diff options
Diffstat (limited to 'drivers/gpu/drm/amd/amdgpu')
56 files changed, 940 insertions, 550 deletions
diff --git a/drivers/gpu/drm/amd/amdgpu/Kconfig b/drivers/gpu/drm/amd/amdgpu/Kconfig index 5341b6b242c3..5d1e28218020 100644 --- a/drivers/gpu/drm/amd/amdgpu/Kconfig +++ b/drivers/gpu/drm/amd/amdgpu/Kconfig @@ -6,6 +6,7 @@ config DRM_AMDGPU select FW_LOADER select DRM_DISPLAY_DP_HELPER select DRM_DISPLAY_HDMI_HELPER + select DRM_DISPLAY_HDCP_HELPER select DRM_DISPLAY_HELPER select DRM_KMS_HELPER select DRM_SCHED @@ -18,6 +19,7 @@ config DRM_AMDGPU select BACKLIGHT_CLASS_DEVICE select INTERVAL_TREE select DRM_BUDDY + select DRM_SUBALLOC_HELPER # amdgpu depends on ACPI_VIDEO when ACPI is enabled, for select to work # ACPI_VIDEO's dependencies must also be selected. select INPUT if ACPI diff --git a/drivers/gpu/drm/amd/amdgpu/Makefile b/drivers/gpu/drm/amd/amdgpu/Makefile index 5df603192cdc..1d72cbc85348 100644 --- a/drivers/gpu/drm/amd/amdgpu/Makefile +++ b/drivers/gpu/drm/amd/amdgpu/Makefile @@ -34,6 +34,7 @@ ccflags-y := -I$(FULL_AMD_PATH)/include/asic_reg \ -I$(FULL_AMD_PATH)/acp/include \ -I$(FULL_AMD_DISPLAY_PATH) \ -I$(FULL_AMD_DISPLAY_PATH)/include \ + -I$(FULL_AMD_DISPLAY_PATH)/modules/inc \ -I$(FULL_AMD_DISPLAY_PATH)/dc \ -I$(FULL_AMD_DISPLAY_PATH)/amdgpu_dm \ -I$(FULL_AMD_PATH)/amdkfd @@ -76,7 +77,7 @@ amdgpu-y += \ vi.o mxgpu_vi.o nbio_v6_1.o soc15.o emu_soc.o mxgpu_ai.o nbio_v7_0.o vega10_reg_init.o \ vega20_reg_init.o nbio_v7_4.o nbio_v2_3.o nv.o arct_reg_init.o mxgpu_nv.o \ nbio_v7_2.o hdp_v4_0.o hdp_v5_0.o aldebaran_reg_init.o aldebaran.o soc21.o \ - sienna_cichlid.o nbio_v4_3.o hdp_v6_0.o nbio_v7_7.o hdp_v5_2.o lsdma_v6_0.o + sienna_cichlid.o smu_v13_0_10.o nbio_v4_3.o hdp_v6_0.o nbio_v7_7.o hdp_v5_2.o lsdma_v6_0.o # add DF block amdgpu-y += \ diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index 4e4efd10cb89..dda88090f044 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -242,6 +242,7 @@ extern int amdgpu_num_kcq; #define AMDGPU_VCNFW_LOG_SIZE (32 * 1024) extern int amdgpu_vcnfw_log; +extern int amdgpu_sg_display; #define AMDGPU_VM_MAX_NUM_CTX 4096 #define AMDGPU_SG_THRESHOLD (256*1024*1024) @@ -423,29 +424,11 @@ struct amdgpu_clock { * alignment). */ -#define AMDGPU_SA_NUM_FENCE_LISTS 32 - struct amdgpu_sa_manager { - wait_queue_head_t wq; - struct amdgpu_bo *bo; - struct list_head *hole; - struct list_head flist[AMDGPU_SA_NUM_FENCE_LISTS]; - struct list_head olist; - unsigned size; - uint64_t gpu_addr; - void *cpu_ptr; - uint32_t domain; - uint32_t align; -}; - -/* sub-allocation buffer */ -struct amdgpu_sa_bo { - struct list_head olist; - struct list_head flist; - struct amdgpu_sa_manager *manager; - unsigned soffset; - unsigned eoffset; - struct dma_fence *fence; + struct drm_suballoc_manager base; + struct amdgpu_bo *bo; + uint64_t gpu_addr; + void *cpu_ptr; }; int amdgpu_fence_slab_init(void); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c index f29c1d0ad4c1..d4196fcb85a0 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c @@ -1073,26 +1073,25 @@ bool amdgpu_acpi_is_s0ix_active(struct amdgpu_device *adev) (pm_suspend_target_state != PM_SUSPEND_TO_IDLE)) return false; + if (adev->asic_type < CHIP_RAVEN) + return false; + /* * If ACPI_FADT_LOW_POWER_S0 is not set in the FADT, it is generally * risky to do any special firmware-related preparations for entering * S0ix even though the system is suspending to idle, so return false * in that case. */ - if (!(acpi_gbl_FADT.flags & ACPI_FADT_LOW_POWER_S0)) { + if (!(acpi_gbl_FADT.flags & ACPI_FADT_LOW_POWER_S0)) dev_warn_once(adev->dev, "Power consumption will be higher as BIOS has not been configured for suspend-to-idle.\n" "To use suspend-to-idle change the sleep mode in BIOS setup.\n"); - return false; - } #if !IS_ENABLED(CONFIG_AMD_PMC) dev_warn_once(adev->dev, "Power consumption will be higher as the kernel has not been compiled with CONFIG_AMD_PMC.\n"); - return false; -#else - return true; #endif /* CONFIG_AMD_PMC */ + return true; } #endif /* CONFIG_SUSPEND */ diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c index 8b7a09b392ac..08eced097bd8 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c @@ -63,6 +63,8 @@ static int amdgpu_cs_parser_init(struct amdgpu_cs_parser *p, amdgpu_ctx_put(p->ctx); return -ECANCELED; } + + amdgpu_sync_create(&p->sync); return 0; } @@ -454,18 +456,6 @@ static int amdgpu_syncobj_lookup_and_add(struct amdgpu_cs_parser *p, } r = amdgpu_sync_fence(&p->sync, fence); - if (r) - goto error; - - /* - * When we have an explicit dependency it might be necessary to insert a - * pipeline sync to make sure that all caches etc are flushed and the - * next job actually sees the results from the previous one. - */ - if (fence->context == p->gang_leader->base.entity->fence_context) - r = amdgpu_sync_fence(&p->gang_leader->explicit_sync, fence); - -error: dma_fence_put(fence); return r; } @@ -1190,10 +1180,19 @@ static int amdgpu_cs_vm_handling(struct amdgpu_cs_parser *p) static int amdgpu_cs_sync_rings(struct amdgpu_cs_parser *p) { struct amdgpu_fpriv *fpriv = p->filp->driver_priv; + struct drm_gpu_scheduler *sched; struct amdgpu_bo_list_entry *e; + struct dma_fence *fence; unsigned int i; int r; + r = amdgpu_ctx_wait_prev_fence(p->ctx, p->entities[p->gang_leader_idx]); + if (r) { + if (r != -ERESTARTSYS) + DRM_ERROR("amdgpu_ctx_wait_prev_fence failed.\n"); + return r; + } + list_for_each_entry(e, &p->validated, tv.head) { struct amdgpu_bo *bo = ttm_to_amdgpu_bo(e->tv.bo); struct dma_resv *resv = bo->tbo.base.resv; @@ -1213,10 +1212,27 @@ static int amdgpu_cs_sync_rings(struct amdgpu_cs_parser *p) return r; } - r = amdgpu_ctx_wait_prev_fence(p->ctx, p->entities[p->gang_leader_idx]); - if (r && r != -ERESTARTSYS) - DRM_ERROR("amdgpu_ctx_wait_prev_fence failed.\n"); - return r; + sched = p->gang_leader->base.entity->rq->sched; + while ((fence = amdgpu_sync_get_fence(&p->sync))) { + struct drm_sched_fence *s_fence = to_drm_sched_fence(fence); + + /* + * When we have an dependency it might be necessary to insert a + * pipeline sync to make sure that all caches etc are flushed and the + * next job actually sees the results from the previous one + * before we start executing on the same scheduler ring. + */ + if (!s_fence || s_fence->sched != sched) { + dma_fence_put(fence); + continue; + } + + r = amdgpu_sync_fence(&p->gang_leader->explicit_sync, fence); + dma_fence_put(fence); + if (r) + return r; + } + return 0; } static void amdgpu_cs_post_dependencies(struct amdgpu_cs_parser *p) @@ -1256,9 +1272,12 @@ static int amdgpu_cs_submit(struct amdgpu_cs_parser *p, continue; fence = &p->jobs[i]->base.s_fence->scheduled; + dma_fence_get(fence); r = drm_sched_job_add_dependency(&leader->base, fence); - if (r) + if (r) { + dma_fence_put(fence); goto error_cleanup; + } } if (p->gang_size > 1) { @@ -1346,6 +1365,7 @@ static void amdgpu_cs_parser_fini(struct amdgpu_cs_parser *parser) { unsigned i; + amdgpu_sync_free(&parser->sync); for (i = 0; i < parser->num_post_deps; i++) { drm_syncobj_put(parser->post_deps[i].syncobj); kfree(parser->post_deps[i].chain); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index 1257745fb202..c4a4e2fe6681 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -164,7 +164,7 @@ static void amdgpu_device_get_pcie_info(struct amdgpu_device *adev); * * The amdgpu driver provides a sysfs API for reporting the product name * for the device - * The file serial_number is used for this and returns the product name + * The file product_name is used for this and returns the product name * as returned from the FRU. * NOTE: This is only available for certain server cards */ @@ -186,7 +186,7 @@ static DEVICE_ATTR(product_name, S_IRUGO, * * The amdgpu driver provides a sysfs API for reporting the part number * for the device - * The file serial_number is used for this and returns the part number + * The file product_number is used for this and returns the part number * as returned from the FRU. * NOTE: This is only available for certain server cards */ @@ -3038,6 +3038,18 @@ static int amdgpu_device_ip_suspend_phase2(struct amdgpu_device *adev) (adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_SDMA)) continue; + /* Once swPSP provides the IMU, RLC FW binaries to TOS during cold-boot. + * These are in TMR, hence are expected to be reused by PSP-TOS to reload + * from this location and RLC Autoload automatically also gets loaded + * from here based on PMFW -> PSP message during re-init sequence. + * Therefore, the psp suspend & resume should be skipped to avoid destroy + * the TMR and reload FWs again for IMU enabled APU ASICs. + */ + if (amdgpu_in_reset(adev) && + (adev->flags & AMD_IS_APU) && adev->gfx.imu.funcs && + adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_PSP) + continue; + /* XXX handle errors */ r = adev->ip_blocks[i].version->funcs->suspend(adev); /* XXX handle errors */ @@ -4019,7 +4031,8 @@ void amdgpu_device_fini_hw(struct amdgpu_device *adev) amdgpu_gart_dummy_page_fini(adev); - amdgpu_device_unmap_mmio(adev); + if (drm_dev_is_unplugged(adev_to_drm(adev))) + amdgpu_device_unmap_mmio(adev); } @@ -4257,6 +4270,9 @@ exit: } adev->in_suspend = false; + if (adev->enable_mes) + amdgpu_mes_self_test(adev); + if (amdgpu_acpi_smart_shift_update(dev, AMDGPU_SS_DEV_D0)) DRM_WARN("smart shift update failed\n"); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c index b719852daa07..1a3cb53d2e0d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c @@ -543,6 +543,7 @@ static void amdgpu_discovery_read_from_harvest_table(struct amdgpu_device *adev, struct harvest_table *harvest_info; u16 offset; int i; + uint32_t umc_harvest_config = 0; bhdr = (struct binary_header *)adev->mman.discovery_bin; offset = le16_to_cpu(bhdr->table_list[HARVEST_INFO].offset); @@ -570,12 +571,17 @@ static void amdgpu_discovery_read_from_harvest_table(struct amdgpu_device *adev, adev->harvest_ip_mask |= AMD_HARVEST_IP_DMU_MASK; break; case UMC_HWID: + umc_harvest_config |= + 1 << (le16_to_cpu(harvest_info->list[i].number_instance)); (*umc_harvest_count)++; break; default: break; } } + + adev->umc.active_mask = ((1 << adev->umc.node_inst_num) - 1) & + ~umc_harvest_config; } /* ================================================== */ @@ -1156,8 +1162,10 @@ static int amdgpu_discovery_reg_base_init(struct amdgpu_device *adev) AMDGPU_MAX_SDMA_INSTANCES); } - if (le16_to_cpu(ip->hw_id) == UMC_HWID) + if (le16_to_cpu(ip->hw_id) == UMC_HWID) { adev->gmc.num_umc++; + adev->umc.node_inst_num++; + } for (k = 0; k < num_base_address; k++) { /* diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c index 7bb12a76631f..f5ffca24def4 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c @@ -107,9 +107,12 @@ * - 3.50.0 - Update AMDGPU_INFO_DEV_INFO IOCTL for minimum engine and memory clock * Update AMDGPU_INFO_SENSOR IOCTL for PEAK_PSTATE engine and memory clock * 3.51.0 - Return the PCIe gen and lanes from the INFO ioctl + * 3.52.0 - Add AMDGPU_IDS_FLAGS_CONFORMANT_TRUNC_COORD, add device_info fields: + * tcp_cache_size, num_sqc_per_wgp, sqc_data_cache_size, sqc_inst_cache_size, + * gl1c_cache_size, gl2c_cache_size, mall_size, enabled_rb_pipes_mask_hi */ #define KMS_DRIVER_MAJOR 3 -#define KMS_DRIVER_MINOR 51 +#define KMS_DRIVER_MINOR 52 #define KMS_DRIVER_PATCHLEVEL 0 unsigned int amdgpu_vram_limit = UINT_MAX; @@ -188,6 +191,7 @@ int amdgpu_num_kcq = -1; int amdgpu_smartshift_bias; int amdgpu_use_xgmi_p2p = 1; int amdgpu_vcnfw_log; +int amdgpu_sg_display = -1; /* auto */ static void amdgpu_drv_delayed_reset_work_handler(struct work_struct *work); @@ -920,7 +924,7 @@ module_param_named(reset_method, amdgpu_reset_method, int, 0444); * result in the GPU entering bad status when the number of total * faulty pages by ECC exceeds the threshold value. */ -MODULE_PARM_DESC(bad_page_threshold, "Bad page threshold(-1 = auto(default value), 0 = disable bad page retirement, -2 = ignore bad page threshold)"); +MODULE_PARM_DESC(bad_page_threshold, "Bad page threshold(-1 = ignore threshold (default value), 0 = disable bad page retirement, -2 = driver sets threshold)"); module_param_named(bad_page_threshold, amdgpu_bad_page_threshold, int, 0444); MODULE_PARM_DESC(num_kcq, "number of kernel compute queue user want to setup (8 if set to greater than 8 or less than 0, only affect gfx 8+)"); @@ -934,6 +938,16 @@ MODULE_PARM_DESC(vcnfw_log, "Enable vcnfw log(0 = disable (default value), 1 = e module_param_named(vcnfw_log, amdgpu_vcnfw_log, int, 0444); /** + * DOC: sg_display (int) + * Disable S/G (scatter/gather) display (i.e., display from system memory). + * This option is only relevant on APUs. Set this option to 0 to disable + * S/G display if you experience flickering or other issues under memory + * pressure and report the issue. + */ +MODULE_PARM_DESC(sg_display, "S/G Display (-1 = auto (default), 0 = disable)"); +module_param_named(sg_display, amdgpu_sg_display, int, 0444); + +/** * DOC: smu_pptable_id (int) * Used to override pptable id. id = 0 use VBIOS pptable. * id > 0 use the soft pptable with specicfied id. @@ -2227,6 +2241,8 @@ amdgpu_pci_remove(struct pci_dev *pdev) struct drm_device *dev = pci_get_drvdata(pdev); struct amdgpu_device *adev = drm_to_adev(dev); + drm_dev_unplug(dev); + if (adev->pm.rpm_mode != AMDGPU_RUNPM_NONE) { pm_runtime_get_sync(dev->dev); pm_runtime_forbid(dev->dev); @@ -2266,8 +2282,6 @@ amdgpu_pci_remove(struct pci_dev *pdev) amdgpu_driver_unload_kms(dev); - drm_dev_unplug(dev); - /* * Flush any in flight DMA operations from device. * Clear the Bus Master Enable bit and then wait on the PCIe Device @@ -2403,8 +2417,10 @@ static int amdgpu_pmops_suspend(struct device *dev) if (amdgpu_acpi_is_s0ix_active(adev)) adev->in_s0ix = true; - else + else if (amdgpu_acpi_is_s3_active(adev)) adev->in_s3 = true; + if (!adev->in_s0ix && !adev->in_s3) + return 0; return amdgpu_device_suspend(drm_dev, true); } @@ -2425,6 +2441,9 @@ static int amdgpu_pmops_resume(struct device *dev) struct amdgpu_device *adev = drm_to_adev(drm_dev); int r; + if (!adev->in_s0ix && !adev->in_s3) + return 0; + /* Avoids registers access if device is physically gone */ if (!pci_device_is_present(adev->pdev)) adev->no_hw_access = true; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c index 00444203220d..faff4a3f96e6 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c @@ -618,7 +618,13 @@ void amdgpu_fence_driver_sw_fini(struct amdgpu_device *adev) if (!ring || !ring->fence_drv.initialized) continue; - if (!ring->no_scheduler) + /* + * Notice we check for sched.ops since there's some + * override on the meaning of sched.ready by amdgpu. + * The natural check would be sched.ready, which is + * set as drm_sched_init() finishes... + */ + if (ring->sched.ops) drm_sched_fini(&ring->sched); for (j = 0; j <= ring->fence_drv.num_fences_mask; ++j) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c index ed1164a87fce..d8e683688daa 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c @@ -258,7 +258,7 @@ static int amdgpu_gem_object_mmap(struct drm_gem_object *obj, struct vm_area_str */ if (is_cow_mapping(vma->vm_flags) && !(vma->vm_flags & VM_ACCESS_FLAGS)) - vma->vm_flags &= ~VM_MAYWRITE; + vm_flags_clear(vma, VM_MAYWRITE); return drm_gem_ttm_mmap(obj, vma); } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h index 86ec9d0d12c8..de9e7a00bb15 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h @@ -178,6 +178,8 @@ struct amdgpu_gfx_config { uint32_t num_sc_per_sh; uint32_t num_packer_per_sc; uint32_t pa_sc_tile_steering_override; + /* Whether texture coordinate truncation is conformant. */ + bool ta_cntl2_truncate_coord_mode; uint64_t tcc_disabled_mask; uint32_t gc_num_tcp_per_sa; uint32_t gc_num_sdp_interface; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c index 94f10ac0eef7..12a6826caef4 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c @@ -552,6 +552,7 @@ void amdgpu_gmc_tmz_set(struct amdgpu_device *adev) case IP_VERSION(10, 3, 2): case IP_VERSION(10, 3, 4): case IP_VERSION(10, 3, 5): + case IP_VERSION(10, 3, 6): /* VANGOGH */ case IP_VERSION(10, 3, 1): /* YELLOW_CARP*/ diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c index bcccc348dbe2..df7eb0b7c4b9 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c @@ -69,7 +69,7 @@ int amdgpu_ib_get(struct amdgpu_device *adev, struct amdgpu_vm *vm, if (size) { r = amdgpu_sa_bo_new(&adev->ib_pools[pool_type], - &ib->sa_bo, size, 256); + &ib->sa_bo, size); if (r) { dev_err(adev->dev, "failed to get a new IB (%d)\n", r); return r; @@ -309,8 +309,7 @@ int amdgpu_ib_pool_init(struct amdgpu_device *adev) for (i = 0; i < AMDGPU_IB_POOL_MAX; i++) { r = amdgpu_sa_bo_manager_init(adev, &adev->ib_pools[i], - AMDGPU_IB_POOL_SIZE, - AMDGPU_GPU_PAGE_SIZE, + AMDGPU_IB_POOL_SIZE, 256, AMDGPU_GEM_DOMAIN_GTT); if (r) goto error; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c index ca945055e683..0efb38539d70 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c @@ -808,6 +808,8 @@ int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file *filp) dev_info->ids_flags |= AMDGPU_IDS_FLAGS_PREEMPTION; if (amdgpu_is_tmz(adev)) dev_info->ids_flags |= AMDGPU_IDS_FLAGS_TMZ; + if (adev->gfx.config.ta_cntl2_truncate_coord_mode) + dev_info->ids_flags |= AMDGPU_IDS_FLAGS_CONFORMANT_TRUNC_COORD; vm_size = adev->vm_manager.max_pfn * AMDGPU_GPU_PAGE_SIZE; vm_size -= AMDGPU_VA_RESERVED_SIZE; @@ -865,6 +867,15 @@ int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file *filp) adev->pm.pcie_mlw_mask & CAIL_PCIE_LINK_WIDTH_SUPPORT_X4 ? 4 : adev->pm.pcie_mlw_mask & CAIL_PCIE_LINK_WIDTH_SUPPORT_X2 ? 2 : 1; + dev_info->tcp_cache_size = adev->gfx.config.gc_tcp_l1_size; + dev_info->num_sqc_per_wgp = adev->gfx.config.gc_num_sqc_per_wgp; + dev_info->sqc_data_cache_size = adev->gfx.config.gc_l1_data_cache_size_per_sqc; + dev_info->sqc_inst_cache_size = adev->gfx.config.gc_l1_instruction_cache_size_per_sqc; + dev_info->gl1c_cache_size = adev->gfx.config.gc_gl1c_size_per_instance * + adev->gfx.config.gc_gl1c_per_sa; + dev_info->gl2c_cache_size = adev->gfx.config.gc_gl2c_per_gpu; + dev_info->mall_size = adev->gmc.mall_size; + ret = copy_to_user(out, dev_info, min((size_t)size, sizeof(*dev_info))) ? -EFAULT : 0; kfree(dev_info); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h index 44c57f4a84c4..32fe05c810c6 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h @@ -549,8 +549,8 @@ struct amdgpu_mst_connector { struct drm_dp_mst_topology_mgr mst_mgr; struct amdgpu_dm_dp_aux dm_dp_aux; - struct drm_dp_mst_port *port; - struct amdgpu_connector *mst_port; + struct drm_dp_mst_port *mst_output_port; + struct amdgpu_connector *mst_root; bool is_mst_connector; struct amdgpu_encoder *mst_encoder; }; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c index 981010de0a28..c842ce635a88 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c @@ -139,7 +139,7 @@ void amdgpu_bo_placement_from_domain(struct amdgpu_bo *abo, u32 domain) if (flags & AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED) places[c].lpfn = visible_pfn; - else + else if (adev->gmc.real_vram_size != adev->gmc.visible_vram_size) places[c].flags |= TTM_PL_FLAG_TOPDOWN; if (flags & AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS) @@ -600,7 +600,7 @@ int amdgpu_bo_create(struct amdgpu_device *adev, if (!amdgpu_gmc_vram_full_visible(&adev->gmc) && bo->tbo.resource->mem_type == TTM_PL_VRAM && - bo->tbo.resource->start < adev->gmc.visible_vram_size >> PAGE_SHIFT) + amdgpu_bo_in_cpu_visible_vram(bo)) amdgpu_cs_report_moved_bytes(adev, ctx.bytes_moved, ctx.bytes_moved); else @@ -1315,7 +1315,7 @@ void amdgpu_bo_release_notify(struct ttm_buffer_object *bo) if (!bo->resource || bo->resource->mem_type != TTM_PL_VRAM || !(abo->flags & AMDGPU_GEM_CREATE_VRAM_WIPE_ON_RELEASE) || - adev->in_suspend || adev->shutdown) + adev->in_suspend || drm_dev_is_unplugged(adev_to_drm(adev))) return; if (WARN_ON_ONCE(!dma_resv_trylock(bo->base.resv))) @@ -1346,7 +1346,6 @@ vm_fault_t amdgpu_bo_fault_reserve_notify(struct ttm_buffer_object *bo) struct amdgpu_device *adev = amdgpu_ttm_adev(bo->bdev); struct ttm_operation_ctx ctx = { false, false }; struct amdgpu_bo *abo = ttm_to_amdgpu_bo(bo); - unsigned long offset; int r; /* Remember that this BO was accessed by the CPU */ @@ -1355,8 +1354,7 @@ vm_fault_t amdgpu_bo_fault_reserve_notify(struct ttm_buffer_object *bo) if (bo->resource->mem_type != TTM_PL_VRAM) return 0; - offset = bo->resource->start << PAGE_SHIFT; - if ((offset + bo->base.size) <= adev->gmc.visible_vram_size) + if (amdgpu_bo_in_cpu_visible_vram(abo)) return 0; /* Can't move a pinned BO to visible VRAM */ @@ -1378,10 +1376,9 @@ vm_fault_t amdgpu_bo_fault_reserve_notify(struct ttm_buffer_object *bo) else if (unlikely(r)) return VM_FAULT_SIGBUS; - offset = bo->resource->start << PAGE_SHIFT; /* this should never happen */ if (bo->resource->mem_type == TTM_PL_VRAM && - (offset + bo->base.size) > adev->gmc.visible_vram_size) + !amdgpu_bo_in_cpu_visible_vram(abo)) return VM_FAULT_SIGBUS; ttm_bo_move_to_lru_tail_unlocked(bo); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h index 93207badf83f..5a85726ce853 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h @@ -336,15 +336,22 @@ uint32_t amdgpu_bo_get_preferred_domain(struct amdgpu_device *adev, /* * sub allocation */ +static inline struct amdgpu_sa_manager * +to_amdgpu_sa_manager(struct drm_suballoc_manager *manager) +{ + return container_of(manager, struct amdgpu_sa_manager, base); +} -static inline uint64_t amdgpu_sa_bo_gpu_addr(struct amdgpu_sa_bo *sa_bo) +static inline uint64_t amdgpu_sa_bo_gpu_addr(struct drm_suballoc *sa_bo) { - return sa_bo->manager->gpu_addr + sa_bo->soffset; + return to_amdgpu_sa_manager(sa_bo->manager)->gpu_addr + + drm_suballoc_soffset(sa_bo); } -static inline void * amdgpu_sa_bo_cpu_addr(struct amdgpu_sa_bo *sa_bo) +static inline void *amdgpu_sa_bo_cpu_addr(struct drm_suballoc *sa_bo) { - return sa_bo->manager->cpu_ptr + sa_bo->soffset; + return to_amdgpu_sa_manager(sa_bo->manager)->cpu_ptr + + drm_suballoc_soffset(sa_bo); } int amdgpu_sa_bo_manager_init(struct amdgpu_device *adev, @@ -355,11 +362,11 @@ void amdgpu_sa_bo_manager_fini(struct amdgpu_device *adev, int amdgpu_sa_bo_manager_start(struct amdgpu_device *adev, struct amdgpu_sa_manager *sa_manager); int amdgpu_sa_bo_new(struct amdgpu_sa_manager *sa_manager, - struct amdgpu_sa_bo **sa_bo, - unsigned size, unsigned align); + struct drm_suballoc **sa_bo, + unsigned int size); void amdgpu_sa_bo_free(struct amdgpu_device *adev, - struct amdgpu_sa_bo **sa_bo, - struct dma_fence *fence); + struct drm_suballoc **sa_bo, + struct dma_fence *fence); #if defined(CONFIG_DEBUG_FS) void amdgpu_sa_bo_dump_debug_info(struct amdgpu_sa_manager *sa_manager, struct seq_file *m); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c index a8391f269cd0..3f5d13035aff 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c @@ -602,7 +602,7 @@ psp_cmd_submit_buf(struct psp_context *psp, struct psp_gfx_cmd_resp *cmd, uint64_t fence_mc_addr) { int ret; - int index, idx; + int index; int timeout = 20000; bool ras_intr = false; bool skip_unsupport = false; @@ -610,9 +610,6 @@ psp_cmd_submit_buf(struct psp_context *psp, if (psp->adev->no_hw_access) return 0; - if (!drm_dev_enter(adev_to_drm(psp->adev), &idx)) - return 0; - memset(psp->cmd_buf_mem, 0, PSP_CMD_BUFFER_SIZE); memcpy(psp->cmd_buf_mem, cmd, sizeof(struct psp_gfx_cmd_resp)); @@ -676,7 +673,6 @@ psp_cmd_submit_buf(struct psp_context *psp, } exit: - drm_dev_exit(idx); return ret; } @@ -1672,7 +1668,7 @@ static int psp_hdcp_initialize(struct psp_context *psp) psp->hdcp_context.context.mem_context.shared_mem_size = PSP_HDCP_SHARED_MEM_SIZE; psp->hdcp_context.context.ta_load_type = GFX_CMD_ID_LOAD_TA; - if (!psp->hdcp_context.context.initialized) { + if (!psp->hdcp_context.context.mem_context.shared_buf) { ret = psp_ta_init_shared_buf(psp, &psp->hdcp_context.context.mem_context); if (ret) return ret; @@ -1739,7 +1735,7 @@ static int psp_dtm_initialize(struct psp_context *psp) psp->dtm_context.context.mem_context.shared_mem_size = PSP_DTM_SHARED_MEM_SIZE; psp->dtm_context.context.ta_load_type = GFX_CMD_ID_LOAD_TA; - if (!psp->dtm_context.context.initialized) { + if (!psp->dtm_context.context.mem_context.shared_buf) { ret = psp_ta_init_shared_buf(psp, &psp->dtm_context.context.mem_context); if (ret) return ret; @@ -1807,7 +1803,7 @@ static int psp_rap_initialize(struct psp_context *psp) psp->rap_context.context.mem_context.shared_mem_size = PSP_RAP_SHARED_MEM_SIZE; psp->rap_context.context.ta_load_type = GFX_CMD_ID_LOAD_TA; - if (!psp->rap_context.context.initialized) { + if (!psp->rap_context.context.mem_context.shared_buf) { ret = psp_ta_init_shared_buf(psp, &psp->rap_context.context.mem_context); if (ret) return ret; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c index 6e543558386d..63dfcc98152d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c @@ -176,7 +176,7 @@ static int amdgpu_reserve_page_direct(struct amdgpu_device *adev, uint64_t addre if (amdgpu_bad_page_threshold != 0) { amdgpu_ras_add_bad_pages(adev, err_data.err_addr, err_data.err_addr_cnt); - amdgpu_ras_save_bad_pages(adev); + amdgpu_ras_save_bad_pages(adev, NULL); } dev_warn(adev->dev, "WARNING: THIS IS ONLY FOR TEST PURPOSES AND WILL CORRUPT RAS EEPROM\n"); @@ -2084,22 +2084,32 @@ out: /* * write error record array to eeprom, the function should be * protected by recovery_lock + * new_cnt: new added UE count, excluding reserved bad pages, can be NULL */ -int amdgpu_ras_save_bad_pages(struct amdgpu_device *adev) +int amdgpu_ras_save_bad_pages(struct amdgpu_device *adev, + unsigned long *new_cnt) { struct amdgpu_ras *con = amdgpu_ras_get_context(adev); struct ras_err_handler_data *data; struct amdgpu_ras_eeprom_control *control; int save_count; - if (!con || !con->eh_data) + if (!con || !con->eh_data) { + if (new_cnt) + *new_cnt = 0; + return 0; + } mutex_lock(&con->recovery_lock); control = &con->eeprom_control; data = con->eh_data; save_count = data->count - control->ras_num_recs; mutex_unlock(&con->recovery_lock); + + if (new_cnt) + *new_cnt = save_count / adev->umc.retire_unit; + /* only new entries are saved */ if (save_count > 0) { if (amdgpu_ras_eeprom_append(control, @@ -2186,11 +2196,12 @@ static void amdgpu_ras_validate_threshold(struct amdgpu_device *adev, /* * Justification of value bad_page_cnt_threshold in ras structure * - * Generally, -1 <= amdgpu_bad_page_threshold <= max record length - * in eeprom, and introduce two scenarios accordingly. + * Generally, 0 <= amdgpu_bad_page_threshold <= max record length + * in eeprom or amdgpu_bad_page_threshold == -2, introduce two + * scenarios accordingly. * * Bad page retirement enablement: - * - If amdgpu_bad_page_threshold = -1, + * - If amdgpu_bad_page_threshold = -2, * bad_page_cnt_threshold = typical value by formula. * * - When the value from user is 0 < amdgpu_bad_page_threshold < diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h index f2ad999993f6..ef38f4c93df0 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h @@ -547,7 +547,8 @@ int amdgpu_ras_query_error_count(struct amdgpu_device *adev, int amdgpu_ras_add_bad_pages(struct amdgpu_device *adev, struct eeprom_table_record *bps, int pages); -int amdgpu_ras_save_bad_pages(struct amdgpu_device *adev); +int amdgpu_ras_save_bad_pages(struct amdgpu_device *adev, + unsigned long *new_cnt); static inline enum ta_ras_block amdgpu_ras_block_to_ta(enum amdgpu_ras_block block) { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c index 2d9f3f4cd79e..2e08fce87521 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c @@ -417,7 +417,8 @@ bool amdgpu_ras_eeprom_check_err_threshold(struct amdgpu_device *adev) { struct amdgpu_ras *con = amdgpu_ras_get_context(adev); - if (!__is_ras_eeprom_supported(adev)) + if (!__is_ras_eeprom_supported(adev) || + !amdgpu_bad_page_threshold) return false; /* skip check eeprom table for VEGA20 Gaming */ @@ -428,10 +429,18 @@ bool amdgpu_ras_eeprom_check_err_threshold(struct amdgpu_device *adev) return false; if (con->eeprom_control.tbl_hdr.header == RAS_TABLE_HDR_BAD) { - dev_warn(adev->dev, "This GPU is in BAD status."); - dev_warn(adev->dev, "Please retire it or set a larger " - "threshold value when reloading driver.\n"); - return true; + if (amdgpu_bad_page_threshold == -1) { + dev_warn(adev->dev, "RAS records:%d exceed threshold:%d", + con->eeprom_control.ras_num_recs, con->bad_page_cnt_threshold); + dev_warn(adev->dev, + "But GPU can be operated due to bad_page_threshold = -1.\n"); + return false; + } else { + dev_warn(adev->dev, "This GPU is in BAD status."); + dev_warn(adev->dev, "Please retire it or set a larger " + "threshold value when reloading driver.\n"); + return true; + } } return false; @@ -1191,8 +1200,8 @@ int amdgpu_ras_eeprom_init(struct amdgpu_ras_eeprom_control *control, } else { dev_err(adev->dev, "RAS records:%d exceed threshold:%d", control->ras_num_recs, ras->bad_page_cnt_threshold); - if (amdgpu_bad_page_threshold == -2) { - dev_warn(adev->dev, "GPU will be initialized due to bad_page_threshold = -2."); + if (amdgpu_bad_page_threshold == -1) { + dev_warn(adev->dev, "GPU will be initialized due to bad_page_threshold = -1."); res = 0; } else { *exceed_err_limit = true; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_reset.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_reset.c index f778466bb9db..6437ead87e5f 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_reset.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_reset.c @@ -24,6 +24,7 @@ #include "amdgpu_reset.h" #include "aldebaran.h" #include "sienna_cichlid.h" +#include "smu_v13_0_10.h" int amdgpu_reset_add_handler(struct amdgpu_reset_control *reset_ctl, struct amdgpu_reset_handler *handler) @@ -44,6 +45,9 @@ int amdgpu_reset_init(struct amdgpu_device *adev) case IP_VERSION(11, 0, 7): ret = sienna_cichlid_reset_init(adev); break; + case IP_VERSION(13, 0, 10): + ret = smu_v13_0_10_reset_init(adev); + break; default: break; } @@ -62,6 +66,9 @@ int amdgpu_reset_fini(struct amdgpu_device *adev) case IP_VERSION(11, 0, 7): ret = sienna_cichlid_reset_fini(adev); break; + case IP_VERSION(13, 0, 10): + ret = smu_v13_0_10_reset_fini(adev); + break; default: break; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h index f752c7ae7f60..018f36b10de8 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h @@ -27,6 +27,7 @@ #include <drm/amdgpu_drm.h> #include <drm/gpu_scheduler.h> #include <drm/drm_print.h> +#include <drm/drm_suballoc.h> struct amdgpu_device; struct amdgpu_ring; @@ -92,7 +93,7 @@ enum amdgpu_ib_pool_type { }; struct amdgpu_ib { - struct amdgpu_sa_bo *sa_bo; + struct drm_suballoc *sa_bo; uint32_t length_dw; uint64_t gpu_addr; uint32_t *ptr; @@ -295,7 +296,7 @@ struct amdgpu_ring { #define amdgpu_ring_parse_cs(r, p, job, ib) ((r)->funcs->parse_cs((p), (job), (ib))) #define amdgpu_ring_patch_cs_in_place(r, p, job, ib) ((r)->funcs->patch_cs_in_place((p), (job), (ib))) #define amdgpu_ring_test_ring(r) (r)->funcs->test_ring((r)) -#define amdgpu_ring_test_ib(r, t) (r)->funcs->test_ib((r), (t)) +#define amdgpu_ring_test_ib(r, t) ((r)->funcs->test_ib ? (r)->funcs->test_ib((r), (t)) : 0) #define amdgpu_ring_get_rptr(r) (r)->funcs->get_rptr((r)) #define amdgpu_ring_get_wptr(r) (r)->funcs->get_wptr((r)) #define amdgpu_ring_set_wptr(r) (r)->funcs->set_wptr((r)) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_sa.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_sa.c index 524d10b21041..c6b4337eb20c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_sa.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_sa.c @@ -44,327 +44,63 @@ #include "amdgpu.h" -static void amdgpu_sa_bo_remove_locked(struct amdgpu_sa_bo *sa_bo); -static void amdgpu_sa_bo_try_free(struct amdgpu_sa_manager *sa_manager); - int amdgpu_sa_bo_manager_init(struct amdgpu_device *adev, struct amdgpu_sa_manager *sa_manager, - unsigned size, u32 align, u32 domain) + unsigned int size, u32 suballoc_align, u32 domain) { - int i, r; - - init_waitqueue_head(&sa_manager->wq); - sa_manager->bo = NULL; - sa_manager->size = size; - sa_manager->domain = domain; - sa_manager->align = align; - sa_manager->hole = &sa_manager->olist; - INIT_LIST_HEAD(&sa_manager->olist); - for (i = 0; i < AMDGPU_SA_NUM_FENCE_LISTS; ++i) - INIT_LIST_HEAD(&sa_manager->flist[i]); + int r; - r = amdgpu_bo_create_kernel(adev, size, align, domain, &sa_manager->bo, - &sa_manager->gpu_addr, &sa_manager->cpu_ptr); + r = amdgpu_bo_create_kernel(adev, size, AMDGPU_GPU_PAGE_SIZE, domain, + &sa_manager->bo, &sa_manager->gpu_addr, + &sa_manager->cpu_ptr); if (r) { dev_err(adev->dev, "(%d) failed to allocate bo for manager\n", r); return r; } - memset(sa_manager->cpu_ptr, 0, sa_manager->size); + memset(sa_manager->cpu_ptr, 0, size); + drm_suballoc_manager_init(&sa_manager->base, size, suballoc_align); return r; } void amdgpu_sa_bo_manager_fini(struct amdgpu_device *adev, struct amdgpu_sa_manager *sa_manager) { - struct amdgpu_sa_bo *sa_bo, *tmp; - if (sa_manager->bo == NULL) { dev_err(adev->dev, "no bo for sa manager\n"); return; } - if (!list_empty(&sa_manager->olist)) { - sa_manager->hole = &sa_manager->olist, - amdgpu_sa_bo_try_free(sa_manager); - if (!list_empty(&sa_manager->olist)) { - dev_err(adev->dev, "sa_manager is not empty, clearing anyway\n"); - } - } - list_for_each_entry_safe(sa_bo, tmp, &sa_manager->olist, olist) { - amdgpu_sa_bo_remove_locked(sa_bo); - } + drm_suballoc_manager_fini(&sa_manager->base); amdgpu_bo_free_kernel(&sa_manager->bo, &sa_manager->gpu_addr, &sa_manager->cpu_ptr); - sa_manager->size = 0; } -static void amdgpu_sa_bo_remove_locked(struct amdgpu_sa_bo *sa_bo) -{ - struct amdgpu_sa_manager *sa_manager = sa_bo->manager; - if (sa_manager->hole == &sa_bo->olist) { - sa_manager->hole = sa_bo->olist.prev; - } - list_del_init(&sa_bo->olist); - list_del_init(&sa_bo->flist); - dma_fence_put(sa_bo->fence); - kfree(sa_bo); -} - -static void amdgpu_sa_bo_try_free(struct amdgpu_sa_manager *sa_manager) +int amdgpu_sa_bo_new(struct amdgpu_sa_manager *sa_manager, + struct drm_suballoc **sa_bo, + unsigned int size) { - struct amdgpu_sa_bo *sa_bo, *tmp; + struct drm_suballoc *sa = drm_suballoc_new(&sa_manager->base, size, + GFP_KERNEL, true, 0); - if (sa_manager->hole->next == &sa_manager->olist) - return; + if (IS_ERR(sa)) { + *sa_bo = NULL; - sa_bo = list_entry(sa_manager->hole->next, struct amdgpu_sa_bo, olist); - list_for_each_entry_safe_from(sa_bo, tmp, &sa_manager->olist, olist) { - if (sa_bo->fence == NULL || - !dma_fence_is_signaled(sa_bo->fence)) { - return; - } - amdgpu_sa_bo_remove_locked(sa_bo); + return PTR_ERR(sa); } -} -static inline unsigned amdgpu_sa_bo_hole_soffset(struct amdgpu_sa_manager *sa_manager) -{ - struct list_head *hole = sa_manager->hole; - - if (hole != &sa_manager->olist) { - return list_entry(hole, struct amdgpu_sa_bo, olist)->eoffset; - } + *sa_bo = sa; return 0; } -static inline unsigned amdgpu_sa_bo_hole_eoffset(struct amdgpu_sa_manager *sa_manager) -{ - struct list_head *hole = sa_manager->hole; - - if (hole->next != &sa_manager->olist) { - return list_entry(hole->next, struct amdgpu_sa_bo, olist)->soffset; - } - return sa_manager->size; -} - -static bool amdgpu_sa_bo_try_alloc(struct amdgpu_sa_manager *sa_manager, - struct amdgpu_sa_bo *sa_bo, - unsigned size, unsigned align) -{ - unsigned soffset, eoffset, wasted; - - soffset = amdgpu_sa_bo_hole_soffset(sa_manager); - eoffset = amdgpu_sa_bo_hole_eoffset(sa_manager); - wasted = (align - (soffset % align)) % align; - - if ((eoffset - soffset) >= (size + wasted)) { - soffset += wasted; - - sa_bo->manager = sa_manager; - sa_bo->soffset = soffset; - sa_bo->eoffset = soffset + size; - list_add(&sa_bo->olist, sa_manager->hole); - INIT_LIST_HEAD(&sa_bo->flist); - sa_manager->hole = &sa_bo->olist; - return true; - } - return false; -} - -/** - * amdgpu_sa_event - Check if we can stop waiting - * - * @sa_manager: pointer to the sa_manager - * @size: number of bytes we want to allocate - * @align: alignment we need to match - * - * Check if either there is a fence we can wait for or - * enough free memory to satisfy the allocation directly - */ -static bool amdgpu_sa_event(struct amdgpu_sa_manager *sa_manager, - unsigned size, unsigned align) -{ - unsigned soffset, eoffset, wasted; - int i; - - for (i = 0; i < AMDGPU_SA_NUM_FENCE_LISTS; ++i) - if (!list_empty(&sa_manager->flist[i])) - return true; - - soffset = amdgpu_sa_bo_hole_soffset(sa_manager); - eoffset = amdgpu_sa_bo_hole_eoffset(sa_manager); - wasted = (align - (soffset % align)) % align; - - if ((eoffset - soffset) >= (size + wasted)) { - return true; - } - - return false; -} - -static bool amdgpu_sa_bo_next_hole(struct amdgpu_sa_manager *sa_manager, - struct dma_fence **fences, - unsigned *tries) -{ - struct amdgpu_sa_bo *best_bo = NULL; - unsigned i, soffset, best, tmp; - - /* if hole points to the end of the buffer */ - if (sa_manager->hole->next == &sa_manager->olist) { - /* try again with its beginning */ - sa_manager->hole = &sa_manager->olist; - return true; - } - - soffset = amdgpu_sa_bo_hole_soffset(sa_manager); - /* to handle wrap around we add sa_manager->size */ - best = sa_manager->size * 2; - /* go over all fence list and try to find the closest sa_bo - * of the current last - */ - for (i = 0; i < AMDGPU_SA_NUM_FENCE_LISTS; ++i) { - struct amdgpu_sa_bo *sa_bo; - - fences[i] = NULL; - - if (list_empty(&sa_manager->flist[i])) - continue; - - sa_bo = list_first_entry(&sa_manager->flist[i], - struct amdgpu_sa_bo, flist); - - if (!dma_fence_is_signaled(sa_bo->fence)) { - fences[i] = sa_bo->fence; - continue; - } - - /* limit the number of tries each ring gets */ - if (tries[i] > 2) { - continue; - } - - tmp = sa_bo->soffset; - if (tmp < soffset) { - /* wrap around, pretend it's after */ - tmp += sa_manager->size; - } - tmp -= soffset; - if (tmp < best) { - /* this sa bo is the closest one */ - best = tmp; - best_bo = sa_bo; - } - } - - if (best_bo) { - uint32_t idx = best_bo->fence->context; - - idx %= AMDGPU_SA_NUM_FENCE_LISTS; - ++tries[idx]; - sa_manager->hole = best_bo->olist.prev; - - /* we knew that this one is signaled, - so it's save to remote it */ - amdgpu_sa_bo_remove_locked(best_bo); - return true; - } - return false; -} - -int amdgpu_sa_bo_new(struct amdgpu_sa_manager *sa_manager, - struct amdgpu_sa_bo **sa_bo, - unsigned size, unsigned align) -{ - struct dma_fence *fences[AMDGPU_SA_NUM_FENCE_LISTS]; - unsigned tries[AMDGPU_SA_NUM_FENCE_LISTS]; - unsigned count; - int i, r; - signed long t; - - if (WARN_ON_ONCE(align > sa_manager->align)) - return -EINVAL; - - if (WARN_ON_ONCE(size > sa_manager->size)) - return -EINVAL; - - *sa_bo = kmalloc(sizeof(struct amdgpu_sa_bo), GFP_KERNEL); - if (!(*sa_bo)) - return -ENOMEM; - (*sa_bo)->manager = sa_manager; - (*sa_bo)->fence = NULL; - INIT_LIST_HEAD(&(*sa_bo)->olist); - INIT_LIST_HEAD(&(*sa_bo)->flist); - - spin_lock(&sa_manager->wq.lock); - do { - for (i = 0; i < AMDGPU_SA_NUM_FENCE_LISTS; ++i) - tries[i] = 0; - - do { - amdgpu_sa_bo_try_free(sa_manager); - - if (amdgpu_sa_bo_try_alloc(sa_manager, *sa_bo, - size, align)) { - spin_unlock(&sa_manager->wq.lock); - return 0; - } - - /* see if we can skip over some allocations */ - } while (amdgpu_sa_bo_next_hole(sa_manager, fences, tries)); - - for (i = 0, count = 0; i < AMDGPU_SA_NUM_FENCE_LISTS; ++i) - if (fences[i]) - fences[count++] = dma_fence_get(fences[i]); - - if (count) { - spin_unlock(&sa_manager->wq.lock); - t = dma_fence_wait_any_timeout(fences, count, false, - MAX_SCHEDULE_TIMEOUT, - NULL); - for (i = 0; i < count; ++i) - dma_fence_put(fences[i]); - - r = (t > 0) ? 0 : t; - spin_lock(&sa_manager->wq.lock); - } else { - /* if we have nothing to wait for block */ - r = wait_event_interruptible_locked( - sa_manager->wq, - amdgpu_sa_event(sa_manager, size, align) - ); - } - - } while (!r); - - spin_unlock(&sa_manager->wq.lock); - kfree(*sa_bo); - *sa_bo = NULL; - return r; -} - -void amdgpu_sa_bo_free(struct amdgpu_device *adev, struct amdgpu_sa_bo **sa_bo, +void amdgpu_sa_bo_free(struct amdgpu_device *adev, struct drm_suballoc **sa_bo, struct dma_fence *fence) { - struct amdgpu_sa_manager *sa_manager; - if (sa_bo == NULL || *sa_bo == NULL) { return; } - sa_manager = (*sa_bo)->manager; - spin_lock(&sa_manager->wq.lock); - if (fence && !dma_fence_is_signaled(fence)) { - uint32_t idx; - - (*sa_bo)->fence = dma_fence_get(fence); - idx = fence->context % AMDGPU_SA_NUM_FENCE_LISTS; - list_add_tail(&(*sa_bo)->flist, &sa_manager->flist[idx]); - } else { - amdgpu_sa_bo_remove_locked(*sa_bo); - } - wake_up_all_locked(&sa_manager->wq); - spin_unlock(&sa_manager->wq.lock); + drm_suballoc_free(*sa_bo, fence); *sa_bo = NULL; } @@ -373,26 +109,8 @@ void amdgpu_sa_bo_free(struct amdgpu_device *adev, struct amdgpu_sa_bo **sa_bo, void amdgpu_sa_bo_dump_debug_info(struct amdgpu_sa_manager *sa_manager, struct seq_file *m) { - struct amdgpu_sa_bo *i; - - spin_lock(&sa_manager->wq.lock); - list_for_each_entry(i, &sa_manager->olist, olist) { - uint64_t soffset = i->soffset + sa_manager->gpu_addr; - uint64_t eoffset = i->eoffset + sa_manager->gpu_addr; - if (&i->olist == sa_manager->hole) { - seq_printf(m, ">"); - } else { - seq_printf(m, " "); - } - seq_printf(m, "[0x%010llx 0x%010llx] size %8lld", - soffset, eoffset, eoffset - soffset); + struct drm_printer p = drm_seq_file_printer(m); - if (i->fence) - seq_printf(m, " protected by 0x%016llx on context %llu", - i->fence->seqno, i->fence->context); - - seq_printf(m, "\n"); - } - spin_unlock(&sa_manager->wq.lock); + drm_suballoc_dump_debug_info(&sa_manager->base, &p, sa_manager->gpu_addr); } #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.c index bac7976975bd..dcd8c066bc1f 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.c @@ -391,8 +391,10 @@ int amdgpu_sync_push_to_job(struct amdgpu_sync *sync, struct amdgpu_job *job) dma_fence_get(f); r = drm_sched_job_add_dependency(&job->base, f); - if (r) + if (r) { + dma_fence_put(f); return r; + } } return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c index c5ef7f7bdc15..2cd081cbf706 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c @@ -466,11 +466,7 @@ static int amdgpu_bo_move(struct ttm_buffer_object *bo, bool evict, return r; } - /* Can't move a pinned BO */ abo = ttm_to_amdgpu_bo(bo); - if (WARN_ON_ONCE(abo->tbo.pin_count > 0)) - return -EINVAL; - adev = amdgpu_ttm_adev(bo->bdev); if (!old_mem || (old_mem->mem_type == TTM_PL_SYSTEM && diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h index bee93ab4298f..b03321e7d2d8 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h @@ -538,6 +538,7 @@ struct amdgpu_firmware { void amdgpu_ucode_print_mc_hdr(const struct common_firmware_header *hdr); void amdgpu_ucode_print_smc_hdr(const struct common_firmware_header *hdr); +void amdgpu_ucode_print_imu_hdr(const struct common_firmware_header *hdr); void amdgpu_ucode_print_gfx_hdr(const struct common_firmware_header *hdr); void amdgpu_ucode_print_rlc_hdr(const struct common_firmware_header *hdr); void amdgpu_ucode_print_sdma_hdr(const struct common_firmware_header *hdr); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.c index 1c7fcb4f2380..1b8574bc4463 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.c @@ -68,7 +68,7 @@ int amdgpu_umc_page_retirement_mca(struct amdgpu_device *adev, if (amdgpu_bad_page_threshold != 0) { amdgpu_ras_add_bad_pages(adev, err_data.err_addr, err_data.err_addr_cnt); - amdgpu_ras_save_bad_pages(adev); + amdgpu_ras_save_bad_pages(adev, NULL); } out: @@ -147,7 +147,7 @@ static int amdgpu_umc_do_page_retirement(struct amdgpu_device *adev, err_data->err_addr_cnt) { amdgpu_ras_add_bad_pages(adev, err_data->err_addr, err_data->err_addr_cnt); - amdgpu_ras_save_bad_pages(adev); + amdgpu_ras_save_bad_pages(adev, &(err_data->ue_count)); amdgpu_dpm_send_hbm_bad_pages_num(adev, con->eeprom_control.ras_num_recs); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.h index a6951160f13a..36e19336f3b3 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.h @@ -42,7 +42,7 @@ #define LOOP_UMC_INST_AND_CH(umc_inst, ch_inst) LOOP_UMC_INST((umc_inst)) LOOP_UMC_CH_INST((ch_inst)) #define LOOP_UMC_NODE_INST(node_inst) \ - for ((node_inst) = 0; (node_inst) < adev->umc.node_inst_num; (node_inst)++) + for_each_set_bit((node_inst), &(adev->umc.active_mask), adev->umc.node_inst_num) #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)) @@ -69,17 +69,22 @@ struct amdgpu_umc { /* number of umc instance with memory map register access */ uint32_t umc_inst_num; - /*number of umc node instance with memory map register access*/ + /* Total number of umc node instance including harvest one */ uint32_t node_inst_num; /* UMC regiser per channel offset */ uint32_t channel_offs; + /* how many pages are retired in one UE */ + uint32_t retire_unit; /* channel index table of interleaved memory */ const uint32_t *channel_idx_tbl; struct ras_common_if *ras_if; const struct amdgpu_umc_funcs *funcs; struct amdgpu_umc_ras *ras; + + /* active mask for umc node instance */ + unsigned long active_mask; }; int amdgpu_umc_ras_late_init(struct amdgpu_device *adev, struct ras_common_if *ras_block); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c index f39391e03d46..f2e2cbaa7fde 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c @@ -983,11 +983,13 @@ static u32 amdgpu_virt_rlcg_reg_rw(struct amdgpu_device *adev, u32 offset, u32 v if (offset == reg_access_ctrl->grbm_cntl) { /* if the target reg offset is grbm_cntl, write to scratch_reg2 */ writel(v, scratch_reg2); - writel(v, ((void __iomem *)adev->rmmio) + (offset * 4)); + if (flag == AMDGPU_RLCG_GC_WRITE_LEGACY) + writel(v, ((void __iomem *)adev->rmmio) + (offset * 4)); } else if (offset == reg_access_ctrl->grbm_idx) { /* if the target reg offset is grbm_idx, write to scratch_reg3 */ writel(v, scratch_reg3); - writel(v, ((void __iomem *)adev->rmmio) + (offset * 4)); + if (flag == AMDGPU_RLCG_GC_WRITE_LEGACY) + writel(v, ((void __iomem *)adev->rmmio) + (offset * 4)); } else { /* * SCRATCH_REG0 = read/write value diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_pt.