diff options
Diffstat (limited to 'drivers/gpu/drm/amd')
188 files changed, 3930 insertions, 3031 deletions
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index ae9371b172e3..8639a4f9c6e8 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -274,9 +274,6 @@ extern int amdgpu_vcnfw_log; #define AMDGPU_RESET_VCE (1 << 13) #define AMDGPU_RESET_VCE1 (1 << 14) -#define AMDGPU_RESET_LEVEL_SOFT_RECOVERY (1 << 0) -#define AMDGPU_RESET_LEVEL_MODE2 (1 << 1) - /* max cursor sizes (in pixels) */ #define CIK_CURSOR_WIDTH 128 #define CIK_CURSOR_HEIGHT 128 @@ -1065,7 +1062,6 @@ struct amdgpu_device { struct work_struct reset_work; - uint32_t amdgpu_reset_level_mask; bool job_hang; }; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c index 3da27436922c..b14800ac179e 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c @@ -1053,6 +1053,10 @@ bool amdgpu_acpi_should_gpu_reset(struct amdgpu_device *adev) { if (adev->flags & AMD_IS_APU) return false; + + if (amdgpu_sriov_vf(adev)) + return false; + return pm_suspend_target_state != PM_SUSPEND_TO_IDLE; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c index c7b1a2dfde13..0561812aa0a4 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c @@ -75,9 +75,6 @@ void amdgpu_amdkfd_device_probe(struct amdgpu_device *adev) return; adev->kfd.dev = kgd2kfd_probe(adev, vf); - - if (adev->kfd.dev) - amdgpu_amdkfd_total_mem_size += adev->gmc.real_vram_size; } /** @@ -137,7 +134,6 @@ static void amdgpu_amdkfd_reset_work(struct work_struct *work) reset_context.method = AMD_RESET_METHOD_NONE; reset_context.reset_req_dev = adev; clear_bit(AMDGPU_NEED_FULL_RESET, &reset_context.flags); - clear_bit(AMDGPU_SKIP_MODE2_RESET, &reset_context.flags); amdgpu_device_gpu_recover(adev, NULL, &reset_context); } @@ -201,6 +197,8 @@ void amdgpu_amdkfd_device_init(struct amdgpu_device *adev) adev->kfd.init_complete = kgd2kfd_device_init(adev->kfd.dev, adev_to_drm(adev), &gpu_resources); + amdgpu_amdkfd_total_mem_size += adev->gmc.real_vram_size; + INIT_WORK(&adev->kfd.reset_work, amdgpu_amdkfd_reset_work); } } @@ -210,6 +208,7 @@ void amdgpu_amdkfd_device_fini_sw(struct amdgpu_device *adev) if (adev->kfd.dev) { kgd2kfd_device_exit(adev->kfd.dev); adev->kfd.dev = NULL; + amdgpu_amdkfd_total_mem_size -= adev->gmc.real_vram_size; } } @@ -756,11 +755,7 @@ void amdgpu_amdkfd_ras_poison_consumption_handler(struct amdgpu_device *adev, bo { struct ras_err_data err_data = {0, 0, 0, NULL}; - /* CPU MCA will handle page retirement if connected_to_cpu is 1 */ - if (!adev->gmc.xgmi.connected_to_cpu) - amdgpu_umc_poison_handler(adev, &err_data, reset); - else if (reset) - amdgpu_amdkfd_gpu_reset(adev); + amdgpu_umc_poison_handler(adev, &err_data, reset); } bool amdgpu_amdkfd_ras_query_utcl2_poison_status(struct amdgpu_device *adev) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v11.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v11.c index 0b0a72ca5695..7e80caa05060 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v11.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v11.c @@ -111,7 +111,7 @@ static int init_interrupts_v11(struct amdgpu_device *adev, uint32_t pipe_id) lock_srbm(adev, mec, pipe, 0, 0); - WREG32(SOC15_REG_OFFSET(GC, 0, regCPC_INT_CNTL), + WREG32_SOC15(GC, 0, regCPC_INT_CNTL, CP_INT_CNTL_RING0__TIME_STAMP_INT_ENABLE_MASK | CP_INT_CNTL_RING0__OPCODE_ERROR_INT_ENABLE_MASK); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c index 6066aebf491c..de61a85c4b02 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c @@ -1954,8 +1954,6 @@ int