diff options
Diffstat (limited to 'drivers/gpu/drm/amd/amdgpu')
42 files changed, 649 insertions, 215 deletions
diff --git a/drivers/gpu/drm/amd/amdgpu/Makefile b/drivers/gpu/drm/amd/amdgpu/Makefile index 6cf0fe871d6c..c56320e78c0e 100644 --- a/drivers/gpu/drm/amd/amdgpu/Makefile +++ b/drivers/gpu/drm/amd/amdgpu/Makefile @@ -75,7 +75,7 @@ amdgpu-y += \ vega20_reg_init.o nbio_v7_4.o nbio_v2_3.o nv.o navi10_reg_init.o navi14_reg_init.o \ arct_reg_init.o navi12_reg_init.o mxgpu_nv.o sienna_cichlid_reg_init.o vangogh_reg_init.o \ nbio_v7_2.o dimgrey_cavefish_reg_init.o hdp_v4_0.o hdp_v5_0.o aldebaran_reg_init.o aldebaran.o \ - beige_goby_reg_init.o + beige_goby_reg_init.o yellow_carp_reg_init.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 267e9e147a6c..c0316eaba547 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -211,6 +211,7 @@ extern int amdgpu_discovery; extern int amdgpu_mes; extern int amdgpu_noretry; extern int amdgpu_force_asic_type; +extern int amdgpu_smartshift_bias; #ifdef CONFIG_HSA_AMD extern int sched_policy; extern bool debug_evictions; @@ -268,6 +269,10 @@ extern int amdgpu_num_kcq; #define CIK_CURSOR_WIDTH 128 #define CIK_CURSOR_HEIGHT 128 +/* smasrt shift bias level limits */ +#define AMDGPU_SMARTSHIFT_MAX_BIAS (100) +#define AMDGPU_SMARTSHIFT_MIN_BIAS (-100) + struct amdgpu_device; struct amdgpu_ib; struct amdgpu_cs_parser; @@ -1280,6 +1285,11 @@ bool amdgpu_device_is_peer_accessible(struct amdgpu_device *adev, int amdgpu_device_baco_enter(struct drm_device *dev); int amdgpu_device_baco_exit(struct drm_device *dev); +void amdgpu_device_flush_hdp(struct amdgpu_device *adev, + struct amdgpu_ring *ring); +void amdgpu_device_invalidate_hdp(struct amdgpu_device *adev, + struct amdgpu_ring *ring); + /* atpx handler */ #if defined(CONFIG_VGA_SWITCHEROO) void amdgpu_register_atpx_handler(void); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h index cf62f43a03da..db16b3e83694 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h @@ -269,7 +269,7 @@ int amdgpu_amdkfd_gpuvm_free_memory_of_gpu( struct kgd_dev *kgd, struct kgd_mem *mem, void *drm_priv, uint64_t *size); int amdgpu_amdkfd_gpuvm_map_memory_to_gpu( - struct kgd_dev *kgd, struct kgd_mem *mem, void *drm_priv); + struct kgd_dev *kgd, struct kgd_mem *mem, void *drm_priv, bool *table_freed); int amdgpu_amdkfd_gpuvm_unmap_memory_from_gpu( struct kgd_dev *kgd, struct kgd_mem *mem, void *drm_priv); int amdgpu_amdkfd_gpuvm_sync_memory( diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v10_3.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v10_3.c index d39cff4a1fe3..1f5620cc3570 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v10_3.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v10_3.c @@ -95,8 +95,8 @@ static void program_sh_mem_settings_v10_3(struct kgd_dev *kgd, uint32_t vmid, lock_srbm(kgd, 0, 0, 0, vmid); - WREG32(SOC15_REG_OFFSET(GC, 0, mmSH_MEM_CONFIG), sh_mem_config); - WREG32(SOC15_REG_OFFSET(GC, 0, mmSH_MEM_BASES), sh_mem_bases); + WREG32_SOC15(GC, 0, mmSH_MEM_CONFIG, sh_mem_config); + WREG32_SOC15(GC, 0, mmSH_MEM_BASES, sh_mem_bases); /* APE1 no longer exists on GFX9 */ unlock_srbm(kgd); @@ -129,7 +129,7 @@ static int init_interrupts_v10_3(struct kgd_dev *kgd, uint32_t pipe_id) lock_srbm(kgd, mec, pipe, 0, 0); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCPC_INT_CNTL), + WREG32_SOC15(GC, 0, mmCPC_INT_CNTL, CP_INT_CNTL_RING0__TIME_STAMP_INT_ENABLE_MASK | CP_INT_CNTL_RING0__OPCODE_ERROR_INT_ENABLE_MASK); @@ -212,10 +212,10 @@ static int hqd_load_v10_3(struct kgd_dev *kgd, void *mqd, uint32_t pipe_id, pr_debug("kfd: set HIQ, mec:%d, pipe:%d, queue:%d.\n", mec, pipe, queue_id); - value = RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CP_SCHEDULERS)); + value = RREG32_SOC15(GC, 0, mmRLC_CP_SCHEDULERS); value = REG_SET_FIELD(value, RLC_CP_SCHEDULERS, scheduler1, ((mec << 5) | (pipe << 3) | queue_id | 0x80)); - WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CP_SCHEDULERS), value); + WREG32_SOC15(GC, 0, mmRLC_CP_SCHEDULERS, value); } /* HQD registers extend from CP_MQD_BASE_ADDR to CP_HQD_EOP_WPTR_MEM. */ @@ -224,13 +224,13 @@ static int hqd_load_v10_3(struct kgd_dev *kgd, void *mqd, uint32_t pipe_id, for (reg = hqd_base; reg <= SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_WPTR_HI); reg++) - WREG32(reg, mqd_hqd[reg - hqd_base]); + WREG32_SOC15_IP(GC, reg, mqd_hqd[reg - hqd_base]); /* Activate doorbell logic before triggering WPTR poll. */ data = REG_SET_FIELD(m->cp_hqd_pq_doorbell_control, CP_HQD_PQ_DOORBELL_CONTROL, DOORBELL_EN, 1); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_DOORBELL_CONTROL), data); + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_DOORBELL_CONTROL, data); if (wptr) { /* Don't read wptr with get_user because the user @@ -259,17 +259,17 @@ static int hqd_load_v10_3(struct kgd_dev *kgd, void *mqd, uint32_t pipe_id, guessed_wptr += m->cp_hqd_pq_wptr_lo & ~(queue_size - 1); guessed_wptr += (uint64_t)m->cp_hqd_pq_wptr_hi << 32; - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_WPTR_LO), + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_WPTR_LO, lower_32_bits(guessed_wptr)); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_WPTR_HI), + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_WPTR_HI, upper_32_bits(guessed_wptr)); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_WPTR_POLL_ADDR), + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_WPTR_POLL_ADDR, lower_32_bits((uint64_t)wptr)); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_WPTR_POLL_ADDR_HI), + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_WPTR_POLL_ADDR_HI, upper_32_bits((uint64_t)wptr)); pr_debug("%s setting CP_PQ_WPTR_POLL_CNTL1 to %x\n", __func__, (uint32_t)get_queue_mask(adev, pipe_id, queue_id)); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_PQ_WPTR_POLL_CNTL1), + WREG32_SOC15(GC, 0, mmCP_PQ_WPTR_POLL_CNTL1, (uint32_t)get_queue_mask(adev, pipe_id, queue_id)); } @@ -279,7 +279,7 @@ static int hqd_load_v10_3(struct kgd_dev *kgd, void *mqd, uint32_t pipe_id, CP_HQD_EOP_RPTR, INIT_FETCHER, 1)); data = REG_SET_FIELD(m->cp_hqd_active, CP_HQD_ACTIVE, ACTIVE, 1); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_ACTIVE), data); + WREG32_SOC15(GC, 0, mmCP_HQD_ACTIVE, data); release_queue(kgd); @@ -350,7 +350,7 @@ static int hqd_dump_v10_3(struct kgd_dev *kgd, if (WARN_ON_ONCE(i >= HQD_N_REGS)) \ break; \ (*dump)[i][0] = (addr) << 2; \ - (*dump)[i++][1] = RREG32(addr); \ + (*dump)[i++][1] = RREG32_SOC15_IP(GC, addr); \ } while (0) *dump = kmalloc(HQD_N_REGS*2*sizeof(uint32_t), GFP_KERNEL); @@ -482,13 +482,13 @@ static bool hqd_is_occupied_v10_3(struct kgd_dev *kgd, uint64_t queue_address, uint32_t low, high; acquire_queue(kgd, pipe_id, queue_id); - act = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_ACTIVE)); + act = RREG32_SOC15(GC, 0, mmCP_HQD_ACTIVE); if (act) { low = lower_32_bits(queue_address >> 8); high = upper_32_bits(queue_address >> 8); - if (low == RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_BASE)) && - high == RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_BASE_HI))) + if (low == RREG32_SOC15(GC, 0, mmCP_HQD_PQ_BASE) && + high == RREG32_SOC15(GC, 0, mmCP_HQD_PQ_BASE_HI)) retval = true; } release_queue(kgd); @@ -542,11 +542,11 @@ static int hqd_destroy_v10_3(struct kgd_dev *kgd, void *mqd, break; } - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_DEQUEUE_REQUEST), type); + WREG32_SOC15(GC, 0, mmCP_HQD_DEQUEUE_REQUEST, type); end_jiffies = (utimeout * HZ / 1000) + jiffies; while (true) { - temp = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_ACTIVE)); + temp = RREG32_SOC15(GC, 0, mmCP_HQD_ACTIVE); if (!(temp & CP_HQD_ACTIVE__ACTIVE_MASK)) break; if (time_after(jiffies, end_jiffies)) { @@ -626,7 +626,7 @@ static int wave_control_execute_v10_3(struct kgd_dev *kgd, mutex_lock(&adev->grbm_idx_mutex); - WREG32(SOC15_REG_OFFSET(GC, 0, mmGRBM_GFX_INDEX), gfx_index_val); + WREG32_SOC15(GC, 0, mmGRBM_GFX_INDEX, gfx_index_val); WREG32(SOC15_REG_OFFSET(GC, 0, mmSQ_CMD), sq_cmd); data = REG_SET_FIELD(data, GRBM_GFX_INDEX, @@ -636,7 +636,7 @@ static int wave_control_execute_v10_3(struct kgd_dev *kgd, data = REG_SET_FIELD(data, GRBM_GFX_INDEX, SE_BROADCAST_WRITES, 1); - WREG32(SOC15_REG_OFFSET(GC, 0, mmGRBM_GFX_INDEX), data); + WREG32_SOC15(GC, 0, mmGRBM_GFX_INDEX, data); mutex_unlock(&adev->grbm_idx_mutex); return 0; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c index 668a28b80a62..fdee98fae2ba 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c @@ -278,7 +278,7 @@ static int amdgpu_amdkfd_remove_eviction_fence(struct amdgpu_bo *bo, write_seqcount_end(&resv->seq); /* Drop the references to the removed fences or move them to ef_list */ - for (i = j, k = 0; i < old->shared_count; ++i) { + for (i = j; i < old->shared_count; ++i) { struct dma_fence *f; f = rcu_dereference_protected(new->shared[i], @@ -1070,7 +1070,8 @@ static void unmap_bo_from_gpuvm(struct kgd_mem *mem, static int update_gpuvm_pte(struct kgd_mem *mem, struct kfd_mem_attachment *entry, - struct amdgpu_sync *sync) + struct amdgpu_sync *sync, + bool *table_freed) { struct amdgpu_bo_va *bo_va = entry->bo_va; struct amdgpu_device *adev = entry->adev; @@ -1081,7 +1082,7 @@ static int update_gpuvm_pte(struct kgd_mem *mem, return ret; /* Update the page tables */ - ret = amdgpu_vm_bo_update(adev, bo_va, false); + ret = amdgpu_vm_bo_update(adev, bo_va, false, table_freed); if (ret) { pr_err("amdgpu_vm_bo_update failed\n"); return ret; @@ -1093,7 +1094,8 @@ static int update_gpuvm_pte(struct kgd_mem *mem, static int map_bo_to_gpuvm(struct kgd_mem *mem, struct kfd_mem_attachment *entry, struct amdgpu_sync *sync, - bool no_update_pte) + bool no_update_pte, + bool *table_freed) { int ret; @@ -1110,7 +1112,7 @@ static int map_bo_to_gpuvm(struct kgd_mem *mem, if (no_update_pte) return 0; - ret = update_gpuvm_pte(mem, entry, sync); + ret = update_gpuvm_pte(mem, entry, sync, table_freed); if (ret) { pr_err("update_gpuvm_pte() failed\n"); goto update_gpuvm_pte_failed; @@ -1608,7 +1610,8 @@ int amdgpu_amdkfd_gpuvm_free_memory_of_gpu( } int amdgpu_amdkfd_gpuvm_map_memory_to_gpu( - struct kgd_dev *kgd, struct kgd_mem *mem, void *drm_priv) + struct