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_pt.c index b5f3bba851db..01e42bdd8e4e 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_pt.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_pt.c @@ -974,7 +974,7 @@ int amdgpu_vm_ptes_update(struct amdgpu_vm_update_params *params, trace_amdgpu_vm_update_ptes(params, frag_start, upd_end, min(nptes, 32u), dst, incr, upd_flags, - vm->task_info.pid, + vm->task_info.tgid, vm->immediate.fence_context); amdgpu_vm_pte_update_flags(params, to_amdgpu_bo_vm(pt), cursor.level, pe_start, dst, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c index 9fa1d814508a..43d6a9d6a538 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c @@ -453,7 +453,8 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man, /* Limit maximum size to 2GiB due to SG table limitations */ size = min(remaining_size, 2ULL << 30); - if (size >= (u64)pages_per_block << PAGE_SHIFT) + if ((size >= (u64)pages_per_block << PAGE_SHIFT) && + !(size & (((u64)pages_per_block << PAGE_SHIFT) - 1))) min_block_size = (u64)pages_per_block << PAGE_SHIFT; cur_size = size; diff --git a/drivers/gpu/drm/amd/amdgpu/df_v1_7.c b/drivers/gpu/drm/amd/amdgpu/df_v1_7.c index b991609f46c1..5dfab80ffff2 100644 --- a/drivers/gpu/drm/amd/amdgpu/df_v1_7.c +++ b/drivers/gpu/drm/amd/amdgpu/df_v1_7.c @@ -94,7 +94,7 @@ static void df_v1_7_update_medium_grain_clock_gating(struct amdgpu_device *adev, WREG32_SOC15(DF, 0, mmDF_PIE_AON0_DfGlobalClkGater, tmp); } - /* Exit boradcast mode */ + /* Exit broadcast mode */ adev->df.funcs->enable_broadcast_mode(adev, false); } diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c index c621b2ad7ba3..3bf697a80cf2 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c @@ -754,8 +754,8 @@ static void gfx_v11_0_read_wave_data(struct amdgpu_device *adev, uint32_t simd, * zero here */ WARN_ON(simd != 0); - /* type 2 wave data */ - dst[(*no_fields)++] = 2; + /* type 3 wave data */ + dst[(*no_fields)++] = 3; dst[(*no_fields)++] = wave_read_ind(adev, wave, ixSQ_WAVE_STATUS); dst[(*no_fields)++] = wave_read_ind(adev, wave, ixSQ_WAVE_PC_LO); dst[(*no_fields)++] = wave_read_ind(adev, wave, ixSQ_WAVE_PC_HI); @@ -1503,44 +1503,70 @@ static void gfx_v11_0_select_se_sh(struct amdgpu_device *adev, u32 se_num, WREG32_SOC15(GC, 0, regGRBM_GFX_INDEX, data); } -static u32 gfx_v11_0_get_rb_active_bitmap(struct amdgpu_device *adev) +static u32 gfx_v11_0_get_sa_active_bitmap(struct amdgpu_device *adev) { - u32 data, mask; + u32 gc_disabled_sa_mask, gc_user_disabled_sa_mask, sa_mask; + + gc_disabled_sa_mask = RREG32_SOC15(GC, 0, regCC_GC_SA_UNIT_DISABLE); + gc_disabled_sa_mask = REG_GET_FIELD(gc_disabled_sa_mask, + CC_GC_SA_UNIT_DISABLE, + SA_DISABLE); + gc_user_disabled_sa_mask = RREG32_SOC15(GC, 0, regGC_USER_SA_UNIT_DISABLE); + gc_user_disabled_sa_mask = REG_GET_FIELD(gc_user_disabled_sa_mask, + GC_USER_SA_UNIT_DISABLE, + SA_DISABLE); + sa_mask = amdgpu_gfx_create_bitmask(adev->gfx.config.max_sh_per_se * + adev->gfx.config.max_shader_engines); - data = RREG32_SOC15(GC, 0, regCC_RB_BACKEND_DISABLE); - data |= RREG32_SOC15(GC, 0, regGC_USER_RB_BACKEND_DISABLE); + return sa_mask & (~(gc_disabled_sa_mask | gc_user_disabled_sa_mask)); +} - data &= CC_RB_BACKEND_DISABLE__BACKEND_DISABLE_MASK; - data >>= GC_USER_RB_BACKEND_DISABLE__BACKEND_DISABLE__SHIFT; +static u32 gfx_v11_0_get_rb_active_bitmap(struct amdgpu_device *adev) +{ + u32 gc_disabled_rb_mask, gc_user_disabled_rb_mask; + u32 rb_mask; - mask = amdgpu_gfx_create_bitmask(adev->gfx.config.max_backends_per_se / - adev->gfx.config.max_sh_per_se); + gc_disabled_rb_mask = RREG32_SOC15(GC, 0, regCC_RB_BACKEND_DISABLE); + gc_disabled_rb_mask = REG_GET_FIELD(gc_disabled_rb_mask, + CC_RB_BACKEND_DISABLE, + BACKEND_DISABLE); + gc_user_disabled_rb_mask = RREG32_SOC15(GC, 0, regGC_USER_RB_BACKEND_DISABLE); + gc_user_disabled_rb_mask = REG_GET_FIELD(gc_user_disabled_rb_mask, + GC_USER_RB_BACKEND_DISABLE, + BACKEND_DISABLE); + rb_mask = amdgpu_gfx_create_bitmask(adev->gfx.config.max_backends_per_se * + adev->gfx.config.max_shader_engines); - return (~data) & mask; + return rb_mask & (~(gc_disabled_rb_mask | gc_user_disabled_rb_mask)); } static void gfx_v11_0_setup_rb(struct amdgpu_device *adev) { - int i, j; - u32 data; - u32 active_rbs = 0; - u32 rb_bitmap_width_per_sh = adev->gfx.config.max_backends_per_se / - adev->gfx.config.max_sh_per_se; + u32 rb_bitmap_width_per_sa; + u32 max_sa; + u32 active_sa_bitmap; + u32 global_active_rb_bitmap; + u32 active_rb_bitmap = 0; + u32 i; - mutex_lock(&adev->grbm_idx_mutex); - for (i = 0; i < adev->gfx.config.max_shader_engines; i++) { - for (j = 0; j < adev->gfx.config.max_sh_per_se; j++) { - gfx_v11_0_select_se_sh(adev, i, j, 0xffffffff); - data = gfx_v11_0_get_rb_active_bitmap(adev); - active_rbs |= data << ((i * adev->gfx.config.max_sh_per_se + j) * - rb_bitmap_width_per_sh); - } + /* query sa bitmap from SA_UNIT_DISABLE registers */ + active_sa_bitmap = gfx_v11_0_get_sa_active_bitmap(adev); + /* query rb bitmap from RB_BACKEND_DISABLE registers */ + global_active_rb_bitmap = gfx_v11_0_get_rb_active_bitmap(adev); + + /* generate active rb bitmap according to active sa bitmap */ + max_sa = adev->gfx.config.max_shader_engines * + adev->gfx.config.max_sh_per_se; + rb_bitmap_width_per_sa = adev->gfx.config.max_backends_per_se / + adev->gfx.config.max_sh_per_se; + for (i = 0; i < max_sa; i++) { + if (active_sa_bitmap & (1 << i)) + active_rb_bitmap |= (0x3 << (i * rb_bitmap_width_per_sa)); } - gfx_v11_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff); - mutex_unlock(&adev->grbm_idx_mutex); - adev->gfx.config.backend_enable_mask = active_rbs; - adev->gfx.config.num_rbs = hweight32(active_rbs); + 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); } #define DEFAULT_SH_MEM_BASES (0x6000) @@ -1633,6 +1659,11 @@ static void gfx_v11_0_constants_init(struct amdgpu_device *adev) gfx_v11_0_get_tcc_info(adev); adev->gfx.config.pa_sc_tile_steering_override = 0; + /* Set whether texture coordinate truncation is conformant. */ + tmp = RREG32_SOC15(GC, 0, regTA_CNTL2); + adev->gfx.config.ta_cntl2_truncate_coord_mode = + REG_GET_FIELD(tmp, TA_CNTL2, TRUNCATE_COORD_MODE); + /* XXX SH_MEM regs */ /* where to put LDS, scratch, GPUVM in FSA64 space */ mutex_lock(&adev->srbm_mutex); diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c index 8ad5c03506f2..ae09fc1cfe6b 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c @@ -6783,7 +6783,6 @@ static const struct amdgpu_ring_funcs gfx_v9_0_ring_funcs_gfx = { .emit_gds_switch = gfx_v9_0_ring_emit_gds_switch, .emit_hdp_flush = gfx_v9_0_ring_emit_hdp_flush, .test_ring = gfx_v9_0_ring_test_ring, - .test_ib = gfx_v9_0_ring_test_ib, .insert_nop = amdgpu_ring_insert_nop, .pad_ib = amdgpu_ring_generic_pad_ib, .emit_switch_buffer = gfx_v9_ring_emit_sb, diff --git a/drivers/gpu/drm/amd/amdgpu/gfxhub_v3_0.c b/drivers/gpu/drm/amd/amdgpu/gfxhub_v3_0.c index fa42d1907dfa..be0d0f47415e 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfxhub_v3_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfxhub_v3_0.c @@ -151,16 +151,17 @@ static void gfxhub_v3_0_init_system_aperture_regs(struct amdgpu_device *adev) { uint64_t value; - /* Disable AGP. */ + /* Program the AGP BAR */ WREG32_SOC15(GC, 0, regGCMC_VM_AGP_BASE, 0); - WREG32_SOC15(GC, 0, regGCMC_VM_AGP_TOP, 0); - WREG32_SOC15(GC, 0, regGCMC_VM_AGP_BOT, 0x00FFFFFF); + WREG32_SOC15(GC, 0, regGCMC_VM_AGP_BOT, adev->gmc.agp_start >> 24); + WREG32_SOC15(GC, 0, regGCMC_VM_AGP_TOP, adev->gmc.agp_end >> 24); + /* Program the system aperture low logical page number. */ WREG32_SOC15(GC, 0, regGCMC_VM_SYSTEM_APERTURE_LOW_ADDR, - adev->gmc.vram_start >> 18); + min(adev->gmc.fb_start, adev->gmc.agp_start) >> 18); WREG32_SOC15(GC, 0, regGCMC_VM_SYSTEM_APERTURE_HIGH_ADDR, - adev->gmc.vram_end >> 18); + max(adev->gmc.fb_end, adev->gmc.agp_end) >> 18); /* Set default page address. */ value = adev->mem_scratch.gpu_addr - adev->gmc.vram_start diff --git a/drivers/gpu/drm/amd/amdgpu/gfxhub_v3_0_3.c b/drivers/gpu/drm/amd/amdgpu/gfxhub_v3_0_3.c index 3dc17a3deedb..6e0bd628c889 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfxhub_v3_0_3.c +++ b/drivers/gpu/drm/amd/amdgpu/gfxhub_v3_0_3.c @@ -159,14 +159,14 @@ static void gfxhub_v3_0_3_init_system_aperture_regs(struct amdgpu_device *adev) /* Disable AGP. */ WREG32_SOC15(GC, 0, regGCMC_VM_AGP_BASE, 0); - WREG32_SOC15(GC, 0, regGCMC_VM_AGP_TOP, 0); - WREG32_SOC15(GC, 0, regGCMC_VM_AGP_BOT, 0x00FFFFFF); + WREG32_SOC15(GC, 0, regGCMC_VM_AGP_BOT, adev->gmc.agp_start >> 24); + WREG32_SOC15(GC, 0, regGCMC_VM_AGP_TOP, adev->gmc.agp_end >> 24); /* Program the system aperture low logical page number. */ WREG32_SOC15(GC, 0, regGCMC_VM_SYSTEM_APERTURE_LOW_ADDR, - adev->gmc.vram_start >> 18); + min(adev->gmc.fb_start, adev->gmc.agp_start) >> 18); WREG32_SOC15(GC, 0, regGCMC_VM_SYSTEM_APERTURE_HIGH_ADDR, - adev->gmc.vram_end >> 18); + max(adev->gmc.fb_end, adev->gmc.agp_end) >> 18); /* Set default page address. */ value = adev->mem_scratch.gpu_addr - adev->gmc.vram_start diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c index 7db1f1a7e33c..ab2556ca984e 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c @@ -692,6 +692,7 @@ static void gmc_v10_0_set_umc_funcs(struct amdgpu_device *adev) adev->umc.channel_inst_num = UMC_V8_7_CHANNEL_INSTANCE_NUM; adev->umc.umc_inst_num = UMC_V8_7_UMC_INSTANCE_NUM; adev->umc.channel_offs = UMC_V8_7_PER_CHANNEL_OFFSET_SIENNA; + adev->umc.retire_unit = 1; adev->umc.channel_idx_tbl = &umc_v8_7_channel_idx_tbl[0][0]; adev->umc.ras = &umc_v8_7_ras; break; diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c index 5e0018fe7e7d..af7b3ba1ca00 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c @@ -567,9 +567,9 @@ static void gmc_v11_0_set_umc_funcs(struct amdgpu_device *adev) case IP_VERSION(8, 10, 0): adev->umc.channel_inst_num = UMC_V8_10_CHANNEL_INSTANCE_NUM; adev->umc.umc_inst_num = UMC_V8_10_UMC_INSTANCE_NUM; - adev->umc.node_inst_num = adev->gmc.num_umc; adev->umc.max_ras_err_cnt_per_query = UMC_V8_10_TOTAL_CHANNEL_NUM(adev); adev->umc.channel_offs = UMC_V8_10_PER_CHANNEL_OFFSET; + adev->umc.retire_unit = UMC_V8_10_NA_COL_2BITS_POWER_OF_2_NUM; if (adev->umc.node_inst_num == 4) adev->umc.channel_idx_tbl = &umc_v8_10_channel_idx_tbl_ext0[0][0][0]; else @@ -673,6 +673,7 @@ static void gmc_v11_0_vram_gtt_location(struct amdgpu_device *adev, amdgpu_gmc_vram_location(adev, &adev->gmc, base); amdgpu_gmc_gart_location(adev, mc); + amdgpu_gmc_agp_location(adev, mc); /* base offset of vram pages */ if (amdgpu_sriov_vf(adev)) diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c index d65c6cea3445..b06170c00dfc 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c @@ -1288,6 +1288,7 @@ static void gmc_v9_0_set_umc_funcs(struct amdgpu_device *adev) adev->umc.channel_inst_num = UMC_V6_1_CHANNEL_INSTANCE_NUM; adev->umc.umc_inst_num = UMC_V6_1_UMC_INSTANCE_NUM; adev->umc.channel_offs = UMC_V6_1_PER_CHANNEL_OFFSET_VG20; + adev->umc.retire_unit = 1; adev->umc.channel_idx_tbl = &umc_v6_1_channel_idx_tbl[0][0]; adev->umc.ras = &umc_v6_1_ras; break; @@ -1296,6 +1297,7 @@ static void gmc_v9_0_set_umc_funcs(struct amdgpu_device *adev) adev->umc.channel_inst_num = UMC_V6_1_CHANNEL_INSTANCE_NUM; adev->umc.umc_inst_num = UMC_V6_1_UMC_INSTANCE_NUM; adev->umc.channel_offs = UMC_V6_1_PER_CHANNEL_OFFSET_ARCT; + adev->umc.retire_unit = 1; adev->umc.channel_idx_tbl = &umc_v6_1_channel_idx_tbl[0][0]; adev->umc.ras = &umc_v6_1_ras; break; @@ -1305,6 +1307,7 @@ static void gmc_v9_0_set_umc_funcs(struct amdgpu_device *adev) adev->umc.channel_inst_num = UMC_V6_7_CHANNEL_INSTANCE_NUM; adev->umc.umc_inst_num = UMC_V6_7_UMC_INSTANCE_NUM; adev->umc.channel_offs = UMC_V6_7_PER_CHANNEL_OFFSET; + adev->umc.retire_unit = (UMC_V6_7_NA_MAP_PA_NUM * 2); if (!adev->gmc.xgmi.connected_to_cpu) adev->umc.ras = &umc_v6_7_ras; if (1 & adev->smuio.funcs->get_die_id(adev)) diff --git a/drivers/gpu/drm/amd/amdgpu/imu_v11_0.c b/drivers/gpu/drm/amd/amdgpu/imu_v11_0.c index ed0d368149aa..4ab90c7852c3 100644 --- a/drivers/gpu/drm/amd/amdgpu/imu_v11_0.c +++ b/drivers/gpu/drm/amd/amdgpu/imu_v11_0.c @@ -35,6 +35,7 @@ MODULE_FIRMWARE("amdgpu/gc_11_0_0_imu.bin"); MODULE_FIRMWARE("amdgpu/gc_11_0_1_imu.bin"); MODULE_FIRMWARE("amdgpu/gc_11_0_2_imu.bin"); MODULE_FIRMWARE("amdgpu/gc_11_0_3_imu.bin"); +MODULE_FIRMWARE("amdgpu/gc_11_0_4_imu.bin"); static int imu_v11_0_init_microcode(struct amdgpu_device *adev) { diff --git a/drivers/gpu/drm/amd/amdgpu/mes_v11_0.c b/drivers/gpu/drm/amd/amdgpu/mes_v11_0.c index bfa305079bfc..5826eac270d7 100644 --- a/drivers/gpu/drm/amd/amdgpu/mes_v11_0.c +++ b/drivers/gpu/drm/amd/amdgpu/mes_v11_0.c @@ -40,6 +40,8 @@ MODULE_FIRMWARE("amdgpu/gc_11_0_2_mes.bin"); MODULE_FIRMWARE("amdgpu/gc_11_0_2_mes1.bin"); MODULE_FIRMWARE("amdgpu/gc_11_0_3_mes.bin"); MODULE_FIRMWARE("amdgpu/gc_11_0_3_mes1.bin"); +MODULE_FIRMWARE("amdgpu/gc_11_0_4_mes.bin"); +MODULE_FIRMWARE("amdgpu/gc_11_0_4_mes1.bin"); static int mes_v11_0_hw_fini(void *handle); static int mes_v11_0_kiq_hw_init(struct amdgpu_device *adev); @@ -196,7 +198,6 @@ static int mes_v11_0_add_hw_queue(struct amdgpu_mes *mes, mes_add_queue_pkt.trap_handler_addr = input->tba_addr; mes_add_queue_pkt.tma_addr = input->tma_addr; mes_add_queue_pkt.is_kfd_process = input->is_kfd_process; - mes_add_queue_pkt.trap_en = 1; /* For KFD, gds_size is re-used for queue size (needed in MES for AQL queues) */ mes_add_queue_pkt.is_aql_queue = input->is_aql_queue; @@ -1283,7 +1284,7 @@ static int mes_v11_0_late_init(void *handle) struct amdgpu_device *adev = (struct amdgpu_device *)handle; /* it's only intended for use in mes_self_test case, not for s0ix and reset */ - if (!amdgpu_in_reset(adev) && !adev->in_s0ix && + if (!amdgpu_in_reset(adev) && !adev->in_s0ix && !adev->in_suspend && (adev->ip_versions[GC_HWIP][0] != IP_VERSION(11, 0, 3))) amdgpu_mes_self_test(adev); diff --git a/drivers/gpu/drm/amd/amdgpu/mmhub_v3_0.c b/drivers/gpu/drm/amd/amdgpu/mmhub_v3_0.c index ae9cd1a4cfee..164948c50ac3 100644 --- a/drivers/gpu/drm/amd/amdgpu/mmhub_v3_0.c +++ b/drivers/gpu/drm/amd/amdgpu/mmhub_v3_0.c @@ -177,15 +177,16 @@ static void mmhub_v3_0_init_system_aperture_regs(struct amdgpu_device *adev) * these regs, and they will be programed at host. * so skip programing these regs. */ - /* Disable AGP. */ + /* Program the AGP BAR */ WREG32_SOC15(MMHUB, 0, regMMMC_VM_AGP_BASE, 0); - WREG32_SOC15(MMHUB, 0, regMMMC_VM_AGP_TOP, 0); - WREG32_SOC15(MMHUB, 0, regMMMC_VM_AGP_BOT, 0x00FFFFFF); + WREG32_SOC15(MMHUB, 0, regMMMC_VM_AGP_BOT, adev->gmc.agp_start >> 24); + WREG32_SOC15(MMHUB, 0, regMMMC_VM_AGP_TOP, adev->gmc.agp_end >> 24); + /* Program the system aperture low logical page number. */ WREG32_SOC15(MMHUB, 0, regMMMC_VM_SYSTEM_APERTURE_LOW_ADDR, - adev->gmc.vram_start >> 18); + min(adev->gmc.fb_start, adev->gmc.agp_start) >> 18); WREG32_SOC15(MMHUB, 0, regMMMC_VM_SYSTEM_APERTURE_HIGH_ADDR, - adev->gmc.vram_end >> 18); + max(adev->gmc.fb_end, adev->gmc.agp_end) >> 18); /* Set default page address. */ value = adev->mem_scratch.gpu_addr - adev->gmc.vram_start + diff --git a/drivers/gpu/drm/amd/amdgpu/mmhub_v3_0_1.c b/drivers/gpu/drm/amd/amdgpu/mmhub_v3_0_1.c index c8d478f2afdc..26509b6b8c24 100644 --- a/drivers/gpu/drm/amd/amdgpu/mmhub_v3_0_1.c +++ b/drivers/gpu/drm/amd/amdgpu/mmhub_v3_0_1.c @@ -183,9 +183,9 @@ static void mmhub_v3_0_1_init_system_aperture_regs(struct amdgpu_device *adev) */ /* Program the system aperture low logical page number. */ WREG32_SOC15(MMHUB, 0, regMMMC_VM_SYSTEM_APERTURE_LOW_ADDR, - adev->gmc.vram_start >> 18); + min(adev->gmc.fb_start, adev->gmc.agp_start) >> 18); WREG32_SOC15(MMHUB, 0, regMMMC_VM_SYSTEM_APERTURE_HIGH_ADDR, - adev->gmc.vram_end >> 18); + max(adev->gmc.fb_end, adev->gmc.agp_end) >> 18); /* Set default page address. */ value = adev->mem_scratch.gpu_addr - adev->gmc.vram_start + diff --git a/drivers/gpu/drm/amd/amdgpu/mmhub_v3_0_2.c b/drivers/gpu/drm/amd/amdgpu/mmhub_v3_0_2.c index c30e40e52fb2..26abbc6a47ab 100644 --- a/drivers/gpu/drm/amd/amdgpu/mmhub_v3_0_2.c +++ b/drivers/gpu/drm/amd/amdgpu/mmhub_v3_0_2.c @@ -162,10 +162,10 @@ static void mmhub_v3_0_2_init_system_aperture_regs(struct amdgpu_device *adev) uint64_t value; uint32_t tmp; - /* Disable AGP. */ + /* Program the AGP BAR */ WREG32_SOC15(MMHUB, 0, regMMMC_VM_AGP_BASE, 0); - WREG32_SOC15(MMHUB, 0, regMMMC_VM_AGP_TOP, 0); - WREG32_SOC15(MMHUB, 0, regMMMC_VM_AGP_BOT, 0x00FFFFFF); + WREG32_SOC15(MMHUB, 0, regMMMC_VM_AGP_BOT, adev->gmc.agp_start >> 24); + WREG32_SOC15(MMHUB, 0, regMMMC_VM_AGP_TOP, adev->gmc.agp_end >> 24); if (!amdgpu_sriov_vf(adev)) { /* @@ -175,9 +175,9 @@ static void mmhub_v3_0_2_init_system_aperture_regs(struct amdgpu_device *adev) */ /* Program the system aperture low logical page number. */ WREG32_SOC15(MMHUB, 0, regMMMC_VM_SYSTEM_APERTURE_LOW_ADDR, - adev->gmc.vram_start >> 18); + min(adev->gmc.fb_start, adev->gmc.agp_start) >> 18); WREG32_SOC15(MMHUB, 0, regMMMC_VM_SYSTEM_APERTURE_HIGH_ADDR, - adev->gmc.vram_end >> 18); + max(adev->gmc.fb_end, adev->gmc.agp_end) >> 18); } /* Set default page address. */ diff --git a/drivers/gpu/drm/amd/amdgpu/nbio_v4_3.c b/drivers/gpu/drm/amd/amdgpu/nbio_v4_3.c index 15eb3658d70e..09fdcd20cb91 100644 --- a/drivers/gpu/drm/amd/amdgpu/nbio_v4_3.c +++ b/drivers/gpu/drm/amd/amdgpu/nbio_v4_3.c @@ -337,7 +337,13 @@ const struct nbio_hdp_flush_reg nbio_v4_3_hdp_flush_reg = { static void nbio_v4_3_init_registers(struct amdgpu_device *adev) { - return; + if (adev->ip_versions[NBIO_HWIP][0] == IP_VERSION(4, 3, 0)) { + uint32_t data; + + data = RREG32_SOC15(NBIO, 0, regRCC_DEV0_EPF2_STRAP2); + data &= ~RCC_DEV0_EPF2_STRAP2__STRAP_NO_SOFT_RESET_DEV0_F2_MASK; + WREG32_SOC15(NBIO, 0, regRCC_DEV0_EPF2_STRAP2, data); + } } static u32 nbio_v4_3_get_rom_offset(struct amdgpu_device *adev) diff --git a/drivers/gpu/drm/amd/amdgpu/nbio_v7_2.c b/drivers/gpu/drm/amd/amdgpu/nbio_v7_2.c index 31776b12e4c4..4ef1fa4603c8 100644 --- a/drivers/gpu/drm/amd/amdgpu/nbio_v7_2.c +++ b/drivers/gpu/drm/amd/amdgpu/nbio_v7_2.c @@ -394,6 +394,15 @@ static void nbio_v7_2_init_registers(struct amdgpu_device *adev) break; } + switch (adev->ip_versions[NBIO_HWIP][0]) { + case IP_VERSION(7, 3, 0): + case IP_VERSION(7, 5, 1): + data = RREG32_SOC15(NBIO, 0, regRCC_DEV2_EPF0_STRAP2); + data &= ~RCC_DEV2_EPF0_STRAP2__STRAP_NO_SOFT_RESET_DEV2_F0_MASK; + WREG32_SOC15(NBIO, 0, regRCC_DEV2_EPF0_STRAP2, data); + break; + } + if (amdgpu_sriov_vf(adev)) adev->rmmio_remap.reg_offset = SOC15_REG_OFFSET(NBIO, 0, regBIF_BX_PF0_HDP_MEM_COHERENCY_FLUSH_CNTL) << 2; diff --git a/drivers/gpu/drm/amd/amdgpu/nv.c b/drivers/gpu/drm/amd/amdgpu/nv.c index d972025f0d20..855d390c41de 100644 --- a/drivers/gpu/drm/amd/amdgpu/nv.c +++ b/drivers/gpu/drm/amd/amdgpu/nv.c @@ -444,9 +444,10 @@ static int nv_read_register(struct amdgpu_device *adev, u32 se_num, *value = 0; for (i = 0; i < ARRAY_SIZE(nv_allowed_read_registers); i++) { en = &nv_allowed_read_registers[i]; - if (adev->reg_offset[en->hwip][en->inst] && - reg_offset != (adev->reg_offset[en->hwip][en->inst][en->seg] - + en->reg_offset)) + if (!adev->reg_offset[en->hwip][en->inst]) + continue; + else if (reg_offset != (adev->reg_offset[en->hwip][en->inst][en->seg] + + en->reg_offset)) continue; *value = nv_get_register_value(adev, diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v6_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v6_0.c index 3d36329be384..40e6b22daa22 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v6_0.c @@ -273,8 +273,6 @@ static void sdma_v6_0_ring_emit_ib(struct amdgpu_ring *ring, * sdma_v6_0_ring_emit_mem_sync - flush the IB by graphics cache rinse * * @ring: amdgpu ring pointer - * @job: job to retrieve vmid from - * @ib: IB object to schedule * * flush the IB by graphics cache rinse. */ @@ -326,7 +324,9 @@ static void sdma_v6_0_ring_emit_hdp_flush(struct amdgpu_ring *ring) * sdma_v6_0_ring_emit_fence - emit a fence on the DMA ring * * @ring: amdgpu ring pointer - * @fence: amdgpu fence object + * @addr: address + * @seq: fence seq number + * @flags: fence flags * * Add a DMA fence packet to the ring to write * the fence seq number and DMA trap packet to generate @@ -1060,10 +1060,9 @@ static void sdma_v6_0_vm_copy_pte(struct amdgpu_ib *ib, * * @ib: indirect buffer to fill with commands * @pe: addr of the page entry - * @addr: dst addr to write into pe + * @value: dst addr to write into pe * @count: number of page entries to update * @incr: increase next addr by incr bytes - * @flags: access flags * * Update PTEs by writing them manually using sDMA. */ @@ -1167,7 +1166,6 @@ static void sdma_v6_0_ring_emit_pipeline_sync(struct amdgpu_ring *ring) * sdma_v6_0_ring_emit_vm_flush - vm flush using sDMA * * @ring: amdgpu_ring pointer - * @vm: amdgpu_vm pointer * * Update the page table base and flush the VM TLB * using sDMA. @@ -1591,10 +1589,11 @@ static void sdma_v6_0_set_irq_funcs(struct amdgpu_device *adev) /** * sdma_v6_0_emit_copy_buffer - copy buffer using the sDMA engine * - * @ring: amdgpu_ring structure holding ring information + * @ib: indirect buffer to fill with commands * @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 GPU buffers using the DMA engine. * Used by the amdgpu ttm implementation to move pages if @@ -1620,7 +1619,7 @@ static void sdma_v6_0_emit_copy_buffer(struct amdgpu_ib *ib, /** * sdma_v6_0_emit_fill_buffer - fill buffer using the sDMA engine * - * @ring: amdgpu_ring structure holding ring information + * @ib: indirect buffer to fill * @src_data: value to write to buffer * @dst_offset: dst GPU address * @byte_count: number of bytes to xfer diff --git a/drivers/gpu/drm/amd/amdgpu/smu_v13_0_10.c b/drivers/gpu/drm/amd/amdgpu/smu_v13_0_10.c new file mode 100644 index 000000000000..ae29620b1ea4 --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/smu_v13_0_10.c @@ -0,0 +1,303 @@ +/* + * 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 "smu_v13_0_10.h" +#include "amdgpu_reset.h" +#include "amdgpu_dpm.h" +#include "amdgpu_job.h" +#include "amdgpu_ring.h" +#include "amdgpu_ras.h" +#include "amdgpu_psp.h" + +static bool smu_v13_0_10_is_mode2_default(struct amdgpu_reset_control *reset_ctl) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)reset_ctl->handle; + if (adev->pm.fw_version >= 0x00502005 && !amdgpu_sriov_vf(adev)) + return true; + + return false; +} + +static struct amdgpu_reset_handler * +smu_v13_0_10_get_reset_handler(struct amdgpu_reset_control *reset_ctl, + struct amdgpu_reset_context *reset_context) +{ + struct amdgpu_reset_handler *handler; + struct amdgpu_device *adev = (struct amdgpu_device *)reset_ctl->handle; + + if (reset_context->method != AMD_RESET_METHOD_NONE) { + list_for_each_entry(handler, &reset_ctl->reset_handlers, + handler_list) { + if (handler->reset_method == reset_context->method) + return handler; + } + } + + if (smu_v13_0_10_is_mode2_default(reset_ctl) && + amdgpu_asic_reset_method(adev) == AMD_RESET_METHOD_MODE2) { + list_for_each_entry (handler, &reset_ctl->reset_handlers, + handler_list) { + if (handler->reset_method == AMD_RESET_METHOD_MODE2) + return handler; + } + } + + return NULL; +} + +static int smu_v13_0_10_mode2_suspend_ip(struct amdgpu_device *adev) +{ + int r, i; + + amdgpu_device_set_pg_state(adev, AMD_PG_STATE_UNGATE); + amdgpu_device_set_cg_state(adev, AMD_CG_STATE_UNGATE); + + for (i = adev->num_ip_blocks - 1; i >= 0; i--) { + if (!