amdgpu_debugfs_init(struct amdgpu_device *adev) return PTR_ERR(ent); } - debugfs_create_u32("amdgpu_reset_level", 0600, root, &adev->amdgpu_reset_level_mask); - /* Register debugfs entries for amdgpu_ttm */ amdgpu_ttm_debugfs_init(adev); amdgpu_debugfs_pm_init(adev); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index c04ea7f1e819..e0445e8cc342 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -2928,6 +2928,14 @@ static int amdgpu_device_ip_suspend_phase1(struct amdgpu_device *adev) amdgpu_device_set_pg_state(adev, AMD_PG_STATE_UNGATE); amdgpu_device_set_cg_state(adev, AMD_CG_STATE_UNGATE); + /* + * Per PMFW team's suggestion, driver needs to handle gfxoff + * and df cstate features disablement for gpu reset(e.g. Mode1Reset) + * scenario. Add the missing df cstate disablement here. + */ + if (amdgpu_dpm_set_df_cstate(adev, DF_CSTATE_DISALLOW)) + dev_warn(adev->dev, "Failed to disallow df cstate"); + for (i = adev->num_ip_blocks - 1; i >= 0; i--) { if (!adev->ip_blocks[i].status.valid) continue; @@ -3154,7 +3162,8 @@ static int amdgpu_device_ip_resume_phase1(struct amdgpu_device *adev) continue; if (adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_COMMON || adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_GMC || - adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_IH) { + adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_IH || + (adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_PSP && amdgpu_sriov_vf(adev))) { r = adev->ip_blocks[i].version->funcs->resume(adev); if (r) { @@ -4069,12 +4078,20 @@ static void amdgpu_device_evict_resources(struct amdgpu_device *adev) int amdgpu_device_suspend(struct drm_device *dev, bool fbcon) { struct amdgpu_device *adev = drm_to_adev(dev); + int r = 0; if (dev->switch_power_state == DRM_SWITCH_POWER_OFF) return 0; adev->in_suspend = true; + if (amdgpu_sriov_vf(adev)) { + amdgpu_virt_fini_data_exchange(adev); + r = amdgpu_virt_request_full_gpu(adev, false); + if (r) + return r; + } + if (amdgpu_acpi_smart_shift_update(dev, AMDGPU_SS_DEV_D3)) DRM_WARN("smart shift update failed\n"); @@ -4098,6 +4115,9 @@ int amdgpu_device_suspend(struct drm_device *dev, bool fbcon) amdgpu_device_ip_suspend_phase2(adev); + if (amdgpu_sriov_vf(adev)) + amdgpu_virt_release_full_gpu(adev, false); + return 0; } @@ -4116,6 +4136,12 @@ int amdgpu_device_resume(struct drm_device *dev, bool fbcon) struct amdgpu_device *adev = drm_to_adev(dev); int r = 0; + if (amdgpu_sriov_vf(adev)) { + r = amdgpu_virt_request_full_gpu(adev, true); + if (r) + return r; + } + if (dev->switch_power_state == DRM_SWITCH_POWER_OFF) return 0; @@ -4130,6 +4156,13 @@ int amdgpu_device_resume(struct drm_device *dev, bool fbcon) } r = amdgpu_device_ip_resume(adev); + + /* no matter what r is, always need to properly release full GPU */ + if (amdgpu_sriov_vf(adev)) { + amdgpu_virt_init_data_exchange(adev); + amdgpu_virt_release_full_gpu(adev, true); + } + if (r) { dev_err(adev->dev, "amdgpu_device_ip_resume failed (%d).\n", r); return r; @@ -5185,7 +5218,6 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev, reset_context->job = job; reset_context->hive = hive; - /* * Build list of devices to reset. * In case we are in XGMI hive mode, resort the device list @@ -5312,11 +5344,8 @@ retry: /* Rest of adevs pre asic reset from XGMI hive. */ amdgpu_ras_resume(adev); } else { r = amdgpu_do_asic_reset(device_list_handle, reset_context); - if (r && r == -EAGAIN) { - set_bit(AMDGPU_SKIP_MODE2_RESET, &reset_context->flags); - adev->asic_reset_res = 0; + if (r && r == -EAGAIN) goto retry; - } if (!r && gpu_reset_for_dev_remove) goto recover_end; @@ -5576,9 +5605,9 @@ bool amdgpu_device_is_peer_accessible(struct amdgpu_device *adev, ~*peer_adev->dev->dma_mask : ~((1ULL << 32) - 1); resource_size_t aper_limit = adev->gmc.aper_base + adev->gmc.aper_size - 1; - bool p2p_access = !adev->gmc.xgmi.connected_to_cpu && - !