kgd_dev *kgd, struct kgd_mem *mem, + void *drm_priv, bool *table_freed) { struct amdgpu_device *adev = get_amdgpu_device(kgd); struct amdgpu_vm *avm = drm_priv_to_vm(drm_priv); @@ -1696,7 +1699,7 @@ int amdgpu_amdkfd_gpuvm_map_memory_to_gpu( entry->va, entry->va + bo_size, entry); ret = map_bo_to_gpuvm(mem, entry, ctx.sync, - is_invalid_userptr); + is_invalid_userptr, table_freed); if (ret) { pr_err("Failed to map bo to gpuvm\n"); goto out_unreserve; @@ -2146,7 +2149,7 @@ static int validate_invalid_user_pages(struct amdkfd_process_info *process_info) continue; kfd_mem_dmaunmap_attachment(mem, attachment); - ret = update_gpuvm_pte(mem, attachment, &sync); + ret = update_gpuvm_pte(mem, attachment, &sync, NULL); if (ret) { pr_err("%s: update PTE failed\n", __func__); /* make sure this gets validated again */ @@ -2352,7 +2355,7 @@ int amdgpu_amdkfd_gpuvm_restore_process_bos(void *info, struct dma_fence **ef) continue; kfd_mem_dmaunmap_attachment(mem, attachment); - ret = update_gpuvm_pte(mem, attachment, &sync_obj); + ret = update_gpuvm_pte(mem, attachment, &sync_obj, NULL); if (ret) { pr_debug("Memory eviction: update PTE failed. Try again\n"); goto validate_map_fail; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c index f6a8f0c5a52f..1476236f5c7c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c @@ -662,7 +662,7 @@ static int amdgpu_cs_sync_rings(struct amdgpu_cs_parser *p) * @error: error number * @backoff: indicator to backoff the reservation * - * If error is set than unvalidate buffer, otherwise just free memory + * If error is set then unvalidate buffer, otherwise just free memory * used by parsing context. **/ static void amdgpu_cs_parser_fini(struct amdgpu_cs_parser *parser, int error, @@ -781,7 +781,7 @@ static int amdgpu_cs_vm_handling(struct amdgpu_cs_parser *p) if (r) return r; - r = amdgpu_vm_bo_update(adev, fpriv->prt_va, false); + r = amdgpu_vm_bo_update(adev, fpriv->prt_va, false, NULL); if (r) return r; @@ -792,7 +792,7 @@ static int amdgpu_cs_vm_handling(struct amdgpu_cs_parser *p) if (amdgpu_mcbp || amdgpu_sriov_vf(adev)) { bo_va = fpriv->csa_va; BUG_ON(!bo_va); - r = amdgpu_vm_bo_update(adev, bo_va, false); + r = amdgpu_vm_bo_update(adev, bo_va, false, NULL); if (r) return r; @@ -811,7 +811,7 @@ static int amdgpu_cs_vm_handling(struct amdgpu_cs_parser *p) if (bo_va == NULL) continue; - r = amdgpu_vm_bo_update(adev, bo_va, false); + r = amdgpu_vm_bo_update(adev, bo_va, false, NULL); if (r) return r; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index 736b8c121979..6a242ec3f7ef 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -84,6 +84,7 @@ MODULE_FIRMWARE("amdgpu/navi10_gpu_info.bin"); MODULE_FIRMWARE("amdgpu/navi14_gpu_info.bin"); MODULE_FIRMWARE("amdgpu/navi12_gpu_info.bin"); MODULE_FIRMWARE("amdgpu/vangogh_gpu_info.bin"); +MODULE_FIRMWARE("amdgpu/yellow_carp_gpu_info.bin"); #define AMDGPU_RESUME_MS 2000 @@ -122,6 +123,7 @@ const char *amdgpu_asic_name[] = { "VANGOGH", "DIMGREY_CAVEFISH", "BEIGE_GOBY", + "YELLOW_CARP", "LAST", }; @@ -313,9 +315,9 @@ void amdgpu_device_vram_access(struct amdgpu_device *adev, loff_t pos, if (write) { memcpy_toio(addr, buf, count); mb(); - amdgpu_asic_flush_hdp(adev, NULL); + amdgpu_device_flush_hdp(adev, NULL); } else { - amdgpu_asic_invalidate_hdp(adev, NULL); + amdgpu_device_invalidate_hdp(adev, NULL); mb(); memcpy_fromio(buf, addr, count); } @@ -1884,6 +1886,9 @@ static int amdgpu_device_parse_gpu_info_fw(struct amdgpu_device *adev) case CHIP_VANGOGH: chip_name = "vangogh"; break; + case CHIP_YELLOW_CARP: + chip_name = "yellow_carp"; + break; } snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_gpu_info.bin", chip_name); @@ -2062,8 +2067,11 @@ static int amdgpu_device_ip_early_init(struct amdgpu_device *adev) case CHIP_DIMGREY_CAVEFISH: case CHIP_BEIGE_GOBY: case CHIP_VANGOGH: + case CHIP_YELLOW_CARP: if (adev->asic_type == CHIP_VANGOGH) adev->family = AMDGPU_FAMILY_VGH; + else if (adev->asic_type == CHIP_YELLOW_CARP) + adev->family = AMDGPU_FAMILY_YC; else adev->family = AMDGPU_FAMILY_NV; @@ -3146,6 +3154,7 @@ bool amdgpu_device_asic_has_dc_support(enum amd_asic_type asic_type) case CHIP_DIMGREY_CAVEFISH: case CHIP_BEIGE_GOBY: case CHIP_VANGOGH: + case CHIP_YELLOW_CARP: #endif return amdgpu_dc != 0; #endif @@ -5475,4 +5484,31 @@ bool amdgpu_device_load_pci_state(struct pci_dev *pdev) return true; } +void amdgpu_device_flush_hdp(struct amdgpu_device *adev, + struct amdgpu_ring *ring) +{ +#ifdef CONFIG_X86_64 + if (adev->flags & AMD_IS_APU) + return; +#endif + if (adev->gmc.xgmi.connected_to_cpu) + return; + + if (ring && ring->funcs->emit_hdp_flush) + amdgpu_ring_emit_hdp_flush(ring); + else + amdgpu_asic_flush_hdp(adev, ring); +} + +void amdgpu_device_invalidate_hdp(struct amdgpu_device *adev, + struct amdgpu_ring *ring) +{ +#ifdef CONFIG_X86_64 + if (adev->flags & AMD_IS_APU) + return; +#endif + if (adev->gmc.xgmi.connected_to_cpu) + return; + amdgpu_asic_invalidate_hdp(adev, ring); +} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c index ac7b37dfff5e..2b6a66c849d4 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c @@ -1056,7 +1056,7 @@ int amdgpu_display_gem_fb_init(struct drm_device *dev, return 0; err: - drm_err(dev, "Failed to init gem fb: %d\n", ret); + drm_dbg_kms(dev, "Failed to init gem fb: %d\n", ret); rfb->base.obj[0] = NULL; return ret; } @@ -1090,7 +1090,7 @@ int amdgpu_display_gem_fb_verify_and_init( return 0; err: - drm_err(dev, "Failed to verify and init gem fb: %d\n", ret); + drm_dbg_kms(dev, "Failed to verify and init gem fb: %d\n", ret); rfb->base.obj[0] = NULL; return ret; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c index 809aa7641d7e..c080ba15ae77 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c @@ -173,6 +173,7 @@ int amdgpu_tmz = -1; /* auto */ uint amdgpu_freesync_vid_mode; int amdgpu_reset_method = -1; /* auto */ int amdgpu_num_kcq = -1; +int amdgpu_smartshift_bias; static void amdgpu_drv_delayed_reset_work_handler(struct work_struct *work); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fw_attestation.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fw_attestation.c index 8d1ad294cb02..2ca3c329de6d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fw_attestation.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fw_attestation.c @@ -121,6 +121,9 @@ static const struct file_operations amdgpu_fw_attestation_debugfs_ops = { static int amdgpu_is_fw_attestation_supported(struct amdgpu_device *adev) { + if (adev->flags & AMD_IS_APU) + return 0; + if (adev->asic_type >= CHIP_SIENNA_CICHLID) return 1; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c index 34243e122d11..1313784605b0 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c @@ -250,7 +250,7 @@ int amdgpu_gart_unbind(struct amdgpu_device *adev, uint64_t offset, } } mb(); - amdgpu_asic_flush_hdp(adev, NULL); + amdgpu_device_flush_hdp(adev, NULL); for (i = 0; i < adev->num_vmhubs; i++) amdgpu_gmc_flush_gpu_tlb(adev, 0, i, 0); @@ -337,7 +337,7 @@ void amdgpu_gart_invalidate_tlb(struct amdgpu_device *adev) int i; mb(); - amdgpu_asic_flush_hdp(adev, NULL); + amdgpu_device_flush_hdp(adev, NULL); for (i = 0; i < adev->num_vmhubs; i++) amdgpu_gmc_flush_gpu_tlb(adev, 0, i, 0); } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c index 1c3e3b608332..9cf4beaf646c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c @@ -612,7 +612,7 @@ static void amdgpu_gem_va_update_vm(struct amdgpu_device *adev, if (operation == AMDGPU_VA_OP_MAP || operation == AMDGPU_VA_OP_REPLACE) { - r = amdgpu_vm_bo_update(adev, bo_va, false); + r = amdgpu_vm_bo_update(adev, bo_va, false, NULL); if (r) goto error; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfxhub.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfxhub.h index 66ebc2e3b2ad..beabab515836 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfxhub.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfxhub.h @@ -34,6 +34,7 @@ struct amdgpu_gfxhub_funcs { void (*set_fault_enable_default)(struct amdgpu_device *adev, bool value); void (*init)(struct amdgpu_device *adev); int (*get_xgmi_info)(struct amdgpu_device *adev); + void (*utcl2_harvest)(struct amdgpu_device *adev); }; struct amdgpu_gfxhub { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c index 85b8bfb8857b..0174f7817ce2 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c @@ -785,3 +785,22 @@ uint64_t amdgpu_gmc_vram_cpu_pa(struct amdgpu_device *adev, struct amdgpu_bo *bo { return amdgpu_bo_gpu_offset(bo) - adev->gmc.vram_start + adev->gmc.aper_base; } + +void amdgpu_gmc_get_reserved_allocation(struct amdgpu_device *adev) +{ + /* Some ASICs need to reserve a region of video memory to avoid access + * from driver */ + adev->mman.stolen_reserved_offset = 0; + adev->mman.stolen_reserved_size = 0; + + switch (adev->asic_type) { + case CHIP_YELLOW_CARP: + if (amdgpu_discovery == 0) { + adev->mman.stolen_reserved_offset = 0x1ffb0000; + adev->mman.stolen_reserved_size = 64 * PAGE_SIZE; + } + break; + default: + break; + } +} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h index 6aa1d52d3aee..e55201134a01 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h @@ -332,6 +332,7 @@ amdgpu_gmc_set_vm_fault_masks(struct amdgpu_device *adev, int hub_type, bool enable); void amdgpu_gmc_get_vbios_allocations(struct amdgpu_device *adev); +void amdgpu_gmc_get_reserved_allocation(struct amdgpu_device *adev); void amdgpu_gmc_init_pdb0(struct amdgpu_device *adev); uint64_t amdgpu_gmc_vram_mc2pa(struct amdgpu_device *adev, uint64_t mc_addr); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c index 77baf9b48d67..ec65ab0ddf89 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c @@ -214,15 +214,7 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs, if (job && ring->funcs->init_cond_exec) patch_offset = amdgpu_ring_init_cond_exec(ring); -#ifdef CONFIG_X86_64 - if (!(adev->flags & AMD_IS_APU)) -#endif - { - if (ring->funcs->emit_hdp_flush) - amdgpu_ring_emit_hdp_flush(ring); - else - amdgpu_asic_flush_hdp(adev, ring); - } + amdgpu_device_flush_hdp(adev, ring); if (need_ctx_switch) status |= AMDGPU_HAVE_CTX_SWITCH; @@ -259,10 +251,7 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs, if (job && ring->funcs->emit_frame_cntl) amdgpu_ring_emit_frame_cntl(ring, false, secure); -#ifdef CONFIG_X86_64 - if (!