(adev->ip_blocks[i].version->type == + AMD_IP_BLOCK_TYPE_GFX || + adev->ip_blocks[i].version->type == + AMD_IP_BLOCK_TYPE_SDMA || + adev->ip_blocks[i].version->type == + AMD_IP_BLOCK_TYPE_MES)) + continue; + + r = adev->ip_blocks[i].version->funcs->suspend(adev); + + if (r) { + dev_err(adev->dev, + "suspend of IP block <%s> failed %d\n", + adev->ip_blocks[i].version->funcs->name, r); + return r; + } + adev->ip_blocks[i].status.hw = false; + } + + return r; +} + +static int +smu_v13_0_10_mode2_prepare_hwcontext(struct amdgpu_reset_control *reset_ctl, + struct amdgpu_reset_context *reset_context) +{ + int r = 0; + struct amdgpu_device *adev = (struct amdgpu_device *)reset_ctl->handle; + + if (!amdgpu_sriov_vf(adev)) + r = smu_v13_0_10_mode2_suspend_ip(adev); + + return r; +} + +static int smu_v13_0_10_mode2_reset(struct amdgpu_device *adev) +{ + return amdgpu_dpm_mode2_reset(adev); +} + +static void smu_v13_0_10_async_reset(struct work_struct *work) +{ + struct amdgpu_reset_handler *handler; + struct amdgpu_reset_control *reset_ctl = + container_of(work, struct amdgpu_reset_control, reset_work); + struct amdgpu_device *adev = (struct amdgpu_device *)reset_ctl->handle; + + list_for_each_entry(handler, &reset_ctl->reset_handlers, + handler_list) { + if (handler->reset_method == reset_ctl->active_reset) { + dev_dbg(adev->dev, "Resetting device\n"); + handler->do_reset(adev); + break; + } + } +} +static int +smu_v13_0_10_mode2_perform_reset(struct amdgpu_reset_control *reset_ctl, + struct amdgpu_reset_context *reset_context) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)reset_ctl->handle; + int r; + + r = smu_v13_0_10_mode2_reset(adev); + if (r) { + dev_err(adev->dev, + "ASIC reset failed with error, %d ", r); + } + return r; +} + +static int smu_v13_0_10_mode2_restore_ip(struct amdgpu_device *adev) +{ + int i, r; + struct psp_context *psp = &adev->psp; + struct amdgpu_firmware_info *ucode; + struct amdgpu_firmware_info *ucode_list[2]; + int ucode_count = 0; + + for (i = 0; i < adev->firmware.max_ucodes; i++) { + ucode = &adev->firmware.ucode[i]; + + switch (ucode->ucode_id) { + case AMDGPU_UCODE_ID_IMU_I: + case AMDGPU_UCODE_ID_IMU_D: + ucode_list[ucode_count++] = ucode; + break; + default: + break; + } + } + + r = psp_load_fw_list(psp, ucode_list, ucode_count); + if (r) { + dev_err(adev->dev, "IMU ucode load failed after mode2 reset\n"); + return r; + } + + r = psp_rlc_autoload_start(psp); + if (r) { + DRM_ERROR("Failed to start rlc autoload after mode2 reset\n"); + return r; + } + + amdgpu_dpm_enable_gfx_features(adev); + + for (i = 0; i < adev->num_ip_blocks; i++) { + if (!(adev->ip_blocks[i].version->type == + AMD_IP_BLOCK_TYPE_GFX || + adev->ip_blocks[i].version->type == + AMD_IP_BLOCK_TYPE_MES || + adev->ip_blocks[i].version->type == + AMD_IP_BLOCK_TYPE_SDMA)) + continue; + r = adev->ip_blocks[i].version->funcs->resume(adev); + if (r) { + dev_err(adev->dev, + "resume of IP block <%s> failed %d\n", + adev->ip_blocks[i].version->funcs->name, r); + return r; + } + + adev->ip_blocks[i].status.hw = true; + } + + for (i = 0; i < adev->num_ip_blocks; i++) { + if (!(adev->ip_blocks[i].version->type == + AMD_IP_BLOCK_TYPE_GFX || + adev->ip_blocks[i].version->type == + AMD_IP_BLOCK_TYPE_MES || + adev->ip_blocks[i].version->type == + AMD_IP_BLOCK_TYPE_SDMA)) + continue; + + if (adev->ip_blocks[i].version->funcs->late_init) { + r = adev->ip_blocks[i].version->funcs->late_init( + (void *)adev); + if (r) { + dev_err(adev->dev, + "late_init of IP block <%s> failed %d after reset\n", + adev->ip_blocks[i].version->funcs->name, + r); + return r; + } + } + adev->ip_blocks[i].status.late_initialized = true; + } + + amdgpu_device_set_cg_state(adev, AMD_CG_STATE_GATE); + amdgpu_device_set_pg_state(adev, AMD_PG_STATE_GATE); + + return r; +} + +static int +smu_v13_0_10_mode2_restore_hwcontext(struct amdgpu_reset_control *reset_ctl, + struct amdgpu_reset_context *reset_context) +{ + int r; + struct amdgpu_device *tmp_adev = (struct amdgpu_device *)reset_ctl->handle; + + dev_info(tmp_adev->dev, + "GPU reset succeeded, trying to resume\n"); + r = smu_v13_0_10_mode2_restore_ip(tmp_adev); + if (r) + goto end; + + amdgpu_register_gpu_instance(tmp_adev); + + /* Resume RAS */ + amdgpu_ras_resume(tmp_adev); + + amdgpu_irq_gpu_reset_resume_helper(tmp_adev); + + r = amdgpu_ib_ring_tests(tmp_adev); + if (r) { + dev_err(tmp_adev->dev, + "ib ring test failed (%d).\n", r); + r = -EAGAIN; + goto end; + } + +end: + if (r) + return -EAGAIN; + else + return r; +} + +static struct amdgpu_reset_handler smu_v13_0_10_mode2_handler = { + .reset_method = AMD_RESET_METHOD_MODE2, + .prepare_env = NULL, + .prepare_hwcontext = smu_v13_0_10_mode2_prepare_hwcontext, + .perform_reset = smu_v13_0_10_mode2_perform_reset, + .restore_hwcontext = smu_v13_0_10_mode2_restore_hwcontext, + .restore_env = NULL, + .do_reset = smu_v13_0_10_mode2_reset, +}; + +int smu_v13_0_10_reset_init(struct amdgpu_device *adev) +{ + struct amdgpu_reset_control *reset_ctl; + + reset_ctl = kzalloc(sizeof(*reset_ctl), GFP_KERNEL); + if (!reset_ctl) + return -ENOMEM; + + reset_ctl->handle = adev; + reset_ctl->async_reset = smu_v13_0_10_async_reset; + reset_ctl->active_reset = AMD_RESET_METHOD_NONE; + reset_ctl->get_reset_handler = smu_v13_0_10_get_reset_handler; + + INIT_LIST_HEAD(&reset_ctl->reset_handlers); + INIT_WORK(&reset_ctl->reset_work, reset_ctl->async_reset); + /* Only mode2 is handled through reset control now */ + amdgpu_reset_add_handler(reset_ctl, &smu_v13_0_10_mode2_handler); + + adev->reset_cntl = reset_ctl; + + return 0; +} + +int smu_v13_0_10_reset_fini(struct amdgpu_device *adev) +{ + kfree(adev->reset_cntl); + adev->reset_cntl = NULL; + return 0; +} diff --git a/drivers/gpu/drm/amd/amdgpu/smu_v13_0_10.h b/drivers/gpu/drm/amd/amdgpu/smu_v13_0_10.h new file mode 100644 index 000000000000..e0cb72a0eec6 --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/smu_v13_0_10.h @@ -0,0 +1,32 @@ +/* + * 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_V13_0_10_H__ +#define __SMU_V13_0_10_H__ + +#include "amdgpu.h" + +int smu_v13_0_10_reset_init(struct amdgpu_device *adev); +int smu_v13_0_10_reset_fini(struct amdgpu_device *adev); + +#endif diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c index 7cd17dda32ce..2eddd7f6cd41 100644 --- a/drivers/gpu/drm/amd/amdgpu/soc15.c +++ b/drivers/gpu/drm/amd/amdgpu/soc15.c @@ -439,8 +439,9 @@ static int soc15_read_register(struct amdgpu_device *adev, u32 se_num, *value = 0; for (i = 0; i < ARRAY_SIZE(soc15_allowed_read_registers); i++) { en = &soc15_allowed_read_registers[i]; - if (adev->reg_offset[en->hwip][en->inst] && - reg_offset != (adev->reg_offset[en->hwip][en->inst][en->seg] + if (!adev->reg_offset[en->hwip][en->inst]) + continue; + else if (reg_offset != (adev->reg_offset[en->hwip][en->inst][en->seg] + en->reg_offset)) continue; diff --git a/drivers/gpu/drm/amd/amdgpu/soc21.c b/drivers/gpu/drm/amd/amdgpu/soc21.c index 9c4a29d50f1c..061793d390cc 100644 --- a/drivers/gpu/drm/amd/amdgpu/soc21.c +++ b/drivers/gpu/drm/amd/amdgpu/soc21.c @@ -52,6 +52,7 @@ static const struct amdgpu_video_codec_info vcn_4_0_0_video_codecs_encode_array_ { {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_MPEG4_AVC, 4096, 2304, 0)}, {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_HEVC, 4096, 2304, 0)}, + {codec_info_build(AMDGPU_INFO_VIDEO_CAPS_CODEC_IDX_AV1, 8192, 4352, 0)}, }; static const struct amdgpu_video_codec_info vcn_4_0_0_video_codecs_encode_array_vcn1[] = @@ -110,6 +111,7 @@ static int soc21_query_video_codecs(struct amdgpu_device *adev, bool encode, switch (adev->ip_versions[UVD_HWIP][0]) { case IP_VERSION(4, 0, 0): case IP_VERSION(4, 0, 2): + case IP_VERSION(4, 0, 4): if (adev->vcn.harvest_config & AMDGPU_VCN_HARVEST_VCN0) { if (encode) *codecs = &vcn_4_0_0_video_codecs_encode_vcn1; @@ -290,9 +292,10 @@ static int soc21_read_register(struct amdgpu_device *adev, u32 se_num, *value = 0; for (i = 0; i < ARRAY_SIZE(soc21_allowed_read_registers); i++) { en = &soc21_allowed_read_registers[i]; - if (adev->reg_offset[en->hwip][en->inst] && - reg_offset != (adev->reg_offset[en->hwip][en->inst][en->seg] - + en->reg_offset)) + if (!adev->reg_offset[en->hwip][en->inst]) + continue; + else if (reg_offset != (adev->reg_offset[en->hwip][en->inst][en->seg] + + en->reg_offset)) continue; *value = soc21_get_register_value(adev, @@ -675,7 +678,10 @@ static int soc21_common_early_init(void *handle) AMD_CG_SUPPORT_GFX_CGCG | AMD_CG_SUPPORT_GFX_CGLS | AMD_CG_SUPPORT_REPEATER_FGCG | - AMD_CG_SUPPORT_GFX_MGCG; + AMD_CG_SUPPORT_GFX_MGCG | + AMD_CG_SUPPORT_HDP_SD | + AMD_CG_SUPPORT_ATHUB_MGCG | + AMD_CG_SUPPORT_ATHUB_LS; adev->pg_flags = AMD_PG_SUPPORT_VCN | AMD_PG_SUPPORT_VCN_DPG | AMD_PG_SUPPORT_JPEG; diff --git a/drivers/gpu/drm/amd/amdgpu/umc_v8_10.c b/drivers/gpu/drm/amd/amdgpu/umc_v8_10.c index da394bc06bba..fb55e8cb9967 100644 --- a/drivers/gpu/drm/amd/amdgpu/umc_v8_10.c +++ b/drivers/gpu/drm/amd/amdgpu/umc_v8_10.c @@ -209,6 +209,45 @@ static int umc_v8_10_swizzle_mode_na_to_pa(struct amdgpu_device *adev, return 0; } +static void umc_v8_10_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, uint64_t mc_umc_status) +{ + uint64_t na_err_addr_base; + uint64_t na_err_addr, retired_page_addr; + uint32_t channel_index, addr_lsb, col = 0; + int ret = 0; + + 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]; + + /* the lowest lsb bits should be ignored */ + addr_lsb = REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, AddrLsb); + err_addr &= ~((0x1ULL << addr_lsb) - 1); + na_err_addr_base = err_addr & ~(0x3ULL << UMC_V8_10_NA_C5_BIT); + + /* loop for all possibilities of [C6 C5] in normal address. */ + for (col = 0; col < UMC_V8_10_NA_COL_2BITS_POWER_OF_2_NUM; col++) { + na_err_addr = na_err_addr_base | (col << UMC_V8_10_NA_C5_BIT); + + /* Mapping normal error address to retired soc physical address. */ + ret = umc_v8_10_swizzle_mode_na_to_pa(adev, channel_index, + na_err_addr, &retired_page_addr); + if (ret) { + dev_err(adev->dev, "Failed to map pa from umc na.