(pci_p2pdma_distance_many(adev->pdev, - &peer_adev->dev, 1, true) < 0); + bool p2p_access = + !adev->gmc.xgmi.connected_to_cpu && + !(pci_p2pdma_distance(adev->pdev, peer_adev->dev, false) < 0); return pcie_p2p && p2p_access && (adev->gmc.visible_vram_size && adev->gmc.real_vram_size == adev->gmc.visible_vram_size && @@ -5752,7 +5781,6 @@ pci_ers_result_t amdgpu_pci_slot_reset(struct pci_dev *pdev) reset_context.reset_req_dev = adev; set_bit(AMDGPU_NEED_FULL_RESET, &reset_context.flags); set_bit(AMDGPU_SKIP_HW_RESET, &reset_context.flags); - set_bit(AMDGPU_SKIP_MODE2_RESET, &reset_context.flags); adev->no_hw_access = true; r = amdgpu_device_pre_asic_reset(adev, &reset_context); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c index 9fa2a5ceb77d..3993e6134914 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c @@ -229,7 +229,7 @@ static int amdgpu_discovery_read_binary_from_file(struct amdgpu_device *adev, ui return r; } - memcpy((u8 *)binary, (u8 *)fw->data, adev->mman.discovery_tmr_size); + memcpy((u8 *)binary, (u8 *)fw->data, fw->size); release_firmware(fw); return 0; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c index 23998f727c7f..1a06b8d724f3 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c @@ -38,8 +38,6 @@ #include <linux/pci.h> #include <linux/pm_runtime.h> #include <drm/drm_crtc_helper.h> -#include <drm/drm_damage_helper.h> -#include <drm/drm_drv.h> #include <drm/drm_edid.h> #include <drm/drm_gem_framebuffer_helper.h> #include <drm/drm_fb_helper.h> @@ -500,12 +498,6 @@ static const struct drm_framebuffer_funcs amdgpu_fb_funcs = { .create_handle = drm_gem_fb_create_handle, }; -static const struct drm_framebuffer_funcs amdgpu_fb_funcs_atomic = { - .destroy = drm_gem_fb_destroy, - .create_handle = drm_gem_fb_create_handle, - .dirty = drm_atomic_helper_dirtyfb, -}; - uint32_t amdgpu_display_supported_domains(struct amdgpu_device *adev, uint64_t bo_flags) { @@ -1108,10 +1100,8 @@ static int amdgpu_display_gem_fb_verify_and_init(struct drm_device *dev, if (ret) goto err; - if (drm_drv_uses_atomic_modeset(dev)) - ret = drm_framebuffer_init(dev, &rfb->base, &amdgpu_fb_funcs_atomic); - else - ret = drm_framebuffer_init(dev, &rfb->base, &amdgpu_fb_funcs); + ret = drm_framebuffer_init(dev, &rfb->base, &amdgpu_fb_funcs); + if (ret) goto err; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c index 782cbca37538..7bd8e33b14be 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c @@ -58,7 +58,7 @@ static int amdgpu_dma_buf_attach(struct dma_buf *dmabuf, struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev); int r; - if (pci_p2pdma_distance_many(adev->pdev, &attach->dev, 1, true) < 0) + if (pci_p2pdma_distance(adev->pdev, attach->dev, false) < 0) attach->peer2peer = false; r = pm_runtime_get_sync(adev_to_drm(adev)->dev); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c index 16f6a313335e..3c9fecdd6b2f 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c @@ -38,6 +38,8 @@ #include <linux/mmu_notifier.h> #include <linux/suspend.h> #include <linux/cc_platform.h> +#include <linux/fb.h> +#include <linux/dynamic_debug.h> #include "amdgpu.h" #include "amdgpu_irq.h" @@ -186,6 +188,18 @@ int amdgpu_vcnfw_log; static void