 0x0 +#define DSC_TOP0_DSC_DEBUG_CONTROL__DSC_TEST_CLOCK_MUX_SEL__SHIFT 0x4 #define DSC_TOP0_DSC_DEBUG_CONTROL__DSC_DBG_EN_MASK 0x00000001L +#define DSC_TOP0_DSC_DEBUG_CONTROL__DSC_TEST_CLOCK_MUX_SEL_MASK 0x00000070L // addressBlock: dce_dc_dsc1_dispdec_dscc_dispdec diff --git a/drivers/gpu/drm/amd/include/asic_reg/dpcs/dpcs_3_0_0_offset.h b/drivers/gpu/drm/amd/include/asic_reg/dpcs/dpcs_3_0_0_offset.h index 0bb47e06eee8..081e726afbf0 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/dpcs/dpcs_3_0_0_offset.h +++ b/drivers/gpu/drm/amd/include/asic_reg/dpcs/dpcs_3_0_0_offset.h @@ -24,6 +24,8 @@ #define mmDPCSTX0_DPCSTX_PLL_UPDATE_ADDR_BASE_IDX 2 #define mmDPCSTX0_DPCSTX_PLL_UPDATE_DATA 0x292d #define mmDPCSTX0_DPCSTX_PLL_UPDATE_DATA_BASE_IDX 2 +#define mmDPCSTX0_DPCSTX_DEBUG_CONFIG 0x292e +#define mmDPCSTX0_DPCSTX_DEBUG_CONFIG_BASE_IDX 2 // addressBlock: dpcssys_dpcs0_rdpcstx0_dispdec @@ -50,6 +52,8 @@ #define mmRDPCSTX0_RDPCSTX_CNTL2_BASE_IDX 2 #define mmRDPCSTX0_RDPCSTX_DMCU_DPALT_DIS_BLOCK_REG 0x293c #define mmRDPCSTX0_RDPCSTX_DMCU_DPALT_DIS_BLOCK_REG_BASE_IDX 2 +#define mmRDPCSTX0_RDPCSTX_DEBUG_CONFIG 0x293d +#define mmRDPCSTX0_RDPCSTX_DEBUG_CONFIG_BASE_IDX 2 #define mmRDPCSTX0_RDPCSTX_PHY_CNTL0 0x2940 #define mmRDPCSTX0_RDPCSTX_PHY_CNTL0_BASE_IDX 2 #define mmRDPCSTX0_RDPCSTX_PHY_CNTL1 0x2941 @@ -120,6 +124,8 @@ #define mmDPCSTX1_DPCSTX_PLL_UPDATE_ADDR_BASE_IDX 2 #define mmDPCSTX1_DPCSTX_PLL_UPDATE_DATA 0x2a05 #define mmDPCSTX1_DPCSTX_PLL_UPDATE_DATA_BASE_IDX 2 +#define mmDPCSTX1_DPCSTX_DEBUG_CONFIG 0x2a06 +#define mmDPCSTX1_DPCSTX_DEBUG_CONFIG_BASE_IDX 2 // addressBlock: dpcssys_dpcs0_rdpcstx1_dispdec @@ -146,6 +152,8 @@ #define mmRDPCSTX1_RDPCSTX_CNTL2_BASE_IDX 2 #define mmRDPCSTX1_RDPCSTX_DMCU_DPALT_DIS_BLOCK_REG 0x2a14 #define mmRDPCSTX1_RDPCSTX_DMCU_DPALT_DIS_BLOCK_REG_BASE_IDX 2 +#define mmRDPCSTX1_RDPCSTX_DEBUG_CONFIG 0x2a15 +#define mmRDPCSTX1_RDPCSTX_DEBUG_CONFIG_BASE_IDX 2 #define mmRDPCSTX1_RDPCSTX_PHY_CNTL0 0x2a18 #define mmRDPCSTX1_RDPCSTX_PHY_CNTL0_BASE_IDX 2 #define mmRDPCSTX1_RDPCSTX_PHY_CNTL1 0x2a19 @@ -216,6 +224,8 @@ #define mmDPCSTX2_DPCSTX_PLL_UPDATE_ADDR_BASE_IDX 2 #define mmDPCSTX2_DPCSTX_PLL_UPDATE_DATA 0x2add #define mmDPCSTX2_DPCSTX_PLL_UPDATE_DATA_BASE_IDX 2 +#define mmDPCSTX2_DPCSTX_DEBUG_CONFIG 0x2ade +#define mmDPCSTX2_DPCSTX_DEBUG_CONFIG_BASE_IDX 2 // addressBlock: dpcssys_dpcs0_rdpcstx2_dispdec @@ -242,6 +252,8 @@ #define mmRDPCSTX2_RDPCSTX_CNTL2_BASE_IDX 2 #define mmRDPCSTX2_RDPCSTX_DMCU_DPALT_DIS_BLOCK_REG 0x2aec #define mmRDPCSTX2_RDPCSTX_DMCU_DPALT_DIS_BLOCK_REG_BASE_IDX 2 +#define mmRDPCSTX2_RDPCSTX_DEBUG_CONFIG 0x2aed +#define mmRDPCSTX2_RDPCSTX_DEBUG_CONFIG_BASE_IDX 2 #define mmRDPCSTX2_RDPCSTX_PHY_CNTL0 0x2af0 #define mmRDPCSTX2_RDPCSTX_PHY_CNTL0_BASE_IDX 2 #define mmRDPCSTX2_RDPCSTX_PHY_CNTL1 0x2af1 @@ -312,6 +324,8 @@ #define mmDPCSTX3_DPCSTX_PLL_UPDATE_ADDR_BASE_IDX 2 #define mmDPCSTX3_DPCSTX_PLL_UPDATE_DATA 0x2bb5 #define mmDPCSTX3_DPCSTX_PLL_UPDATE_DATA_BASE_IDX 2 +#define mmDPCSTX3_DPCSTX_DEBUG_CONFIG 0x2bb6 +#define mmDPCSTX3_DPCSTX_DEBUG_CONFIG_BASE_IDX 2 // addressBlock: dpcssys_dpcs0_rdpcstx3_dispdec @@ -338,6 +352,8 @@ #define mmRDPCSTX3_RDPCSTX_CNTL2_BASE_IDX 2 #define mmRDPCSTX3_RDPCSTX_DMCU_DPALT_DIS_BLOCK_REG 0x2bc4 #define mmRDPCSTX3_RDPCSTX_DMCU_DPALT_DIS_BLOCK_REG_BASE_IDX 2 +#define mmRDPCSTX3_RDPCSTX_DEBUG_CONFIG 0x2bc5 +#define mmRDPCSTX3_RDPCSTX_DEBUG_CONFIG_BASE_IDX 2 #define mmRDPCSTX3_RDPCSTX_PHY_CNTL0 0x2bc8 #define mmRDPCSTX3_RDPCSTX_PHY_CNTL0_BASE_IDX 2 #define mmRDPCSTX3_RDPCSTX_PHY_CNTL1 0x2bc9 @@ -408,6 +424,8 @@ #define mmDPCSTX4_DPCSTX_PLL_UPDATE_ADDR_BASE_IDX 2 #define mmDPCSTX4_DPCSTX_PLL_UPDATE_DATA 0x2c8d #define mmDPCSTX4_DPCSTX_PLL_UPDATE_DATA_BASE_IDX 2 +#define mmDPCSTX4_DPCSTX_DEBUG_CONFIG 0x2c8e +#define mmDPCSTX4_DPCSTX_DEBUG_CONFIG_BASE_IDX 2 // addressBlock: dpcssys_dpcs0_rdpcstx4_dispdec @@ -434,6 +452,8 @@ #define mmRDPCSTX4_RDPCSTX_CNTL2_BASE_IDX 2 #define mmRDPCSTX4_RDPCSTX_DMCU_DPALT_DIS_BLOCK_REG 0x2c9c #define mmRDPCSTX4_RDPCSTX_DMCU_DPALT_DIS_BLOCK_REG_BASE_IDX 2 +#define mmRDPCSTX4_RDPCSTX_DEBUG_CONFIG 0x2c9d +#define mmRDPCSTX4_RDPCSTX_DEBUG_CONFIG_BASE_IDX 2 #define mmRDPCSTX4_RDPCSTX_PHY_CNTL0 0x2ca0 #define mmRDPCSTX4_RDPCSTX_PHY_CNTL0_BASE_IDX 2 #define mmRDPCSTX4_RDPCSTX_PHY_CNTL1 0x2ca1 @@ -504,6 +524,8 @@ #define mmDPCSTX5_DPCSTX_PLL_UPDATE_ADDR_BASE_IDX 2 #define mmDPCSTX5_DPCSTX_PLL_UPDATE_DATA 0x2d65 #define mmDPCSTX5_DPCSTX_PLL_UPDATE_DATA_BASE_IDX 2 +#define mmDPCSTX5_DPCSTX_DEBUG_CONFIG 0x2d66 +#define mmDPCSTX5_DPCSTX_DEBUG_CONFIG_BASE_IDX 2 // addressBlock: dpcssys_dpcs0_rdpcstx5_dispdec @@ -530,6 +552,8 @@ #define mmRDPCSTX5_RDPCSTX_CNTL2_BASE_IDX 2 #define mmRDPCSTX5_RDPCSTX_DMCU_DPALT_DIS_BLOCK_REG 0x2d74 #define mmRDPCSTX5_RDPCSTX_DMCU_DPALT_DIS_BLOCK_REG_BASE_IDX 2 +#define mmRDPCSTX5_RDPCSTX_DEBUG_CONFIG 0x2d75 +#define mmRDPCSTX5_RDPCSTX_DEBUG_CONFIG_BASE_IDX 2 #define mmRDPCSTX5_RDPCSTX_PHY_CNTL0 0x2d78 #define mmRDPCSTX5_RDPCSTX_PHY_CNTL0_BASE_IDX 2 #define mmRDPCSTX5_RDPCSTX_PHY_CNTL1 0x2d79 diff --git a/drivers/gpu/drm/amd/include/asic_reg/dpcs/dpcs_3_0_0_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/dpcs/dpcs_3_0_0_sh_mask.h index 23fa1121a967..1f846fa6c1a2 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/dpcs/dpcs_3_0_0_sh_mask.h +++ b/drivers/gpu/drm/amd/include/asic_reg/dpcs/dpcs_3_0_0_sh_mask.h @@ -70,7 +70,9 @@ //DPCSTX0_DPCSTX_PLL_UPDATE_DATA #define DPCSTX0_DPCSTX_PLL_UPDATE_DATA__DPCS_PLL_UPDATE_DATA__SHIFT 0x0 #define DPCSTX0_DPCSTX_PLL_UPDATE_DATA__DPCS_PLL_UPDATE_DATA_MASK 0xFFFFFFFFL - +//DPCSTX0_DPCSTX_DEBUG_CONFIG +#define DPCSTX0_DPCSTX_DEBUG_CONFIG__DPCS_DBG_CBUS_DIS__SHIFT 0xe +#define DPCSTX0_DPCSTX_DEBUG_CONFIG__DPCS_DBG_CBUS_DIS_MASK 0x00004000L // addressBlock: dpcssys_dpcs0_rdpcstx0_dispdec //RDPCSTX0_RDPCSTX_CNTL diff --git a/drivers/gpu/drm/amd/include/asic_reg/dpcs/dpcs_3_0_3_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/dpcs/dpcs_3_0_3_sh_mask.h index 55743d06f728..e55ff0e8d74c 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/dpcs/dpcs_3_0_3_sh_mask.h +++ b/drivers/gpu/drm/amd/include/asic_reg/dpcs/dpcs_3_0_3_sh_mask.h @@ -70,7 +70,9 @@ //DPCSTX0_DPCSTX_PLL_UPDATE_DATA #define DPCSTX0_DPCSTX_PLL_UPDATE_DATA__DPCS_PLL_UPDATE_DATA__SHIFT 0x0 #define DPCSTX0_DPCSTX_PLL_UPDATE_DATA__DPCS_PLL_UPDATE_DATA_MASK 0xFFFFFFFFL - +//DPCSTX0_DPCSTX_DEBUG_CONFIG +#define DPCSTX0_DPCSTX_DEBUG_CONFIG__DPCS_DBG_CBUS_DIS__SHIFT 0xe +#define DPCSTX0_DPCSTX_DEBUG_CONFIG__DPCS_DBG_CBUS_DIS_MASK 0x00004000L // addressBlock: dpcssys_dpcs0_rdpcstx0_dispdec //RDPCSTX0_RDPCSTX_CNTL diff --git a/drivers/gpu/drm/amd/include/asic_reg/dpcs/dpcs_4_2_0_offset.h b/drivers/gpu/drm/amd/include/asic_reg/dpcs/dpcs_4_2_0_offset.h index 01a56556cde1..5b4fdeda1040 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/dpcs/dpcs_4_2_0_offset.h +++ b/drivers/gpu/drm/amd/include/asic_reg/dpcs/dpcs_4_2_0_offset.h @@ -155,6 +155,8 @@ #define regRDPCSTX0_RDPCSTX_CNTL2_BASE_IDX 2 #define regRDPCSTX0_RDPCSTX_DMCU_DPALT_DIS_BLOCK_REG 0x293c #define regRDPCSTX0_RDPCSTX_DMCU_DPALT_DIS_BLOCK_REG_BASE_IDX 2 +#define regRDPCSTX0_RDPCSTX_DEBUG_CONFIG 0x293d +#define regRDPCSTX0_RDPCSTX_DEBUG_CONFIG_BASE_IDX 2 #define regRDPCSTX0_RDPCSTX_PHY_CNTL0 0x2940 #define regRDPCSTX0_RDPCSTX_PHY_CNTL0_BASE_IDX 2 #define regRDPCSTX0_RDPCSTX_PHY_CNTL1 0x2941 @@ -239,6 +241,8 @@ #define regRDPCSTX1_RDPCSTX_CNTL2_BASE_IDX 2 #define regRDPCSTX1_RDPCSTX_DMCU_DPALT_DIS_BLOCK_REG 0x2a14 #define regRDPCSTX1_RDPCSTX_DMCU_DPALT_DIS_BLOCK_REG_BASE_IDX 2 +#define regRDPCSTX1_RDPCSTX_DEBUG_CONFIG 0x2a15 +#define regRDPCSTX1_RDPCSTX_DEBUG_CONFIG_BASE_IDX 2 #define regRDPCSTX1_RDPCSTX_PHY_CNTL0 0x2a18 #define regRDPCSTX1_RDPCSTX_PHY_CNTL0_BASE_IDX 2 #define regRDPCSTX1_RDPCSTX_PHY_CNTL1 0x2a19 @@ -323,6 +327,8 @@ #define regRDPCSTX2_RDPCSTX_CNTL2_BASE_IDX 2 #define regRDPCSTX2_RDPCSTX_DMCU_DPALT_DIS_BLOCK_REG 0x2aec #define regRDPCSTX2_RDPCSTX_DMCU_DPALT_DIS_BLOCK_REG_BASE_IDX 2 +#define regRDPCSTX2_RDPCSTX_DEBUG_CONFIG 0x2aed +#define regRDPCSTX2_RDPCSTX_DEBUG_CONFIG_BASE_IDX 2 #define regRDPCSTX2_RDPCSTX_PHY_CNTL0 0x2af0 #define regRDPCSTX2_RDPCSTX_PHY_CNTL0_BASE_IDX 2 #define regRDPCSTX2_RDPCSTX_PHY_CNTL1 0x2af1 @@ -407,6 +413,8 @@ #define regRDPCSTX3_RDPCSTX_CNTL2_BASE_IDX 2 #define regRDPCSTX3_RDPCSTX_DMCU_DPALT_DIS_BLOCK_REG 0x2bc4 #define regRDPCSTX3_RDPCSTX_DMCU_DPALT_DIS_BLOCK_REG_BASE_IDX 2 +#define regRDPCSTX3_RDPCSTX_DEBUG_CONFIG 0x2bc5 +#define regRDPCSTX3_RDPCSTX_DEBUG_CONFIG_BASE_IDX 2 #define regRDPCSTX3_RDPCSTX_PHY_CNTL0 0x2bc8 #define regRDPCSTX3_RDPCSTX_PHY_CNTL0_BASE_IDX 2 #define regRDPCSTX3_RDPCSTX_PHY_CNTL1 0x2bc9 @@ -491,6 +499,8 @@ #define regRDPCSTX4_RDPCSTX_CNTL2_BASE_IDX 2 #define regRDPCSTX4_RDPCSTX_DMCU_DPALT_DIS_BLOCK_REG 0x2c9c #define regRDPCSTX4_RDPCSTX_DMCU_DPALT_DIS_BLOCK_REG_BASE_IDX 2 +#define regRDPCSTX4_RDPCSTX_DEBUG_CONFIG 0x2c9d +#define regRDPCSTX4_RDPCSTX_DEBUG_CONFIG_BASE_IDX 2 #define regRDPCSTX4_RDPCSTX_PHY_CNTL0 0x2ca0 #define regRDPCSTX4_RDPCSTX_PHY_CNTL0_BASE_IDX 2 #define regRDPCSTX4_RDPCSTX_PHY_CNTL1 0x2ca1 diff --git a/drivers/gpu/drm/amd/include/asic_reg/gc/gc_10_1_0_offset.h b/drivers/gpu/drm/amd/include/asic_reg/gc/gc_10_1_0_offset.h index 4908044f7409..4c8e7fdb6976 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/gc/gc_10_1_0_offset.h +++ b/drivers/gpu/drm/amd/include/asic_reg/gc/gc_10_1_0_offset.h @@ -4830,6 +4830,8 @@ #define mmCP_ECC_FIRSTOCCURRENCE_RING2_BASE_IDX 0 #define mmGB_EDC_MODE 0x1e1e #define mmGB_EDC_MODE_BASE_IDX 0 +#define mmCP_DEBUG 0x1e1f +#define mmCP_DEBUG_BASE_IDX 0 #define mmCP_FETCHER_SOURCE 0x1e22 #define mmCP_FETCHER_SOURCE_BASE_IDX 0 #define mmCP_PQ_WPTR_POLL_CNTL 0x1e23 @@ -7778,6 +7780,8 @@ #define mmCP_MES_DOORBELL_CONTROL5_BASE_IDX 1 #define mmCP_MES_DOORBELL_CONTROL6 0x2841 #define mmCP_MES_DOORBELL_CONTROL6_BASE_IDX 1 +#define mmCP_MES_DEBUG_INTERRUPT_INSTR_PNTR 0x2842 +#define mmCP_MES_DEBUG_INTERRUPT_INSTR_PNTR_BASE_IDX 1 #define mmCP_MES_GP0_LO 0x2843 #define mmCP_MES_GP0_LO_BASE_IDX 1 #define mmCP_MES_GP0_HI 0x2844 @@ -9332,10 +9336,16 @@ #define mmRLC_LB_CNTR_INIT_1_BASE_IDX 1 #define mmRLC_LB_CNTR_1 0x4c1c #define mmRLC_LB_CNTR_1_BASE_IDX 1 +#define mmRLC_GPM_DEBUG_INST_ADDR 0x4c1d +#define mmRLC_GPM_DEBUG_INST_ADDR_BASE_IDX 1 #define mmRLC_JUMP_TABLE_RESTORE 0x4c1e #define mmRLC_JUMP_TABLE_RESTORE_BASE_IDX 1 #define mmRLC_PG_DELAY_2 0x4c1f #define mmRLC_PG_DELAY_2_BASE_IDX 1 +#define mmRLC_GPM_DEBUG_INST_A 0x4c22 +#define mmRLC_GPM_DEBUG_INST_A_BASE_IDX 1 +#define mmRLC_GPM_DEBUG_INST_B 0x4c23 +#define mmRLC_GPM_DEBUG_INST_B_BASE_IDX 1 #define mmRLC_GPU_CLOCK_COUNT_LSB 0x4c24 #define mmRLC_GPU_CLOCK_COUNT_LSB_BASE_IDX 1 #define mmRLC_GPU_CLOCK_COUNT_MSB 0x4c25 @@ -9720,6 +9730,8 @@ #define mmRLC_SPM_THREAD_TRACE_CTRL_BASE_IDX 1 #define mmRLC_LB_CNTR_2 0x4de7 #define mmRLC_LB_CNTR_2_BASE_IDX 1 +#define mmRLC_LX6_CORE_PDEBUG_INST 0x4deb +#define mmRLC_LX6_CORE_PDEBUG_INST_BASE_IDX 1 #define mmRLC_CPAXI_DOORBELL_MON_CTRL 0x4df1 #define mmRLC_CPAXI_DOORBELL_MON_CTRL_BASE_IDX 1 #define mmRLC_CPAXI_DOORBELL_MON_STAT 0x4df2 diff --git a/drivers/gpu/drm/amd/include/asic_reg/gc/gc_9_0_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/gc/gc_9_0_sh_mask.h index efc16ddf274a..2dfa0e5b1aa3 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/gc/gc_9_0_sh_mask.h +++ b/drivers/gpu/drm/amd/include/asic_reg/gc/gc_9_0_sh_mask.h @@ -6822,6 +6822,8 @@ #define VM_L2_PROTECTION_FAULT_STATUS__VMID__SHIFT 0x14 #define VM_L2_PROTECTION_FAULT_STATUS__VF__SHIFT 0x18 #define VM_L2_PROTECTION_FAULT_STATUS__VFID__SHIFT 0x19 +#define VM_L2_PROTECTION_FAULT_STATUS__UCE__SHIFT 0x1d +#define VM_L2_PROTECTION_FAULT_STATUS__FED__SHIFT 0x1e #define VM_L2_PROTECTION_FAULT_STATUS__MORE_FAULTS_MASK 0x00000001L #define VM_L2_PROTECTION_FAULT_STATUS__WALKER_ERROR_MASK 0x0000000EL #define VM_L2_PROTECTION_FAULT_STATUS__PERMISSION_FAULTS_MASK 0x000000F0L @@ -6832,6 +6834,8 @@ #define VM_L2_PROTECTION_FAULT_STATUS__VMID_MASK 0x00F00000L #define VM_L2_PROTECTION_FAULT_STATUS__VF_MASK 0x01000000L #define VM_L2_PROTECTION_FAULT_STATUS__VFID_MASK 0x1E000000L +#define VM_L2_PROTECTION_FAULT_STATUS__UCE_MASK 0x20000000L +#define VM_L2_PROTECTION_FAULT_STATUS__FED_MASK 0x40000000L //VM_L2_PROTECTION_FAULT_ADDR_LO32 #define VM_L2_PROTECTION_FAULT_ADDR_LO32__LOGICAL_PAGE_ADDR_LO32__SHIFT 0x0 #define VM_L2_PROTECTION_FAULT_ADDR_LO32__LOGICAL_PAGE_ADDR_LO32_MASK 0xFFFFFFFFL diff --git a/drivers/gpu/drm/amd/include/asic_reg/oss/osssys_6_0_0_offset.h b/drivers/gpu/drm/amd/include/asic_reg/oss/osssys_6_0_0_offset.h index 8b931bbabe70..969e006b859b 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/oss/osssys_6_0_0_offset.h +++ b/drivers/gpu/drm/amd/include/asic_reg/oss/osssys_6_0_0_offset.h @@ -237,6 +237,10 @@ #define regSEM_REGISTER_LAST_PART2_BASE_IDX 0 #define regIH_CLIENT_CFG 0x0184 #define regIH_CLIENT_CFG_BASE_IDX 0 +#define regIH_RING1_CLIENT_CFG_INDEX 0x0185 +#define regIH_RING1_CLIENT_CFG_INDEX_BASE_IDX 0 +#define regIH_RING1_CLIENT_CFG_DATA 0x0186 +#define regIH_RING1_CLIENT_CFG_DATA_BASE_IDX 0 #define regIH_CLIENT_CFG_INDEX 0x0188 #define regIH_CLIENT_CFG_INDEX_BASE_IDX 0 #define regIH_CLIENT_CFG_DATA 0x0189 diff --git a/drivers/gpu/drm/amd/include/asic_reg/oss/osssys_6_0_0_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/oss/osssys_6_0_0_sh_mask.h index f262f44fa68c..a672a91e58f0 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/oss/osssys_6_0_0_sh_mask.h +++ b/drivers/gpu/drm/amd/include/asic_reg/oss/osssys_6_0_0_sh_mask.h @@ -888,6 +888,16 @@ //IH_CLIENT_CFG #define IH_CLIENT_CFG__TOTAL_CLIENT_NUM__SHIFT 0x0 #define IH_CLIENT_CFG__TOTAL_CLIENT_NUM_MASK 0x0000003FL +//IH_RING1_CLIENT_CFG_INDEX +#define IH_RING1_CLIENT_CFG_INDEX__INDEX__SHIFT 0x0 +#define IH_RING1_CLIENT_CFG_INDEX__INDEX_MASK 0x00000007L +//IH_RING1_CLIENT_CFG_DATA +#define IH_RING1_CLIENT_CFG_DATA__CLIENT_ID__SHIFT 0x0 +#define IH_RING1_CLIENT_CFG_DATA__SOURCE_ID__SHIFT 0x8 +#define IH_RING1_CLIENT_CFG_DATA__SOURCE_ID_MATCH_ENABLE__SHIFT 0x10 +#define IH_RING1_CLIENT_CFG_DATA__CLIENT_ID_MASK 0x000000FFL +#define IH_RING1_CLIENT_CFG_DATA__SOURCE_ID_MASK 0x0000FF00L +#define IH_RING1_CLIENT_CFG_DATA__SOURCE_ID_MATCH_ENABLE_MASK 0x00010000L //IH_CLIENT_CFG_INDEX #define IH_CLIENT_CFG_INDEX__INDEX__SHIFT 0x0 #define IH_CLIENT_CFG_INDEX__INDEX_MASK 0x0000001FL diff --git a/drivers/gpu/drm/amd/include/asic_reg/smuio/smuio_14_0_2_offset.h b/drivers/gpu/drm/amd/include/asic_reg/smuio/smuio_14_0_2_offset.h new file mode 100644 index 000000000000..da7e31fedd58 --- /dev/null +++ b/drivers/gpu/drm/amd/include/asic_reg/smuio/smuio_14_0_2_offset.h @@ -0,0 +1,511 @@ +/* + * Copyright 2023 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ +#ifndef _smuio_14_0_2_OFFSET_HEADER +#define _smuio_14_0_2_OFFSET_HEADER + + + +// addressBlock: smuio_smuio_tsc_SmuSmuioDec +// base address: 0x5a8a0 +#define regPWROK_REFCLK_GAP_CYCLES 0x0028 +#define regPWROK_REFCLK_GAP_CYCLES_BASE_IDX 1 +#define regGOLDEN_TSC_INCREMENT_UPPER 0x002b +#define regGOLDEN_TSC_INCREMENT_UPPER_BASE_IDX 1 +#define regGOLDEN_TSC_INCREMENT_LOWER 0x002c +#define regGOLDEN_TSC_INCREMENT_LOWER_BASE_IDX 1 +#define regGOLDEN_TSC_COUNT_UPPER 0x002d +#define regGOLDEN_TSC_COUNT_UPPER_BASE_IDX 1 +#define regGOLDEN_TSC_COUNT_LOWER 0x002e +#define regGOLDEN_TSC_COUNT_LOWER_BASE_IDX 1 +#define regSOC_GOLDEN_TSC_SHADOW_UPPER 0x002f +#define regSOC_GOLDEN_TSC_SHADOW_UPPER_BASE_IDX 1 +#define regSOC_GOLDEN_TSC_SHADOW_LOWER 0x0030 +#define regSOC_GOLDEN_TSC_SHADOW_LOWER_BASE_IDX 1 +#define regSOC_GAP_PWROK 0x0031 +#define regSOC_GAP_PWROK_BASE_IDX 1 + + +// addressBlock: smuio_smuio_swtimer_SmuSmuioDec +// base address: 0x5aca8 +#define regPWR_VIRT_RESET_REQ 0x012a +#define regPWR_VIRT_RESET_REQ_BASE_IDX 1 +#define regPWR_DISP_TIMER_CONTROL 0x012b +#define regPWR_DISP_TIMER_CONTROL_BASE_IDX 1 +#define regPWR_DISP_TIMER_DEBUG 0x012c +#define regPWR_DISP_TIMER_DEBUG_BASE_IDX 1 +#define regPWR_DISP_TIMER2_CONTROL 0x012d +#define regPWR_DISP_TIMER2_CONTROL_BASE_IDX 1 +#define regPWR_DISP_TIMER2_DEBUG 0x012e +#define regPWR_DISP_TIMER2_DEBUG_BASE_IDX 1 +#define regPWR_DISP_TIMER_GLOBAL_CONTROL 0x012f +#define regPWR_DISP_TIMER_GLOBAL_CONTROL_BASE_IDX 1 +#define regPWR_IH_CONTROL 0x0130 +#define regPWR_IH_CONTROL_BASE_IDX 1 + + +// addressBlock: smuio_smuio_misc_SmuSmuioDec +// base address: 0x5a000 +#define regSMUIO_MCM_CONFIG 0x0023 +#define regSMUIO_MCM_CONFIG_BASE_IDX 0 +#define regIP_DISCOVERY_VERSION 0x0000 +#define regIP_DISCOVERY_VERSION_BASE_IDX 1 +#define regSCRATCH_REGISTER0 0x01bd +#define regSCRATCH_REGISTER0_BASE_IDX 1 +#define regSCRATCH_REGISTER1 0x01be +#define regSCRATCH_REGISTER1_BASE_IDX 1 +#define regSCRATCH_REGISTER2 0x01bf +#define regSCRATCH_REGISTER2_BASE_IDX 1 +#define regSCRATCH_REGISTER3 0x01c0 +#define regSCRATCH_REGISTER3_BASE_IDX 1 +#define regSCRATCH_REGISTER4 0x01c1 +#define regSCRATCH_REGISTER4_BASE_IDX 1 +#define regSCRATCH_REGISTER5 0x01c2 +#define regSCRATCH_REGISTER5_BASE_IDX 1 +#define regSCRATCH_REGISTER6 0x01c3 +#define regSCRATCH_REGISTER6_BASE_IDX 1 +#define regSCRATCH_REGISTER7 0x01c4 +#define regSCRATCH_REGISTER7_BASE_IDX 1 + + +// addressBlock: smuio_smuio_i2c_SmuSmuioDec +// base address: 0x5a100 +#define regCKSVII2C_IC_CON 0x0040 +#define regCKSVII2C_IC_CON_BASE_IDX 0 +#define regCKSVII2C_IC_TAR 0x0041 +#define regCKSVII2C_IC_TAR_BASE_IDX 0 +#define regCKSVII2C_IC_SAR 0x0042 +#define regCKSVII2C_IC_SAR_BASE_IDX 0 +#define regCKSVII2C_IC_HS_MADDR 0x0043 +#define regCKSVII2C_IC_HS_MADDR_BASE_IDX 0 +#define regCKSVII2C_IC_DATA_CMD 0x0044 +#define regCKSVII2C_IC_DATA_CMD_BASE_IDX 0 +#define regCKSVII2C_IC_SS_SCL_HCNT 0x0045 +#define regCKSVII2C_IC_SS_SCL_HCNT_BASE_IDX 0 +#define regCKSVII2C_IC_SS_SCL_LCNT 0x0046 +#define regCKSVII2C_IC_SS_SCL_LCNT_BASE_IDX 0 +#define regCKSVII2C_IC_FS_SCL_HCNT 0x0047 +#define regCKSVII2C_IC_FS_SCL_HCNT_BASE_IDX 0 +#define regCKSVII2C_IC_FS_SCL_LCNT 0x0048 +#define regCKSVII2C_IC_FS_SCL_LCNT_BASE_IDX 0 +#define regCKSVII2C_IC_HS_SCL_HCNT 0x0049 +#define regCKSVII2C_IC_HS_SCL_HCNT_BASE_IDX 0 +#define regCKSVII2C_IC_HS_SCL_LCNT 0x004a +#define regCKSVII2C_IC_HS_SCL_LCNT_BASE_IDX 0 +#define regCKSVII2C_IC_INTR_STAT 0x004b +#define regCKSVII2C_IC_INTR_STAT_BASE_IDX 0 +#define regCKSVII2C_IC_INTR_MASK 0x004c +#define regCKSVII2C_IC_INTR_MASK_BASE_IDX 0 +#define regCKSVII2C_IC_RAW_INTR_STAT 0x004d +#define regCKSVII2C_IC_RAW_INTR_STAT_BASE_IDX 0 +#define regCKSVII2C_IC_RX_TL 0x004e +#define regCKSVII2C_IC_RX_TL_BASE_IDX 0 +#define regCKSVII2C_IC_TX_TL 0x004f +#define regCKSVII2C_IC_TX_TL_BASE_IDX 0 +#define regCKSVII2C_IC_CLR_INTR 0x0050 +#define regCKSVII2C_IC_CLR_INTR_BASE_IDX 0 +#define regCKSVII2C_IC_CLR_RX_UNDER 0x0051 +#define regCKSVII2C_IC_CLR_RX_UNDER_BASE_IDX 0 +#define regCKSVII2C_IC_CLR_RX_OVER 0x0052 +#define regCKSVII2C_IC_CLR_RX_OVER_BASE_IDX 0 +#define regCKSVII2C_IC_CLR_TX_OVER 0x0053 +#define regCKSVII2C_IC_CLR_TX_OVER_BASE_IDX 0 +#define regCKSVII2C_IC_CLR_RD_REQ 0x0054 +#define regCKSVII2C_IC_CLR_RD_REQ_BASE_IDX 0 +#define regCKSVII2C_IC_CLR_TX_ABRT 0x0055 +#define regCKSVII2C_IC_CLR_TX_ABRT_BASE_IDX 0 +#define regCKSVII2C_IC_CLR_RX_DONE 0x0056 +#define regCKSVII2C_IC_CLR_RX_DONE_BASE_IDX 0 +#define regCKSVII2C_IC_CLR_ACTIVITY 0x0057 +#define regCKSVII2C_IC_CLR_ACTIVITY_BASE_IDX 0 +#define regCKSVII2C_IC_CLR_STOP_DET 0x0058 +#define regCKSVII2C_IC_CLR_STOP_DET_BASE_IDX 0 +#define regCKSVII2C_IC_CLR_START_DET 0x0059 +#define regCKSVII2C_IC_CLR_START_DET_BASE_IDX 0 +#define regCKSVII2C_IC_CLR_GEN_CALL 0x005a +#define regCKSVII2C_IC_CLR_GEN_CALL_BASE_IDX 0 +#define regCKSVII2C_IC_ENABLE 0x005b +#define regCKSVII2C_IC_ENABLE_BASE_IDX 0 +#define regCKSVII2C_IC_STATUS 0x005c +#define regCKSVII2C_IC_STATUS_BASE_IDX 0 +#define regCKSVII2C_IC_TXFLR 0x005d +#define regCKSVII2C_IC_TXFLR_BASE_IDX 0 +#define regCKSVII2C_IC_RXFLR 0x005e +#define regCKSVII2C_IC_RXFLR_BASE_IDX 0 +#define regCKSVII2C_IC_SDA_HOLD 0x005f +#define regCKSVII2C_IC_SDA_HOLD_BASE_IDX 0 +#define regCKSVII2C_IC_TX_ABRT_SOURCE 0x0060 +#define regCKSVII2C_IC_TX_ABRT_SOURCE_BASE_IDX 0 +#define regCKSVII2C_IC_SLV_DATA_NACK_ONLY 0x0061 +#define regCKSVII2C_IC_SLV_DATA_NACK_ONLY_BASE_IDX 0 +#define regCKSVII2C_IC_DMA_CR 0x0062 +#define regCKSVII2C_IC_DMA_CR_BASE_IDX 0 +#define regCKSVII2C_IC_DMA_TDLR 0x0063 +#define regCKSVII2C_IC_DMA_TDLR_BASE_IDX 0 +#define regCKSVII2C_IC_DMA_RDLR 0x0064 +#define regCKSVII2C_IC_DMA_RDLR_BASE_IDX 0 +#define regCKSVII2C_IC_SDA_SETUP 0x0065 +#define regCKSVII2C_IC_SDA_SETUP_BASE_IDX 0 +#define regCKSVII2C_IC_ACK_GENERAL_CALL 0x0066 +#define regCKSVII2C_IC_ACK_GENERAL_CALL_BASE_IDX 0 +#define regCKSVII2C_IC_ENABLE_STATUS 0x0067 +#define regCKSVII2C_IC_ENABLE_STATUS_BASE_IDX 0 +#define regCKSVII2C_IC_FS_SPKLEN 0x0068 +#define regCKSVII2C_IC_FS_SPKLEN_BASE_IDX 0 +#define regCKSVII2C_IC_HS_SPKLEN 0x0069 +#define regCKSVII2C_IC_HS_SPKLEN_BASE_IDX 0 +#define regCKSVII2C_IC_CLR_RESTART_DET 0x006a +#define regCKSVII2C_IC_CLR_RESTART_DET_BASE_IDX 0 +#define regCKSVII2C_IC_COMP_PARAM_1 0x006d +#define regCKSVII2C_IC_COMP_PARAM_1_BASE_IDX 0 +#define regCKSVII2C_IC_COMP_VERSION 0x006e +#define regCKSVII2C_IC_COMP_VERSION_BASE_IDX 0 +#define regCKSVII2C_IC_COMP_TYPE 0x006f +#define regCKSVII2C_IC_COMP_TYPE_BASE_IDX 0 +#define regCKSVII2C1_IC_CON 0x0080 +#define regCKSVII2C1_IC_CON_BASE_IDX 0 +#define regCKSVII2C1_IC_TAR 0x0081 +#define regCKSVII2C1_IC_TAR_BASE_IDX 0 +#define regCKSVII2C1_IC_SAR 0x0082 +#define regCKSVII2C1_IC_SAR_BASE_IDX 0 +#define regCKSVII2C1_IC_HS_MADDR 0x0083 +#define regCKSVII2C1_IC_HS_MADDR_BASE_IDX 0 +#define regCKSVII2C1_IC_DATA_CMD 0x0084 +#define regCKSVII2C1_IC_DATA_CMD_BASE_IDX 0 +#define regCKSVII2C1_IC_SS_SCL_HCNT 0x0085 +#define regCKSVII2C1_IC_SS_SCL_HCNT_BASE_IDX 0 +#define regCKSVII2C1_IC_SS_SCL_LCNT 0x0086 +#define regCKSVII2C1_IC_SS_SCL_LCNT_BASE_IDX 0 +#define regCKSVII2C1_IC_FS_SCL_HCNT 0x0087 +#define regCKSVII2C1_IC_FS_SCL_HCNT_BASE_IDX 0 +#define regCKSVII2C1_IC_FS_SCL_LCNT 0x0088 +#define regCKSVII2C1_IC_FS_SCL_LCNT_BASE_IDX 0 +#define regCKSVII2C1_IC_HS_SCL_HCNT 0x0089 +#define regCKSVII2C1_IC_HS_SCL_HCNT_BASE_IDX 0 +#define regCKSVII2C1_IC_HS_SCL_LCNT 0x008a +#define regCKSVII2C1_IC_HS_SCL_LCNT_BASE_IDX 0 +#define regCKSVII2C1_IC_INTR_STAT 0x008b +#define regCKSVII2C1_IC_INTR_STAT_BASE_IDX 0 +#define regCKSVII2C1_IC_INTR_MASK 0x008c +#define regCKSVII2C1_IC_INTR_MASK_BASE_IDX 0 +#define regCKSVII2C1_IC_RAW_INTR_STAT 0x008d +#define regCKSVII2C1_IC_RAW_INTR_STAT_BASE_IDX 0 +#define regCKSVII2C1_IC_RX_TL 0x008e +#define regCKSVII2C1_IC_RX_TL_BASE_IDX 0 +#define regCKSVII2C1_IC_TX_TL 0x008f +#define regCKSVII2C1_IC_TX_TL_BASE_IDX 0 +#define regCKSVII2C1_IC_CLR_INTR 0x0090 +#define regCKSVII2C1_IC_CLR_INTR_BASE_IDX 0 +#define regCKSVII2C1_IC_CLR_RX_UNDER 0x0091 +#define regCKSVII2C1_IC_CLR_RX_UNDER_BASE_IDX 0 +#define regCKSVII2C1_IC_CLR_RX_OVER 0x0092 +#define regCKSVII2C1_IC_CLR_RX_OVER_BASE_IDX 0 +#define regCKSVII2C1_IC_CLR_TX_OVER 0x0093 +#define regCKSVII2C1_IC_CLR_TX_OVER_BASE_IDX 0 +#define regCKSVII2C1_IC_CLR_RD_REQ 0x0094 +#define regCKSVII2C1_IC_CLR_RD_REQ_BASE_IDX 0 +#define regCKSVII2C1_IC_CLR_TX_ABRT 0x0095 +#define regCKSVII2C1_IC_CLR_TX_ABRT_BASE_IDX 0 +#define regCKSVII2C1_IC_CLR_RX_DONE 0x0096 +#define regCKSVII2C1_IC_CLR_RX_DONE_BASE_IDX 0 +#define regCKSVII2C1_IC_CLR_ACTIVITY 0x0097 +#define regCKSVII2C1_IC_CLR_ACTIVITY_BASE_IDX 0 +#define regCKSVII2C1_IC_CLR_STOP_DET 0x0098 +#define regCKSVII2C1_IC_CLR_STOP_DET_BASE_IDX 0 +#define regCKSVII2C1_IC_CLR_START_DET 0x0099 +#define regCKSVII2C1_IC_CLR_START_DET_BASE_IDX 0 +#define regCKSVII2C1_IC_CLR_GEN_CALL 0x009a +#define regCKSVII2C1_IC_CLR_GEN_CALL_BASE_IDX 0 +#define regCKSVII2C1_IC_ENABLE 0x009b +#define regCKSVII2C1_IC_ENABLE_BASE_IDX 0 +#define regCKSVII2C1_IC_STATUS 0x009c +#define regCKSVII2C1_IC_STATUS_BASE_IDX 0 +#define regCKSVII2C1_IC_TXFLR 0x009d +#define regCKSVII2C1_IC_TXFLR_BASE_IDX 0 +#define regCKSVII2C1_IC_RXFLR 0x009e +#define regCKSVII2C1_IC_RXFLR_BASE_IDX 0 +#define regCKSVII2C1_IC_SDA_HOLD 0x009f +#define regCKSVII2C1_IC_SDA_HOLD_BASE_IDX 0 +#define regCKSVII2C1_IC_TX_ABRT_SOURCE 0x00a0 +#define regCKSVII2C1_IC_TX_ABRT_SOURCE_BASE_IDX 0 +#define regCKSVII2C1_IC_SLV_DATA_NACK_ONLY 0x00a1 +#define regCKSVII2C1_IC_SLV_DATA_NACK_ONLY_BASE_IDX 0 +#define regCKSVII2C1_IC_DMA_CR 0x00a2 +#define regCKSVII2C1_IC_DMA_CR_BASE_IDX 0 +#define regCKSVII2C1_IC_DMA_TDLR 0x00a3 +#define regCKSVII2C1_IC_DMA_TDLR_BASE_IDX 0 +#define regCKSVII2C1_IC_DMA_RDLR 0x00a4 +#define regCKSVII2C1_IC_DMA_RDLR_BASE_IDX 0 +#define regCKSVII2C1_IC_SDA_SETUP 0x00a5 +#define regCKSVII2C1_IC_SDA_SETUP_BASE_IDX 0 +#define regCKSVII2C1_IC_ACK_GENERAL_CALL 0x00a6 +#define regCKSVII2C1_IC_ACK_GENERAL_CALL_BASE_IDX 0 +#define regCKSVII2C1_IC_ENABLE_STATUS 0x00a7 +#define regCKSVII2C1_IC_ENABLE_STATUS_BASE_IDX 0 +#define regCKSVII2C1_IC_FS_SPKLEN 0x00a8 +#define regCKSVII2C1_IC_FS_SPKLEN_BASE_IDX 0 +#define regCKSVII2C1_IC_HS_SPKLEN 0x00a9 +#define regCKSVII2C1_IC_HS_SPKLEN_BASE_IDX 0 +#define regCKSVII2C1_IC_CLR_RESTART_DET 0x00aa +#define regCKSVII2C1_IC_CLR_RESTART_DET_BASE_IDX 0 +#define regCKSVII2C1_IC_COMP_PARAM_1 0x00ad +#define regCKSVII2C1_IC_COMP_PARAM_1_BASE_IDX 0 +#define regCKSVII2C1_IC_COMP_VERSION 0x00ae +#define regCKSVII2C1_IC_COMP_VERSION_BASE_IDX 0 +#define regCKSVII2C1_IC_COMP_TYPE 0x00af +#define regCKSVII2C1_IC_COMP_TYPE_BASE_IDX 0 +#define regSMUIO_PWRMGT 0x018c +#define regSMUIO_PWRMGT_BASE_IDX 0 + + +// addressBlock: smuio_smuio_rom_SmuSmuioDec +// base address: 0x5a380 +#define regROM_CNTL 0x00e0 +#define regROM_CNTL_BASE_IDX 0 +#define regPAGE_MIRROR_CNTL 0x00e1 +#define regPAGE_MIRROR_CNTL_BASE_IDX 0 +#define regROM_STATUS 0x00e2 +#define regROM_STATUS_BASE_IDX 0 +#define regCGTT_ROM_CLK_CTRL0 0x00e3 +#define regCGTT_ROM_CLK_CTRL0_BASE_IDX 0 +#define regROM_INDEX 0x00e4 +#define regROM_INDEX_BASE_IDX 0 +#define regROM_DATA 0x00e5 +#define regROM_DATA_BASE_IDX 0 +#define regROM_START 0x00e6 +#define regROM_START_BASE_IDX 0 +#define regROM_SW_CNTL 0x00e8 +#define regROM_SW_CNTL_BASE_IDX 0 +#define regROM_SW_STATUS 0x00e9 +#define regROM_SW_STATUS_BASE_IDX 0 +#define regROM_SW_COMMAND 0x00ea +#define regROM_SW_COMMAND_BASE_IDX 0 +#define regROM_SW_DATA_1 0x00ec +#define regROM_SW_DATA_1_BASE_IDX 0 +#define regROM_SW_DATA_2 0x00ed +#define regROM_SW_DATA_2_BASE_IDX 0 +#define regROM_SW_DATA_3 0x00ee +#define regROM_SW_DATA_3_BASE_IDX 0 +#define regROM_SW_DATA_4 0x00ef +#define regROM_SW_DATA_4_BASE_IDX 0 +#define regROM_SW_DATA_5 0x00f0 +#define regROM_SW_DATA_5_BASE_IDX 0 +#define regROM_SW_DATA_6 0x00f1 +#define regROM_SW_DATA_6_BASE_IDX 0 +#define regROM_SW_DATA_7 0x00f2 +#define regROM_SW_DATA_7_BASE_IDX 0 +#define regROM_SW_DATA_8 0x00f3 +#define regROM_SW_DATA_8_BASE_IDX 0 +#define regROM_SW_DATA_9 0x00f4 +#define regROM_SW_DATA_9_BASE_IDX 0 +#define regROM_SW_DATA_10 0x00f5 +#define regROM_SW_DATA_10_BASE_IDX 0 +#define regROM_SW_DATA_11 0x00f6 +#define regROM_SW_DATA_11_BASE_IDX 0 +#define regROM_SW_DATA_12 0x00f7 +#define regROM_SW_DATA_12_BASE_IDX 0 +#define regROM_SW_DATA_13 0x00f8 +#define regROM_SW_DATA_13_BASE_IDX 0 +#define regROM_SW_DATA_14 0x00f9 +#define regROM_SW_DATA_14_BASE_IDX 0 +#define regROM_SW_DATA_15 0x00fa +#define regROM_SW_DATA_15_BASE_IDX 0 +#define regROM_SW_DATA_16 0x00fb +#define regROM_SW_DATA_16_BASE_IDX 0 +#define regROM_SW_DATA_17 0x00fc +#define regROM_SW_DATA_17_BASE_IDX 0 +#define regROM_SW_DATA_18 0x00fd +#define regROM_SW_DATA_18_BASE_IDX 0 +#define regROM_SW_DATA_19 0x00fe +#define regROM_SW_DATA_19_BASE_IDX 0 +#define regROM_SW_DATA_20 0x00ff +#define regROM_SW_DATA_20_BASE_IDX 0 +#define regROM_SW_DATA_21 0x0100 +#define regROM_SW_DATA_21_BASE_IDX 0 +#define regROM_SW_DATA_22 0x0101 +#define regROM_SW_DATA_22_BASE_IDX 0 +#define regROM_SW_DATA_23 0x0102 +#define regROM_SW_DATA_23_BASE_IDX 0 +#define regROM_SW_DATA_24 0x0103 +#define regROM_SW_DATA_24_BASE_IDX 0 +#define regROM_SW_DATA_25 0x0104 +#define regROM_SW_DATA_25_BASE_IDX 0 +#define regROM_SW_DATA_26 0x0105 +#define regROM_SW_DATA_26_BASE_IDX 0 +#define regROM_SW_DATA_27 0x0106 +#define regROM_SW_DATA_27_BASE_IDX 0 +#define regROM_SW_DATA_28 0x0107 +#define regROM_SW_DATA_28_BASE_IDX 0 +#define regROM_SW_DATA_29 0x0108 +#define regROM_SW_DATA_29_BASE_IDX 0 +#define regROM_SW_DATA_30 0x0109 +#define regROM_SW_DATA_30_BASE_IDX 0 +#define regROM_SW_DATA_31 0x010a +#define regROM_SW_DATA_31_BASE_IDX 0 +#define regROM_SW_DATA_32 0x010b +#define regROM_SW_DATA_32_BASE_IDX 0 +#define regROM_SW_DATA_33 0x010c +#define regROM_SW_DATA_33_BASE_IDX 0 +#define regROM_SW_DATA_34 0x010d +#define regROM_SW_DATA_34_BASE_IDX 0 +#define regROM_SW_DATA_35 0x010e +#define regROM_SW_DATA_35_BASE_IDX 0 +#define regROM_SW_DATA_36 0x010f +#define regROM_SW_DATA_36_BASE_IDX 0 +#define regROM_SW_DATA_37 0x0110 +#define regROM_SW_DATA_37_BASE_IDX 0 +#define regROM_SW_DATA_38 0x0111 +#define regROM_SW_DATA_38_BASE_IDX 0 +#define regROM_SW_DATA_39 0x0112 +#define regROM_SW_DATA_39_BASE_IDX 0 +#define regROM_SW_DATA_40 0x0113 +#define regROM_SW_DATA_40_BASE_IDX 0 +#define regROM_SW_DATA_41 0x0114 +#define regROM_SW_DATA_41_BASE_IDX 0 +#define regROM_SW_DATA_42 0x0115 +#define regROM_SW_DATA_42_BASE_IDX 0 +#define regROM_SW_DATA_43 0x0116 +#define regROM_SW_DATA_43_BASE_IDX 0 +#define regROM_SW_DATA_44 0x0117 +#define regROM_SW_DATA_44_BASE_IDX 0 +#define regROM_SW_DATA_45 0x0118 +#define regROM_SW_DATA_45_BASE_IDX 0 +#define regROM_SW_DATA_46 0x0119 +#define regROM_SW_DATA_46_BASE_IDX 0 +#define regROM_SW_DATA_47 0x011a +#define regROM_SW_DATA_47_BASE_IDX 0 +#define regROM_SW_DATA_48 0x011b +#define regROM_SW_DATA_48_BASE_IDX 0 +#define regROM_SW_DATA_49 0x011c +#define regROM_SW_DATA_49_BASE_IDX 0 +#define regROM_SW_DATA_50 0x011d +#define regROM_SW_DATA_50_BASE_IDX 0 +#define regROM_SW_DATA_51 0x011e +#define regROM_SW_DATA_51_BASE_IDX 0 +#define regROM_SW_DATA_52 0x011f +#define regROM_SW_DATA_52_BASE_IDX 0 +#define regROM_SW_DATA_53 0x0120 +#define regROM_SW_DATA_53_BASE_IDX 0 +#define regROM_SW_DATA_54 0x0121 +#define regROM_SW_DATA_54_BASE_IDX 0 +#define regROM_SW_DATA_55 0x0122 +#define regROM_SW_DATA_55_BASE_IDX 0 +#define regROM_SW_DATA_56 0x0123 +#define regROM_SW_DATA_56_BASE_IDX 0 +#define regROM_SW_DATA_57 0x0124 +#define regROM_SW_DATA_57_BASE_IDX 0 +#define regROM_SW_DATA_58 0x0125 +#define regROM_SW_DATA_58_BASE_IDX 0 +#define regROM_SW_DATA_59 0x0126 +#define regROM_SW_DATA_59_BASE_IDX 0 +#define regROM_SW_DATA_60 0x0127 +#define regROM_SW_DATA_60_BASE_IDX 0 +#define regROM_SW_DATA_61 0x0128 +#define regROM_SW_DATA_61_BASE_IDX 0 +#define regROM_SW_DATA_62 0x0129 +#define regROM_SW_DATA_62_BASE_IDX 0 +#define regROM_SW_DATA_63 0x012a +#define regROM_SW_DATA_63_BASE_IDX 0 +#define regROM_SW_DATA_64 0x012b +#define regROM_SW_DATA_64_BASE_IDX 0 + + +// addressBlock: smuio_smuio_gpio_SmuSmuioDec +// base address: 0x5a500 +#define regSMU_GPIOPAD_SW_INT_STAT 0x0140 +#define regSMU_GPIOPAD_SW_INT_STAT_BASE_IDX 0 +#define regSMU_GPIOPAD_MASK 0x0141 +#define regSMU_GPIOPAD_MASK_BASE_IDX 0 +#define regSMU_GPIOPAD_A 0x0142 +#define regSMU_GPIOPAD_A_BASE_IDX 0 +#define regSMU_GPIOPAD_TXIMPSEL 0x0143 +#define regSMU_GPIOPAD_TXIMPSEL_BASE_IDX 0 +#define regSMU_GPIOPAD_EN 0x0144 +#define regSMU_GPIOPAD_EN_BASE_IDX 0 +#define regSMU_GPIOPAD_Y 0x0145 +#define regSMU_GPIOPAD_Y_BASE_IDX 0 +#define regSMU_GPIOPAD_RXEN 0x0146 +#define regSMU_GPIOPAD_RXEN_BASE_IDX 0 +#define regSMU_GPIOPAD_RCVR_SEL0 0x0147 +#define regSMU_GPIOPAD_RCVR_SEL0_BASE_IDX 0 +#define regSMU_GPIOPAD_RCVR_SEL1 0x0148 +#define regSMU_GPIOPAD_RCVR_SEL1_BASE_IDX 0 +#define regSMU_GPIOPAD_PU_EN 0x0149 +#define regSMU_GPIOPAD_PU_EN_BASE_IDX 0 +#define regSMU_GPIOPAD_PD_EN 0x014a +#define regSMU_GPIOPAD_PD_EN_BASE_IDX 0 +#define regSMU_GPIOPAD_PINSTRAPS 0x014b +#define regSMU_GPIOPAD_PINSTRAPS_BASE_IDX 0 +#define regDFT_PINSTRAPS 0x014c +#define regDFT_PINSTRAPS_BASE_IDX 0 +#define regSMU_GPIOPAD_INT_STAT_EN 0x014d +#define regSMU_GPIOPAD_INT_STAT_EN_BASE_IDX 0 +#define regSMU_GPIOPAD_INT_STAT 0x014e +#define regSMU_GPIOPAD_INT_STAT_BASE_IDX 0 +#define regSMU_GPIOPAD_INT_STAT_AK 0x014f +#define regSMU_GPIOPAD_INT_STAT_AK_BASE_IDX 0 +#define regSMU_GPIOPAD_INT_EN 0x0150 +#define regSMU_GPIOPAD_INT_EN_BASE_IDX 0 +#define regSMU_GPIOPAD_INT_TYPE 0x0151 +#define regSMU_GPIOPAD_INT_TYPE_BASE_IDX 0 +#define regSMU_GPIOPAD_INT_POLARITY 0x0152 +#define regSMU_GPIOPAD_INT_POLARITY_BASE_IDX 0 +#define regSMUIO_PCC_GPIO_SELECT 0x0155 +#define regSMUIO_PCC_GPIO_SELECT_BASE_IDX 0 +#define regSMU_GPIOPAD_S0 0x0156 +#define regSMU_GPIOPAD_S0_BASE_IDX 0 +#define regSMU_GPIOPAD_S1 0x0157 +#define regSMU_GPIOPAD_S1_BASE_IDX 0 +#define regSMU_GPIOPAD_SCHMEN 0x0158 +#define regSMU_GPIOPAD_SCHMEN_BASE_IDX 0 +#define regSMU_GPIOPAD_SCL_EN 0x0159 +#define regSMU_GPIOPAD_SCL_EN_BASE_IDX 0 +#define regSMU_GPIOPAD_SDA_EN 0x015a +#define regSMU_GPIOPAD_SDA_EN_BASE_IDX 0 +#define regSMUIO_GPIO_INT0_SELECT 0x015b +#define regSMUIO_GPIO_INT0_SELECT_BASE_IDX 0 +#define regSMUIO_GPIO_INT1_SELECT 0x015c +#define regSMUIO_GPIO_INT1_SELECT_BASE_IDX 0 +#define regSMUIO_GPIO_INT2_SELECT 0x015d +#define regSMUIO_GPIO_INT2_SELECT_BASE_IDX 0 +#define regSMUIO_GPIO_INT3_SELECT 0x015e +#define regSMUIO_GPIO_INT3_SELECT_BASE_IDX 0 +#define regSMU_GPIOPAD_MP_INT0_STAT 0x015f +#define regSMU_GPIOPAD_MP_INT0_STAT_BASE_IDX 0 +#define regSMU_GPIOPAD_MP_INT1_STAT 0x0160 +#define regSMU_GPIOPAD_MP_INT1_STAT_BASE_IDX 0 +#define regSMU_GPIOPAD_MP_INT2_STAT 0x0161 +#define regSMU_GPIOPAD_MP_INT2_STAT_BASE_IDX 0 +#define regSMU_GPIOPAD_MP_INT3_STAT 0x0162 +#define regSMU_GPIOPAD_MP_INT3_STAT_BASE_IDX 0 +#define regSMIO_INDEX 0x0163 +#define regSMIO_INDEX_BASE_IDX 0 +#define regS0_VID_SMIO_CNTL 0x0164 +#define regS0_VID_SMIO_CNTL_BASE_IDX 0 +#define regS1_VID_SMIO_CNTL 0x0165 +#define regS1_VID_SMIO_CNTL_BASE_IDX 0 +#define regOPEN_DRAIN_SELECT 0x0166 +#define regOPEN_DRAIN_SELECT_BASE_IDX 0 +#define regSMIO_ENABLE 0x0167 +#define regSMIO_ENABLE_BASE_IDX 0 + +#endif diff --git a/drivers/gpu/drm/amd/include/asic_reg/smuio/smuio_14_0_2_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/smuio/smuio_14_0_2_sh_mask.h new file mode 100644 index 000000000000..6204505e553b --- /dev/null +++ b/drivers/gpu/drm/amd/include/asic_reg/smuio/smuio_14_0_2_sh_mask.h @@ -0,0 +1,1106 @@ +/* + * Copyright 2023 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ +#ifndef _smuio_14_0_2_SH_MASK_HEADER +#define _smuio_14_0_2_SH_MASK_HEADER + + +// addressBlock: smuio_smuio_tsc_SmuSmuioDec +//PWROK_REFCLK_GAP_CYCLES +#define PWROK_REFCLK_GAP_CYCLES__Pwrok_PreAssertion_clkgap_cycles__SHIFT 0x0 +#define PWROK_REFCLK_GAP_CYCLES__Pwrok_PostAssertion_clkgap_cycles__SHIFT 0x8 +#define PWROK_REFCLK_GAP_CYCLES__Pwrok_PreAssertion_clkgap_cycles_MASK 0x000000FFL +#define PWROK_REFCLK_GAP_CYCLES__Pwrok_PostAssertion_clkgap_cycles_MASK 0x0000FF00L +//GOLDEN_TSC_INCREMENT_UPPER +#define GOLDEN_TSC_INCREMENT_UPPER__GoldenTscIncrementUpper__SHIFT 0x0 +#define GOLDEN_TSC_INCREMENT_UPPER__GoldenTscIncrementUpper_MASK 0x00FFFFFFL +//GOLDEN_TSC_INCREMENT_LOWER +#define GOLDEN_TSC_INCREMENT_LOWER__GoldenTscIncrementLower__SHIFT 0x0 +#define GOLDEN_TSC_INCREMENT_LOWER__GoldenTscIncrementLower_MASK 0xFFFFFFFFL +//GOLDEN_TSC_COUNT_UPPER +#define GOLDEN_TSC_COUNT_UPPER__GoldenTscCountUpper__SHIFT 0x0 +#define GOLDEN_TSC_COUNT_UPPER__GoldenTscCountUpper_MASK 0x00FFFFFFL +//GOLDEN_TSC_COUNT_LOWER +#define GOLDEN_TSC_COUNT_LOWER__GoldenTscCountLower__SHIFT 0x0 +#define GOLDEN_TSC_COUNT_LOWER__GoldenTscCountLower_MASK 0xFFFFFFFFL +//SOC_GOLDEN_TSC_SHADOW_UPPER +#define SOC_GOLDEN_TSC_SHADOW_UPPER__SocGoldenTscShadowUpper__SHIFT 0x0 +#define SOC_GOLDEN_TSC_SHADOW_UPPER__SocGoldenTscShadowUpper_MASK 0x00FFFFFFL +//SOC_GOLDEN_TSC_SHADOW_LOWER +#define SOC_GOLDEN_TSC_SHADOW_LOWER__SocGoldenTscShadowLower__SHIFT 0x0 +#define SOC_GOLDEN_TSC_SHADOW_LOWER__SocGoldenTscShadowLower_MASK 0xFFFFFFFFL +//SOC_GAP_PWROK +#define SOC_GAP_PWROK__soc_gap_pwrok__SHIFT 0x0 +#define SOC_GAP_PWROK__soc_gap_pwrok_MASK 0x00000001L + + +// addressBlock: smuio_smuio_swtimer_SmuSmuioDec +//PWR_VIRT_RESET_REQ +#define PWR_VIRT_RESET_REQ__VF_FLR__SHIFT 0x0 +#define PWR_VIRT_RESET_REQ__PF_FLR__SHIFT 0x1f +#define PWR_VIRT_RESET_REQ__VF_FLR_MASK 0x7FFFFFFFL +#define PWR_VIRT_RESET_REQ__PF_FLR_MASK 0x80000000L +//PWR_DISP_TIMER_CONTROL +#define PWR_DISP_TIMER_CONTROL__DISP_TIMER_INT_COUNT__SHIFT 0x0 +#define PWR_DISP_TIMER_CONTROL__DISP_TIMER_INT_ENABLE__SHIFT 0x19 +#define PWR_DISP_TIMER_CONTROL__DISP_TIMER_INT_DISABLE__SHIFT 0x1a +#define PWR_DISP_TIMER_CONTROL__DISP_TIMER_INT_MASK__SHIFT 0x1b +#define PWR_DISP_TIMER_CONTROL__DISP_TIMER_INT_STAT_AK__SHIFT 0x1c +#define PWR_DISP_TIMER_CONTROL__DISP_TIMER_INT_TYPE__SHIFT 0x1d +#define PWR_DISP_TIMER_CONTROL__DISP_TIMER_INT_MODE__SHIFT 0x1e +#define PWR_DISP_TIMER_CONTROL__DISP_TIMER_INT_COUNT_MASK 0x01FFFFFFL +#define PWR_DISP_TIMER_CONTROL__DISP_TIMER_INT_ENABLE_MASK 0x02000000L +#define PWR_DISP_TIMER_CONTROL__DISP_TIMER_INT_DISABLE_MASK 0x04000000L +#define PWR_DISP_TIMER_CONTROL__DISP_TIMER_INT_MASK_MASK 0x08000000L +#define PWR_DISP_TIMER_CONTROL__DISP_TIMER_INT_STAT_AK_MASK 0x10000000L +#define PWR_DISP_TIMER_CONTROL__DISP_TIMER_INT_TYPE_MASK 0x20000000L +#define PWR_DISP_TIMER_CONTROL__DISP_TIMER_INT_MODE_MASK 0x40000000L +//PWR_DISP_TIMER_DEBUG +#define PWR_DISP_TIMER_DEBUG__DISP_TIMER_INT_RUNNING__SHIFT 0x0 +#define PWR_DISP_TIMER_DEBUG__DISP_TIMER_INT_STAT__SHIFT 0x1 +#define PWR_DISP_TIMER_DEBUG__DISP_TIMER_INT__SHIFT 0x2 +#define PWR_DISP_TIMER_DEBUG__DISP_TIMER_RUN_VAL__SHIFT 0x7 +#define PWR_DISP_TIMER_DEBUG__DISP_TIMER_INT_RUNNING_MASK 0x00000001L +#define PWR_DISP_TIMER_DEBUG__DISP_TIMER_INT_STAT_MASK 0x00000002L +#define PWR_DISP_TIMER_DEBUG__DISP_TIMER_INT_MASK 0x00000004L +#define PWR_DISP_TIMER_DEBUG__DISP_TIMER_RUN_VAL_MASK 0xFFFFFF80L +//PWR_DISP_TIMER2_CONTROL +#define PWR_DISP_TIMER2_CONTROL__DISP_TIMER_INT_COUNT__SHIFT 0x0 +#define PWR_DISP_TIMER2_CONTROL__DISP_TIMER_INT_ENABLE__SHIFT 0x19 +#define PWR_DISP_TIMER2_CONTROL__DISP_TIMER_INT_DISABLE__SHIFT 0x1a +#define PWR_DISP_TIMER2_CONTROL__DISP_TIMER_INT_MASK__SHIFT 0x1b +#define PWR_DISP_TIMER2_CONTROL__DISP_TIMER_INT_STAT_AK__SHIFT 0x1c +#define PWR_DISP_TIMER2_CONTROL__DISP_TIMER_INT_TYPE__SHIFT 0x1d +#define PWR_DISP_TIMER2_CONTROL__DISP_TIMER_INT_MODE__SHIFT 0x1e +#define PWR_DISP_TIMER2_CONTROL__DISP_TIMER_INT_COUNT_MASK 0x01FFFFFFL +#define PWR_DISP_TIMER2_CONTROL__DISP_TIMER_INT_ENABLE_MASK 0x02000000L +#define PWR_DISP_TIMER2_CONTROL__DISP_TIMER_INT_DISABLE_MASK 0x04000000L +#define PWR_DISP_TIMER2_CONTROL__DISP_TIMER_INT_MASK_MASK 0x08000000L +#define PWR_DISP_TIMER2_CONTROL__DISP_TIMER_INT_STAT_AK_MASK 0x10000000L +#define PWR_DISP_TIMER2_CONTROL__DISP_TIMER_INT_TYPE_MASK 0x20000000L +#define PWR_DISP_TIMER2_CONTROL__DISP_TIMER_INT_MODE_MASK 0x40000000L +//PWR_DISP_TIMER2_DEBUG +#define PWR_DISP_TIMER2_DEBUG__DISP_TIMER_INT_RUNNING__SHIFT 0x0 +#define PWR_DISP_TIMER2_DEBUG__DISP_TIMER_INT_STAT__SHIFT 0x1 +#define PWR_DISP_TIMER2_DEBUG__DISP_TIMER_INT__SHIFT 0x2 +#define PWR_DISP_TIMER2_DEBUG__DISP_TIMER_RUN_VAL__SHIFT 0x7 +#define PWR_DISP_TIMER2_DEBUG__DISP_TIMER_INT_RUNNING_MASK 0x00000001L +#define PWR_DISP_TIMER2_DEBUG__DISP_TIMER_INT_STAT_MASK 0x00000002L +#define PWR_DISP_TIMER2_DEBUG__DISP_TIMER_INT_MASK 0x00000004L +#define PWR_DISP_TIMER2_DEBUG__DISP_TIMER_RUN_VAL_MASK 0xFFFFFF80L +//PWR_DISP_TIMER_GLOBAL_CONTROL +#define PWR_DISP_TIMER_GLOBAL_CONTROL__DISP_TIMER_PULSE_WIDTH__SHIFT 0x0 +#define PWR_DISP_TIMER_GLOBAL_CONTROL__DISP_TIMER_PULSE_EN__SHIFT 0xa +#define PWR_DISP_TIMER_GLOBAL_CONTROL__DISP_TIMER_PULSE_WIDTH_MASK 0x000003FFL +#define PWR_DISP_TIMER_GLOBAL_CONTROL__DISP_TIMER_PULSE_EN_MASK 0x00000400L +//PWR_IH_CONTROL +#define PWR_IH_CONTROL__MAX_CREDIT__SHIFT 0x0 +#define PWR_IH_CONTROL__DISP_TIMER_TRIGGER_MASK__SHIFT 0x5 +#define PWR_IH_CONTROL__DISP_TIMER2_TRIGGER_MASK__SHIFT 0x6 +#define PWR_IH_CONTROL__PWR_IH_CLK_GATE_EN__SHIFT 0x1f +#define PWR_IH_CONTROL__MAX_CREDIT_MASK 0x0000001FL +#define PWR_IH_CONTROL__DISP_TIMER_TRIGGER_MASK_MASK 0x00000020L +#define PWR_IH_CONTROL__DISP_TIMER2_TRIGGER_MASK_MASK 0x00000040L +#define PWR_IH_CONTROL__PWR_IH_CLK_GATE_EN_MASK 0x80000000L + + +// addressBlock: smuio_smuio_misc_SmuSmuioDec +//SMUIO_MCM_CONFIG +#define SMUIO_MCM_CONFIG__DIE_ID__SHIFT 0x0 +#define SMUIO_MCM_CONFIG__PKG_TYPE__SHIFT 0x2 +#define SMUIO_MCM_CONFIG__SOCKET_ID__SHIFT 0x8 +#define SMUIO_MCM_CONFIG__PKG_SUBTYPE__SHIFT 0xc +#define SMUIO_MCM_CONFIG__DIE_CONFIG__SHIFT 0xd +#define SMUIO_MCM_CONFIG__CONSOLE_K__SHIFT 0x10 +#define SMUIO_MCM_CONFIG__CONSOLE_A__SHIFT 0x11 +#define SMUIO_MCM_CONFIG__DIE_ID_MASK 0x00000003L +#define SMUIO_MCM_CONFIG__PKG_TYPE_MASK 0x0000001CL +#define SMUIO_MCM_CONFIG__SOCKET_ID_MASK 0x00000300L +#define SMUIO_MCM_CONFIG__PKG_SUBTYPE_MASK 0x00001000L +#define SMUIO_MCM_CONFIG__CONSOLE_K_MASK 0x00010000L +#define SMUIO_MCM_CONFIG__CONSOLE_A_MASK 0x00020000L +//IP_DISCOVERY_VERSION +#define IP_DISCOVERY_VERSION__IP_DISCOVERY_VERSION__SHIFT 0x0 +#define IP_DISCOVERY_VERSION__IP_DISCOVERY_VERSION_MASK 0xFFFFFFFFL +//SCRATCH_REGISTER0 +#define SCRATCH_REGISTER0__ScratchPad0__SHIFT 0x0 +#define SCRATCH_REGISTER0__ScratchPad0_MASK 0xFFFFFFFFL +//SCRATCH_REGISTER1 +#define SCRATCH_REGISTER1__ScratchPad1__SHIFT 0x0 +#define SCRATCH_REGISTER1__ScratchPad1_MASK 0xFFFFFFFFL +//SCRATCH_REGISTER2 +#define SCRATCH_REGISTER2__ScratchPad2__SHIFT 0x0 +#define SCRATCH_REGISTER2__ScratchPad2_MASK 0xFFFFFFFFL +//SCRATCH_REGISTER3 +#define SCRATCH_REGISTER3__ScratchPad3__SHIFT 0x0 +#define