\n"); + break; + } + dev_info(adev->dev, "Error Address(PA): 0x%llx\n", + retired_page_addr); + amdgpu_umc_fill_error_record(err_data, na_err_addr, + retired_page_addr, channel_index, umc_inst); + } +} + static void umc_v8_10_query_error_address(struct amdgpu_device *adev, struct ras_err_data *err_data, uint32_t umc_reg_offset, @@ -218,10 +257,7 @@ static void umc_v8_10_query_error_address(struct amdgpu_device *adev, { uint64_t mc_umc_status_addr; uint64_t mc_umc_status, err_addr; - uint64_t mc_umc_addrt0, na_err_addr_base; - uint64_t na_err_addr, retired_page_addr; - uint32_t channel_index, addr_lsb, col = 0; - int ret = 0; + uint64_t mc_umc_addrt0; mc_umc_status_addr = SOC15_REG_OFFSET(UMC, 0, regMCA_UMC_UMC0_MCUMC_STATUST0); @@ -236,12 +272,6 @@ static void umc_v8_10_query_error_address(struct amdgpu_device *adev, return; } - 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]; - /* calculate error address if ue error is detected */ if (REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, Val) == 1 && REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, AddrV) == 1 && @@ -251,27 +281,8 @@ static void umc_v8_10_query_error_address(struct amdgpu_device *adev, err_addr = RREG64_PCIE((mc_umc_addrt0 + umc_reg_offset) * 4); err_addr = REG_GET_FIELD(err_addr, MCA_UMC_UMC0_MCUMC_ADDRT0, ErrorAddr); - /* the lowest lsb bits should be ignored */ - addr_lsb = REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, AddrLsb); - err_addr &= ~((0x1ULL << addr_lsb) - 1); - na_err_addr_base = err_addr & ~(0x3ULL << UMC_V8_10_NA_C5_BIT); - - /* loop for all possibilities of [C6 C5] in normal address. */ - for (col = 0; col < UMC_V8_10_NA_COL_2BITS_POWER_OF_2_NUM; col++) { - na_err_addr = na_err_addr_base | (col << UMC_V8_10_NA_C5_BIT); - - /* Mapping normal error address to retired soc physical address. */ - ret = umc_v8_10_swizzle_mode_na_to_pa(adev, channel_index, - na_err_addr, &retired_page_addr); - if (ret) { - dev_err(adev->dev, "Failed to map pa from umc na.\n"); - break; - } - dev_info(adev->dev, "Error Address(PA): 0x%llx\n", - retired_page_addr); - amdgpu_umc_fill_error_record(err_data, na_err_addr, - retired_page_addr, channel_index, umc_inst); - } + umc_v8_10_convert_error_address(adev, err_data, err_addr, + ch_inst, umc_inst, node_inst, mc_umc_status); } /* clear umc status */ @@ -349,6 +360,133 @@ static bool umc_v8_10_query_ras_poison_mode(struct amdgpu_device *adev) return true; } +static void umc_v8_10_ecc_info_query_correctable_error_count(struct amdgpu_device *adev, + uint32_t node_inst, uint32_t umc_inst, uint32_t ch_inst, + unsigned long *error_count) +{ + uint64_t mc_umc_status; + uint32_t eccinfo_table_idx; + struct amdgpu_ras *ras = amdgpu_ras_get_context(adev); + + eccinfo_table_idx = node_inst * adev->umc.umc_inst_num * + adev->umc.channel_inst_num + + umc_inst * adev->umc.channel_inst_num + + ch_inst; + + /* check the MCUMC_STATUS */ + mc_umc_status = ras->umc_ecc.ecc[eccinfo_table_idx].mca_umc_status; + if (REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, Val) == 1 && + REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, CECC) == 1) { + *error_count += 1; + } +} + +static void umc_v8_10_ecc_info_query_uncorrectable_error_count(struct amdgpu_device *adev, + uint32_t node_inst, uint32_t umc_inst, uint32_t ch_inst, + unsigned long *error_count) +{ + uint64_t mc_umc_status; + uint32_t eccinfo_table_idx; + struct amdgpu_ras *ras = amdgpu_ras_get_context(adev); + + eccinfo_table_idx = node_inst * adev->umc.umc_inst_num * + adev->umc.channel_inst_num + + umc_inst * adev->umc.channel_inst_num + + ch_inst; + + /* check the MCUMC_STATUS */ + mc_umc_status = ras->umc_ecc.ecc[eccinfo_table_idx].mca_umc_status; + if ((REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, Val) == 1) && + (REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, Deferred) == 1 || + REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, UECC) == 1 || + REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, PCC) == 1 || + REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, UC) == 1 || + REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, TCC) == 1)) { + *error_count += 1; + } +} + +static void umc_v8_10_ecc_info_query_ras_error_count(struct amdgpu_device *adev, + void *ras_error_status) +{ + struct ras_err_data *err_data = (struct ras_err_data *)ras_error_status; + + uint32_t node_inst = 0; + uint32_t umc_inst = 0; + uint32_t ch_inst = 0; + + /* TODO: driver needs to toggle DF Cstate to ensure + * safe access of UMC registers. Will add the protection + */ + LOOP_UMC_EACH_NODE_INST_AND_CH(node_inst, umc_inst, ch_inst) { + umc_v8_10_ecc_info_query_correctable_error_count(adev, + node_inst, umc_inst, ch_inst, + &(err_data->ce_count)); + umc_v8_10_ecc_info_query_uncorrectable_error_count(adev, + node_inst, umc_inst, ch_inst, + &(err_data->ue_count)); + } +} + +static void umc_v8_10_ecc_info_query_error_address(struct amdgpu_device *adev, + struct ras_err_data *err_data, + uint32_t ch_inst, + uint32_t umc_inst, + uint32_t node_inst) +{ + uint32_t eccinfo_table_idx; + uint64_t mc_umc_status, err_addr; + + struct amdgpu_ras *ras = amdgpu_ras_get_context(adev); + + eccinfo_table_idx = node_inst * adev->umc.umc_inst_num * + adev->umc.channel_inst_num + + umc_inst * adev->umc.channel_inst_num + + ch_inst; + + mc_umc_status = ras->umc_ecc.ecc[eccinfo_table_idx].mca_umc_status; + + if (mc_umc_status == 0) + return; + + if (!err_data->err_addr) + return; + + /* calculate error address if ue error is detected */ + if (REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, Val) == 1 && + REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, AddrV) == 1 && + (REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, UECC) == 1)) { + + err_addr = ras->umc_ecc.ecc[eccinfo_table_idx].mca_umc_addr; + err_addr = REG_GET_FIELD(err_addr, MCA_UMC_UMC0_MCUMC_ADDRT0, ErrorAddr); + + umc_v8_10_convert_error_address(adev, err_data, err_addr, + ch_inst, umc_inst, node_inst, mc_umc_status); + } +} + +static void umc_v8_10_ecc_info_query_ras_error_address(struct amdgpu_device *adev, + void *ras_error_status) +{ + struct ras_err_data *err_data = (struct ras_err_data *)ras_error_status; + + uint32_t node_inst = 0; + uint32_t umc_inst = 0; + uint32_t ch_inst = 0; + + /* TODO: driver needs to toggle DF Cstate to ensure + * safe access of UMC resgisters. Will add the protection + * when firmware interface is ready + */ + LOOP_UMC_EACH_NODE_INST_AND_CH(node_inst, umc_inst, ch_inst) { + umc_v8_10_ecc_info_query_error_address(adev, + err_data, + ch_inst, + umc_inst, + node_inst); + } +} + 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, @@ -360,4 +498,6 @@ struct amdgpu_umc_ras umc_v8_10_ras = { }, .err_cnt_init = umc_v8_10_err_cnt_init, .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, }; diff --git a/drivers/gpu/drm/amd/amdgpu/umc_v8_10.h b/drivers/gpu/drm/amd/amdgpu/umc_v8_10.h index 25eaf4af5fcf..c6dfd433fec7 100644 --- a/drivers/gpu/drm/amd/amdgpu/umc_v8_10.h +++ b/drivers/gpu/drm/amd/amdgpu/umc_v8_10.h @@ -31,9 +31,9 @@ /* number of umc instance with memory map register access */ #define UMC_V8_10_UMC_INSTANCE_NUM 2 -/* Total channel instances for all umc nodes */ +/* Total channel instances for all available umc nodes */ #define UMC_V8_10_TOTAL_CHANNEL_NUM(adev) \ - (UMC_V8_10_CHANNEL_INSTANCE_NUM * UMC_V8_10_UMC_INSTANCE_NUM * (adev)->umc.node_inst_num) + (UMC_V8_10_CHANNEL_INSTANCE_NUM * UMC_V8_10_UMC_INSTANCE_NUM * (adev)->gmc.num_umc) /* UMC regiser per channel offset */ #define UMC_V8_10_PER_CHANNEL_OFFSET 0x400 diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v4_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v4_0.c index efb22d0975b3..43d587404c3e 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v4_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v4_0.c @@ -78,9 +78,17 @@ static void vcn_v4_0_set_ras_funcs(struct amdgpu_device *adev); static int vcn_v4_0_early_init(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; + int i; - if (amdgpu_sriov_vf(adev)) + if (amdgpu_sriov_vf(adev)) { adev->vcn.harvest_config = VCN_HARVEST_MMSCH; + for (i = 0; i < adev->vcn.num_vcn_inst; ++i) { + if (amdgpu_vcn_is_disabled_vcn(adev, VCN_ENCODE_RING, i)) { + adev->vcn.harvest_config |= 1 << i; + dev_info(adev->dev, "VCN%d is disabled by hypervisor\n", i); + } + } + } /* re-use enc ring as unified ring */ adev->vcn.num_enc_rings = 1; @@ -238,16 +246,11 @@ static int vcn_v4_0_hw_init(void *handle) continue; ring = &adev->vcn.inst[i].ring_enc[0]; - if (amdgpu_vcn_is_disabled_vcn(adev, VCN_ENCODE_RING, i)) { - ring->sched.ready = false; - ring->no_scheduler = true; - dev_info(adev->dev, "ring %s is disabled by hypervisor\n", ring->name); - } else { - ring->wptr = 0; - ring->wptr_old = 0; - vcn_v4_0_unified_ring_set_wptr(ring); - ring->sched.ready = true; - } + ring->wptr = 0; + ring->wptr_old = 0; + vcn_v4_0_unified_ring_set_wptr(ring); + ring->sched.ready = true; + } } else { for (i = 0; i < adev->vcn.num_vcn_inst; ++i) { @@ -1710,7 +1713,7 @@ static int vcn_v4_0_dec_msg(struct amdgpu_cs_parser *p, struct amdgpu_job *job, create = ptr + addr + offset - start; - /* H246, HEVC and VP9 can run on any instance */ + /* H264, HEVC and VP9 can run on any instance */ if (create[0] == 0x7 || create[0] == 0x10 || create[0] == 0x11) continue; @@ -1724,7 +1727,29 @@ out: return r; } -#define RADEON_VCN_ENGINE_TYPE_DECODE (0x00000003) +#define RADEON_VCN_ENGINE_TYPE_ENCODE (0x00000002) +#define RADEON_VCN_ENGINE_TYPE_DECODE (0x00000003) + +#define RADEON_VCN_ENGINE_INFO (0x30000001) +#define RADEON_VCN_ENGINE_INFO_MAX_OFFSET 16 + +#define RENCODE_ENCODE_STANDARD_AV1 2 +#define RENCODE_IB_PARAM_SESSION_INIT 0x00000003 +#define RENCODE_IB_PARAM_SESSION_INIT_MAX_OFFSET 64 + +/* return the offset in ib if id is found, -1 otherwise + * to speed up the searching we only search upto max_offset + */ +static int vcn_v4_0_enc_find_ib_param(struct amdgpu_ib *ib, uint32_t id, int max_offset) +{ + int i; + + for (i = 0; i < ib->length_dw && i < max_offset && ib->ptr[i] >= 8; i += ib->ptr[i]/4) { + if (ib->ptr[i + 1] == id) + return i; + } + return -1; +} static int vcn_v4_0_ring_patch_cs_in_place(struct amdgpu_cs_parser *p, struct amdgpu_job *job, @@ -1734,27 +1759,35 @@ static int vcn_v4_0_ring_patch_cs_in_place(struct amdgpu_cs_parser *p, struct amdgpu_vcn_decode_buffer *decode_buffer; uint64_t addr; uint32_t val; + int idx; /* The first instance can decode anything */ if (!ring->me) return 0; - /* unified queue ib header has 8 double words. */ - if (ib->length_dw < 8) + /* RADEON_VCN_ENGINE_INFO is at the top of ib block */ + idx = vcn_v4_0_enc_find_ib_param(ib, RADEON_VCN_ENGINE_INFO, + RADEON_VCN_ENGINE_INFO_MAX_OFFSET); + if (idx < 0) /* engine info is missing */ return 0; - val = amdgpu_ib_get_value(ib, 6); //RADEON_VCN_ENGINE_TYPE - if (val != RADEON_VCN_ENGINE_TYPE_DECODE) - return 0; - - decode_buffer = (struct amdgpu_vcn_decode_buffer *)&ib->ptr[10]; - - if (!(decode_buffer->valid_buf_flag & 0x1)) - return 0; - - addr = ((u64)decode_buffer->msg_buffer_address_hi) << 32 | - decode_buffer->msg_buffer_address_lo; - return vcn_v4_0_dec_msg(p, job, addr); + val = amdgpu_ib_get_value(ib, idx + 2); /* RADEON_VCN_ENGINE_TYPE */ + if (val == RADEON_VCN_ENGINE_TYPE_DECODE) { + decode_buffer = (struct amdgpu_vcn_decode_buffer *)&ib->ptr[idx + 6]; + + if (!(decode_buffer->valid_buf_flag & 0x1)) + return 0; + + addr = ((u64)decode_buffer->msg_buffer_address_hi) << 32 | + decode_buffer->msg_buffer_address_lo; + return vcn_v4_0_dec_msg(p, job, addr); + } else if (val == RADEON_VCN_ENGINE_TYPE_ENCODE) { + idx = vcn_v4_0_enc_find_ib_param(ib, RENCODE_IB_PARAM_SESSION_INIT, + RENCODE_IB_PARAM_SESSION_INIT_MAX_OFFSET); + if (idx >= 0 && ib->ptr[idx + 2] == RENCODE_ENCODE_STANDARD_AV1) + return vcn_v4_0_limit_sched(p, job); + } + return 0; } static const struct amdgpu_ring_funcs vcn_v4_0_unified_ring_vm_funcs = { |