(adev->flags & AMD_IS_APU)) -#endif - amdgpu_asic_invalidate_hdp(adev, ring); + amdgpu_device_invalidate_hdp(adev, ring); if (ib->flags & AMDGPU_IB_FLAG_TC_WB_NOT_INVALIDATE) fence_flags |= AMDGPU_FENCE_FLAG_TC_WB_ONLY; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c index c83f116edde2..b7a2070d90af 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c @@ -76,7 +76,7 @@ static void amdgpu_bo_destroy(struct ttm_buffer_object *tbo) kfree(ubo->metadata); } - kfree(bo); + kvfree(bo); } /** @@ -541,7 +541,7 @@ int amdgpu_bo_create(struct amdgpu_device *adev, BUG_ON(bp->bo_ptr_size < sizeof(struct amdgpu_bo)); *bo_ptr = NULL; - bo = kzalloc(bp->bo_ptr_size, GFP_KERNEL); + bo = kvzalloc(bp->bo_ptr_size, GFP_KERNEL); if (bo == NULL) return -ENOMEM; drm_gem_private_object_init(adev_to_drm(adev), &bo->tbo.base, size); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c index 3ff76cbaec8d..6046123d0562 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c @@ -118,6 +118,10 @@ static int psp_early_init(void *handle) case CHIP_ALDEBARAN: psp_v13_0_set_psp_funcs(psp); break; + case CHIP_YELLOW_CARP: + psp_v13_0_set_psp_funcs(psp); + psp->autoload_supported = true; + break; default: return -EINVAL; } @@ -179,6 +183,12 @@ static int psp_sw_init(void *handle) DRM_ERROR("Failed to load psp firmware!\n"); return ret; } + } else if (amdgpu_sriov_vf(adev) && adev->asic_type == CHIP_ALDEBARAN) { + ret = psp_init_ta_microcode(psp, "aldebaran"); + if (ret) { + DRM_ERROR("Failed to initialize ta microcode!\n"); + return ret; + } } ret = psp_memory_training_init(psp); @@ -282,7 +292,7 @@ psp_cmd_submit_buf(struct psp_context *psp, goto exit; } - amdgpu_asic_invalidate_hdp(psp->adev, NULL); + amdgpu_device_invalidate_hdp(psp->adev, NULL); while (*((unsigned int *)psp->fence_buf) != index) { if (--timeout == 0) break; @@ -295,7 +305,7 @@ psp_cmd_submit_buf(struct psp_context *psp, if (ras_intr) break; usleep_range(10, 100); - amdgpu_asic_invalidate_hdp(psp->adev, NULL); + amdgpu_device_invalidate_hdp(psp->adev, NULL); } /* We allow TEE_ERROR_NOT_SUPPORTED for VMR command and PSP_ERR_UNKNOWN_COMMAND in SRIOV */ @@ -684,6 +694,8 @@ int psp_reg_program(struct psp_context *psp, enum psp_reg_prog_id reg, psp_prep_reg_prog_cmd_buf(cmd, reg, value); ret = psp_cmd_submit_buf(psp, NULL, cmd, psp->fence_buf_mc_addr); + if (ret) + DRM_ERROR("PSP failed to program reg id %d", reg); kfree(cmd); return ret; @@ -2319,11 +2331,20 @@ static int psp_load_fw(struct amdgpu_device *adev) if (!psp->cmd) return -ENOMEM; - ret = amdgpu_bo_create_kernel(adev, PSP_1_MEG, PSP_1_MEG, - AMDGPU_GEM_DOMAIN_GTT, - &psp->fw_pri_bo, - &psp->fw_pri_mc_addr, - &psp->fw_pri_buf); + if (amdgpu_sriov_vf(adev)) { + ret = amdgpu_bo_create_kernel(adev, PSP_1_MEG, PSP_1_MEG, + AMDGPU_GEM_DOMAIN_VRAM, + &psp->fw_pri_bo, + &psp->fw_pri_mc_addr, + &psp->fw_pri_buf); + } else { + ret = amdgpu_bo_create_kernel(adev, PSP_1_MEG, PSP_1_MEG, + AMDGPU_GEM_DOMAIN_GTT, + &psp->fw_pri_bo, + &psp->fw_pri_mc_addr, + &psp->fw_pri_buf); + } + if (ret) goto failed; @@ -2696,7 +2717,7 @@ int psp_ring_cmd_submit(struct psp_context *psp, write_frame->fence_addr_hi = upper_32_bits(fence_mc_addr); write_frame->fence_addr_lo = lower_32_bits(fence_mc_addr); write_frame->fence_value = index; - amdgpu_asic_flush_hdp(adev, NULL); + amdgpu_device_flush_hdp(adev, NULL); /* Update the write Pointer in DWORDs */ psp_write_ptr_reg = (psp_write_ptr_reg + rb_frame_size_dw) % ring_size_dw; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c index 5d0c57b9917f..80dff29f2bc7 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c @@ -1725,6 +1725,13 @@ int amdgpu_ttm_init(struct amdgpu_device *adev) NULL); if (r) return r; + r = amdgpu_bo_create_kernel_at(adev, adev->mman.stolen_reserved_offset, + adev->mman.stolen_reserved_size, + AMDGPU_GEM_DOMAIN_VRAM, + &adev->mman.stolen_reserved_memory, + NULL); + if (r) + return r; DRM_INFO("amdgpu: %uM of VRAM memory ready\n", (unsigned) (adev->gmc.real_vram_size / (1024 * 1024))); @@ -1794,6 +1801,9 @@ void amdgpu_ttm_fini(struct amdgpu_device *adev) amdgpu_bo_free_kernel(&adev->mman.stolen_extended_memory, NULL, NULL); /* return the IP Discovery TMR memory back to VRAM */ amdgpu_bo_free_kernel(&adev->mman.discovery_memory, NULL, NULL); + if (adev->mman.stolen_reserved_size) + amdgpu_bo_free_kernel(&adev->mman.stolen_reserved_memory, + NULL, NULL); amdgpu_ttm_fw_reserve_vram_fini(adev); amdgpu_vram_mgr_fini(adev); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h index 74a7021592c3..e69f3e8e06e5 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h @@ -84,6 +84,10 @@ struct amdgpu_mman { struct amdgpu_bo *stolen_extended_memory; bool keep_stolen_vga_memory; + struct amdgpu_bo *stolen_reserved_memory; + uint64_t stolen_reserved_offset; + uint64_t stolen_reserved_size; + /* discovery */ uint8_t *discovery_bin; uint32_t discovery_tmr_size; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c index 8b502ffd58dd..5eb84de588eb 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c @@ -404,6 +404,7 @@ amdgpu_ucode_get_load_type(struct amdgpu_device *adev, int load_type) case CHIP_DIMGREY_CAVEFISH: case CHIP_ALDEBARAN: case CHIP_BEIGE_GOBY: + case CHIP_YELLOW_CARP: if (!load_type) return AMDGPU_FW_LOAD_DIRECT; else diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c index b812938c9322..647d2c31e8bd 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c @@ -50,6 +50,7 @@ #define FIRMWARE_DIMGREY_CAVEFISH "amdgpu/dimgrey_cavefish_vcn.bin" #define FIRMWARE_ALDEBARAN "amdgpu/aldebaran_vcn.bin" #define FIRMWARE_BEIGE_GOBY "amdgpu/beige_goby_vcn.bin" +#define FIRMWARE_YELLOW_CARP "amdgpu/yellow_carp_vcn.bin" MODULE_FIRMWARE(FIRMWARE_RAVEN); MODULE_FIRMWARE(FIRMWARE_PICASSO); @@ -66,6 +67,7 @@ MODULE_FIRMWARE(FIRMWARE_NAVY_FLOUNDER); MODULE_FIRMWARE(FIRMWARE_VANGOGH); MODULE_FIRMWARE(FIRMWARE_DIMGREY_CAVEFISH); MODULE_FIRMWARE(FIRMWARE_BEIGE_GOBY); +MODULE_FIRMWARE(FIRMWARE_YELLOW_CARP); static void amdgpu_vcn_idle_work_handler(struct work_struct *work); @@ -160,6 +162,12 @@ int amdgpu_vcn_sw_init(struct amdgpu_device *adev) (adev->pg_flags & AMD_PG_SUPPORT_VCN_DPG)) adev->vcn.indirect_sram = true; break; + case CHIP_YELLOW_CARP: + fw_name = FIRMWARE_YELLOW_CARP; + if ((adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) && + (adev->pg_flags & AMD_PG_SUPPORT_VCN_DPG)) + adev->vcn.indirect_sram = true; + break; default: return -EINVAL; } @@ -385,7 +393,6 @@ static void amdgpu_vcn_idle_work_handler(struct work_struct *work) } if (!fences && !atomic_read(&adev->vcn.total_submission_cnt)) { - amdgpu_gfx_off_ctrl(adev, true); amdgpu_device_ip_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_VCN, AMD_PG_STATE_GATE); r = amdgpu_dpm_switch_power_profile(adev, PP_SMC_POWER_PROFILE_VIDEO, @@ -405,7 +412,6 @@ void amdgpu_vcn_ring_begin_use(struct amdgpu_ring *ring) atomic_inc(&adev->vcn.total_submission_cnt); if (!cancel_delayed_work_sync(&adev->vcn.idle_work)) { - amdgpu_gfx_off_ctrl(adev, false); r = amdgpu_dpm_switch_power_profile(adev, PP_SMC_POWER_PROFILE_VIDEO, true); if (r) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index 777e8922ecf3..18246b5b6ee3 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -910,7 +910,7 @@ static int amdgpu_vm_pt_create(struct amdgpu_device *adev, return r; bo = &(*vmbo)->bo; - if (vm->is_compute_context && (adev->flags & AMD_IS_APU)) { + if (vm->is_compute_context || (adev->flags & AMD_IS_APU)) { (*vmbo)->shadow = NULL; return 0; } @@ -1713,8 +1713,8 @@ int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev, if (r) goto error_unlock; - amdgpu_res_first(res, offset, (last - start + 1) * AMDGPU_GPU_PAGE_SIZE, - &cursor); + amdgpu_res_first(pages_addr ? NULL : res, offset, + (last - start + 1) * AMDGPU_GPU_PAGE_SIZE, &cursor); while (cursor.remaining) { uint64_t tmp, num_entries, addr; @@ -1764,12 +1764,12 @@ int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev, amdgpu_res_next(&cursor, num_entries * AMDGPU_GPU_PAGE_SIZE); start = tmp; - }; + } r = vm->update_funcs->commit(¶ms, fence); if (table_freed) - *table_freed = params.table_freed; + *table_freed = *table_freed || params.table_freed; error_unlock: amdgpu_vm_eviction_unlock(vm); @@ -1827,6 +1827,7 @@ void amdgpu_vm_get_memory(struct amdgpu_vm *vm, uint64_t *vram_mem, * @adev: amdgpu_device pointer * @bo_va: requested BO and VM object * @clear: if true clear the entries + * @table_freed: return true if page table is freed * * Fill in the page table entries for @bo_va. * @@ -1834,7 +1835,7 @@ void amdgpu_vm_get_memory(struct amdgpu_vm *vm, uint64_t *vram_mem, * 0 for success, -EINVAL for failure. */ int amdgpu_vm_bo_update(struct amdgpu_device *adev, struct amdgpu_bo_va *bo_va, - bool clear) + bool clear, bool *table_freed) { struct amdgpu_bo *bo = bo_va->base.bo; struct amdgpu_vm *vm = bo_va->base.vm; @@ -1913,7 +1914,7 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev, struct amdgpu_bo_va *bo_va, resv, mapping->start, mapping->last, update_flags, mapping->offset, mem, - pages_addr, last_update, NULL); + pages_addr, last_update, table_freed); if (r) return r; } @@ -2165,7 +2166,7 @@ int amdgpu_vm_handle_moved(struct amdgpu_device *adev, list_for_each_entry_safe(bo_va, tmp, &vm->moved, base.vm_status) { /* Per VM BOs never need to bo cleared in the page tables */ - r = amdgpu_vm_bo_update(adev, bo_va, false); + r = amdgpu_vm_bo_update(adev, bo_va, false, NULL); if (r) return r; } @@ -2184,7 +2185,7 @@ int amdgpu_vm_handle_moved(struct amdgpu_device *adev, else clear = true; - r = amdgpu_vm_bo_update(adev, bo_va, clear); + r = amdgpu_vm_bo_update(adev, bo_va, clear, NULL); if (r) return r; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h index bee439dd673a..1f089da1e615 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h @@ -413,7 +413,7 @@ int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev, struct dma_fence **fence, bool *free_table); int amdgpu_vm_bo_update(struct amdgpu_device *adev, struct amdgpu_bo_va *bo_va, - bool clear); + bool clear, bool *table_freed); bool amdgpu_vm_evictable(struct amdgpu_bo *bo); void amdgpu_vm_bo_invalidate(struct amdgpu_device *adev, struct amdgpu_bo *bo, bool evicted); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_cpu.