SCRATCH_REGISTER3__ScratchPad3_MASK 0xFFFFFFFFL +//SCRATCH_REGISTER4 +#define SCRATCH_REGISTER4__ScratchPad4__SHIFT 0x0 +#define SCRATCH_REGISTER4__ScratchPad4_MASK 0xFFFFFFFFL +//SCRATCH_REGISTER5 +#define SCRATCH_REGISTER5__ScratchPad5__SHIFT 0x0 +#define SCRATCH_REGISTER5__ScratchPad5_MASK 0xFFFFFFFFL +//SCRATCH_REGISTER6 +#define SCRATCH_REGISTER6__ScratchPad6__SHIFT 0x0 +#define SCRATCH_REGISTER6__ScratchPad6_MASK 0xFFFFFFFFL +//SCRATCH_REGISTER7 +#define SCRATCH_REGISTER7__ScratchPad7__SHIFT 0x0 +#define SCRATCH_REGISTER7__ScratchPad7_MASK 0xFFFFFFFFL + + +// addressBlock: smuio_smuio_i2c_SmuSmuioDec +//CKSVII2C_IC_CON +#define CKSVII2C_IC_CON__IC_MASTER_MODE__SHIFT 0x0 +#define CKSVII2C_IC_CON__IC_MAX_SPEED_MODE__SHIFT 0x1 +#define CKSVII2C_IC_CON__IC_10BITADDR_SLAVE__SHIFT 0x3 +#define CKSVII2C_IC_CON__IC_10BITADDR_MASTER__SHIFT 0x4 +#define CKSVII2C_IC_CON__IC_RESTART_EN__SHIFT 0x5 +#define CKSVII2C_IC_CON__IC_SLAVE_DISABLE__SHIFT 0x6 +#define CKSVII2C_IC_CON__STOP_DET_IFADDRESSED__SHIFT 0x7 +#define CKSVII2C_IC_CON__TX_EMPTY_CTRL__SHIFT 0x8 +#define CKSVII2C_IC_CON__RX_FIFO_FULL_HLD_CTRL__SHIFT 0x9 +#define CKSVII2C_IC_CON__BUS_CLEAR_FEATURE_CTRL__SHIFT 0xb +#define CKSVII2C_IC_CON__IC_MASTER_MODE_MASK 0x00000001L +#define CKSVII2C_IC_CON__IC_MAX_SPEED_MODE_MASK 0x00000006L +#define CKSVII2C_IC_CON__IC_10BITADDR_SLAVE_MASK 0x00000008L +#define CKSVII2C_IC_CON__IC_10BITADDR_MASTER_MASK 0x00000010L +#define CKSVII2C_IC_CON__IC_RESTART_EN_MASK 0x00000020L +#define CKSVII2C_IC_CON__IC_SLAVE_DISABLE_MASK 0x00000040L +#define CKSVII2C_IC_CON__STOP_DET_IFADDRESSED_MASK 0x00000080L +#define CKSVII2C_IC_CON__TX_EMPTY_CTRL_MASK 0x00000100L +#define CKSVII2C_IC_CON__RX_FIFO_FULL_HLD_CTRL_MASK 0x00000200L +//CKSVII2C_IC_TAR +#define CKSVII2C_IC_TAR__IC_TAR__SHIFT 0x0 +#define CKSVII2C_IC_TAR__GC_OR_START__SHIFT 0xa +#define CKSVII2C_IC_TAR__SPECIAL__SHIFT 0xb +#define CKSVII2C_IC_TAR__IC_10BITADDR_MASTER__SHIFT 0xc +#define CKSVII2C_IC_TAR__IC_TAR_MASK 0x000003FFL +#define CKSVII2C_IC_TAR__GC_OR_START_MASK 0x00000400L +#define CKSVII2C_IC_TAR__SPECIAL_MASK 0x00000800L +#define CKSVII2C_IC_TAR__IC_10BITADDR_MASTER_MASK 0x00001000L +//CKSVII2C_IC_SAR +#define CKSVII2C_IC_SAR__IC_SAR__SHIFT 0x0 +#define CKSVII2C_IC_SAR__IC_SAR_MASK 0x000003FFL +//CKSVII2C_IC_HS_MADDR +#define CKSVII2C_IC_HS_MADDR__IC_HS_MADDR__SHIFT 0x0 +#define CKSVII2C_IC_HS_MADDR__IC_HS_MADDR_MASK 0x00000007L +//CKSVII2C_IC_DATA_CMD +#define CKSVII2C_IC_DATA_CMD__DAT__SHIFT 0x0 +#define CKSVII2C_IC_DATA_CMD__CMD__SHIFT 0x8 +#define CKSVII2C_IC_DATA_CMD__STOP__SHIFT 0x9 +#define CKSVII2C_IC_DATA_CMD__RESTART__SHIFT 0xa +#define CKSVII2C_IC_DATA_CMD__FIRST_DATA_BYTE__SHIFT 0xb +#define CKSVII2C_IC_DATA_CMD__DAT_MASK 0x000000FFL +#define CKSVII2C_IC_DATA_CMD__CMD_MASK 0x00000100L +#define CKSVII2C_IC_DATA_CMD__STOP_MASK 0x00000200L +#define CKSVII2C_IC_DATA_CMD__RESTART_MASK 0x00000400L +//CKSVII2C_IC_SS_SCL_HCNT +#define CKSVII2C_IC_SS_SCL_HCNT__IC_SS_SCL_HCNT__SHIFT 0x0 +#define CKSVII2C_IC_SS_SCL_HCNT__IC_SS_SCL_HCNT_MASK 0x0000FFFFL +//CKSVII2C_IC_SS_SCL_LCNT +#define CKSVII2C_IC_SS_SCL_LCNT__IC_SS_SCL_LCNT__SHIFT 0x0 +#define CKSVII2C_IC_SS_SCL_LCNT__IC_SS_SCL_LCNT_MASK 0x0000FFFFL +//CKSVII2C_IC_FS_SCL_HCNT +#define CKSVII2C_IC_FS_SCL_HCNT__IC_FS_SCL_HCNT__SHIFT 0x0 +#define CKSVII2C_IC_FS_SCL_HCNT__IC_FS_SCL_HCNT_MASK 0x0000FFFFL +//CKSVII2C_IC_FS_SCL_LCNT +#define CKSVII2C_IC_FS_SCL_LCNT__IC_FS_SCL_LCNT__SHIFT 0x0 +#define CKSVII2C_IC_FS_SCL_LCNT__IC_FS_SCL_LCNT_MASK 0x0000FFFFL +//CKSVII2C_IC_HS_SCL_HCNT +#define CKSVII2C_IC_HS_SCL_HCNT__IC_HS_SCL_HCNT__SHIFT 0x0 +#define CKSVII2C_IC_HS_SCL_HCNT__IC_HS_SCL_HCNT_MASK 0x0000FFFFL +//CKSVII2C_IC_HS_SCL_LCNT +#define CKSVII2C_IC_HS_SCL_LCNT__IC_HS_SCL_LCNT__SHIFT 0x0 +#define CKSVII2C_IC_HS_SCL_LCNT__IC_HS_SCL_LCNT_MASK 0x0000FFFFL +//CKSVII2C_IC_INTR_STAT +#define CKSVII2C_IC_INTR_STAT__R_RX_UNDER__SHIFT 0x0 +#define CKSVII2C_IC_INTR_STAT__R_RX_OVER__SHIFT 0x1 +#define CKSVII2C_IC_INTR_STAT__R_RX_FULL__SHIFT 0x2 +#define CKSVII2C_IC_INTR_STAT__R_TX_OVER__SHIFT 0x3 +#define CKSVII2C_IC_INTR_STAT__R_TX_EMPTY__SHIFT 0x4 +#define CKSVII2C_IC_INTR_STAT__R_RD_REQ__SHIFT 0x5 +#define CKSVII2C_IC_INTR_STAT__R_TX_ABRT__SHIFT 0x6 +#define CKSVII2C_IC_INTR_STAT__R_RX_DONE__SHIFT 0x7 +#define CKSVII2C_IC_INTR_STAT__R_ACTIVITY__SHIFT 0x8 +#define CKSVII2C_IC_INTR_STAT__R_STOP_DET__SHIFT 0x9 +#define CKSVII2C_IC_INTR_STAT__R_START_DET__SHIFT 0xa +#define CKSVII2C_IC_INTR_STAT__R_GEN_CALL__SHIFT 0xb +#define CKSVII2C_IC_INTR_STAT__R_RESTART_DET__SHIFT 0xc +#define CKSVII2C_IC_INTR_STAT__R_MST_ON_HOLD__SHIFT 0xd +#define CKSVII2C_IC_INTR_STAT__R_SCL_STUCK_AT_LOW__SHIFT 0xe +#define CKSVII2C_IC_INTR_STAT__R_RX_UNDER_MASK 0x00000001L +#define CKSVII2C_IC_INTR_STAT__R_RX_OVER_MASK 0x00000002L +#define CKSVII2C_IC_INTR_STAT__R_RX_FULL_MASK 0x00000004L +#define CKSVII2C_IC_INTR_STAT__R_TX_OVER_MASK 0x00000008L +#define CKSVII2C_IC_INTR_STAT__R_TX_EMPTY_MASK 0x00000010L +#define CKSVII2C_IC_INTR_STAT__R_RD_REQ_MASK 0x00000020L +#define CKSVII2C_IC_INTR_STAT__R_TX_ABRT_MASK 0x00000040L +#define CKSVII2C_IC_INTR_STAT__R_RX_DONE_MASK 0x00000080L +#define CKSVII2C_IC_INTR_STAT__R_ACTIVITY_MASK 0x00000100L +#define CKSVII2C_IC_INTR_STAT__R_STOP_DET_MASK 0x00000200L +#define CKSVII2C_IC_INTR_STAT__R_START_DET_MASK 0x00000400L +#define CKSVII2C_IC_INTR_STAT__R_GEN_CALL_MASK 0x00000800L +#define CKSVII2C_IC_INTR_STAT__R_RESTART_DET_MASK 0x00001000L +#define CKSVII2C_IC_INTR_STAT__R_MST_ON_HOLD_MASK 0x00002000L +//CKSVII2C_IC_INTR_MASK +#define CKSVII2C_IC_INTR_MASK__M_RX_UNDER__SHIFT 0x0 +#define CKSVII2C_IC_INTR_MASK__M_RX_OVER__SHIFT 0x1 +#define CKSVII2C_IC_INTR_MASK__M_RX_FULL__SHIFT 0x2 +#define CKSVII2C_IC_INTR_MASK__M_TX_OVER__SHIFT 0x3 +#define CKSVII2C_IC_INTR_MASK__M_TX_EMPTY__SHIFT 0x4 +#define CKSVII2C_IC_INTR_MASK__M_RD_REQ__SHIFT 0x5 +#define CKSVII2C_IC_INTR_MASK__M_TX_ABRT__SHIFT 0x6 +#define CKSVII2C_IC_INTR_MASK__M_RX_DONE__SHIFT 0x7 +#define CKSVII2C_IC_INTR_MASK__M_ACTIVITY__SHIFT 0x8 +#define CKSVII2C_IC_INTR_MASK__M_STOP_DET__SHIFT 0x9 +#define CKSVII2C_IC_INTR_MASK__M_START_DET__SHIFT 0xa +#define CKSVII2C_IC_INTR_MASK__M_GEN_CALL__SHIFT 0xb +#define CKSVII2C_IC_INTR_MASK__M_RESTART_DET__SHIFT 0xc +#define CKSVII2C_IC_INTR_MASK__M_MST_ON_HOLD__SHIFT 0xd +#define CKSVII2C_IC_INTR_MASK__M_SCL_STUCK_AT_LOW__SHIFT 0xe +#define CKSVII2C_IC_INTR_MASK__M_RX_UNDER_MASK 0x00000001L +#define CKSVII2C_IC_INTR_MASK__M_RX_OVER_MASK 0x00000002L +#define CKSVII2C_IC_INTR_MASK__M_RX_FULL_MASK 0x00000004L +#define CKSVII2C_IC_INTR_MASK__M_TX_OVER_MASK 0x00000008L +#define CKSVII2C_IC_INTR_MASK__M_TX_EMPTY_MASK 0x00000010L +#define CKSVII2C_IC_INTR_MASK__M_RD_REQ_MASK 0x00000020L +#define CKSVII2C_IC_INTR_MASK__M_TX_ABRT_MASK 0x00000040L +#define CKSVII2C_IC_INTR_MASK__M_RX_DONE_MASK 0x00000080L +#define CKSVII2C_IC_INTR_MASK__M_ACTIVITY_MASK 0x00000100L +#define CKSVII2C_IC_INTR_MASK__M_STOP_DET_MASK 0x00000200L +#define CKSVII2C_IC_INTR_MASK__M_START_DET_MASK 0x00000400L +#define CKSVII2C_IC_INTR_MASK__M_GEN_CALL_MASK 0x00000800L +#define CKSVII2C_IC_INTR_MASK__M_RESTART_DET_MASK 0x00001000L +#define CKSVII2C_IC_INTR_MASK__M_MST_ON_HOLD_MASK 0x00002000L +//CKSVII2C_IC_RAW_INTR_STAT +//CKSVII2C_IC_RX_TL +#define CKSVII2C_IC_RX_TL__RX_TL__SHIFT 0x0 +//CKSVII2C_IC_TX_TL +#define CKSVII2C_IC_TX_TL__TX_TL__SHIFT 0x0 +//CKSVII2C_IC_CLR_INTR +//CKSVII2C_IC_CLR_RX_UNDER +//CKSVII2C_IC_CLR_RX_OVER +//CKSVII2C_IC_CLR_TX_OVER +//CKSVII2C_IC_CLR_RD_REQ +//CKSVII2C_IC_CLR_TX_ABRT +//CKSVII2C_IC_CLR_RX_DONE +//CKSVII2C_IC_CLR_ACTIVITY +//CKSVII2C_IC_CLR_STOP_DET +//CKSVII2C_IC_CLR_START_DET +//CKSVII2C_IC_CLR_GEN_CALL +//CKSVII2C_IC_ENABLE +#define CKSVII2C_IC_ENABLE__ENABLE__SHIFT 0x0 +#define CKSVII2C_IC_ENABLE__ABORT__SHIFT 0x1 +#define CKSVII2C_IC_ENABLE__TX_CMD_BLOCK__SHIFT 0x2 +#define CKSVII2C_IC_ENABLE__SDA_STUCK_RECOVERY_ENABLE__SHIFT 0x3 +#define CKSVII2C_IC_ENABLE__ENABLE_MASK 0x00000001L +#define CKSVII2C_IC_ENABLE__ABORT_MASK 0x00000002L +//CKSVII2C_IC_STATUS +#define CKSVII2C_IC_STATUS__ACTIVITY__SHIFT 0x0 +#define CKSVII2C_IC_STATUS__TFNF__SHIFT 0x1 +#define CKSVII2C_IC_STATUS__TFE__SHIFT 0x2 +#define CKSVII2C_IC_STATUS__RFNE__SHIFT 0x3 +#define CKSVII2C_IC_STATUS__RFF__SHIFT 0x4 +#define CKSVII2C_IC_STATUS__MST_ACTIVITY__SHIFT 0x5 +#define CKSVII2C_IC_STATUS__SLV_ACTIVITY__SHIFT 0x6 +#define CKSVII2C_IC_STATUS__MST_HOLD_TX_FIFO_EMPTY__SHIFT 0x7 +#define CKSVII2C_IC_STATUS__MST_HOLD_RX_FIFO_FULL__SHIFT 0x8 +#define CKSVII2C_IC_STATUS__SLV_HOLD_TX_FIFO_EMPTY__SHIFT 0x9 +#define CKSVII2C_IC_STATUS__SLV_HOLD_RX_FIFO_FULL__SHIFT 0xa +#define CKSVII2C_IC_STATUS__SDA_STUCK_NOT_RECOVERED__SHIFT 0xb +#define CKSVII2C_IC_STATUS__ACTIVITY_MASK 0x00000001L +#define CKSVII2C_IC_STATUS__TFNF_MASK 0x00000002L +#define CKSVII2C_IC_STATUS__TFE_MASK 0x00000004L +#define CKSVII2C_IC_STATUS__RFNE_MASK 0x00000008L +#define CKSVII2C_IC_STATUS__RFF_MASK 0x00000010L +#define CKSVII2C_IC_STATUS__MST_ACTIVITY_MASK 0x00000020L +#define CKSVII2C_IC_STATUS__SLV_ACTIVITY_MASK 0x00000040L +//CKSVII2C_IC_TXFLR +#define CKSVII2C_IC_TXFLR__TXFLR__SHIFT 0x0 +//CKSVII2C_IC_RXFLR +#define CKSVII2C_IC_RXFLR__RXFLR__SHIFT 0x0 +//CKSVII2C_IC_SDA_HOLD +#define CKSVII2C_IC_SDA_HOLD__IC_SDA_TX_HOLD__SHIFT 0x0 +#define CKSVII2C_IC_SDA_HOLD__IC_SDA_RX_HOLD__SHIFT 0x10 +//CKSVII2C_IC_TX_ABRT_SOURCE +//CKSVII2C_IC_SLV_DATA_NACK_ONLY +//CKSVII2C_IC_DMA_CR +//CKSVII2C_IC_DMA_TDLR +//CKSVII2C_IC_DMA_RDLR +//CKSVII2C_IC_SDA_SETUP +#define CKSVII2C_IC_SDA_SETUP__SDA_SETUP__SHIFT 0x0 +#define CKSVII2C_IC_SDA_SETUP__SDA_SETUP_MASK 0x000000FFL +//CKSVII2C_IC_ACK_GENERAL_CALL +#define CKSVII2C_IC_ACK_GENERAL_CALL__ACK_GENERAL_CALL__SHIFT 0x0 +#define CKSVII2C_IC_ACK_GENERAL_CALL__ACK_GENERAL_CALL_MASK 0x00000001L +//CKSVII2C_IC_ENABLE_STATUS +#define CKSVII2C_IC_ENABLE_STATUS__IC_EN__SHIFT 0x0 +#define CKSVII2C_IC_ENABLE_STATUS__SLV_DISABLED_WHILE_BUSY__SHIFT 0x1 +#define CKSVII2C_IC_ENABLE_STATUS__SLV_RX_DATA_LOST__SHIFT 0x2 +#define CKSVII2C_IC_ENABLE_STATUS__IC_EN_MASK 0x00000001L +//CKSVII2C_IC_FS_SPKLEN +#define CKSVII2C_IC_FS_SPKLEN__FS_SPKLEN__SHIFT 0x0 +#define CKSVII2C_IC_FS_SPKLEN__FS_SPKLEN_MASK 0x000000FFL +//CKSVII2C_IC_HS_SPKLEN +#define CKSVII2C_IC_HS_SPKLEN__HS_SPKLEN__SHIFT 0x0 +#define CKSVII2C_IC_HS_SPKLEN__HS_SPKLEN_MASK 0x000000FFL +//CKSVII2C_IC_CLR_RESTART_DET +//CKSVII2C_IC_COMP_PARAM_1 +#define CKSVII2C_IC_COMP_PARAM_1__APB_DATA_WIDTH__SHIFT 0x0 +#define CKSVII2C_IC_COMP_PARAM_1__MAX_SPEED_MODE__SHIFT 0x2 +#define CKSVII2C_IC_COMP_PARAM_1__HC_COUNT_VALUES__SHIFT 0x4 +#define CKSVII2C_IC_COMP_PARAM_1__INTR_IO__SHIFT 0x5 +#define CKSVII2C_IC_COMP_PARAM_1__HAS_DMA__SHIFT 0x6 +#define CKSVII2C_IC_COMP_PARAM_1__ADD_ENCODED_PARAMS__SHIFT 0x7 +#define CKSVII2C_IC_COMP_PARAM_1__RX_BUFFER_DEPTH__SHIFT 0x8 +#define CKSVII2C_IC_COMP_PARAM_1__TX_BUFFER_DEPTH__SHIFT 0x10 +//CKSVII2C_IC_COMP_VERSION +#define CKSVII2C_IC_COMP_VERSION__COMP_VERSION__SHIFT 0x0 +#define CKSVII2C_IC_COMP_VERSION__COMP_VERSION_MASK 0xFFFFFFFFL +//CKSVII2C_IC_COMP_TYPE +#define CKSVII2C_IC_COMP_TYPE__COMP_TYPE__SHIFT 0x0 +#define CKSVII2C_IC_COMP_TYPE__COMP_TYPE_MASK 0xFFFFFFFFL +//CKSVII2C1_IC_CON +#define CKSVII2C1_IC_CON__IC1_MASTER_MODE__SHIFT 0x0 +#define CKSVII2C1_IC_CON__IC1_MAX_SPEED_MODE__SHIFT 0x1 +#define CKSVII2C1_IC_CON__IC1_10BITADDR_SLAVE__SHIFT 0x3 +#define CKSVII2C1_IC_CON__IC1_10BITADDR_MASTER__SHIFT 0x4 +#define CKSVII2C1_IC_CON__IC1_RESTART_EN__SHIFT 0x5 +#define CKSVII2C1_IC_CON__IC1_SLAVE_DISABLE__SHIFT 0x6 +#define CKSVII2C1_IC_CON__STOP1_DET_IFADDRESSED__SHIFT 0x7 +#define CKSVII2C1_IC_CON__TX1_EMPTY_CTRL__SHIFT 0x8 +#define CKSVII2C1_IC_CON__RX1_FIFO_FULL_HLD_CTRL__SHIFT 0x9 +#define CKSVII2C1_IC_CON__BUS_CLEAR_FEATURE_CTRL1__SHIFT 0xb +#define CKSVII2C1_IC_CON__IC1_MASTER_MODE_MASK 0x00000001L +#define CKSVII2C1_IC_CON__IC1_MAX_SPEED_MODE_MASK 0x00000006L +#define CKSVII2C1_IC_CON__IC1_10BITADDR_SLAVE_MASK 0x00000008L +#define CKSVII2C1_IC_CON__IC1_10BITADDR_MASTER_MASK 0x00000010L +#define CKSVII2C1_IC_CON__IC1_RESTART_EN_MASK 0x00000020L +#define CKSVII2C1_IC_CON__IC1_SLAVE_DISABLE_MASK 0x00000040L +#define CKSVII2C1_IC_CON__STOP1_DET_IFADDRESSED_MASK 0x00000080L +#define CKSVII2C1_IC_CON__TX1_EMPTY_CTRL_MASK 0x00000100L +#define CKSVII2C1_IC_CON__RX1_FIFO_FULL_HLD_CTRL_MASK 0x00000200L +//CKSVII2C1_IC_TAR +#define CKSVII2C1_IC_TAR__IC1_TAR__SHIFT 0x0 +#define CKSVII2C1_IC_TAR__GC1_OR_START__SHIFT 0xa +#define CKSVII2C1_IC_TAR__SPECIAL1__SHIFT 0xb +#define CKSVII2C1_IC_TAR__IC1_10BITADDR_MASTER__SHIFT 0xc +#define CKSVII2C1_IC_TAR__IC1_TAR_MASK 0x000003FFL +#define CKSVII2C1_IC_TAR__GC1_OR_START_MASK 0x00000400L +#define CKSVII2C1_IC_TAR__SPECIAL1_MASK 0x00000800L +#define CKSVII2C1_IC_TAR__IC1_10BITADDR_MASTER_MASK 0x00001000L +//CKSVII2C1_IC_SAR +#define CKSVII2C1_IC_SAR__IC1_SAR__SHIFT 0x0 +#define CKSVII2C1_IC_SAR__IC1_SAR_MASK 0x000003FFL +//CKSVII2C1_IC_HS_MADDR +#define CKSVII2C1_IC_HS_MADDR__IC1_HS_MADDR__SHIFT 0x0 +#define CKSVII2C1_IC_HS_MADDR__IC1_HS_MADDR_MASK 0x00000007L +//CKSVII2C1_IC_DATA_CMD +#define CKSVII2C1_IC_DATA_CMD__DAT1__SHIFT 0x0 +#define CKSVII2C1_IC_DATA_CMD__CMD1__SHIFT 0x8 +#define CKSVII2C1_IC_DATA_CMD__STOP1__SHIFT 0x9 +#define CKSVII2C1_IC_DATA_CMD__RESTART1__SHIFT 0xa +#define CKSVII2C1_IC_DATA_CMD__FIRST1_DATA_BYTE__SHIFT 0xb +#define CKSVII2C1_IC_DATA_CMD__DAT1_MASK 0x000000FFL +#define CKSVII2C1_IC_DATA_CMD__CMD1_MASK 0x00000100L +#define CKSVII2C1_IC_DATA_CMD__STOP1_MASK 0x00000200L +#define CKSVII2C1_IC_DATA_CMD__RESTART1_MASK 0x00000400L +//CKSVII2C1_IC_SS_SCL_HCNT +#define CKSVII2C1_IC_SS_SCL_HCNT__IC1_SS_SCL_HCNT__SHIFT 0x0 +#define CKSVII2C1_IC_SS_SCL_HCNT__IC1_SS_SCL_HCNT_MASK 0x0000FFFFL +//CKSVII2C1_IC_SS_SCL_LCNT +#define CKSVII2C1_IC_SS_SCL_LCNT__IC1_SS_SCL_LCNT__SHIFT 0x0 +#define CKSVII2C1_IC_SS_SCL_LCNT__IC1_SS_SCL_LCNT_MASK 0x0000FFFFL +//CKSVII2C1_IC_FS_SCL_HCNT +#define CKSVII2C1_IC_FS_SCL_HCNT__IC1_FS_SCL_HCNT__SHIFT 0x0 +#define CKSVII2C1_IC_FS_SCL_HCNT__IC1_FS_SCL_HCNT_MASK 0x0000FFFFL +//CKSVII2C1_IC_FS_SCL_LCNT +#define CKSVII2C1_IC_FS_SCL_LCNT__IC1_FS_SCL_LCNT__SHIFT 0x0 +#define CKSVII2C1_IC_FS_SCL_LCNT__IC1_FS_SCL_LCNT_MASK 0x0000FFFFL +//CKSVII2C1_IC_HS_SCL_HCNT +#define CKSVII2C1_IC_HS_SCL_HCNT__IC1_HS_SCL_HCNT__SHIFT 0x0 +#define CKSVII2C1_IC_HS_SCL_HCNT__IC1_HS_SCL_HCNT_MASK 0x0000FFFFL +//CKSVII2C1_IC_HS_SCL_LCNT +#define CKSVII2C1_IC_HS_SCL_LCNT__IC1_HS_SCL_LCNT__SHIFT 0x0 +#define CKSVII2C1_IC_HS_SCL_LCNT__IC1_HS_SCL_LCNT_MASK 0x0000FFFFL +//CKSVII2C1_IC_INTR_STAT +#define CKSVII2C1_IC_INTR_STAT__R1_RX_UNDER__SHIFT 0x0 +#define CKSVII2C1_IC_INTR_STAT__R1_RX_OVER__SHIFT 0x1 +#define CKSVII2C1_IC_INTR_STAT__R1_RX_FULL__SHIFT 0x2 +#define CKSVII2C1_IC_INTR_STAT__R1_TX_OVER__SHIFT 0x3 +#define CKSVII2C1_IC_INTR_STAT__R1_TX_EMPTY__SHIFT 0x4 +#define CKSVII2C1_IC_INTR_STAT__R1_RD_REQ__SHIFT 0x5 +#define CKSVII2C1_IC_INTR_STAT__R1_TX_ABRT__SHIFT 0x6 +#define CKSVII2C1_IC_INTR_STAT__R1_RX_DONE__SHIFT 0x7 +#define CKSVII2C1_IC_INTR_STAT__R1_ACTIVITY__SHIFT 0x8 +#define CKSVII2C1_IC_INTR_STAT__R1_STOP_DET__SHIFT 0x9 +#define CKSVII2C1_IC_INTR_STAT__R1_START_DET__SHIFT 0xa +#define CKSVII2C1_IC_INTR_STAT__R1_GEN_CALL__SHIFT 0xb +#define CKSVII2C1_IC_INTR_STAT__R1_RESTART_DET__SHIFT 0xc +#define CKSVII2C1_IC_INTR_STAT__R1_MST_ON_HOLD__SHIFT 0xd +#define CKSVII2C1_IC_INTR_STAT__R1_SCL_STUCK_AT_LOW__SHIFT 0xe +#define CKSVII2C1_IC_INTR_STAT__R1_RX_UNDER_MASK 0x00000001L +#define CKSVII2C1_IC_INTR_STAT__R1_RX_OVER_MASK 0x00000002L +#define CKSVII2C1_IC_INTR_STAT__R1_RX_FULL_MASK 0x00000004L +#define CKSVII2C1_IC_INTR_STAT__R1_TX_OVER_MASK 0x00000008L +#define CKSVII2C1_IC_INTR_STAT__R1_TX_EMPTY_MASK 0x00000010L +#define CKSVII2C1_IC_INTR_STAT__R1_RD_REQ_MASK 0x00000020L +#define CKSVII2C1_IC_INTR_STAT__R1_TX_ABRT_MASK 0x00000040L +#define CKSVII2C1_IC_INTR_STAT__R1_RX_DONE_MASK 0x00000080L +#define CKSVII2C1_IC_INTR_STAT__R1_ACTIVITY_MASK 0x00000100L +#define CKSVII2C1_IC_INTR_STAT__R1_STOP_DET_MASK 0x00000200L +#define CKSVII2C1_IC_INTR_STAT__R1_START_DET_MASK 0x00000400L +#define CKSVII2C1_IC_INTR_STAT__R1_GEN_CALL_MASK 0x00000800L +#define CKSVII2C1_IC_INTR_STAT__R1_RESTART_DET_MASK 0x00001000L +#define CKSVII2C1_IC_INTR_STAT__R1_MST_ON_HOLD_MASK 0x00002000L +//CKSVII2C1_IC_INTR_MASK +#define CKSVII2C1_IC_INTR_MASK__M1_RX_UNDER__SHIFT 0x0 +#define CKSVII2C1_IC_INTR_MASK__M1_RX_OVER__SHIFT 0x1 +#define CKSVII2C1_IC_INTR_MASK__M1_RX_FULL__SHIFT 0x2 +#define CKSVII2C1_IC_INTR_MASK__M1_TX_OVER__SHIFT 0x3 +#define CKSVII2C1_IC_INTR_MASK__M1_TX_EMPTY__SHIFT 0x4 +#define CKSVII2C1_IC_INTR_MASK__M1_RD_REQ__SHIFT 0x5 +#define CKSVII2C1_IC_INTR_MASK__M1_TX_ABRT__SHIFT 0x6 +#define CKSVII2C1_IC_INTR_MASK__M1_RX_DONE__SHIFT 0x7 +#define CKSVII2C1_IC_INTR_MASK__M1_ACTIVITY__SHIFT 0x8 +#define CKSVII2C1_IC_INTR_MASK__M1_STOP_DET__SHIFT 0x9 +#define CKSVII2C1_IC_INTR_MASK__M1_START_DET__SHIFT 0xa +#define CKSVII2C1_IC_INTR_MASK__M1_GEN_CALL__SHIFT 0xb +#define CKSVII2C1_IC_INTR_MASK__M1_RESTART_DET__SHIFT 0xc +#define CKSVII2C1_IC_INTR_MASK__M1_MST_ON_HOLD__SHIFT 0xd +#define CKSVII2C1_IC_INTR_MASK__M1_SCL_STUCK_AT_LOW__SHIFT 0xe +#define CKSVII2C1_IC_INTR_MASK__M1_RX_UNDER_MASK 0x00000001L +#define CKSVII2C1_IC_INTR_MASK__M1_RX_OVER_MASK 0x00000002L +#define CKSVII2C1_IC_INTR_MASK__M1_RX_FULL_MASK 0x00000004L +#define CKSVII2C1_IC_INTR_MASK__M1_TX_OVER_MASK 0x00000008L +#define CKSVII2C1_IC_INTR_MASK__M1_TX_EMPTY_MASK 0x00000010L +#define CKSVII2C1_IC_INTR_MASK__M1_RD_REQ_MASK 0x00000020L +#define CKSVII2C1_IC_INTR_MASK__M1_TX_ABRT_MASK 0x00000040L +#define CKSVII2C1_IC_INTR_MASK__M1_RX_DONE_MASK 0x00000080L +#define CKSVII2C1_IC_INTR_MASK__M1_ACTIVITY_MASK 0x00000100L +#define CKSVII2C1_IC_INTR_MASK__M1_STOP_DET_MASK 0x00000200L +#define CKSVII2C1_IC_INTR_MASK__M1_START_DET_MASK 0x00000400L +#define CKSVII2C1_IC_INTR_MASK__M1_GEN_CALL_MASK 0x00000800L +#define CKSVII2C1_IC_INTR_MASK__M1_RESTART_DET_MASK 0x00001000L +#define CKSVII2C1_IC_INTR_MASK__M1_MST_ON_HOLD_MASK 0x00002000L +//CKSVII2C1_IC_RAW_INTR_STAT +//CKSVII2C1_IC_RX_TL +#define CKSVII2C1_IC_RX_TL__RX1_TL__SHIFT 0x0 +//CKSVII2C1_IC_TX_TL +#define CKSVII2C1_IC_TX_TL__TX1_TL__SHIFT 0x0 +//CKSVII2C1_IC_CLR_INTR +//CKSVII2C1_IC_CLR_RX_UNDER +//CKSVII2C1_IC_CLR_RX_OVER +//CKSVII2C1_IC_CLR_TX_OVER +//CKSVII2C1_IC_CLR_RD_REQ +//CKSVII2C1_IC_CLR_TX_ABRT +//CKSVII2C1_IC_CLR_RX_DONE +//CKSVII2C1_IC_CLR_ACTIVITY +//CKSVII2C1_IC_CLR_STOP_DET +//CKSVII2C1_IC_CLR_START_DET +//CKSVII2C1_IC_CLR_GEN_CALL +//CKSVII2C1_IC_ENABLE +#define CKSVII2C1_IC_ENABLE__ENABLE1__SHIFT 0x0 +#define CKSVII2C1_IC_ENABLE__ABORT1__SHIFT 0x1 +#define CKSVII2C1_IC_ENABLE__TX1_CMD_BLOCK__SHIFT 0x2 +#define CKSVII2C1_IC_ENABLE__SDA1_STUCK_RECOVERY_ENABLE__SHIFT 0x3 +#define CKSVII2C1_IC_ENABLE__ENABLE1_MASK 0x00000001L +#define CKSVII2C1_IC_ENABLE__ABORT1_MASK 0x00000002L +//CKSVII2C1_IC_STATUS +#define CKSVII2C1_IC_STATUS__ACTIVITY1__SHIFT 0x0 +#define CKSVII2C1_IC_STATUS__TFNF1__SHIFT 0x1 +#define CKSVII2C1_IC_STATUS__TFE1__SHIFT 0x2 +#define CKSVII2C1_IC_STATUS__RFNE1__SHIFT 0x3 +#define CKSVII2C1_IC_STATUS__RFF1__SHIFT 0x4 +#define CKSVII2C1_IC_STATUS__MST1_ACTIVITY__SHIFT 0x5 +#define CKSVII2C1_IC_STATUS__SLV1_ACTIVITY__SHIFT 0x6 +#define CKSVII2C1_IC_STATUS__MST1_HOLD_TX_FIFO_EMPTY__SHIFT 0x7 +#define CKSVII2C1_IC_STATUS__MST1_HOLD_RX_FIFO_FULL__SHIFT 0x8 +#define CKSVII2C1_IC_STATUS__SLV1_HOLD_TX_FIFO_EMPTY__SHIFT 0x9 +#define CKSVII2C1_IC_STATUS__SLV1_HOLD_RX_FIFO_FULL__SHIFT 0xa +#define CKSVII2C1_IC_STATUS__SDA1_STUCK_NOT_RECOVERED__SHIFT 0xb +#define CKSVII2C1_IC_STATUS__ACTIVITY1_MASK 0x00000001L +#define CKSVII2C1_IC_STATUS__TFNF1_MASK 0x00000002L +#define CKSVII2C1_IC_STATUS__TFE1_MASK 0x00000004L +#define CKSVII2C1_IC_STATUS__RFNE1_MASK 0x00000008L +#define CKSVII2C1_IC_STATUS__RFF1_MASK 0x00000010L +#define CKSVII2C1_IC_STATUS__MST1_ACTIVITY_MASK 0x00000020L +#define CKSVII2C1_IC_STATUS__SLV1_ACTIVITY_MASK 0x00000040L +//CKSVII2C1_IC_TXFLR +#define CKSVII2C1_IC_TXFLR__TXFLR1__SHIFT 0x0 +//CKSVII2C1_IC_RXFLR +#define CKSVII2C1_IC_RXFLR__RXFLR1__SHIFT 0x0 +//CKSVII2C1_IC_SDA_HOLD +#define CKSVII2C1_IC_SDA_HOLD__IC1_SDA_TX_HOLD__SHIFT 0x0 +#define CKSVII2C1_IC_SDA_HOLD__IC1_SDA_RX_HOLD__SHIFT 0x10 +//CKSVII2C1_IC_TX_ABRT_SOURCE +//CKSVII2C1_IC_SLV_DATA_NACK_ONLY +//CKSVII2C1_IC_DMA_CR +//CKSVII2C1_IC_DMA_TDLR +//CKSVII2C1_IC_DMA_RDLR +//CKSVII2C1_IC_SDA_SETUP +#define CKSVII2C1_IC_SDA_SETUP__SDA1_SETUP__SHIFT 0x0 +#define CKSVII2C1_IC_SDA_SETUP__SDA1_SETUP_MASK 0x000000FFL +//CKSVII2C1_IC_ACK_GENERAL_CALL +#define CKSVII2C1_IC_ACK_GENERAL_CALL__ACK1_GENERAL_CALL__SHIFT 0x0 +#define CKSVII2C1_IC_ACK_GENERAL_CALL__ACK1_GENERAL_CALL_MASK 0x00000001L +//CKSVII2C1_IC_ENABLE_STATUS +#define CKSVII2C1_IC_ENABLE_STATUS__IC1_EN__SHIFT 0x0 +#define CKSVII2C1_IC_ENABLE_STATUS__SLV1_DISABLED_WHILE_BUSY__SHIFT 0x1 +#define CKSVII2C1_IC_ENABLE_STATUS__SLV1_RX_DATA_LOST__SHIFT 0x2 +#define CKSVII2C1_IC_ENABLE_STATUS__IC1_EN_MASK 0x00000001L +//CKSVII2C1_IC_FS_SPKLEN +#define CKSVII2C1_IC_FS_SPKLEN__FS1_SPKLEN__SHIFT 0x0 +//CKSVII2C1_IC_HS_SPKLEN +#define CKSVII2C1_IC_HS_SPKLEN__HS1_SPKLEN__SHIFT 0x0 +//CKSVII2C1_IC_CLR_RESTART_DET +//CKSVII2C1_IC_COMP_PARAM_1 +#define CKSVII2C1_IC_COMP_PARAM_1__APB1_DATA_WIDTH__SHIFT 0x0 +#define CKSVII2C1_IC_COMP_PARAM_1__MAX1_SPEED_MODE__SHIFT 0x2 +#define CKSVII2C1_IC_COMP_PARAM_1__HC1_COUNT_VALUES__SHIFT 0x4 +#define CKSVII2C1_IC_COMP_PARAM_1__INTR1_IO__SHIFT 0x5 +#define CKSVII2C1_IC_COMP_PARAM_1__HAS1_DMA__SHIFT 0x6 +#define CKSVII2C1_IC_COMP_PARAM_1__ADD1_ENCODED_PARAMS__SHIFT 0x7 +#define CKSVII2C1_IC_COMP_PARAM_1__RX1_BUFFER_DEPTH__SHIFT 0x8 +#define CKSVII2C1_IC_COMP_PARAM_1__TX1_BUFFER_DEPTH__SHIFT 0x10 +//CKSVII2C1_IC_COMP_VERSION +#define CKSVII2C1_IC_COMP_VERSION__COMP1_VERSION__SHIFT 0x0 +//CKSVII2C1_IC_COMP_TYPE +#define CKSVII2C1_IC_COMP_TYPE__COMP1_TYPE__SHIFT 0x0 +//SMUIO_PWRMGT +#define SMUIO_PWRMGT__i2c_clk_gate_en__SHIFT 0x0 +#define SMUIO_PWRMGT__i2c1_clk_gate_en__SHIFT 0x4 +#define SMUIO_PWRMGT__i2c_clk_gate_en_MASK 0x00000001L +#define SMUIO_PWRMGT__i2c1_clk_gate_en_MASK 0x00000010L + + +// addressBlock: smuio_smuio_rom_SmuSmuioDec +//ROM_CNTL +#define ROM_CNTL__CLOCK_GATING_EN__SHIFT 0x0 +#define ROM_CNTL__READ_MODE__SHIFT 0x1 +#define ROM_CNTL__READ_MODE_OVERRIDE__SHIFT 0x3 +#define ROM_CNTL__SPI_TIMING_RELAX_SCK__SHIFT 0x4 +#define ROM_CNTL__SPI_TIMING_RELAX_SCK_OVERRIDE__SHIFT 0x5 +#define ROM_CNTL__FOUR_BYTE_ADDRESS_MODE__SHIFT 0x6 +#define ROM_CNTL__DUMMY_CYCLE_NUM__SHIFT 0x8 +#define ROM_CNTL__SPI_TIMING_RELAX__SHIFT 0x13 +#define ROM_CNTL__SPI_TIMING_RELAX_OVERRIDE__SHIFT 0x14 +#define ROM_CNTL__SPI_FAST_MODE__SHIFT 0x15 +#define ROM_CNTL__SPI_FAST_MODE_OVERRIDE__SHIFT 0x16 +#define ROM_CNTL__SCK_PRESCALE_REFCLK__SHIFT 0x17 +#define ROM_CNTL__SCK_PRESCALE_REFCLK_OVERRIDE__SHIFT 0x1c +#define ROM_CNTL__ROM_INDEX_ADDRESS_AUTO_INCREASE__SHIFT 0x1d +#define ROM_CNTL__PAD_SAMPLE_MODE__SHIFT 0x1e +#define ROM_CNTL__PAD_SAMPLE_MODE_OVERRIDE__SHIFT 0x1f +#define ROM_CNTL__CLOCK_GATING_EN_MASK 0x00000001L +#define ROM_CNTL__SPI_TIMING_RELAX_MASK 0x00080000L +#define ROM_CNTL__SPI_TIMING_RELAX_OVERRIDE_MASK 0x00100000L +#define ROM_CNTL__SPI_FAST_MODE_MASK 0x00200000L +#define ROM_CNTL__SPI_FAST_MODE_OVERRIDE_MASK 0x00400000L +#define ROM_CNTL__SCK_PRESCALE_REFCLK_MASK 0x0F800000L +#define ROM_CNTL__SCK_PRESCALE_REFCLK_OVERRIDE_MASK 0x10000000L +//PAGE_MIRROR_CNTL +#define PAGE_MIRROR_CNTL__PAGE_MIRROR_BASE_ADDR__SHIFT 0x0 +#define PAGE_MIRROR_CNTL__PAGE_MIRROR_ENABLE__SHIFT 0x19 +#define PAGE_MIRROR_CNTL__PAGE_MIRROR_USAGE__SHIFT 0x1a +#define PAGE_MIRROR_CNTL__PAGE_MIRROR_INVALIDATE__SHIFT 0x1c +#define PAGE_MIRROR_CNTL__PAGE_MIRROR_BASE_ADDR_MASK 0x01FFFFFFL +#define PAGE_MIRROR_CNTL__PAGE_MIRROR_ENABLE_MASK 0x02000000L +#define PAGE_MIRROR_CNTL__PAGE_MIRROR_USAGE_MASK 0x0C000000L +#define PAGE_MIRROR_CNTL__PAGE_MIRROR_INVALIDATE_MASK 0x10000000L +//ROM_STATUS +#define ROM_STATUS__ROM_BUSY__SHIFT 0x0 +#define ROM_STATUS__ROM_BUSY_MASK 0x00000001L +//CGTT_ROM_CLK_CTRL0 +#define CGTT_ROM_CLK_CTRL0__ON_DELAY__SHIFT 0x0 +#define CGTT_ROM_CLK_CTRL0__OFF_HYSTERESIS__SHIFT 0x4 +#define CGTT_ROM_CLK_CTRL0__SOFT_OVERRIDE1__SHIFT 0x1e +#define CGTT_ROM_CLK_CTRL0__SOFT_OVERRIDE0__SHIFT 0x1f +#define CGTT_ROM_CLK_CTRL0__ON_DELAY_MASK 0x0000000FL +#define CGTT_ROM_CLK_CTRL0__OFF_HYSTERESIS_MASK 0x00000FF0L +#define CGTT_ROM_CLK_CTRL0__SOFT_OVERRIDE1_MASK 0x40000000L +#define CGTT_ROM_CLK_CTRL0__SOFT_OVERRIDE0_MASK 0x80000000L +//ROM_INDEX +#define ROM_INDEX__ROM_INDEX__SHIFT 0x0 +#define ROM_INDEX__ROM_INDEX_MASK 0x01FFFFFFL +//ROM_DATA +#define ROM_DATA__ROM_DATA__SHIFT 0x0 +#define ROM_DATA__ROM_DATA_MASK 0xFFFFFFFFL +//ROM_START +#define ROM_START__ROM_START__SHIFT 0x0 +#define ROM_START__ROM_START_MASK 0x01FFFFFFL +//ROM_SW_CNTL +#define ROM_SW_CNTL__DATA_SIZE__SHIFT 0x0 +#define ROM_SW_CNTL__COMMAND_SIZE__SHIFT 0x10 +#define ROM_SW_CNTL__ROM_SW_RETURN_DATA_ENABLE__SHIFT 0x13 +#define ROM_SW_CNTL__DATA_SIZE_MASK 0x0000FFFFL +#define ROM_SW_CNTL__COMMAND_SIZE_MASK 0x00070000L +#define ROM_SW_CNTL__ROM_SW_RETURN_DATA_ENABLE_MASK 0x00080000L +//ROM_SW_STATUS +#define ROM_SW_STATUS__ROM_SW_DONE__SHIFT 0x0 +#define ROM_SW_STATUS__ROM_SW_DONE_MASK 0x00000001L +//ROM_SW_COMMAND +#define ROM_SW_COMMAND__ROM_SW_INSTRUCTION__SHIFT 0x0 +#define ROM_SW_COMMAND__ROM_SW_ADDRESS__SHIFT 0x8 +#define ROM_SW_COMMAND__ROM_SW_INSTRUCTION_MASK 0x000000FFL +#define ROM_SW_COMMAND__ROM_SW_ADDRESS_MASK 0xFFFFFF00L +//ROM_SW_DATA_1 +#define ROM_SW_DATA_1__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_1__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_2 +#define ROM_SW_DATA_2__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_2__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_3 +#define ROM_SW_DATA_3__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_3__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_4 +#define ROM_SW_DATA_4__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_4__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_5 +#define ROM_SW_DATA_5__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_5__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_6 +#define ROM_SW_DATA_6__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_6__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_7 +#define ROM_SW_DATA_7__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_7__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_8 +#define ROM_SW_DATA_8__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_8__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_9 +#define ROM_SW_DATA_9__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_9__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_10 +#define ROM_SW_DATA_10__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_10__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_11 +#define ROM_SW_DATA_11__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_11__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_12 +#define ROM_SW_DATA_12__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_12__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_13 +#define ROM_SW_DATA_13__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_13__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_14 +#define ROM_SW_DATA_14__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_14__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_15 +#define ROM_SW_DATA_15__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_15__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_16 +#define ROM_SW_DATA_16__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_16__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_17 +#define ROM_SW_DATA_17__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_17__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_18 +#define ROM_SW_DATA_18__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_18__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_19 +#define ROM_SW_DATA_19__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_19__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_20 +#define ROM_SW_DATA_20__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_20__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_21 +#define ROM_SW_DATA_21__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_21__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_22 +#define ROM_SW_DATA_22__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_22__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_23 +#define ROM_SW_DATA_23__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_23__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_24 +#define ROM_SW_DATA_24__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_24__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_25 +#define ROM_SW_DATA_25__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_25__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_26 +#define ROM_SW_DATA_26__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_26__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_27 +#define ROM_SW_DATA_27__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_27__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_28 +#define ROM_SW_DATA_28__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_28__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_29 +#define ROM_SW_DATA_29__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_29__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_30 +#define ROM_SW_DATA_30__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_30__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_31 +#define ROM_SW_DATA_31__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_31__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_32 +#define ROM_SW_DATA_32__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_32__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_33 +#define ROM_SW_DATA_33__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_33__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_34 +#define ROM_SW_DATA_34__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_34__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_35 +#define ROM_SW_DATA_35__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_35__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_36 +#define ROM_SW_DATA_36__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_36__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_37 +#define ROM_SW_DATA_37__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_37__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_38 +#define ROM_SW_DATA_38__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_38__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_39 +#define ROM_SW_DATA_39__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_39__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_40 +#define ROM_SW_DATA_40__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_40__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_41 +#define ROM_SW_DATA_41__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_41__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_42 +#define ROM_SW_DATA_42__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_42__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_43 +#define ROM_SW_DATA_43__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_43__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_44 +#define ROM_SW_DATA_44__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_44__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_45 +#define ROM_SW_DATA_45__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_45__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_46 +#define ROM_SW_DATA_46__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_46__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_47 +#define ROM_SW_DATA_47__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_47__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_48 +#define ROM_SW_DATA_48__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_48__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_49 +#define ROM_SW_DATA_49__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_49__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_50 +#define ROM_SW_DATA_50__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_50__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_51 +#define ROM_SW_DATA_51__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_51__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_52 +#define ROM_SW_DATA_52__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_52__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_53 +#define ROM_SW_DATA_53__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_53__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_54 +#define ROM_SW_DATA_54__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_54__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_55 +#define ROM_SW_DATA_55__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_55__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_56 +#define ROM_SW_DATA_56__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_56__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_57 +#define ROM_SW_DATA_57__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_57__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_58 +#define ROM_SW_DATA_58__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_58__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_59 +#define ROM_SW_DATA_59__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_59__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_60 +#define ROM_SW_DATA_60__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_60__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_61 +#define ROM_SW_DATA_61__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_61__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_62 +#define ROM_SW_DATA_62__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_62__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_63 +#define ROM_SW_DATA_63__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_63__ROM_SW_DATA_MASK 0xFFFFFFFFL +//ROM_SW_DATA_64 +#define ROM_SW_DATA_64__ROM_SW_DATA__SHIFT 0x0 +#define ROM_SW_DATA_64__ROM_SW_DATA_MASK 0xFFFFFFFFL + + +// addressBlock: smuio_smuio_gpio_SmuSmuioDec +//SMU_GPIOPAD_SW_INT_STAT +#define SMU_GPIOPAD_SW_INT_STAT__SW_INT_STAT__SHIFT 0x0 +#define SMU_GPIOPAD_SW_INT_STAT__SW_INT_STAT_MASK 0x00000001L +//SMU_GPIOPAD_MASK +#define SMU_GPIOPAD_MASK__GPIO_MASK__SHIFT 0x0 +#define SMU_GPIOPAD_MASK__GPIO_MASK_MASK 0x7FFFFFFFL +//SMU_GPIOPAD_A +#define SMU_GPIOPAD_A__GPIO_A__SHIFT 0x0 +#define SMU_GPIOPAD_A__GPIO_A_MASK 0x7FFFFFFFL +//SMU_GPIOPAD_TXIMPSEL +#define SMU_GPIOPAD_TXIMPSEL__GPIO_TXIMPSEL__SHIFT 0x0 +#define SMU_GPIOPAD_TXIMPSEL__GPIO_TXIMPSEL_MASK 0x7FFFFFFFL +//SMU_GPIOPAD_EN +#define SMU_GPIOPAD_EN__GPIO_EN__SHIFT 0x0 +#define SMU_GPIOPAD_EN__GPIO_EN_MASK 0x7FFFFFFFL +//SMU_GPIOPAD_Y +#define SMU_GPIOPAD_Y__GPIO_Y__SHIFT 0x0 +#define SMU_GPIOPAD_Y__GPIO_Y_MASK 0x7FFFFFFFL +//SMU_GPIOPAD_RXEN +#define SMU_GPIOPAD_RXEN__GPIO_RXEN__SHIFT 0x0 +#define SMU_GPIOPAD_RXEN__GPIO_RXEN_MASK 0x7FFFFFFFL +//SMU_GPIOPAD_RCVR_SEL0 +#define SMU_GPIOPAD_RCVR_SEL0__GPIO_RCVR_SEL0__SHIFT 0x0 +#define SMU_GPIOPAD_RCVR_SEL0__GPIO_RCVR_SEL0_MASK 0x7FFFFFFFL +//SMU_GPIOPAD_RCVR_SEL1 +#define SMU_GPIOPAD_RCVR_SEL1__GPIO_RCVR_SEL1__SHIFT 0x0 +#define SMU_GPIOPAD_RCVR_SEL1__GPIO_RCVR_SEL1_MASK 0x7FFFFFFFL +//SMU_GPIOPAD_PU_EN +#define SMU_GPIOPAD_PU_EN__GPIO_PU_EN__SHIFT 0x0 +#define SMU_GPIOPAD_PU_EN__GPIO_PU_EN_MASK 0x7FFFFFFFL +//SMU_GPIOPAD_PD_EN +#define SMU_GPIOPAD_PD_EN__GPIO_PD_EN__SHIFT 0x0 +#define SMU_GPIOPAD_PD_EN__GPIO_PD_EN_MASK 0x7FFFFFFFL +//SMU_GPIOPAD_PINSTRAPS +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_0__SHIFT 0x0 +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_1__SHIFT 0x1 +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_2__SHIFT 0x2 +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_3__SHIFT 0x3 +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_4__SHIFT 0x4 +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_5__SHIFT 0x5 +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_6__SHIFT 0x6 +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_7__SHIFT 0x7 +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_8__SHIFT 0x8 +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_9__SHIFT 0x9 +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_10__SHIFT 0xa +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_11__SHIFT 0xb +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_12__SHIFT 0xc +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_13__SHIFT 0xd +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_14__SHIFT 0xe +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_15__SHIFT 0xf +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_16__SHIFT 0x10 +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_17__SHIFT 0x11 +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_18__SHIFT 0x12 +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_19__SHIFT 0x13 +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_20__SHIFT 0x14 +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_21__SHIFT 0x15 +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_22__SHIFT 0x16 +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_23__SHIFT 0x17 +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_24__SHIFT 0x18 +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_25__SHIFT 0x19 +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_26__SHIFT 0x1a +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_27__SHIFT 0x1b +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_28__SHIFT 0x1c +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_29__SHIFT 0x1d +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_30__SHIFT 0x1e +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_0_MASK 0x00000001L +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_1_MASK 0x00000002L +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_2_MASK 0x00000004L +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_3_MASK 0x00000008L +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_4_MASK 0x00000010L +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_5_MASK 0x00000020L +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_6_MASK 0x00000040L +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_7_MASK 0x00000080L +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_8_MASK 0x00000100L +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_9_MASK 0x00000200L +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_10_MASK 0x00000400L +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_11_MASK 0x00000800L +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_12_MASK 0x00001000L +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_13_MASK 0x00002000L +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_14_MASK 0x00004000L +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_15_MASK 0x00008000L +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_16_MASK 0x00010000L +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_17_MASK 0x00020000L +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_18_MASK 0x00040000L +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_19_MASK 0x00080000L +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_20_MASK 0x00100000L +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_21_MASK 0x00200000L +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_22_MASK 0x00400000L +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_23_MASK 0x00800000L +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_24_MASK 0x01000000L +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_25_MASK 0x02000000L +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_26_MASK 0x04000000L +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_27_MASK 0x08000000L +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_28_MASK 0x10000000L +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_29_MASK 0x20000000L +#define SMU_GPIOPAD_PINSTRAPS__GPIO_PINSTRAP_30_MASK 0x40000000L +//DFT_PINSTRAPS +#define DFT_PINSTRAPS__DFT_PINSTRAPS__SHIFT 0x0 +#define DFT_PINSTRAPS__DFT_PINSTRAPS_MASK 0x000000FFL +//SMU_GPIOPAD_INT_STAT_EN +#define SMU_GPIOPAD_INT_STAT_EN__GPIO_INT_STAT_EN__SHIFT 0x0 +#define SMU_GPIOPAD_INT_STAT_EN__SW_INITIATED_INT_STAT_EN__SHIFT 0x1f +#define SMU_GPIOPAD_INT_STAT_EN__GPIO_INT_STAT_EN_MASK 0x1FFFFFFFL +#define SMU_GPIOPAD_INT_STAT_EN__SW_INITIATED_INT_STAT_EN_MASK 0x80000000L +//SMU_GPIOPAD_INT_STAT +#define SMU_GPIOPAD_INT_STAT__GPIO_INT_STAT__SHIFT 0x0 +#define SMU_GPIOPAD_INT_STAT__SW_INITIATED_INT_STAT__SHIFT 0x1f +#define SMU_GPIOPAD_INT_STAT__GPIO_INT_STAT_MASK 0x1FFFFFFFL +#define SMU_GPIOPAD_INT_STAT__SW_INITIATED_INT_STAT_MASK 0x80000000L +//SMU_GPIOPAD_INT_STAT_AK +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_0__SHIFT 0x0 +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_1__SHIFT 0x1 +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_2__SHIFT 0x2 +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_3__SHIFT 0x3 +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_4__SHIFT 0x4 +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_5__SHIFT 0x5 +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_6__SHIFT 0x6 +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_7__SHIFT 0x7 +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_8__SHIFT 0x8 +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_9__SHIFT 0x9 +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_10__SHIFT 0xa +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_11__SHIFT 0xb +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_12__SHIFT 0xc +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_13__SHIFT 0xd +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_14__SHIFT 0xe +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_15__SHIFT 0xf +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_16__SHIFT 0x10 +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_17__SHIFT 0x11 +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_18__SHIFT 0x12 +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_19__SHIFT 0x13 +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_20__SHIFT 0x14 +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_21__SHIFT 0x15 +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_22__SHIFT 0x16 +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_23__SHIFT 0x17 +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_24__SHIFT 0x18 +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_25__SHIFT 0x19 +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_26__SHIFT 0x1a +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_27__SHIFT 0x1b +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_28__SHIFT 0x1c +#define SMU_GPIOPAD_INT_STAT_AK__SW_INITIATED_INT_STAT_AK__SHIFT 0x1f +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_0_MASK 0x00000001L +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_1_MASK 0x00000002L +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_2_MASK 0x00000004L +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_3_MASK 0x00000008L +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_4_MASK 0x00000010L +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_5_MASK 0x00000020L +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_6_MASK 0x00000040L +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_7_MASK 0x00000080L +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_8_MASK 0x00000100L +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_9_MASK 0x00000200L +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_10_MASK 0x00000400L +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_11_MASK 0x00000800L +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_12_MASK 0x00001000L +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_13_MASK 0x00002000L +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_14_MASK 0x00004000L +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_15_MASK 0x00008000L +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_16_MASK 0x00010000L +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_17_MASK 0x00020000L +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_18_MASK 0x00040000L +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_19_MASK 0x00080000L +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_20_MASK 0x00100000L +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_21_MASK 0x00200000L +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_22_MASK 0x00400000L +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_23_MASK 0x00800000L +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_24_MASK 0x01000000L +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_25_MASK 0x02000000L +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_26_MASK 0x04000000L +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_27_MASK 0x08000000L +#define SMU_GPIOPAD_INT_STAT_AK__GPIO_INT_STAT_AK_28_MASK 0x10000000L +#define SMU_GPIOPAD_INT_STAT_AK__SW_INITIATED_INT_STAT_AK_MASK 0x80000000L +//SMU_GPIOPAD_INT_EN +#define SMU_GPIOPAD_INT_EN__GPIO_INT_EN__SHIFT 0x0 +#define SMU_GPIOPAD_INT_EN__SW_INITIATED_INT_EN__SHIFT 0x1f +#define SMU_GPIOPAD_INT_EN__GPIO_INT_EN_MASK 0x1FFFFFFFL +#define SMU_GPIOPAD_INT_EN__SW_INITIATED_INT_EN_MASK 0x80000000L +//SMU_GPIOPAD_INT_TYPE +#define SMU_GPIOPAD_INT_TYPE__GPIO_INT_TYPE__SHIFT 0x0 +#define SMU_GPIOPAD_INT_TYPE__SW_INITIATED_INT_TYPE__SHIFT 0x1f +#define SMU_GPIOPAD_INT_TYPE__GPIO_INT_TYPE_MASK 0x1FFFFFFFL +#define SMU_GPIOPAD_INT_TYPE__SW_INITIATED_INT_TYPE_MASK 0x80000000L +//SMU_GPIOPAD_INT_POLARITY +#define SMU_GPIOPAD_INT_POLARITY__GPIO_INT_POLARITY__SHIFT 0x0 +#define SMU_GPIOPAD_INT_POLARITY__SW_INITIATED_INT_POLARITY__SHIFT 0x1f +#define SMU_GPIOPAD_INT_POLARITY__GPIO_INT_POLARITY_MASK 0x1FFFFFFFL +#define SMU_GPIOPAD_INT_POLARITY__SW_INITIATED_INT_POLARITY_MASK 0x80000000L +//SMUIO_PCC_GPIO_SELECT +#define SMUIO_PCC_GPIO_SELECT__GPIO__SHIFT 0x0 +#define SMUIO_PCC_GPIO_SELECT__GPIO_MASK 0xFFFFFFFFL +//SMU_GPIOPAD_S0 +#define SMU_GPIOPAD_S0__GPIO_S0__SHIFT 0x0 +#define SMU_GPIOPAD_S0__GPIO_S0_MASK 0x7FFFFFFFL +//SMU_GPIOPAD_S1 +#define SMU_GPIOPAD_S1__GPIO_S1__SHIFT 0x0 +#define SMU_GPIOPAD_S1__GPIO_S1_MASK 0x7FFFFFFFL +//SMU_GPIOPAD_SCHMEN +#define SMU_GPIOPAD_SCHMEN__GPIO_SCHMEN__SHIFT 0x0 +#define SMU_GPIOPAD_SCHMEN__GPIO_SCHMEN_MASK 0x7FFFFFFFL +//SMU_GPIOPAD_SCL_EN +#define SMU_GPIOPAD_SCL_EN__GPIO_SCL_EN__SHIFT 0x0 +#define SMU_GPIOPAD_SCL_EN__GPIO_SCL_EN_MASK 0x7FFFFFFFL +//SMU_GPIOPAD_SDA_EN +#define SMU_GPIOPAD_SDA_EN__GPIO_SDA_EN__SHIFT 0x0 +#define SMU_GPIOPAD_SDA_EN__GPIO_SDA_EN_MASK 0x7FFFFFFFL +//SMUIO_GPIO_INT0_SELECT +#define SMUIO_GPIO_INT0_SELECT__GPIO_INT0_SELECT__SHIFT 0x0 +#define SMUIO_GPIO_INT0_SELECT__GPIO_INT0_SELECT_MASK 0xFFFFFFFFL +//SMUIO_GPIO_INT1_SELECT +#define SMUIO_GPIO_INT1_SELECT__GPIO_INT1_SELECT__SHIFT 0x0 +#define SMUIO_GPIO_INT1_SELECT__GPIO_INT1_SELECT_MASK 0xFFFFFFFFL +//SMUIO_GPIO_INT2_SELECT +#define SMUIO_GPIO_INT2_SELECT__GPIO_INT2_SELECT__SHIFT 0x0 +#define SMUIO_GPIO_INT2_SELECT__GPIO_INT2_SELECT_MASK 0xFFFFFFFFL +//SMUIO_GPIO_INT3_SELECT +#define SMUIO_GPIO_INT3_SELECT__GPIO_INT3_SELECT__SHIFT 0x0 +#define SMUIO_GPIO_INT3_SELECT__GPIO_INT3_SELECT_MASK 0xFFFFFFFFL +//SMU_GPIOPAD_MP_INT0_STAT +#define SMU_GPIOPAD_MP_INT0_STAT__GPIO_MP_INT0_STAT__SHIFT 0x0 +#define SMU_GPIOPAD_MP_INT0_STAT__GPIO_MP_INT0_STAT_MASK 0x1FFFFFFFL +//SMU_GPIOPAD_MP_INT1_STAT +#define SMU_GPIOPAD_MP_INT1_STAT__GPIO_MP_INT1_STAT__SHIFT 0x0 +#define SMU_GPIOPAD_MP_INT1_STAT__GPIO_MP_INT1_STAT_MASK 0x1FFFFFFFL +//SMU_GPIOPAD_MP_INT2_STAT +#define SMU_GPIOPAD_MP_INT2_STAT__GPIO_MP_INT2_STAT__SHIFT 0x0 +#define SMU_GPIOPAD_MP_INT2_STAT__GPIO_MP_INT2_STAT_MASK 0x1FFFFFFFL +//SMU_GPIOPAD_MP_INT3_STAT +#define SMU_GPIOPAD_MP_INT3_STAT__GPIO_MP_INT3_STAT__SHIFT 0x0 +#define SMU_GPIOPAD_MP_INT3_STAT__GPIO_MP_INT3_STAT_MASK 0x1FFFFFFFL +//SMIO_INDEX +#define SMIO_INDEX__SW_SMIO_INDEX__SHIFT 0x0 +#define SMIO_INDEX__SW_SMIO_INDEX_MASK 0x00000001L +//S0_VID_SMIO_CNTL +#define S0_VID_SMIO_CNTL__S0_SMIO_VALUES__SHIFT 0x0 +#define S0_VID_SMIO_CNTL__S0_SMIO_VALUES_MASK 0xFFFFFFFFL +//S1_VID_SMIO_CNTL +#define S1_VID_SMIO_CNTL__S1_SMIO_VALUES__SHIFT 0x0 +#define S1_VID_SMIO_CNTL__S1_SMIO_VALUES_MASK 0xFFFFFFFFL +//OPEN_DRAIN_SELECT +#define OPEN_DRAIN_SELECT__OPEN_DRAIN_SELECT__SHIFT 0x0 +#define OPEN_DRAIN_SELECT__RESERVED__SHIFT 0x1f +#define OPEN_DRAIN_SELECT__OPEN_DRAIN_SELECT_MASK 0x7FFFFFFFL +#define OPEN_DRAIN_SELECT__RESERVED_MASK 0x80000000L +//SMIO_ENABLE +#define SMIO_ENABLE__SMIO_ENABLE__SHIFT 0x0 +#define SMIO_ENABLE__SMIO_ENABLE_MASK 0xFFFFFFFFL + +#endif diff --git a/drivers/gpu/drm/amd/include/atomfirmware.h b/drivers/gpu/drm/amd/include/atomfirmware.h index af3eebb4c9bc..571691837200 100644 --- a/drivers/gpu/drm/amd/include/atomfirmware.h +++ b/drivers/gpu/drm/amd/include/atomfirmware.h @@ -1657,6 +1657,49 @@ struct atom_integrated_system_info_v2_2 uint32_t reserved4[189]; }; +struct uma_carveout_option { + char optionName[29]; //max length of string is 28chars + '\0'. Current design is for "minimum", "Medium", "High". This makes entire struct size 64bits + uint8_t memoryCarvedGb; //memory carved out with setting + uint8_t memoryRemainingGb; //memory remaining on system + union { + struct _flags { + uint8_t Auto : 1; + uint8_t Custom : 1; + uint8_t Reserved : 6; + } flags; + uint8_t all8; + } uma_carveout_option_flags; +}; + +struct atom_integrated_system_info_v2_3 { + struct atom_common_table_header table_header; + uint32_t vbios_misc; // enum of atom_system_vbiosmisc_def + uint32_t gpucapinfo; // enum of atom_system_gpucapinf_def + uint32_t system_config; + uint32_t cpucapinfo; + uint16_t gpuclk_ss_percentage; // unit of 0.001%, 1000 mean 1% + uint16_t gpuclk_ss_type; + uint16_t dpphy_override; // bit vector, enum of atom_sysinfo_dpphy_override_def + uint8_t memorytype; // enum of atom_dmi_t17_mem_type_def, APU memory type indication. + uint8_t umachannelnumber; // number of memory channels + uint8_t htc_hyst_limit; + uint8_t htc_tmp_limit; + uint8_t reserved1; // dp_ss_control + uint8_t gpu_package_id; + struct edp_info_table edp1_info; + struct edp_info_table edp2_info; + uint32_t reserved2[8]; + struct atom_external_display_connection_info extdispconninfo; + uint8_t UMACarveoutVersion; + uint8_t UMACarveoutIndexMax; + uint8_t UMACarveoutTypeDefault; + uint8_t UMACarveoutIndexDefault; + uint8_t UMACarveoutType; //Auto or Custom + uint8_t UMACarveoutIndex; + struct uma_carveout_option UMASizeControlOption[20]; + uint8_t reserved3[110]; +}; + // system_config enum atom_system_vbiosmisc_def{ INTEGRATED_SYSTEM_INFO__GET_EDID_CALLBACK_FUNC_SUPPORT = 0x01, @@ -3540,7 +3583,7 @@ struct atom_gpio_voltage_object_v4 uint8_t phase_delay_us; // phase delay in unit of micro second uint8_t reserved; uint32_t gpio_mask_val; // GPIO Mask value - struct atom_voltage_gpio_map_lut voltage_gpio_lut[1]; + struct atom_voltage_gpio_map_lut voltage_gpio_lut[] __counted_by(gpio_entry_num); }; struct atom_svid2_voltage_object_v4 diff --git a/drivers/gpu/drm/amd/include/kgd_pp_interface.h b/drivers/gpu/drm/amd/include/kgd_pp_interface.h index 32054ecf0b87..805c9d37a2b4 100644 --- a/drivers/gpu/drm/amd/include/kgd_pp_interface.h +++ b/drivers/gpu/drm/amd/include/kgd_pp_interface.h @@ -150,6 +150,7 @@ enum amd_pp_sensors { AMDGPU_PP_SENSOR_VCN_POWER_STATE, AMDGPU_PP_SENSOR_PEAK_PSTATE_SCLK, AMDGPU_PP_SENSOR_PEAK_PSTATE_MCLK, + AMDGPU_PP_SENSOR_VCN_LOAD, }; enum amd_pp_task { @@ -420,7 +421,7 @@ struct amd_pm_funcs { int (*set_hard_min_dcefclk_by_freq)(void *handle, uint32_t clock); int (*set_hard_min_fclk_by_freq)(void *handle, uint32_t clock); int (*set_min_deep_sleep_dcefclk)(void *handle, uint32_t clock); - bool (*get_asic_baco_capability)(void *handle); + int (*get_asic_baco_capability)(void *handle); int (*get_asic_baco_state)(void *handle, int *state); int (*set_asic_baco_state)(void *handle, int state); int (*get_ppfeature_status)(void *handle, char *buf); diff --git a/drivers/gpu/drm/amd/include/mes_v11_api_def.h b/drivers/gpu/drm/amd/include/mes_v11_api_def.h index ec5b9ab67c5e..b72d5d362251 100644 --- a/drivers/gpu/drm/amd/include/mes_v11_api_def.h +++ b/drivers/gpu/drm/amd/include/mes_v11_api_def.h @@ -61,6 +61,7 @@ enum MES_SCH_API_OPCODE { MES_SCH_API_MISC = 14, MES_SCH_API_UPDATE_ROOT_PAGE_TABLE = 15, MES_SCH_API_AMD_LOG = 16, + MES_SCH_API_SET_HW_RSRC_1 = 19, MES_SCH_API_MAX = 0xFF }; @@ -238,6 +239,26 @@ union MESAPI_SET_HW_RESOURCES { uint32_t max_dwords_in_api[API_FRAME_SIZE_IN_DWORDS]; }; +union MESAPI_SET_HW_RESOURCES_1 { + struct { + union MES_API_HEADER header; + struct MES_API_STATUS api_status; + uint64_t timestamp; + union { + struct { + uint32_t enable_mes_info_ctx : 1; + uint32_t reserved : 31; + }; + uint32_t uint32_all; + }; + uint64_t mes_info_ctx_mc_addr; + uint32_t mes_info_ctx_size; + uint32_t mes_kiq_unmap_timeout; // unit is 100ms + }; + + uint32_t max_dwords_in_api[API_FRAME_SIZE_IN_DWORDS]; +}; + union MESAPI__ADD_QUEUE { struct { union MES_API_HEADER header; @@ -278,10 +299,21 @@ union MESAPI__ADD_QUEUE { uint32_t skip_process_ctx_clear : 1; uint32_t map_legacy_kq : 1; uint32_t exclusively_scheduled : 1; - uint32_t reserved : 17; + uint32_t is_long_running : 1; + uint32_t is_dwm_queue : 1; + uint32_t is_video_blit_queue : 1; + uint32_t reserved : 14; }; - struct MES_API_STATUS api_status; - uint64_t tma_addr; + struct MES_API_STATUS api_status; + uint64_t tma_addr; + uint32_t sch_id; + uint64_t timestamp; + uint32_t process_context_array_index; + uint32_t gang_context_array_index; + uint32_t pipe_id; + uint32_t queue_id; + uint32_t alignment_mode_setting; + uint64_t unmap_flag_addr; }; uint32_t max_dwords_in_api[API_FRAME_SIZE_IN_DWORDS]; diff --git a/drivers/gpu/drm/amd/include/umsch_mm_4_0_api_def.h b/drivers/gpu/drm/amd/include/umsch_mm_4_0_api_def.h index beadb9e42850..ca83e9e5c3ff 100644 --- a/drivers/gpu/drm/amd/include/umsch_mm_4_0_api_def.h +++ b/drivers/gpu/drm/amd/include/umsch_mm_4_0_api_def.h @@ -234,7 +234,8 @@ union UMSCHAPI__SET_HW_RESOURCES { uint32_t enable_level_process_quantum_check : 1; uint32_t is_vcn0_enabled : 1; uint32_t is_vcn1_enabled : 1; - uint32_t reserved : 27; + uint32_t use_rs64mem_for_proc_ctx_csa : 1; + uint32_t reserved : 26; }; uint32_t uint32_all; }; @@ -297,9 +298,12 @@ union UMSCHAPI__ADD_QUEUE { struct { uint32_t is_context_suspended : 1; - uint32_t reserved : 31; + uint32_t collaboration_mode : 1; + uint32_t reserved : 30; }; struct UMSCH_API_STATUS api_status; + uint32_t process_csa_array_index; + uint32_t context_csa_array_index; }; uint32_t max_dwords_in_api[API_FRAME_SIZE_IN_DWORDS]; @@ -314,6 +318,7 @@ union UMSCHAPI__REMOVE_QUEUE { uint64_t context_csa_addr; struct UMSCH_API_STATUS api_status; + uint32_t context_csa_array_index; }; uint32_t max_dwords_in_api[API_FRAME_SIZE_IN_DWORDS]; @@ -337,6 +342,7 @@ union UMSCHAPI__SUSPEND { uint32_t suspend_fence_value; struct UMSCH_API_STATUS api_status; + uint32_t context_csa_array_index; }; uint32_t max_dwords_in_api[API_FRAME_SIZE_IN_DWORDS]; @@ -356,6 +362,7 @@ union UMSCHAPI__RESUME { enum UMSCH_ENGINE_TYPE engine_type; struct UMSCH_API_STATUS api_status; + uint32_t context_csa_array_index; }; uint32_t max_dwords_in_api[API_FRAME_SIZE_IN_DWORDS]; @@ -404,6 +411,7 @@ union UMSCHAPI__UPDATE_AFFINITY { union UMSCH_AFFINITY affinity; uint64_t context_csa_addr; struct UMSCH_API_STATUS api_status; + uint32_t context_csa_array_index; }; uint32_t max_dwords_in_api[API_FRAME_SIZE_IN_DWORDS]; @@ -417,6 +425,7 @@ union UMSCHAPI__CHANGE_CONTEXT_PRIORITY_LEVEL { uint64_t context_quantum; uint64_t context_csa_addr; struct UMSCH_API_STATUS api_status; + uint32_t context_csa_array_index; }; uint32_t max_dwords_in_api[API_FRAME_SIZE_IN_DWORDS]; diff --git a/drivers/gpu/drm/amd/pm/amdgpu_dpm.c b/drivers/gpu/drm/amd/pm/amdgpu_dpm.c index f84bfed50681..eee919577b44 100644 --- a/drivers/gpu/drm/amd/pm/amdgpu_dpm.c +++ b/drivers/gpu/drm/amd/pm/amdgpu_dpm.c @@ -199,14 +199,14 @@ int amdgpu_dpm_notify_rlc_state(struct amdgpu_device *adev, bool en) return ret; } -bool amdgpu_dpm_is_baco_supported(struct amdgpu_device *adev) +int amdgpu_dpm_is_baco_supported(struct amdgpu_device *adev) { const struct amd_pm_funcs *pp_funcs = adev->powerplay.pp_funcs; void *pp_handle = adev->powerplay.pp_handle; - bool ret; + int ret; if (!pp_funcs || !pp_funcs->get_asic_baco_capability) - return false; + return 0; /* Don't use baco for reset in S3. * This is a workaround for some platforms * where entering BACO during suspend @@ -217,7 +217,7 @@ bool amdgpu_dpm_is_baco_supported(struct amdgpu_device *adev) * devices. Needs more investigation. */ if (adev->in_s3) - return false; + return 0; mutex_lock(&adev->pm.mutex); diff --git a/drivers/gpu/drm/amd/pm/amdgpu_pm.c b/drivers/gpu/drm/amd/pm/amdgpu_pm.c index f09b9d49297e..c11952a4389b 100644 --- a/drivers/gpu/drm/amd/pm/amdgpu_pm.c +++ b/drivers/gpu/drm/amd/pm/amdgpu_pm.c @@ -38,6 +38,8 @@ #define MAX_NUM_OF_FEATURES_PER_SUBSET 8 #define MAX_NUM_OF_SUBSETS 8 +#define DEVICE_ATTR_IS(_name) (attr_id == device_attr_id__##_name) + struct od_attribute { struct kobj_attribute attribute; struct list_head entry; @@ -1582,6 +1584,30 @@ static ssize_t amdgpu_get_mem_busy_percent(struct device *dev, } /** + * DOC: vcn_busy_percent + * + * The amdgpu driver provides a sysfs API for reading how busy the VCN + * is as a percentage. The file vcn_busy_percent is used for this. + * The SMU firmware computes a percentage of load based on the + * aggregate activity level in the IP cores. + */ +static ssize_t amdgpu_get_vcn_busy_percent(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + struct drm_device *ddev = dev_get_drvdata(dev); + struct amdgpu_device *adev = drm_to_adev(ddev); + unsigned int value; + int r; + + r = amdgpu_hwmon_get_sensor_generic(adev, AMDGPU_PP_SENSOR_VCN_LOAD, &value); + if (r) + return r; + + return sysfs_emit(buf, "%d\n", value); +} + +/** * DOC: pcie_bw * * The amdgpu driver provides a sysfs API for estimating how much data @@ -2091,6 +2117,99 @@ static int pp_dpm_dcefclk_attr_update(struct amdgpu_device *adev, struct amdgpu_ return 0; } +static int pp_dpm_clk_default_attr_update(struct amdgpu_device *adev, struct amdgpu_device_attr *attr, + uint32_t mask, enum amdgpu_device_attr_states *states) +{ + struct device_attribute *dev_attr = &attr->dev_attr; + enum amdgpu_device_attr_id attr_id = attr->attr_id; + uint32_t mp1_ver = amdgpu_ip_version(adev, MP1_HWIP, 0); + uint32_t gc_ver = amdgpu_ip_version(adev, GC_HWIP, 0); + + *states = ATTR_STATE_SUPPORTED; + + if (!(attr->flags & mask)) { + *states = ATTR_STATE_UNSUPPORTED; + return 0; + } + + if (DEVICE_ATTR_IS(pp_dpm_socclk)) { + if (gc_ver < IP_VERSION(9, 0, 0)) + *states = ATTR_STATE_UNSUPPORTED; + } else if (DEVICE_ATTR_IS(pp_dpm_fclk)) { + if (mp1_ver < IP_VERSION(10, 0, 0)) + *states = ATTR_STATE_UNSUPPORTED; + } else if (DEVICE_ATTR_IS(pp_dpm_vclk)) { + if (!(gc_ver == IP_VERSION(10, 3, 1) || + gc_ver == IP_VERSION(10, 3, 3) || + gc_ver == IP_VERSION(10, 3, 6) || + gc_ver == IP_VERSION(10, 3, 7) || + gc_ver == IP_VERSION(10, 3, 0) || + gc_ver == IP_VERSION(10, 1, 2) || + gc_ver == IP_VERSION(11, 0, 0) || + gc_ver == IP_VERSION(11, 0, 1) || + gc_ver == IP_VERSION(11, 0, 4) || + gc_ver == IP_VERSION(11, 5, 0) || + gc_ver == IP_VERSION(11, 0, 2) || + gc_ver == IP_VERSION(11, 0, 3) || + gc_ver == IP_VERSION(9, 4, 3))) + *states = ATTR_STATE_UNSUPPORTED; + } else if (DEVICE_ATTR_IS(pp_dpm_vclk1)) { + if (!((gc_ver == IP_VERSION(10, 3, 1) || + gc_ver == IP_VERSION(10, 3, 0) || + gc_ver == IP_VERSION(11, 0, 2) || + gc_ver == IP_VERSION(11, 0, 3)) && adev->vcn.num_vcn_inst >= 2)) + *states = ATTR_STATE_UNSUPPORTED; + } else if (DEVICE_ATTR_IS(pp_dpm_dclk)) { + if (!(gc_ver == IP_VERSION(10, 3, 1) || + gc_ver == IP_VERSION(10, 3, 3) || + gc_ver == IP_VERSION(10, 3, 6) || + gc_ver == IP_VERSION(10, 3, 7) || + gc_ver == IP_VERSION(10, 3, 0) || + gc_ver == IP_VERSION(10, 1, 2) || + gc_ver == IP_VERSION(11, 0, 0) || + gc_ver == IP_VERSION(11, 0, 1) || + gc_ver == IP_VERSION(11, 0, 4) || + gc_ver == IP_VERSION(11, 5, 0) || + gc_ver == IP_VERSION(11, 0, 2) || + gc_ver == IP_VERSION(11, 0, 3) || + gc_ver == IP_VERSION(9, 4, 3))) + *states = ATTR_STATE_UNSUPPORTED; + } else if (DEVICE_ATTR_IS(pp_dpm_dclk1)) { + if (!((gc_ver == IP_VERSION(10, 3, 1) || + gc_ver == IP_VERSION(10, 3, 0) || + gc_ver == IP_VERSION(11, 0, 2) || + gc_ver == IP_VERSION(11, 0, 3)) && adev->vcn.num_vcn_inst >= 2)) + *states = ATTR_STATE_UNSUPPORTED; + } else if (DEVICE_ATTR_IS(pp_dpm_pcie)) { + if (gc_ver == IP_VERSION(9, 4, 2) || + gc_ver == IP_VERSION(9, 4, 3)) + *states = ATTR_STATE_UNSUPPORTED; + } + + switch (gc_ver) { + case IP_VERSION(9, 4, 1): + case IP_VERSION(9, 4, 2): + /* the Mi series card does not support standalone mclk/socclk/fclk level setting */ + if (DEVICE_ATTR_IS(pp_dpm_mclk) || + DEVICE_ATTR_IS(pp_dpm_socclk) || + DEVICE_ATTR_IS(pp_dpm_fclk)) { + dev_attr->attr.mode &= ~S_IWUGO; + dev_attr->store = NULL; + } + break; + default: + break; + } + + /* setting should not be allowed from VF if not in one VF mode */ + if (amdgpu_sriov_vf(adev) && amdgpu_sriov_is_pp_one_vf(adev)) { + dev_attr->attr.mode &= ~S_IWUGO; + dev_attr->store = NULL; + } + + return 0; +} + /* Following items will be read out to indicate current plpd policy: * - -1: none * - 0: disallow @@ -2162,17 +2281,26 @@ static struct amdgpu_device_attr amdgpu_device_attrs[] = { AMDGPU_DEVICE_ATTR_RO(pp_cur_state, ATTR_FLAG_BASIC|ATTR_FLAG_ONEVF), AMDGPU_DEVICE_ATTR_RW(pp_force_state, ATTR_FLAG_BASIC|ATTR_FLAG_ONEVF), AMDGPU_DEVICE_ATTR_RW(pp_table, ATTR_FLAG_BASIC|ATTR_FLAG_ONEVF), - AMDGPU_DEVICE_ATTR_RW(pp_dpm_sclk, ATTR_FLAG_BASIC|ATTR_FLAG_ONEVF), - AMDGPU_DEVICE_ATTR_RW(pp_dpm_mclk, ATTR_FLAG_BASIC|ATTR_FLAG_ONEVF), - AMDGPU_DEVICE_ATTR_RW(pp_dpm_socclk, ATTR_FLAG_BASIC|ATTR_FLAG_ONEVF), - AMDGPU_DEVICE_ATTR_RW(pp_dpm_fclk, ATTR_FLAG_BASIC|ATTR_FLAG_ONEVF), - AMDGPU_DEVICE_ATTR_RW(pp_dpm_vclk, ATTR_FLAG_BASIC|ATTR_FLAG_ONEVF), - AMDGPU_DEVICE_ATTR_RW(pp_dpm_vclk1, ATTR_FLAG_BASIC|ATTR_FLAG_ONEVF), - AMDGPU_DEVICE_ATTR_RW(pp_dpm_dclk, ATTR_FLAG_BASIC|ATTR_FLAG_ONEVF), - AMDGPU_DEVICE_ATTR_RW(pp_dpm_dclk1, ATTR_FLAG_BASIC|ATTR_FLAG_ONEVF), + AMDGPU_DEVICE_ATTR_RW(pp_dpm_sclk, ATTR_FLAG_BASIC|ATTR_FLAG_ONEVF, + .attr_update = pp_dpm_clk_default_attr_update), + AMDGPU_DEVICE_ATTR_RW(pp_dpm_mclk, ATTR_FLAG_BASIC|ATTR_FLAG_ONEVF, + .attr_update = pp_dpm_clk_default_attr_update), + AMDGPU_DEVICE_ATTR_RW(pp_dpm_socclk, ATTR_FLAG_BASIC|ATTR_FLAG_ONEVF, + .attr_update = pp_dpm_clk_default_attr_update), + AMDGPU_DEVICE_ATTR_RW(pp_dpm_fclk, ATTR_FLAG_BASIC|ATTR_FLAG_ONEVF, + .attr_update = pp_dpm_clk_default_attr_update), + AMDGPU_DEVICE_ATTR_RW(pp_dpm_vclk, ATTR_FLAG_BASIC|ATTR_FLAG_ONEVF, + .attr_update = pp_dpm_clk_default_attr_update), + AMDGPU_DEVICE_ATTR_RW(pp_dpm_vclk1, ATTR_FLAG_BASIC|ATTR_FLAG_ONEVF, + .attr_update = pp_dpm_clk_default_attr_update), + AMDGPU_DEVICE_ATTR_RW(pp_dpm_dclk, ATTR_FLAG_BASIC|ATTR_FLAG_ONEVF, + .attr_update = pp_dpm_clk_default_attr_update), + AMDGPU_DEVICE_ATTR_RW(pp_dpm_dclk1, ATTR_FLAG_BASIC|ATTR_FLAG_ONEVF, + .attr_update = pp_dpm_clk_default_attr_update), AMDGPU_DEVICE_ATTR_RW(pp_dpm_dcefclk, ATTR_FLAG_BASIC|ATTR_FLAG_ONEVF, .attr_update = pp_dpm_dcefclk_attr_update), - AMDGPU_DEVICE_ATTR_RW(pp_dpm_pcie, ATTR_FLAG_BASIC|ATTR_FLAG_ONEVF), + AMDGPU_DEVICE_ATTR_RW(pp_dpm_pcie, ATTR_FLAG_BASIC|ATTR_FLAG_ONEVF, + .attr_update = pp_dpm_clk_default_attr_update), AMDGPU_DEVICE_ATTR_RW(pp_sclk_od, ATTR_FLAG_BASIC), AMDGPU_DEVICE_ATTR_RW(pp_mclk_od, ATTR_FLAG_BASIC), AMDGPU_DEVICE_ATTR_RW(pp_power_profile_mode, ATTR_FLAG_BASIC|ATTR_FLAG_ONEVF), @@ -2180,6 +2308,7 @@ static struct amdgpu_device_attr amdgpu_device_attrs[] = { .attr_update = pp_od_clk_voltage_attr_update), AMDGPU_DEVICE_ATTR_RO(gpu_busy_percent, ATTR_FLAG_BASIC|ATTR_FLAG_ONEVF), AMDGPU_DEVICE_ATTR_RO(mem_busy_percent, ATTR_FLAG_BASIC|ATTR_FLAG_ONEVF), + AMDGPU_DEVICE_ATTR_RO(vcn_busy_percent, ATTR_FLAG_BASIC|ATTR_FLAG_ONEVF), AMDGPU_DEVICE_ATTR_RO(pcie_bw, ATTR_FLAG_BASIC), AMDGPU_DEVICE_ATTR_RW(pp_features, ATTR_FLAG_BASIC|ATTR_FLAG_ONEVF), AMDGPU_DEVICE_ATTR_RO(unique_id, ATTR_FLAG_BASIC|ATTR_FLAG_ONEVF), @@ -2201,28 +2330,28 @@ static int default_attr_update(struct amdgpu_device *adev, struct amdgpu_device_ uint32_t mask, enum amdgpu_device_attr_states *states) { struct device_attribute *dev_attr = &attr->dev_attr; - uint32_t mp1_ver = amdgpu_ip_version(adev, MP1_HWIP, 0); + enum amdgpu_device_attr_id attr_id = attr->attr_id; uint32_t gc_ver = amdgpu_ip_version(adev, GC_HWIP, 0); - const char *attr_name = dev_attr->attr.name; if (!(attr->flags & mask)) { *states = ATTR_STATE_UNSUPPORTED; return 0; } -#define DEVICE_ATTR_IS(_name) (!strcmp(attr_name, #_name)) - - if (DEVICE_ATTR_IS(pp_dpm_socclk)) { - if (gc_ver < IP_VERSION(9, 0, 0)) - *states = ATTR_STATE_UNSUPPORTED; - } else if (DEVICE_ATTR_IS(pp_dpm_fclk)) { - if (mp1_ver < IP_VERSION(10, 0, 0)) - *states = ATTR_STATE_UNSUPPORTED; - } else if (DEVICE_ATTR_IS(mem_busy_percent)) { + if (DEVICE_ATTR_IS(mem_busy_percent)) { if ((adev->flags & AMD_IS_APU && gc_ver != IP_VERSION(9, 4, 3)) || gc_ver == IP_VERSION(9, 0, 1)) *states = ATTR_STATE_UNSUPPORTED; + } else if (DEVICE_ATTR_IS(vcn_busy_percent)) { + if (!(gc_ver == IP_VERSION(10, 3, 1) || + gc_ver == IP_VERSION(10, 3, 3) || + gc_ver == IP_VERSION(10, 3, 6) || + gc_ver == IP_VERSION(10, 3, 7) || + gc_ver == IP_VERSION(11, 0, 1) || + gc_ver == IP_VERSION(11, 0, 4) || + gc_ver == IP_VERSION(11, 5, 0))) + *states = ATTR_STATE_UNSUPPORTED; } else if (DEVICE_ATTR_IS(pcie_bw)) { /* PCIe Perf counters won't work on APU nodes */ if (adev->flags & AMD_IS_APU || @@ -2253,36 +2382,6 @@ static int default_attr_update(struct amdgpu_device *adev, struct amdgpu_device_ } else if (DEVICE_ATTR_IS(gpu_metrics)) { if (gc_ver < IP_VERSION(9, 1, 0)) *states = ATTR_STATE_UNSUPPORTED; - } else if (DEVICE_ATTR_IS(pp_dpm_vclk)) { - if (!(gc_ver == IP_VERSION(10, 3, 1) || - gc_ver == IP_VERSION(10, 3, 0) || - gc_ver == IP_VERSION(10, 1, 2) || - gc_ver == IP_VERSION(11, 0, 0) || - gc_ver == IP_VERSION(11, 0, 2) || - gc_ver == IP_VERSION(11, 0, 3) || - gc_ver == IP_VERSION(9, 4, 3))) - *states = ATTR_STATE_UNSUPPORTED; - } else if (DEVICE_ATTR_IS(pp_dpm_vclk1)) { - if (!((gc_ver == IP_VERSION(10, 3, 1) || - gc_ver == IP_VERSION(10, 3, 0) || - gc_ver == IP_VERSION(11, 0, 2) || - gc_ver == IP_VERSION(11, 0, 3)) && adev->vcn.num_vcn_inst >= 2)) - *states = ATTR_STATE_UNSUPPORTED; - } else if (DEVICE_ATTR_IS(pp_dpm_dclk)) { - if (!(gc_ver == IP_VERSION(10, 3, 1) || - gc_ver == IP_VERSION(10, 3, 0) || - gc_ver == IP_VERSION(10, 1, 2) || - gc_ver == IP_VERSION(11, 0, 0) || - gc_ver == IP_VERSION(11, 0, 2) || - gc_ver == IP_VERSION(11, 0, 3) || - gc_ver == IP_VERSION(9, 4, 3))) - *states = ATTR_STATE_UNSUPPORTED; - } else if (DEVICE_ATTR_IS(pp_dpm_dclk1)) { - if (!((gc_ver == IP_VERSION(10, 3, 1) || - gc_ver == IP_VERSION(10, 3, 0) || - gc_ver == IP_VERSION(11, 0, 2) || - gc_ver == IP_VERSION(11, 0, 3)) && adev->vcn.num_vcn_inst >= 2)) - *states = ATTR_STATE_UNSUPPORTED; } else if (DEVICE_ATTR_IS(pp_power_profile_mode)) { if (amdgpu_dpm_get_power_profile_mode(adev, NULL) == -EOPNOTSUPP) *states = ATTR_STATE_UNSUPPORTED; @@ -2304,23 +2403,9 @@ static int default_attr_update(struct amdgpu_device *adev, struct amdgpu_device_ if (amdgpu_dpm_get_apu_thermal_limit(adev, &limit) == -EOPNOTSUPP) *states = ATTR_STATE_UNSUPPORTED; - } else if (DEVICE_ATTR_IS(pp_dpm_pcie)) { - if (gc_ver == IP_VERSION(9, 4, 2) || - gc_ver == IP_VERSION(9, 4, 3)) - *states = ATTR_STATE_UNSUPPORTED; } switch (gc_ver) { - case IP_VERSION(9, 4, 1): - case IP_VERSION(9, 4, 2): - /* the Mi series card does not support standalone mclk/socclk/fclk level setting */ - if (DEVICE_ATTR_IS(pp_dpm_mclk) || - DEVICE_ATTR_IS(pp_dpm_socclk) || - DEVICE_ATTR_IS(pp_dpm_fclk)) { - dev_attr->attr.mode &= ~S_IWUGO; - dev_attr->store = NULL; - } - break; case IP_VERSION(10, 3, 0): if (DEVICE_ATTR_IS(power_dpm_force_performance_level) && amdgpu_sriov_vf(adev)) { @@ -2332,14 +2417,6 @@ static int default_attr_update(struct amdgpu_device *adev, struct amdgpu_device_ break; } - /* setting should not be allowed from VF if not in one VF mode */ - if (amdgpu_sriov_vf(adev) && !amdgpu_sriov_is_pp_one_vf(adev)) { - dev_attr->attr.mode &= ~S_IWUGO; - dev_attr->store = NULL; - } - -#undef DEVICE_ATTR_IS - return 0; } @@ -4261,6 +4338,13 @@ static int amdgpu_od_set_init(struct amdgpu_device *adev) } } + /* + * If gpu_od is the only member in the list, that means gpu_od is an + * empty directory, so remove it. + */ + if (list_is_singular(&adev->pm.od_kobj_list)) + goto err_out; + return 0; err_out: @@ -4322,6 +4406,8 @@ int amdgpu_pm_sysfs_init(struct amdgpu_device *adev) ret = amdgpu_od_set_init(adev); if (ret) goto err_out1; + } else if (adev->pm.pp_feature & PP_OVERDRIVE_MASK) { + dev_info(adev->dev, "overdrive feature is not supported\n"); } adev->pm.sysfs_initialized = true; @@ -4429,6 +4515,9 @@ static int amdgpu_debugfs_pm_info_pp(struct seq_file *m, struct amdgpu_device *a /* MEM Load */ if (!amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_MEM_LOAD, (void *)&value, &size)) seq_printf(m, "MEM Load: %u %%\n", value); + /* VCN Load */ + if (!amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_VCN_LOAD, (void *)&value, &size)) + seq_printf(m, "VCN Load: %u %%\n", value); seq_printf(m, "\n"); diff --git a/drivers/gpu/drm/amd/pm/inc/amdgpu_dpm.h b/drivers/gpu/drm/amd/pm/inc/amdgpu_dpm.h index 621200e0823f..501f8c726e8d 100644 --- a/drivers/gpu/drm/amd/pm/inc/amdgpu_dpm.h +++ b/drivers/gpu/drm/amd/pm/inc/amdgpu_dpm.h @@ -50,8 +50,12 @@ enum amdgpu_runpm_mode { AMDGPU_RUNPM_PX, AMDGPU_RUNPM_BOCO, AMDGPU_RUNPM_BACO, + AMDGPU_RUNPM_BAMACO, }; +#define BACO_SUPPORT (1<<0) +#define MACO_SUPPORT (1<<1) + struct amdgpu_ps { u32 caps; /* vbios flags */ u32 class; /* vbios flags */ @@ -407,7 +411,7 @@ int amdgpu_dpm_baco_reset(struct amdgpu_device *adev); int amdgpu_dpm_mode2_reset(struct amdgpu_device *adev); int amdgpu_dpm_enable_gfx_features(struct amdgpu_device *adev); -bool amdgpu_dpm_is_baco_supported(struct amdgpu_device *adev); +int amdgpu_dpm_is_baco_supported(struct amdgpu_device *adev); bool amdgpu_dpm_is_mode1_reset_supported(struct amdgpu_device *adev); int amdgpu_dpm_mode1_reset(struct amdgpu_device *adev); diff --git a/drivers/gpu/drm/amd/pm/inc/amdgpu_pm.h b/drivers/gpu/drm/amd/pm/inc/amdgpu_pm.h index eec816f0cbf9..448ba3a14584 100644 --- a/drivers/gpu/drm/amd/pm/inc/amdgpu_pm.h +++ b/drivers/gpu/drm/amd/pm/inc/amdgpu_pm.h @@ -43,8 +43,48 @@ enum amdgpu_device_attr_states { ATTR_STATE_SUPPORTED, }; +enum amdgpu_device_attr_id { + device_attr_id__unknown = -1, + device_attr_id__power_dpm_state = 0, + device_attr_id__power_dpm_force_performance_level, + device_attr_id__pp_num_states, + device_attr_id__pp_cur_state, + device_attr_id__pp_force_state, + device_attr_id__pp_table, + device_attr_id__pp_dpm_sclk, + device_attr_id__pp_dpm_mclk, + device_attr_id__pp_dpm_socclk, + device_attr_id__pp_dpm_fclk, + device_attr_id__pp_dpm_vclk, + device_attr_id__pp_dpm_vclk1, + device_attr_id__pp_dpm_dclk, + device_attr_id__pp_dpm_dclk1, + device_attr_id__pp_dpm_dcefclk, + device_attr_id__pp_dpm_pcie, + device_attr_id__pp_sclk_od, + device_attr_id__pp_mclk_od, + device_attr_id__pp_power_profile_mode, + device_attr_id__pp_od_clk_voltage, + device_attr_id__gpu_busy_percent, + device_attr_id__mem_busy_percent, + device_attr_id__vcn_busy_percent, + device_attr_id__pcie_bw, + device_attr_id__pp_features, + device_attr_id__unique_id, + device_attr_id__thermal_throttling_logging, + device_attr_id__apu_thermal_cap, + device_attr_id__gpu_metrics, + device_attr_id__smartshift_apu_power, + device_attr_id__smartshift_dgpu_power, + device_attr_id__smartshift_bias, + device_attr_id__xgmi_plpd_policy, + device_attr_id__pm_metrics, + device_attr_id__count, +}; + struct amdgpu_device_attr { struct device_attribute dev_attr; + enum amdgpu_device_attr_id attr_id; enum amdgpu_device_attr_flags flags; int (*attr_update)(struct amdgpu_device *adev, struct amdgpu_device_attr *attr, uint32_t mask, enum amdgpu_device_attr_states *states); @@ -61,6 +101,7 @@ struct amdgpu_device_attr_entry { #define __AMDGPU_DEVICE_ATTR(_name, _mode, _show, _store, _flags, ...) \ { .dev_attr = __ATTR(_name, _mode, _show, _store), \ + .attr_id = device_attr_id__##_name, \ .flags = _flags, \ ##__VA_ARGS__, } diff --git a/drivers/gpu/drm/amd/pm/legacy-dpm/kv_dpm.c b/drivers/gpu/drm/amd/pm/legacy-dpm/kv_dpm.c index 5cb4725c773f..6bb42d04b247 100644 --- a/drivers/gpu/drm/amd/pm/legacy-dpm/kv_dpm.c +++ b/drivers/gpu/drm/amd/pm/legacy-dpm/kv_dpm.c @@ -3316,6 +3316,8 @@ static const struct amd_ip_funcs kv_dpm_ip_funcs = { .soft_reset = kv_dpm_soft_reset, .set_clockgating_state = kv_dpm_set_clockgating_state, .set_powergating_state = kv_dpm_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; const struct amdgpu_ip_block_version kv_smu_ip_block = { diff --git a/drivers/gpu/drm/amd/pm/legacy-dpm/si_dpm.c b/drivers/gpu/drm/amd/pm/legacy-dpm/si_dpm.c index eb4da3666e05..f245fc0bc6d3 100644 --- a/drivers/gpu/drm/amd/pm/legacy-dpm/si_dpm.c +++ b/drivers/gpu/drm/amd/pm/legacy-dpm/si_dpm.c @@ -8060,6 +8060,8 @@ static const struct amd_ip_funcs si_dpm_ip_funcs = { .soft_reset = si_dpm_soft_reset, .set_clockgating_state = si_dpm_set_clockgating_state, .set_powergating_state = si_dpm_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; const struct amdgpu_ip_block_version si_smu_ip_block = diff --git a/drivers/gpu/drm/amd/pm/powerplay/amd_powerplay.c b/drivers/gpu/drm/amd/pm/powerplay/amd_powerplay.c index aed0e2cefbf9..5fb21a0508cd 100644 --- a/drivers/gpu/drm/amd/pm/powerplay/amd_powerplay.c +++ b/drivers/gpu/drm/amd/pm/powerplay/amd_powerplay.c @@ -302,6 +302,8 @@ static const struct amd_ip_funcs pp_ip_funcs = { .soft_reset = pp_sw_reset, .set_clockgating_state = pp_set_clockgating_state, .set_powergating_state = pp_set_powergating_state, + .dump_ip_state = NULL, + .print_ip_state = NULL, }; const struct amdgpu_ip_block_version pp_smu_ip_block = @@ -1371,7 +1373,7 @@ static int pp_set_active_display_count(void *handle, uint32_t count) return phm_set_active_display_count(hwmgr, count); } -static bool pp_get_asic_baco_capability(void *handle) +static int pp_get_asic_baco_capability(void *handle) { struct pp_hwmgr *hwmgr = handle; @@ -1379,10 +1381,10 @@ static bool pp_get_asic_baco_capability(void *handle) return false; if (!(hwmgr->not_vf && amdgpu_dpm) || - !hwmgr->hwmgr_func->get_asic_baco_capability) + !hwmgr->hwmgr_func->get_bamaco_support) return false; - return hwmgr->hwmgr_func->get_asic_baco_capability(hwmgr); + return hwmgr->hwmgr_func->get_bamaco_support(hwmgr); } static int pp_get_asic_baco_state(void *handle, int *state) diff --git a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/smu7_baco.c b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/smu7_baco.c index e8a9471c1898..ad60918aaae1 100644 --- a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/smu7_baco.c +++ b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/smu7_baco.c @@ -33,7 +33,7 @@ #include "smu/smu_7_1_2_d.h" #include "smu/smu_7_1_2_sh_mask.h" -bool smu7_baco_get_capability(struct pp_hwmgr *hwmgr) +int smu7_get_bamaco_support(struct pp_hwmgr *hwmgr) { struct amdgpu_device *adev = (struct amdgpu_device *)(hwmgr->adev); uint32_t reg; @@ -44,9 +44,9 @@ bool smu7_baco_get_capability(struct pp_hwmgr *hwmgr) reg = RREG32(mmCC_BIF_BX_FUSESTRAP0); if (reg & CC_BIF_BX_FUSESTRAP0__STRAP_BIF_PX_CAPABLE_MASK) - return true; + return BACO_SUPPORT; - return false; + return 0; } int smu7_baco_get_state(struct pp_hwmgr *hwmgr, enum BACO_STATE *state) diff --git a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/smu7_baco.h b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/smu7_baco.h index 73a773f4ce2e..750082ea74d8 100644 --- a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/smu7_baco.h +++ b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/smu7_baco.h @@ -25,7 +25,7 @@ #include "hwmgr.h" #include "common_baco.h" -extern bool smu7_baco_get_capability(struct pp_hwmgr *hwmgr); +extern int smu7_get_bamaco_support(struct pp_hwmgr *hwmgr); extern int smu7_baco_get_state(struct pp_hwmgr *hwmgr, enum BACO_STATE *state); extern int smu7_baco_set_state(struct pp_hwmgr *hwmgr, enum BACO_STATE state); diff --git a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/smu7_hwmgr.c b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/smu7_hwmgr.c index aa91730e4eaf..1fcd4451001f 100644 --- a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/smu7_hwmgr.c +++ b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/smu7_hwmgr.c @@ -5791,7 +5791,7 @@ static const struct pp_hwmgr_func smu7_hwmgr_funcs = { .get_power_profile_mode = smu7_get_power_profile_mode, .set_power_profile_mode = smu7_set_power_profile_mode, .get_performance_level = smu7_get_performance_level, - .get_asic_baco_capability = smu7_baco_get_capability, + .get_bamaco_support = smu7_get_bamaco_support, .get_asic_baco_state = smu7_baco_get_state, .set_asic_baco_state = smu7_baco_set_state, .power_off_asic = smu7_power_off_asic, diff --git a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/smu9_baco.c b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/smu9_baco.c index c66ef9741535..c1ce1d7cae48 100644 --- a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/smu9_baco.c +++ b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/smu9_baco.c @@ -28,13 +28,13 @@ #include "vega10_inc.h" #include "smu9_baco.h" -bool smu9_baco_get_capability(struct pp_hwmgr *hwmgr) +int smu9_get_bamaco_support(struct pp_hwmgr *hwmgr) { struct amdgpu_device *adev = (struct amdgpu_device *)(hwmgr->adev); uint32_t reg, data; if (!phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_BACO)) - return false; + return 0; WREG32(0x12074, 0xFFF0003B); data = RREG32(0x12075); @@ -43,10 +43,10 @@ bool smu9_baco_get_capability(struct pp_hwmgr *hwmgr) reg = RREG32_SOC15(NBIF, 0, mmRCC_BIF_STRAP0); if (reg & RCC_BIF_STRAP0__STRAP_PX_CAPABLE_MASK) - return true; + return BACO_SUPPORT; } - return false; + return 0; } int smu9_baco_get_state(struct pp_hwmgr *hwmgr, enum BACO_STATE *state) diff --git a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/smu9_baco.h b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/smu9_baco.h index 9ff7c2ea1b58..2c100482084c 100644 --- a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/smu9_baco.h +++ b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/smu9_baco.h @@ -25,7 +25,7 @@ #include "hwmgr.h" #include "common_baco.h" -extern bool smu9_baco_get_capability(struct pp_hwmgr *hwmgr); +extern int smu9_get_bamaco_support(struct pp_hwmgr *hwmgr); extern int smu9_baco_get_state(struct pp_hwmgr *hwmgr, enum BACO_STATE *state); #endif diff --git a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega10_hwmgr.c b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega10_hwmgr.c index 6d6bc6a380b3..9f5bd998c6bf 100644 --- a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega10_hwmgr.c +++ b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega10_hwmgr.c @@ -5756,7 +5756,7 @@ static const struct pp_hwmgr_func vega10_hwmgr_funcs = { .set_power_limit = vega10_set_power_limit, .odn_edit_dpm_table = vega10_odn_edit_dpm_table, .get_performance_level = vega10_get_performance_level, - .get_asic_baco_capability = smu9_baco_get_capability, + .get_bamaco_support = smu9_get_bamaco_support, .get_asic_baco_state = smu9_baco_get_state, .set_asic_baco_state = vega10_baco_set_state, .enable_mgpu_fan_boost = vega10_enable_mgpu_fan_boost, diff --git a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega12_hwmgr.c b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega12_hwmgr.c index 460067933de2..c223e3a6bfca 100644 --- a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega12_hwmgr.c +++ b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega12_hwmgr.c @@ -2966,7 +2966,7 @@ static const struct pp_hwmgr_func vega12_hwmgr_funcs = { .start_thermal_controller = vega12_start_thermal_controller, .powergate_gfx = vega12_gfx_off_control, .get_performance_level = vega12_get_performance_level, - .get_asic_baco_capability = smu9_baco_get_capability, + .get_bamaco_support = smu9_get_bamaco_support, .get_asic_baco_state = smu9_baco_get_state, .set_asic_baco_state = vega12_baco_set_state, .get_ppfeature_status = vega12_get_ppfeature_status, diff --git a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega20_baco.c b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega20_baco.c index dad4c80aee58..424e4ec9e389 100644 --- a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega20_baco.c +++ b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega20_baco.c @@ -36,22 +36,22 @@ static const struct soc15_baco_cmd_entry clean_baco_tbl[] = { {CMD_WRITE, SOC15_REG_ENTRY(NBIF, 0, mmBIOS_SCRATCH_7), 0, 0, 0, 0}, }; -bool vega20_baco_get_capability(struct pp_hwmgr *hwmgr) +int vega20_get_bamaco_support(struct pp_hwmgr *hwmgr) { struct amdgpu_device *adev = (struct amdgpu_device *)(hwmgr->adev); uint32_t reg; if (!phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_BACO)) - return false; + return 0; if (((RREG32(0x17569) & 0x20000000) >> 29) == 0x1) { reg = RREG32_SOC15(NBIF, 0, mmRCC_BIF_STRAP0); if (reg & RCC_BIF_STRAP0__STRAP_PX_CAPABLE_MASK) - return true; + return BACO_SUPPORT; } - return false; + return 0; } int vega20_baco_get_state(struct pp_hwmgr *hwmgr, enum BACO_STATE *state) diff --git a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega20_baco.h b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega20_baco.h index bdad9c915631..0f2dd8c008ba 100644 --- a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega20_baco.h +++ b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega20_baco.h @@ -25,7 +25,7 @@ #include "hwmgr.h" #include "common_baco.h" -extern bool vega20_baco_get_capability(struct pp_hwmgr *hwmgr); +extern int vega20_get_bamaco_support(struct pp_hwmgr *hwmgr); extern int vega20_baco_get_state(struct pp_hwmgr *hwmgr, enum BACO_STATE *state); extern int vega20_baco_set_state(struct pp_hwmgr *hwmgr, enum BACO_STATE state); extern int vega20_baco_apply_vdci_flush_workaround(struct pp_hwmgr *hwmgr); diff --git a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega20_hwmgr.c b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega20_hwmgr.c index 3b33af30eb0f..f9efb0bad807 100644 --- a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega20_hwmgr.c +++ b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega20_hwmgr.c @@ -4422,7 +4422,7 @@ static const struct pp_hwmgr_func vega20_hwmgr_funcs = { .notify_cac_buffer_info = vega20_notify_cac_buffer_info, .enable_mgpu_fan_boost = vega20_enable_mgpu_fan_boost, /* BACO related */ - .get_asic_baco_capability = vega20_baco_get_capability, + .get_bamaco_support = vega20_get_bamaco_support, .get_asic_baco_state = vega20_baco_get_state, .set_asic_baco_state = vega20_baco_set_state, .set_mp1_state = vega20_set_mp1_state, diff --git a/drivers/gpu/drm/amd/pm/powerplay/inc/hwmgr.h b/drivers/gpu/drm/amd/pm/powerplay/inc/hwmgr.h index 6f536159df4d..69928a4a074b 100644 --- a/drivers/gpu/drm/amd/pm/powerplay/inc/hwmgr.h +++ b/drivers/gpu/drm/amd/pm/powerplay/inc/hwmgr.h @@ -351,7 +351,7 @@ struct pp_hwmgr_func { int (*set_hard_min_fclk_by_freq)(struct pp_hwmgr *hwmgr, uint32_t clock); int (*set_hard_min_gfxclk_by_freq)(struct pp_hwmgr *hwmgr, uint32_t clock); int (*set_soft_max_gfxclk_by_freq)(struct pp_hwmgr *hwmgr, uint32_t clock); - bool (*get_asic_baco_capability)(struct pp_hwmgr *hwmgr); + int (*get_bamaco_support)(struct pp_hwmgr *hwmgr); int (*get_asic_baco_state)(struct pp_hwmgr *hwmgr, enum BACO_STATE *state); int (*set_asic_baco_state)(struct pp_hwmgr *hwmgr, enum BACO_STATE state); int (*get_ppfeature_status)(struct pp_hwmgr *hwmgr, char *buf); diff --git a/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c b/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c index 246b211b1e85..7789b313285c 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c +++ b/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c @@ -45,6 +45,7 @@ #include "smu_v13_0_6_ppt.h" #include "smu_v13_0_7_ppt.h" #include "smu_v14_0_0_ppt.h" +#include "smu_v14_0_2_ppt.h" #include "amd_pcie.h" /* @@ -715,6 +716,10 @@ static int smu_set_funcs(struct amdgpu_device *adev) case IP_VERSION(14, 0, 1): smu_v14_0_0_set_ppt_funcs(smu); break; + case IP_VERSION(14, 0, 2): + case IP_VERSION(14, 0, 3): + smu_v14_0_2_set_ppt_funcs(smu); + break; default: return -EINVAL; } @@ -735,8 +740,9 @@ static int smu_early_init(void *handle) smu->adev = adev; smu->pm_enabled = !!amdgpu_dpm; smu->is_apu = false; - smu->smu_baco.state = SMU_BACO_STATE_EXIT; + smu->smu_baco.state = SMU_BACO_STATE_NONE; smu->smu_baco.platform_support = false; + smu->smu_baco.maco_support = false; smu->user_dpm_profile.fan_mode = -1; mutex_init(&smu->message_lock); @@ -1966,10 +1972,25 @@ static int smu_smc_hw_cleanup(struct smu_context *smu) return 0; } +static int smu_reset_mp1_state(struct smu_context *smu) +{ + struct amdgpu_device *adev = smu->adev; + int ret = 0; + + if ((!adev->in_runpm) && (!adev->in_suspend) && + (!amdgpu_in_reset(adev)) && amdgpu_ip_version(adev, MP1_HWIP, 0) == + IP_VERSION(13, 0, 10) && + !amdgpu_device_has_display_hardware(adev)) + ret = smu_set_mp1_state(smu, PP_MP1_STATE_UNLOAD); + + return ret; +} + static int smu_hw_fini(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; struct smu_context *smu = adev->powerplay.pp_handle; + int ret; if (amdgpu_sriov_vf(adev) && !amdgpu_sriov_is_pp_one_vf(adev)) return 0; @@ -1987,7 +2008,15 @@ static int smu_hw_fini(void *handle) adev->pm.dpm_enabled = false; - return smu_smc_hw_cleanup(smu); + ret = smu_smc_hw_cleanup(smu); + if (ret) + return ret; + + ret = smu_reset_mp1_state(smu); + if (ret) + return ret; + + return 0; } static void smu_late_fini(void *handle) @@ -3200,17 +3229,17 @@ static int smu_set_xgmi_pstate(void *handle, return ret; } -static bool smu_get_baco_capability(void *handle) +static int smu_get_baco_capability(void *handle) { struct smu_context *smu = handle; if (!smu->pm_enabled) return false; - if (!smu->ppt_funcs || !smu->ppt_funcs->baco_is_support) + if (!smu->ppt_funcs || !smu->ppt_funcs->get_bamaco_support) return false; - return smu->ppt_funcs->baco_is_support(smu); + return smu->ppt_funcs->get_bamaco_support(smu); } static int smu_baco_set_state(void *handle, int state) diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h b/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h index a870bdd49a4e..0917dec8efe3 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h @@ -424,6 +424,7 @@ enum smu_reset_mode { enum smu_baco_state { SMU_BACO_STATE_ENTER = 0, SMU_BACO_STATE_EXIT, + SMU_BACO_STATE_NONE, }; struct smu_baco_context { @@ -458,7 +459,7 @@ struct smu_umd_pstate_table { struct cmn2asic_msg_mapping { int valid_mapping; int map_to; - int valid_in_vf; + uint32_t flags; }; struct cmn2asic_mapping { @@ -538,6 +539,7 @@ struct smu_context { uint32_t smc_driver_if_version; uint32_t smc_fw_if_version; uint32_t smc_fw_version; + uint32_t smc_fw_caps; bool uploading_custom_pp_table; bool dc_controlled_by_gpio; @@ -1173,9 +1175,11 @@ struct pptable_funcs { int (*get_max_sustainable_clocks_by_dc)(struct smu_context *smu, struct pp_smu_nv_clock_table *max_clocks); /** - * @baco_is_support: Check if GPU supports BACO (Bus Active, Chip Off). + * @get_bamaco_support: Check if GPU supports BACO/MACO + * BACO: Bus Active, Chip Off + * MACO: Memory Active, Chip Off */ - bool (*baco_is_support)(struct smu_context *smu); + int (*get_bamaco_support)(struct smu_context *smu); /** * @baco_get_state: Get the current BACO state. @@ -1482,8 +1486,8 @@ enum smu_baco_seq { BACO_SEQ_COUNT, }; -#define MSG_MAP(msg, index, valid_in_vf) \ - [SMU_MSG_##msg] = {1, (index), (valid_in_vf)} +#define MSG_MAP(msg, index, flags) \ + [SMU_MSG_##msg] = {1, (index), (flags)} #define CLK_MAP(clk, index) \ [SMU_##clk] = {1, (index)} diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu14_driver_if_v14_0.h b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu14_driver_if_v14_0.h new file mode 100644 index 000000000000..97a29b80fb13 --- /dev/null +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu14_driver_if_v14_0.h @@ -0,0 +1,1836 @@ +/* + * Copyright 2023 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ + +#ifndef SMU14_DRIVER_IF_V14_0_H +#define SMU14_DRIVER_IF_V14_0_H + +//Increment this version if SkuTable_t or BoardTable_t change +#define PPTABLE_VERSION 0x18 + +#define NUM_GFXCLK_DPM_LEVELS 16 +#define NUM_SOCCLK_DPM_LEVELS 8 +#define NUM_MP0CLK_DPM_LEVELS 2 +#define NUM_DCLK_DPM_LEVELS 8 +#define NUM_VCLK_DPM_LEVELS 8 +#define NUM_DISPCLK_DPM_LEVELS 8 +#define NUM_DPPCLK_DPM_LEVELS 8 +#define NUM_DPREFCLK_DPM_LEVELS 8 +#define NUM_DCFCLK_DPM_LEVELS 8 +#define NUM_DTBCLK_DPM_LEVELS 8 +#define NUM_UCLK_DPM_LEVELS 6 +#define NUM_LINK_LEVELS 3 +#define NUM_FCLK_DPM_LEVELS 8 +#define NUM_OD_FAN_MAX_POINTS 6 + +// Feature Control Defines +#define FEATURE_FW_DATA_READ_BIT 0 +#define FEATURE_DPM_GFXCLK_BIT 1 +#define FEATURE_DPM_GFX_POWER_OPTIMIZER_BIT 2 +#define FEATURE_DPM_UCLK_BIT 3 +#define FEATURE_DPM_FCLK_BIT 4 +#define FEATURE_DPM_SOCCLK_BIT 5 +#define FEATURE_DPM_LINK_BIT 6 +#define FEATURE_DPM_DCN_BIT 7 +#define FEATURE_VMEMP_SCALING_BIT 8 +#define FEATURE_VDDIO_MEM_SCALING_BIT 9 +#define FEATURE_DS_GFXCLK_BIT 10 +#define FEATURE_DS_SOCCLK_BIT 11 +#define FEATURE_DS_FCLK_BIT 12 +#define FEATURE_DS_LCLK_BIT 13 +#define FEATURE_DS_DCFCLK_BIT 14 +#define FEATURE_DS_UCLK_BIT 15 +#define FEATURE_GFX_ULV_BIT 16 +#define FEATURE_FW_DSTATE_BIT 17 +#define FEATURE_GFXOFF_BIT 18 +#define FEATURE_BACO_BIT 19 +#define FEATURE_MM_DPM_BIT 20 +#define FEATURE_SOC_MPCLK_DS_BIT 21 +#define FEATURE_BACO_MPCLK_DS_BIT 22 +#define FEATURE_THROTTLERS_BIT 23 +#define FEATURE_SMARTSHIFT_BIT 24 +#define FEATURE_GTHR_BIT 25 +#define FEATURE_ACDC_BIT 26 +#define FEATURE_VR0HOT_BIT 27 +#define FEATURE_FW_CTF_BIT 28 +#define FEATURE_FAN_CONTROL_BIT 29 +#define FEATURE_GFX_DCS_BIT 30 +#define FEATURE_GFX_READ_MARGIN_BIT 31 +#define FEATURE_LED_DISPLAY_BIT 32 +#define FEATURE_GFXCLK_SPREAD_SPECTRUM_BIT 33 +#define FEATURE_OUT_OF_BAND_MONITOR_BIT 34 +#define FEATURE_OPTIMIZED_VMIN_BIT 35 +#define FEATURE_GFX_IMU_BIT 36 +#define FEATURE_BOOT_TIME_CAL_BIT 37 +#define FEATURE_GFX_PCC_DFLL_BIT 38 +#define FEATURE_SOC_CG_BIT 39 +#define FEATURE_DF_CSTATE_BIT 40 +#define FEATURE_GFX_EDC_BIT 41 +#define FEATURE_BOOT_POWER_OPT_BIT 42 +#define FEATURE_CLOCK_POWER_DOWN_BYPASS_BIT 43 +#define FEATURE_DS_VCN_BIT 44 +#define FEATURE_BACO_CG_BIT 45 +#define FEATURE_MEM_TEMP_READ_BIT 46 +#define FEATURE_ATHUB_MMHUB_PG_BIT 47 +#define FEATURE_SOC_PCC_BIT 48 +#define FEATURE_EDC_PWRBRK_BIT 49 +#define FEATURE_SOC_EDC_XVMIN_BIT 50 +#define FEATURE_GFX_PSM_DIDT_BIT 51 +#define FEATURE_APT_ALL_ENABLE_BIT 52 +#define FEATURE_APT_SQ_THROTTLE_BIT 53 +#define FEATURE_APT_PF_DCS_BIT 54 +#define FEATURE_GFX_EDC_XVMIN_BIT 55 +#define FEATURE_GFX_DIDT_XVMIN_BIT 56 +#define FEATURE_FAN_ABNORMAL_BIT 57 +#define FEATURE_CLOCK_STRETCH_COMPENSATOR 58 +#define FEATURE_SPARE_59_BIT 59 +#define FEATURE_SPARE_60_BIT 60 +#define FEATURE_SPARE_61_BIT 61 +#define FEATURE_SPARE_62_BIT 62 +#define FEATURE_SPARE_63_BIT 63 +#define NUM_FEATURES 64 + +#define ALLOWED_FEATURE_CTRL_DEFAULT 0xFFFFFFFFFFFFFFFFULL +#define ALLOWED_FEATURE_CTRL_SCPM (1 << FEATURE_DPM_GFXCLK_BIT) | \ + (1 << FEATURE_DPM_GFX_POWER_OPTIMIZER_BIT) | \ + (1 << FEATURE_DPM_UCLK_BIT) | \ + (1 << FEATURE_DPM_FCLK_BIT) | \ + (1 << FEATURE_DPM_SOCCLK_BIT) | \ + (1 << FEATURE_DPM_LINK_BIT) | \ + (1 << FEATURE_DPM_DCN_BIT) | \ + (1 << FEATURE_DS_GFXCLK_BIT) | \ + (1 << FEATURE_DS_SOCCLK_BIT) | \ + (1 << FEATURE_DS_FCLK_BIT) | \ + (1 << FEATURE_DS_LCLK_BIT) | \ + (1 << FEATURE_DS_DCFCLK_BIT) | \ + (1 << FEATURE_DS_UCLK_BIT) | \ + (1ULL << FEATURE_DS_VCN_BIT) + + +//For use with feature control messages +typedef enum { + FEATURE_PWR_ALL, + FEATURE_PWR_S5, + FEATURE_PWR_BACO, + FEATURE_PWR_SOC, + FEATURE_PWR_GFX, + FEATURE_PWR_DOMAIN_COUNT, +} FEATURE_PWR_DOMAIN_e; + +//For use with feature control + BTC save restore +typedef enum { + FEATURE_BTC_NOP, + FEATURE_BTC_SAVE, + FEATURE_BTC_RESTORE, + FEATURE_BTC_COUNT, +} FEATURE_BTC_e; + +// Debug Overrides Bitmask +#define DEBUG_OVERRIDE_DISABLE_VOLT_LINK_VCN_FCLK 0x00000001 +#define DEBUG_OVERRIDE_DISABLE_VOLT_LINK_DCN_FCLK 0x00000002 +#define DEBUG_OVERRIDE_DISABLE_VOLT_LINK_MP0_FCLK 0x00000004 +#define DEBUG_OVERRIDE_DISABLE_VOLT_LINK_VCN_DCFCLK 0x00000008 +#define DEBUG_OVERRIDE_DISABLE_FAST_FCLK_TIMER 0x00000010 +#define DEBUG_OVERRIDE_DISABLE_VCN_PG 0x00000020 +#define DEBUG_OVERRIDE_DISABLE_FMAX_VMAX 0x00000040 +#define DEBUG_OVERRIDE_DISABLE_IMU_FW_CHECKS 0x00000080 +#define DEBUG_OVERRIDE_DISABLE_D0i2_REENTRY_HSR_TIMER_CHECK 0x00000100 +#define DEBUG_OVERRIDE_DISABLE_DFLL 0x00000200 +#define DEBUG_OVERRIDE_ENABLE_RLC_VF_BRINGUP_MODE 0x00000400 +#define DEBUG_OVERRIDE_DFLL_MASTER_MODE 0x00000800 +#define DEBUG_OVERRIDE_ENABLE_PROFILING_MODE 0x00001000 +#define DEBUG_OVERRIDE_ENABLE_SOC_VF_BRINGUP_MODE 0x00002000 +#define DEBUG_OVERRIDE_ENABLE_PER_WGP_RESIENCY 0x00004000 +#define DEBUG_OVERRIDE_DISABLE_MEMORY_VOLTAGE_SCALING 0x00008000 + +// VR Mapping Bit Defines +#define VR_MAPPING_VR_SELECT_MASK 0x01 +#define VR_MAPPING_VR_SELECT_SHIFT 0x00 + +#define VR_MAPPING_PLANE_SELECT_MASK 0x02 +#define VR_MAPPING_PLANE_SELECT_SHIFT 0x01 + +// PSI Bit Defines +#define PSI_SEL_VR0_PLANE0_PSI0 0x01 +#define PSI_SEL_VR0_PLANE0_PSI1 0x02 +#define PSI_SEL_VR0_PLANE1_PSI0 0x04 +#define PSI_SEL_VR0_PLANE1_PSI1 0x08 +#define PSI_SEL_VR1_PLANE0_PSI0 0x10 +#define PSI_SEL_VR1_PLANE0_PSI1 0x20 +#define PSI_SEL_VR1_PLANE1_PSI0 0x40 +#define PSI_SEL_VR1_PLANE1_PSI1 0x80 + +typedef enum { + SVI_PSI_0, // Full phase count (default) + SVI_PSI_1, // Phase count 1st level + SVI_PSI_2, // Phase count 2nd level + SVI_PSI_3, // Single phase operation + active diode emulation + SVI_PSI_4, // Single phase operation + passive diode emulation *optional* + SVI_PSI_5, // Reserved + SVI_PSI_6, // Power down to 0V (voltage regulation disabled) + SVI_PSI_7, // Automated phase shedding and diode emulation +} SVI_PSI_e; + +// Throttler Control/Status Bits +#define THROTTLER_TEMP_EDGE_BIT 0 +#define THROTTLER_TEMP_HOTSPOT_BIT 1 +#define THROTTLER_TEMP_HOTSPOT_GFX_BIT 2 +#define THROTTLER_TEMP_HOTSPOT_SOC_BIT 3 +#define THROTTLER_TEMP_MEM_BIT 4 +#define THROTTLER_TEMP_VR_GFX_BIT 5 +#define THROTTLER_TEMP_VR_SOC_BIT 6 +#define THROTTLER_TEMP_VR_MEM0_BIT 7 +#define THROTTLER_TEMP_VR_MEM1_BIT 8 +#define THROTTLER_TEMP_LIQUID0_BIT 9 +#define THROTTLER_TEMP_LIQUID1_BIT 10 +#define THROTTLER_TEMP_PLX_BIT 11 +#define THROTTLER_TDC_GFX_BIT 12 +#define THROTTLER_TDC_SOC_BIT 13 +#define THROTTLER_PPT0_BIT 14 +#define THROTTLER_PPT1_BIT 15 +#define THROTTLER_PPT2_BIT 16 +#define THROTTLER_PPT3_BIT 17 +#define THROTTLER_FIT_BIT 18 +#define THROTTLER_GFX_APCC_PLUS_BIT 19 +#define THROTTLER_GFX_DVO_BIT 20 +#define THROTTLER_COUNT 21 + +// FW DState Features Control Bits +#define FW_DSTATE_SOC_ULV_BIT 0 +#define FW_DSTATE_G6_HSR_BIT 1 +#define FW_DSTATE_G6_PHY_VMEMP_OFF_BIT 2 +#define FW_DSTATE_SMN_DS_BIT 3 +#define FW_DSTATE_MP1_WHISPER_MODE_BIT 4 +#define FW_DSTATE_SOC_LIV_MIN_BIT 5 +#define FW_DSTATE_SOC_PLL_PWRDN_BIT 6 +#define FW_DSTATE_MEM_PLL_PWRDN_BIT 7 +#define FW_DSTATE_MALL_ALLOC_BIT 8 +#define FW_DSTATE_MEM_PSI_BIT 9 +#define FW_DSTATE_HSR_NON_STROBE_BIT 10 +#define FW_DSTATE_MP0_ENTER_WFI_BIT 11 +#define FW_DSTATE_MALL_FLUSH_BIT 12 +#define FW_DSTATE_SOC_PSI_BIT 13 +#define FW_DSTATE_MMHUB_INTERLOCK_BIT 14 +#define FW_DSTATE_D0i3_2_QUIET_FW_BIT 15 +#define FW_DSTATE_CLDO_PRG_BIT 16 +#define FW_DSTATE_DF_PLL_PWRDN_BIT 17 + +//LED Display Mask & Control Bits +#define LED_DISPLAY_GFX_DPM_BIT 0 +#define LED_DISPLAY_PCIE_BIT 1 +#define LED_DISPLAY_ERROR_BIT 2 + + +#define MEM_TEMP_READ_OUT_OF_BAND_BIT 0 +#define MEM_TEMP_READ_IN_BAND_REFRESH_BIT 1 +#define MEM_TEMP_READ_IN_BAND_DUMMY_PSTATE_BIT 2 + +typedef enum { + SMARTSHIFT_VERSION_1, + SMARTSHIFT_VERSION_2, + SMARTSHIFT_VERSION_3, +} SMARTSHIFT_VERSION_e; + +typedef enum { + FOPT_CALC_AC_CALC_DC, + FOPT_PPTABLE_AC_CALC_DC, + FOPT_CALC_AC_PPTABLE_DC, + FOPT_PPTABLE_AC_PPTABLE_DC, +} FOPT_CALC_e; + +typedef enum { + DRAM_BIT_WIDTH_DISABLED = 0, + DRAM_BIT_WIDTH_X_8 = 8, + DRAM_BIT_WIDTH_X_16 = 16, + DRAM_BIT_WIDTH_X_32 = 32, + DRAM_BIT_WIDTH_X_64 = 64, + DRAM_BIT_WIDTH_X_128 = 128, + DRAM_BIT_WIDTH_COUNT, +} DRAM_BIT_WIDTH_TYPE_e; + +//I2C Interface +#define NUM_I2C_CONTROLLERS 8 + +#define I2C_CONTROLLER_ENABLED 1 +#define I2C_CONTROLLER_DISABLED 0 + +#define MAX_SW_I2C_COMMANDS 24 + +typedef enum { + I2C_CONTROLLER_PORT_0 = 0, //CKSVII2C0 + I2C_CONTROLLER_PORT_1 = 1, //CKSVII2C1 + I2C_CONTROLLER_PORT_COUNT, +} I2cControllerPort_e; + +typedef enum { + I2C_CONTROLLER_NAME_VR_GFX = 0, + I2C_CONTROLLER_NAME_VR_SOC, + I2C_CONTROLLER_NAME_VR_VMEMP, + I2C_CONTROLLER_NAME_VR_VDDIO, + I2C_CONTROLLER_NAME_LIQUID0, + I2C_CONTROLLER_NAME_LIQUID1, + I2C_CONTROLLER_NAME_PLX, + I2C_CONTROLLER_NAME_FAN_INTAKE, + I2C_CONTROLLER_NAME_COUNT, +} I2cControllerName_e; + +typedef enum { + I2C_CONTROLLER_THROTTLER_TYPE_NONE = 0, + I2C_CONTROLLER_THROTTLER_VR_GFX, + I2C_CONTROLLER_THROTTLER_VR_SOC, + I2C_CONTROLLER_THROTTLER_VR_VMEMP, + I2C_CONTROLLER_THROTTLER_VR_VDDIO, + I2C_CONTROLLER_THROTTLER_LIQUID0, + I2C_CONTROLLER_THROTTLER_LIQUID1, + I2C_CONTROLLER_THROTTLER_PLX, + I2C_CONTROLLER_THROTTLER_FAN_INTAKE, + I2C_CONTROLLER_THROTTLER_INA3221, + I2C_CONTROLLER_THROTTLER_COUNT, +} I2cControllerThrottler_e; + +typedef enum { + I2C_CONTROLLER_PROTOCOL_VR_XPDE132G5, + I2C_CONTROLLER_PROTOCOL_VR_IR35217, + I2C_CONTROLLER_PROTOCOL_TMP_MAX31875, + I2C_CONTROLLER_PROTOCOL_INA3221, + I2C_CONTROLLER_PROTOCOL_TMP_MAX6604, + I2C_CONTROLLER_PROTOCOL_COUNT, +} I2cControllerProtocol_e; + +typedef struct { + uint8_t Enabled; + uint8_t Speed; + uint8_t SlaveAddress; + uint8_t ControllerPort; + uint8_t ControllerName; + uint8_t ThermalThrotter; + uint8_t I2cProtocol; + uint8_t PaddingConfig; +} I2cControllerConfig_t; + +typedef enum { + I2C_PORT_SVD_SCL = 0, + I2C_PORT_GPIO, +} I2cPort_e; + +typedef enum { + I2C_SPEED_FAST_50K = 0, //50 Kbits/s + I2C_SPEED_FAST_100K, //100 Kbits/s + I2C_SPEED_FAST_400K, //400 Kbits/s + I2C_SPEED_FAST_PLUS_1M, //1 Mbits/s (in fast mode) + I2C_SPEED_HIGH_1M, //1 Mbits/s (in high speed mode) + I2C_SPEED_HIGH_2M, //2.3 Mbits/s + I2C_SPEED_COUNT, +} I2cSpeed_e; + +typedef enum { + I2C_CMD_READ = 0, + I2C_CMD_WRITE, + I2C_CMD_COUNT, +} I2cCmdType_e; + +#define CMDCONFIG_STOP_BIT 0 +#define CMDCONFIG_RESTART_BIT 1 +#define CMDCONFIG_READWRITE_BIT 2 //bit should be 0 for read, 1 for write + +#define CMDCONFIG_STOP_MASK (1 << CMDCONFIG_STOP_BIT) +#define CMDCONFIG_RESTART_MASK (1 << CMDCONFIG_RESTART_BIT) +#define CMDCONFIG_READWRITE_MASK (1 << CMDCONFIG_READWRITE_BIT) + +typedef struct { + uint8_t ReadWriteData; //Return data for read. Data to send for write + uint8_t CmdConfig; //Includes whether associated command should have a stop or restart command, and is a read or write +} SwI2cCmd_t; //SW I2C Command Table + +typedef struct { + uint8_t I2CcontrollerPort; //CKSVII2C0(0) or //CKSVII2C1(1) + uint8_t I2CSpeed; //Use I2cSpeed_e to indicate speed to select + uint8_t SlaveAddress; //Slave address of device + uint8_t NumCmds; //Number of commands + + SwI2cCmd_t SwI2cCmds[MAX_SW_I2C_COMMANDS]; +} SwI2cRequest_t; // SW I2C Request Table + +typedef struct { + SwI2cRequest_t SwI2cRequest; + + uint32_t Spare[8]; + uint32_t MmHubPadding[8]; // SMU internal use +} SwI2cRequestExternal_t; + +typedef struct { + uint64_t mca_umc_status; + uint64_t mca_umc_addr; + + uint16_t ce_count_lo_chip; + uint16_t ce_count_hi_chip; + + uint32_t eccPadding; +} EccInfo_t; + +typedef struct { + EccInfo_t EccInfo[24]; +} EccInfoTable_t; + +//D3HOT sequences +typedef enum { + BACO_SEQUENCE, + MSR_SEQUENCE, + BAMACO_SEQUENCE, + ULPS_SEQUENCE, + D3HOT_SEQUENCE_COUNT, +} D3HOTSequence_e; + +//This is aligned with RSMU PGFSM Register Mapping +typedef enum { + PG_DYNAMIC_MODE = 0, + PG_STATIC_MODE, +} PowerGatingMode_e; + +//This is aligned with RSMU PGFSM Register Mapping +typedef enum { + PG_POWER_DOWN = 0, + PG_POWER_UP, +} PowerGatingSettings_e; + +typedef struct { + uint32_t a; // store in IEEE float format in this variable + uint32_t b; // store in IEEE float format in this variable + uint32_t c; // store in IEEE float format in this variable +} QuadraticInt_t; + +typedef struct { + uint32_t m; // store in IEEE float format in this variable + uint32_t b; // store in IEEE float format in this variable +} LinearInt_t; + +typedef struct { + uint32_t a; // store in IEEE float format in this variable + uint32_t b; // store in IEEE float format in this variable + uint32_t c; // store in IEEE float format in this variable +} DroopInt_t; + +typedef enum { + DCS_ARCH_DISABLED, + DCS_ARCH_FADCS, + DCS_ARCH_ASYNC, +} DCS_ARCH_e; + +//Only Clks that have DPM descriptors are listed here +typedef enum { + PPCLK_GFXCLK = 0, + PPCLK_SOCCLK, + PPCLK_UCLK, + PPCLK_FCLK, + PPCLK_DCLK_0, + PPCLK_VCLK_0, + PPCLK_DISPCLK, + PPCLK_DPPCLK, + PPCLK_DPREFCLK, + PPCLK_DCFCLK, + PPCLK_DTBCLK, + PPCLK_COUNT, +} PPCLK_e; + +typedef enum { + VOLTAGE_MODE_PPTABLE = 0, + VOLTAGE_MODE_FUSES, + VOLTAGE_MODE_COUNT, +} VOLTAGE_MODE_e; + +typedef enum { + AVFS_VOLTAGE_GFX = 0, + AVFS_VOLTAGE_SOC, + AVFS_VOLTAGE_COUNT, +} AVFS_VOLTAGE_TYPE_e; + +typedef enum { + AVFS_TEMP_COLD = 0, + AVFS_TEMP_HOT, + AVFS_TEMP_COUNT, +} AVFS_TEMP_e; + +typedef enum { + AVFS_D_G, + AVFS_D_COUNT, +} AVFS_D_e; + + +typedef enum { + UCLK_DIV_BY_1 = 0, + UCLK_DIV_BY_2, + UCLK_DIV_BY_4, + UCLK_DIV_BY_8, +} UCLK_DIV_e; + +typedef enum { + GPIO_INT_POLARITY_ACTIVE_LOW = 0, + GPIO_INT_POLARITY_ACTIVE_HIGH, +} GpioIntPolarity_e; + +typedef enum { + PWR_CONFIG_TDP = 0, + PWR_CONFIG_TGP, + PWR_CONFIG_TCP_ESTIMATED, + PWR_CONFIG_TCP_MEASURED, + PWR_CONFIG_TBP_DESKTOP, + PWR_CONFIG_TBP_MOBILE, +} PwrConfig_e; + +typedef struct { + uint8_t Padding; + uint8_t SnapToDiscrete; // 0 - Fine grained DPM, 1 - Discrete DPM + uint8_t NumDiscreteLevels; // Set to 2 (Fmin, Fmax) when using fine grained DPM, otherwise set to # discrete levels used + uint8_t CalculateFopt; // Indication whether FW should calculate Fopt or use values below. Reference FOPT_CALC_e + LinearInt_t ConversionToAvfsClk; // Transfer function to AVFS Clock (GHz->GHz) + uint32_t Padding3[3]; + uint16_t Padding4; + uint16_t FoptimalDc; //Foptimal frequency in DC power mode. + uint16_t FoptimalAc; //Foptimal frequency in AC power mode. + uint16_t Padding2; +} DpmDescriptor_t; + +typedef enum { + PPT_THROTTLER_PPT0, + PPT_THROTTLER_PPT1, + PPT_THROTTLER_PPT2, + PPT_THROTTLER_PPT3, + PPT_THROTTLER_COUNT +} PPT_THROTTLER_e; + +typedef enum { + TEMP_EDGE, + TEMP_HOTSPOT, + TEMP_HOTSPOT_GFX, + TEMP_HOTSPOT_SOC, + TEMP_MEM, + TEMP_VR_GFX, + TEMP_VR_SOC, + TEMP_VR_MEM0, + TEMP_VR_MEM1, + TEMP_LIQUID0, + TEMP_LIQUID1, + TEMP_PLX, + TEMP_COUNT, +} TEMP_e; + +typedef enum { + TDC_THROTTLER_GFX, + TDC_THROTTLER_SOC, + TDC_THROTTLER_COUNT +} TDC_THROTTLER_e; + +typedef enum { + SVI_PLANE_VDD_GFX, + SVI_PLANE_VDD_SOC, + SVI_PLANE_VDDCI_MEM, + SVI_PLANE_VDDIO_MEM, + SVI_PLANE_COUNT, +} SVI_PLANE_e; + +typedef enum { + PMFW_VOLT_PLANE_GFX, + PMFW_VOLT_PLANE_SOC, + PMFW_VOLT_PLANE_COUNT +} PMFW_VOLT_PLANE_e; + +typedef enum { + CUSTOMER_VARIANT_ROW, + CUSTOMER_VARIANT_FALCON, + CUSTOMER_VARIANT_COUNT, +} CUSTOMER_VARIANT_e; + +typedef enum { + POWER_SOURCE_AC, + POWER_SOURCE_DC, + POWER_SOURCE_COUNT, +} POWER_SOURCE_e; + +typedef enum { + MEM_VENDOR_PLACEHOLDER0, // 0 + MEM_VENDOR_SAMSUNG, // 1 + MEM_VENDOR_INFINEON, // 2 + MEM_VENDOR_ELPIDA, // 3 + MEM_VENDOR_ETRON, // 4 + MEM_VENDOR_NANYA, // 5 + MEM_VENDOR_HYNIX, // 6 + MEM_VENDOR_MOSEL, // 7 + MEM_VENDOR_WINBOND, // 8 + MEM_VENDOR_ESMT, // 9 + MEM_VENDOR_PLACEHOLDER1, // 10 + MEM_VENDOR_PLACEHOLDER2, // 11 + MEM_VENDOR_PLACEHOLDER3, // 12 + MEM_VENDOR_PLACEHOLDER4, // 13 + MEM_VENDOR_PLACEHOLDER5, // 14 + MEM_VENDOR_MICRON, // 15 + MEM_VENDOR_COUNT, +} MEM_VENDOR_e; + +typedef enum { + PP_GRTAVFS_HW_CPO_CTL_ZONE0, + PP_GRTAVFS_HW_CPO_CTL_ZONE1, + PP_GRTAVFS_HW_CPO_CTL_ZONE2, + PP_GRTAVFS_HW_CPO_CTL_ZONE3, + PP_GRTAVFS_HW_CPO_CTL_ZONE4, + PP_GRTAVFS_HW_CPO_EN_0_31_ZONE0, + PP_GRTAVFS_HW_CPO_EN_32_63_ZONE0, + PP_GRTAVFS_HW_CPO_EN_0_31_ZONE1, + PP_GRTAVFS_HW_CPO_EN_32_63_ZONE1, + PP_GRTAVFS_HW_CPO_EN_0_31_ZONE2, + PP_GRTAVFS_HW_CPO_EN_32_63_ZONE2, + PP_GRTAVFS_HW_CPO_EN_0_31_ZONE3, + PP_GRTAVFS_HW_CPO_EN_32_63_ZONE3, + PP_GRTAVFS_HW_CPO_EN_0_31_ZONE4, + PP_GRTAVFS_HW_CPO_EN_32_63_ZONE4, + PP_GRTAVFS_HW_ZONE0_VF, + PP_GRTAVFS_HW_ZONE1_VF1, + PP_GRTAVFS_HW_ZONE2_VF2, + PP_GRTAVFS_HW_ZONE3_VF3, + PP_GRTAVFS_HW_VOLTAGE_GB, + PP_GRTAVFS_HW_CPOSCALINGCTRL_ZONE0, + PP_GRTAVFS_HW_CPOSCALINGCTRL_ZONE1, + PP_GRTAVFS_HW_CPOSCALINGCTRL_ZONE2, + PP_GRTAVFS_HW_CPOSCALINGCTRL_ZONE3, + PP_GRTAVFS_HW_CPOSCALINGCTRL_ZONE4, + PP_GRTAVFS_HW_RESERVED_0, + PP_GRTAVFS_HW_RESERVED_1, + PP_GRTAVFS_HW_RESERVED_2, + PP_GRTAVFS_HW_RESERVED_3, + PP_GRTAVFS_HW_RESERVED_4, + PP_GRTAVFS_HW_RESERVED_5, + PP_GRTAVFS_HW_RESERVED_6, + PP_GRTAVFS_HW_FUSE_COUNT, +} PP_GRTAVFS_HW_FUSE_e; + +typedef enum { + PP_GRTAVFS_FW_COMMON_PPVMIN_Z1_HOT_T0, + PP_GRTAVFS_FW_COMMON_PPVMIN_Z1_COLD_T0, + PP_GRTAVFS_FW_COMMON_PPVMIN_Z2_HOT_T0, + PP_GRTAVFS_FW_COMMON_PPVMIN_Z2_COLD_T0, + PP_GRTAVFS_FW_COMMON_PPVMIN_Z3_HOT_T0, + PP_GRTAVFS_FW_COMMON_PPVMIN_Z3_COLD_T0, + PP_GRTAVFS_FW_COMMON_PPVMIN_Z4_HOT_T0, + PP_GRTAVFS_FW_COMMON_PPVMIN_Z4_COLD_T0, + PP_GRTAVFS_FW_COMMON_SRAM_RM_Z0, + PP_GRTAVFS_FW_COMMON_SRAM_RM_Z1, + PP_GRTAVFS_FW_COMMON_SRAM_RM_Z2, + PP_GRTAVFS_FW_COMMON_SRAM_RM_Z3, + PP_GRTAVFS_FW_COMMON_SRAM_RM_Z4, + PP_GRTAVFS_FW_COMMON_FUSE_COUNT, +} PP_GRTAVFS_FW_COMMON_FUSE_e; + +typedef enum { + PP_GRTAVFS_FW_SEP_FUSE_GB1_PWL_VOLTAGE_NEG_1, + PP_GRTAVFS_FW_SEP_FUSE_GB1_PWL_VOLTAGE_0, + PP_GRTAVFS_FW_SEP_FUSE_GB1_PWL_VOLTAGE_1, + PP_GRTAVFS_FW_SEP_FUSE_GB1_PWL_VOLTAGE_2, + PP_GRTAVFS_FW_SEP_FUSE_GB1_PWL_VOLTAGE_3, + PP_GRTAVFS_FW_SEP_FUSE_GB1_PWL_VOLTAGE_4, + PP_GRTAVFS_FW_SEP_FUSE_GB2_PWL_VOLTAGE_NEG_1, + PP_GRTAVFS_FW_SEP_FUSE_GB2_PWL_VOLTAGE_0, + PP_GRTAVFS_FW_SEP_FUSE_GB2_PWL_VOLTAGE_1, + PP_GRTAVFS_FW_SEP_FUSE_GB2_PWL_VOLTAGE_2, + PP_GRTAVFS_FW_SEP_FUSE_GB2_PWL_VOLTAGE_3, + PP_GRTAVFS_FW_SEP_FUSE_GB2_PWL_VOLTAGE_4, + PP_GRTAVFS_FW_SEP_FUSE_VF_NEG_1_FREQUENCY, + PP_GRTAVFS_FW_SEP_FUSE_VF4_FREQUENCY, + PP_GRTAVFS_FW_SEP_FUSE_FREQUENCY_TO_COUNT_SCALER_0, + PP_GRTAVFS_FW_SEP_FUSE_FREQUENCY_TO_COUNT_SCALER_1, + PP_GRTAVFS_FW_SEP_FUSE_FREQUENCY_TO_COUNT_SCALER_2, + PP_GRTAVFS_FW_SEP_FUSE_FREQUENCY_TO_COUNT_SCALER_3, + PP_GRTAVFS_FW_SEP_FUSE_FREQUENCY_TO_COUNT_SCALER_4, + PP_GRTAVFS_FW_SEP_FUSE_COUNT, +} PP_GRTAVFS_FW_SEP_FUSE_e; + +#define PP_NUM_RTAVFS_PWL_ZONES 5 + + +// VBIOS or PPLIB configures telemetry slope and offset. Only slope expected to be set for SVI3 +// Slope Q1.7, Offset Q1.2 +typedef struct { + int8_t Offset; // in Amps + uint8_t Padding; + uint16_t MaxCurrent; // in Amps +} SviTelemetryScale_t; + +#define PP_NUM_OD_VF_CURVE_POINTS PP_NUM_RTAVFS_PWL_ZONES + 1 + +#define PP_OD_FEATURE_GFX_VF_CURVE_BIT 0 +#define PP_OD_FEATURE_GFX_VMAX_BIT 1 +#define PP_OD_FEATURE_SOC_VMAX_BIT 2 +#define PP_OD_FEATURE_PPT_BIT 3 +#define PP_OD_FEATURE_FAN_CURVE_BIT 4 +#define PP_OD_FEATURE_FAN_LEGACY_BIT 5 +#define PP_OD_FEATURE_FULL_CTRL_BIT 6 +#define PP_OD_FEATURE_TDC_BIT 7 +#define PP_OD_FEATURE_GFXCLK_BIT 8 +#define PP_OD_FEATURE_UCLK_BIT 9 +#define PP_OD_FEATURE_FCLK_BIT 10 +#define PP_OD_FEATURE_ZERO_FAN_BIT 11 +#define PP_OD_FEATURE_TEMPERATURE_BIT 12 +#define PP_OD_FEATURE_EDC_BIT 13 +#define PP_OD_FEATURE_COUNT 14 + +typedef enum { + PP_OD_POWER_FEATURE_ALWAYS_ENABLED, + PP_OD_POWER_FEATURE_DISABLED_WHILE_GAMING, + PP_OD_POWER_FEATURE_ALWAYS_DISABLED, +} PP_OD_POWER_FEATURE_e; + +typedef enum { + FAN_MODE_AUTO = 0, + FAN_MODE_MANUAL_LINEAR, +} FanMode_e; + +typedef enum { + OD_NO_ERROR, + OD_REQUEST_ADVANCED_NOT_SUPPORTED, + OD_UNSUPPORTED_FEATURE, + OD_INVALID_FEATURE_COMBO_ERROR, + OD_GFXCLK_VF_CURVE_OFFSET_ERROR, + OD_VDD_GFX_VMAX_ERROR, + OD_VDD_SOC_VMAX_ERROR, + OD_PPT_ERROR, + OD_FAN_MIN_PWM_ERROR, + OD_FAN_ACOUSTIC_TARGET_ERROR, + OD_FAN_ACOUSTIC_LIMIT_ERROR, + OD_FAN_TARGET_TEMP_ERROR, + OD_FAN_ZERO_RPM_STOP_TEMP_ERROR, + OD_FAN_CURVE_PWM_ERROR, + OD_FAN_CURVE_TEMP_ERROR, + OD_FULL_CTRL_GFXCLK_ERROR, + OD_FULL_CTRL_UCLK_ERROR, + OD_FULL_CTRL_FCLK_ERROR, + OD_FULL_CTRL_VDD_GFX_ERROR, + OD_FULL_CTRL_VDD_SOC_ERROR, + OD_TDC_ERROR, + OD_GFXCLK_ERROR, + OD_UCLK_ERROR, + OD_FCLK_ERROR, + OD_OP_TEMP_ERROR, + OD_OP_GFX_EDC_ERROR, + OD_OP_GFX_PCC_ERROR, + OD_POWER_FEATURE_CTRL_ERROR, +} OD_FAIL_e; + +typedef struct { + uint32_t FeatureCtrlMask; + + //Voltage control + int16_t VoltageOffsetPerZoneBoundary[PP_NUM_OD_VF_CURVE_POINTS]; + + uint16_t VddGfxVmax; // in mV + uint16_t VddSocVmax; + + uint8_t IdlePwrSavingFeaturesCtrl; + uint8_t RuntimePwrSavingFeaturesCtrl; + uint16_t Padding; + + //Frequency changes + int16_t GfxclkFmin; // MHz + int16_t GfxclkFmax; // MHz + uint16_t UclkFmin; // MHz + uint16_t UclkFmax; // MHz + uint16_t FclkFmin; + uint16_t FclkFmax; + + //PPT + int16_t Ppt; // % + int16_t Tdc; + + //Fan control + uint8_t FanLinearPwmPoints[NUM_OD_FAN_MAX_POINTS]; + uint8_t FanLinearTempPoints[NUM_OD_FAN_MAX_POINTS]; + uint16_t FanMinimumPwm; + uint16_t AcousticTargetRpmThreshold; + uint16_t AcousticLimitRpmThreshold; + uint16_t FanTargetTemperature; // Degree Celcius + uint8_t FanZeroRpmEnable; + uint8_t FanZeroRpmStopTemp; + uint8_t FanMode; + uint8_t MaxOpTemp; + + uint8_t AdvancedOdModeEnabled; + uint8_t Padding1[3]; + + uint16_t GfxVoltageFullCtrlMode; + uint16_t SocVoltageFullCtrlMode; + uint16_t GfxclkFullCtrlMode; + uint16_t UclkFullCtrlMode; + uint16_t FclkFullCtrlMode; + uint16_t Padding2; + + int16_t GfxEdc; + int16_t GfxPccLimitControl; + + uint32_t Spare[10]; + uint32_t MmHubPadding[8]; // SMU internal use. Adding here instead of external as a workaround +} OverDriveTable_t; + +typedef struct { + OverDriveTable_t OverDriveTable; + +} OverDriveTableExternal_t; + +typedef struct { + uint32_t FeatureCtrlMask; + + //Gfx Vf Curve + int16_t VoltageOffsetPerZoneBoundary[PP_NUM_OD_VF_CURVE_POINTS]; + //gfx Vmax + uint16_t VddGfxVmax; // in mV + //soc Vmax + uint16_t VddSocVmax; + + //gfxclk + int16_t GfxclkFmin; // MHz + int16_t GfxclkFmax; // MHz + //uclk + uint16_t UclkFmin; // MHz + uint16_t UclkFmax; // MHz + //fclk + uint16_t FclkFmin; + uint16_t FclkFmax; + + //PPT + int16_t Ppt; // % + //TDC + int16_t Tdc; + + //Fan Curve + uint8_t FanLinearPwmPoints[NUM_OD_FAN_MAX_POINTS]; + uint8_t FanLinearTempPoints[NUM_OD_FAN_MAX_POINTS]; + //Fan Legacy + uint16_t FanMinimumPwm; + uint16_t AcousticTargetRpmThreshold; + uint16_t AcousticLimitRpmThreshold; + uint16_t FanTargetTemperature; // Degree Celcius + //zero fan + uint8_t FanZeroRpmEnable; + //temperature + uint8_t MaxOpTemp; + uint8_t Padding[2]; + + //Full Ctrl + uint16_t GfxVoltageFullCtrlMode; + uint16_t SocVoltageFullCtrlMode; + uint16_t GfxclkFullCtrlMode; + uint16_t UclkFullCtrlMode; + uint16_t FclkFullCtrlMode; + //EDC + int16_t GfxEdc; + int16_t GfxPccLimitControl; + int16_t Padding1; + + uint32_t Spare[5]; +} OverDriveLimits_t; + +typedef enum { + BOARD_GPIO_SMUIO_0, + BOARD_GPIO_SMUIO_1, + BOARD_GPIO_SMUIO_2, + BOARD_GPIO_SMUIO_3, + BOARD_GPIO_SMUIO_4, + BOARD_GPIO_SMUIO_5, + BOARD_GPIO_SMUIO_6, + BOARD_GPIO_SMUIO_7, + BOARD_GPIO_SMUIO_8, + BOARD_GPIO_SMUIO_9, + BOARD_GPIO_SMUIO_10, + BOARD_GPIO_SMUIO_11, + BOARD_GPIO_SMUIO_12, + BOARD_GPIO_SMUIO_13, + BOARD_GPIO_SMUIO_14, + BOARD_GPIO_SMUIO_15, + BOARD_GPIO_SMUIO_16, + BOARD_GPIO_SMUIO_17, + BOARD_GPIO_SMUIO_18, + BOARD_GPIO_SMUIO_19, + BOARD_GPIO_SMUIO_20, + BOARD_GPIO_SMUIO_21, + BOARD_GPIO_SMUIO_22, + BOARD_GPIO_SMUIO_23, + BOARD_GPIO_SMUIO_24, + BOARD_GPIO_SMUIO_25, + BOARD_GPIO_SMUIO_26, + BOARD_GPIO_SMUIO_27, + BOARD_GPIO_SMUIO_28, + BOARD_GPIO_SMUIO_29, + BOARD_GPIO_SMUIO_30, + BOARD_GPIO_SMUIO_31, + MAX_BOARD_GPIO_SMUIO_NUM, + BOARD_GPIO_DC_GEN_A, + BOARD_GPIO_DC_GEN_B, + BOARD_GPIO_DC_GEN_C, + BOARD_GPIO_DC_GEN_D, + BOARD_GPIO_DC_GEN_E, + BOARD_GPIO_DC_GEN_F, + BOARD_GPIO_DC_GEN_G, + BOARD_GPIO_DC_GENLK_CLK, + BOARD_GPIO_DC_GENLK_VSYNC, + BOARD_GPIO_DC_SWAPLOCK_A, + BOARD_GPIO_DC_SWAPLOCK_B, + MAX_BOARD_DC_GPIO_NUM, + BOARD_GPIO_LV_EN, +} BOARD_GPIO_TYPE_e; + +#define INVALID_BOARD_GPIO 0xFF + + +typedef struct { + //PLL 0 + uint16_t InitImuClk; + uint16_t InitSocclk; + uint16_t InitMpioclk; + uint16_t InitSmnclk; + //PLL 1 + uint16_t InitDispClk; + uint16_t InitDppClk; + uint16_t InitDprefclk; + uint16_t InitDcfclk; + uint16_t InitDtbclk; + uint16_t InitDbguSocClk; + //PLL 2 + uint16_t InitGfxclk_bypass; + uint16_t InitMp1clk; + uint16_t InitLclk; + uint16_t InitDbguBacoClk; + uint16_t InitBaco400clk; + uint16_t InitBaco1200clk_bypass; + uint16_t InitBaco700clk_bypass; + uint16_t InitBaco500clk; + // PLL 3 + uint16_t InitDclk0; + uint16_t InitVclk0; + // PLL 4 + uint16_t InitFclk; + uint16_t Padding1; + // PLL 5 + //UCLK clocks, assumed all UCLK instances will be the same. + uint8_t InitUclkLevel; // =0,1,2,3,4,5 frequency from FreqTableUclk + + uint8_t Padding[3]; + + uint32_t InitVcoFreqPll0; //smu_socclk_t + uint32_t InitVcoFreqPll1; //smu_displayclk_t + uint32_t InitVcoFreqPll2; //smu_nbioclk_t + uint32_t InitVcoFreqPll3; //smu_vcnclk_t + uint32_t InitVcoFreqPll4; //smu_fclk_t + uint32_t InitVcoFreqPll5; //smu_uclk_01_t + uint32_t InitVcoFreqPll6; //smu_uclk_23_t + uint32_t InitVcoFreqPll7; //smu_uclk_45_t + uint32_t InitVcoFreqPll8; //smu_uclk_67_t + + //encoding will be SVI3 + uint16_t InitGfx; // In mV(Q2) , should be 0? + uint16_t InitSoc; // In mV(Q2) + uint16_t InitVddIoMem; // In mV(Q2) MemVdd + uint16_t InitVddCiMem; // In mV(Q2) VMemP + + //uint16_t Padding2; + + uint32_t Spare[8]; +} BootValues_t; + +typedef struct { + uint16_t Power[PPT_THROTTLER_COUNT][POWER_SOURCE_COUNT]; // Watts + uint16_t Tdc[TDC_THROTTLER_COUNT]; // Amps + + uint16_t Temperature[TEMP_COUNT]; // Celsius + + uint8_t PwmLimitMin; + uint8_t PwmLimitMax; + uint8_t FanTargetTemperature; + uint8_t Spare1[1]; + + uint16_t AcousticTargetRpmThresholdMin; + uint16_t AcousticTargetRpmThresholdMax; + + uint16_t AcousticLimitRpmThresholdMin; + uint16_t AcousticLimitRpmThresholdMax; + + uint16_t PccLimitMin; + uint16_t PccLimitMax; + + uint16_t FanStopTempMin; + uint16_t FanStopTempMax; + uint16_t FanStartTempMin; + uint16_t FanStartTempMax; + + uint16_t PowerMinPpt0[POWER_SOURCE_COUNT]; + uint32_t Spare[11]; +} MsgLimits_t; + +typedef struct { + uint16_t BaseClockAc; + uint16_t GameClockAc; + uint16_t BoostClockAc; + uint16_t BaseClockDc; + uint16_t GameClockDc; + uint16_t BoostClockDc; + + uint32_t Reserved[4]; +} DriverReportedClocks_t; + +typedef struct { + uint8_t DcBtcEnabled; + uint8_t Padding[3]; + + uint16_t DcTol; // mV Q2 + uint16_t DcBtcGb; // mV Q2 + + uint16_t DcBtcMin; // mV Q2 + uint16_t DcBtcMax; // mV Q2 + + LinearInt_t DcBtcGbScalar; +} AvfsDcBtcParams_t; + +typedef struct { + uint16_t AvfsTemp[AVFS_TEMP_COUNT]; //in degrees C + uint16_t VftFMin; // in MHz + uint16_t VInversion; // in mV Q2 + QuadraticInt_t qVft[AVFS_TEMP_COUNT]; + QuadraticInt_t qAvfsGb; + QuadraticInt_t qAvfsGb2; +} AvfsFuseOverride_t; + +//all settings maintained by PFE team +typedef struct { + uint8_t Version; + uint8_t Spare8[3]; + // SECTION: Feature Control + uint32_t FeaturesToRun[NUM_FEATURES / 32]; // Features that PMFW will attempt to enable. Use FEATURE_*_BIT as mapping + // SECTION: FW DSTATE Settings + uint32_t FwDStateMask; // See FW_DSTATE_*_BIT for mapping + // SECTION: Advanced Options + uint32_t DebugOverrides; + + uint32_t Spare[2]; +} PFE_Settings_t; + +typedef struct { + // SECTION: Version + uint32_t Version; // should be unique to each SKU(i.e if any value changes in below structure then this value must be different) + + // SECTION: Miscellaneous Configuration + uint8_t TotalPowerConfig; // Determines how PMFW calculates the power. Use defines from PwrConfig_e + uint8_t CustomerVariant; //To specify if this PPTable is intended for a particular customer. Use defines from CUSTOMER_VARIANT_e + uint8_t MemoryTemperatureTypeMask; // Bit mapping indicating which methods of memory temperature reading are enabled. Use defines from MEM_TEMP_*BIT + uint8_t SmartShiftVersion; // Determine what SmartShift feature version is supported Use defines from SMARTSHIFT_VERSION_e + + // SECTION: Infrastructure Limits + uint8_t SocketPowerLimitSpare[10]; + + //if set to 1, SocketPowerLimitAc and SocketPowerLimitDc will be interpreted as legacy programs(i.e absolute power). If 0, all except index 0 will be scalars + //relative index 0 + uint8_t EnableLegacyPptLimit; + uint8_t UseInputTelemetry; //applicable to SVI3 only and only to be set if VRs support + + uint8_t SmartShiftMinReportedPptinDcs; //minimum possible active power consumption for this SKU. Used for SmartShift power reporting + + uint8_t PaddingPpt[7]; + + uint16_t HwCtfTempLimit; // In degrees Celsius. Temperature above which HW will trigger CTF. Consumed by VBIOS only + + uint16_t PaddingInfra; + + // Per year normalized Vmax state failure rates (sum of the two domains divided by life time in years) + uint32_t FitControllerFailureRateLimit; //in IEEE float + //Expected GFX Duty Cycle at Vmax. + uint32_t FitControllerGfxDutyCycle; // in IEEE float + //Expected SOC Duty Cycle at Vmax. + uint32_t FitControllerSocDutyCycle; // in IEEE float + + //This offset will be deducted from the controller output to before it goes through the SOC Vset limiter block. + uint32_t FitControllerSocOffset; //in IEEE float + + uint32_t GfxApccPlusResidencyLimit; // Percentage value. Used by APCC+ controller to control PCC residency to some value + + // SECTION: Throttler settings + uint32_t ThrottlerControlMask; // See THROTTLER_*_BIT for mapping + + + // SECTION: Voltage Control Parameters + uint16_t UlvVoltageOffset[PMFW_VOLT_PLANE_COUNT]; // In mV(Q2). ULV offset used in either GFX_ULV or SOC_ULV(part of FW_DSTATE) + + uint8_t Padding[2]; + uint16_t DeepUlvVoltageOffsetSoc; // In mV(Q2) Long Idle Vmin (deep ULV), for VDD_SOC as part of FW_DSTATE + + // Voltage Limits + uint16_t DefaultMaxVoltage[PMFW_VOLT_PLANE_COUNT]; // In mV(Q2) Maximum voltage without FIT controller enabled + uint16_t BoostMaxVoltage[PMFW_VOLT_PLANE_COUNT]; // In mV(Q2) Maximum voltage with FIT controller enabled + + //Vmin Optimizations + int16_t VminTempHystersis[PMFW_VOLT_PLANE_COUNT]; // Celsius Temperature hysteresis for switching between low/high temperature values for Vmin + int16_t VminTempThreshold[PMFW_VOLT_PLANE_COUNT]; // Celsius Temperature threshold for switching between low/high temperature values for Vmin + uint16_t Vmin_Hot_T0[PMFW_VOLT_PLANE_COUNT]; //In mV(Q2) Initial (pre-aging) Vset to be used at hot. + uint16_t Vmin_Cold_T0[PMFW_VOLT_PLANE_COUNT]; //In mV(Q2) Initial (pre-aging) Vset to be used at cold. + uint16_t Vmin_Hot_Eol[PMFW_VOLT_PLANE_COUNT]; //In mV(Q2) End-of-life Vset to be used at hot. + uint16_t Vmin_Cold_Eol[PMFW_VOLT_PLANE_COUNT]; //In mV(Q2) End-of-life Vset to be used at cold. + uint16_t Vmin_Aging_Offset[PMFW_VOLT_PLANE_COUNT]; //In mV(Q2) Worst-case aging margin + uint16_t Spare_Vmin_Plat_Offset_Hot[PMFW_VOLT_PLANE_COUNT]; //In mV(Q2) Platform offset apply to T0 Hot + uint16_t Spare_Vmin_Plat_Offset_Cold[PMFW_VOLT_PLANE_COUNT]; //In mV(Q2) Platform offset apply to T0 Cold + + //This is a fixed/minimum VMIN aging degradation offset which is applied at T0. This reflects the minimum amount of aging already accounted for. + uint16_t VcBtcFixedVminAgingOffset[PMFW_VOLT_PLANE_COUNT]; + //Linear offset or GB term to account for mis-correlation between PSM and Vmin shift trends across parts. + uint16_t VcBtcVmin2PsmDegrationGb[PMFW_VOLT_PLANE_COUNT]; + //Scalar coefficient of the PSM aging degradation function + uint32_t VcBtcPsmA[PMFW_VOLT_PLANE_COUNT]; // A_PSM + //Exponential coefficient of the PSM aging degradation function + uint32_t VcBtcPsmB[PMFW_VOLT_PLANE_COUNT]; // B_PSM + //Scalar coefficient of the VMIN aging degradation function. Specified as worst case between hot and cold. + uint32_t VcBtcVminA[PMFW_VOLT_PLANE_COUNT]; // A_VMIN + //Exponential coefficient of the VMIN aging degradation function. Specified as worst case between hot and cold. + uint32_t VcBtcVminB[PMFW_VOLT_PLANE_COUNT]; // B_VMIN + + uint8_t PerPartVminEnabled[PMFW_VOLT_PLANE_COUNT]; + uint8_t VcBtcEnabled[PMFW_VOLT_PLANE_COUNT]; + + uint16_t SocketPowerLimitAcTau[PPT_THROTTLER_COUNT]; // Time constant of LPF in ms + uint16_t SocketPowerLimitDcTau[PPT_THROTTLER_COUNT]; // Time constant of LPF in ms + + QuadraticInt_t Gfx_Vmin_droop; + QuadraticInt_t Soc_Vmin_droop; + uint32_t SpareVmin[6]; + + //SECTION: DPM Configuration 1 + DpmDescriptor_t DpmDescriptor[PPCLK_COUNT]; + + uint16_t FreqTableGfx [NUM_GFXCLK_DPM_LEVELS ]; // In MHz + uint16_t FreqTableVclk [NUM_VCLK_DPM_LEVELS ]; // In MHz + uint16_t FreqTableDclk [NUM_DCLK_DPM_LEVELS ]; // In MHz + uint16_t FreqTableSocclk [NUM_SOCCLK_DPM_LEVELS ]; // In MHz + uint16_t FreqTableUclk [NUM_UCLK_DPM_LEVELS ]; // In MHz + uint16_t FreqTableShadowUclk [NUM_UCLK_DPM_LEVELS ]; // In MHz + uint16_t FreqTableDispclk [NUM_DISPCLK_DPM_LEVELS ]; // In MHz + uint16_t FreqTableDppClk [NUM_DPPCLK_DPM_LEVELS ]; // In MHz + uint16_t FreqTableDprefclk [NUM_DPREFCLK_DPM_LEVELS]; // In MHz + uint16_t FreqTableDcfclk [NUM_DCFCLK_DPM_LEVELS ]; // In MHz + uint16_t FreqTableDtbclk [NUM_DTBCLK_DPM_LEVELS ]; // In MHz + uint16_t FreqTableFclk [NUM_FCLK_DPM_LEVELS ]; // In MHz + + uint32_t DcModeMaxFreq [PPCLK_COUNT ]; // In MHz + + uint16_t GfxclkAibFmax; + uint16_t GfxclkFreqCap; + + //GFX Idle Power Settings + uint16_t GfxclkFgfxoffEntry; // Entry in RLC stage (PLL), in Mhz + uint16_t GfxclkFgfxoffExitImu; // Exit/Entry in IMU stage (BYPASS), in Mhz + uint16_t GfxclkFgfxoffExitRlc; // Exit in RLC stage (PLL), in Mhz + uint16_t GfxclkThrottleClock; //Used primarily in DCS + uint8_t EnableGfxPowerStagesGpio; //Genlk_vsync GPIO flag used to control gfx power stages + uint8_t GfxIdlePadding; + + uint8_t SmsRepairWRCKClkDivEn; + uint8_t SmsRepairWRCKClkDivVal; + uint8_t GfxOffEntryEarlyMGCGEn; + uint8_t GfxOffEntryForceCGCGEn; + uint8_t GfxOffEntryForceCGCGDelayEn; + uint8_t GfxOffEntryForceCGCGDelayVal; // in microseconds + + uint16_t GfxclkFreqGfxUlv; // in MHz + uint8_t GfxIdlePadding2[2]; + uint32_t GfxOffEntryHysteresis; //For RLC to count after it enters CGCG, and before triggers GFXOFF entry + uint32_t GfxoffSpare[15]; + + // DFLL + uint16_t DfllMstrOscConfigA; //Used for voltage sensitivity slope tuning: 0 = (en_leaker << 9) | (en_vint1_reduce << 8) | (gain_code << 6) | (bias_code << 3) | (vint1_code << 1) | en_bias + uint16_t DfllSlvOscConfigA; //Used for voltage sensitivity slope tuning: 0 = (en_leaker << 9) | (en_vint1_reduce << 8) | (gain_code << 6) | (bias_code << 3) | (vint1_code << 1) | en_bias + uint32_t DfllBtcMasterScalerM; + int32_t DfllBtcMasterScalerB; + uint32_t DfllBtcSlaveScalerM; + int32_t DfllBtcSlaveScalerB; + + uint32_t DfllPccAsWaitCtrl; //GDFLL_AS_WAIT_CTRL_PCC register value to be passed to RLC msg + uint32_t DfllPccAsStepCtrl; //GDFLL_AS_STEP_CTRL_PCC register value to be passed to RLC msg + uint32_t GfxDfllSpare[9]; + + // DVO + uint32_t DvoPsmDownThresholdVoltage; //Voltage float + uint32_t DvoPsmUpThresholdVoltage; //Voltage float + uint32_t DvoFmaxLowScaler; //Unitless float + + // GFX DCS + uint16_t DcsGfxOffVoltage; //Voltage in mV(Q2) applied to VDDGFX when entering DCS GFXOFF phase + uint16_t PaddingDcs; + + uint16_t DcsMinGfxOffTime; //Minimum amount of time PMFW shuts GFX OFF as part of GFX DCS phase + uint16_t DcsMaxGfxOffTime; //Maximum amount of time PMFW can shut GFX OFF as part of GFX DCS phase at a stretch. + + uint32_t DcsMinCreditAccum; //Min amount of positive credit accumulation before waking GFX up as part of DCS. + + uint16_t DcsExitHysteresis; //The min amount of time power credit accumulator