amdgpu_drv_delayed_reset_work_handler(struct work_struct *work); +DECLARE_DYNDBG_CLASSMAP(drm_debug_classes, DD_CLASS_TYPE_DISJOINT_BITS, 0, + "DRM_UT_CORE", + "DRM_UT_DRIVER", + "DRM_UT_KMS", + "DRM_UT_PRIME", + "DRM_UT_ATOMIC", + "DRM_UT_VBL", + "DRM_UT_STATE", + "DRM_UT_LEASE", + "DRM_UT_DP", + "DRM_UT_DRMRES"); + struct amdgpu_mgpu_info mgpu_info = { .mutex = __MUTEX_INITIALIZER(mgpu_info.mutex), .delayed_reset_work = __DELAYED_WORK_INITIALIZER( diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c index 8adeb7469f1e..d0d99ed607dd 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c @@ -400,7 +400,6 @@ unsigned amdgpu_fence_count_emitted(struct amdgpu_ring *ring) /* We are not protected by ring lock when reading the last sequence * but it's ok to report slightly wrong fence count here. */ - amdgpu_fence_process(ring); emitted = 0x100000000ull; emitted -= atomic_read(&ring->fence_drv.last_seq); emitted += READ_ONCE(ring->fence_drv.sync_seq); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c index ceb91469958a..9546adc8a76f 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c @@ -23,6 +23,7 @@ * */ +#include <linux/firmware.h> #include "amdgpu.h" #include "amdgpu_gfx.h" #include "amdgpu_rlc.h" @@ -865,3 +866,142 @@ int amdgpu_gfx_get_num_kcq(struct amdgpu_device *adev) } return amdgpu_num_kcq; } + +void amdgpu_gfx_cp_init_microcode(struct amdgpu_device *adev, + uint32_t ucode_id) +{ + const struct gfx_firmware_header_v1_0 *cp_hdr; + const struct gfx_firmware_header_v2_0 *cp_hdr_v2_0; + struct amdgpu_firmware_info *info = NULL; + const struct firmware *ucode_fw; + unsigned int fw_size; + + switch (ucode_id) { + case AMDGPU_UCODE_ID_CP_PFP: + cp_hdr = (const struct gfx_firmware_header_v1_0 *) + adev->gfx.pfp_fw->data; + adev->gfx.pfp_fw_version = + le32_to_cpu(cp_hdr->header.ucode_version); + adev->gfx.pfp_feature_version = + le32_to_cpu(cp_hdr->ucode_feature_version); + ucode_fw = adev->gfx.pfp_fw; + fw_size = le32_to_cpu(cp_hdr->header.ucode_size_bytes); + break; + case AMDGPU_UCODE_ID_CP_RS64_PFP: + cp_hdr_v2_0 = (const struct gfx_firmware_header_v2_0 *) + adev->gfx.pfp_fw->data; + adev->gfx.pfp_fw_version = + le32_to_cpu(cp_hdr_v2_0->header.ucode_version); + adev->gfx.pfp_feature_version = + le32_to_cpu(cp_hdr_v2_0->ucode_feature_version); + ucode_fw = adev->gfx.pfp_fw; + fw_size = le32_to_cpu(cp_hdr_v2_0->ucode_size_bytes); + break; + case AMDGPU_UCODE_ID_CP_RS64_PFP_P0_STACK: + case AMDGPU_UCODE_ID_CP_RS64_PFP_P1_STACK: + cp_hdr_v2_0 = (const struct gfx_firmware_header_v2_0 *) + adev->gfx.pfp_fw->data; + ucode_fw = adev->gfx.pfp_fw; + fw_size = le32_to_cpu(cp_hdr_v2_0->data_size_bytes); + break; + case AMDGPU_UCODE_ID_CP_ME: + cp_hdr = (const struct gfx_firmware_header_v1_0 *) + adev->gfx.me_fw->data; + adev->gfx.me_fw_version = + le32_to_cpu(cp_hdr->header.ucode_version); + adev->gfx.me_feature_version = + le32_to_cpu(cp_hdr->ucode_feature_version); + ucode_fw = adev->gfx.me_fw; + fw_size = le32_to_cpu(cp_hdr->header.ucode_size_bytes); + break; + case AMDGPU_UCODE_ID_CP_RS64_ME: + cp_hdr_v2_0 = (const struct gfx_firmware_header_v2_0 *) + adev->gfx.me_fw->data; + adev->gfx.me_fw_version = + le32_to_cpu(cp_hdr_v2_0->header.ucode_version); + adev->gfx.me_feature_version = + le32_to_cpu(cp_hdr_v2_0->ucode_feature_version); + ucode_fw = adev->gfx.me_fw; + fw_size = le32_to_cpu(cp_hdr_v2_0->ucode_size_bytes); + break; + case AMDGPU_UCODE_ID_CP_RS64_ME_P0_STACK: + case AMDGPU_UCODE_ID_CP_RS64_ME_P1_STACK: + cp_hdr_v2_0 = (const struct gfx_firmware_header_v2_0 *) + adev->gfx.me_fw->data; + ucode_fw = adev->gfx.me_fw; + fw_size = le32_to_cpu(cp_hdr_v2_0->data_size_bytes); + break; + case AMDGPU_UCODE_ID_CP_CE: + cp_hdr = (const struct gfx_firmware_header_v1_0 *) + adev->gfx.ce_fw->data; + adev->gfx.ce_fw_version = + le32_to_cpu(cp_hdr->header.ucode_version); + adev->gfx.ce_feature_version = + le32_to_cpu(cp_hdr->ucode_feature_version); + ucode_fw = adev->gfx.ce_fw; + fw_size = le32_to_cpu(cp_hdr->header.ucode_size_bytes); + break; + case AMDGPU_UCODE_ID_CP_MEC1: + cp_hdr = (const struct gfx_firmware_header_v1_0 *) + adev->gfx.mec_fw->data; + adev->gfx.mec_fw_version = + le32_to_cpu(cp_hdr->header.ucode_version); + adev->gfx.mec_feature_version = + le32_to_cpu(cp_hdr->ucode_feature_version); + ucode_fw = adev->gfx.mec_fw; + fw_size = le32_to_cpu(cp_hdr->header.ucode_size_bytes) - + le32_to_cpu(cp_hdr->jt_size) * 4; + break; + case AMDGPU_UCODE_ID_CP_MEC1_JT: + cp_hdr = (const struct gfx_firmware_header_v1_0 *) + adev->gfx.mec_fw->data; + ucode_fw = adev->gfx.mec_fw; + fw_size = le32_to_cpu(cp_hdr->jt_size) * 4; + break; + case AMDGPU_UCODE_ID_CP_MEC2: + cp_hdr = (const struct gfx_firmware_header_v1_0 *) + adev->gfx.mec2_fw->data; + adev->gfx.mec2_fw_version = + le32_to_cpu(cp_hdr->header.ucode_version); + adev->gfx.mec2_feature_version = + le32_to_cpu(cp_hdr->ucode_feature_version); + ucode_fw = adev->gfx.mec2_fw; + fw_size = le32_to_cpu(cp_hdr->header.ucode_size_bytes) - + le32_to_cpu(cp_hdr->jt_size) * 4; + break; + case AMDGPU_UCODE_ID_CP_MEC2_JT: + cp_hdr = (const struct gfx_firmware_header_v1_0 *) + adev->gfx.mec2_fw->data; + ucode_fw = adev->gfx.mec2_fw; + fw_size = le32_to_cpu(cp_hdr->jt_size) * 4; + break; + case AMDGPU_UCODE_ID_CP_RS64_MEC: + cp_hdr_v2_0 = (const struct gfx_firmware_header_v2_0 *) + adev->gfx.mec_fw->data; + adev->gfx.mec_fw_version = + le32_to_cpu(cp_hdr_v2_0->header.ucode_version); + adev->gfx.mec_feature_version = + le32_to_cpu(cp_hdr_v2_0->ucode_feature_version); + ucode_fw = adev->gfx.mec_fw; + fw_size = le32_to_cpu(cp_hdr_v2_0->ucode_size_bytes); + break; + case AMDGPU_UCODE_ID_CP_RS64_MEC_P0_STACK: + case AMDGPU_UCODE_ID_CP_RS64_MEC_P1_STACK: + case AMDGPU_UCODE_ID_CP_RS64_MEC_P2_STACK: + case AMDGPU_UCODE_ID_CP_RS64_MEC_P3_STACK: + cp_hdr_v2_0 = (const struct gfx_firmware_header_v2_0 *) + adev->gfx.mec_fw->data; + ucode_fw = adev->gfx.mec_fw; + fw_size = le32_to_cpu(cp_hdr_v2_0->data_size_bytes); + break; + default: + break; + } + + if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) { + info = &adev->firmware.ucode[ucode_id]; + info->ucode_id = ucode_id; + info->fw = ucode_fw; + adev->firmware.fw_size += ALIGN(fw_size, PAGE_SIZE); + } +} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h index b91ab919ee70..832b3807f1d6 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h @@ -426,4 +426,6 @@ int amdgpu_gfx_cp_ecc_error_irq(struct amdgpu_device *adev, uint32_t amdgpu_kiq_rreg(struct amdgpu_device *adev, uint32_t reg); void amdgpu_kiq_wreg(struct amdgpu_device *adev, uint32_t reg, uint32_t v); int amdgpu_gfx_get_num_kcq(struct amdgpu_device *adev); +void amdgpu_gfx_cp_init_microcode(struct amdgpu_device *adev, uint32_t ucode_id); + #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c index aebc384531ac..34233a74248c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c @@ -572,45 +572,15 @@ void amdgpu_gmc_tmz_set(struct amdgpu_device *adev) void amdgpu_gmc_noretry_set(struct amdgpu_device *adev) { struct amdgpu_gmc *gmc = &adev->gmc; - - switch (adev->ip_versions[GC_HWIP][0]) { - case IP_VERSION(9, 0, 