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_cpu.c index 03a44be50dd7..e3fbf0f10add 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_cpu.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_cpu.c @@ -110,7 +110,7 @@ static int amdgpu_vm_cpu_commit(struct amdgpu_vm_update_params *p, { /* Flush HDP */ mb(); - amdgpu_asic_flush_hdp(p->adev, NULL); + amdgpu_device_flush_hdp(p->adev, NULL); return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c index 102f31526204..2d56b60bc058 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c @@ -173,6 +173,9 @@ #define mmGC_THROTTLE_CTRL_Sienna_Cichlid 0x2030 #define mmGC_THROTTLE_CTRL_Sienna_Cichlid_BASE_IDX 0 +#define mmRLC_SPARE_INT_0_Sienna_Cichlid 0x4ca5 +#define mmRLC_SPARE_INT_0_Sienna_Cichlid_BASE_IDX 1 + #define GFX_RLCG_GC_WRITE_OLD (0x8 << 28) #define GFX_RLCG_GC_WRITE (0x0 << 28) #define GFX_RLCG_GC_READ (0x1 << 28) @@ -239,6 +242,13 @@ MODULE_FIRMWARE("amdgpu/beige_goby_mec.bin"); MODULE_FIRMWARE("amdgpu/beige_goby_mec2.bin"); MODULE_FIRMWARE("amdgpu/beige_goby_rlc.bin"); +MODULE_FIRMWARE("amdgpu/yellow_carp_ce.bin"); +MODULE_FIRMWARE("amdgpu/yellow_carp_pfp.bin"); +MODULE_FIRMWARE("amdgpu/yellow_carp_me.bin"); +MODULE_FIRMWARE("amdgpu/yellow_carp_mec.bin"); +MODULE_FIRMWARE("amdgpu/yellow_carp_mec2.bin"); +MODULE_FIRMWARE("amdgpu/yellow_carp_rlc.bin"); + static const struct soc15_reg_golden golden_settings_gc_10_1[] = { SOC15_REG_GOLDEN_VALUE(GC, 0, mmCB_HW_CONTROL_4, 0xffffffff, 0x00400014), @@ -1485,8 +1495,15 @@ static u32 gfx_v10_rlcg_rw(struct amdgpu_device *adev, u32 offset, u32 v, uint32 (adev->reg_offset[GC_HWIP][0][mmSCRATCH_REG0_BASE_IDX] + mmSCRATCH_REG2) * 4; scratch_reg3 = adev->rmmio + (adev->reg_offset[GC_HWIP][0][mmSCRATCH_REG1_BASE_IDX] + mmSCRATCH_REG3) * 4; - spare_int = adev->rmmio + - (adev->reg_offset[GC_HWIP][0][mmRLC_SPARE_INT_BASE_IDX] + mmRLC_SPARE_INT) * 4; + + if (adev->asic_type >= CHIP_SIENNA_CICHLID) { + spare_int = adev->rmmio + + (adev->reg_offset[GC_HWIP][0][mmRLC_SPARE_INT_0_Sienna_Cichlid_BASE_IDX] + + mmRLC_SPARE_INT_0_Sienna_Cichlid) * 4; + } else { + spare_int = adev->rmmio + + (adev->reg_offset[GC_HWIP][0][mmRLC_SPARE_INT_BASE_IDX] + mmRLC_SPARE_INT) * 4; + } grbm_cntl = adev->reg_offset[GC_HWIP][0][mmGRBM_GFX_CNTL_BASE_IDX] + mmGRBM_GFX_CNTL; grbm_idx = adev->reg_offset[GC_HWIP][0][mmGRBM_GFX_INDEX_BASE_IDX] + mmGRBM_GFX_INDEX; @@ -3370,6 +3387,30 @@ static const struct soc15_reg_golden golden_settings_gc_10_3_vangogh[] = SOC15_REG_GOLDEN_VALUE(GC, 0, mmLDS_CONFIG, 0x00000020, 0x00000020), }; +static const struct soc15_reg_golden golden_settings_gc_10_3_3[] = +{ + SOC15_REG_GOLDEN_VALUE(GC, 0, mmCGTT_SPI_CS_CLK_CTRL, 0xff7f0fff, 0x78000100), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmCH_PIPE_STEER, 0x000000ff, 0x000000e4), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmCPF_GCR_CNTL, 0x0007ffff, 0x0000c200), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmDB_DEBUG3, 0xffffffff, 0x00000280), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmDB_DEBUG4, 0xffffffff, 0x00800000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGB_ADDR_CONFIG, 0x0c1807ff, 0x00000242), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGCR_GENERAL_CNTL, 0x1ff1ffff, 0x00000500), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGL1_PIPE_STEER, 0x000000ff, 0x000000e4), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGL2_PIPE_STEER_0, 0x77777777, 0x32103210), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGL2_PIPE_STEER_1, 0x77777777, 0x32103210), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGL2A_ADDR_MATCH_MASK, 0xffffffff, 0xfffffff3), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGL2C_ADDR_MATCH_MASK, 0xffffffff, 0xfffffff3), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGL2C_CM_CTRL1, 0xff8fff0f, 0x580f1008), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGL2C_CTRL3, 0xf7ffffff, 0x00f80988), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmLDS_CONFIG, 0x000001ff, 0x00000020), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_CL_ENHANCE, 0xf17fffff, 0x01200007), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_BINNER_TIMEOUT_COUNTER, 0xffffffff, 0x00000800), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_ENHANCE_2, 0xffffffbf, 0x00000820), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmTA_CNTL_AUX, 0xfff7ffff, 0x01030000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmUTCL1_CTRL, 0xffffffff, 0x00100000) +}; + static const struct soc15_reg_golden golden_settings_gc_10_3_4[] = { SOC15_REG_GOLDEN_VALUE(GC, 0, mmCGTT_SPI_CS_CLK_CTRL, 0x78000000, 0x78000100), @@ -3659,6 +3700,11 @@ static void gfx_v10_0_init_golden_registers(struct amdgpu_device *adev) golden_settings_gc_10_3_vangogh, (const u32)ARRAY_SIZE(golden_settings_gc_10_3_vangogh)); break; + case CHIP_YELLOW_CARP: + soc15_program_register_sequence(adev, + golden_settings_gc_10_3_3, + (const u32)ARRAY_SIZE(golden_settings_gc_10_3_3)); + break; case CHIP_DIMGREY_CAVEFISH: soc15_program_register_sequence(adev, golden_settings_gc_10_3_4, @@ -3855,6 +3901,7 @@ static void gfx_v10_0_check_fw_write_wait(struct amdgpu_device *adev) case CHIP_VANGOGH: case CHIP_DIMGREY_CAVEFISH: case CHIP_BEIGE_GOBY: + case CHIP_YELLOW_CARP: adev->gfx.cp_fw_write_wait = true; break; default: @@ -3972,6 +4019,9 @@ static int gfx_v10_0_init_microcode(struct amdgpu_device *adev) case CHIP_BEIGE_GOBY: chip_name = "beige_goby"; break; + case CHIP_YELLOW_CARP: + chip_name = "yellow_carp"; + break; default: BUG(); } @@ -4541,6 +4591,7 @@ static void gfx_v10_0_gpu_early_init(struct amdgpu_device *adev) case CHIP_VANGOGH: case CHIP_DIMGREY_CAVEFISH: case CHIP_BEIGE_GOBY: + case CHIP_YELLOW_CARP: adev->gfx.config.max_hw_contexts = 8; adev->gfx.config.sc_prim_fifo_size_frontend = 0x20; adev->gfx.config.sc_prim_fifo_size_backend = 0x100; @@ -4666,6 +4717,7 @@ static int gfx_v10_0_sw_init(void *handle) case CHIP_VANGOGH: case CHIP_DIMGREY_CAVEFISH: case CHIP_BEIGE_GOBY: + case CHIP_YELLOW_CARP: adev->gfx.me.num_me = 1; adev->gfx.me.num_pipe_per_me = 1; adev->gfx.me.num_queue_per_pipe = 1; @@ -4898,7 +4950,8 @@ static void gfx_v10_0_setup_rb(struct amdgpu_device *adev) for (i = 0; i < adev->gfx.config.max_shader_engines; i++) { for (j = 0; j < adev->gfx.config.max_sh_per_se; j++) { bitmap = i * adev->gfx.config.max_sh_per_se + j; - if ((adev->asic_type == CHIP_SIENNA_CICHLID) && + if (((adev->asic_type == CHIP_SIENNA_CICHLID) || + (adev->asic_type == CHIP_YELLOW_CARP)) && ((gfx_v10_3_get_disabled_sa(adev) >> bitmap) & 1)) continue; gfx_v10_0_select_se_sh(adev, i, j, 0xffffffff); @@ -6174,6 +6227,7 @@ static void gfx_v10_0_cp_gfx_set_doorbell(struct amdgpu_device *adev, case CHIP_VANGOGH: case CHIP_DIMGREY_CAVEFISH: case CHIP_BEIGE_GOBY: + case CHIP_YELLOW_CARP: tmp = REG_SET_FIELD(0, CP_RB_DOORBELL_RANGE_LOWER, DOORBELL_RANGE_LOWER_Sienna_Cichlid, ring->doorbell_index); WREG32_SOC15(GC, 0, mmCP_RB_DOORBELL_RANGE_LOWER, tmp); @@ -6310,6 +6364,7 @@ static void gfx_v10_0_cp_compute_enable(struct amdgpu_device *adev, bool enable) case CHIP_VANGOGH: case CHIP_DIMGREY_CAVEFISH: case CHIP_BEIGE_GOBY: + case CHIP_YELLOW_CARP: WREG32_SOC15(GC, 0, mmCP_MEC_CNTL_Sienna_Cichlid, 0); break; default: @@ -6323,6 +6378,7 @@ static void gfx_v10_0_cp_compute_enable(struct amdgpu_device *adev, bool enable) case CHIP_VANGOGH: case CHIP_DIMGREY_CAVEFISH: case CHIP_BEIGE_GOBY: + case CHIP_YELLOW_CARP: WREG32_SOC15(GC, 0, mmCP_MEC_CNTL_Sienna_Cichlid, (CP_MEC_CNTL__MEC_ME1_HALT_MASK | CP_MEC_CNTL__MEC_ME2_HALT_MASK)); @@ -6420,6 +6476,7 @@ static void gfx_v10_0_kiq_setting(struct amdgpu_ring *ring) case CHIP_VANGOGH: case CHIP_DIMGREY_CAVEFISH: case CHIP_BEIGE_GOBY: + case CHIP_YELLOW_CARP: tmp = RREG32_SOC15(GC, 0, mmRLC_CP_SCHEDULERS_Sienna_Cichlid); tmp &= 0xffffff00; tmp |= (ring->me << 5) | (ring->pipe << 3) | (ring->queue); @@ -7148,6 +7205,7 @@ static bool gfx_v10_0_check_grbm_cam_remapping(struct amdgpu_device *adev) } break; case CHIP_VANGOGH: + case CHIP_YELLOW_CARP: return true; default: data = RREG32_SOC15(GC, 0, mmVGT_ESGS_RING_SIZE); @@ -7182,6 +7240,7 @@ static void gfx_v10_0_setup_grbm_cam_remapping(struct amdgpu_device *adev) case CHIP_VANGOGH: case CHIP_DIMGREY_CAVEFISH: case CHIP_BEIGE_GOBY: + case CHIP_YELLOW_CARP: /* mmVGT_TF_RING_SIZE_UMD -> mmVGT_TF_RING_SIZE */ data = (SOC15_REG_OFFSET(GC, 0, mmVGT_TF_RING_SIZE_UMD) << GRBM_CAM_DATA__CAM_ADDR__SHIFT) | @@ -7406,9 +7465,15 @@ static int gfx_v10_0_hw_fini(void *handle) if (amdgpu_sriov_vf(adev)) { gfx_v10_0_cp_gfx_enable(adev, false); /* Program KIQ position of RLC_CP_SCHEDULERS during destroy */ - tmp = RREG32_SOC15(GC, 0, mmRLC_CP_SCHEDULERS); - tmp &= 0xffffff00; - WREG32_SOC15(GC, 0, mmRLC_CP_SCHEDULERS, tmp); + if (adev->asic_type >= CHIP_SIENNA_CICHLID) { + tmp = RREG32_SOC15(GC, 0, mmRLC_CP_SCHEDULERS_Sienna_Cichlid); + tmp &= 0xffffff00; + WREG32_SOC15(GC, 0, mmRLC_CP_SCHEDULERS_Sienna_Cichlid, tmp); + } else { + tmp = RREG32_SOC15(GC, 0, mmRLC_CP_SCHEDULERS); + tmp &= 0xffffff00; + WREG32_SOC15(GC, 0, mmRLC_CP_SCHEDULERS, tmp); + } return 0; } @@ -7492,6 +7557,7 @@ static int gfx_v10_0_soft_reset(void *handle) case CHIP_VANGOGH: case CHIP_DIMGREY_CAVEFISH: case CHIP_BEIGE_GOBY: + case CHIP_YELLOW_CARP: if (REG_GET_FIELD(tmp, GRBM_STATUS2, RLC_BUSY_Sienna_Cichlid)) grbm_soft_reset = REG_SET_FIELD(grbm_soft_reset, GRBM_SOFT_RESET, @@ -7545,6 +7611,7 @@ static uint64_t gfx_v10_0_get_gpu_clock_counter(struct amdgpu_device *adev) mutex_lock(&adev->gfx.gpu_clock_mutex); switch (adev->asic_type) { case CHIP_VANGOGH: + case CHIP_YELLOW_CARP: clock = (uint64_t)RREG32_SOC15(SMUIO, 0, mmGOLDEN_TSC_COUNT_LOWER_Vangogh) | ((uint64_t)RREG32_SOC15(SMUIO, 0, mmGOLDEN_TSC_COUNT_UPPER_Vangogh) << 32ULL); break; @@ -7602,6 +7669,7 @@ static int gfx_v10_0_early_init(void *handle) case CHIP_VANGOGH: case CHIP_DIMGREY_CAVEFISH: case CHIP_BEIGE_GOBY: + case CHIP_YELLOW_CARP: adev->gfx.num_gfx_rings = GFX10_NUM_GFX_RINGS_Sienna_Cichlid; break; default: @@ -7659,6 +7727,7 @@ static void gfx_v10_0_set_safe_mode(struct amdgpu_device *adev) case