should have a value > 0 before SMU exits the DCS throttling phase. + uint16_t DcsTimeout; //This is the amount of time SMU FW waits for RLC to put GFX into GFXOFF before reverting to the fallback mechanism of throttling GFXCLK to Fmin. + + uint32_t DcsPfGfxFopt; //Default to GFX FMIN + uint32_t DcsPfUclkFopt; //Default to UCLK FMIN + + uint8_t FoptEnabled; + uint8_t DcsSpare2[3]; + uint32_t DcsFoptM; //Tuning paramters to shift Fopt calculation, IEEE754 float + uint32_t DcsFoptB; //Tuning paramters to shift Fopt calculation, IEEE754 float + uint32_t DcsSpare[9]; + + // UCLK section + uint8_t UseStrobeModeOptimizations; //Set to indicate that FW should use strobe mode optimizations + uint8_t PaddingMem[3]; + + uint8_t UclkDpmPstates [NUM_UCLK_DPM_LEVELS]; // 6 Primary SW DPM states (6 + 6 Shadow) + uint8_t UclkDpmShadowPstates [NUM_UCLK_DPM_LEVELS]; // 6 Shadow SW DPM states (6 + 6 Shadow) + uint8_t FreqTableUclkDiv [NUM_UCLK_DPM_LEVELS]; // 0:Div-1, 1:Div-1/2, 2:Div-1/4, 3:Div-1/8 + uint8_t FreqTableShadowUclkDiv [NUM_UCLK_DPM_LEVELS]; // 0:Div-1, 1:Div-1/2, 2:Div-1/4, 3:Div-1/8 + uint16_t MemVmempVoltage [NUM_UCLK_DPM_LEVELS]; // mV(Q2) + uint16_t MemVddioVoltage [NUM_UCLK_DPM_LEVELS]; // mV(Q2) + uint16_t DalDcModeMaxUclkFreq; + uint8_t PaddingsMem[2]; + //FCLK Section + uint16_t FclkDpmDisallowPstateFreq; //Frequency which FW will target when indicated that display config cannot support P-state. Set to 0 use FW calculated value + uint16_t PaddingFclk; + + // Link DPM Settings + uint8_t PcieGenSpeed[NUM_LINK_LEVELS]; ///< 0:PciE-gen1 1:PciE-gen2 2:PciE-gen3 3:PciE-gen4 4:PciE-gen5 + uint8_t PcieLaneCount[NUM_LINK_LEVELS]; ///< 1=x1, 2=x2, 3=x4, 4=x8, 5=x12, 6=x16 + uint16_t LclkFreq[NUM_LINK_LEVELS]; + + // SECTION: VDD_GFX AVFS + uint8_t OverrideGfxAvfsFuses; + uint8_t GfxAvfsPadding[3]; + + uint32_t SocHwRtAvfsFuses[PP_GRTAVFS_HW_FUSE_COUNT]; //new added for Soc domain + uint32_t GfxL2HwRtAvfsFuses[PP_GRTAVFS_HW_FUSE_COUNT]; //see fusedoc for encoding + //uint32_t GfxSeHwRtAvfsFuses[PP_GRTAVFS_HW_FUSE_COUNT]; + uint32_t spare_HwRtAvfsFuses[PP_GRTAVFS_HW_FUSE_COUNT]; + + uint32_t SocCommonRtAvfs[PP_GRTAVFS_FW_COMMON_FUSE_COUNT]; + uint32_t GfxCommonRtAvfs[PP_GRTAVFS_FW_COMMON_FUSE_COUNT]; + + uint32_t SocFwRtAvfsFuses[PP_GRTAVFS_FW_SEP_FUSE_COUNT]; + uint32_t GfxL2FwRtAvfsFuses[PP_GRTAVFS_FW_SEP_FUSE_COUNT]; + //uint32_t GfxSeFwRtAvfsFuses[PP_GRTAVFS_FW_SEP_FUSE_COUNT]; + uint32_t spare_FwRtAvfsFuses[PP_GRTAVFS_FW_SEP_FUSE_COUNT]; + + uint32_t Soc_Droop_PWL_F[PP_NUM_RTAVFS_PWL_ZONES]; + uint32_t Soc_Droop_PWL_a[PP_NUM_RTAVFS_PWL_ZONES]; + uint32_t Soc_Droop_PWL_b[PP_NUM_RTAVFS_PWL_ZONES]; + uint32_t Soc_Droop_PWL_c[PP_NUM_RTAVFS_PWL_ZONES]; + + uint32_t Gfx_Droop_PWL_F[PP_NUM_RTAVFS_PWL_ZONES]; + uint32_t Gfx_Droop_PWL_a[PP_NUM_RTAVFS_PWL_ZONES]; + uint32_t Gfx_Droop_PWL_b[PP_NUM_RTAVFS_PWL_ZONES]; + uint32_t Gfx_Droop_PWL_c[PP_NUM_RTAVFS_PWL_ZONES]; + + uint32_t Gfx_Static_PWL_Offset[PP_NUM_RTAVFS_PWL_ZONES]; + uint32_t Soc_Static_PWL_Offset[PP_NUM_RTAVFS_PWL_ZONES]; + + uint32_t dGbV_dT_vmin; + uint32_t dGbV_dT_vmax; + + //Unused: PMFW-9370 + uint32_t V2F_vmin_range_low; + uint32_t V2F_vmin_range_high; + uint32_t V2F_vmax_range_low; + uint32_t V2F_vmax_range_high; + + AvfsDcBtcParams_t DcBtcGfxParams; + QuadraticInt_t SSCurve_GFX; + uint32_t GfxAvfsSpare[29]; + + //SECTION: VDD_SOC AVFS + uint8_t OverrideSocAvfsFuses; + uint8_t MinSocAvfsRevision; + uint8_t SocAvfsPadding[2]; + + AvfsFuseOverride_t SocAvfsFuseOverride[AVFS_D_COUNT]; + + DroopInt_t dBtcGbSoc[AVFS_D_COUNT]; // GHz->V BtcGb + + LinearInt_t qAgingGb[AVFS_D_COUNT]; // GHz->V + + QuadraticInt_t qStaticVoltageOffset[AVFS_D_COUNT]; // GHz->V + + AvfsDcBtcParams_t DcBtcSocParams[AVFS_D_COUNT]; + + QuadraticInt_t SSCurve_SOC; + uint32_t SocAvfsSpare[29]; + + //SECTION: Boot clock and voltage values + BootValues_t BootValues; + + //SECTION: Driver Reported Clocks + DriverReportedClocks_t DriverReportedClocks; + + //SECTION: Message Limits + MsgLimits_t MsgLimits; + + //SECTION: OverDrive Limits + OverDriveLimits_t OverDriveLimitsBasicMin; + OverDriveLimits_t OverDriveLimitsBasicMax; + OverDriveLimits_t OverDriveLimitsAdvancedMin; + OverDriveLimits_t OverDriveLimitsAdvancedMax; + + // Section: Total Board Power idle vs active coefficients + uint8_t TotalBoardPowerSupport; + uint8_t TotalBoardPowerPadding[1]; + uint16_t TotalBoardPowerRoc; + + //PMFW-11158 + QuadraticInt_t qFeffCoeffGameClock[POWER_SOURCE_COUNT]; + QuadraticInt_t qFeffCoeffBaseClock[POWER_SOURCE_COUNT]; + QuadraticInt_t qFeffCoeffBoostClock[POWER_SOURCE_COUNT]; + + // APT GFX to UCLK mapping + int32_t AptUclkGfxclkLookup[POWER_SOURCE_COUNT][6]; + uint32_t AptUclkGfxclkLookupHyst[POWER_SOURCE_COUNT][6]; + uint32_t AptPadding; + + // Xvmin didt + QuadraticInt_t GfxXvminDidtDroopThresh; + uint32_t GfxXvminDidtResetDDWait; + uint32_t GfxXvminDidtClkStopWait; + uint32_t GfxXvminDidtFcsStepCtrl; + uint32_t GfxXvminDidtFcsWaitCtrl; + + // PSM based didt controller + uint32_t PsmModeEnabled; //0: all disabled 1: static mode only 2: dynamic mode only 3:static + dynamic mode + uint32_t P2v_a; // floating point in U32 format + uint32_t P2v_b; + uint32_t P2v_c; + uint32_t T2p_a; + uint32_t T2p_b; + uint32_t T2p_c; + uint32_t P2vTemp; + QuadraticInt_t PsmDidtStaticSettings; + QuadraticInt_t PsmDidtDynamicSettings; + uint8_t PsmDidtAvgDiv; + uint8_t PsmDidtForceStall; + uint16_t PsmDidtReleaseTimer; + uint32_t PsmDidtStallPattern; //Will be written to both pattern 1 and didt_static_level_prog + // CAC EDC + uint32_t Leakage_C0; // in IEEE float + uint32_t Leakage_C1; // in IEEE float + uint32_t Leakage_C2; // in IEEE float + uint32_t Leakage_C3; // in IEEE float + uint32_t Leakage_C4; // in IEEE float + uint32_t Leakage_C5; // in IEEE float + uint32_t GFX_CLK_SCALAR; // in IEEE float + uint32_t GFX_CLK_INTERCEPT; // in IEEE float + uint32_t GFX_CAC_M; // in IEEE float + uint32_t GFX_CAC_B; // in IEEE float + uint32_t VDD_GFX_CurrentLimitGuardband; // in IEEE float + uint32_t DynToTotalCacScalar; // in IEEE + // GFX EDC XVMIN + uint32_t XVmin_Gfx_EdcThreshScalar; + uint32_t XVmin_Gfx_EdcEnableFreq; + uint32_t XVmin_Gfx_EdcPccAsStepCtrl; + uint32_t XVmin_Gfx_EdcPccAsWaitCtrl; + uint16_t XVmin_Gfx_EdcThreshold; + uint16_t XVmin_Gfx_EdcFiltHysWaitCtrl; + // SOC EDC XVMIN + uint32_t XVmin_Soc_EdcThreshScalar; + uint32_t XVmin_Soc_EdcEnableFreq; + uint32_t XVmin_Soc_EdcThreshold; // LPF: number of cycles Xvmin_trig_filt will react. + uint16_t XVmin_Soc_EdcStepUpTime; // 10 bit, refclk count to step up throttle when PCC remains asserted. + uint16_t XVmin_Soc_EdcStepDownTime;// 10 bit, refclk count to step down throttle when PCC remains asserted. + uint8_t XVmin_Soc_EdcInitPccStep; // 3 bit, First Pcc Step number that will applied when PCC asserts. + uint8_t PaddingSocEdc[3]; + + // Fuse Override for SOC and GFX XVMIN + uint8_t GfxXvminFuseOverride; + uint8_t SocXvminFuseOverride; + uint8_t PaddingXvminFuseOverride[2]; + uint8_t GfxXvminFddTempLow; // bit 7: sign, bit 0-6: ABS value + uint8_t GfxXvminFddTempHigh; // bit 7: sign, bit 0-6: ABS value + uint8_t SocXvminFddTempLow; // bit 7: sign, bit 0-6: ABS value + uint8_t SocXvminFddTempHigh; // bit 7: sign, bit 0-6: ABS value + + + uint16_t GfxXvminFddVolt0; // low voltage, in VID + uint16_t GfxXvminFddVolt1; // mid voltage, in VID + uint16_t GfxXvminFddVolt2; // high voltage, in VID + uint16_t SocXvminFddVolt0; // low voltage, in VID + uint16_t SocXvminFddVolt1; // mid voltage, in VID + uint16_t SocXvminFddVolt2; // high voltage, in VID + uint16_t GfxXvminDsFddDsm[6]; // XVMIN DS, same organization with fuse + uint16_t GfxXvminEdcFddDsm[6];// XVMIN GFX EDC, same organization with fuse + uint16_t SocXvminEdcFddDsm[6];// XVMIN SOC EDC, same organization with fuse + + // SECTION: Sku Reserved + uint32_t Spare; + + // Padding for MMHUB - do not modify this + uint32_t MmHubPadding[8]; +} SkuTable_t; + +typedef struct { + uint8_t SlewRateConditions; + uint8_t LoadLineAdjust; + uint8_t VoutOffset; + uint8_t VidMax; + uint8_t VidMin; + uint8_t TenBitTelEn; + uint8_t SixteenBitTelEn; + uint8_t OcpThresh; + uint8_t OcpWarnThresh; + uint8_t OcpSettings; + uint8_t VrhotThresh; + uint8_t OtpThresh; + uint8_t UvpOvpDeltaRef; + uint8_t PhaseShed; + uint8_t Padding[10]; + uint32_t SettingOverrideMask; +} Svi3RegulatorSettings_t; + +typedef struct { + // SECTION: Version + uint32_t Version; //should be unique to each board type + + // SECTION: I2C Control + I2cControllerConfig_t I2cControllers[NUM_I2C_CONTROLLERS]; + + //SECTION SVI3 Board Parameters + uint8_t SlaveAddrMapping[SVI_PLANE_COUNT]; + uint8_t VrPsiSupport[SVI_PLANE_COUNT]; + + uint32_t Svi3SvcSpeed; + uint8_t EnablePsi6[SVI_PLANE_COUNT]; // only applicable in SVI3 + + // SECTION: Voltage Regulator Settings + Svi3RegulatorSettings_t Svi3RegSettings[SVI_PLANE_COUNT]; + + // SECTION: GPIO Settings + uint8_t LedOffGpio; + uint8_t FanOffGpio; + uint8_t GfxVrPowerStageOffGpio; + + uint8_t AcDcGpio; // GPIO pin configured for AC/DC switching + uint8_t AcDcPolarity; // GPIO polarity for AC/DC switching + uint8_t VR0HotGpio; // GPIO pin configured for VR0 HOT event + uint8_t VR0HotPolarity; // GPIO polarity for VR0 HOT event + + uint8_t GthrGpio; // GPIO pin configured for GTHR Event + uint8_t GthrPolarity; // replace GPIO polarity for GTHR + + // LED Display Settings + uint8_t LedPin0; // GPIO number for LedPin[0] + uint8_t LedPin1; // GPIO number for LedPin[1] + uint8_t LedPin2; // GPIO number for LedPin[2] + uint8_t LedEnableMask; + + uint8_t LedPcie; // GPIO number for PCIE results + uint8_t LedError; // GPIO number for Error Cases + uint8_t PaddingLed; + + // SECTION: Clock Spread Spectrum + + // UCLK Spread Spectrum + uint8_t UclkTrainingModeSpreadPercent; // Q4.4 + uint8_t UclkSpreadPadding; + uint16_t UclkSpreadFreq; // kHz + + // UCLK Spread Spectrum + uint8_t UclkSpreadPercent[MEM_VENDOR_COUNT]; + + // DFLL Spread Spectrum + uint8_t GfxclkSpreadEnable; + + // FCLK Spread Spectrum + uint8_t FclkSpreadPercent; // Q4.4 + uint16_t FclkSpreadFreq; // kHz + + // Section: Memory Config + uint8_t DramWidth; // Width of interface to the channel for each DRAM module. See DRAM_BIT_WIDTH_TYPE_e + uint8_t PaddingMem1[7]; + + // SECTION: UMC feature flags + uint8_t HsrEnabled; + uint8_t VddqOffEnabled; + uint8_t PaddingUmcFlags[2]; + + uint32_t PostVoltageSetBacoDelay; // in microseconds. Amount of time FW will wait after power good is established or PSI0 command is issued + uint32_t BacoEntryDelay; // in milliseconds. Amount of time FW will wait to trigger BACO entry after receiving entry notification from OS + + uint8_t FuseWritePowerMuxPresent; + uint8_t FuseWritePadding[3]; + + // SECTION: EDC Params + uint32_t LoadlineGfx; + uint32_t LoadlineSoc; + uint32_t GfxEdcLimit; + uint32_t SocEdcLimit; + + uint32_t RestBoardPower; //power consumed by board that is not captured by the SVI3 input telemetry + uint32_t ConnectorsImpedance; // impedance of the input ATX power connectors + + uint8_t EpcsSens0; //GPIO number for External Power Connector Support Sense0 + uint8_t EpcsSens1; //GPIO Number for External Power Connector Support Sense1 + uint8_t PaddingEpcs[2]; + + // SECTION: Board Reserved + uint32_t BoardSpare[52]; + + // SECTION: Structure Padding + + // Padding for MMHUB - do not modify this + uint32_t MmHubPadding[8]; +} BoardTable_t; + +typedef struct { + // SECTION: Infrastructure Limits + uint16_t SocketPowerLimitAc[PPT_THROTTLER_COUNT]; // In Watts. Power limit that PMFW attempts to control to in AC mode. Multiple limits supported + + uint16_t VrTdcLimit[TDC_THROTTLER_COUNT]; // In Amperes. Current limit associated with VR regulator maximum temperature + + int16_t TotalIdleBoardPowerM; + int16_t TotalIdleBoardPowerB; + int16_t TotalBoardPowerM; + int16_t TotalBoardPowerB; + + uint16_t TemperatureLimit[TEMP_COUNT]; // In degrees Celsius. Temperature limit associated with each input + + // SECTION: Fan Control + uint16_t FanStopTemp[TEMP_COUNT]; //Celsius + uint16_t FanStartTemp[TEMP_COUNT]; //Celsius + + uint16_t FanGain[TEMP_COUNT]; + + uint16_t FanPwmMin; + uint16_t AcousticTargetRpmThreshold; + uint16_t AcousticLimitRpmThreshold; + uint16_t FanMaximumRpm; + uint16_t MGpuAcousticLimitRpmThreshold; + uint16_t FanTargetGfxclk; + uint32_t TempInputSelectMask; + uint8_t FanZeroRpmEnable; + uint8_t FanTachEdgePerRev; + uint16_t FanPadding; + uint16_t FanTargetTemperature[TEMP_COUNT]; + + // The following are AFC override parameters. Leave at 0 to use FW defaults. + int16_t FuzzyFan_ErrorSetDelta; + int16_t FuzzyFan_ErrorRateSetDelta; + int16_t FuzzyFan_PwmSetDelta; + uint16_t FuzzyFan_Reserved; + + uint16_t FwCtfLimit[TEMP_COUNT]; + + uint16_t IntakeTempEnableRPM; + int16_t IntakeTempOffsetTemp; + uint16_t IntakeTempReleaseTemp; + uint16_t IntakeTempHighIntakeAcousticLimit; + + uint16_t IntakeTempAcouticLimitReleaseRate; + int16_t FanAbnormalTempLimitOffset; // FanStalledTempLimitOffset + uint16_t FanStalledTriggerRpm; // + uint16_t FanAbnormalTriggerRpmCoeff; // FanAbnormalTriggerRpm + + uint16_t FanSpare[1]; + uint8_t FanIntakeSensorSupport; + uint8_t FanIntakePadding; + uint32_t FanAmbientPerfBoostThreshold; + uint32_t FanSpare2[12]; + + uint16_t TemperatureLimit_Hynix; // In degrees Celsius. Memory temperature limit associated with Hynix + uint16_t TemperatureLimit_Micron; // In degrees Celsius. Memory temperature limit associated with Micron + uint16_t TemperatureFwCtfLimit_Hynix; + uint16_t TemperatureFwCtfLimit_Micron; + + // SECTION: Board Reserved + uint16_t PlatformTdcLimit[TDC_THROTTLER_COUNT]; // In Amperes. Current limit associated with platform maximum temperature per VR current rail + uint16_t SocketPowerLimitDc[PPT_THROTTLER_COUNT]; // In Watts. Power limit that PMFW attempts to control to in DC mode. Multiple limits supported + uint16_t SocketPowerLimitSmartShift2; // In Watts. Power limit used SmartShift + uint16_t CustomSkuSpare16b; + uint32_t CustomSkuSpare32b[10]; + + // SECTION: Structure Padding + + // Padding for MMHUB - do not modify this + uint32_t MmHubPadding[8]; +} CustomSkuTable_t; + +typedef struct { + PFE_Settings_t PFE_Settings; + SkuTable_t SkuTable; + CustomSkuTable_t CustomSkuTable; + BoardTable_t BoardTable; +} PPTable_t; + +typedef struct { + // Time constant parameters for clock averages in ms + uint16_t GfxclkAverageLpfTau; + uint16_t FclkAverageLpfTau; + uint16_t UclkAverageLpfTau; + uint16_t GfxActivityLpfTau; + uint16_t UclkActivityLpfTau; + uint16_t UclkMaxActivityLpfTau; + uint16_t SocketPowerLpfTau; + uint16_t VcnClkAverageLpfTau; + uint16_t VcnUsageAverageLpfTau; + uint16_t PcieActivityLpTau; +} DriverSmuConfig_t; + +typedef struct { + DriverSmuConfig_t DriverSmuConfig; + + uint32_t Spare[8]; + // Padding - ignore + uint32_t MmHubPadding[8]; // SMU internal use +} DriverSmuConfigExternal_t; + + +typedef struct { + + uint16_t FreqTableGfx [NUM_GFXCLK_DPM_LEVELS ]; // In MHz + uint16_t FreqTableVclk [NUM_VCLK_DPM_LEVELS ]; // In MHz + uint16_t FreqTableDclk [NUM_DCLK_DPM_LEVELS ]; // In MHz + uint16_t FreqTableSocclk [NUM_SOCCLK_DPM_LEVELS ]; // In MHz + uint16_t FreqTableUclk [NUM_UCLK_DPM_LEVELS ]; // In MHz + uint16_t FreqTableDispclk [NUM_DISPCLK_DPM_LEVELS ]; // In MHz + uint16_t FreqTableDppClk [NUM_DPPCLK_DPM_LEVELS ]; // In MHz + uint16_t FreqTableDprefclk [NUM_DPREFCLK_DPM_LEVELS]; // In MHz + uint16_t FreqTableDcfclk [NUM_DCFCLK_DPM_LEVELS ]; // In MHz + uint16_t FreqTableDtbclk [NUM_DTBCLK_DPM_LEVELS ]; // In MHz + uint16_t FreqTableFclk [NUM_FCLK_DPM_LEVELS ]; // In MHz + + uint16_t DcModeMaxFreq [PPCLK_COUNT ]; // In MHz + + uint16_t Padding; + + uint32_t Spare[32]; + + // Padding - ignore + uint32_t MmHubPadding[8]; // SMU internal use + +} DriverInfoTable_t; + +typedef struct { + uint32_t CurrClock[PPCLK_COUNT]; + + uint16_t AverageGfxclkFrequencyTarget; + uint16_t AverageGfxclkFrequencyPreDs; + uint16_t AverageGfxclkFrequencyPostDs; + uint16_t AverageFclkFrequencyPreDs; + uint16_t AverageFclkFrequencyPostDs; + uint16_t AverageMemclkFrequencyPreDs ; // this is scaled to actual memory clock + uint16_t AverageMemclkFrequencyPostDs ; // this is scaled to actual memory clock + uint16_t AverageVclk0Frequency ; + uint16_t AverageDclk0Frequency ; + uint16_t AverageVclk1Frequency ; + uint16_t AverageDclk1Frequency ; + uint16_t PCIeBusy ; + uint16_t dGPU_W_MAX ; + uint16_t padding ; + + uint32_t MetricsCounter ; + + uint16_t AvgVoltage[SVI_PLANE_COUNT]; + uint16_t AvgCurrent[SVI_PLANE_COUNT]; + + uint16_t AverageGfxActivity ; + uint16_t AverageUclkActivity ; + uint16_t Vcn0ActivityPercentage ; + uint16_t Vcn1ActivityPercentage ; + + uint32_t EnergyAccumulator; + uint16_t AverageSocketPower; + uint16_t AverageTotalBoardPower; + + uint16_t AvgTemperature[TEMP_COUNT]; + uint16_t AvgTemperatureFanIntake; + + uint8_t PcieRate ; + uint8_t PcieWidth ; + + uint8_t AvgFanPwm; + uint8_t Padding[1]; + uint16_t AvgFanRpm; + + + uint8_t ThrottlingPercentage[THROTTLER_COUNT]; + uint8_t padding1[3]; + + //metrics for D3hot entry/exit and driver ARM msgs + uint32_t D3HotEntryCountPerMode[D3HOT_SEQUENCE_COUNT]; + uint32_t D3HotExitCountPerMode[D3HOT_SEQUENCE_COUNT]; + uint32_t ArmMsgReceivedCountPerMode[D3HOT_SEQUENCE_COUNT]; + + uint16_t ApuSTAPMSmartShiftLimit; + uint16_t ApuSTAPMLimit; + uint16_t AvgApuSocketPower; + + uint16_t AverageUclkActivity_MAX; + + uint32_t PublicSerialNumberLower; + uint32_t PublicSerialNumberUpper; + +} SmuMetrics_t; + +typedef struct { + SmuMetrics_t SmuMetrics; + uint32_t Spare[30]; + + // Padding - ignore + uint32_t MmHubPadding[8]; // SMU internal use +} SmuMetricsExternal_t; + +typedef struct { + uint8_t WmSetting; + uint8_t Flags; + uint8_t Padding[2]; + +} WatermarkRowGeneric_t; + +#define NUM_WM_RANGES 4 + +typedef enum { + WATERMARKS_CLOCK_RANGE = 0, + WATERMARKS_DUMMY_PSTATE, + WATERMARKS_MALL, + WATERMARKS_COUNT, +} WATERMARKS_FLAGS_e; + +typedef struct { + // Watermarks + WatermarkRowGeneric_t WatermarkRow[NUM_WM_RANGES]; +} Watermarks_t; + +typedef struct { + Watermarks_t Watermarks; + uint32_t Spare[16]; + + uint32_t MmHubPadding[8]; // SMU internal use +} WatermarksExternal_t; + +typedef struct { + uint16_t avgPsmCount[76]; + uint16_t minPsmCount[76]; + uint16_t maxPsmCount[76]; + float avgPsmVoltage[76]; + float minPsmVoltage[76]; + float maxPsmVoltage[76]; +} AvfsDebugTable_t; + +typedef struct { + AvfsDebugTable_t AvfsDebugTable; + + uint32_t MmHubPadding[8]; // SMU internal use +} AvfsDebugTableExternal_t; + + +typedef struct { + uint8_t Gfx_ActiveHystLimit; + uint8_t Gfx_IdleHystLimit; + uint8_t Gfx_FPS; + uint8_t Gfx_MinActiveFreqType; + uint8_t Gfx_BoosterFreqType; + uint8_t PaddingGfx; + uint16_t Gfx_MinActiveFreq; // MHz + uint16_t Gfx_BoosterFreq; // MHz + uint16_t Gfx_PD_Data_time_constant; // Time constant of PD controller in ms + uint32_t Gfx_PD_Data_limit_a; // Q16 + uint32_t Gfx_PD_Data_limit_b; // Q16 + uint32_t Gfx_PD_Data_limit_c; // Q16 + uint32_t Gfx_PD_Data_error_coeff; // Q16 + uint32_t Gfx_PD_Data_error_rate_coeff; // Q16 + + uint8_t Fclk_ActiveHystLimit; + uint8_t Fclk_IdleHystLimit; + uint8_t Fclk_FPS; + uint8_t Fclk_MinActiveFreqType; + uint8_t Fclk_BoosterFreqType; + uint8_t PaddingFclk; + uint16_t Fclk_MinActiveFreq; // MHz + uint16_t Fclk_BoosterFreq; // MHz + uint16_t Fclk_PD_Data_time_constant; // Time constant of PD controller in ms + uint32_t Fclk_PD_Data_limit_a; // Q16 + uint32_t Fclk_PD_Data_limit_b; // Q16 + uint32_t Fclk_PD_Data_limit_c; // Q16 + uint32_t Fclk_PD_Data_error_coeff; // Q16 + uint32_t Fclk_PD_Data_error_rate_coeff; // Q16 + + uint32_t Mem_UpThreshold_Limit[NUM_UCLK_DPM_LEVELS]; // Q16 + uint8_t Mem_UpHystLimit[NUM_UCLK_DPM_LEVELS]; + uint16_t Mem_DownHystLimit[NUM_UCLK_DPM_LEVELS]; + uint16_t Mem_Fps; + +} DpmActivityMonitorCoeffInt_t; + + +typedef struct { + DpmActivityMonitorCoeffInt_t DpmActivityMonitorCoeffInt; + uint32_t MmHubPadding[8]; // SMU internal use +} DpmActivityMonitorCoeffIntExternal_t; + + + +// Workload bits +#define WORKLOAD_PPLIB_DEFAULT_BIT 0 +#define WORKLOAD_PPLIB_FULL_SCREEN_3D_BIT 1 +#define WORKLOAD_PPLIB_POWER_SAVING_BIT 2 +#define WORKLOAD_PPLIB_VIDEO_BIT 3 +#define WORKLOAD_PPLIB_VR_BIT 4 +#define WORKLOAD_PPLIB_COMPUTE_BIT 5 +#define WORKLOAD_PPLIB_CUSTOM_BIT 6 +#define WORKLOAD_PPLIB_WINDOW_3D_BIT 7 +#define WORKLOAD_PPLIB_DIRECT_ML_BIT 8 +#define WORKLOAD_PPLIB_CGVDI_BIT 9 +#define WORKLOAD_PPLIB_COUNT 10 + + +// These defines are used with the following messages: +// SMC_MSG_TransferTableDram2Smu +// SMC_MSG_TransferTableSmu2Dram + +// Table transfer status +#define TABLE_TRANSFER_OK 0x0 +#define TABLE_TRANSFER_FAILED 0xFF +#define TABLE_TRANSFER_PENDING 0xAB + +// Table types +#define TABLE_PPTABLE 0 +#define TABLE_COMBO_PPTABLE 1 +#define TABLE_WATERMARKS 2 +#define TABLE_AVFS_PSM_DEBUG 3 +#define TABLE_PMSTATUSLOG 4 +#define TABLE_SMU_METRICS 5 +#define TABLE_DRIVER_SMU_CONFIG 6 +#define TABLE_ACTIVITY_MONITOR_COEFF 7 +#define TABLE_OVERDRIVE 8 +#define TABLE_I2C_COMMANDS 9 +#define TABLE_DRIVER_INFO 10 +#define TABLE_ECCINFO 11 +#define TABLE_CUSTOM_SKUTABLE 12 +#define TABLE_COUNT 13 + +//IH Interupt ID +#define IH_INTERRUPT_ID_TO_DRIVER 0xFE +#define IH_INTERRUPT_CONTEXT_ID_BACO 0x2 +#define IH_INTERRUPT_CONTEXT_ID_AC 0x3 +#define IH_INTERRUPT_CONTEXT_ID_DC 0x4 +#define IH_INTERRUPT_CONTEXT_ID_AUDIO_D0 0x5 +#define IH_INTERRUPT_CONTEXT_ID_AUDIO_D3 0x6 +#define IH_INTERRUPT_CONTEXT_ID_THERMAL_THROTTLING 0x7 +#define IH_INTERRUPT_CONTEXT_ID_FAN_ABNORMAL 0x8 +#define IH_INTERRUPT_CONTEXT_ID_FAN_RECOVERY 0x9 + +#endif diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu14_driver_if_v14_0_0.h b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu14_driver_if_v14_0_0.h index 5bb7a63c0602..97522c085258 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu14_driver_if_v14_0_0.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu14_driver_if_v14_0_0.h @@ -144,6 +144,37 @@ typedef struct { uint32_t MaxGfxClk; } DpmClocks_t; +//Freq in MHz +//Voltage in milli volts with 2 fractional bits +typedef struct { + uint32_t DcfClocks[NUM_DCFCLK_DPM_LEVELS]; + uint32_t DispClocks[NUM_DISPCLK_DPM_LEVELS]; + uint32_t DppClocks[NUM_DPPCLK_DPM_LEVELS]; + uint32_t SocClocks[NUM_SOCCLK_DPM_LEVELS]; + uint32_t VClocks0[NUM_VCN_DPM_LEVELS]; + uint32_t VClocks1[NUM_VCN_DPM_LEVELS]; + uint32_t DClocks0[NUM_VCN_DPM_LEVELS]; + uint32_t DClocks1[NUM_VCN_DPM_LEVELS]; + uint32_t VPEClocks[NUM_VPE_DPM_LEVELS]; + uint32_t FclkClocks_Freq[NUM_FCLK_DPM_LEVELS]; + uint32_t FclkClocks_Voltage[NUM_FCLK_DPM_LEVELS]; + uint32_t SocVoltage[NUM_SOC_VOLTAGE_LEVELS]; + MemPstateTable_t MemPstateTable[NUM_MEM_PSTATE_LEVELS]; + + uint8_t NumDcfClkLevelsEnabled; + uint8_t NumDispClkLevelsEnabled; //Applies to both Dispclk and Dppclk + uint8_t NumSocClkLevelsEnabled; + uint8_t Vcn0ClkLevelsEnabled; //Applies to both Vclk0 and Dclk0 + uint8_t Vcn1ClkLevelsEnabled; //Applies to both Vclk1 and Dclk1 + uint8_t VpeClkLevelsEnabled; + uint8_t NumMemPstatesEnabled; + uint8_t NumFclkLevelsEnabled; + uint8_t spare; + + uint32_t MinGfxClk; + uint32_t MaxGfxClk; +} DpmClocks_t_v14_0_1; + typedef struct { uint16_t CoreFrequency[16]; //Target core frequency [MHz] uint16_t CorePower[16]; //CAC calculated core power [mW] @@ -224,7 +255,7 @@ typedef enum { #define TABLE_CUSTOM_DPM 2 // Called by Driver #define TABLE_BIOS_GPIO_CONFIG 3 // Called by BIOS #define TABLE_DPMCLOCKS 4 // Called by Driver and VBIOS -#define TABLE_SPARE0 5 // Unused +#define TABLE_MOMENTARY_PM 5 // Called by Tools #define TABLE_MODERN_STDBY 6 // Called by Tools for Modern Standby Log #define TABLE_SMU_METRICS 7 // Called by Driver and SMF/PMF #define TABLE_COUNT 8 diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v13_0_6_pmfw.h b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v13_0_6_pmfw.h index 7b812b9994d7..0b3c2f54a343 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v13_0_6_pmfw.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v13_0_6_pmfw.h @@ -123,7 +123,7 @@ typedef enum { VOLTAGE_GUARDBAND_COUNT } GFX_GUARDBAND_e; -#define SMU_METRICS_TABLE_VERSION 0xB +#define SMU_METRICS_TABLE_VERSION 0xC typedef struct __attribute__((packed, aligned(4))) { uint32_t AccumulationCounter; @@ -223,6 +223,10 @@ typedef struct __attribute__((packed, aligned(4))) { // VCN/JPEG ACTIVITY uint32_t VcnBusy[4]; uint32_t JpegBusy[32]; + + // PCIE LINK Speed and width + uint32_t PCIeLinkSpeed; + uint32_t PCIeLinkWidth; } MetricsTableX_t; typedef struct __attribute__((packed, aligned(4))) { diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v14_0_0_pmfw.h b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v14_0_0_pmfw.h index 356e0f57a426..ddb625860083 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v14_0_0_pmfw.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v14_0_0_pmfw.h @@ -42,7 +42,7 @@ #define FEATURE_EDC_BIT 7 #define FEATURE_PLL_POWER_DOWN_BIT 8 #define FEATURE_VDDOFF_BIT 9 -#define FEATURE_VCN_DPM_BIT 10 +#define FEATURE_VCN_DPM_BIT 10 /* this is for both VCN0 and VCN1 */ #define FEATURE_DS_MPM_BIT 11 #define FEATURE_FCLK_DPM_BIT 12 #define FEATURE_SOCCLK_DPM_BIT 13 @@ -56,9 +56,9 @@ #define FEATURE_DS_GFXCLK_BIT 21 #define FEATURE_DS_SOCCLK_BIT 22 #define FEATURE_DS_LCLK_BIT 23 -#define FEATURE_LOW_POWER_DCNCLKS_BIT 24 // for all DISP clks +#define FEATURE_LOW_POWER_DCNCLKS_BIT 24 #define FEATURE_DS_SHUBCLK_BIT 25 -#define FEATURE_SPARE0_BIT 26 //SPARE +#define FEATURE_RESERVED0_BIT 26 #define FEATURE_ZSTATES_BIT 27 #define FEATURE_IOMMUL2_PG_BIT 28 #define FEATURE_DS_FCLK_BIT 29 @@ -66,8 +66,8 @@ #define FEATURE_DS_MP1CLK_BIT 31 #define FEATURE_WHISPER_MODE_BIT 32 #define FEATURE_SMU_LOW_POWER_BIT 33 -#define FEATURE_SMART_L3_RINSER_BIT 34 -#define FEATURE_SPARE1_BIT 35 //SPARE +#define FEATURE_RESERVED1_BIT 34 /* v14_0_0 SMART_L3_RINSER; v14_0_1 RESERVED1 */ +#define FEATURE_GFX_DEM_BIT 35 /* v14_0_0 SPARE; v14_0_1 GFX_DEM */ #define FEATURE_PSI_BIT 36 #define FEATURE_PROCHOT_BIT 37 #define FEATURE_CPUOFF_BIT 38 @@ -77,11 +77,11 @@ #define FEATURE_PERF_LIMIT_BIT 42 #define FEATURE_CORE_DLDO_BIT 43 #define FEATURE_DVO_BIT 44 -#define FEATURE_DS_VCN_BIT 45 +#define FEATURE_DS_VCN_BIT 45 /* v14_0_1 this is for both VCN0 and VCN1 */ #define FEATURE_CPPC_BIT 46 #define FEATURE_CPPC_PREFERRED_CORES 47 #define FEATURE_DF_CSTATES_BIT 48 -#define FEATURE_SPARE2_BIT 49 //SPARE +#define FEATURE_FAST_PSTATE_CLDO_BIT 49 /* v14_0_0 SPARE */ #define FEATURE_ATHUB_PG_BIT 50 #define FEATURE_VDDOFF_ECO_BIT 51 #define FEATURE_ZSTATES_ECO_BIT 52 @@ -93,8 +93,8 @@ #define FEATURE_DS_IPUCLK_BIT 58 #define FEATURE_DS_VPECLK_BIT 59 #define FEATURE_VPE_DPM_BIT 60 -#define FEATURE_SPARE_61 61 -#define FEATURE_FP_DIDT 62 +#define FEATURE_SMART_L3_RINSER_BIT 61 /* v14_0_0 SPARE*/ +#define FEATURE_PCC_BIT 62 /* v14_0_0 FP_DIDT v14_0_1 PCC_BIT */ #define NUM_FEATURES 63 // Firmware Header/Footer @@ -151,6 +151,43 @@ typedef struct { // MP1_EXT_SCRATCH7 = RTOS Current Job } FwStatus_t; +typedef struct { + // MP1_EXT_SCRATCH0 + uint32_t DpmHandlerID : 8; + uint32_t ActivityMonitorID : 8; + uint32_t DpmTimerID : 8; + uint32_t DpmHubID : 4; + uint32_t DpmHubTask : 4; + // MP1_EXT_SCRATCH1 + uint32_t CclkSyncStatus : 8; + uint32_t ZstateStatus : 4; + uint32_t Cpu1VddOff : 4; + uint32_t DstateFun : 4; + uint32_t DstateDev : 4; + uint32_t GfxOffStatus : 2; + uint32_t Cpu0Off : 2; + uint32_t Cpu1Off : 2; + uint32_t Cpu0VddOff : 2; + // MP1_EXT_SCRATCH2 + uint32_t P2JobHandler :32; + // MP1_EXT_SCRATCH3 + uint32_t PostCode :32; + // MP1_EXT_SCRATCH4 + uint32_t MsgPortBusy :15; + uint32_t RsmuPmiP1Pending : 1; + uint32_t RsmuPmiP2PendingCnt : 8; + uint32_t DfCstateExitPending : 1; + uint32_t Pc6EntryPending : 1; + uint32_t Pc6ExitPending : 1; + uint32_t WarmResetPending : 1; + uint32_t Mp0ClkPending : 1; + uint32_t InWhisperMode : 1; + uint32_t spare2 : 2; + // MP1_EXT_SCRATCH5 + uint32_t IdleMask :32; + // MP1_EXT_SCRATCH6 = RTOS threads' status + // MP1_EXT_SCRATCH7 = RTOS Current Job +} FwStatus_t_v14_0_1; #pragma pack(pop) diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v14_0_0_ppsmc.h b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v14_0_0_ppsmc.h index 8a8a57c56bc0..c4dc5881d8df 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v14_0_0_ppsmc.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v14_0_0_ppsmc.h @@ -54,14 +54,14 @@ #define PPSMC_MSG_TestMessage 0x01 ///< To check if PMFW is alive and responding. Requirement specified by PMFW team #define PPSMC_MSG_GetPmfwVersion 0x02 ///< Get PMFW version #define PPSMC_MSG_GetDriverIfVersion 0x03 ///< Get PMFW_DRIVER_IF version -#define PPSMC_MSG_SPARE0 0x04 ///< SPARE -#define PPSMC_MSG_SPARE1 0x05 ///< SPARE -#define PPSMC_MSG_PowerDownVcn 0x06 ///< Power down VCN -#define PPSMC_MSG_PowerUpVcn 0x07 ///< Power up VCN; VCN is power gated by default -#define PPSMC_MSG_SetHardMinVcn 0x08 ///< For wireless display +#define PPSMC_MSG_PowerDownVcn1 0x04 ///< Power down VCN1 +#define PPSMC_MSG_PowerUpVcn1 0x05 ///< Power up VCN1; VCN1 is power gated by default +#define PPSMC_MSG_PowerDownVcn0 0x06 ///< Power down VCN0 +#define PPSMC_MSG_PowerUpVcn0 0x07 ///< Power up VCN0; VCN0 is power gated by default +#define PPSMC_MSG_SetHardMinVcn0 0x08 ///< For wireless display #define PPSMC_MSG_SetSoftMinGfxclk 0x09 ///< Set SoftMin for GFXCLK, argument is frequency in MHz -#define PPSMC_MSG_SPARE2 0x0A ///< SPARE -#define PPSMC_MSG_SPARE3 0x0B ///< SPARE +#define PPSMC_MSG_SetHardMinVcn1 0x0A ///< For wireless display +#define PPSMC_MSG_SetSoftMinVcn1 0x0B ///< Set soft min for VCN1 clocks (VCLK1 and DCLK1) #define PPSMC_MSG_PrepareMp1ForUnload 0x0C ///< Prepare PMFW for GFX driver unload #define PPSMC_MSG_SetDriverDramAddrHigh 0x0D ///< Set high 32 bits of DRAM address for Driver table transfer #define PPSMC_MSG_SetDriverDramAddrLow 0x0E ///< Set low 32 bits of DRAM address for Driver table transfer @@ -71,36 +71,32 @@ #define PPSMC_MSG_GetEnabledSmuFeatures 0x12 ///< Get enabled features in PMFW #define PPSMC_MSG_SetHardMinSocclkByFreq 0x13 ///< Set hard min for SOC CLK #define PPSMC_MSG_SetSoftMinFclk 0x14 ///< Set hard min for FCLK -#define PPSMC_MSG_SetSoftMinVcn 0x15 ///< Set soft min for VCN clocks (VCLK and DCLK) - +#define PPSMC_MSG_SetSoftMinVcn0 0x15 ///< Set soft min for VCN0 clocks (VCLK0 and DCLK0) #define PPSMC_MSG_EnableGfxImu 0x16 ///< Enable GFX IMU - -#define PPSMC_MSG_spare_0x17 0x17 -#define PPSMC_MSG_spare_0x18 0x18 +#define PPSMC_MSG_spare_0x17 0x17 ///< Get GFX clock frequency +#define PPSMC_MSG_spare_0x18 0x18 ///< Get FCLK frequency #define PPSMC_MSG_AllowGfxOff 0x19 ///< Inform PMFW of allowing GFXOFF entry #define PPSMC_MSG_DisallowGfxOff 0x1A ///< Inform PMFW of disallowing GFXOFF entry #define PPSMC_MSG_SetSoftMaxGfxClk 0x1B ///< Set soft max for GFX CLK #define PPSMC_MSG_SetHardMinGfxClk 0x1C ///< Set hard min for GFX CLK - #define PPSMC_MSG_SetSoftMaxSocclkByFreq 0x1D ///< Set soft max for SOC CLK #define PPSMC_MSG_SetSoftMaxFclkByFreq 0x1E ///< Set soft max for FCLK -#define PPSMC_MSG_SetSoftMaxVcn 0x1F ///< Set soft max for VCN clocks (VCLK and DCLK) -#define PPSMC_MSG_spare_0x20 0x20 -#define PPSMC_MSG_PowerDownJpeg 0x21 ///< Power down Jpeg -#define PPSMC_MSG_PowerUpJpeg 0x22 ///< Power up Jpeg; VCN is power gated by default - +#define PPSMC_MSG_SetSoftMaxVcn0 0x1F ///< Set soft max for VCN0 clocks (VCLK0 and DCLK0) +#define PPSMC_MSG_spare_0x20 0x20 ///< Set power limit percentage +#define PPSMC_MSG_PowerDownJpeg0 0x21 ///< Power down Jpeg of VCN0 +#define PPSMC_MSG_PowerUpJpeg0 0x22 ///< Power up Jpeg of VCN0; VCN0 is power gated by default #define PPSMC_MSG_SetHardMinFclkByFreq 0x23 ///< Set hard min for FCLK #define PPSMC_MSG_SetSoftMinSocclkByFreq 0x24 ///< Set soft min for SOC CLK #define PPSMC_MSG_AllowZstates 0x25 ///< Inform PMFM of allowing Zstate entry, i.e. no Miracast activity -#define PPSMC_MSG_Reserved 0x26 ///< Not used -#define PPSMC_MSG_Reserved1 0x27 ///< Not used, previously PPSMC_MSG_RequestActiveWgp -#define PPSMC_MSG_Reserved2 0x28 ///< Not used, previously PPSMC_MSG_QueryActiveWgp +#define PPSMC_MSG_PowerDownJpeg1 0x26 ///< Power down Jpeg of VCN1 +#define PPSMC_MSG_PowerUpJpeg1 0x27 ///< Power up Jpeg of VCN1; VCN1 is power gated by default +#define PPSMC_MSG_SetSoftMaxVcn1 0x28 ///< Set soft max for VCN1 clocks (VCLK1 and DCLK1) #define PPSMC_MSG_PowerDownIspByTile 0x29 ///< ISP is power gated by default #define PPSMC_MSG_PowerUpIspByTile 0x2A ///< This message is used to power up ISP tiles and enable the ISP DPM #define PPSMC_MSG_SetHardMinIspiclkByFreq 0x2B ///< Set HardMin by frequency for ISPICLK #define PPSMC_MSG_SetHardMinIspxclkByFreq 0x2C ///< Set HardMin by frequency for ISPXCLK -#define PPSMC_MSG_PowerDownUmsch 0x2D ///< Power down VCN.UMSCH (aka VSCH) scheduler -#define PPSMC_MSG_PowerUpUmsch 0x2E ///< Power up VCN.UMSCH (aka VSCH) scheduler +#define PPSMC_MSG_PowerDownUmsch 0x2D ///< Power down VCN0.UMSCH (aka VSCH) scheduler +#define PPSMC_MSG_PowerUpUmsch 0x2E ///< Power up VCN0.UMSCH (aka VSCH) scheduler #define PPSMC_Message_IspStutterOn_MmhubPgDis 0x2F ///< ISP StutterOn mmHub PgDis #define PPSMC_Message_IspStutterOff_MmhubPgEn 0x30 ///< ISP StufferOff mmHub PgEn #define PPSMC_MSG_PowerUpVpe 0x31 ///< Power up VPE @@ -110,7 +106,9 @@ #define PPSMC_MSG_DisableLSdma 0x35 ///< Disable LSDMA #define PPSMC_MSG_SetSoftMaxVpe 0x36 ///< #define PPSMC_MSG_SetSoftMinVpe 0x37 ///< -#define PPSMC_Message_Count 0x38 ///< Total number of PPSMC messages +#define PPSMC_MSG_AllocMALLCache 0x38 ///< Allocating MALL Cache +#define PPSMC_MSG_ReleaseMALLCache 0x39 ///< Releasing MALL Cache +#define PPSMC_Message_Count 0x3A ///< Total number of PPSMC messages /** @}*/ /** diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v14_0_2_ppsmc.h b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v14_0_2_ppsmc.h new file mode 100644 index 000000000000..de2e442281ff --- /dev/null +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v14_0_2_ppsmc.h @@ -0,0 +1,140 @@ +/* + * Copyright 2023 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ + +#ifndef SMU_V14_0_2_PPSMC_H +#define SMU_V14_0_2_PPSMC_H + +#define PPSMC_VERSION 0x1 + +// SMU Response Codes: +#define PPSMC_Result_OK 0x1 +#define PPSMC_Result_Failed 0xFF +#define PPSMC_Result_UnknownCmd 0xFE +#define PPSMC_Result_CmdRejectedPrereq 0xFD +#define PPSMC_Result_CmdRejectedBusy 0xFC + +// Message Definitions: +// BASIC +#define PPSMC_MSG_TestMessage 0x1 +#define PPSMC_MSG_GetSmuVersion 0x2 +#define PPSMC_MSG_GetDriverIfVersion 0x3 +#define PPSMC_MSG_SetAllowedFeaturesMaskLow 0x4 +#define PPSMC_MSG_SetAllowedFeaturesMaskHigh 0x5 +#define PPSMC_MSG_EnableAllSmuFeatures 0x6 +#define PPSMC_MSG_DisableAllSmuFeatures 0x7 +#define PPSMC_MSG_EnableSmuFeaturesLow 0x8 +#define PPSMC_MSG_EnableSmuFeaturesHigh 0x9 +#define PPSMC_MSG_DisableSmuFeaturesLow 0xA +#define PPSMC_MSG_DisableSmuFeaturesHigh 0xB +#define PPSMC_MSG_GetRunningSmuFeaturesLow 0xC +#define PPSMC_MSG_GetRunningSmuFeaturesHigh 0xD +#define PPSMC_MSG_SetDriverDramAddrHigh 0xE +#define PPSMC_MSG_SetDriverDramAddrLow 0xF +#define PPSMC_MSG_SetToolsDramAddrHigh 0x10 +#define PPSMC_MSG_SetToolsDramAddrLow 0x11 +#define PPSMC_MSG_TransferTableSmu2Dram 0x12 +#define PPSMC_MSG_TransferTableDram2Smu 0x13 +#define PPSMC_MSG_UseDefaultPPTable 0x14 + +//BACO/BAMACO/BOMACO +#define PPSMC_MSG_EnterBaco 0x15 +#define PPSMC_MSG_ExitBaco 0x16 +#define PPSMC_MSG_ArmD3 0x17 +#define PPSMC_MSG_BacoAudioD3PME 0x18 + +//DPM +#define PPSMC_MSG_SetSoftMinByFreq 0x19 +#define PPSMC_MSG_SetSoftMaxByFreq 0x1A +#define PPSMC_MSG_SetHardMinByFreq 0x1B +#define PPSMC_MSG_SetHardMaxByFreq 0x1C +#define PPSMC_MSG_GetMinDpmFreq 0x1D +#define PPSMC_MSG_GetMaxDpmFreq 0x1E +#define PPSMC_MSG_GetDpmFreqByIndex 0x1F +#define PPSMC_MSG_OverridePcieParameters 0x20 + +//DramLog Set DramAddr +#define PPSMC_MSG_DramLogSetDramAddrHigh 0x21 +#define PPSMC_MSG_DramLogSetDramAddrLow 0x22 +#define PPSMC_MSG_DramLogSetDramSize 0x23 +#define PPSMC_MSG_SetWorkloadMask 0x24 + +#define PPSMC_MSG_GetVoltageByDpm 0x25 // Can be removed +#define PPSMC_MSG_SetVideoFps 0x26 // Can be removed +#define PPSMC_MSG_GetDcModeMaxDpmFreq 0x27 + +//Power Gating +#define PPSMC_MSG_AllowGfxOff 0x28 +#define PPSMC_MSG_DisallowGfxOff 0x29 +#define PPSMC_MSG_PowerUpVcn 0x2A +#define PPSMC_MSG_PowerDownVcn 0x2B +#define PPSMC_MSG_PowerUpJpeg 0x2C +#define PPSMC_MSG_PowerDownJpeg 0x2D + +//Resets +#define PPSMC_MSG_PrepareMp1ForUnload 0x2E +#define PPSMC_MSG_Mode1Reset 0x2F + +//Set SystemVirtual DramAddrHigh +#define PPSMC_MSG_SetSystemVirtualDramAddrHigh 0x30 +#define PPSMC_MSG_SetSystemVirtualDramAddrLow 0x31 +//ACDC Power Source +#define PPSMC_MSG_SetPptLimit 0x32 +#define PPSMC_MSG_GetPptLimit 0x33 +#define PPSMC_MSG_ReenableAcDcInterrupt 0x34 +#define PPSMC_MSG_NotifyPowerSource 0x35 + +//BTC +#define PPSMC_MSG_RunDcBtc 0x36 + +// 0x37 + +//Others +#define PPSMC_MSG_SetTemperatureInputSelect 0x38 // Can be removed +#define PPSMC_MSG_SetFwDstatesMask 0x39 +#define PPSMC_MSG_SetThrottlerMask 0x3A + +#define PPSMC_MSG_SetExternalClientDfCstateAllow 0x3B + +#define PPSMC_MSG_SetMGpuFanBoostLimitRpm 0x3C + +//STB to dram log +#define PPSMC_MSG_DumpSTBtoDram 0x3D +#define PPSMC_MSG_STBtoDramLogSetDramAddrHigh 0x3E +#define PPSMC_MSG_STBtoDramLogSetDramAddrLow 0x3F +#define PPSMC_MSG_STBtoDramLogSetDramSize 0x40 +#define PPSMC_MSG_SetOBMTraceBufferLogging 0x41 + +#define PPSMC_MSG_AllowGfxDcs 0x43 +#define PPSMC_MSG_DisallowGfxDcs 0x44 +#define PPSMC_MSG_EnableAudioStutterWA 0x45 +#define PPSMC_MSG_PowerUpUmsch 0x46 +#define PPSMC_MSG_PowerDownUmsch 0x47 +#define PPSMC_MSG_SetDcsArch 0x48 +#define PPSMC_MSG_TriggerVFFLR 0x49 +#define PPSMC_MSG_SetNumBadMemoryPagesRetired 0x4A +#define PPSMC_MSG_SetBadMemoryPagesRetiredFlagsPerChannel 0x4B +#define PPSMC_MSG_SetPriorityDeltaGain 0x4C +#define PPSMC_MSG_AllowIHHostInterrupt 0x4D +#define PPSMC_MSG_Mode3Reset 0x4F +#define PPSMC_Message_Count 0x50 +#endif diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_types.h b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_types.h index a941fdbf78b6..c48214e3dc8e 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_types.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_types.h @@ -115,6 +115,10 @@ __SMU_DUMMY_MAP(PowerDownVcn), \ __SMU_DUMMY_MAP(PowerUpJpeg), \ __SMU_DUMMY_MAP(PowerDownJpeg), \ + __SMU_DUMMY_MAP(PowerUpJpeg0), \ + __SMU_DUMMY_MAP(PowerDownJpeg0), \ + __SMU_DUMMY_MAP(PowerUpJpeg1), \ + __SMU_DUMMY_MAP(PowerDownJpeg1), \ __SMU_DUMMY_MAP(BacoAudioD3PME), \ __SMU_DUMMY_MAP(ArmD3), \ __SMU_DUMMY_MAP(RunDcBtc), \ @@ -135,6 +139,8 @@ __SMU_DUMMY_MAP(PowerUpSdma), \ __SMU_DUMMY_MAP(SetHardMinIspclkByFreq), \ __SMU_DUMMY_MAP(SetHardMinVcn), \ + __SMU_DUMMY_MAP(SetHardMinVcn0), \ + __SMU_DUMMY_MAP(SetHardMinVcn1), \ __SMU_DUMMY_MAP(SetAllowFclkSwitch), \ __SMU_DUMMY_MAP(SetMinVideoGfxclkFreq), \ __SMU_DUMMY_MAP(ActiveProcessNotify), \ @@ -150,6 +156,8 @@ __SMU_DUMMY_MAP(SetPhyclkVoltageByFreq), \ __SMU_DUMMY_MAP(SetDppclkVoltageByFreq), \ __SMU_DUMMY_MAP(SetSoftMinVcn), \ + __SMU_DUMMY_MAP(SetSoftMinVcn0), \ + __SMU_DUMMY_MAP(SetSoftMinVcn1), \ __SMU_DUMMY_MAP(EnablePostCode), \ __SMU_DUMMY_MAP(GetGfxclkFrequency), \ __SMU_DUMMY_MAP(GetFclkFrequency), \ @@ -161,6 +169,8 @@ __SMU_DUMMY_MAP(SetSoftMaxSocclkByFreq), \ __SMU_DUMMY_MAP(SetSoftMaxFclkByFreq), \ __SMU_DUMMY_MAP(SetSoftMaxVcn), \ + __SMU_DUMMY_MAP(SetSoftMaxVcn0), \ + __SMU_DUMMY_MAP(SetSoftMaxVcn1), \ __SMU_DUMMY_MAP(PowerGateMmHub), \ __SMU_DUMMY_MAP(UpdatePmeRestore), \ __SMU_DUMMY_MAP(GpuChangeState), \ @@ -435,4 +445,11 @@ enum smu_feature_mask { SMU_FEATURE_COUNT, }; +/* Message category flags */ +#define SMU_MSG_VF_FLAG (1U << 0) +#define SMU_MSG_RAS_PRI (1U << 1) + +/* Firmware capability flags */ +#define SMU_FW_CAP_RAS_PRI (1U << 0) + #endif diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v11_0.h b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v11_0.h index a0e5ad0381d6..c2ab336bb530 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v11_0.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v11_0.h @@ -237,7 +237,7 @@ int smu_v11_0_set_azalia_d3_pme(struct smu_context *smu); int smu_v11_0_get_max_sustainable_clocks_by_dc(struct smu_context *smu, struct pp_smu_nv_clock_table *max_clocks); -bool smu_v11_0_baco_is_support(struct smu_context *smu); +int smu_v11_0_get_bamaco_support(struct smu_context *smu); enum smu_baco_state smu_v11_0_baco_get_state(struct smu_context *smu); diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v13_0.h b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v13_0.h index fbd57fa1a004..d9700a3f28d2 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v13_0.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v13_0.h @@ -210,7 +210,7 @@ int smu_v13_0_set_azalia_d3_pme(struct smu_context *smu); int smu_v13_0_get_max_sustainable_clocks_by_dc(struct smu_context *smu, struct pp_smu_nv_clock_table *max_clocks); -bool smu_v13_0_baco_is_support(struct smu_context *smu); +int smu_v13_0_get_bamaco_support(struct smu_context *smu); int smu_v13_0_baco_enter(struct smu_context *smu); int smu_v13_0_baco_exit(struct smu_context *smu); diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v14_0.h b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v14_0.h index 3f7463c1c1a9..1fc4557e6fb4 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v14_0.