1): - case IP_VERSION(9, 3, 0): - case IP_VERSION(9, 4, 0): - case IP_VERSION(9, 4, 1): - case IP_VERSION(9, 4, 2): - case IP_VERSION(10, 3, 3): - case IP_VERSION(10, 3, 4): - case IP_VERSION(10, 3, 5): - case IP_VERSION(10, 3, 6): - case IP_VERSION(10, 3, 7): - /* - * noretry = 0 will cause kfd page fault tests fail - * for some ASICs, so set default to 1 for these ASICs. - */ - if (amdgpu_noretry == -1) - gmc->noretry = 1; - else - gmc->noretry = amdgpu_noretry; - break; - default: - /* Raven currently has issues with noretry - * regardless of what we decide for other - * asics, we should leave raven with - * noretry = 0 until we root cause the - * issues. - * - * default this to 0 for now, but we may want - * to change this in the future for certain - * GPUs as it can increase performance in - * certain cases. - */ - if (amdgpu_noretry == -1) - gmc->noretry = 0; - else - gmc->noretry = amdgpu_noretry; - break; - } + uint32_t gc_ver = adev->ip_versions[GC_HWIP][0]; + bool noretry_default = (gc_ver == IP_VERSION(9, 0, 1) || + gc_ver == IP_VERSION(9, 3, 0) || + gc_ver == IP_VERSION(9, 4, 0) || + gc_ver == IP_VERSION(9, 4, 1) || + gc_ver == IP_VERSION(9, 4, 2) || + gc_ver >= IP_VERSION(10, 3, 0)); + + gmc->noretry = (amdgpu_noretry == -1) ? noretry_default : amdgpu_noretry; } void amdgpu_gmc_set_vm_fault_masks(struct amdgpu_device *adev, int hub_type, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gtt_mgr.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gtt_mgr.c index 1f3302aebeff..44367f03316f 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gtt_mgr.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gtt_mgr.c @@ -144,7 +144,7 @@ static int amdgpu_gtt_mgr_new(struct ttm_resource_manager *man, node->base.start = node->mm_nodes[0].start; } else { node->mm_nodes[0].start = 0; - node->mm_nodes[0].size = node->base.num_pages; + node->mm_nodes[0].size = PFN_UP(node->base.size); node->base.start = AMDGPU_BO_INVALID_OFFSET; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c index 46c99331d7f1..cd968e781077 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c @@ -72,7 +72,6 @@ static enum drm_gpu_sched_stat amdgpu_job_timedout(struct drm_sched_job *s_job) reset_context.method = AMD_RESET_METHOD_NONE; reset_context.reset_req_dev = adev; clear_bit(AMDGPU_NEED_FULL_RESET, &reset_context.flags); - clear_bit(AMDGPU_SKIP_MODE2_RESET, &reset_context.flags); r = amdgpu_device_gpu_recover(ring->adev, job, &reset_context); if (r) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.h index 7b46f6bf4187..ad980f4b66e1 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.h @@ -222,6 +222,8 @@ struct mes_add_queue_input { uint64_t tba_addr; uint64_t tma_addr; uint32_t is_kfd_process; + uint32_t is_aql_queue; + uint32_t queue_size; }; struct mes_remove_queue_input { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c index e6a9b9fc9e0b..974e85d8b6cc 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c @@ -542,6 +542,7 @@ int amdgpu_bo_create(struct amdgpu_device *adev, /* GWS and OA don't need any alignment. */ page_align = bp->byte_align; size <<= PAGE_SHIFT; + } else if (bp->domain & AMDGPU_GEM_DOMAIN_GDS) { /* Both size and alignment must be a multiple of 4. */ page_align = ALIGN(bp->byte_align, 4); @@ -688,13 +689,16 @@ int amdgpu_bo_create_vm(struct amdgpu_device *adev, * num of amdgpu_vm_pt entries. */ BUG_ON(bp->bo_ptr_size < sizeof(struct amdgpu_bo_vm)); - bp->destroy = &amdgpu_bo_vm_destroy; r = amdgpu_bo_create(adev, bp, &bo_ptr); if (r) return r; *vmbo_ptr = to_amdgpu_bo_vm(bo_ptr); INIT_LIST_HEAD(&(*vmbo_ptr)->shadow_list); + /* Set destroy callback to amdgpu_bo_vm_destroy after vmbo->shadow_list + * is initialized. + */ + bo_ptr->tbo.destroy = &amdgpu_bo_vm_destroy; return r; } @@ -773,7 +777,7 @@ int amdgpu_bo_kmap(struct amdgpu_bo *bo, void **ptr) return 0; } - r = ttm_bo_kmap(&bo->tbo, 0, bo->tbo.resource->num_pages, &bo->kmap); + r = ttm_bo_kmap(&bo->tbo, 0, PFN_UP(bo->tbo.base.size), &bo->kmap); if (r) return r; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c index e55f106621ef..a4b47e1bd111 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c @@ -1950,7 +1950,6 @@ static void amdgpu_ras_do_recovery(struct work_struct *work) reset_context.method = AMD_RESET_METHOD_NONE; reset_context.reset_req_dev = adev; clear_bit(AMDGPU_NEED_FULL_RESET, &reset_context.flags); - clear_bit(AMDGPU_SKIP_MODE2_RESET, &reset_context.flags); amdgpu_device_gpu_recover(ras->adev, NULL, &reset_context); } @@ -2268,6 +2267,25 @@ static int amdgpu_ras_recovery_fini(struct amdgpu_device *adev) static bool amdgpu_ras_asic_supported(struct amdgpu_device *adev) { + if (amdgpu_sriov_vf(adev)) { + switch (adev->ip_versions[MP0_HWIP][0]) { + case IP_VERSION(13, 0, 2): + return true; + default: + return false; + } + } + + if (adev->asic_type == CHIP_IP_DISCOVERY) { + switch (adev->ip_versions[MP0_HWIP][0]) { + case IP_VERSION(13, 0, 0): + case IP_VERSION(13, 0, 10): + return true; + default: + return false; + } + } + return adev->asic_type == CHIP_VEGA10 || adev->asic_type == CHIP_VEGA20 || adev->asic_type == CHIP_ARCTURUS || @@ -2311,11 +2329,6 @@ static void amdgpu_ras_check_supported(struct amdgpu_device *adev) !amdgpu_ras_asic_supported(adev)) return; - /* If driver run on sriov guest side, only enable ras for aldebaran */ - if (amdgpu_sriov_vf(adev) && - adev->ip_versions[MP1_HWIP][0] != IP_VERSION(13, 0, 2)) - return; - if (!adev->gmc.xgmi.connected_to_cpu) { if (amdgpu_atomfirmware_mem_ecc_supported(adev)) { dev_info(adev->dev, "MEM ECC is active.\n"); @@ -2720,7 +2733,8 @@ int amdgpu_ras_pre_fini(struct amdgpu_device *adev) /* Need disable ras on all IPs here before ip [hw/sw]fini */ - amdgpu_ras_disable_all_features(adev, 0); + if (con->features) + amdgpu_ras_disable_all_features(adev, 0); amdgpu_ras_recovery_fini(adev); return 0; } @@ -2833,11 +2847,8 @@ static int amdgpu_bad_page_notifier(struct notifier_block *nb, struct mce *m = (struct mce *)data; struct amdgpu_device *adev = NULL; uint32_t gpu_id = 0; - uint32_t umc_inst = 0; - uint32_t ch_inst, channel_index = 0; + uint32_t umc_inst = 0, ch_inst = 0; struct ras_err_data err_data = {0, 0, 0, NULL}; - struct eeprom_table_record err_rec; - uint64_t retired_page; /* * If the error was generated in UMC_V2, which belongs to GPU UMCs, @@ -2876,21 +2887,22 @@ static int amdgpu_bad_page_notifier(struct notifier_block *nb, dev_info(adev->dev, "Uncorrectable error detected in UMC inst: %d, chan_idx: %d", umc_inst, ch_inst); + err_data.err_addr = + kcalloc(adev->umc.max_ras_err_cnt_per_query, + sizeof(struct eeprom_table_record), GFP_KERNEL); + if (!err_data.err_addr) { + dev_warn(adev->dev, + "Failed to alloc memory for umc error record in mca notifier!