CHIP_VANGOGH: case CHIP_DIMGREY_CAVEFISH: case CHIP_BEIGE_GOBY: + case CHIP_YELLOW_CARP: WREG32_SOC15(GC, 0, mmRLC_SAFE_MODE_Sienna_Cichlid, data); /* wait for RLC_SAFE_MODE */ @@ -7694,6 +7763,7 @@ static void gfx_v10_0_unset_safe_mode(struct amdgpu_device *adev) case CHIP_VANGOGH: case CHIP_DIMGREY_CAVEFISH: case CHIP_BEIGE_GOBY: + case CHIP_YELLOW_CARP: WREG32_SOC15(GC, 0, mmRLC_SAFE_MODE_Sienna_Cichlid, data); break; default: @@ -7996,12 +8066,23 @@ static void gfx_v10_cntl_power_gating(struct amdgpu_device *adev, bool enable) * in refclk count. Note that RLC FW is modified to take 16 bits from * RLC_PG_DELAY_3[15:0] as the hysteresis instead of just 8 bits. * - * The recommendation from RLC team is setting RLC_PG_DELAY_3 to 200us(0x4E20) - * as part of CGPG enablement starting point. + * The recommendation from RLC team is setting RLC_PG_DELAY_3 to 200us as part) + * of CGPG enablement starting point. + * Power/performance team will optimize it and might give a new value later. */ - if (enable && (adev->pg_flags & AMD_PG_SUPPORT_GFX_PG) && adev->asic_type == CHIP_VANGOGH) { - data = 0x4E20 & RLC_PG_DELAY_3__CGCG_ACTIVE_BEFORE_CGPG_MASK_Vangogh; - WREG32_SOC15(GC, 0, mmRLC_PG_DELAY_3, data); + if (enable && (adev->pg_flags & AMD_PG_SUPPORT_GFX_PG)) { + switch (adev->asic_type) { + case CHIP_VANGOGH: + data = 0x4E20 & RLC_PG_DELAY_3__CGCG_ACTIVE_BEFORE_CGPG_MASK_Vangogh; + WREG32_SOC15(GC, 0, mmRLC_PG_DELAY_3, data); + break; + case CHIP_YELLOW_CARP: + data = 0x1388 & RLC_PG_DELAY_3__CGCG_ACTIVE_BEFORE_CGPG_MASK_Vangogh; + WREG32_SOC15(GC, 0, mmRLC_PG_DELAY_3, data); + break; + default: + break; + } } } @@ -8065,6 +8146,7 @@ static int gfx_v10_0_set_powergating_state(void *handle, amdgpu_gfx_off_ctrl(adev, enable); break; case CHIP_VANGOGH: + case CHIP_YELLOW_CARP: gfx_v10_cntl_pg(adev, enable); amdgpu_gfx_off_ctrl(adev, enable); break; @@ -8091,6 +8173,7 @@ static int gfx_v10_0_set_clockgating_state(void *handle, case CHIP_VANGOGH: case CHIP_DIMGREY_CAVEFISH: case CHIP_BEIGE_GOBY: + case CHIP_YELLOW_CARP: gfx_v10_0_update_gfx_clock_gating(adev, state == AMD_CG_STATE_GATE); break; @@ -9197,14 +9280,15 @@ static void gfx_v10_0_set_rlc_funcs(struct amdgpu_device *adev) switch (adev->asic_type) { case CHIP_NAVI10: case CHIP_NAVI14: - case CHIP_SIENNA_CICHLID: case CHIP_NAVY_FLOUNDER: case CHIP_VANGOGH: case CHIP_DIMGREY_CAVEFISH: case CHIP_BEIGE_GOBY: + case CHIP_YELLOW_CARP: adev->gfx.rlc.funcs = &gfx_v10_0_rlc_funcs; break; case CHIP_NAVI12: + case CHIP_SIENNA_CICHLID: adev->gfx.rlc.funcs = &gfx_v10_0_rlc_funcs_sriov; break; default: @@ -9287,7 +9371,8 @@ static int gfx_v10_0_get_cu_info(struct amdgpu_device *adev, for (i = 0; i < adev->gfx.config.max_shader_engines; i++) { for (j = 0; j < adev->gfx.config.max_sh_per_se; j++) { bitmap = i * adev->gfx.config.max_sh_per_se + j; - if ((adev->asic_type == CHIP_SIENNA_CICHLID) && + if (((adev->asic_type == CHIP_SIENNA_CICHLID) || + (adev->asic_type == CHIP_YELLOW_CARP)) && ((gfx_v10_3_get_disabled_sa(adev) >> bitmap) & 1)) continue; mask = 1; diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c index fe5908f708cc..044076ec1d03 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c @@ -790,7 +790,8 @@ static void gfx_v9_0_rlcg_w(struct amdgpu_device *adev, u32 offset, u32 v, u32 f static void gfx_v9_0_rlcg_wreg(struct amdgpu_device *adev, u32 offset, u32 v, u32 acc_flags, u32 hwip) { - if (amdgpu_sriov_fullaccess(adev)) { + if ((acc_flags & AMDGPU_REGS_RLC) && + amdgpu_sriov_fullaccess(adev)) { gfx_v9_0_rlcg_w(adev, offset, v, acc_flags); return; diff --git a/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c b/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c index a03fdd41212b..bda1542ef1dd 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c @@ -321,18 +321,6 @@ static void gfxhub_v1_0_program_invalidation(struct amdgpu_device *adev) static int gfxhub_v1_0_gart_enable(struct amdgpu_device *adev) { - if (amdgpu_sriov_vf(adev) && adev->asic_type != CHIP_ARCTURUS) { - /* - * MC_VM_FB_LOCATION_BASE/TOP is NULL for VF, becuase they are - * VF copy registers so vbios post doesn't program them, for - * SRIOV driver need to program them - */ - WREG32_SOC15_RLC(GC, 0, mmMC_VM_FB_LOCATION_BASE, - adev->gmc.vram_start >> 24); - WREG32_SOC15_RLC(GC, 0, mmMC_VM_FB_LOCATION_TOP, - adev->gmc.vram_end >> 24); - } - /* GART Enable. */ gfxhub_v1_0_init_gart_aperture_regs(adev); gfxhub_v1_0_init_system_aperture_regs(adev); diff --git a/drivers/gpu/drm/amd/amdgpu/gfxhub_v2_1.c b/drivers/gpu/drm/amd/amdgpu/gfxhub_v2_1.c index 41807817de7d..1a374ec0514a 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfxhub_v2_1.c +++ b/drivers/gpu/drm/amd/amdgpu/gfxhub_v2_1.c @@ -31,6 +31,9 @@ #include "soc15_common.h" +#define mmGCUTCL2_HARVEST_BYPASS_GROUPS_YELLOW_CARP 0x16f8 +#define mmGCUTCL2_HARVEST_BYPASS_GROUPS_YELLOW_CARP_BASE_IDX 0 + static const char *gfxhub_client_ids[] = { "CB/DB", "Reserved", @@ -531,6 +534,42 @@ static int gfxhub_v2_1_get_xgmi_info(struct amdgpu_device *adev) return 0; } +static void gfxhub_v2_1_utcl2_harvest(struct amdgpu_device *adev) +{ + int i; + u32 tmp = 0, disabled_sa = 0; + u32 efuse_setting, vbios_setting; + + u32 max_sa_mask = amdgpu_gfx_create_bitmask( + adev->gfx.config.max_sh_per_se * + adev->gfx.config.max_shader_engines); + + if (adev->asic_type == CHIP_YELLOW_CARP) { + /* Get SA disabled bitmap from eFuse setting */ + efuse_setting = RREG32_SOC15(GC, 0, mmCC_GC_SA_UNIT_DISABLE); + efuse_setting &= CC_GC_SA_UNIT_DISABLE__SA_DISABLE_MASK; + efuse_setting >>= CC_GC_SA_UNIT_DISABLE__SA_DISABLE__SHIFT; + + /* Get SA disabled bitmap from VBIOS setting */ + vbios_setting = RREG32_SOC15(GC, 0, mmGC_USER_SA_UNIT_DISABLE); + vbios_setting &= GC_USER_SA_UNIT_DISABLE__SA_DISABLE_MASK; + vbios_setting >>= GC_USER_SA_UNIT_DISABLE__SA_DISABLE__SHIFT; + + disabled_sa |= efuse_setting | vbios_setting; + /* Make sure not to report harvested SAs beyond the max SA count */ + disabled_sa &= max_sa_mask; + + for (i = 0; disabled_sa > 0; i++) { + if (disabled_sa & 1) + tmp |= 0x3 << (i * 2); + disabled_sa >>= 1; + } + disabled_sa = tmp; + + WREG32_SOC15(GC, 0, mmGCUTCL2_HARVEST_BYPASS_GROUPS_YELLOW_CARP, disabled_sa); + } +} + const struct amdgpu_gfxhub_funcs gfxhub_v2_1_funcs = { .get_fb_location = gfxhub_v2_1_get_fb_location, .get_mc_fb_offset = gfxhub_v2_1_get_mc_fb_offset, @@ -540,4 +579,5 @@ const struct amdgpu_gfxhub_funcs gfxhub_v2_1_funcs = { .set_fault_enable_default = gfxhub_v2_1_set_fault_enable_default, .init = gfxhub_v2_1_init, .get_xgmi_info = gfxhub_v2_1_get_xgmi_info, + .utcl2_harvest = gfxhub_v2_1_utcl2_harvest, }; diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c index db154bf083e2..4523df2785d6 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c @@ -675,6 +675,7 @@ static void gmc_v10_0_set_mmhub_funcs(struct amdgpu_device *adev) { switch (adev->asic_type) { case CHIP_VANGOGH: + case CHIP_YELLOW_CARP: adev->mmhub.funcs = &mmhub_v2_3_funcs; break; default: @@ -691,6 +692,7 @@ static void gmc_v10_0_set_gfxhub_funcs(struct amdgpu_device *adev) case CHIP_VANGOGH: case CHIP_DIMGREY_CAVEFISH: case CHIP_BEIGE_GOBY: + case CHIP_YELLOW_CARP: adev->gfxhub.funcs = &gfxhub_v2_1_funcs; break; default: @@ -807,6 +809,7 @@ static int gmc_v10_0_mc_init(struct amdgpu_device *adev) case CHIP_VANGOGH: case CHIP_DIMGREY_CAVEFISH: case CHIP_BEIGE_GOBY: + case CHIP_YELLOW_CARP: default: adev->gmc.gart_size = 512ULL << 20; break; @@ -875,6 +878,7 @@ static int gmc_v10_0_sw_init(void *handle) case CHIP_VANGOGH: case CHIP_DIMGREY_CAVEFISH: case CHIP_BEIGE_GOBY: + case CHIP_YELLOW_CARP: adev->num_vmhubs = 2; /* * To fulfill 4-level page support, @@ -932,6 +936,7 @@ static int gmc_v10_0_sw_init(void *handle) return r; amdgpu_gmc_get_vbios_allocations(adev); + amdgpu_gmc_get_reserved_allocation(adev); /* Memory manager */ r = amdgpu_bo_init(adev); @@ -990,6 +995,7 @@ static void gmc_v10_0_init_golden_registers(struct amdgpu_device *adev) case CHIP_VANGOGH: case CHIP_DIMGREY_CAVEFISH: case CHIP_BEIGE_GOBY: + case CHIP_YELLOW_CARP: break; default: break; @@ -1053,6 +1059,13 @@ static int gmc_v10_0_hw_init(void *handle) /* The sequence of these two function calls matters.*/ gmc_v10_0_init_golden_registers(adev); + /* + * harvestable groups in gc_utcl2 need to be programmed before any GFX block + * register setup within GMC, or else system hang when harvesting SA. + */ + if (adev->gfxhub.funcs && adev->gfxhub.funcs->utcl2_harvest) + adev->gfxhub.funcs->utcl2_harvest(adev); + r = gmc_v10_0_gart_enable(adev); if (r) return r; @@ -1145,7 +1158,7 @@ static int gmc_v10_0_set_clockgating_state(void *handle, return r; if (adev->asic_type >= CHIP_SIENNA_CICHLID && - adev->asic_type <= CHIP_DIMGREY_CAVEFISH) + adev->asic_type <= CHIP_YELLOW_CARP) return athub_v2_1_set_clockgating(adev, state); else return athub_v2_0_set_clockgating(adev, state); @@ -1158,7 +1171,7 @@ static void gmc_v10_0_get_clockgating_state(void *handle, u32 *flags) adev->mmhub.funcs->get_clockgating(adev, flags); if (adev->asic_type >= CHIP_SIENNA_CICHLID && - adev->asic_type <= CHIP_DIMGREY_CAVEFISH) + adev->asic_type <= CHIP_YELLOW_CARP) athub_v2_1_get_clockgating(adev, flags); else athub_v2_0_get_clockgating(adev, flags); diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c index 68e6b6b8989a..7eb70d69f760 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c @@ -1288,10 +1288,7 @@ static int gmc_v9_0_late_init(void *handle) static void gmc_v9_0_vram_gtt_location(struct amdgpu_device *adev, struct amdgpu_gmc *mc) { - u64 base = 0; - - if (!amdgpu_sriov_vf(adev)) - base = adev->mmhub.funcs->get_fb_location(adev); + u64 base = adev->mmhub.funcs->get_fb_location(adev); /* add the xgmi offset of the physical node */ base += adev->gmc.xgmi.physical_node_id * adev->gmc.xgmi.node_segment_size; diff --git a/drivers/gpu/drm/amd/amdgpu/jpeg_v3_0.c b/drivers/gpu/drm/amd/amdgpu/jpeg_v3_0.c index bd77794315bc..01c242c5abc3 100644 --- a/drivers/gpu/drm/amd/amdgpu/jpeg_v3_0.c +++ b/drivers/gpu/drm/amd/amdgpu/jpeg_v3_0.c @@ -49,10 +49,13 @@ static int jpeg_v3_0_set_powergating_state(void *handle, static int jpeg_v3_0_early_init(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; - u32 harvest = RREG32_SOC15(JPEG, 0, mmCC_UVD_HARVESTING); - if (harvest & CC_UVD_HARVESTING__UVD_DISABLE_MASK) - return -ENOENT; + if (adev->asic_type != CHIP_YELLOW_CARP) { + u32 harvest = RREG32_SOC15(JPEG, 0, mmCC_UVD_HARVESTING); + + if (harvest & CC_UVD_HARVESTING__UVD_DISABLE_MASK) + return -ENOENT; + } adev->jpeg.num_jpeg_inst = 1; diff --git a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_7.c b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_7.c index 6264934b67ac..f80a14a1b82d 100644 --- a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_7.c +++ b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_7.c @@ -111,6 +111,9 @@ static void mmhub_v1_7_init_system_aperture_regs(struct amdgpu_device *adev) WREG32_SOC15(MMHUB, 0, regMC_VM_AGP_BOT, adev->gmc.agp_start >> 24); WREG32_SOC15(MMHUB, 0, regMC_VM_AGP_TOP, adev->gmc.agp_end >> 24); + if (amdgpu_sriov_vf(adev)) + return; + /* Program the system aperture low logical page number. */ WREG32_SOC15(MMHUB, 0, regMC_VM_SYSTEM_APERTURE_LOW_ADDR, min(adev->gmc.fb_start, adev->gmc.agp_start) >> 18); @@ -129,8 +132,6 @@ static void mmhub_v1_7_init_system_aperture_regs(struct amdgpu_device *adev) WREG32_SOC15(MMHUB, 0, regMC_VM_SYSTEM_APERTURE_LOW_ADDR, 0x3FFFFFFF); WREG32_SOC15(MMHUB, 0, regMC_VM_SYSTEM_APERTURE_HIGH_ADDR, 0); } - if (amdgpu_sriov_vf(adev)) - return; /* Set default page address. */ value = amdgpu_gmc_vram_mc2pa(adev, adev->vram_scratch.gpu_addr); @@ -332,18 +333,6 @@ static void mmhub_v1_7_program_invalidation(struct amdgpu_device *adev) static int mmhub_v1_7_gart_enable(struct amdgpu_device *adev) { - if (amdgpu_sriov_vf(adev)) { - /* - * MC_VM_FB_LOCATION_BASE/TOP is NULL for VF, becuase they are - * VF copy registers so vbios post doesn't program them, for - * SRIOV driver need to program them - */ - WREG32_SOC15(MMHUB, 0, regMC_VM_FB_LOCATION_BASE, - adev->gmc.vram_start >> 24); - WREG32_SOC15(MMHUB, 0, regMC_VM_FB_LOCATION_TOP, - adev->gmc.vram_end >> 24); - } - /* GART Enable. */ mmhub_v1_7_init_gart_aperture_regs(adev); mmhub_v1_7_init_system_aperture_regs(adev); diff --git a/drivers/gpu/drm/amd/amdgpu/mmhub_v2_3.c b/drivers/gpu/drm/amd/amdgpu/mmhub_v2_3.c index 709ac576ac7e..88e457a150e0 100644 --- a/drivers/gpu/drm/amd/amdgpu/mmhub_v2_3.c +++ b/drivers/gpu/drm/amd/amdgpu/mmhub_v2_3.c @@ -92,6 +92,7 @@ mmhub_v2_3_print_l2_protection_fault_status(struct amdgpu_device *adev, status); switch (adev->asic_type) { case CHIP_VANGOGH: + case CHIP_YELLOW_CARP: mmhub_cid = mmhub_client_ids_vangogh[cid][rw]; break; default: diff --git a/drivers/gpu/drm/amd/amdgpu/navi10_ih.c b/drivers/gpu/drm/amd/amdgpu/navi10_ih.c index baa2fd119e6e..530011622801 100644 --- a/drivers/gpu/drm/amd/amdgpu/navi10_ih.c +++ b/drivers/gpu/drm/amd/amdgpu/navi10_ih.c @@ -120,11 +120,23 @@ force_update_wptr_for_self_int(struct amdgpu_device *adev, ih_rb_cntl = REG_SET_FIELD(ih_rb_cntl, IH_RB_CNTL_RING1, RB_USED_INT_THRESHOLD, threshold); - WREG32_SOC15(OSSSYS, 0, mmIH_RB_CNTL_RING1, ih_rb_cntl); + if (amdgpu_sriov_vf(adev) && amdgpu_sriov_reg_indirect_ih(adev)) { + if (psp_reg_program(&adev->psp, PSP_REG_IH_RB_CNTL_RING1, ih_rb_cntl)) + return; + } else { + WREG32_SOC15(OSSSYS, 0, mmIH_RB_CNTL_RING1, ih_rb_cntl); + } + ih_rb_cntl = RREG32_SOC15(OSSSYS, 0, mmIH_RB_CNTL_RING2); ih_rb_cntl = REG_SET_FIELD(ih_rb_cntl, IH_RB_CNTL_RING2, RB_USED_INT_THRESHOLD, threshold); - WREG32_SOC15(OSSSYS, 0, mmIH_RB_CNTL_RING2, ih_rb_cntl); + if (amdgpu_sriov_vf(adev) && amdgpu_sriov_reg_indirect_ih(adev)) { + if (psp_reg_program(&adev->psp, PSP_REG_IH_RB_CNTL_RING2, ih_rb_cntl)) + return; + } else { + WREG32_SOC15(OSSSYS, 0, mmIH_RB_CNTL_RING2, ih_rb_cntl); + } + WREG32_SOC15(OSSSYS, 0, mmIH_CNTL2, ih_cntl); } @@ -153,10 +165,8 @@ static int navi10_ih_toggle_ring_interrupts(struct amdgpu_device *adev, tmp = REG_SET_FIELD(tmp, IH_RB_CNTL, ENABLE_INTR, (enable ? 1 : 0)); if (amdgpu_sriov_vf(adev) && amdgpu_sriov_reg_indirect_ih(adev)) { - if (psp_reg_program(&adev->psp, ih_regs->psp_reg_id, tmp)) { - DRM_ERROR("PSP program IH_RB_CNTL failed!\n"); + if (psp_reg_program(&adev->psp, ih_regs->psp_reg_id, tmp)) return -ETIMEDOUT; - } } else { WREG32(ih_regs->ih_rb_cntl, tmp); } @@ -328,6 +338,7 @@ static int navi10_ih_irq_init(struct amdgpu_device *adev) case CHIP_VANGOGH: case CHIP_DIMGREY_CAVEFISH: case CHIP_BEIGE_GOBY: + case CHIP_YELLOW_CARP: ih_chicken = RREG32_SOC15(OSSSYS, 0, mmIH_CHICKEN_Sienna_Cichlid); ih_chicken = REG_SET_FIELD(ih_chicken, IH_CHICKEN, MC_SPACE_GPA_ENABLE, 1); diff --git a/drivers/gpu/drm/amd/amdgpu/nbio_v7_2.c b/drivers/gpu/drm/amd/amdgpu/nbio_v7_2.c index 598ce0e93627..8f2a315e7c73 100644 --- a/drivers/gpu/drm/amd/amdgpu/nbio_v7_2.c +++ b/drivers/gpu/drm/amd/amdgpu/nbio_v7_2.c @@ -28,6 +28,25 @@ #include "nbio/nbio_7_2_0_sh_mask.h" #include <uapi/linux/kfd_ioctl.h> +#define regRCC_STRAP0_RCC_DEV0_EPF0_STRAP0_YC 0x0015 +#define regRCC_STRAP0_RCC_DEV0_EPF0_STRAP0_YC_BASE_IDX 2 +#define regBIF_BX0_BIF_FB_EN_YC 0x0100 +#define regBIF_BX0_BIF_FB_EN_YC_BASE_IDX 2 +#define regBIF1_PCIE_MST_CTRL_3 0x4601c6 +#define regBIF1_PCIE_MST_CTRL_3_BASE_IDX 5 +#define BIF1_PCIE_MST_CTRL_3__CI_SWUS_MAX_READ_REQUEST_SIZE_MODE__SHIFT \ + 0x1b +#define BIF1_PCIE_MST_CTRL_3__CI_SWUS_MAX_READ_REQUEST_SIZE_PRIV__SHIFT \ + 0x1c +#define BIF1_PCIE_MST_CTRL_3__CI_SWUS_MAX_READ_REQUEST_SIZE_MODE_MASK \ + 0x08000000L +#define BIF1_PCIE_MST_CTRL_3__CI_SWUS_MAX_READ_REQUEST_SIZE_PRIV_MASK \ + 0x30000000L +#define regBIF1_PCIE_TX_POWER_CTRL_1 0x460187 +#define regBIF1_PCIE_TX_POWER_CTRL_1_BASE_IDX 5 +#define BIF1_PCIE_TX_POWER_CTRL_1__MST_MEM_LS_EN_MASK 0x00000001L +#define BIF1_PCIE_TX_POWER_CTRL_1__REPLAY_MEM_LS_EN_MASK 0x00000008L + static void nbio_v7_2_remap_hdp_registers(struct amdgpu_device *adev) { WREG32_SOC15(NBIO, 0, regBIF_BX0_REMAP_HDP_MEM_FLUSH_CNTL, @@ -38,7 +57,12 @@ static void nbio_v7_2_remap_hdp_registers(struct amdgpu_device *adev) static u32 nbio_v7_2_get_rev_id(struct amdgpu_device *adev) { - u32 tmp = RREG32_SOC15(NBIO, 0, regRCC_STRAP0_RCC_DEV0_EPF0_STRAP0); + u32 tmp; + + if (adev->asic_type == CHIP_YELLOW_CARP) + tmp = RREG32_SOC15(NBIO, 0, regRCC_STRAP0_RCC_DEV0_EPF0_STRAP0_YC); + else + tmp = RREG32_SOC15(NBIO, 0, regRCC_STRAP0_RCC_DEV0_EPF0_STRAP0); tmp &= RCC_STRAP0_RCC_DEV0_EPF0_STRAP0__STRAP_ATI_REV_ID_DEV0_F0_MASK; tmp >>= RCC_STRAP0_RCC_DEV0_EPF0_STRAP0__STRAP_ATI_REV_ID_DEV0_F0__SHIFT; @@ -49,11 +73,19 @@ static u32 nbio_v7_2_get_rev_id(struct amdgpu_device *adev) static void nbio_v7_2_mc_access_enable(struct amdgpu_device *adev, bool enable) { if (enable) - WREG32_SOC15(NBIO, 0, regBIF_BX0_BIF_FB_EN, - BIF_BX0_BIF_FB_EN__FB_READ_EN_MASK | - BIF_BX0_BIF_FB_EN__FB_WRITE_EN_MASK); + if (adev->asic_type == CHIP_YELLOW_CARP) + WREG32_SOC15(NBIO, 0, regBIF_BX0_BIF_FB_EN_YC, + BIF_BX0_BIF_FB_EN__FB_READ_EN_MASK | + BIF_BX0_BIF_FB_EN__FB_WRITE_EN_MASK); + else + WREG32_SOC15(NBIO, 0, regBIF_BX0_BIF_FB_EN, + BIF_BX0_BIF_FB_EN__FB_READ_EN_MASK | + BIF_BX0_BIF_FB_EN__FB_WRITE_EN_MASK); else - WREG32_SOC15(NBIO, 0, regBIF_BX0_BIF_FB_EN, 0); + if (adev->asic_type == CHIP_YELLOW_CARP) + WREG32_SOC15(NBIO, 0, regBIF_BX0_BIF_FB_EN_YC, 0); + else + WREG32_SOC15(NBIO, 0, regBIF_BX0_BIF_FB_EN, 0); } static u32 nbio_v7_2_get_memsize(struct amdgpu_device *adev) @@ -92,13 +124,13 @@ static void nbio_v7_2_vcn_doorbell_range(struct amdgpu_device *adev, bool use_do if (use_doorbell) { doorbell_range = REG_SET_FIELD(doorbell_range, - GDC0_BIF_VCN0_DOORBELL_RANGE, OFFSET, - doorbell_index); + GDC0_BIF_VCN0_DOORBELL_RANGE, OFFSET, + doorbell_index); doorbell_range = REG_SET_FIELD(doorbell_range, - GDC0_BIF_VCN0_DOORBELL_RANGE, SIZE, 8); + GDC0_BIF_VCN0_DOORBELL_RANGE, SIZE, 8); } else { doorbell_range = REG_SET_FIELD(doorbell_range, - GDC0_BIF_VCN0_DOORBELL_RANGE, SIZE, 0); + GDC0_BIF_VCN0_DOORBELL_RANGE, SIZE, 0); } WREG32_PCIE_PORT(reg, doorbell_range); @@ -123,22 +155,22 @@ static void nbio_v7_2_enable_doorbell_selfring_aperture(struct amdgpu_device *ad if (enable) { tmp = REG_SET_FIELD(tmp, BIF_BX_PF0_DOORBELL_SELFRING_GPA_APER_CNTL, - DOORBELL_SELFRING_GPA_APER_EN, 1) | - REG_SET_FIELD(tmp, BIF_BX_PF0_DOORBELL_SELFRING_GPA_APER_CNTL, - DOORBELL_SELFRING_GPA_APER_MODE, 1) | - REG_SET_FIELD(tmp, BIF_BX_PF0_DOORBELL_SELFRING_GPA_APER_CNTL, - DOORBELL_SELFRING_GPA_APER_SIZE, 0); + DOORBELL_SELFRING_GPA_APER_EN, 1) | + REG_SET_FIELD(tmp, BIF_BX_PF0_DOORBELL_SELFRING_GPA_APER_CNTL, + DOORBELL_SELFRING_GPA_APER_MODE, 1) | + REG_SET_FIELD(tmp, BIF_BX_PF0_DOORBELL_SELFRING_GPA_APER_CNTL, + DOORBELL_SELFRING_GPA_APER_SIZE, 0); WREG32_SOC15(NBIO, 0, - regBIF_BX_PF0_DOORBELL_SELFRING_GPA_APER_BASE_LOW, - lower_32_bits(adev->doorbell.base)); + regBIF_BX_PF0_DOORBELL_SELFRING_GPA_APER_BASE_LOW, + lower_32_bits(adev->doorbell.base)); WREG32_SOC15(NBIO, 0, - regBIF_BX_PF0_DOORBELL_SELFRING_GPA_APER_BASE_HIGH, - upper_32_bits(adev->doorbell.base)); + regBIF_BX_PF0_DOORBELL_SELFRING_GPA_APER_BASE_HIGH, + upper_32_bits(adev->doorbell.base)); } WREG32_SOC15(NBIO, 0, regBIF_BX_PF0_DOORBELL_SELFRING_GPA_APER_CNTL, - tmp); + tmp); } @@ -218,19 +250,42 @@ static void nbio_v7_2_update_medium_grain_light_sleep(struct amdgpu_device *adev { uint32_t def, data; - def = data = RREG32_PCIE_PORT(SOC15_REG_OFFSET(NBIO, 0, regPCIE_CNTL2)); - if (enable && (adev->cg_flags & AMD_CG_SUPPORT_BIF_LS)) { - data |= (PCIE_CNTL2__SLV_MEM_LS_EN_MASK | - PCIE_CNTL2__MST_MEM_LS_EN_MASK | - PCIE_CNTL2__REPLAY_MEM_LS_EN_MASK); + if (adev->asic_type == CHIP_YELLOW_CARP) { + def = data = RREG32_PCIE_PORT(SOC15_REG_OFFSET(NBIO, 0, regPCIE_CNTL2)); + if (enable && (adev->cg_flags & AMD_CG_SUPPORT_BIF_LS)) + data |= PCIE_CNTL2__SLV_MEM_LS_EN_MASK; + else + data &= ~PCIE_CNTL2__SLV_MEM_LS_EN_MASK; + + if (def != data) + WREG32_PCIE_PORT(SOC15_REG_OFFSET(NBIO, 0, regPCIE_CNTL2), data); + + data = RREG32_PCIE_PORT(SOC15_REG_OFFSET(NBIO, 0, regBIF1_PCIE_TX_POWER_CTRL_1)); + def = data; + if (enable && (adev->cg_flags & AMD_CG_SUPPORT_BIF_LS)) + data |= (BIF1_PCIE_TX_POWER_CTRL_1__MST_MEM_LS_EN_MASK | + BIF1_PCIE_TX_POWER_CTRL_1__REPLAY_MEM_LS_EN_MASK); + else + data &= ~(BIF1_PCIE_TX_POWER_CTRL_1__MST_MEM_LS_EN_MASK | + BIF1_PCIE_TX_POWER_CTRL_1__REPLAY_MEM_LS_EN_MASK); + + if (def != data) + WREG32_PCIE_PORT(SOC15_REG_OFFSET(NBIO, 