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v14_0.h @@ -27,7 +27,8 @@ #define SMU14_DRIVER_IF_VERSION_INV 0xFFFFFFFF #define SMU14_DRIVER_IF_VERSION_SMU_V14_0_0 0x7 -#define SMU14_DRIVER_IF_VERSION_SMU_V14_0_2 0x1 +#define SMU14_DRIVER_IF_VERSION_SMU_V14_0_1 0x6 +#define SMU14_DRIVER_IF_VERSION_SMU_V14_0_2 0x25 #define FEATURE_MASK(feature) (1ULL << feature) @@ -38,7 +39,8 @@ #define MP1_SRAM 0x03c00004 /* address block */ -#define smnMP1_FIRMWARE_FLAGS 0x3010028 +#define smnMP1_FIRMWARE_FLAGS_14_0_0 0x3010028 +#define smnMP1_FIRMWARE_FLAGS 0x3010024 #define smnMP1_PUB_CTRL 0x3010d10 #define MAX_DPM_LEVELS 16 @@ -159,7 +161,7 @@ int smu_v14_0_register_irq_handler(struct smu_context *smu); int smu_v14_0_baco_set_armd3_sequence(struct smu_context *smu, enum smu_baco_seq baco_seq); -bool smu_v14_0_baco_is_support(struct smu_context *smu); +int smu_v14_0_get_bamaco_support(struct smu_context *smu); enum smu_baco_state smu_v14_0_baco_get_state(struct smu_context *smu); diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v14_0_2_pptable.h b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v14_0_2_pptable.h new file mode 100644 index 000000000000..4a3fde89aed7 --- /dev/null +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v14_0_2_pptable.h @@ -0,0 +1,164 @@ +/* + * Copyright 2023 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ + +#ifndef SMU_14_0_2_PPTABLE_H +#define SMU_14_0_2_PPTABLE_H + + +#pragma pack(push, 1) + +#define SMU_14_0_2_TABLE_FORMAT_REVISION 3 + +// POWERPLAYTABLE::ulPlatformCaps +#define SMU_14_0_2_PP_PLATFORM_CAP_POWERPLAY 0x1 // This cap indicates whether CCC need to show Powerplay page. +#define SMU_14_0_2_PP_PLATFORM_CAP_SBIOSPOWERSOURCE 0x2 // This cap indicates whether power source notificaiton is done by SBIOS instead of OS. +#define SMU_14_0_2_PP_PLATFORM_CAP_HARDWAREDC 0x4 // This cap indicates whether DC mode notificaiton is done by GPIO pin directly. +#define SMU_14_0_2_PP_PLATFORM_CAP_BACO 0x8 // This cap indicates whether board supports the BACO circuitry. +#define SMU_14_0_2_PP_PLATFORM_CAP_MACO 0x10 // This cap indicates whether board supports the MACO circuitry. +#define SMU_14_0_2_PP_PLATFORM_CAP_SHADOWPSTATE 0x20 // This cap indicates whether board supports the Shadow Pstate. +#define SMU_14_0_2_PP_PLATFORM_CAP_LEDSUPPORTED 0x40 // This cap indicates whether board supports the LED. +#define SMU_14_0_2_PP_PLATFORM_CAP_MOBILEOVERDRIVE 0x80 // This cap indicates whether board supports the Mobile Overdrive. + +// SMU_14_0_2_PP_THERMALCONTROLLER - Thermal Controller Type +#define SMU_14_0_2_PP_THERMALCONTROLLER_NONE 0 + +#define SMU_14_0_2_PP_OVERDRIVE_VERSION 0x1 // TODO: FIX OverDrive Version TBD +#define SMU_14_0_2_PP_POWERSAVINGCLOCK_VERSION 0x01 // Power Saving Clock Table Version 1.00 + +enum SMU_14_0_2_OD_SW_FEATURE_CAP +{ + SMU_14_0_2_ODCAP_AUTO_FAN_ACOUSTIC_LIMIT = 0, + SMU_14_0_2_ODCAP_POWER_MODE = 1, + SMU_14_0_2_ODCAP_AUTO_UV_ENGINE = 2, + SMU_14_0_2_ODCAP_AUTO_OC_ENGINE = 3, + SMU_14_0_2_ODCAP_AUTO_OC_MEMORY = 4, + SMU_14_0_2_ODCAP_MEMORY_TIMING_TUNE = 5, + SMU_14_0_2_ODCAP_MANUAL_AC_TIMING = 6, + SMU_14_0_2_ODCAP_AUTO_VF_CURVE_OPTIMIZER = 7, + SMU_14_0_2_ODCAP_AUTO_SOC_UV = 8, + SMU_14_0_2_ODCAP_COUNT = 9, +}; + +enum SMU_14_0_2_OD_SW_FEATURE_ID +{ + SMU_14_0_2_ODFEATURE_AUTO_FAN_ACOUSTIC_LIMIT = 1 << SMU_14_0_2_ODCAP_AUTO_FAN_ACOUSTIC_LIMIT, // Auto Fan Acoustic RPM + SMU_14_0_2_ODFEATURE_POWER_MODE = 1 << SMU_14_0_2_ODCAP_POWER_MODE, // Optimized GPU Power Mode + SMU_14_0_2_ODFEATURE_AUTO_UV_ENGINE = 1 << SMU_14_0_2_ODCAP_AUTO_UV_ENGINE, // Auto Under Volt GFXCLK + SMU_14_0_2_ODFEATURE_AUTO_OC_ENGINE = 1 << SMU_14_0_2_ODCAP_AUTO_OC_ENGINE, // Auto Over Clock GFXCLK + SMU_14_0_2_ODFEATURE_AUTO_OC_MEMORY = 1 << SMU_14_0_2_ODCAP_AUTO_OC_MEMORY, // Auto Over Clock MCLK + SMU_14_0_2_ODFEATURE_MEMORY_TIMING_TUNE = 1 << SMU_14_0_2_ODCAP_MEMORY_TIMING_TUNE, // Auto AC Timing Tuning + SMU_14_0_2_ODFEATURE_MANUAL_AC_TIMING = 1 << SMU_14_0_2_ODCAP_MANUAL_AC_TIMING, // Manual fine grain AC Timing tuning + SMU_14_0_2_ODFEATURE_AUTO_VF_CURVE_OPTIMIZER = 1 << SMU_14_0_2_ODCAP_AUTO_VF_CURVE_OPTIMIZER, // Fine grain auto VF curve tuning + SMU_14_0_2_ODFEATURE_AUTO_SOC_UV = 1 << SMU_14_0_2_ODCAP_AUTO_SOC_UV, // Auto Unver Volt VDDSOC +}; + +#define SMU_14_0_2_MAX_ODFEATURE 32 // Maximum Number of OD Features + +enum SMU_14_0_2_OD_SW_FEATURE_SETTING_ID +{ + SMU_14_0_2_ODSETTING_AUTO_FAN_ACOUSTIC_LIMIT = 0, + SMU_14_0_2_ODSETTING_POWER_MODE = 1, + SMU_14_0_2_ODSETTING_AUTOUVENGINE = 2, + SMU_14_0_2_ODSETTING_AUTOOCENGINE = 3, + SMU_14_0_2_ODSETTING_AUTOOCMEMORY = 4, + SMU_14_0_2_ODSETTING_ACTIMING = 5, + SMU_14_0_2_ODSETTING_MANUAL_AC_TIMING = 6, + SMU_14_0_2_ODSETTING_AUTO_VF_CURVE_OPTIMIZER = 7, + SMU_14_0_2_ODSETTING_AUTO_SOC_UV = 8, + SMU_14_0_2_ODSETTING_COUNT = 9, +}; +#define SMU_14_0_2_MAX_ODSETTING 64 // Maximum Number of ODSettings + +enum SMU_14_0_2_PWRMODE_SETTING +{ + SMU_14_0_2_PMSETTING_POWER_LIMIT_QUIET = 0, + SMU_14_0_2_PMSETTING_POWER_LIMIT_BALANCE, + SMU_14_0_2_PMSETTING_POWER_LIMIT_TURBO, + SMU_14_0_2_PMSETTING_POWER_LIMIT_RAGE, + SMU_14_0_2_PMSETTING_ACOUSTIC_TEMP_QUIET, + SMU_14_0_2_PMSETTING_ACOUSTIC_TEMP_BALANCE, + SMU_14_0_2_PMSETTING_ACOUSTIC_TEMP_TURBO, + SMU_14_0_2_PMSETTING_ACOUSTIC_TEMP_RAGE, + SMU_14_0_2_PMSETTING_ACOUSTIC_TARGET_RPM_QUIET, + SMU_14_0_2_PMSETTING_ACOUSTIC_TARGET_RPM_BALANCE, + SMU_14_0_2_PMSETTING_ACOUSTIC_TARGET_RPM_TURBO, + SMU_14_0_2_PMSETTING_ACOUSTIC_TARGET_RPM_RAGE, + SMU_14_0_2_PMSETTING_ACOUSTIC_LIMIT_RPM_QUIET, + SMU_14_0_2_PMSETTING_ACOUSTIC_LIMIT_RPM_BALANCE, + SMU_14_0_2_PMSETTING_ACOUSTIC_LIMIT_RPM_TURBO, + SMU_14_0_2_PMSETTING_ACOUSTIC_LIMIT_RPM_RAGE, +}; +#define SMU_14_0_2_MAX_PMSETTING 32 // Maximum Number of PowerMode Settings + +enum SMU_14_0_2_overdrive_table_id +{ + SMU_14_0_2_OVERDRIVE_TABLE_BASIC = 0, + SMU_14_0_2_OVERDRIVE_TABLE_ADVANCED = 1, + SMU_14_0_2_OVERDRIVE_TABLE_COUNT = 2, +}; + +struct smu_14_0_2_overdrive_table +{ + uint8_t revision; // Revision = SMU_14_0_2_PP_OVERDRIVE_VERSION + uint8_t reserve[3]; // Zero filled field reserved for future use + uint8_t cap[SMU_14_0_2_OVERDRIVE_TABLE_COUNT][SMU_14_0_2_MAX_ODFEATURE]; // OD feature support flags + int32_t max[SMU_14_0_2_OVERDRIVE_TABLE_COUNT][SMU_14_0_2_MAX_ODSETTING]; // maximum settings + int32_t min[SMU_14_0_2_OVERDRIVE_TABLE_COUNT][SMU_14_0_2_MAX_ODSETTING]; // minimum settings + int16_t pm_setting[SMU_14_0_2_MAX_PMSETTING]; // Optimized power mode feature settings +}; + +struct smu_14_0_2_powerplay_table +{ + struct atom_common_table_header header; // header.format_revision = 3 (HAS TO MATCH SMU_14_0_2_TABLE_FORMAT_REVISION), header.content_revision = ? structuresize is calculated by PPGen. + uint8_t table_revision; // PPGen use only: table_revision = 3 + uint8_t padding; // Padding 1 byte to align table_size offset to 6 bytes (pmfw_start_offset, for PMFW to know the starting offset of PPTable_t). + uint16_t pmfw_pptable_start_offset; // The start offset of the pmfw portion. i.e. start of PPTable_t (start of SkuTable_t) + uint16_t pmfw_pptable_size; // The total size of pmfw_pptable, i.e PPTable_t. + uint16_t pmfw_pfe_table_start_offset; // The start offset of the PFE_Settings_t within pmfw_pptable. + uint16_t pmfw_pfe_table_size; // The size of PFE_Settings_t. + uint16_t pmfw_board_table_start_offset; // The start offset of the BoardTable_t within pmfw_pptable. + uint16_t pmfw_board_table_size; // The size of BoardTable_t. + uint16_t pmfw_custom_sku_table_start_offset; // The start offset of the CustomSkuTable_t within pmfw_pptable. + uint16_t pmfw_custom_sku_table_size; // The size of the CustomSkuTable_t. + uint32_t golden_pp_id; // PPGen use only: PP Table ID on the Golden Data Base + uint32_t golden_revision; // PPGen use only: PP Table Revision on the Golden Data Base + uint16_t format_id; // PPGen use only: PPTable for different ASICs. + uint32_t platform_caps; // POWERPLAYTABLE::ulPlatformCaps + + uint8_t thermal_controller_type; // one of smu_14_0_2_PP_THERMALCONTROLLER + + uint16_t small_power_limit1; + uint16_t small_power_limit2; + uint16_t boost_power_limit; // For Gemini Board, when the slave adapter is in BACO mode, the master adapter will use this boost power limit instead of the default power limit to boost the power limit. + uint16_t software_shutdown_temp; + + uint8_t reserve[143]; // Zero filled field reserved for future use + + struct smu_14_0_2_overdrive_table overdrive_table; + + PPTable_t smc_pptable; // PPTable_t in driver_if.h -- as requested by PMFW, this offset should start at a 32-byte boundary, and the table_size above should remain at offset=6 bytes +}; + +#pragma pack(pop) + +#endif diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu11/arcturus_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu11/arcturus_ppt.c index 0c2d04f978ac..6d334a2aff67 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu11/arcturus_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/arcturus_ppt.c @@ -2387,7 +2387,7 @@ static const struct pptable_funcs arcturus_ppt_funcs = { .register_irq_handler = smu_v11_0_register_irq_handler, .set_azalia_d3_pme = smu_v11_0_set_azalia_d3_pme, .get_max_sustainable_clocks_by_dc = smu_v11_0_get_max_sustainable_clocks_by_dc, - .baco_is_support = smu_v11_0_baco_is_support, + .get_bamaco_support = smu_v11_0_get_bamaco_support, .baco_enter = smu_v11_0_baco_enter, .baco_exit = smu_v11_0_baco_exit, .get_dpm_ultimate_freq = smu_v11_0_get_dpm_ultimate_freq, diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu11/navi10_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu11/navi10_ppt.c index 836b1df79928..5a68d365967f 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu11/navi10_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/navi10_ppt.c @@ -3538,7 +3538,7 @@ static const struct pptable_funcs navi10_ppt_funcs = { .register_irq_handler = smu_v11_0_register_irq_handler, .set_azalia_d3_pme = smu_v11_0_set_azalia_d3_pme, .get_max_sustainable_clocks_by_dc = smu_v11_0_get_max_sustainable_clocks_by_dc, - .baco_is_support = smu_v11_0_baco_is_support, + .get_bamaco_support = smu_v11_0_get_bamaco_support, .baco_enter = navi10_baco_enter, .baco_exit = navi10_baco_exit, .get_dpm_ultimate_freq = smu_v11_0_get_dpm_ultimate_freq, diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c index 1f18b61884f3..e426f457a017 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c @@ -4431,7 +4431,7 @@ static const struct pptable_funcs sienna_cichlid_ppt_funcs = { .register_irq_handler = smu_v11_0_register_irq_handler, .set_azalia_d3_pme = smu_v11_0_set_azalia_d3_pme, .get_max_sustainable_clocks_by_dc = smu_v11_0_get_max_sustainable_clocks_by_dc, - .baco_is_support = smu_v11_0_baco_is_support, + .get_bamaco_support = smu_v11_0_get_bamaco_support, .baco_enter = sienna_cichlid_baco_enter, .baco_exit = sienna_cichlid_baco_exit, .mode1_reset_is_support = sienna_cichlid_is_mode1_reset_supported, diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu11/smu_v11_0.c b/drivers/gpu/drm/amd/pm/swsmu/smu11/smu_v11_0.c index f6545093bfc1..9d5ab2ea643a 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu11/smu_v11_0.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/smu_v11_0.c @@ -93,7 +93,7 @@ static void smu_v11_0_poll_baco_exit(struct smu_context *smu) int smu_v11_0_init_microcode(struct smu_context *smu) { struct amdgpu_device *adev = smu->adev; - char ucode_prefix[30]; + char ucode_prefix[25]; char fw_name[SMU_FW_NAME_LEN]; int err = 0; const struct smc_firmware_header_v1_0 *hdr; @@ -1557,23 +1557,27 @@ int smu_v11_0_baco_set_armd3_sequence(struct smu_context *smu, return smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_ArmD3, baco_seq, NULL); } -bool smu_v11_0_baco_is_support(struct smu_context *smu) +int smu_v11_0_get_bamaco_support(struct smu_context *smu) { struct smu_baco_context *smu_baco = &smu->smu_baco; + int bamaco_support = 0; if (amdgpu_sriov_vf(smu->adev) || !smu_baco->platform_support) - return false; + return 0; + + if (smu_baco->maco_support) + bamaco_support |= MACO_SUPPORT; /* return true if ASIC is in BACO state already */ if (smu_v11_0_baco_get_state(smu) == SMU_BACO_STATE_ENTER) - return true; + return bamaco_support |= BACO_SUPPORT; /* Arcturus does not support this bit mask */ if (smu_cmn_feature_is_supported(smu, SMU_FEATURE_BACO_BIT) && !smu_cmn_feature_is_enabled(smu, SMU_FEATURE_BACO_BIT)) - return false; + return 0; - return true; + return (bamaco_support |= BACO_SUPPORT); } enum smu_baco_state smu_v11_0_baco_get_state(struct smu_context *smu) @@ -1603,7 +1607,7 @@ int smu_v11_0_baco_set_state(struct smu_context *smu, enum smu_baco_state state) case IP_VERSION(11, 0, 11): case IP_VERSION(11, 0, 12): case IP_VERSION(11, 0, 13): - if (amdgpu_runtime_pm == 2) + if (adev->pm.rpm_mode == AMDGPU_RUNPM_BAMACO) ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_EnterBaco, D3HOT_BAMACO_SEQUENCE, diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu11/vangogh_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu11/vangogh_ppt.c index da1f43999d09..379e44eb0019 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu11/vangogh_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/vangogh_ppt.c @@ -301,7 +301,7 @@ static int vangogh_get_legacy_smu_metrics_data(struct smu_context *smu, *value = metrics->GfxActivity / 100; break; case METRICS_AVERAGE_VCNACTIVITY: - *value = metrics->UvdActivity; + *value = metrics->UvdActivity / 100; break; case METRICS_AVERAGE_SOCKETPOWER: *value = (metrics->CurrentSocketPower << 8) / @@ -1507,6 +1507,12 @@ static int vangogh_read_sensor(struct smu_context *smu, (uint32_t *)data); *size = 4; break; + case AMDGPU_PP_SENSOR_VCN_LOAD: + ret = vangogh_common_get_smu_metrics_data(smu, + METRICS_AVERAGE_VCNACTIVITY, + (uint32_t *)data); + *size = 4; + break; case AMDGPU_PP_SENSOR_GPU_AVG_POWER: ret = vangogh_common_get_smu_metrics_data(smu, METRICS_AVERAGE_SOCKETPOWER, diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/aldebaran_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/aldebaran_ppt.c index f41ac6465f2a..ce941fbb9cfb 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu13/aldebaran_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/aldebaran_ppt.c @@ -759,8 +759,11 @@ static int aldebaran_emit_clk_levels(struct smu_context *smu, switch (type) { case SMU_OD_SCLK: - *offset += sysfs_emit_at(buf, *offset, "%s:\n", "GFXCLK"); - fallthrough; + *offset += sysfs_emit_at(buf, *offset, "%s:\n", "OD_SCLK"); + *offset += sysfs_emit_at(buf, *offset, "0: %uMhz\n1: %uMhz\n", + pstate_table->gfxclk_pstate.curr.min, + pstate_table->gfxclk_pstate.curr.max); + return 0; case SMU_SCLK: ret = aldebaran_get_current_clk_freq_by_table(smu, SMU_GFXCLK, &cur_value); if (ret) { @@ -788,8 +791,11 @@ static int aldebaran_emit_clk_levels(struct smu_context *smu, break; case SMU_OD_MCLK: - *offset += sysfs_emit_at(buf, *offset, "%s:\n", "MCLK"); - fallthrough; + *offset += sysfs_emit_at(buf, *offset, "%s:\n", "OD_MCLK"); + *offset += sysfs_emit_at(buf, *offset, "0: %uMhz\n1: %uMhz\n", + pstate_table->uclk_pstate.curr.min, + pstate_table->uclk_pstate.curr.max); + return 0; case SMU_MCLK: ret = aldebaran_get_current_clk_freq_by_table(smu, SMU_UCLK, &cur_value); if (ret) { @@ -850,7 +856,6 @@ static int aldebaran_emit_clk_levels(struct smu_context *smu, } switch (type) { - case SMU_OD_SCLK: case SMU_SCLK: for (i = 0; i < display_levels; i++) { clock_mhz = freq_values[i]; @@ -863,7 +868,6 @@ static int aldebaran_emit_clk_levels(struct smu_context *smu, } break; - case SMU_OD_MCLK: case SMU_MCLK: case SMU_SOCCLK: case SMU_FCLK: @@ -1581,11 +1585,11 @@ out: adev->unique_id = ((uint64_t)upper32 << 32) | lower32; } -static bool aldebaran_is_baco_supported(struct smu_context *smu) +static int aldebaran_get_bamaco_support(struct smu_context *smu) { /* aldebaran is not support baco */ - return false; + return 0; } static int aldebaran_set_df_cstate(struct smu_context *smu, @@ -2059,7 +2063,7 @@ static const struct pptable_funcs aldebaran_ppt_funcs = { .register_irq_handler = smu_v13_0_register_irq_handler, .set_azalia_d3_pme = smu_v13_0_set_azalia_d3_pme, .get_max_sustainable_clocks_by_dc = smu_v13_0_get_max_sustainable_clocks_by_dc, - .baco_is_support = aldebaran_is_baco_supported, + .get_bamaco_support = aldebaran_get_bamaco_support, .get_dpm_ultimate_freq = smu_v13_0_get_dpm_ultimate_freq, .set_soft_freq_limited_range = aldebaran_set_soft_freq_limited_range, .od_edit_dpm_table = aldebaran_usr_edit_dpm_table, diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0.c index 48170bb5112e..a8d34adc7d3f 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0.c @@ -93,7 +93,7 @@ int smu_v13_0_init_microcode(struct smu_context *smu) { struct amdgpu_device *adev = smu->adev; char fw_name[30]; - char ucode_prefix[30]; + char ucode_prefix[15]; int err = 0; const struct smc_firmware_header_v1_0 *hdr; const struct common_firmware_header *header; @@ -2247,7 +2247,7 @@ static int smu_v13_0_baco_set_state(struct smu_context *smu, if (state == SMU_BACO_STATE_ENTER) { ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_EnterBaco, - (smu_baco->maco_support && amdgpu_runtime_pm != 1) ? + (adev->pm.rpm_mode == AMDGPU_RUNPM_BAMACO) ? BACO_SEQ_BAMACO : BACO_SEQ_BACO, NULL); } else { @@ -2268,33 +2268,36 @@ static int smu_v13_0_baco_set_state(struct smu_context *smu, return ret; } -bool smu_v13_0_baco_is_support(struct smu_context *smu) +int smu_v13_0_get_bamaco_support(struct smu_context *smu) { struct smu_baco_context *smu_baco = &smu->smu_baco; + int bamaco_support = 0; if (amdgpu_sriov_vf(smu->adev) || !smu_baco->platform_support) - return false; + return 0; + + if (smu_baco->maco_support) + bamaco_support |= MACO_SUPPORT; /* return true if ASIC is in BACO state already */ if (smu_v13_0_baco_get_state(smu) == SMU_BACO_STATE_ENTER) - return true; + return bamaco_support |= BACO_SUPPORT; if (smu_cmn_feature_is_supported(smu, SMU_FEATURE_BACO_BIT) && !smu_cmn_feature_is_enabled(smu, SMU_FEATURE_BACO_BIT)) - return false; + return 0; - return true; + return (bamaco_support |= BACO_SUPPORT); } int smu_v13_0_baco_enter(struct smu_context *smu) { - struct smu_baco_context *smu_baco = &smu->smu_baco; struct amdgpu_device *adev = smu->adev; int ret; if (adev->in_runpm && smu_cmn_is_audio_func_enabled(adev)) { return smu_v13_0_baco_set_armd3_sequence(smu, - (smu_baco->maco_support && amdgpu_runtime_pm != 1) ? + (adev->pm.rpm_mode == AMDGPU_RUNPM_BAMACO) ? BACO_SEQ_BAMACO : BACO_SEQ_BACO); } else { ret = smu_v13_0_baco_set_state(smu, SMU_BACO_STATE_ENTER); diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c index 9c03296f92cd..1e09d5f2d82f 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c @@ -2751,7 +2751,13 @@ static int smu_v13_0_0_set_mp1_state(struct smu_context *smu, switch (mp1_state) { case PP_MP1_STATE_UNLOAD: - ret = smu_cmn_set_mp1_state(smu, mp1_state); + ret = smu_cmn_send_smc_msg_with_param(smu, + SMU_MSG_PrepareMp1ForUnload, + 0x55, NULL); + + if (!ret && smu->smu_baco.state == SMU_BACO_STATE_EXIT) + ret = smu_v13_0_disable_pmfw_state(smu); + break; default: /* Ignore others */ @@ -3070,7 +3076,7 @@ static const struct pptable_funcs smu_v13_0_0_ppt_funcs = { .set_tool_table_location = smu_v13_0_set_tool_table_location, .deep_sleep_control = smu_v13_0_deep_sleep_control, .gfx_ulv_control = smu_v13_0_gfx_ulv_control, - .baco_is_support = smu_v13_0_baco_is_support, + .get_bamaco_support = smu_v13_0_get_bamaco_support, .baco_enter = smu_v13_0_baco_enter, .baco_exit = smu_v13_0_baco_exit, .mode1_reset_is_support = smu_v13_0_0_is_mode1_reset_supported, diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_4_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_4_ppt.c index bb98156b2fa1..bc241b593db1 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_4_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_4_ppt.c @@ -226,8 +226,18 @@ static int smu_v13_0_4_system_features_control(struct smu_context *smu, bool en) struct amdgpu_device *adev = smu->adev; int ret = 0; - if (!en && !adev->in_s0ix) + if (!en && adev->in_s4) { + /* Adds a GFX reset as workaround just before sending the + * MP1_UNLOAD message to prevent GC/RLC/PMFW from entering + * an invalid state. + */ + ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_GfxDeviceDriverReset, + SMU_RESET_MODE_2, NULL); + if (ret) + return ret; + ret = smu_cmn_send_smc_msg(smu, SMU_MSG_PrepareMp1ForUnload, NULL); + } return ret; } @@ -318,7 +328,7 @@ static int smu_v13_0_4_get_smu_metrics_data(struct smu_context *smu, *value = metrics->GfxActivity / 100; break; case METRICS_AVERAGE_VCNACTIVITY: - *value = metrics->UvdActivity; + *value = metrics->UvdActivity / 100; break; case METRICS_AVERAGE_SOCKETPOWER: *value = (metrics->AverageSocketPower << 8) / 1000; @@ -572,6 +582,12 @@ static int smu_v13_0_4_read_sensor(struct smu_context *smu, (uint32_t *)data); *size = 4; break; + case AMDGPU_PP_SENSOR_VCN_LOAD: + ret = smu_v13_0_4_get_smu_metrics_data(smu, + METRICS_AVERAGE_VCNACTIVITY, + (uint32_t *)data); + *size = 4; + break; case AMDGPU_PP_SENSOR_GPU_AVG_POWER: ret = smu_v13_0_4_get_smu_metrics_data(smu, METRICS_AVERAGE_SOCKETPOWER, diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_5_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_5_ppt.c index 0dce672ac1b9..218f209c3775 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_5_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_5_ppt.c @@ -286,7 +286,7 @@ static int smu_v13_0_5_get_smu_metrics_data(struct smu_context *smu, *value = metrics->GfxActivity / 100; break; case METRICS_AVERAGE_VCNACTIVITY: - *value = metrics->UvdActivity; + *value = metrics->UvdActivity / 100; break; case METRICS_CURR_SOCKETPOWER: *value = (metrics->CurrentSocketPower << 8) / 1000; @@ -332,6 +332,12 @@ static int smu_v13_0_5_read_sensor(struct smu_context *smu, (uint32_t *)data); *size = 4; break; + case AMDGPU_PP_SENSOR_VCN_LOAD: + ret = smu_v13_0_5_get_smu_metrics_data(smu, + METRICS_AVERAGE_VCNACTIVITY, + (uint32_t *)data); + *size = 4; + break; case AMDGPU_PP_SENSOR_GPU_INPUT_POWER: ret = smu_v13_0_5_get_smu_metrics_data(smu, METRICS_CURR_SOCKETPOWER, diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_6_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_6_ppt.c index 3957af057d54..4d3eca2fc3f1 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_6_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_6_ppt.c @@ -138,13 +138,13 @@ static const struct cmn2asic_msg_mapping smu_v13_0_6_message_map[SMU_MSG_MAX_COU MSG_MAP(SetToolsDramAddrHigh, PPSMC_MSG_SetToolsDramAddrHigh, 0), MSG_MAP(SetToolsDramAddrLow, PPSMC_MSG_SetToolsDramAddrLow, 0), MSG_MAP(SetSoftMinByFreq, PPSMC_MSG_SetSoftMinByFreq, 0), - MSG_MAP(SetSoftMaxByFreq, PPSMC_MSG_SetSoftMaxByFreq, 0), + MSG_MAP(SetSoftMaxByFreq, PPSMC_MSG_SetSoftMaxByFreq, 1), MSG_MAP(GetMinDpmFreq, PPSMC_MSG_GetMinDpmFreq, 1), MSG_MAP(GetMaxDpmFreq, PPSMC_MSG_GetMaxDpmFreq, 1), MSG_MAP(GetDpmFreqByIndex, PPSMC_MSG_GetDpmFreqByIndex, 1), MSG_MAP(SetPptLimit, PPSMC_MSG_SetPptLimit, 0), MSG_MAP(GetPptLimit, PPSMC_MSG_GetPptLimit, 1), - MSG_MAP(GfxDeviceDriverReset, PPSMC_MSG_GfxDriverReset, 0), + MSG_MAP(GfxDeviceDriverReset, PPSMC_MSG_GfxDriverReset, SMU_MSG_RAS_PRI), MSG_MAP(DramLogSetDramAddrHigh, PPSMC_MSG_DramLogSetDramAddrHigh, 0), MSG_MAP(DramLogSetDramAddrLow, PPSMC_MSG_DramLogSetDramAddrLow, 0), MSG_MAP(DramLogSetDramSize, PPSMC_MSG_DramLogSetDramSize, 0), @@ -167,10 +167,10 @@ static const struct cmn2asic_msg_mapping smu_v13_0_6_message_map[SMU_MSG_MAX_COU MSG_MAP(GetCTFLimit, PPSMC_MSG_GetCTFLimit, 0), MSG_MAP(GetThermalLimit, PPSMC_MSG_ReadThrottlerLimit, 0), MSG_MAP(ClearMcaOnRead, PPSMC_MSG_ClearMcaOnRead, 0), - MSG_MAP(QueryValidMcaCount, PPSMC_MSG_QueryValidMcaCount, 0), - MSG_MAP(QueryValidMcaCeCount, PPSMC_MSG_QueryValidMcaCeCount, 0), - MSG_MAP(McaBankDumpDW, PPSMC_MSG_McaBankDumpDW, 0), - MSG_MAP(McaBankCeDumpDW, PPSMC_MSG_McaBankCeDumpDW, 0), + MSG_MAP(QueryValidMcaCount, PPSMC_MSG_QueryValidMcaCount, SMU_MSG_RAS_PRI), + MSG_MAP(QueryValidMcaCeCount, PPSMC_MSG_QueryValidMcaCeCount, SMU_MSG_RAS_PRI), + MSG_MAP(McaBankDumpDW, PPSMC_MSG_McaBankDumpDW, SMU_MSG_RAS_PRI), + MSG_MAP(McaBankCeDumpDW, PPSMC_MSG_McaBankCeDumpDW, SMU_MSG_RAS_PRI), MSG_MAP(SelectPLPDMode, PPSMC_MSG_SelectPLPDMode, 0), MSG_MAP(RmaDueToBadPageThreshold, PPSMC_MSG_RmaDueToBadPageThreshold, 0), }; @@ -1010,8 +1010,11 @@ static int smu_v13_0_6_print_clk_levels(struct smu_context *smu, switch (type) { case SMU_OD_SCLK: - size += sysfs_emit_at(buf, size, "%s:\n", "GFXCLK"); - fallthrough; + size += sysfs_emit_at(buf, size, "%s:\n", "OD_SCLK"); + size += sysfs_emit_at(buf, size, "0: %uMhz\n1: %uMhz\n", + pstate_table->gfxclk_pstate.curr.min, + pstate_table->gfxclk_pstate.curr.max); + break; case SMU_SCLK: ret = smu_v13_0_6_get_current_clk_freq_by_table(smu, SMU_GFXCLK, &now); @@ -1052,8 +1055,11 @@ static int smu_v13_0_6_print_clk_levels(struct smu_context *smu, break; case SMU_OD_MCLK: - size += sysfs_emit_at(buf, size, "%s:\n", "MCLK"); - fallthrough; + size += sysfs_emit_at(buf, size, "%s:\n", "OD_MCLK"); + size += sysfs_emit_at(buf, size, "0: %uMhz\n1: %uMhz\n", + pstate_table->uclk_pstate.curr.min, + pstate_table->uclk_pstate.curr.max); + break; case SMU_MCLK: ret = smu_v13_0_6_get_current_clk_freq_by_table(smu, SMU_UCLK, &now); @@ -1670,6 +1676,11 @@ static int smu_v13_0_6_set_soft_freq_limited_range(struct smu_context *smu, if (clk_type == SMU_UCLK) { if (max == pstate_table->uclk_pstate.curr.max) return 0; + /* For VF, only allowed in FW versions 85.102 or greater */ + if (amdgpu_sriov_vf(adev) && + ((smu->smc_fw_version < 0x556600) || + (adev->flags & AMD_IS_APU))) + return -EOPNOTSUPP; /* Only max clock limiting is allowed for UCLK */ ret = smu_v13_0_set_soft_freq_limited_range( smu, SMU_UCLK, 0, max); @@ -2077,11 +2088,11 @@ static void smu_v13_0_6_get_unique_id(struct smu_context *smu) adev->unique_id = pptable->PublicSerialNumber_AID; } -static bool smu_v13_0_6_is_baco_supported(struct smu_context *smu) +static int smu_v13_0_6_get_bamaco_support(struct smu_context *smu) { /* smu_13_0_6 does not support baco */ - return false; + return 0; } static const char *const throttling_logging_label[] = { @@ -2228,7 +2239,15 @@ static ssize_t smu_v13_0_6_get_gpu_metrics(struct smu_context *smu, void **table gpu_metrics->gfxclk_lock_status = GET_METRIC_FIELD(GfxLockXCDMak) >> GET_INST(GC, 0); if (!(adev->flags & AMD_IS_APU)) { - if (!amdgpu_sriov_vf(adev)) { + /*Check smu version, PCIE link speed and width will be reported from pmfw metric + * table for both pf & one vf for smu version 85.99.0 or higher else report only + * for pf from registers + */ + if (smu->smc_fw_version >= 0x556300) { + gpu_metrics->pcie_link_width = metrics_x->PCIeLinkWidth; + gpu_metrics->pcie_link_speed = + pcie_gen_to_speed(metrics_x->PCIeLinkSpeed); + } else if (!amdgpu_sriov_vf(adev)) { link_width_level = smu_v13_0_6_get_current_pcie_link_width_level(smu); if (link_width_level > MAX_LINK_WIDTH) link_width_level = 0; @@ -2238,6 +2257,7 @@ static ssize_t smu_v13_0_6_get_gpu_metrics(struct smu_context *smu, void **table gpu_metrics->pcie_link_speed = smu_v13_0_6_get_current_pcie_link_speed(smu); } + gpu_metrics->pcie_bandwidth_acc = SMUQ10_ROUND(metrics_x->PcieBandwidthAcc[0]); gpu_metrics->pcie_bandwidth_inst = @@ -2294,6 +2314,17 @@ static ssize_t smu_v13_0_6_get_gpu_metrics(struct smu_context *smu, void **table return sizeof(*gpu_metrics); } +static void smu_v13_0_6_restore_pci_config(struct smu_context *smu) +{ + struct amdgpu_device *adev = smu->adev; + int i; + + for (i = 0; i < 16; i++) + pci_write_config_dword(adev->pdev, i * 4, + adev->pdev->saved_config_space[i]); + pci_restore_msi_state(adev->pdev); +} + static int smu_v13_0_6_mode2_reset(struct smu_context *smu) { int ret = 0, index; @@ -2315,6 +2346,20 @@ static int smu_v13_0_6_mode2_reset(struct smu_context *smu) /* Restore the config space saved during init */ amdgpu_device_load_pci_state(adev->pdev); + /* Certain platforms have switches which assign virtual BAR values to + * devices. OS uses the virtual BAR values and device behind the switch + * is assgined another BAR value. When device's config space registers + * are queried, switch returns the virtual BAR values. When mode-2 reset + * is performed, switch is unaware of it, and will continue to return + * the same virtual values to the OS.This affects + * pci_restore_config_space() API as it doesn't write the value saved if + * the current value read from config space is the same as what is + * saved. As a workaround, make sure the config space is restored + * always. + */ + if (!(adev->flags & AMD_IS_APU)) + smu_v13_0_6_restore_pci_config(smu); + dev_dbg(smu->adev->dev, "wait for reset ack\n"); do { ret = smu_cmn_wait_for_response(smu); @@ -2671,6 +2716,11 @@ static int mca_umc_mca_get_err_count(const struct mca_ras_info *mca_ras, struct umc_v12_0_is_correctable_error(adev, status0)) *count = (ext_error_code == 0) ? odecc_err_cnt : 1; + amdgpu_umc_update_ecc_status(adev, + entry->regs[MCA_REG_IDX_STATUS], + entry->regs[MCA_REG_IDX_IPID], + entry->regs[MCA_REG_IDX_ADDR]); + return 0; } @@ -2684,7 +2734,8 @@ static int mca_pcs_xgmi_mca_get_err_count(const struct mca_ras_info *mca_ras, st ext_error_code = MCA_REG__STATUS__ERRORCODEEXT(entry->regs[MCA_REG_IDX_STATUS]); err_cnt = MCA_REG__MISC0__ERRCNT(entry->regs[MCA_REG_IDX_MISC0]); - if (type == AMDGPU_MCA_ERROR_TYPE_UE && ext_error_code == 0) + if (type == AMDGPU_MCA_ERROR_TYPE_UE && + (ext_error_code == 0 || ext_error_code == 9)) *count = err_cnt; else if (type == AMDGPU_MCA_ERROR_TYPE_CE && ext_error_code == 6) *count = err_cnt; @@ -2975,7 +3026,7 @@ static int aca_smu_set_debug_mode(struct amdgpu_device *adev, bool enable) return smu_v13_0_6_mca_set_debug_mode(smu, enable); } -static int smu_v13_0_6_get_valid_aca_count(struct smu_context *smu, enum aca_error_type type, u32 *count) +static int smu_v13_0_6_get_valid_aca_count(struct smu_context *smu, enum aca_smu_type type, u32 *count) { uint32_t msg; int ret; @@ -2984,10 +3035,10 @@ static int smu_v13_0_6_get_valid_aca_count(struct smu_context *smu, enum aca_err return -EINVAL; switch (type) { - case ACA_ERROR_TYPE_UE: + case ACA_SMU_TYPE_UE: msg = SMU_MSG_QueryValidMcaCount; break; - case ACA_ERROR_TYPE_CE: + case ACA_SMU_TYPE_CE: msg = SMU_MSG_QueryValidMcaCeCount; break; default: @@ -3004,14 +3055,14 @@ static int smu_v13_0_6_get_valid_aca_count(struct smu_context *smu, enum aca_err } static int aca_smu_get_valid_aca_count(struct amdgpu_device *adev, - enum aca_error_type type, u32 *count) + enum aca_smu_type type, u32 *count) { struct smu_context *smu = adev->powerplay.pp_handle; int ret; switch (type) { - case ACA_ERROR_TYPE_UE: - case ACA_ERROR_TYPE_CE: + case ACA_SMU_TYPE_UE: + case ACA_SMU_TYPE_CE: ret = smu_v13_0_6_get_valid_aca_count(smu, type, count); break; default: @@ -3022,16 +3073,16 @@ static int aca_smu_get_valid_aca_count(struct amdgpu_device *adev, return ret; } -static int __smu_v13_0_6_aca_bank_dump(struct smu_context *smu, enum aca_error_type type, +static int __smu_v13_0_6_aca_bank_dump(struct smu_context *smu, enum aca_smu_type type, int idx, int offset, u32 *val) { uint32_t msg, param; switch (type) { - case ACA_ERROR_TYPE_UE: + case ACA_SMU_TYPE_UE: msg = SMU_MSG_McaBankDumpDW; break; - case ACA_ERROR_TYPE_CE: + case ACA_SMU_TYPE_CE: msg = SMU_MSG_McaBankCeDumpDW; break; default: @@ -3043,7 +3094,7 @@ static int __smu_v13_0_6_aca_bank_dump(struct smu_context *smu, enum aca_error_t return smu_cmn_send_smc_msg_with_param(smu, msg, param, (uint32_t *)val); } -static int smu_v13_0_6_aca_bank_dump(struct smu_context *smu, enum aca_error_type type, +static int smu_v13_0_6_aca_bank_dump(struct smu_context *smu, enum aca_smu_type type, int idx, int offset, u32 *val, int count) { int ret, i; @@ -3060,7 +3111,7 @@ static int smu_v13_0_6_aca_bank_dump(struct smu_context *smu, enum aca_error_typ return 0; } -static int aca_bank_read_reg(struct amdgpu_device *adev, enum aca_error_type type, +static int aca_bank_read_reg(struct amdgpu_device *adev, enum aca_smu_type type, int idx, int reg_idx, u64 *val) { struct smu_context *smu = adev->powerplay.pp_handle; @@ -3077,13 +3128,13 @@ static int aca_bank_read_reg(struct amdgpu_device *adev, enum aca_error_type typ *val = (u64)data[1] << 32 | data[0]; dev_dbg(adev->dev, "mca read bank reg: type:%s, index: %d, reg_idx: %d, val: 0x%016llx\n", - type == ACA_ERROR_TYPE_UE ? "UE" : "CE", idx, reg_idx, *val); + type == ACA_SMU_TYPE_UE ? "UE" : "CE", idx, reg_idx, *val); return 0; } static int aca_smu_get_valid_aca_bank(struct amdgpu_device *adev, - enum aca_error_type type, int idx, struct aca_bank *bank) + enum aca_smu_type type, int idx, struct aca_bank *bank) { int i, ret, count; @@ -3097,12 +3148,25 @@ static int aca_smu_get_valid_aca_bank(struct amdgpu_device *adev, return 0; } +static int aca_smu_parse_error_code(struct amdgpu_device *adev, struct aca_bank *bank) +{ + int error_code; + + if (!(adev->flags & AMD_IS_APU) && adev->pm.fw_version >= 0x00555600) + error_code = ACA_REG__SYND__ERRORINFORMATION(bank->regs[ACA_REG_IDX_SYND]); + else + error_code = ACA_REG__STATUS__ERRORCODE(bank->regs[ACA_REG_IDX_STATUS]); + + return error_code & 0xff; +} + static const struct aca_smu_funcs smu_v13_0_6_aca_smu_funcs = { .max_ue_bank_count = 12, .max_ce_bank_count = 12, .set_debug_mode = aca_smu_set_debug_mode, .get_valid_aca_count = aca_smu_get_valid_aca_count, .get_valid_aca_bank = aca_smu_get_valid_aca_bank, + .parse_error_code = aca_smu_parse_error_code, }; static int smu_v13_0_6_select_xgmi_plpd_policy(struct smu_context *smu, @@ -3179,7 +3243,7 @@ static const struct pptable_funcs smu_v13_0_6_ppt_funcs = { .enable_thermal_alert = smu_v13_0_enable_thermal_alert, .disable_thermal_alert = smu_v13_0_disable_thermal_alert, .setup_pptable = smu_v13_0_6_setup_pptable, - .baco_is_support = smu_v13_0_6_is_baco_supported, + .get_bamaco_support = smu_v13_0_6_get_bamaco_support, .get_dpm_ultimate_freq = smu_v13_0_6_get_dpm_ultimate_freq, .set_soft_freq_limited_range = smu_v13_0_6_set_soft_freq_limited_range, .od_edit_dpm_table = smu_v13_0_6_usr_edit_dpm_table, @@ -3208,6 +3272,7 @@ void smu_v13_0_6_set_ppt_funcs(struct smu_context *smu) smu->feature_map = smu_v13_0_6_feature_mask_map; smu->table_map = smu_v13_0_6_table_map; smu->smc_driver_if_version = SMU13_0_6_DRIVER_IF_VERSION; + smu->smc_fw_caps |= SMU_FW_CAP_RAS_PRI; smu_v13_0_set_smu_mailbox_registers(smu); amdgpu_mca_smu_init_funcs(smu->adev, &smu_v13_0_6_mca_smu_funcs); amdgpu_aca_set_smu_funcs(smu->adev, &smu_v13_0_6_aca_smu_funcs); diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c index 7318964f1f14..e996a0a4d33e 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c @@ -2650,7 +2650,7 @@ static const struct pptable_funcs smu_v13_0_7_ppt_funcs = { .set_tool_table_location = smu_v13_0_set_tool_table_location, .get_pp_feature_mask = smu_cmn_get_pp_feature_mask, .set_pp_feature_mask = smu_cmn_set_pp_feature_mask, - .baco_is_support = smu_v13_0_baco_is_support, + .get_bamaco_support = smu_v13_0_get_bamaco_support, .baco_enter = smu_v13_0_baco_enter, .baco_exit = smu_v13_0_baco_exit, .mode1_reset_is_support = smu_v13_0_7_is_mode1_reset_supported, diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/yellow_carp_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/yellow_carp_ppt.c index 2d1736234b4a..d8bcf765a803 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu13/yellow_carp_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/yellow_carp_ppt.c @@ -363,7 +363,7 @@ static int yellow_carp_get_smu_metrics_data(struct smu_context *smu, *value = metrics->GfxActivity / 100; break; case METRICS_AVERAGE_VCNACTIVITY: - *value = metrics->UvdActivity; + *value = metrics->UvdActivity / 100; break; case METRICS_CURR_SOCKETPOWER: *value = (metrics->CurrentSocketPower << 8) / 1000; @@ -423,6 +423,12 @@ static int yellow_carp_read_sensor(struct smu_context *smu, (uint32_t *)data); *size = 4; break; + case AMDGPU_PP_SENSOR_VCN_LOAD: + ret = yellow_carp_get_smu_metrics_data(smu, + METRICS_AVERAGE_VCNACTIVITY, + (uint32_t *)data); + *size = 4; + break; case AMDGPU_PP_SENSOR_GPU_INPUT_POWER: ret = yellow_carp_get_smu_metrics_data(smu, METRICS_CURR_SOCKETPOWER, diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu14/Makefile b/drivers/gpu/drm/amd/pm/swsmu/smu14/Makefile index ddbac5c655f7..4593e29e8ff8 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu14/Makefile +++ b/drivers/gpu/drm/amd/pm/swsmu/smu14/Makefile @@ -23,7 +23,7 @@ # Makefile for the 'smu manager' sub-component of powerplay. # It provides the smu management services for the driver. -SMU14_MGR = smu_v14_0.o smu_v14_0_0_ppt.o +SMU14_MGR = smu_v14_0.o smu_v14_0_0_ppt.o smu_v14_0_2_ppt.o AMD_SWSMU_SMU14MGR = $(addprefix $(AMD_SWSMU_PATH)/smu14/,$(SMU14_MGR)) diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0.c b/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0.c index b06a3cc43305..68b9bf822e8d 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0.c @@ -38,8 +38,13 @@ #include "amdgpu_ras.h" #include "smu_cmn.h" -#include "asic_reg/mp/mp_14_0_0_offset.h" -#include "asic_reg/mp/mp_14_0_0_sh_mask.h" +#include "asic_reg/mp/mp_14_0_2_offset.h" +#include "asic_reg/mp/mp_14_0_2_sh_mask.h" + +#define regMP1_SMN_IH_SW_INT_mp1_14_0_0 0x0341 +#define regMP1_SMN_IH_SW_INT_mp1_14_0_0_BASE_IDX 0 +#define regMP1_SMN_IH_SW_INT_CTRL_mp1_14_0_0 0x0342 +#define regMP1_SMN_IH_SW_INT_CTRL_mp1_14_0_0_BASE_IDX 0 /* * DO NOT use these for err/warn/info/debug messages. @@ -52,6 +57,7 @@ #undef pr_debug MODULE_FIRMWARE("amdgpu/smu_14_0_2.bin"); +MODULE_FIRMWARE("amdgpu/smu_14_0_3.bin"); #define ENABLE_IMU_ARG_GFXOFF_ENABLE 1 @@ -106,7 +112,6 @@ void smu_v14_0_fini_microcode(struct smu_context *smu) int smu_v14_0_load_microcode(struct smu_context *smu) { -#if 0 struct amdgpu_device *adev = smu->adev; const uint32_t *src; const struct smc_firmware_header_v1_0 *hdr; @@ -131,8 +136,13 @@ int smu_v14_0_load_microcode(struct smu_context *smu) 1 & ~MP1_SMN_PUB_CTRL__LX3_RESET_MASK); for (i = 0; i < adev->usec_timeout; i++) { - mp1_fw_flags = RREG32_PCIE(MP1_Public | - (smnMP1_FIRMWARE_FLAGS & 0xffffffff)); + if (amdgpu_ip_version(adev, MP1_HWIP, 0) == IP_VERSION(14, 0, 0) || + amdgpu_ip_version(adev, MP1_HWIP, 0) == IP_VERSION(14, 0, 1)) + mp1_fw_flags = RREG32_PCIE(MP1_Public | + (smnMP1_FIRMWARE_FLAGS_14_0_0 & 0xffffffff)); + else + mp1_fw_flags = RREG32_PCIE(MP1_Public | + (smnMP1_FIRMWARE_FLAGS & 0xffffffff)); if ((mp1_fw_flags & MP1_CRU1_MP1_FIRMWARE_FLAGS__INTERRUPTS_ENABLED_MASK) >> MP1_CRU1_MP1_FIRMWARE_FLAGS__INTERRUPTS_ENABLED__SHIFT) break; @@ -142,9 +152,7 @@ int smu_v14_0_load_microcode(struct smu_context *smu) if (i == adev->usec_timeout) return -ETIME; -#endif return 0; - } int smu_v14_0_init_pptable_microcode(struct smu_context *smu) @@ -165,6 +173,10 @@ int smu_v14_0_init_pptable_microcode(struct smu_context *smu) if (!adev->scpm_enabled) return 0; + if ((amdgpu_ip_version(adev, MP1_HWIP, 0) == IP_VERSION(14, 0, 2)) || + (amdgpu_ip_version(adev, MP1_HWIP, 0) == IP_VERSION(14, 0, 3))) + return 0; + /* override pptable_id from driver parameter */ if (amdgpu_smu_pptable_id >= 0) { pptable_id = amdgpu_smu_pptable_id; @@ -198,7 +210,12 @@ int smu_v14_0_check_fw_status(struct smu_context *smu) struct amdgpu_device *adev = smu->adev; uint32_t mp1_fw_flags; - mp1_fw_flags = RREG32_PCIE(MP1_Public | + if (amdgpu_ip_version(adev, MP1_HWIP, 0) == IP_VERSION(14, 0, 0) || + amdgpu_ip_version(adev, MP1_HWIP, 0) == IP_VERSION(14, 0, 1)) + mp1_fw_flags = RREG32_PCIE(MP1_Public | + (smnMP1_FIRMWARE_FLAGS_14_0_0 & 0xffffffff)); + else + mp1_fw_flags = RREG32_PCIE(MP1_Public | (smnMP1_FIRMWARE_FLAGS & 0xffffffff)); if ((mp1_fw_flags & MP1_CRU1_MP1_FIRMWARE_FLAGS__INTERRUPTS_ENABLED_MASK) >> @@ -227,16 +244,16 @@ int smu_v14_0_check_fw_version(struct smu_context *smu) adev->pm.fw_version = smu_version; switch (amdgpu_ip_version(adev, MP1_HWIP, 0)) { - case IP_VERSION(14, 0, 2): - smu->smc_driver_if_version = SMU14_DRIVER_IF_VERSION_SMU_V14_0_2; - break; case IP_VERSION(14, 0, 0): smu->smc_driver_if_version = SMU14_DRIVER_IF_VERSION_SMU_V14_0_0; break; case IP_VERSION(14, 0, 1): - smu->smc_driver_if_version = SMU14_DRIVER_IF_VERSION_SMU_V14_0_0; + smu->smc_driver_if_version = SMU14_DRIVER_IF_VERSION_SMU_V14_0_1; + break; + case IP_VERSION(14, 0, 2): + case IP_VERSION(14, 0, 3): + smu->smc_driver_if_version = SMU14_DRIVER_IF_VERSION_SMU_V14_0_2; break; - default: dev_err(adev->dev, "smu unsupported IP version: 0x%x.\n", amdgpu_ip_version(adev, MP1_HWIP, 0)); @@ -738,9 +755,9 @@ int smu_v14_0_gfx_off_control(struct smu_context *smu, bool enable) struct amdgpu_device *adev = smu->adev; switch (amdgpu_ip_version(adev, MP1_HWIP, 0)) { - case IP_VERSION(14, 0, 2): case IP_VERSION(14, 0, 0): case IP_VERSION(14, 0, 1): + case IP_VERSION(14, 0, 2): if (!