0, regBIF1_PCIE_TX_POWER_CTRL_1), + data); } else { - data &= ~(PCIE_CNTL2__SLV_MEM_LS_EN_MASK | - PCIE_CNTL2__MST_MEM_LS_EN_MASK | - PCIE_CNTL2__REPLAY_MEM_LS_EN_MASK); + def = data = RREG32_PCIE_PORT(SOC15_REG_OFFSET(NBIO, 0, regPCIE_CNTL2)); + if (enable && (adev->cg_flags & AMD_CG_SUPPORT_BIF_LS)) + data |= (PCIE_CNTL2__SLV_MEM_LS_EN_MASK | + PCIE_CNTL2__MST_MEM_LS_EN_MASK | + PCIE_CNTL2__REPLAY_MEM_LS_EN_MASK); + else + data &= ~(PCIE_CNTL2__SLV_MEM_LS_EN_MASK | + PCIE_CNTL2__MST_MEM_LS_EN_MASK | + PCIE_CNTL2__REPLAY_MEM_LS_EN_MASK); + + if (def != data) + WREG32_PCIE_PORT(SOC15_REG_OFFSET(NBIO, 0, regPCIE_CNTL2), data); } - - if (def != data) - WREG32_PCIE_PORT(SOC15_REG_OFFSET(NBIO, 0, regPCIE_CNTL2), data); } static void nbio_v7_2_get_clockgating_state(struct amdgpu_device *adev, @@ -297,14 +352,25 @@ const struct nbio_hdp_flush_reg nbio_v7_2_hdp_flush_reg = { static void nbio_v7_2_init_registers(struct amdgpu_device *adev) { uint32_t def, data; - - def = data = RREG32_PCIE_PORT(SOC15_REG_OFFSET(NBIO, 0, regPCIE_CONFIG_CNTL)); - data = REG_SET_FIELD(data, PCIE_CONFIG_CNTL, CI_SWUS_MAX_READ_REQUEST_SIZE_MODE, 1); - data = REG_SET_FIELD(data, PCIE_CONFIG_CNTL, CI_SWUS_MAX_READ_REQUEST_SIZE_PRIV, 1); - - if (def != data) - WREG32_PCIE_PORT(SOC15_REG_OFFSET(NBIO, 0, regPCIE_CONFIG_CNTL), - data); + if (adev->asic_type == CHIP_YELLOW_CARP) { + def = data = RREG32_PCIE_PORT(SOC15_REG_OFFSET(NBIO, 0, regBIF1_PCIE_MST_CTRL_3)); + data = REG_SET_FIELD(data, BIF1_PCIE_MST_CTRL_3, + CI_SWUS_MAX_READ_REQUEST_SIZE_MODE, 1); + data = REG_SET_FIELD(data, BIF1_PCIE_MST_CTRL_3, + CI_SWUS_MAX_READ_REQUEST_SIZE_PRIV, 1); + + if (def != data) + WREG32_PCIE_PORT(SOC15_REG_OFFSET(NBIO, 0, regBIF1_PCIE_MST_CTRL_3), data); + } else { + def = data = RREG32_PCIE_PORT(SOC15_REG_OFFSET(NBIO, 0, regPCIE_CONFIG_CNTL)); + data = REG_SET_FIELD(data, PCIE_CONFIG_CNTL, + CI_SWUS_MAX_READ_REQUEST_SIZE_MODE, 1); + data = REG_SET_FIELD(data, PCIE_CONFIG_CNTL, + CI_SWUS_MAX_READ_REQUEST_SIZE_PRIV, 1); + + if (def != data) + WREG32_PCIE_PORT(SOC15_REG_OFFSET(NBIO, 0, regPCIE_CONFIG_CNTL), data); + } } const struct amdgpu_nbio_funcs nbio_v7_2_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/nv.c b/drivers/gpu/drm/amd/amdgpu/nv.c index 27ba0408a2aa..455d0425787c 100644 --- a/drivers/gpu/drm/amd/amdgpu/nv.c +++ b/drivers/gpu/drm/amd/amdgpu/nv.c @@ -329,6 +329,7 @@ static int nv_query_video_codecs(struct amdgpu_device *adev, bool encode, case CHIP_NAVY_FLOUNDER: case CHIP_DIMGREY_CAVEFISH: case CHIP_VANGOGH: + case CHIP_YELLOW_CARP: if (encode) *codecs = &nv_video_codecs_encode; else @@ -633,6 +634,7 @@ nv_asic_reset_method(struct amdgpu_device *adev) switch (adev->asic_type) { case CHIP_VANGOGH: + case CHIP_YELLOW_CARP: return AMD_RESET_METHOD_MODE2; case CHIP_SIENNA_CICHLID: case CHIP_NAVY_FLOUNDER: @@ -782,6 +784,9 @@ legacy_init: case CHIP_BEIGE_GOBY: beige_goby_reg_base_init(adev); break; + case CHIP_YELLOW_CARP: + yellow_carp_reg_base_init(adev); + break; default: return -EINVAL; } @@ -985,6 +990,26 @@ int nv_set_ip_blocks(struct amdgpu_device *adev) amdgpu_device_ip_block_add(adev, &smu_v11_0_ip_block); amdgpu_device_ip_block_add(adev, &vcn_v3_0_ip_block); break; + case CHIP_YELLOW_CARP: + amdgpu_device_ip_block_add(adev, &nv_common_ip_block); + amdgpu_device_ip_block_add(adev, &gmc_v10_0_ip_block); + amdgpu_device_ip_block_add(adev, &navi10_ih_ip_block); + if (likely(adev->firmware.load_type == AMDGPU_FW_LOAD_PSP)) + amdgpu_device_ip_block_add(adev, &psp_v13_0_ip_block); + amdgpu_device_ip_block_add(adev, &smu_v13_0_ip_block); + if (adev->enable_virtual_display || amdgpu_sriov_vf(adev)) + amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block); + amdgpu_device_ip_block_add(adev, &gfx_v10_0_ip_block); + amdgpu_device_ip_block_add(adev, &sdma_v5_2_ip_block); + if (adev->enable_virtual_display || amdgpu_sriov_vf(adev)) + amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block); +#if defined(CONFIG_DRM_AMD_DC) + else if (amdgpu_device_has_dc_support(adev)) + amdgpu_device_ip_block_add(adev, &dm_ip_block); +#endif + amdgpu_device_ip_block_add(adev, &vcn_v3_0_ip_block); + amdgpu_device_ip_block_add(adev, &jpeg_v3_0_ip_block); + break; default: return -EINVAL; } @@ -1309,6 +1334,32 @@ static int nv_common_early_init(void *handle) AMD_PG_SUPPORT_MMHUB; adev->external_rev_id = adev->rev_id + 0x46; break; + case CHIP_YELLOW_CARP: + adev->cg_flags = AMD_CG_SUPPORT_GFX_MGCG | + AMD_CG_SUPPORT_GFX_MGLS | + AMD_CG_SUPPORT_GFX_CGCG | + AMD_CG_SUPPORT_GFX_CGLS | + AMD_CG_SUPPORT_GFX_3D_CGCG | + AMD_CG_SUPPORT_GFX_3D_CGLS | + AMD_CG_SUPPORT_GFX_RLC_LS | + AMD_CG_SUPPORT_GFX_CP_LS | + AMD_CG_SUPPORT_GFX_FGCG | + AMD_CG_SUPPORT_MC_MGCG | + AMD_CG_SUPPORT_MC_LS | + AMD_CG_SUPPORT_SDMA_LS | + AMD_CG_SUPPORT_HDP_MGCG | + AMD_CG_SUPPORT_HDP_LS | + AMD_CG_SUPPORT_ATHUB_MGCG | + AMD_CG_SUPPORT_ATHUB_LS | + AMD_CG_SUPPORT_IH_CG | + AMD_CG_SUPPORT_VCN_MGCG | + AMD_CG_SUPPORT_JPEG_MGCG; + adev->pg_flags = AMD_PG_SUPPORT_GFX_PG | + AMD_PG_SUPPORT_VCN | + AMD_PG_SUPPORT_VCN_DPG | + AMD_PG_SUPPORT_JPEG; + adev->external_rev_id = adev->rev_id + 0x01; + break; default: /* FIXME: not supported yet */ return -EINVAL; diff --git a/drivers/gpu/drm/amd/amdgpu/nv.h b/drivers/gpu/drm/amd/amdgpu/nv.h index 330438388fde..eb9aff1e7253 100644 --- a/drivers/gpu/drm/amd/amdgpu/nv.h +++ b/drivers/gpu/drm/amd/amdgpu/nv.h @@ -37,4 +37,6 @@ int sienna_cichlid_reg_base_init(struct amdgpu_device *adev); void vangogh_reg_base_init(struct amdgpu_device *adev); int dimgrey_cavefish_reg_base_init(struct amdgpu_device *adev); int beige_goby_reg_base_init(struct amdgpu_device *adev); +int yellow_carp_reg_base_init(struct amdgpu_device *adev); + #endif diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v13_0.c b/drivers/gpu/drm/amd/amdgpu/psp_v13_0.c index fcdce46445d6..c8949add88f3 100644 --- a/drivers/gpu/drm/amd/amdgpu/psp_v13_0.c +++ b/drivers/gpu/drm/amd/amdgpu/psp_v13_0.c @@ -31,6 +31,9 @@ MODULE_FIRMWARE("amdgpu/aldebaran_sos.bin"); MODULE_FIRMWARE("amdgpu/aldebaran_ta.bin"); +MODULE_FIRMWARE("amdgpu/yellow_carp_asd.bin"); +MODULE_FIRMWARE("amdgpu/yellow_carp_toc.bin"); +MODULE_FIRMWARE("amdgpu/yellow_carp_ta.bin"); static int psp_v13_0_init_microcode(struct psp_context *psp) { @@ -42,17 +45,37 @@ static int psp_v13_0_init_microcode(struct psp_context *psp) case CHIP_ALDEBARAN: chip_name = "aldebaran"; break; + case CHIP_YELLOW_CARP: + chip_name = "yellow_carp"; + break; + default: + BUG(); + } + switch (adev->asic_type) { + case CHIP_ALDEBARAN: + err = psp_init_sos_microcode(psp, chip_name); + if (err) + return err; + err = psp_init_ta_microcode(&adev->psp, chip_name); + if (err) + return err; + break; + case CHIP_YELLOW_CARP: + err = psp_init_asd_microcode(psp, chip_name); + if (err) + return err; + err = psp_init_toc_microcode(psp, chip_name); + if (err) + return err; + err = psp_init_ta_microcode(psp, chip_name); + if (err) + return err; + break; default: BUG(); } - err = psp_init_sos_microcode(psp, chip_name); - if (err) - return err; - - err = psp_init_ta_microcode(&adev->psp, chip_name); - - return err; + return 0; } static bool psp_v13_0_is_sos_alive(struct psp_context *psp) diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c b/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c index 98059bce692f..7486e5306786 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c @@ -50,6 +50,7 @@ MODULE_FIRMWARE("amdgpu/dimgrey_cavefish_sdma.bin"); MODULE_FIRMWARE("amdgpu/beige_goby_sdma.bin"); MODULE_FIRMWARE("amdgpu/vangogh_sdma.bin"); +MODULE_FIRMWARE("amdgpu/yellow_carp_sdma.bin"); #define SDMA1_REG_OFFSET 0x600 #define SDMA3_REG_OFFSET 0x400 @@ -94,6 +95,7 @@ static void sdma_v5_2_init_golden_registers(struct amdgpu_device *adev) case CHIP_VANGOGH: case CHIP_DIMGREY_CAVEFISH: case CHIP_BEIGE_GOBY: + case CHIP_YELLOW_CARP: break; default: break; @@ -165,6 +167,9 @@ static int sdma_v5_2_init_microcode(struct amdgpu_device *adev) case CHIP_BEIGE_GOBY: chip_name = "beige_goby"; break; + case CHIP_YELLOW_CARP: + chip_name = "yellow_carp"; + break; default: BUG(); } @@ -495,12 +500,12 @@ static void sdma_v5_2_gfx_stop(struct amdgpu_device *adev) amdgpu_ttm_set_buffer_funcs_status(adev, false); for (i = 0; i < adev->sdma.num_instances; i++) { - rb_cntl = RREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_CNTL)); + rb_cntl = RREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_CNTL)); rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_GFX_RB_CNTL, RB_ENABLE, 0); - WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_CNTL), rb_cntl); - ib_cntl = RREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_IB_CNTL)); + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_CNTL), rb_cntl); + ib_cntl = RREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_IB_CNTL)); ib_cntl = REG_SET_FIELD(ib_cntl, SDMA0_GFX_IB_CNTL, IB_ENABLE, 0); - WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_IB_CNTL), ib_cntl); + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_IB_CNTL), ib_cntl); } } @@ -558,11 +563,11 @@ static void sdma_v5_2_ctx_switch_enable(struct amdgpu_device *adev, bool enable) f32_cntl = REG_SET_FIELD(f32_cntl, SDMA0_CNTL, AUTO_CTXSW_ENABLE, enable ? 1 : 0); if (enable && amdgpu_sdma_phase_quantum) { - WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_PHASE0_QUANTUM), + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_PHASE0_QUANTUM), phase_quantum); - WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_PHASE1_QUANTUM), + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_PHASE1_QUANTUM), phase_quantum); - WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_PHASE2_QUANTUM), + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_PHASE2_QUANTUM), phase_quantum); } WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_CNTL), f32_cntl); @@ -620,62 +625,62 @@ static int sdma_v5_2_gfx_resume(struct amdgpu_device *adev) ring = &adev->sdma.instance[i].ring; wb_offset = (ring->rptr_offs * 4); - WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_SEM_WAIT_FAIL_TIMER_CNTL), 0); + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_SEM_WAIT_FAIL_TIMER_CNTL), 0); /* Set ring buffer size in dwords */ rb_bufsz = order_base_2(ring->ring_size / 4); - rb_cntl = RREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_CNTL)); + rb_cntl = RREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_CNTL)); rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_GFX_RB_CNTL, RB_SIZE, rb_bufsz); #ifdef __BIG_ENDIAN rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_GFX_RB_CNTL, RB_SWAP_ENABLE, 1); rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_GFX_RB_CNTL, RPTR_WRITEBACK_SWAP_ENABLE, 1); #endif - WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_CNTL), rb_cntl); + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_CNTL), rb_cntl); /* Initialize the ring buffer's read and write pointers */ - WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_RPTR), 0); - WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_RPTR_HI), 0); - WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR), 0); - WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR_HI), 0); + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_RPTR), 0); + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_RPTR_HI), 0); + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR), 0); + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR_HI), 0); /* setup the wptr shadow polling */ wptr_gpu_addr = adev->wb.gpu_addr + (ring->wptr_offs * 4); - WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR_POLL_ADDR_LO), + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR_POLL_ADDR_LO), lower_32_bits(wptr_gpu_addr)); - WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR_POLL_ADDR_HI), + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR_POLL_ADDR_HI), upper_32_bits(wptr_gpu_addr)); - wptr_poll_cntl = RREG32(sdma_v5_2_get_reg_offset(adev, i, + wptr_poll_cntl = RREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR_POLL_CNTL)); wptr_poll_cntl = REG_SET_FIELD(wptr_poll_cntl, SDMA0_GFX_RB_WPTR_POLL_CNTL, F32_POLL_ENABLE, 1); - WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR_POLL_CNTL), + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR_POLL_CNTL), wptr_poll_cntl); /* set the wb address whether it's enabled or not */ - WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_RPTR_ADDR_HI), + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_RPTR_ADDR_HI), upper_32_bits(adev->wb.gpu_addr + wb_offset) & 0xFFFFFFFF); - WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_RPTR_ADDR_LO), + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_RPTR_ADDR_LO), lower_32_bits(adev->wb.gpu_addr + wb_offset) & 0xFFFFFFFC); rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_GFX_RB_CNTL, RPTR_WRITEBACK_ENABLE, 1); - WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_BASE), ring->gpu_addr >> 8); - WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_BASE_HI), ring->gpu_addr >> 40); + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_BASE), ring->gpu_addr >> 8); + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_BASE_HI), ring->gpu_addr >> 40); ring->wptr = 0; /* before programing wptr to a less value, need set minor_ptr_update first */ - WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_MINOR_PTR_UPDATE), 1); + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_MINOR_PTR_UPDATE), 1); if (!amdgpu_sriov_vf(adev)) { /* only bare-metal use register write for wptr */ WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR), lower_32_bits(ring->wptr) << 2); WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR_HI), upper_32_bits(ring->wptr) << 2); } - doorbell = RREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_DOORBELL)); - doorbell_offset = RREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_DOORBELL_OFFSET)); + doorbell = RREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_DOORBELL)); + doorbell_offset = RREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_DOORBELL_OFFSET)); if (ring->use_doorbell) { doorbell = REG_SET_FIELD(doorbell, SDMA0_GFX_DOORBELL, ENABLE, 1); @@ -684,8 +689,8 @@ static int sdma_v5_2_gfx_resume(struct amdgpu_device *adev) } else { doorbell = REG_SET_FIELD(doorbell, SDMA0_GFX_DOORBELL, ENABLE, 0); } - WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_DOORBELL), doorbell); - WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_DOORBELL_OFFSET), doorbell_offset); + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_DOORBELL), doorbell); + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_DOORBELL_OFFSET), doorbell_offset); adev->nbio.funcs->sdma_doorbell_range(adev, i, ring->use_doorbell, ring->doorbell_index, @@ -695,7 +700,7 @@ static int sdma_v5_2_gfx_resume(struct amdgpu_device *adev) sdma_v5_2_ring_set_wptr(ring); /* set minor_ptr_update to 0 after wptr programed */ - WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_MINOR_PTR_UPDATE), 0); + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_MINOR_PTR_UPDATE), 0); /* set utc l1 enable flag always to 1 */ temp = RREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_CNTL)); @@ -706,19 +711,19 @@ static int sdma_v5_2_gfx_resume(struct amdgpu_device *adev) WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_CNTL), temp); /* Set up RESP_MODE to non-copy addresses */ - temp = RREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_UTCL1_CNTL)); + temp = RREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_UTCL1_CNTL)); temp = REG_SET_FIELD(temp, SDMA0_UTCL1_CNTL, RESP_MODE, 3); temp = REG_SET_FIELD(temp, SDMA0_UTCL1_CNTL, REDO_DELAY, 9); - WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_UTCL1_CNTL), temp); + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_UTCL1_CNTL), temp); /* program default cache read and write policy */ - temp = RREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_UTCL1_PAGE)); + temp = RREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_UTCL1_PAGE)); /* clean read policy and write policy bits */ temp &= 0xFF0FFF; temp |= ((CACHE_READ_POLICY_L2__DEFAULT << 12) | (CACHE_WRITE_POLICY_L2__DEFAULT << 14) | SDMA0_UTCL1_PAGE__LLC_NOALLOC_MASK); - WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_UTCL1_PAGE), temp); + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_UTCL1_PAGE), temp); if (!amdgpu_sriov_vf(adev)) { /* unhalt engine */ @@ -729,15 +734,15 @@ static int sdma_v5_2_gfx_resume(struct amdgpu_device *adev) /* enable DMA RB */ rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_GFX_RB_CNTL, RB_ENABLE, 1); - WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_CNTL), rb_cntl); + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_CNTL), rb_cntl); - ib_cntl = RREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_IB_CNTL)); + ib_cntl = RREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_IB_CNTL)); ib_cntl = REG_SET_FIELD(ib_cntl, SDMA0_GFX_IB_CNTL, IB_ENABLE, 1); #ifdef __BIG_ENDIAN ib_cntl = REG_SET_FIELD(ib_cntl, SDMA0_GFX_IB_CNTL, IB_SWAP_ENABLE, 1); #endif /* enable DMA IBs */ - WREG32(sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_IB_CNTL), ib_cntl); + WREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_IB_CNTL), ib_cntl); ring->sched.ready = true; @@ -1229,6 +1234,7 @@ static int sdma_v5_2_early_init(void *handle) break; case CHIP_VANGOGH: case CHIP_BEIGE_GOBY: + case CHIP_YELLOW_CARP: adev->sdma.num_instances = 1; break; default: @@ -1630,6 +1636,7 @@ static int sdma_v5_2_set_clockgating_state(void *handle, case CHIP_VANGOGH: case CHIP_DIMGREY_CAVEFISH: case CHIP_BEIGE_GOBY: + case CHIP_YELLOW_CARP: sdma_v5_2_update_medium_grain_clock_gating(adev, state == AMD_CG_STATE_GATE); sdma_v5_2_update_medium_grain_light_sleep(adev, diff --git a/drivers/gpu/drm/amd/amdgpu/soc15_common.h b/drivers/gpu/drm/amd/amdgpu/soc15_common.h index c781808e4dc3..0eeb5e073be8 100644 --- a/drivers/gpu/drm/amd/amdgpu/soc15_common.h +++ b/drivers/gpu/drm/amd/amdgpu/soc15_common.h @@ -28,12 +28,12 @@ #define SOC15_REG_OFFSET(ip, inst, reg) (adev->reg_offset[ip##_HWIP][inst][reg##_BASE_IDX] + reg) #define __WREG32_SOC15_RLC__(reg, value, flag, hwip) \ - ((amdgpu_sriov_vf(adev) && adev->gfx.rlc.funcs->rlcg_wreg) ? \ + ((amdgpu_sriov_vf(adev) && adev->gfx.rlc.funcs && adev->gfx.rlc.funcs->rlcg_wreg) ? \ adev->gfx.rlc.funcs->rlcg_wreg(adev, reg, value, flag, hwip) : \ WREG32(reg, value)) #define __RREG32_SOC15_RLC__(reg, flag, hwip) \ - ((amdgpu_sriov_vf(adev) && adev->gfx.rlc.funcs->rlcg_rreg) ? \ + ((amdgpu_sriov_vf(adev) && adev->gfx.rlc.funcs && adev->gfx.rlc.funcs->rlcg_rreg) ? \ adev->gfx.rlc.funcs->rlcg_rreg(adev, reg, flag, hwip) : \ RREG32(reg)) diff --git a/drivers/gpu/drm/amd/amdgpu/yellow_carp_reg_init.c b/drivers/gpu/drm/amd/amdgpu/yellow_carp_reg_init.c new file mode 100644 index 000000000000..3d89421275ed --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/yellow_carp_reg_init.c @@ -0,0 +1,51 @@ +/* + * Copyright 2019 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ +#include "amdgpu.h" +#include "nv.h" + +#include "soc15_common.h" +#include "soc15_hw_ip.h" +#include "yellow_carp_offset.h" + +int yellow_carp_reg_base_init(struct amdgpu_device *adev) +{ + /* HW has more IP blocks, only initialized the block needed by driver */ + uint32_t i; + for (i = 0 ; i < MAX_INSTANCE ; ++i) { + adev->reg_offset[GC_HWIP][i] = (uint32_t *)(&(GC_BASE.instance[i])); + adev->reg_offset[HDP_HWIP][i] = (uint32_t *)(&(HDP_BASE.instance[i])); + adev->reg_offset[MMHUB_HWIP][i] = (uint32_t *)(&(MMHUB_BASE.instance[i])); + adev->reg_offset[ATHUB_HWIP][i] = (uint32_t *)(&(ATHUB_BASE.instance[i])); + adev->reg_offset[NBIO_HWIP][i] = (uint32_t *)(&(NBIO_BASE.instance[i])); + adev->reg_offset[MP0_HWIP][i] = (uint32_t *)(&(MP0_BASE.instance[i])); + adev->reg_offset[MP1_HWIP][i] = (uint32_t *)(&(MP1_BASE.instance[i])); + adev->reg_offset[VCN_HWIP][i] = (uint32_t *)(&(VCN_BASE.instance[i])); + adev->reg_offset[DF_HWIP][i] = (uint32_t *)(&(DF_BASE.instance[i])); + adev->reg_offset[DCE_HWIP][i] = (uint32_t *)(&(DCN_BASE.instance[i])); + adev->reg_offset[OSSSYS_HWIP][i] = (uint32_t *)(&(OSSSYS_BASE.instance[i])); + adev->reg_offset[SDMA0_HWIP][i] = (uint32_t *)(&(SDMA0_BASE.instance[i])); + adev->reg_offset[SMUIO_HWIP][i] = (uint32_t *)(&(SMUIO_BASE.instance[i])); + adev->reg_offset[THM_HWIP][i] = (uint32_t *)(&(THM_BASE.instance[i])); + } + return 0; +} |