(adev->pm.pp_feature & PP_GFXOFF_MASK)) return 0; if (enable) @@ -841,9 +858,16 @@ static int smu_v14_0_set_irq_state(struct amdgpu_device *adev, // TODO /* For MP1 SW irqs */ - val = RREG32_SOC15(MP1, 0, regMP1_SMN_IH_SW_INT_CTRL); - val = REG_SET_FIELD(val, MP1_SMN_IH_SW_INT_CTRL, INT_MASK, 1); - WREG32_SOC15(MP1, 0, regMP1_SMN_IH_SW_INT_CTRL, val); + if (amdgpu_ip_version(adev, MP1_HWIP, 0) == IP_VERSION(14, 0, 0) || + amdgpu_ip_version(adev, MP1_HWIP, 0) == IP_VERSION(14, 0, 1)) { + val = RREG32_SOC15(MP1, 0, regMP1_SMN_IH_SW_INT_CTRL_mp1_14_0_0); + val = REG_SET_FIELD(val, MP1_SMN_IH_SW_INT_CTRL, INT_MASK, 1); + WREG32_SOC15(MP1, 0, regMP1_SMN_IH_SW_INT_CTRL_mp1_14_0_0, val); + } else { + val = RREG32_SOC15(MP1, 0, regMP1_SMN_IH_SW_INT_CTRL); + val = REG_SET_FIELD(val, MP1_SMN_IH_SW_INT_CTRL, INT_MASK, 1); + WREG32_SOC15(MP1, 0, regMP1_SMN_IH_SW_INT_CTRL, val); + } break; case AMDGPU_IRQ_STATE_ENABLE: @@ -851,14 +875,26 @@ static int smu_v14_0_set_irq_state(struct amdgpu_device *adev, // TODO /* For MP1 SW irqs */ - val = RREG32_SOC15(MP1, 0, regMP1_SMN_IH_SW_INT); - val = REG_SET_FIELD(val, MP1_SMN_IH_SW_INT, ID, 0xFE); - val = REG_SET_FIELD(val, MP1_SMN_IH_SW_INT, VALID, 0); - WREG32_SOC15(MP1, 0, regMP1_SMN_IH_SW_INT, val); - - val = RREG32_SOC15(MP1, 0, regMP1_SMN_IH_SW_INT_CTRL); - val = REG_SET_FIELD(val, MP1_SMN_IH_SW_INT_CTRL, INT_MASK, 0); - WREG32_SOC15(MP1, 0, regMP1_SMN_IH_SW_INT_CTRL, val); + if (amdgpu_ip_version(adev, MP1_HWIP, 0) == IP_VERSION(14, 0, 0) || + amdgpu_ip_version(adev, MP1_HWIP, 0) == IP_VERSION(14, 0, 1)) { + val = RREG32_SOC15(MP1, 0, regMP1_SMN_IH_SW_INT_mp1_14_0_0); + val = REG_SET_FIELD(val, MP1_SMN_IH_SW_INT, ID, 0xFE); + val = REG_SET_FIELD(val, MP1_SMN_IH_SW_INT, VALID, 0); + WREG32_SOC15(MP1, 0, regMP1_SMN_IH_SW_INT_mp1_14_0_0, val); + + val = RREG32_SOC15(MP1, 0, regMP1_SMN_IH_SW_INT_CTRL_mp1_14_0_0); + val = REG_SET_FIELD(val, MP1_SMN_IH_SW_INT_CTRL, INT_MASK, 0); + WREG32_SOC15(MP1, 0, regMP1_SMN_IH_SW_INT_CTRL_mp1_14_0_0, val); + } else { + val = RREG32_SOC15(MP1, 0, regMP1_SMN_IH_SW_INT); + val = REG_SET_FIELD(val, MP1_SMN_IH_SW_INT, ID, 0xFE); + val = REG_SET_FIELD(val, MP1_SMN_IH_SW_INT, VALID, 0); + WREG32_SOC15(MP1, 0, regMP1_SMN_IH_SW_INT, val); + + val = RREG32_SOC15(MP1, 0, regMP1_SMN_IH_SW_INT_CTRL); + val = REG_SET_FIELD(val, MP1_SMN_IH_SW_INT_CTRL, INT_MASK, 0); + WREG32_SOC15(MP1, 0, regMP1_SMN_IH_SW_INT_CTRL, val); + } break; default: @@ -868,11 +904,32 @@ static int smu_v14_0_set_irq_state(struct amdgpu_device *adev, return 0; } +#define THM_11_0__SRCID__THM_DIG_THERM_L2H 0 /* ASIC_TEMP > CG_THERMAL_INT.DIG_THERM_INTH */ +#define THM_11_0__SRCID__THM_DIG_THERM_H2L 1 /* ASIC_TEMP < CG_THERMAL_INT.DIG_THERM_INTL */ + static int smu_v14_0_irq_process(struct amdgpu_device *adev, struct amdgpu_irq_src *source, struct amdgpu_iv_entry *entry) { - // TODO + struct smu_context *smu = adev->powerplay.pp_handle; + uint32_t client_id = entry->client_id; + uint32_t src_id = entry->src_id; + + if (client_id == SOC15_IH_CLIENTID_THM) { + switch (src_id) { + case THM_11_0__SRCID__THM_DIG_THERM_L2H: + schedule_delayed_work(&smu->swctf_delayed_work, + msecs_to_jiffies(AMDGPU_SWCTF_EXTRA_DELAY)); + break; + case THM_11_0__SRCID__THM_DIG_THERM_H2L: + dev_emerg(adev->dev, "ERROR: GPU under temperature range detected\n"); + break; + default: + dev_emerg(adev->dev, "ERROR: GPU under temperature range unknown src id (%d)\n", + src_id); + break; + } + } return 0; } @@ -894,7 +951,17 @@ int smu_v14_0_register_irq_handler(struct smu_context *smu) irq_src->num_types = 1; irq_src->funcs = &smu_v14_0_irq_funcs; - // TODO: THM related + ret = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_THM, + THM_11_0__SRCID__THM_DIG_THERM_L2H, + irq_src); + if (ret) + return ret; + + ret = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_THM, + THM_11_0__SRCID__THM_DIG_THERM_H2L, + irq_src); + if (ret) + return ret; ret = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_MP1, SMU_IH_INTERRUPT_ID_TO_DRIVER, @@ -1402,9 +1469,22 @@ int smu_v14_0_set_vcn_enable(struct smu_context *smu, if (adev->vcn.harvest_config & (1 << i)) continue; - ret = smu_cmn_send_smc_msg_with_param(smu, enable ? - SMU_MSG_PowerUpVcn : SMU_MSG_PowerDownVcn, - i << 16U, NULL); + if (amdgpu_ip_version(adev, MP1_HWIP, 0) == IP_VERSION(14, 0, 0) || + amdgpu_ip_version(adev, MP1_HWIP, 0) == IP_VERSION(14, 0, 1)) { + if (i == 0) + ret = smu_cmn_send_smc_msg_with_param(smu, enable ? + SMU_MSG_PowerUpVcn0 : SMU_MSG_PowerDownVcn0, + i << 16U, NULL); + else if (i == 1) + ret = smu_cmn_send_smc_msg_with_param(smu, enable ? + SMU_MSG_PowerUpVcn1 : SMU_MSG_PowerDownVcn1, + i << 16U, NULL); + } else { + ret = smu_cmn_send_smc_msg_with_param(smu, enable ? + SMU_MSG_PowerUpVcn : SMU_MSG_PowerDownVcn, + i << 16U, NULL); + } + if (ret) return ret; } @@ -1415,9 +1495,34 @@ int smu_v14_0_set_vcn_enable(struct smu_context *smu, int smu_v14_0_set_jpeg_enable(struct smu_context *smu, bool enable) { - return smu_cmn_send_smc_msg_with_param(smu, enable ? - SMU_MSG_PowerUpJpeg : SMU_MSG_PowerDownJpeg, - 0, NULL); + struct amdgpu_device *adev = smu->adev; + int i, ret = 0; + + for (i = 0; i < adev->jpeg.num_jpeg_inst; i++) { + if (adev->jpeg.harvest_config & (1 << i)) + continue; + + if (amdgpu_ip_version(adev, MP1_HWIP, 0) == IP_VERSION(14, 0, 0) || + amdgpu_ip_version(adev, MP1_HWIP, 0) == IP_VERSION(14, 0, 1)) { + if (i == 0) + ret = smu_cmn_send_smc_msg_with_param(smu, enable ? + SMU_MSG_PowerUpJpeg0 : SMU_MSG_PowerDownJpeg0, + i << 16U, NULL); + else if (i == 1 && amdgpu_ip_version(adev, MP1_HWIP, 0) == IP_VERSION(14, 0, 1)) + ret = smu_cmn_send_smc_msg_with_param(smu, enable ? + SMU_MSG_PowerUpJpeg1 : SMU_MSG_PowerDownJpeg1, + i << 16U, NULL); + } else { + ret = smu_cmn_send_smc_msg_with_param(smu, enable ? + SMU_MSG_PowerUpJpeg : SMU_MSG_PowerDownJpeg, + i << 16U, NULL); + } + + if (ret) + return ret; + } + + return ret; } int smu_v14_0_run_btc(struct smu_context *smu) @@ -1552,23 +1657,27 @@ int smu_v14_0_baco_set_armd3_sequence(struct smu_context *smu, return 0; } -bool smu_v14_0_baco_is_support(struct smu_context *smu) +int smu_v14_0_get_bamaco_support(struct smu_context *smu) { struct smu_baco_context *smu_baco = &smu->smu_baco; + int bamaco_support = 0; if (amdgpu_sriov_vf(smu->adev) || !smu_baco->platform_support) - return false; + return 0; + + if (smu_baco->maco_support) + bamaco_support |= MACO_SUPPORT; /* return true if ASIC is in BACO state already */ if (smu_v14_0_baco_get_state(smu) == SMU_BACO_STATE_ENTER) - return true; + return (bamaco_support |= BACO_SUPPORT); if (smu_cmn_feature_is_supported(smu, SMU_FEATURE_BACO_BIT) && !smu_cmn_feature_is_enabled(smu, SMU_FEATURE_BACO_BIT)) - return false; + return 0; - return true; + return (bamaco_support |= BACO_SUPPORT); } enum smu_baco_state smu_v14_0_baco_get_state(struct smu_context *smu) @@ -1591,7 +1700,7 @@ int smu_v14_0_baco_set_state(struct smu_context *smu, if (state == SMU_BACO_STATE_ENTER) { ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_EnterBaco, - smu_baco->maco_support ? + (adev->pm.rpm_mode == AMDGPU_RUNPM_BAMACO) ? BACO_SEQ_BAMACO : BACO_SEQ_BACO, NULL); } else { diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0_0_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0_0_ppt.c index 9310c4758e38..e4419e1561ef 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0_0_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0_0_ppt.c @@ -70,9 +70,12 @@ static struct cmn2asic_msg_mapping smu_v14_0_0_message_map[SMU_MSG_MAX_COUNT] = MSG_MAP(TestMessage, PPSMC_MSG_TestMessage, 1), MSG_MAP(GetSmuVersion, PPSMC_MSG_GetPmfwVersion, 1), MSG_MAP(GetDriverIfVersion, PPSMC_MSG_GetDriverIfVersion, 1), - MSG_MAP(PowerDownVcn, PPSMC_MSG_PowerDownVcn, 1), - MSG_MAP(PowerUpVcn, PPSMC_MSG_PowerUpVcn, 1), - MSG_MAP(SetHardMinVcn, PPSMC_MSG_SetHardMinVcn, 1), + MSG_MAP(PowerDownVcn0, PPSMC_MSG_PowerDownVcn0, 1), + MSG_MAP(PowerUpVcn0, PPSMC_MSG_PowerUpVcn0, 1), + MSG_MAP(SetHardMinVcn0, PPSMC_MSG_SetHardMinVcn0, 1), + MSG_MAP(PowerDownVcn1, PPSMC_MSG_PowerDownVcn1, 1), + MSG_MAP(PowerUpVcn1, PPSMC_MSG_PowerUpVcn1, 1), + MSG_MAP(SetHardMinVcn1, PPSMC_MSG_SetHardMinVcn1, 1), MSG_MAP(SetSoftMinGfxclk, PPSMC_MSG_SetSoftMinGfxclk, 1), MSG_MAP(PrepareMp1ForUnload, PPSMC_MSG_PrepareMp1ForUnload, 1), MSG_MAP(SetDriverDramAddrHigh, PPSMC_MSG_SetDriverDramAddrHigh, 1), @@ -83,7 +86,8 @@ static struct cmn2asic_msg_mapping smu_v14_0_0_message_map[SMU_MSG_MAX_COUNT] = MSG_MAP(GetEnabledSmuFeatures, PPSMC_MSG_GetEnabledSmuFeatures, 1), MSG_MAP(SetHardMinSocclkByFreq, PPSMC_MSG_SetHardMinSocclkByFreq, 1), MSG_MAP(SetSoftMinFclk, PPSMC_MSG_SetSoftMinFclk, 1), - MSG_MAP(SetSoftMinVcn, PPSMC_MSG_SetSoftMinVcn, 1), + MSG_MAP(SetSoftMinVcn0, PPSMC_MSG_SetSoftMinVcn0, 1), + MSG_MAP(SetSoftMinVcn1, PPSMC_MSG_SetSoftMinVcn1, 1), MSG_MAP(EnableGfxImu, PPSMC_MSG_EnableGfxImu, 1), MSG_MAP(AllowGfxOff, PPSMC_MSG_AllowGfxOff, 1), MSG_MAP(DisallowGfxOff, PPSMC_MSG_DisallowGfxOff, 1), @@ -91,9 +95,12 @@ static struct cmn2asic_msg_mapping smu_v14_0_0_message_map[SMU_MSG_MAX_COUNT] = MSG_MAP(SetHardMinGfxClk, PPSMC_MSG_SetHardMinGfxClk, 1), MSG_MAP(SetSoftMaxSocclkByFreq, PPSMC_MSG_SetSoftMaxSocclkByFreq, 1), MSG_MAP(SetSoftMaxFclkByFreq, PPSMC_MSG_SetSoftMaxFclkByFreq, 1), - MSG_MAP(SetSoftMaxVcn, PPSMC_MSG_SetSoftMaxVcn, 1), - MSG_MAP(PowerDownJpeg, PPSMC_MSG_PowerDownJpeg, 1), - MSG_MAP(PowerUpJpeg, PPSMC_MSG_PowerUpJpeg, 1), + MSG_MAP(SetSoftMaxVcn0, PPSMC_MSG_SetSoftMaxVcn0, 1), + MSG_MAP(SetSoftMaxVcn1, PPSMC_MSG_SetSoftMaxVcn1, 1), + MSG_MAP(PowerDownJpeg0, PPSMC_MSG_PowerDownJpeg0, 1), + MSG_MAP(PowerUpJpeg0, PPSMC_MSG_PowerUpJpeg0, 1), + MSG_MAP(PowerDownJpeg1, PPSMC_MSG_PowerDownJpeg1, 1), + MSG_MAP(PowerUpJpeg1, PPSMC_MSG_PowerUpJpeg1, 1), MSG_MAP(SetHardMinFclkByFreq, PPSMC_MSG_SetHardMinFclkByFreq, 1), MSG_MAP(SetSoftMinSocclkByFreq, PPSMC_MSG_SetSoftMinSocclkByFreq, 1), MSG_MAP(PowerDownIspByTile, PPSMC_MSG_PowerDownIspByTile, 1), @@ -154,7 +161,7 @@ static int smu_v14_0_0_init_smc_tables(struct smu_context *smu) SMU_TABLE_INIT(tables, SMU_TABLE_WATERMARKS, sizeof(Watermarks_t), PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM); - SMU_TABLE_INIT(tables, SMU_TABLE_DPMCLOCKS, sizeof(DpmClocks_t), + SMU_TABLE_INIT(tables, SMU_TABLE_DPMCLOCKS, max(sizeof(DpmClocks_t), sizeof(DpmClocks_t_v14_0_1)), PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM); SMU_TABLE_INIT(tables, SMU_TABLE_SMU_METRICS, sizeof(SmuMetrics_t), PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM); @@ -164,7 +171,7 @@ static int smu_v14_0_0_init_smc_tables(struct smu_context *smu) goto err0_out; smu_table->metrics_time = 0; - smu_table->clocks_table = kzalloc(sizeof(DpmClocks_t), GFP_KERNEL); + smu_table->clocks_table = kzalloc(max(sizeof(DpmClocks_t), sizeof(DpmClocks_t_v14_0_1)), GFP_KERNEL); if (!smu_table->clocks_table) goto err1_out; @@ -355,6 +362,12 @@ static int smu_v14_0_0_read_sensor(struct smu_context *smu, (uint32_t *)data); *size = 4; break; + case AMDGPU_PP_SENSOR_VCN_LOAD: + ret = smu_v14_0_0_get_smu_metrics_data(smu, + METRICS_AVERAGE_VCNACTIVITY, + (uint32_t *)data); + *size = 4; + break; case AMDGPU_PP_SENSOR_GPU_AVG_POWER: ret = smu_v14_0_0_get_smu_metrics_data(smu, METRICS_AVERAGE_SOCKETPOWER, @@ -586,6 +599,60 @@ static int smu_v14_0_0_mode2_reset(struct smu_context *smu) return ret; } +static int smu_v14_0_1_get_dpm_freq_by_index(struct smu_context *smu, + enum smu_clk_type clk_type, + uint32_t dpm_level, + uint32_t *freq) +{ + DpmClocks_t_v14_0_1 *clk_table = smu->smu_table.clocks_table; + + if (!clk_table || clk_type >= SMU_CLK_COUNT) + return -EINVAL; + + switch (clk_type) { + case SMU_SOCCLK: + if (dpm_level >= clk_table->NumSocClkLevelsEnabled) + return -EINVAL; + *freq = clk_table->SocClocks[dpm_level]; + break; + case SMU_VCLK: + if (dpm_level >= clk_table->Vcn0ClkLevelsEnabled) + return -EINVAL; + *freq = clk_table->VClocks0[dpm_level]; + break; + case SMU_DCLK: + if (dpm_level >= clk_table->Vcn0ClkLevelsEnabled) + return -EINVAL; + *freq = clk_table->DClocks0[dpm_level]; + break; + case SMU_VCLK1: + if (dpm_level >= clk_table->Vcn1ClkLevelsEnabled) + return -EINVAL; + *freq = clk_table->VClocks1[dpm_level]; + break; + case SMU_DCLK1: + if (dpm_level >= clk_table->Vcn1ClkLevelsEnabled) + return -EINVAL; + *freq = clk_table->DClocks1[dpm_level]; + break; + case SMU_UCLK: + case SMU_MCLK: + if (dpm_level >= clk_table->NumMemPstatesEnabled) + return -EINVAL; + *freq = clk_table->MemPstateTable[dpm_level].MemClk; + break; + case SMU_FCLK: + if (dpm_level >= clk_table->NumFclkLevelsEnabled) + return -EINVAL; + *freq = clk_table->FclkClocks_Freq[dpm_level]; + break; + default: + return -EINVAL; + } + + return 0; +} + static int smu_v14_0_0_get_dpm_freq_by_index(struct smu_context *smu, enum smu_clk_type clk_type, uint32_t dpm_level, @@ -630,6 +697,19 @@ static int smu_v14_0_0_get_dpm_freq_by_index(struct smu_context *smu, return 0; } +static int smu_v14_0_common_get_dpm_freq_by_index(struct smu_context *smu, + enum smu_clk_type clk_type, + uint32_t dpm_level, + uint32_t *freq) +{ + if (amdgpu_ip_version(smu->adev, MP1_HWIP, 0) == IP_VERSION(14, 0, 0)) + smu_v14_0_0_get_dpm_freq_by_index(smu, clk_type, dpm_level, freq); + else if (amdgpu_ip_version(smu->adev, MP1_HWIP, 0) == IP_VERSION(14, 0, 1)) + smu_v14_0_1_get_dpm_freq_by_index(smu, clk_type, dpm_level, freq); + + return 0; +} + static bool smu_v14_0_0_clk_dpm_is_enabled(struct smu_context *smu, enum smu_clk_type clk_type) { @@ -650,6 +730,8 @@ static bool smu_v14_0_0_clk_dpm_is_enabled(struct smu_context *smu, break; case SMU_VCLK: case SMU_DCLK: + case SMU_VCLK1: + case SMU_DCLK1: feature_id = SMU_FEATURE_VCN_DPM_BIT; break; default: @@ -659,6 +741,126 @@ static bool smu_v14_0_0_clk_dpm_is_enabled(struct smu_context *smu, return smu_cmn_feature_is_enabled(smu, feature_id); } +static int smu_v14_0_1_get_dpm_ultimate_freq(struct smu_context *smu, + enum smu_clk_type clk_type, + uint32_t *min, + uint32_t *max) +{ + DpmClocks_t_v14_0_1 *clk_table = smu->smu_table.clocks_table; + uint32_t clock_limit; + uint32_t max_dpm_level, min_dpm_level; + int ret = 0; + + if (!smu_v14_0_0_clk_dpm_is_enabled(smu, clk_type)) { + switch (clk_type) { + case SMU_MCLK: + case SMU_UCLK: + clock_limit = smu->smu_table.boot_values.uclk; + break; + case SMU_FCLK: + clock_limit = smu->smu_table.boot_values.fclk; + break; + case SMU_GFXCLK: + case SMU_SCLK: + clock_limit = smu->smu_table.boot_values.gfxclk; + break; + case SMU_SOCCLK: + clock_limit = smu->smu_table.boot_values.socclk; + break; + case SMU_VCLK: + case SMU_VCLK1: + clock_limit = smu->smu_table.boot_values.vclk; + break; + case SMU_DCLK: + case SMU_DCLK1: + clock_limit = smu->smu_table.boot_values.dclk; + break; + default: + clock_limit = 0; + break; + } + + /* clock in Mhz unit */ + if (min) + *min = clock_limit / 100; + if (max) + *max = clock_limit / 100; + + return 0; + } + + if (max) { + switch (clk_type) { + case SMU_GFXCLK: + case SMU_SCLK: + *max = clk_table->MaxGfxClk; + break; + case SMU_MCLK: + case SMU_UCLK: + case SMU_FCLK: + max_dpm_level = 0; + break; + case SMU_SOCCLK: + max_dpm_level = clk_table->NumSocClkLevelsEnabled - 1; + break; + case SMU_VCLK: + case SMU_DCLK: + max_dpm_level = clk_table->Vcn0ClkLevelsEnabled - 1; + break; + case SMU_VCLK1: + case SMU_DCLK1: + max_dpm_level = clk_table->Vcn1ClkLevelsEnabled - 1; + break; + default: + ret = -EINVAL; + goto failed; + } + + if (clk_type != SMU_GFXCLK && clk_type != SMU_SCLK) { + ret = smu_v14_0_common_get_dpm_freq_by_index(smu, clk_type, max_dpm_level, max); + if (ret) + goto failed; + } + } + + if (min) { + switch (clk_type) { + case SMU_GFXCLK: + case SMU_SCLK: + *min = clk_table->MinGfxClk; + break; + case SMU_MCLK: + case SMU_UCLK: + min_dpm_level = clk_table->NumMemPstatesEnabled - 1; + break; + case SMU_FCLK: + min_dpm_level = clk_table->NumFclkLevelsEnabled - 1; + break; + case SMU_SOCCLK: + min_dpm_level = 0; + break; + case SMU_VCLK: + case SMU_DCLK: + case SMU_VCLK1: + case SMU_DCLK1: + min_dpm_level = 0; + break; + default: + ret = -EINVAL; + goto failed; + } + + if (clk_type != SMU_GFXCLK && clk_type != SMU_SCLK) { + ret = smu_v14_0_common_get_dpm_freq_by_index(smu, clk_type, min_dpm_level, min); + if (ret) + goto failed; + } + } + +failed: + return ret; +} + static int smu_v14_0_0_get_dpm_ultimate_freq(struct smu_context *smu, enum smu_clk_type clk_type, uint32_t *min, @@ -729,7 +931,7 @@ static int smu_v14_0_0_get_dpm_ultimate_freq(struct smu_context *smu, } if (clk_type != SMU_GFXCLK && clk_type != SMU_SCLK) { - ret = smu_v14_0_0_get_dpm_freq_by_index(smu, clk_type, max_dpm_level, max); + ret = smu_v14_0_common_get_dpm_freq_by_index(smu, clk_type, max_dpm_level, max); if (ret) goto failed; } @@ -761,7 +963,7 @@ static int smu_v14_0_0_get_dpm_ultimate_freq(struct smu_context *smu, } if (clk_type != SMU_GFXCLK && clk_type != SMU_SCLK) { - ret = smu_v14_0_0_get_dpm_freq_by_index(smu, clk_type, min_dpm_level, min); + ret = smu_v14_0_common_get_dpm_freq_by_index(smu, clk_type, min_dpm_level, min); if (ret) goto failed; } @@ -771,6 +973,19 @@ failed: return ret; } +static int smu_v14_0_common_get_dpm_ultimate_freq(struct smu_context *smu, + enum smu_clk_type clk_type, + uint32_t *min, + uint32_t *max) +{ + if (amdgpu_ip_version(smu->adev, MP1_HWIP, 0) == IP_VERSION(14, 0, 0)) + smu_v14_0_0_get_dpm_ultimate_freq(smu, clk_type, min, max); + else if (amdgpu_ip_version(smu->adev, MP1_HWIP, 0) == IP_VERSION(14, 0, 1)) + smu_v14_0_1_get_dpm_ultimate_freq(smu, clk_type, min, max); + + return 0; +} + static int smu_v14_0_0_get_current_clk_freq(struct smu_context *smu, enum smu_clk_type clk_type, uint32_t *value) @@ -804,6 +1019,37 @@ static int smu_v14_0_0_get_current_clk_freq(struct smu_context *smu, return smu_v14_0_0_get_smu_metrics_data(smu, member_type, value); } +static int smu_v14_0_1_get_dpm_level_count(struct smu_context *smu, + enum smu_clk_type clk_type, + uint32_t *count) +{ + DpmClocks_t_v14_0_1 *clk_table = smu->smu_table.clocks_table; + + switch (clk_type) { + case SMU_SOCCLK: + *count = clk_table->NumSocClkLevelsEnabled; + break; + case SMU_VCLK: + case SMU_DCLK: + *count = clk_table->Vcn0ClkLevelsEnabled; + break; + case SMU_VCLK1: + case SMU_DCLK1: + *count = clk_table->Vcn1ClkLevelsEnabled; + break; + case SMU_MCLK: + *count = clk_table->NumMemPstatesEnabled; + break; + case SMU_FCLK: + *count = clk_table->NumFclkLevelsEnabled; + break; + default: + break; + } + + return 0; +} + static int smu_v14_0_0_get_dpm_level_count(struct smu_context *smu, enum smu_clk_type clk_type, uint32_t *count) @@ -833,6 +1079,18 @@ static int smu_v14_0_0_get_dpm_level_count(struct smu_context *smu, return 0; } +static int smu_v14_0_common_get_dpm_level_count(struct smu_context *smu, + enum smu_clk_type clk_type, + uint32_t *count) +{ + if (amdgpu_ip_version(smu->adev, MP1_HWIP, 0) == IP_VERSION(14, 0, 0)) + smu_v14_0_0_get_dpm_level_count(smu, clk_type, count); + else if (amdgpu_ip_version(smu->adev, MP1_HWIP, 0) == IP_VERSION(14, 0, 1)) + smu_v14_0_1_get_dpm_level_count(smu, clk_type, count); + + return 0; +} + static int smu_v14_0_0_print_clk_levels(struct smu_context *smu, enum smu_clk_type clk_type, char *buf) { @@ -859,18 +1117,20 @@ static int smu_v14_0_0_print_clk_levels(struct smu_context *smu, case SMU_SOCCLK: case SMU_VCLK: case SMU_DCLK: + case SMU_VCLK1: + case SMU_DCLK1: case SMU_MCLK: case SMU_FCLK: ret = smu_v14_0_0_get_current_clk_freq(smu, clk_type, &cur_value); if (ret) break; - ret = smu_v14_0_0_get_dpm_level_count(smu, clk_type, &count); + ret = smu_v14_0_common_get_dpm_level_count(smu, clk_type, &count); if (ret) break; for (i = 0; i < count; i++) { - ret = smu_v14_0_0_get_dpm_freq_by_index(smu, clk_type, i, &value); + ret = smu_v14_0_common_get_dpm_freq_by_index(smu, clk_type, i, &value); if (ret) break; @@ -933,8 +1193,13 @@ static int smu_v14_0_0_set_soft_freq_limited_range(struct smu_context *smu, break; case SMU_VCLK: case SMU_DCLK: - msg_set_min = SMU_MSG_SetHardMinVcn; - msg_set_max = SMU_MSG_SetSoftMaxVcn; + msg_set_min = SMU_MSG_SetHardMinVcn0; + msg_set_max = SMU_MSG_SetSoftMaxVcn0; + break; + case SMU_VCLK1: + case SMU_DCLK1: + msg_set_min = SMU_MSG_SetHardMinVcn1; + msg_set_max = SMU_MSG_SetSoftMaxVcn1; break; default: return -EINVAL; @@ -964,11 +1229,11 @@ static int smu_v14_0_0_force_clk_levels(struct smu_context *smu, case SMU_FCLK: case SMU_VCLK: case SMU_DCLK: - ret = smu_v14_0_0_get_dpm_freq_by_index(smu, clk_type, soft_min_level, &min_freq); + ret = smu_v14_0_common_get_dpm_freq_by_index(smu, clk_type, soft_min_level, &min_freq); if (ret) break; - ret = smu_v14_0_0_get_dpm_freq_by_index(smu, clk_type, soft_max_level, &max_freq); + ret = smu_v14_0_common_get_dpm_freq_by_index(smu, clk_type, soft_max_level, &max_freq); if (ret) break; @@ -993,25 +1258,25 @@ static int smu_v14_0_0_set_performance_level(struct smu_context *smu, switch (level) { case AMD_DPM_FORCED_LEVEL_HIGH: - smu_v14_0_0_get_dpm_ultimate_freq(smu, SMU_SCLK, NULL, &sclk_max); - smu_v14_0_0_get_dpm_ultimate_freq(smu, SMU_FCLK, NULL, &fclk_max); - smu_v14_0_0_get_dpm_ultimate_freq(smu, SMU_SOCCLK, NULL, &socclk_max); + smu_v14_0_common_get_dpm_ultimate_freq(smu, SMU_SCLK, NULL, &sclk_max); + smu_v14_0_common_get_dpm_ultimate_freq(smu, SMU_FCLK, NULL, &fclk_max); + smu_v14_0_common_get_dpm_ultimate_freq(smu, SMU_SOCCLK, NULL, &socclk_max); sclk_min = sclk_max; fclk_min = fclk_max; socclk_min = socclk_max; break; case AMD_DPM_FORCED_LEVEL_LOW: - smu_v14_0_0_get_dpm_ultimate_freq(smu, SMU_SCLK, &sclk_min, NULL); - smu_v14_0_0_get_dpm_ultimate_freq(smu, SMU_FCLK, &fclk_min, NULL); - smu_v14_0_0_get_dpm_ultimate_freq(smu, SMU_SOCCLK, &socclk_min, NULL); + smu_v14_0_common_get_dpm_ultimate_freq(smu, SMU_SCLK, &sclk_min, NULL); + smu_v14_0_common_get_dpm_ultimate_freq(smu, SMU_FCLK, &fclk_min, NULL); + smu_v14_0_common_get_dpm_ultimate_freq(smu, SMU_SOCCLK, &socclk_min, NULL); sclk_max = sclk_min; fclk_max = fclk_min; socclk_max = socclk_min; break; case AMD_DPM_FORCED_LEVEL_AUTO: - smu_v14_0_0_get_dpm_ultimate_freq(smu, SMU_SCLK, &sclk_min, &sclk_max); - smu_v14_0_0_get_dpm_ultimate_freq(smu, SMU_FCLK, &fclk_min, &fclk_max); - smu_v14_0_0_get_dpm_ultimate_freq(smu, SMU_SOCCLK, &socclk_min, &socclk_max); + smu_v14_0_common_get_dpm_ultimate_freq(smu, SMU_SCLK, &sclk_min, &sclk_max); + smu_v14_0_common_get_dpm_ultimate_freq(smu, SMU_FCLK, &fclk_min, &fclk_max); + smu_v14_0_common_get_dpm_ultimate_freq(smu, SMU_SOCCLK, &socclk_min, &socclk_max); break; case AMD_DPM_FORCED_LEVEL_PROFILE_STANDARD: case AMD_DPM_FORCED_LEVEL_PROFILE_MIN_SCLK: @@ -1060,6 +1325,18 @@ static int smu_v14_0_0_set_performance_level(struct smu_context *smu, return ret; } +static int smu_v14_0_1_set_fine_grain_gfx_freq_parameters(struct smu_context *smu) +{ + DpmClocks_t_v14_0_1 *clk_table = smu->smu_table.clocks_table; + + smu->gfx_default_hard_min_freq = clk_table->MinGfxClk; + smu->gfx_default_soft_max_freq = clk_table->MaxGfxClk; + smu->gfx_actual_hard_min_freq = 0; + smu->gfx_actual_soft_max_freq = 0; + + return 0; +} + static int smu_v14_0_0_set_fine_grain_gfx_freq_parameters(struct smu_context *smu) { DpmClocks_t *clk_table = smu->smu_table.clocks_table; @@ -1072,6 +1349,16 @@ static int smu_v14_0_0_set_fine_grain_gfx_freq_parameters(struct smu_context *sm return 0; } +static int smu_v14_0_common_set_fine_grain_gfx_freq_parameters(struct smu_context *smu) +{ + if (amdgpu_ip_version(smu->adev, MP1_HWIP, 0) == IP_VERSION(14, 0, 0)) + smu_v14_0_0_set_fine_grain_gfx_freq_parameters(smu); + else if (amdgpu_ip_version(smu->adev, MP1_HWIP, 0) == IP_VERSION(14, 0, 1)) + smu_v14_0_1_set_fine_grain_gfx_freq_parameters(smu); + + return 0; +} + static int smu_v14_0_0_set_vpe_enable(struct smu_context *smu, bool enable) { @@ -1088,6 +1375,25 @@ static int smu_v14_0_0_set_umsch_mm_enable(struct smu_context *smu, 0, NULL); } +static int smu_14_0_1_get_dpm_table(struct smu_context *smu, struct dpm_clocks *clock_table) +{ + DpmClocks_t_v14_0_1 *clk_table = smu->smu_table.clocks_table; + uint8_t idx; + + /* Only the Clock information of SOC and VPE is copied to provide VPE DPM settings for use. */ + for (idx = 0; idx < NUM_SOCCLK_DPM_LEVELS; idx++) { + clock_table->SocClocks[idx].Freq = (idx < clk_table->NumSocClkLevelsEnabled) ? clk_table->SocClocks[idx]:0; + clock_table->SocClocks[idx].Vol = 0; + } + + for (idx = 0; idx < NUM_VPE_DPM_LEVELS; idx++) { + clock_table->VPEClocks[idx].Freq = (idx < clk_table->VpeClkLevelsEnabled) ? clk_table->VPEClocks[idx]:0; + clock_table->VPEClocks[idx].Vol = 0; + } + + return 0; +} + static int smu_14_0_0_get_dpm_table(struct smu_context *smu, struct dpm_clocks *clock_table) { DpmClocks_t *clk_table = smu->smu_table.clocks_table; @@ -1107,6 +1413,16 @@ static int smu_14_0_0_get_dpm_table(struct smu_context *smu, struct dpm_clocks * return 0; } +static int smu_v14_0_common_get_dpm_table(struct smu_context *smu, struct dpm_clocks *clock_table) +{ + if (amdgpu_ip_version(smu->adev, MP1_HWIP, 0) == IP_VERSION(14, 0, 0)) + smu_14_0_0_get_dpm_table(smu, clock_table); + else if (amdgpu_ip_version(smu->adev, MP1_HWIP, 0) == IP_VERSION(14, 0, 1)) + smu_14_0_1_get_dpm_table(smu, clock_table); + + return 0; +} + static const struct pptable_funcs smu_v14_0_0_ppt_funcs = { .check_fw_status = smu_v14_0_check_fw_status, .check_fw_version = smu_v14_0_check_fw_version, @@ -1128,16 +1444,16 @@ static const struct pptable_funcs smu_v14_0_0_ppt_funcs = { .set_driver_table_location = smu_v14_0_set_driver_table_location, .gfx_off_control = smu_v14_0_gfx_off_control, .mode2_reset = smu_v14_0_0_mode2_reset, - .get_dpm_ultimate_freq = smu_v14_0_0_get_dpm_ultimate_freq, + .get_dpm_ultimate_freq = smu_v14_0_common_get_dpm_ultimate_freq, .od_edit_dpm_table = smu_v14_0_od_edit_dpm_table, .print_clk_levels = smu_v14_0_0_print_clk_levels, .force_clk_levels = smu_v14_0_0_force_clk_levels, .set_performance_level = smu_v14_0_0_set_performance_level, - .set_fine_grain_gfx_freq_parameters = smu_v14_0_0_set_fine_grain_gfx_freq_parameters, + .set_fine_grain_gfx_freq_parameters = smu_v14_0_common_set_fine_grain_gfx_freq_parameters, .set_gfx_power_up_by_imu = smu_v14_0_set_gfx_power_up_by_imu, .dpm_set_vpe_enable = smu_v14_0_0_set_vpe_enable, .dpm_set_umsch_mm_enable = smu_v14_0_0_set_umsch_mm_enable, - .get_dpm_clock_table = smu_14_0_0_get_dpm_table, + .get_dpm_clock_table = smu_v14_0_common_get_dpm_table, }; static void smu_v14_0_0_set_smu_mailbox_registers(struct smu_context *smu) diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0_2_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0_2_ppt.c new file mode 100644 index 000000000000..90703f4542ab --- /dev/null +++ b/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0_2_ppt.c @@ -0,0 +1,1795 @@ +/* + * Copyright 2023 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. + * + */ + +#define SWSMU_CODE_LAYER_L2 + +#include <linux/firmware.h> +#include <linux/pci.h> +#include <linux/i2c.h> +#include "amdgpu.h" +#include "amdgpu_smu.h" +#include "atomfirmware.h" +#include "amdgpu_atomfirmware.h" +#include "amdgpu_atombios.h" +#include "smu_v14_0.h" +#include "smu14_driver_if_v14_0.h" +#include "soc15_common.h" +#include "atom.h" +#include "smu_v14_0_2_ppt.h" +#include "smu_v14_0_2_pptable.h" +#include "smu_v14_0_2_ppsmc.h" +#include "mp/mp_14_0_2_offset.h" +#include "mp/mp_14_0_2_sh_mask.h" + +#include "smu_cmn.h" +#include "amdgpu_ras.h" + +/* + * DO NOT use these for err/warn/info/debug messages. + * Use dev_err, dev_warn, dev_info and dev_dbg instead. + * They are more MGPU friendly. + */ +#undef pr_err +#undef pr_warn +#undef pr_info +#undef pr_debug + +#define to_amdgpu_device(x) (container_of(x, struct amdgpu_device, pm.smu_i2c)) + +#define FEATURE_MASK(feature) (1ULL << feature) +#define SMC_DPM_FEATURE ( \ + FEATURE_MASK(FEATURE_DPM_GFXCLK_BIT) | \ + FEATURE_MASK(FEATURE_DPM_UCLK_BIT) | \ + FEATURE_MASK(FEATURE_DPM_LINK_BIT) | \ + FEATURE_MASK(FEATURE_DPM_SOCCLK_BIT) | \ + FEATURE_MASK(FEATURE_DPM_FCLK_BIT)) + +#define MP0_MP1_DATA_REGION_SIZE_COMBOPPTABLE 0x4000 + +static struct cmn2asic_msg_mapping smu_v14_0_2_message_map[SMU_MSG_MAX_COUNT] = { + MSG_MAP(TestMessage, PPSMC_MSG_TestMessage, 1), + MSG_MAP(GetSmuVersion, PPSMC_MSG_GetSmuVersion, 1), + MSG_MAP(GetDriverIfVersion, PPSMC_MSG_GetDriverIfVersion, 1), + MSG_MAP(SetAllowedFeaturesMaskLow, PPSMC_MSG_SetAllowedFeaturesMaskLow, 0), + MSG_MAP(SetAllowedFeaturesMaskHigh, PPSMC_MSG_SetAllowedFeaturesMaskHigh, 0), + MSG_MAP(EnableAllSmuFeatures, PPSMC_MSG_EnableAllSmuFeatures, 0), + MSG_MAP(DisableAllSmuFeatures, PPSMC_MSG_DisableAllSmuFeatures, 0), + MSG_MAP(EnableSmuFeaturesLow, PPSMC_MSG_EnableSmuFeaturesLow, 1), + MSG_MAP(EnableSmuFeaturesHigh, PPSMC_MSG_EnableSmuFeaturesHigh, 1), + MSG_MAP(DisableSmuFeaturesLow, PPSMC_MSG_DisableSmuFeaturesLow, 1), + MSG_MAP(DisableSmuFeaturesHigh, PPSMC_MSG_DisableSmuFeaturesHigh, 1), + MSG_MAP(GetEnabledSmuFeaturesLow, PPSMC_MSG_GetRunningSmuFeaturesLow, 1), + MSG_MAP(GetEnabledSmuFeaturesHigh, PPSMC_MSG_GetRunningSmuFeaturesHigh, 1), + MSG_MAP(SetWorkloadMask, PPSMC_MSG_SetWorkloadMask, 1), + MSG_MAP(SetPptLimit, PPSMC_MSG_SetPptLimit, 0), + MSG_MAP(SetDriverDramAddrHigh, PPSMC_MSG_SetDriverDramAddrHigh, 1), + MSG_MAP(SetDriverDramAddrLow, PPSMC_MSG_SetDriverDramAddrLow, 1), + MSG_MAP(SetToolsDramAddrHigh, PPSMC_MSG_SetToolsDramAddrHigh, 0), + MSG_MAP(SetToolsDramAddrLow, PPSMC_MSG_SetToolsDramAddrLow, 0), + MSG_MAP(TransferTableSmu2Dram, PPSMC_MSG_TransferTableSmu2Dram, 1), + MSG_MAP(TransferTableDram2Smu, PPSMC_MSG_TransferTableDram2Smu, 0), + MSG_MAP(UseDefaultPPTable, PPSMC_MSG_UseDefaultPPTable, 0), + MSG_MAP(RunDcBtc, PPSMC_MSG_RunDcBtc, 0), + MSG_MAP(EnterBaco, PPSMC_MSG_EnterBaco, 0), + MSG_MAP(ExitBaco, PPSMC_MSG_ExitBaco, 0), + MSG_MAP(SetSoftMinByFreq, PPSMC_MSG_SetSoftMinByFreq, 1), + MSG_MAP(SetSoftMaxByFreq, PPSMC_MSG_SetSoftMaxByFreq, 1), + MSG_MAP(SetHardMinByFreq, PPSMC_MSG_SetHardMinByFreq, 1), + MSG_MAP(SetHardMaxByFreq, PPSMC_MSG_SetHardMaxByFreq, 0), + MSG_MAP(GetMinDpmFreq, PPSMC_MSG_GetMinDpmFreq, 1), + MSG_MAP(GetMaxDpmFreq, PPSMC_MSG_GetMaxDpmFreq, 1), + MSG_MAP(GetDpmFreqByIndex, PPSMC_MSG_GetDpmFreqByIndex, 1), + MSG_MAP(PowerUpVcn, PPSMC_MSG_PowerUpVcn, 0), + MSG_MAP(PowerDownVcn, PPSMC_MSG_PowerDownVcn, 0), + MSG_MAP(PowerUpJpeg, PPSMC_MSG_PowerUpJpeg, 0), + MSG_MAP(PowerDownJpeg, PPSMC_MSG_PowerDownJpeg, 0), + MSG_MAP(GetDcModeMaxDpmFreq, PPSMC_MSG_GetDcModeMaxDpmFreq, 1), + MSG_MAP(OverridePcieParameters, PPSMC_MSG_OverridePcieParameters, 0), + MSG_MAP(DramLogSetDramAddrHigh, PPSMC_MSG_DramLogSetDramAddrHigh, 0), + MSG_MAP(DramLogSetDramAddrLow, PPSMC_MSG_DramLogSetDramAddrLow, 0), + MSG_MAP(DramLogSetDramSize, PPSMC_MSG_DramLogSetDramSize, 0), + MSG_MAP(AllowGfxOff, PPSMC_MSG_AllowGfxOff, 0), + MSG_MAP(DisallowGfxOff, PPSMC_MSG_DisallowGfxOff, 0), + MSG_MAP(SetMGpuFanBoostLimitRpm, PPSMC_MSG_SetMGpuFanBoostLimitRpm, 0), + MSG_MAP(GetPptLimit, PPSMC_MSG_GetPptLimit, 0), + MSG_MAP(NotifyPowerSource, PPSMC_MSG_NotifyPowerSource, 0), + MSG_MAP(Mode1Reset, PPSMC_MSG_Mode1Reset, 0), + MSG_MAP(PrepareMp1ForUnload, PPSMC_MSG_PrepareMp1ForUnload, 0), + MSG_MAP(DFCstateControl, PPSMC_MSG_SetExternalClientDfCstateAllow, 0), + MSG_MAP(ArmD3, PPSMC_MSG_ArmD3, 0), + MSG_MAP(SetNumBadMemoryPagesRetired, PPSMC_MSG_SetNumBadMemoryPagesRetired, 0), + MSG_MAP(SetBadMemoryPagesRetiredFlagsPerChannel, + PPSMC_MSG_SetBadMemoryPagesRetiredFlagsPerChannel, 0), + MSG_MAP(AllowIHHostInterrupt, PPSMC_MSG_AllowIHHostInterrupt, 0), + MSG_MAP(ReenableAcDcInterrupt, PPSMC_MSG_ReenableAcDcInterrupt, 0), +}; + +static struct cmn2asic_mapping smu_v14_0_2_clk_map[SMU_CLK_COUNT] = { + CLK_MAP(GFXCLK, PPCLK_GFXCLK), + CLK_MAP(SCLK, PPCLK_GFXCLK), + CLK_MAP(SOCCLK, PPCLK_SOCCLK), + CLK_MAP(FCLK, PPCLK_FCLK), + CLK_MAP(UCLK, PPCLK_UCLK), + CLK_MAP(MCLK, PPCLK_UCLK), + CLK_MAP(VCLK, PPCLK_VCLK_0), + CLK_MAP(DCLK, PPCLK_DCLK_0), +}; + +static struct cmn2asic_mapping smu_v14_0_2_feature_mask_map[SMU_FEATURE_COUNT] = { + FEA_MAP(FW_DATA_READ), + FEA_MAP(DPM_GFXCLK), + FEA_MAP(DPM_GFX_POWER_OPTIMIZER), + FEA_MAP(DPM_UCLK), + FEA_MAP(DPM_FCLK), + FEA_MAP(DPM_SOCCLK), + FEA_MAP(DPM_LINK), + FEA_MAP(DPM_DCN), + FEA_MAP(VMEMP_SCALING), + FEA_MAP(VDDIO_MEM_SCALING), + FEA_MAP(DS_GFXCLK), + FEA_MAP(DS_SOCCLK), + FEA_MAP(DS_FCLK), + FEA_MAP(DS_LCLK), + FEA_MAP(DS_DCFCLK), + FEA_MAP(DS_UCLK), + FEA_MAP(GFX_ULV), + FEA_MAP(FW_DSTATE), + FEA_MAP(GFXOFF), + FEA_MAP(BACO), + FEA_MAP(MM_DPM), + FEA_MAP(SOC_MPCLK_DS), + FEA_MAP(BACO_MPCLK_DS), + FEA_MAP(THROTTLERS), + FEA_MAP(SMARTSHIFT), + FEA_MAP(GTHR), + FEA_MAP(ACDC), + FEA_MAP(VR0HOT), + FEA_MAP(FW_CTF), + FEA_MAP(FAN_CONTROL), + FEA_MAP(GFX_DCS), + FEA_MAP(GFX_READ_MARGIN), + FEA_MAP(LED_DISPLAY), + FEA_MAP(GFXCLK_SPREAD_SPECTRUM), + FEA_MAP(OUT_OF_BAND_MONITOR), + FEA_MAP(OPTIMIZED_VMIN), + FEA_MAP(GFX_IMU), + FEA_MAP(BOOT_TIME_CAL), + FEA_MAP(GFX_PCC_DFLL), + FEA_MAP(SOC_CG), + FEA_MAP(DF_CSTATE), + FEA_MAP(GFX_EDC), + FEA_MAP(BOOT_POWER_OPT), + FEA_MAP(CLOCK_POWER_DOWN_BYPASS), + FEA_MAP(DS_VCN), + FEA_MAP(BACO_CG), + FEA_MAP(MEM_TEMP_READ), + FEA_MAP(ATHUB_MMHUB_PG), + FEA_MAP(SOC_PCC), + [SMU_FEATURE_DPM_VCLK_BIT] = {1, FEATURE_MM_DPM_BIT}, + [SMU_FEATURE_DPM_DCLK_BIT] = {1, FEATURE_MM_DPM_BIT}, + [SMU_FEATURE_PPT_BIT] = {1, FEATURE_THROTTLERS_BIT}, +}; + +static struct cmn2asic_mapping smu_v14_0_2_table_map[SMU_TABLE_COUNT] = { + TAB_MAP(PPTABLE), + TAB_MAP(WATERMARKS), + TAB_MAP(AVFS_PSM_DEBUG), + TAB_MAP(PMSTATUSLOG), + TAB_MAP(SMU_METRICS), + TAB_MAP(DRIVER_SMU_CONFIG), + TAB_MAP(ACTIVITY_MONITOR_COEFF), + [SMU_TABLE_COMBO_PPTABLE] = {1, TABLE_COMBO_PPTABLE}, + TAB_MAP(I2C_COMMANDS), + TAB_MAP(ECCINFO), +}; + +static struct cmn2asic_mapping smu_v14_0_2_pwr_src_map[SMU_POWER_SOURCE_COUNT] = { + PWR_MAP(AC), + PWR_MAP(DC), +}; + +static struct cmn2asic_mapping smu_v14_0_2_workload_map[PP_SMC_POWER_PROFILE_COUNT] = { + WORKLOAD_MAP(PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT, WORKLOAD_PPLIB_DEFAULT_BIT), + WORKLOAD_MAP(PP_SMC_POWER_PROFILE_FULLSCREEN3D, WORKLOAD_PPLIB_FULL_SCREEN_3D_BIT), + WORKLOAD_MAP(PP_SMC_POWER_PROFILE_POWERSAVING, WORKLOAD_PPLIB_POWER_SAVING_BIT), + WORKLOAD_MAP(PP_SMC_POWER_PROFILE_VIDEO, WORKLOAD_PPLIB_VIDEO_BIT), + WORKLOAD_MAP(PP_SMC_POWER_PROFILE_VR, WORKLOAD_PPLIB_VR_BIT), + WORKLOAD_MAP(PP_SMC_POWER_PROFILE_COMPUTE, WORKLOAD_PPLIB_COMPUTE_BIT), + WORKLOAD_MAP(PP_SMC_POWER_PROFILE_CUSTOM, WORKLOAD_PPLIB_CUSTOM_BIT), + WORKLOAD_MAP(PP_SMC_POWER_PROFILE_WINDOW3D, WORKLOAD_PPLIB_WINDOW_3D_BIT), +}; + +#if 0 +static const uint8_t smu_v14_0_2_throttler_map[] = { + [THROTTLER_PPT0_BIT] = (SMU_THROTTLER_PPT0_BIT), + [THROTTLER_PPT1_BIT] = (SMU_THROTTLER_PPT1_BIT), + [THROTTLER_PPT2_BIT] = (SMU_THROTTLER_PPT2_BIT), + [THROTTLER_PPT3_BIT] = (SMU_THROTTLER_PPT3_BIT), + [THROTTLER_TDC_GFX_BIT] = (SMU_THROTTLER_TDC_GFX_BIT), + [THROTTLER_TDC_SOC_BIT] = (SMU_THROTTLER_TDC_SOC_BIT), + [THROTTLER_TEMP_EDGE_BIT] = (SMU_THROTTLER_TEMP_EDGE_BIT), + [THROTTLER_TEMP_HOTSPOT_BIT] = (SMU_THROTTLER_TEMP_HOTSPOT_BIT), + [THROTTLER_TEMP_MEM_BIT] = (SMU_THROTTLER_TEMP_MEM_BIT), + [THROTTLER_TEMP_VR_GFX_BIT] = (SMU_THROTTLER_TEMP_VR_GFX_BIT), + [THROTTLER_TEMP_VR_SOC_BIT] = (SMU_THROTTLER_TEMP_VR_SOC_BIT), + [THROTTLER_TEMP_VR_MEM0_BIT] = (SMU_THROTTLER_TEMP_VR_MEM0_BIT), + [THROTTLER_TEMP_VR_MEM1_BIT] = (SMU_THROTTLER_TEMP_VR_MEM1_BIT), + [THROTTLER_TEMP_LIQUID0_BIT] = (SMU_THROTTLER_TEMP_LIQUID0_BIT), + [THROTTLER_TEMP_LIQUID1_BIT] = (SMU_THROTTLER_TEMP_LIQUID1_BIT), + [THROTTLER_GFX_APCC_PLUS_BIT] = (SMU_THROTTLER_APCC_BIT), + [THROTTLER_FIT_BIT] = (SMU_THROTTLER_FIT_BIT), +}; +#endif + +static int +smu_v14_0_2_get_allowed_feature_mask(struct smu_context *smu, + uint32_t *feature_mask, uint32_t num) +{ + struct amdgpu_device *adev = smu->adev; + /*u32 smu_version;*/ + + if (num > 2) + return -EINVAL; + + memset(feature_mask, 0xff, sizeof(uint32_t) * num); + + if (adev->pm.pp_feature & PP_SCLK_DPM_MASK) { + *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DPM_GFXCLK_BIT); + *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_GFX_IMU_BIT); + } +#if 0 + if (!(adev->pg_flags & AMD_PG_SUPPORT_ATHUB) || + !(adev->pg_flags & AMD_PG_SUPPORT_MMHUB)) + *(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_ATHUB_MMHUB_PG_BIT); + + if (!(adev->pm.pp_feature & PP_SOCCLK_DPM_MASK)) + *(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_DPM_SOCCLK_BIT); + + /* PMFW 78.58 contains a critical fix for gfxoff feature */ + smu_cmn_get_smc_version(smu, NULL, &smu_version); + if ((smu_version < 0x004e3a00) || + !(adev->pm.pp_feature & PP_GFXOFF_MASK)) + *(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_GFXOFF_BIT); + + if (!(adev->pm.pp_feature & PP_MCLK_DPM_MASK)) { + *(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_DPM_UCLK_BIT); + *(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_VMEMP_SCALING_BIT); + *(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_VDDIO_MEM_SCALING_BIT); + } + + if (!(adev->pm.pp_feature & PP_SCLK_DEEP_SLEEP_MASK)) + *(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_DS_GFXCLK_BIT); + + if (!(adev->pm.pp_feature & PP_PCIE_DPM_MASK)) { + *(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_DPM_LINK_BIT); + *(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_DS_LCLK_BIT); + } + + if (!(adev->pm.pp_feature & PP_ULV_MASK)) + *(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_GFX_ULV_BIT); +#endif + + return 0; +} + +static int smu_v14_0_2_check_powerplay_table(struct smu_context *smu) +{ + struct smu_table_context *table_context = &smu->smu_table; + struct smu_14_0_2_powerplay_table *powerplay_table = + table_context->power_play_table; + struct smu_baco_context *smu_baco = &smu->smu_baco; + PPTable_t *pptable = smu->smu_table.driver_pptable; + const OverDriveLimits_t * const overdrive_upperlimits = + &pptable->SkuTable.OverDriveLimitsBasicMax; + const OverDriveLimits_t * const overdrive_lowerlimits = + &pptable->SkuTable.OverDriveLimitsBasicMin; + + if (powerplay_table->platform_caps & SMU_14_0_2_PP_PLATFORM_CAP_HARDWAREDC) + smu->dc_controlled_by_gpio = true; + + if (powerplay_table->platform_caps & SMU_14_0_2_PP_PLATFORM_CAP_BACO) { + smu_baco->platform_support = true; + + if (powerplay_table->platform_caps & SMU_14_0_2_PP_PLATFORM_CAP_MACO) + smu_baco->maco_support = true; + } + + if (!overdrive_lowerlimits->FeatureCtrlMask || + !overdrive_upperlimits->FeatureCtrlMask) + smu->od_enabled = false; + + table_context->thermal_controller_type = + powerplay_table->thermal_controller_type; + + /* + * Instead of having its own buffer space and get overdrive_table copied, + * smu->od_settings just points to the actual overdrive_table + */ + smu->od_settings = &powerplay_table->overdrive_table; + + smu->adev->pm.no_fan = + !(pptable->PFE_Settings.FeaturesToRun[0] & (1 << FEATURE_FAN_CONTROL_BIT)); + + return 0; +} + +static int smu_v14_0_2_store_powerplay_table(struct smu_context *smu) +{ + struct smu_table_context *table_context = &smu->smu_table; + struct smu_14_0_2_powerplay_table *powerplay_table = + table_context->power_play_table; + + memcpy(table_context->driver_pptable, &powerplay_table->smc_pptable, + sizeof(PPTable_t)); + + return 0; +} + +#ifndef atom_smc_dpm_info_table_14_0_0 +struct atom_smc_dpm_info_table_14_0_0 { + struct atom_common_table_header table_header; + BoardTable_t BoardTable; +}; +#endif + +static int smu_v14_0_2_append_powerplay_table(struct smu_context *smu) +{ + struct smu_table_context *table_context = &smu->smu_table; + PPTable_t *smc_pptable = table_context->driver_pptable; + struct atom_smc_dpm_info_table_14_0_0 *smc_dpm_table; + BoardTable_t *BoardTable = &smc_pptable->BoardTable; + int index, ret; + + index = get_index_into_master_table(atom_master_list_of_data_tables_v2_1, + smc_dpm_info); + + ret = amdgpu_atombios_get_data_table(smu->adev, index, NULL, NULL, NULL, + (uint8_t **)&smc_dpm_table); + if (ret) + return ret; + + memcpy(BoardTable, &smc_dpm_table->BoardTable, sizeof(BoardTable_t)); + + return 0; +} + +#if 0 +static int smu_v14_0_2_get_pptable_from_pmfw(struct smu_context *smu, + void **table, + uint32_t *size) +{ + struct smu_table_context *smu_table = &smu->smu_table; + void *combo_pptable = smu_table->combo_pptable; + int ret = 0; + + ret = smu_cmn_get_combo_pptable(smu); + if (ret) + return ret; + + *table = combo_pptable; + *size = sizeof(struct smu_14_0_powerplay_table); + + return 0; +} +#endif + +static int smu_v14_0_2_get_pptable_from_pmfw(struct smu_context *smu, + void **table, + uint32_t *size) +{ + struct smu_table_context *smu_table = &smu->smu_table; + void *combo_pptable = smu_table->combo_pptable; + int ret = 0; + + ret = smu_cmn_get_combo_pptable(smu); + if (ret) + return ret; + + *table = combo_pptable; + *size = sizeof(struct smu_14_0_2_powerplay_table); + + return 0; +} + +static int smu_v14_0_2_setup_pptable(struct smu_context *smu) +{ + struct smu_table_context *smu_table = &smu->smu_table; + struct amdgpu_device *adev = smu->adev; + int ret = 0; + + if (amdgpu_sriov_vf(smu->adev)) + return 0; + + if (!adev->scpm_enabled) + ret = smu_v14_0_setup_pptable(smu); + else + ret = smu_v14_0_2_get_pptable_from_pmfw(smu, + &smu_table->power_play_table, + &smu_table->power_play_table_size); + if (ret) + return ret; + + ret = smu_v14_0_2_store_powerplay_table(smu); + if (ret) + return ret; + + /* + * With SCPM enabled, the operation below will be handled + * by PSP. Driver involvment is unnecessary and useless. + */ + if (!adev->scpm_enabled) { + ret = smu_v14_0_2_append_powerplay_table(smu); + if (ret) + return ret; + } + + ret = smu_v14_0_2_check_powerplay_table(smu); + if (ret) + return ret; + + return ret; +} + +static int smu_v14_0_2_tables_init(struct smu_context *smu) +{ + struct smu_table_context *smu_table = &smu->smu_table; + struct smu_table *tables = smu_table->tables; + + SMU_TABLE_INIT(tables, SMU_TABLE_PPTABLE, sizeof(PPTable_t), + PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM); + SMU_TABLE_INIT(tables, SMU_TABLE_WATERMARKS, sizeof(Watermarks_t), + PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM); + SMU_TABLE_INIT(tables, SMU_TABLE_SMU_METRICS, sizeof(SmuMetricsExternal_t), + PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM); + SMU_TABLE_INIT(tables, SMU_TABLE_I2C_COMMANDS, sizeof(SwI2cRequest_t), + PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM); + SMU_TABLE_INIT(tables, SMU_TABLE_OVERDRIVE, sizeof(OverDriveTable_t), + PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM); + SMU_TABLE_INIT(tables, SMU_TABLE_ACTIVITY_MONITOR_COEFF, + sizeof(DpmActivityMonitorCoeffIntExternal_t), PAGE_SIZE, + AMDGPU_GEM_DOMAIN_VRAM); + SMU_TABLE_INIT(tables, SMU_TABLE_COMBO_PPTABLE, MP0_MP1_DATA_REGION_SIZE_COMBOPPTABLE, + PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM); + SMU_TABLE_INIT(tables, SMU_TABLE_ECCINFO, sizeof(EccInfoTable_t), + PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM); + + smu_table->metrics_table = kzalloc(sizeof(SmuMetricsExternal_t), GFP_KERNEL); + if (!smu_table->metrics_table) + goto err0_out; + smu_table->metrics_time = 0; + + smu_table->gpu_metrics_table_size = sizeof(struct gpu_metrics_v1_3); + smu_table->gpu_metrics_table = kzalloc(smu_table->gpu_metrics_table_size, GFP_KERNEL); + if (!smu_table->gpu_metrics_table) + goto err1_out; + + smu_table->watermarks_table = kzalloc(sizeof(Watermarks_t), GFP_KERNEL); + if (!smu_table->watermarks_table) + goto err2_out; + + smu_table->ecc_table = kzalloc(tables[SMU_TABLE_ECCINFO].size, GFP_KERNEL); + if (!smu_table->ecc_table) + goto err3_out; + + return 0; + +err3_out: + kfree(smu_table->watermarks_table); +err2_out: + kfree(smu_table->gpu_metrics_table); +err1_out: + kfree(smu_table->metrics_table); +err0_out: + return -ENOMEM; +} + +static int smu_v14_0_2_allocate_dpm_context(struct smu_context *smu) +{ + struct smu_dpm_context *smu_dpm = &smu->smu_dpm; + + smu_dpm->dpm_context = kzalloc(sizeof(struct smu_14_0_dpm_context), + GFP_KERNEL); + if (!smu_dpm->dpm_context) + return -ENOMEM; + + smu_dpm->dpm_context_size = sizeof(struct smu_14_0_dpm_context); + + return 0; +} + +static int smu_v14_0_2_init_smc_tables(struct smu_context *smu) +{ + int ret = 0; + + ret = smu_v14_0_2_tables_init(smu); + if (ret) + return ret; + + ret = smu_v14_0_2_allocate_dpm_context(smu); + if (ret) + return ret; + + return smu_v14_0_init_smc_tables(smu); +} + +static int smu_v14_0_2_set_default_dpm_table(struct smu_context *smu) +{ + struct smu_14_0_dpm_context *dpm_context = smu->smu_dpm.dpm_context; + struct smu_table_context *table_context = &smu->smu_table; + PPTable_t *pptable = table_context->driver_pptable; + SkuTable_t *skutable = &pptable->SkuTable; + struct smu_14_0_dpm_table *dpm_table; + struct smu_14_0_pcie_table *pcie_table; + uint32_t link_level; + int ret = 0; + + /* socclk dpm table setup */ + dpm_table = &dpm_context->dpm_tables.soc_table; + if (smu_cmn_feature_is_enabled(smu, SMU_FEATURE_DPM_SOCCLK_BIT)) { + ret = smu_v14_0_set_single_dpm_table(smu, + SMU_SOCCLK, + dpm_table); + if (ret) + return ret; + } else { + dpm_table->count = 1; + dpm_table->dpm_levels[0].value = smu->smu_table.boot_values.socclk / 100; + dpm_table->dpm_levels[0].enabled = true; + dpm_table->min = dpm_table->dpm_levels[0].value; + dpm_table->max = dpm_table->dpm_levels[0].value; + } + + /* gfxclk dpm table setup */ + dpm_table = &dpm_context->dpm_tables.gfx_table; + if (smu_cmn_feature_is_enabled(smu, SMU_FEATURE_DPM_GFXCLK_BIT)) { + ret = smu_v14_0_set_single_dpm_table(smu, + SMU_GFXCLK, + dpm_table); + if (ret) + return ret; + + /* + * Update the reported maximum shader clock to the value + * which can be guarded to be achieved on all cards. This + * is aligned with Window setting. And considering that value + * might be not the peak frequency the card can achieve, it + * is normal some real-time clock frequency can overtake this + * labelled maximum clock frequency(for example in pp_dpm_sclk + * sysfs output). + */ + if (skutable->DriverReportedClocks.GameClockAc && + (dpm_table->dpm_levels[dpm_table->count - 1].value > + skutable->DriverReportedClocks.GameClockAc)) { + dpm_table->dpm_levels[dpm_table->count - 1].value = + skutable->DriverReportedClocks.GameClockAc; + dpm_table->max = skutable->DriverReportedClocks.GameClockAc; + } + } else { + dpm_table->count = 1; + dpm_table->dpm_levels[0].value = smu->smu_table.boot_values.gfxclk / 100; + dpm_table->dpm_levels[0].enabled = true; + dpm_table->min = dpm_table->dpm_levels[0].value; + dpm_table->max = dpm_table->dpm_levels[0].value; + } + + /* uclk dpm table setup */ + dpm_table = &dpm_context->dpm_tables.uclk_table; + if (smu_cmn_feature_is_enabled(smu, SMU_FEATURE_DPM_UCLK_BIT)) { + ret = smu_v14_0_set_single_dpm_table(smu, + SMU_UCLK, + dpm_table); + if (ret) + return ret; + } else { + dpm_table->count = 1; + dpm_table->dpm_levels[0].value = smu->smu_table.boot_values.uclk / 100; + dpm_table->dpm_levels[0].enabled = true; + dpm_table->min = dpm_table->dpm_levels[0].value; + dpm_table->max = dpm_table->dpm_levels[0].value; + } + + /* fclk dpm table setup */ + dpm_table = &dpm_context->dpm_tables.fclk_table; + if (smu_cmn_feature_is_enabled(smu, SMU_FEATURE_DPM_FCLK_BIT)) { + ret = smu_v14_0_set_single_dpm_table(smu, + SMU_FCLK, + dpm_table); + if (ret) + return ret; + } else { + dpm_table->count = 1; + dpm_table->dpm_levels[0].value = smu->smu_table.boot_values.fclk / 100; + dpm_table->dpm_levels[0].enabled = true; + dpm_table->min = dpm_table->dpm_levels[0].value; + dpm_table->max = dpm_table->dpm_levels[0].value; + } + + /* vclk dpm table setup */ + dpm_table = &dpm_context->dpm_tables.vclk_table; + if (smu_cmn_feature_is_enabled(smu, SMU_FEATURE_DPM_VCLK_BIT)) { + ret = smu_v14_0_set_single_dpm_table(smu, + SMU_VCLK, + dpm_table); + if (ret) + return ret; + } else { + dpm_table->count = 1; + dpm_table->dpm_levels[0].value = smu->smu_table.boot_values.vclk / 100; + dpm_table->dpm_levels[0].enabled = true; + dpm_table->min = dpm_table->dpm_levels[0].value; + dpm_table->max = dpm_table->dpm_levels[0].value; + } + + /* dclk dpm table setup */ + dpm_table = &dpm_context->dpm_tables.dclk_table; + if (smu_cmn_feature_is_enabled(smu, SMU_FEATURE_DPM_DCLK_BIT)) { + ret = smu_v14_0_set_single_dpm_table(smu, + SMU_DCLK, + dpm_table); + if (ret) + return ret; + } else { + dpm_table->count = 1; + dpm_table->dpm_levels[0].value = smu->smu_table.boot_values.dclk / 100; + dpm_table->dpm_levels[0].enabled = true; + dpm_table->min = dpm_table->dpm_levels[0].value; + dpm_table->max = dpm_table->dpm_levels[0].value; + } + + /* lclk dpm table setup */ + pcie_table = &dpm_context->dpm_tables.pcie_table; + pcie_table->num_of_link_levels = 0; + for (link_level = 0; link_level < NUM_LINK_LEVELS; link_level++) { + if (!skutable->PcieGenSpeed[link_level] && + !skutable->PcieLaneCount[link_level] && + !skutable->LclkFreq[link_level]) + continue; + + pcie_table->pcie_gen[pcie_table->num_of_link_levels] = + skutable->PcieGenSpeed[link_level]; + pcie_table->pcie_lane[pcie_table->num_of_link_levels] = + skutable->PcieLaneCount[link_level]; + pcie_table->clk_freq[pcie_table->num_of_link_levels] = + skutable->LclkFreq[link_level]; + pcie_table->num_of_link_levels++; + } + + return 0; +} + +static bool smu_v14_0_2_is_dpm_running(struct smu_context *smu) +{ + int ret = 0; + uint64_t feature_enabled; + + ret = smu_cmn_get_enabled_mask(smu, &feature_enabled); + if (ret) + return false; + + return !!(feature_enabled & SMC_DPM_FEATURE); +} + +static void smu_v14_0_2_dump_pptable(struct smu_context *smu) +{ + struct smu_table_context *table_context = &smu->smu_table; + PPTable_t *pptable = table_context->driver_pptable; + PFE_Settings_t *PFEsettings = &pptable->PFE_Settings; + + dev_info(smu->adev->dev, "Dumped PPTable:\n"); + + dev_info(smu->adev->dev, "Version = 0x%08x\n", PFEsettings->Version); + dev_info(smu->adev->dev, "FeaturesToRun[0] = 0x%08x\n", PFEsettings->FeaturesToRun[0]); + dev_info(smu->adev->dev, "FeaturesToRun[1] = 0x%08x\n", PFEsettings->FeaturesToRun[1]); +} + +static uint32_t smu_v14_0_2_get_throttler_status(SmuMetrics_t *metrics) +{ + uint32_t throttler_status = 0; + int i; + + for (i = 0; i < THROTTLER_COUNT; i++) + throttler_status |= + (metrics->ThrottlingPercentage[i] ? 1U << i : 0); + + return throttler_status; +} + +#define SMU_14_0_2_BUSY_THRESHOLD 5 +static int smu_v14_0_2_get_smu_metrics_data(struct smu_context *smu, + MetricsMember_t member, + uint32_t *value) +{ + struct smu_table_context *smu_table = &smu->smu_table; + SmuMetrics_t *metrics = + &(((SmuMetricsExternal_t *)(smu_table->metrics_table))->SmuMetrics); + int ret = 0; + + ret = smu_cmn_get_metrics_table(smu, + NULL, + false); + if (ret) + return ret; + + switch (member) { + case METRICS_CURR_GFXCLK: + *value = metrics->CurrClock[PPCLK_GFXCLK]; + break; + case METRICS_CURR_SOCCLK: + *value = metrics->CurrClock[PPCLK_SOCCLK]; + break; + case METRICS_CURR_UCLK: + *value = metrics->CurrClock[PPCLK_UCLK]; + break; + case METRICS_CURR_VCLK: + *value = metrics->CurrClock[PPCLK_VCLK_0]; + break; + case METRICS_CURR_DCLK: + *value = metrics->CurrClock[PPCLK_DCLK_0]; + break; + case METRICS_CURR_FCLK: + *value = metrics->CurrClock[PPCLK_FCLK]; + break; + case METRICS_CURR_DCEFCLK: + *value = metrics->CurrClock[PPCLK_DCFCLK]; + break; + case METRICS_AVERAGE_GFXCLK: + if (metrics->AverageGfxActivity <= SMU_14_0_2_BUSY_THRESHOLD) + *value = metrics->AverageGfxclkFrequencyPostDs; + else + *value = metrics->AverageGfxclkFrequencyPreDs; + break; + case METRICS_AVERAGE_FCLK: + if (metrics->AverageUclkActivity <= SMU_14_0_2_BUSY_THRESHOLD) + *value = metrics->AverageFclkFrequencyPostDs; + else + *value = metrics->AverageFclkFrequencyPreDs; + break; + case METRICS_AVERAGE_UCLK: + if (metrics->AverageUclkActivity <= SMU_14_0_2_BUSY_THRESHOLD) + *value = metrics->AverageMemclkFrequencyPostDs; + else + *value = metrics->AverageMemclkFrequencyPreDs; + break; + case METRICS_AVERAGE_VCLK: + *value = metrics->AverageVclk0Frequency; + break; + case METRICS_AVERAGE_DCLK: + *value = metrics->AverageDclk0Frequency; + break; + case METRICS_AVERAGE_VCLK1: + *value = metrics->AverageVclk1Frequency; + break; + case METRICS_AVERAGE_DCLK1: + *value = metrics->AverageDclk1Frequency; + break; + case METRICS_AVERAGE_GFXACTIVITY: + *value = metrics->AverageGfxActivity; + break; + case METRICS_AVERAGE_MEMACTIVITY: + *value = metrics->AverageUclkActivity; + break; + case METRICS_AVERAGE_SOCKETPOWER: + *value = metrics->AverageSocketPower << 8; + break; + case METRICS_TEMPERATURE_EDGE: + *value = metrics->AvgTemperature[TEMP_EDGE] * + SMU_TEMPERATURE_UNITS_PER_CENTIGRADES; + break; + case METRICS_TEMPERATURE_HOTSPOT: + *value = metrics->AvgTemperature[TEMP_HOTSPOT] * + SMU_TEMPERATURE_UNITS_PER_CENTIGRADES; + break; + case METRICS_TEMPERATURE_MEM: + *value = metrics->AvgTemperature[TEMP_MEM] * + SMU_TEMPERATURE_UNITS_PER_CENTIGRADES; + break; + case METRICS_TEMPERATURE_VRGFX: + *value = metrics->AvgTemperature[TEMP_VR_GFX] * + SMU_TEMPERATURE_UNITS_PER_CENTIGRADES; + break; + case METRICS_TEMPERATURE_VRSOC: + *value = metrics->AvgTemperature[TEMP_VR_SOC] * + SMU_TEMPERATURE_UNITS_PER_CENTIGRADES; + break; + case METRICS_THROTTLER_STATUS: + *value = smu_v14_0_2_get_throttler_status(metrics); + break; + case METRICS_CURR_FANSPEED: + *value = metrics->AvgFanRpm; + break; + case METRICS_CURR_FANPWM: + *value = metrics->AvgFanPwm; + break; + case METRICS_VOLTAGE_VDDGFX: + *value = metrics->AvgVoltage[SVI_PLANE_VDD_GFX]; + break; + case METRICS_PCIE_RATE: + *value = metrics->PcieRate; + break; + case METRICS_PCIE_WIDTH: + *value = metrics->PcieWidth; + break; + default: + *value = UINT_MAX; + break; + } + + return ret; +} + +static int smu_v14_0_2_get_dpm_ultimate_freq(struct smu_context *smu, + enum smu_clk_type clk_type, + uint32_t *min, + uint32_t *max) +{ + struct smu_14_0_dpm_context *dpm_context = + smu->smu_dpm.dpm_context; + struct smu_14_0_dpm_table *dpm_table; + + switch (clk_type) { + case SMU_MCLK: + case SMU_UCLK: + /* uclk dpm table */ + dpm_table = &dpm_context->dpm_tables.uclk_table; + break; + case SMU_GFXCLK: + case SMU_SCLK: + /* gfxclk dpm table */ + dpm_table = &dpm_context->dpm_tables.gfx_table; + break; + case SMU_SOCCLK: + /* socclk dpm table */ + dpm_table = &dpm_context->dpm_tables.soc_table; + break; + case SMU_FCLK: + /* fclk dpm table */ + dpm_table = &dpm_context->dpm_tables.fclk_table; + break; + case SMU_VCLK: + case SMU_VCLK1: + /* vclk dpm table */ + dpm_table = &dpm_context->dpm_tables.vclk_table; + break; + case SMU_DCLK: + case SMU_DCLK1: + /* dclk dpm table */ + dpm_table = &dpm_context->dpm_tables.dclk_table; + break; + default: + dev_err(smu->adev->dev, "Unsupported clock type!\n"); + return -EINVAL; + } + + if (min) + *min = dpm_table->min; + if (max) + *max = dpm_table->max; + + return 0; +} + +static int smu_v14_0_2_read_sensor(struct smu_context *smu, + enum amd_pp_sensors sensor, + void *data, + uint32_t *size) +{ + struct smu_table_context *table_context = &smu->smu_table; + PPTable_t *smc_pptable = table_context->driver_pptable; + int ret = 0; + + switch (sensor) { + case AMDGPU_PP_SENSOR_MAX_FAN_RPM: + *(uint16_t *)data = smc_pptable->CustomSkuTable.FanMaximumRpm; + *size = 4; + break; + case AMDGPU_PP_SENSOR_MEM_LOAD: + ret = smu_v14_0_2_get_smu_metrics_data(smu, + METRICS_AVERAGE_MEMACTIVITY, + (uint32_t *)data); + *size = 4; + break; + case AMDGPU_PP_SENSOR_GPU_LOAD: + ret = smu_v14_0_2_get_smu_metrics_data(smu, + METRICS_AVERAGE_GFXACTIVITY, + (uint32_t *)data); + *size = 4; + break; + case AMDGPU_PP_SENSOR_GPU_AVG_POWER: + ret = smu_v14_0_2_get_smu_metrics_data(smu, + METRICS_AVERAGE_SOCKETPOWER, + (uint32_t *)data); + *size = 4; + break; + case AMDGPU_PP_SENSOR_HOTSPOT_TEMP: + ret = smu_v14_0_2_get_smu_metrics_data(smu, + METRICS_TEMPERATURE_HOTSPOT, + (uint32_t *)data); + *size = 4; + break; + case AMDGPU_PP_SENSOR_EDGE_TEMP: + ret = smu_v14_0_2_get_smu_metrics_data(smu, + METRICS_TEMPERATURE_EDGE, + (uint32_t *)data); + *size = 4; + break; + case AMDGPU_PP_SENSOR_MEM_TEMP: + ret = smu_v14_0_2_get_smu_metrics_data(smu, + METRICS_TEMPERATURE_MEM, + (uint32_t *)data); + *size = 4; + break; + case AMDGPU_PP_SENSOR_GFX_MCLK: + ret = smu_v14_0_2_get_smu_metrics_data(smu, + METRICS_CURR_UCLK, + (uint32_t *)data); + *(uint32_t *)data *= 100; + *size = 4; + break; + case AMDGPU_PP_SENSOR_GFX_SCLK: + ret = smu_v14_0_2_get_smu_metrics_data(smu, + METRICS_AVERAGE_GFXCLK, + (uint32_t *)data); + *(uint32_t *)data *= 100; + *size = 4; + break; + case AMDGPU_PP_SENSOR_VDDGFX: + ret = smu_v14_0_2_get_smu_metrics_data(smu, + METRICS_VOLTAGE_VDDGFX, + (uint32_t *)data); + *size = 4; + break; + default: + ret = -EOPNOTSUPP; + break; + } + + return ret; +} + +static int smu_v14_0_2_get_current_clk_freq_by_table(struct smu_context *smu, + enum smu_clk_type clk_type, + uint32_t *value) +{ + MetricsMember_t member_type; + int clk_id = 0; + + clk_id = smu_cmn_to_asic_specific_index(smu, + CMN2ASIC_MAPPING_CLK, + clk_type); + if (clk_id < 0) + return -EINVAL; + + switch (clk_id) { + case PPCLK_GFXCLK: + member_type = METRICS_AVERAGE_GFXCLK; + break; + case PPCLK_UCLK: + member_type = METRICS_CURR_UCLK; + break; + case PPCLK_FCLK: + member_type = METRICS_CURR_FCLK; + break; + case PPCLK_SOCCLK: + member_type = METRICS_CURR_SOCCLK; + break; + case PPCLK_VCLK_0: + member_type = METRICS_AVERAGE_VCLK; + break; + case PPCLK_DCLK_0: + member_type = METRICS_AVERAGE_DCLK; + break; + default: + return -EINVAL; + } + + return smu_v14_0_2_get_smu_metrics_data(smu, + member_type, + value); +} + +static int smu_v14_0_2_print_clk_levels(struct smu_context *smu, + enum smu_clk_type clk_type, + char *buf) +{ + struct smu_dpm_context *smu_dpm = &smu->smu_dpm; + struct smu_14_0_dpm_context *dpm_context = smu_dpm->dpm_context; + struct smu_14_0_dpm_table *single_dpm_table; + int i, curr_freq, size = 0; + int ret = 0; + + smu_cmn_get_sysfs_buf(&buf, &size); + + if (amdgpu_ras_intr_triggered()) { + size += sysfs_emit_at(buf, size, "unavailable\n"); + return size; + } + + switch (clk_type) { + case SMU_SCLK: + single_dpm_table = &(dpm_context->dpm_tables.gfx_table); + break; + case SMU_MCLK: + single_dpm_table = &(dpm_context->dpm_tables.uclk_table); + break; + case SMU_SOCCLK: + single_dpm_table = &(dpm_context->dpm_tables.soc_table); + break; + case SMU_FCLK: + single_dpm_table = &(dpm_context->dpm_tables.fclk_table); + break; + case SMU_VCLK: + case SMU_VCLK1: + single_dpm_table = &(dpm_context->dpm_tables.vclk_table); + break; + case SMU_DCLK: + case SMU_DCLK1: + single_dpm_table = &(dpm_context->dpm_tables.dclk_table); + break; + default: + break; + } + + switch (clk_type) { + case SMU_SCLK: + case SMU_MCLK: + case SMU_SOCCLK: + case SMU_FCLK: + case SMU_VCLK: + case SMU_VCLK1: + case SMU_DCLK: + case SMU_DCLK1: + ret = smu_v14_0_2_get_current_clk_freq_by_table(smu, clk_type, &curr_freq); + if (ret) { + dev_err(smu->adev->dev, "Failed to get current clock freq!"); + return ret; + } + + if (single_dpm_table->is_fine_grained) { + /* + * For fine grained dpms, there are only two dpm levels: + * - level 0 -> min clock freq + * - level 1 -> max clock freq + * And the current clock frequency can be any value between them. + * So, if the current clock frequency is not at level 0 or level 1, + * we will fake it as three dpm levels: + * - level 0 -> min clock freq + * - level 1 -> current actual clock freq + * - level 2 -> max clock freq + */ + if ((single_dpm_table->dpm_levels[0].value != curr_freq) && + (single_dpm_table->dpm_levels[1].value != curr_freq)) { + size += sysfs_emit_at(buf, size, "0: %uMhz\n", + single_dpm_table->dpm_levels[0].value); + size += sysfs_emit_at(buf, size, "1: %uMhz *\n", + curr_freq); + size += sysfs_emit_at(buf, size, "2: %uMhz\n", + single_dpm_table->dpm_levels[1].value); + } else { + size += sysfs_emit_at(buf, size, "0: %uMhz %s\n", + single_dpm_table->dpm_levels[0].value, + single_dpm_table->dpm_levels[0].value == curr_freq ? "*" : ""); + size += sysfs_emit_at(buf, size, "1: %uMhz %s\n", + single_dpm_table->dpm_levels[1].value, + single_dpm_table->dpm_levels[1].value == curr_freq ? "*" : ""); + } + } else { + for (i = 0; i < single_dpm_table->count; i++) + size += sysfs_emit_at(buf, size, "%d: %uMhz %s\n", + i, single_dpm_table->dpm_levels[i].value, + single_dpm_table->dpm_levels[i].value == curr_freq ? "*" : ""); + } + break; + case SMU_PCIE: + // TODO + break; + + default: + break; + } + + return size; +} + +static int smu_v14_0_2_force_clk_levels(struct smu_context *smu, + enum smu_clk_type clk_type, + uint32_t mask) +{ + struct smu_dpm_context *smu_dpm = &smu->smu_dpm; + struct smu_14_0_dpm_context *dpm_context = smu_dpm->dpm_context; + struct smu_14_0_dpm_table *single_dpm_table; + uint32_t soft_min_level, soft_max_level; + uint32_t min_freq, max_freq; + int ret = 0; + + soft_min_level = mask ? (ffs(mask) - 1) : 0; + soft_max_level = mask ? (fls(mask) - 1) : 0; + + switch (clk_type) { + case SMU_GFXCLK: + case SMU_SCLK: + single_dpm_table = &(dpm_context->dpm_tables.gfx_table); + break; + case SMU_MCLK: + case SMU_UCLK: + single_dpm_table = &(dpm_context->dpm_tables.uclk_table); + break; + case SMU_SOCCLK: + single_dpm_table = &(dpm_context->dpm_tables.soc_table); + break; + case SMU_FCLK: + single_dpm_table = &(dpm_context->dpm_tables.fclk_table); + break; + case SMU_VCLK: + case SMU_VCLK1: + single_dpm_table = &(dpm_context->dpm_tables.vclk_table); + break; + case SMU_DCLK: + case SMU_DCLK1: + single_dpm_table = &(dpm_context->dpm_tables.dclk_table); + break; + default: + break; + } + + switch (clk_type) { + case SMU_GFXCLK: + case SMU_SCLK: + case SMU_MCLK: + case SMU_UCLK: + case SMU_SOCCLK: + case SMU_FCLK: + case SMU_VCLK: + case SMU_VCLK1: + case SMU_DCLK: + case SMU_DCLK1: + if (single_dpm_table->is_fine_grained) { + /* There is only 2 levels for fine grained DPM */ + soft_max_level = (soft_max_level >= 1 ? 1 : 0); + soft_min_level = (soft_min_level >= 1 ? 1 : 0); + } else { + if ((soft_max_level >= single_dpm_table->count) || + (soft_min_level >= single_dpm_table->count)) + return -EINVAL; + } + + min_freq = single_dpm_table->dpm_levels[soft_min_level].value; + max_freq = single_dpm_table->dpm_levels[soft_max_level].value; + + ret = smu_v14_0_set_soft_freq_limited_range(smu, + clk_type, + min_freq, + max_freq); + break; + case SMU_DCEFCLK: + case SMU_PCIE: + default: + break; + } + + return ret; +} + +static int smu_v14_0_2_update_pcie_parameters(struct smu_context *smu, + uint8_t pcie_gen_cap, + uint8_t pcie_width_cap) +{ + struct smu_14_0_dpm_context *dpm_context = smu->smu_dpm.dpm_context; + struct smu_14_0_pcie_table *pcie_table = + &dpm_context->dpm_tables.pcie_table; + uint32_t smu_pcie_arg; + int ret, i; + + for (i = 0; i < pcie_table->num_of_link_levels; i++) { + if (pcie_table->pcie_gen[i] > pcie_gen_cap) + pcie_table->pcie_gen[i] = pcie_gen_cap; + if (pcie_table->pcie_lane[i] > pcie_width_cap) + pcie_table->pcie_lane[i] = pcie_width_cap; + + smu_pcie_arg = i << 16; + smu_pcie_arg |= pcie_table->pcie_gen[i] << 8; + smu_pcie_arg |= pcie_table->pcie_lane[i]; + + ret = smu_cmn_send_smc_msg_with_param(smu, + SMU_MSG_OverridePcieParameters, + smu_pcie_arg, + NULL); + if (ret) + return ret; + } + + return 0; +} + +static int smu_v14_0_2_get_thermal_temperature_range(struct smu_context *smu, + struct smu_temperature_range *range) +{ + // TODO + + return 0; +} + +static int smu_v14_0_2_populate_umd_state_clk(struct smu_context *smu) +{ + // TODO + + return 0; +} + +static void smu_v14_0_2_get_unique_id(struct smu_context *smu) +{ + struct smu_table_context *smu_table = &smu->smu_table; + SmuMetrics_t *metrics = + &(((SmuMetricsExternal_t *)(smu_table->metrics_table))->SmuMetrics); + struct amdgpu_device *adev = smu->adev; + uint32_t upper32 = 0, lower32 = 0; + int ret; + + ret = smu_cmn_get_metrics_table(smu, NULL, false); + if (ret) + goto out; + + upper32 = metrics->PublicSerialNumberUpper; + lower32 = metrics->PublicSerialNumberLower; + +out: + adev->unique_id = ((uint64_t)upper32 << 32) | lower32; +} + +static int smu_v14_0_2_get_power_limit(struct smu_context *smu, + uint32_t *current_power_limit, + uint32_t *default_power_limit, + uint32_t *max_power_limit, + uint32_t *min_power_limit) +{ + // TODO + + return 0; +} + +static int smu_v14_0_2_get_power_profile_mode(struct smu_context *smu, + char *buf) +{ + DpmActivityMonitorCoeffIntExternal_t activity_monitor_external; + DpmActivityMonitorCoeffInt_t *activity_monitor = + &(activity_monitor_external.DpmActivityMonitorCoeffInt); + static const char *title[] = { + "PROFILE_INDEX(NAME)", + "CLOCK_TYPE(NAME)", + "FPS", + "MinActiveFreqType", + "MinActiveFreq", + "BoosterFreqType", + "BoosterFreq", + "PD_Data_limit_c", + "PD_Data_error_coeff", + "PD_Data_error_rate_coeff"}; + int16_t workload_type = 0; + uint32_t i, size = 0; + int result = 0; + + if (!buf) + return -EINVAL; + + size += sysfs_emit_at(buf, size, "%16s %s %s %s %s %s %s %s %s %s\n", + title[0], title[1], title[2], title[3], title[4], title[5], + title[6], title[7], title[8], title[9]); + + for (i = 0; i < PP_SMC_POWER_PROFILE_COUNT; i++) { + /* conv PP_SMC_POWER_PROFILE* to WORKLOAD_PPLIB_*_BIT */ + workload_type = smu_cmn_to_asic_specific_index(smu, + CMN2ASIC_MAPPING_WORKLOAD, + i); + if (workload_type == -ENOTSUPP) + continue; + else if (workload_type < 0) + return -EINVAL; + + result = smu_cmn_update_table(smu, + SMU_TABLE_ACTIVITY_MONITOR_COEFF, + workload_type, + (void *)(&activity_monitor_external), + false); + if (result) { + dev_err(smu->adev->dev, "[%s] Failed to get activity monitor!", __func__); + return result; + } + + size += sysfs_emit_at(buf, size, "%2d %14s%s:\n", + i, amdgpu_pp_profile_name[i], (i == smu->power_profile_mode) ? "*" : " "); + + size += sysfs_emit_at(buf, size, "%19s %d(%13s) %7d %7d %7d %7d %7d %7d %7d %7d\n", + " ", + 0, + "GFXCLK", + activity_monitor->Gfx_FPS, + activity_monitor->Gfx_MinActiveFreqType, + activity_monitor->Gfx_MinActiveFreq, + activity_monitor->Gfx_BoosterFreqType, + activity_monitor->Gfx_BoosterFreq, + activity_monitor->Gfx_PD_Data_limit_c, + activity_monitor->Gfx_PD_Data_error_coeff, + activity_monitor->Gfx_PD_Data_error_rate_coeff); + + size += sysfs_emit_at(buf, size, "%19s %d(%13s) %7d %7d %7d %7d %7d %7d %7d %7d\n", + " ", + 1, + "FCLK", + activity_monitor->Fclk_FPS, + activity_monitor->Fclk_MinActiveFreqType, + activity_monitor->Fclk_MinActiveFreq, + activity_monitor->Fclk_BoosterFreqType, + activity_monitor->Fclk_BoosterFreq, + activity_monitor->Fclk_PD_Data_limit_c, + activity_monitor->Fclk_PD_Data_error_coeff, + activity_monitor->Fclk_PD_Data_error_rate_coeff); + } + + return size; +} + +static int smu_v14_0_2_set_power_profile_mode(struct smu_context *smu, + long *input, + uint32_t size) +{ + DpmActivityMonitorCoeffIntExternal_t activity_monitor_external; + DpmActivityMonitorCoeffInt_t *activity_monitor = + &(activity_monitor_external.DpmActivityMonitorCoeffInt); + int workload_type, ret = 0; + + smu->power_profile_mode = input[size]; + + if (smu->power_profile_mode >= PP_SMC_POWER_PROFILE_COUNT) { + dev_err(smu->adev->dev, "Invalid power profile mode %d\n", smu->power_profile_mode); + return -EINVAL; + } + + if (smu->power_profile_mode == PP_SMC_POWER_PROFILE_CUSTOM) { + ret = smu_cmn_update_table(smu, + SMU_TABLE_ACTIVITY_MONITOR_COEFF, + WORKLOAD_PPLIB_CUSTOM_BIT, + (void *)(&activity_monitor_external), + false); + if (ret) { + dev_err(smu->adev->dev, "[%s] Failed to get activity monitor!", __func__); + return ret; + } + + switch (input[0]) { + case 0: /* Gfxclk */ + activity_monitor->Gfx_FPS = input[1]; + activity_monitor->Gfx_MinActiveFreqType = input[2]; + activity_monitor->Gfx_MinActiveFreq = input[3]; + activity_monitor->Gfx_BoosterFreqType = input[4]; + activity_monitor->Gfx_BoosterFreq = input[5]; + activity_monitor->Gfx_PD_Data_limit_c = input[6]; + activity_monitor->Gfx_PD_Data_error_coeff = input[7]; + activity_monitor->Gfx_PD_Data_error_rate_coeff = input[8]; + break; + case 1: /* Fclk */ + activity_monitor->Fclk_FPS = input[1]; + activity_monitor->Fclk_MinActiveFreqType = input[2]; + activity_monitor->Fclk_MinActiveFreq = input[3]; + activity_monitor->Fclk_BoosterFreqType = input[4]; + activity_monitor->Fclk_BoosterFreq = input[5]; + activity_monitor->Fclk_PD_Data_limit_c = input[6]; + activity_monitor->Fclk_PD_Data_error_coeff = input[7]; + activity_monitor->Fclk_PD_Data_error_rate_coeff = input[8]; + break; + } + + ret = smu_cmn_update_table(smu, + SMU_TABLE_ACTIVITY_MONITOR_COEFF, + WORKLOAD_PPLIB_CUSTOM_BIT, + (void *)(&activity_monitor_external), + true); + if (ret) { + dev_err(smu->adev->dev, "[%s] Failed to set activity monitor!", __func__); + return ret; + } + } + + /* conv PP_SMC_POWER_PROFILE* to WORKLOAD_PPLIB_*_BIT */ + workload_type = smu_cmn_to_asic_specific_index(smu, + CMN2ASIC_MAPPING_WORKLOAD, + smu->power_profile_mode); + if (workload_type < 0) + return -EINVAL; + + return smu_cmn_send_smc_msg_with_param(smu, + SMU_MSG_SetWorkloadMask, + 1 << workload_type, + NULL); +} + +static int smu_v14_0_2_baco_enter(struct smu_context *smu) +{ + struct smu_baco_context *smu_baco = &smu->smu_baco; + struct amdgpu_device *adev = smu->adev; + + if (adev->in_runpm && smu_cmn_is_audio_func_enabled(adev)) + return smu_v14_0_baco_set_armd3_sequence(smu, + smu_baco->maco_support ? BACO_SEQ_BAMACO : BACO_SEQ_BACO); + else + return smu_v14_0_baco_enter(smu); +} + +static int smu_v14_0_2_baco_exit(struct smu_context *smu) +{ + struct amdgpu_device *adev = smu->adev; + + if (adev->in_runpm && smu_cmn_is_audio_func_enabled(adev)) { + /* Wait for PMFW handling for the Dstate change */ + usleep_range(10000, 11000); + return smu_v14_0_baco_set_armd3_sequence(smu, BACO_SEQ_ULPS); + } else { + return smu_v14_0_baco_exit(smu); + } +} + +static bool smu_v14_0_2_is_mode1_reset_supported(struct smu_context *smu) +{ + // TODO + + return true; +} + +static int smu_v14_0_2_i2c_xfer(struct i2c_adapter *i2c_adap, + struct i2c_msg *msg, int num_msgs) +{ + struct amdgpu_smu_i2c_bus *smu_i2c = i2c_get_adapdata(i2c_adap); + struct amdgpu_device *adev = smu_i2c->adev; + struct smu_context *smu = adev->powerplay.pp_handle; + struct smu_table_context *smu_table = &smu->smu_table; + struct smu_table *table = &smu_table->driver_table; + SwI2cRequest_t *req, *res = (SwI2cRequest_t *)table->cpu_addr; + int i, j, r, c; + u16 dir; + + if (!adev->pm.dpm_enabled) + return -EBUSY; + + req = kzalloc(sizeof(*req), GFP_KERNEL); + if (!req) + return -ENOMEM; + + req->I2CcontrollerPort = smu_i2c->port; + req->I2CSpeed = I2C_SPEED_FAST_400K; + req->SlaveAddress = msg[0].addr << 1; /* wants an 8-bit address */ + dir = msg[0].flags & I2C_M_RD; + + for (c = i = 0; i < num_msgs; i++) { + for (j = 0; j < msg[i].len; j++, c++) { + SwI2cCmd_t *cmd = &req->SwI2cCmds[c]; + + if (!(msg[i].flags & I2C_M_RD)) { + /* write */ + cmd->CmdConfig |= CMDCONFIG_READWRITE_MASK; + cmd->ReadWriteData = msg[i].buf[j]; + } + + if ((dir ^ msg[i].flags) & I2C_M_RD) { + /* The direction changes. + */ + dir = msg[i].flags & I2C_M_RD; + cmd->CmdConfig |= CMDCONFIG_RESTART_MASK; + } + + req->NumCmds++; + + /* + * Insert STOP if we are at the last byte of either last + * message for the transaction or the client explicitly + * requires a STOP at this particular message. + */ + if ((j == msg[i].len - 1) && + ((i == num_msgs - 1) || (msg[i].flags & I2C_M_STOP))) { + cmd->CmdConfig &= ~CMDCONFIG_RESTART_MASK; + cmd->CmdConfig |= CMDCONFIG_STOP_MASK; + } + } + } + mutex_lock(&adev->pm.mutex); + r = smu_cmn_update_table(smu, SMU_TABLE_I2C_COMMANDS, 0, req, true); + mutex_unlock(&adev->pm.mutex); + if (r) + goto fail; + + for (c = i = 0; i < num_msgs; i++) { + if (!(msg[i].flags & I2C_M_RD)) { + c += msg[i].len; + continue; + } + for (j = 0; j < msg[i].len; j++, c++) { + SwI2cCmd_t *cmd = &res->SwI2cCmds[c]; + + msg[i].buf[j] = cmd->ReadWriteData; + } + } + r = num_msgs; +fail: + kfree(req); + return r; +} + +static u32 smu_v14_0_2_i2c_func(struct i2c_adapter *adap) +{ + return I2C_FUNC_I2C | I2C_FUNC_SMBUS_EMUL; +} + +static const struct i2c_algorithm smu_v14_0_2_i2c_algo = { + .master_xfer = smu_v14_0_2_i2c_xfer, + .functionality = smu_v14_0_2_i2c_func, +}; + +static const struct i2c_adapter_quirks smu_v14_0_2_i2c_control_quirks = { + .flags = I2C_AQ_COMB | I2C_AQ_COMB_SAME_ADDR | I2C_AQ_NO_ZERO_LEN, + .max_read_len = MAX_SW_I2C_COMMANDS, + .max_write_len = MAX_SW_I2C_COMMANDS, + .max_comb_1st_msg_len = 2, + .max_comb_2nd_msg_len = MAX_SW_I2C_COMMANDS - 2, +}; + +static int smu_v14_0_2_i2c_control_init(struct smu_context *smu) +{ + struct amdgpu_device *adev = smu->adev; + int res, i; + + for (i = 0; i < MAX_SMU_I2C_BUSES; i++) { + struct amdgpu_smu_i2c_bus *smu_i2c = &adev->pm.smu_i2c[i]; + struct i2c_adapter *control = &smu_i2c->adapter; + + smu_i2c->adev = adev; + smu_i2c->port = i; + mutex_init(&smu_i2c->mutex); + control->owner = THIS_MODULE; + control->dev.parent = &adev->pdev->dev; + control->algo = &smu_v14_0_2_i2c_algo; + snprintf(control->name, sizeof(control->name), "AMDGPU SMU %d", i); + control->quirks = &smu_v14_0_2_i2c_control_quirks; + i2c_set_adapdata(control, smu_i2c); + + res = i2c_add_adapter(control); + if (res) { + DRM_ERROR("Failed to register hw i2c, err: %d\n", res); + goto Out_err; + } + } + + /* assign the buses used for the FRU EEPROM and RAS EEPROM */ + /* XXX ideally this would be something in a vbios data table */ + adev->pm.ras_eeprom_i2c_bus = &adev->pm.smu_i2c[1].adapter; + adev->pm.fru_eeprom_i2c_bus = &adev->pm.smu_i2c[0].adapter; + + return 0; +Out_err: + for ( ; i >= 0; i--) { + struct amdgpu_smu_i2c_bus *smu_i2c = &adev->pm.smu_i2c[i]; + struct i2c_adapter *control = &smu_i2c->adapter; + + i2c_del_adapter(control); + } + return res; +} + +static void smu_v14_0_2_i2c_control_fini(struct smu_context *smu) +{ + struct amdgpu_device *adev = smu->adev; + int i; + + for (i = 0; i < MAX_SMU_I2C_BUSES; i++) { + struct amdgpu_smu_i2c_bus *smu_i2c = &adev->pm.smu_i2c[i]; + struct i2c_adapter *control = &smu_i2c->adapter; + + i2c_del_adapter(control); + } + adev->pm.ras_eeprom_i2c_bus = NULL; + adev->pm.fru_eeprom_i2c_bus = NULL; +} + +static int smu_v14_0_2_set_mp1_state(struct smu_context *smu, + enum pp_mp1_state mp1_state) +{ + int ret; + + switch (mp1_state) { + case PP_MP1_STATE_UNLOAD: + ret = smu_cmn_set_mp1_state(smu, mp1_state); + break; + default: + /* Ignore others */ + ret = 0; + } + + return ret; +} + +static int smu_v14_0_2_set_df_cstate(struct smu_context *smu, + enum pp_df_cstate state) +{ + return smu_cmn_send_smc_msg_with_param(smu, + SMU_MSG_DFCstateControl, + state, + NULL); +} + +static int smu_v14_0_2_mode1_reset(struct smu_context *smu) +{ + int ret = 0; + + // TODO + + return ret; +} + +static int smu_v14_0_2_mode2_reset(struct smu_context *smu) +{ + int ret = 0; + + // TODO + + return ret; +} + +static int smu_v14_0_2_enable_gfx_features(struct smu_context *smu) +{ + struct amdgpu_device *adev = smu->adev; + + if (adev->ip_versions[MP1_HWIP][0] == IP_VERSION(14, 0, 2)) + return smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_EnableAllSmuFeatures, + FEATURE_PWR_GFX, NULL); + else + return -EOPNOTSUPP; +} + +static void smu_v14_0_2_set_smu_mailbox_registers(struct smu_context *smu) +{ + struct amdgpu_device *adev = smu->adev; + + smu->param_reg = SOC15_REG_OFFSET(MP1, 0, regMP1_SMN_C2PMSG_82); + smu->msg_reg = SOC15_REG_OFFSET(MP1, 0, regMP1_SMN_C2PMSG_66); + smu->resp_reg = SOC15_REG_OFFSET(MP1, 0, regMP1_SMN_C2PMSG_90); +} + +static int smu_v14_0_2_smu_send_bad_mem_page_num(struct smu_context *smu, + uint32_t size) +{ + int ret = 0; + + /* message SMU to update the bad page number on SMUBUS */ + ret = smu_cmn_send_smc_msg_with_param(smu, + SMU_MSG_SetNumBadMemoryPagesRetired, + size, NULL); + if (ret) + dev_err(smu->adev->dev, + "[%s] failed to message SMU to update bad memory pages number\n", + __func__); + + return ret; +} + +static int smu_v14_0_2_send_bad_mem_channel_flag(struct smu_context *smu, + uint32_t size) +{ + int ret = 0; + + /* message SMU to update the bad channel info on SMUBUS */ + ret = smu_cmn_send_smc_msg_with_param(smu, + SMU_MSG_SetBadMemoryPagesRetiredFlagsPerChannel, + size, NULL); + if (ret) + dev_err(smu->adev->dev, + "[%s] failed to message SMU to update bad memory pages channel info\n", + __func__); + + return ret; +} + +static ssize_t smu_v14_0_2_get_ecc_info(struct smu_context *smu, + void *table) +{ + int ret = 0; + + // TODO + + return ret; +} + +static const struct pptable_funcs smu_v14_0_2_ppt_funcs = { + .get_allowed_feature_mask = smu_v14_0_2_get_allowed_feature_mask, + .set_default_dpm_table = smu_v14_0_2_set_default_dpm_table, + .i2c_init = smu_v14_0_2_i2c_control_init, + .i2c_fini = smu_v14_0_2_i2c_control_fini, + .is_dpm_running = smu_v14_0_2_is_dpm_running, + .dump_pptable = smu_v14_0_2_dump_pptable, + .init_microcode = smu_v14_0_init_microcode, + .load_microcode = smu_v14_0_load_microcode, + .fini_microcode = smu_v14_0_fini_microcode, + .init_smc_tables = smu_v14_0_2_init_smc_tables, + .fini_smc_tables = smu_v14_0_fini_smc_tables, + .init_power = smu_v14_0_init_power, + .fini_power = smu_v14_0_fini_power, + .check_fw_status = smu_v14_0_check_fw_status, + .setup_pptable = smu_v14_0_2_setup_pptable, + .check_fw_version = smu_v14_0_check_fw_version, + .write_pptable = smu_cmn_write_pptable, + .set_driver_table_location = smu_v14_0_set_driver_table_location, + .system_features_control = smu_v14_0_system_features_control, + .set_allowed_mask = smu_v14_0_set_allowed_mask, + .get_enabled_mask = smu_cmn_get_enabled_mask, + .dpm_set_vcn_enable = smu_v14_0_set_vcn_enable, + .dpm_set_jpeg_enable = smu_v14_0_set_jpeg_enable, + .get_dpm_ultimate_freq = smu_v14_0_2_get_dpm_ultimate_freq, + .get_vbios_bootup_values = smu_v14_0_get_vbios_bootup_values, + .read_sensor = smu_v14_0_2_read_sensor, + .feature_is_enabled = smu_cmn_feature_is_enabled, + .print_clk_levels = smu_v14_0_2_print_clk_levels, + .force_clk_levels = smu_v14_0_2_force_clk_levels, + .update_pcie_parameters = smu_v14_0_2_update_pcie_parameters, + .get_thermal_temperature_range = smu_v14_0_2_get_thermal_temperature_range, + .register_irq_handler = smu_v14_0_register_irq_handler, + .notify_memory_pool_location = smu_v14_0_notify_memory_pool_location, + .set_soft_freq_limited_range = smu_v14_0_set_soft_freq_limited_range, + .init_pptable_microcode = smu_v14_0_init_pptable_microcode, + .populate_umd_state_clk = smu_v14_0_2_populate_umd_state_clk, + .set_performance_level = smu_v14_0_set_performance_level, + .gfx_off_control = smu_v14_0_gfx_off_control, + .get_unique_id = smu_v14_0_2_get_unique_id, + .get_power_limit = smu_v14_0_2_get_power_limit, + .set_power_limit = smu_v14_0_set_power_limit, + .set_power_source = smu_v14_0_set_power_source, + .get_power_profile_mode = smu_v14_0_2_get_power_profile_mode, + .set_power_profile_mode = smu_v14_0_2_set_power_profile_mode, + .run_btc = smu_v14_0_run_btc, + .get_pp_feature_mask = smu_cmn_get_pp_feature_mask, + .set_pp_feature_mask = smu_cmn_set_pp_feature_mask, + .set_tool_table_location = smu_v14_0_set_tool_table_location, + .deep_sleep_control = smu_v14_0_deep_sleep_control, + .gfx_ulv_control = smu_v14_0_gfx_ulv_control, + .get_bamaco_support = smu_v14_0_get_bamaco_support, + .baco_get_state = smu_v14_0_baco_get_state, + .baco_set_state = smu_v14_0_baco_set_state, + .baco_enter = smu_v14_0_2_baco_enter, + .baco_exit = smu_v14_0_2_baco_exit, + .mode1_reset_is_support = smu_v14_0_2_is_mode1_reset_supported, + .mode1_reset = smu_v14_0_2_mode1_reset, + .mode2_reset = smu_v14_0_2_mode2_reset, + .enable_gfx_features = smu_v14_0_2_enable_gfx_features, + .set_mp1_state = smu_v14_0_2_set_mp1_state, + .set_df_cstate = smu_v14_0_2_set_df_cstate, + .send_hbm_bad_pages_num = smu_v14_0_2_smu_send_bad_mem_page_num, + .send_hbm_bad_channel_flag = smu_v14_0_2_send_bad_mem_channel_flag, + .gpo_control = smu_v14_0_gpo_control, + .get_ecc_info = smu_v14_0_2_get_ecc_info, +}; + +void smu_v14_0_2_set_ppt_funcs(struct smu_context *smu) +{ + smu->ppt_funcs = &smu_v14_0_2_ppt_funcs; + smu->message_map = smu_v14_0_2_message_map; + smu->clock_map = smu_v14_0_2_clk_map; + smu->feature_map = smu_v14_0_2_feature_mask_map; + smu->table_map = smu_v14_0_2_table_map; + smu->pwr_src_map = smu_v14_0_2_pwr_src_map; + smu->workload_map = smu_v14_0_2_workload_map; + smu_v14_0_2_set_smu_mailbox_registers(smu); +} diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0_2_ppt.h b/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0_2_ppt.h new file mode 100644 index 000000000000..b83729e5d6f9 --- /dev/null +++ b/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0_2_ppt.h @@ -0,0 +1,28 @@ +/* + * Copyright 2023 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ +#ifndef __SMU_V14_0_2_PPT_H__ +#define __SMU_V14_0_2_PPT_H__ + +extern void smu_v14_0_2_set_ppt_funcs(struct smu_context *smu); + +#endif diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.c b/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.c index b8dbd4e25348..6d1c3af927ca 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.c @@ -235,6 +235,50 @@ static void __smu_cmn_send_msg(struct smu_context *smu, WREG32(smu->msg_reg, msg); } +static inline uint32_t __smu_cmn_get_msg_flags(struct smu_context *smu, + enum smu_message_type msg) +{ + return smu->message_map[msg].flags; +} + +static int __smu_cmn_ras_filter_msg(struct smu_context *smu, + enum smu_message_type msg, bool *poll) +{ + struct amdgpu_device *adev = smu->adev; + uint32_t flags, resp; + bool fed_status; + + flags = __smu_cmn_get_msg_flags(smu, msg); + *poll = true; + + /* When there is RAS fatal error, FW won't process non-RAS priority + * messages. Don't allow any messages other than RAS priority messages. + */ + fed_status = amdgpu_ras_get_fed_status(adev); + if (fed_status) { + if (!(flags & SMU_MSG_RAS_PRI)) { + dev_dbg(adev->dev, + "RAS error detected, skip sending %s", + smu_get_message_name(smu, msg)); + return -EACCES; + } + + /* FW will ignore non-priority messages when a RAS fatal error + * is detected. Hence it is possible that a previous message + * wouldn't have got response. Allow to continue without polling + * for response status for priority messages. + */ + resp = RREG32(smu->resp_reg); + dev_dbg(adev->dev, + "Sending RAS priority message %s response status: %x", + smu_get_message_name(smu, msg), resp); + if (resp == 0) + *poll = false; + } + + return 0; +} + static int __smu_cmn_send_debug_msg(struct smu_context *smu, u32 msg, u32 param) @@ -354,6 +398,7 @@ int smu_cmn_send_smc_msg_with_param(struct smu_context *smu, { struct amdgpu_device *adev = smu->adev; int res, index; + bool poll = true; u32 reg; if (adev->no_hw_access) @@ -366,12 +411,20 @@ int smu_cmn_send_smc_msg_with_param(struct smu_context *smu, return index == -EACCES ? 0 : index; mutex_lock(&smu->message_lock); - reg = __smu_cmn_poll_stat(smu); - res = __smu_cmn_reg2errno(smu, reg); - if (reg == SMU_RESP_NONE || - res == -EREMOTEIO) { - __smu_cmn_reg_print_error(smu, reg, index, param, msg); - goto Out; + + if (smu->smc_fw_caps & SMU_FW_CAP_RAS_PRI) { + res = __smu_cmn_ras_filter_msg(smu, msg, &poll); + if (res) + goto Out; + } + + if (poll) { + reg = __smu_cmn_poll_stat(smu); + res = __smu_cmn_reg2errno(smu, reg); + if (reg == SMU_RESP_NONE || res == -EREMOTEIO) { + __smu_cmn_reg_print_error(smu, reg, index, param, msg); + goto Out; + } } __smu_cmn_send_msg(smu, (uint16_t) index, param); reg = __smu_cmn_poll_stat(smu); @@ -437,7 +490,7 @@ int smu_cmn_to_asic_specific_index(struct smu_context *smu, return -EINVAL; if (amdgpu_sriov_vf(smu->adev) && - !msg_mapping.valid_in_vf) + !(msg_mapping.flags & SMU_MSG_VF_FLAG)) return -EACCES; return msg_mapping.map_to; |