\n"); + return NOTIFY_DONE; + } + /* * Translate UMC channel address to Physical address */ - channel_index = - adev->umc.channel_idx_tbl[umc_inst * adev->umc.channel_inst_num - + ch_inst]; - - retired_page = ADDR_OF_8KB_BLOCK(m->addr) | - ADDR_OF_256B_BLOCK(channel_index) | - OFFSET_IN_256B_BLOCK(m->addr); - - memset(&err_rec, 0x0, sizeof(struct eeprom_table_record)); - err_data.err_addr = &err_rec; - amdgpu_umc_fill_error_record(&err_data, m->addr, - retired_page, channel_index, umc_inst); + if (adev->umc.ras && + adev->umc.ras->convert_ras_error_address) + adev->umc.ras->convert_ras_error_address(adev, + &err_data, m->addr, ch_inst, umc_inst); if (amdgpu_bad_page_threshold != 0) { amdgpu_ras_add_bad_pages(adev, err_data.err_addr, @@ -2898,6 +2910,7 @@ static int amdgpu_bad_page_notifier(struct notifier_block *nb, amdgpu_ras_save_bad_pages(adev); } + kfree(err_data.err_addr); return NOTIFY_OK; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_res_cursor.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_res_cursor.h index 6546552e596c..5c4f93ee0c57 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_res_cursor.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_res_cursor.h @@ -62,7 +62,7 @@ static inline void amdgpu_res_first(struct ttm_resource *res, if (!res) goto fallback; - BUG_ON(start + size > res->num_pages << PAGE_SHIFT); + BUG_ON(start + size > res->size); cur->mem_type = res->mem_type; @@ -110,7 +110,7 @@ fallback: cur->size = size; cur->remaining = size; cur->node = NULL; - WARN_ON(res && start + size > res->num_pages << PAGE_SHIFT); + WARN_ON(res && start + size > res->size); return; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_reset.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_reset.c index 9da5ead50c90..f778466bb9db 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_reset.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_reset.c @@ -37,8 +37,6 @@ int amdgpu_reset_init(struct amdgpu_device *adev) { int ret = 0; - adev->amdgpu_reset_level_mask = 0x1; - switch (adev->ip_versions[MP1_HWIP][0]) { case IP_VERSION(13, 0, 2): ret = aldebaran_reset_init(adev); @@ -76,12 +74,6 @@ int amdgpu_reset_prepare_hwcontext(struct amdgpu_device *adev, { struct amdgpu_reset_handler *reset_handler = NULL; - if (!(adev->amdgpu_reset_level_mask & AMDGPU_RESET_LEVEL_MODE2)) - return -ENOSYS; - - if (test_bit(AMDGPU_SKIP_MODE2_RESET, &reset_context->flags)) - return -ENOSYS; - if (adev->reset_cntl && adev->reset_cntl->get_reset_handler) reset_handler = adev->reset_cntl->get_reset_handler( adev->reset_cntl, reset_context); @@ -98,12 +90,6 @@ int amdgpu_reset_perform_reset(struct amdgpu_device *adev, int ret; struct amdgpu_reset_handler *reset_handler = NULL; - if (!(adev->amdgpu_reset_level_mask & AMDGPU_RESET_LEVEL_MODE2)) - return -ENOSYS; - - if (test_bit(AMDGPU_SKIP_MODE2_RESET, &reset_context->flags)) - return -ENOSYS; - if (adev->reset_cntl) reset_handler = adev->reset_cntl->get_reset_handler( adev->reset_cntl, reset_context); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_reset.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_reset.h index dc43fcb93eac..f4a501ff87d9 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_reset.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_reset.h @@ -30,8 +30,7 @@ enum AMDGPU_RESET_FLAGS { AMDGPU_NEED_FULL_RESET = 0, AMDGPU_SKIP_HW_RESET = 1, - AMDGPU_SKIP_MODE2_RESET = 2, - AMDGPU_RESET_FOR_DEVICE_REMOVE = 3, + AMDGPU_RESET_FOR_DEVICE_REMOVE = 2, }; struct amdgpu_reset_context { @@ -113,7 +112,8 @@ static inline bool amdgpu_reset_get_reset_domain(struct amdgpu_reset_domain *dom static inline void amdgpu_reset_put_reset_domain(struct amdgpu_reset_domain *domain) { - kref_put(&domain->refcount, amdgpu_reset_destroy_reset_domain); + if (domain) + kref_put(&domain->refcount, amdgpu_reset_destroy_reset_domain); } static inline bool amdgpu_reset_domain_schedule(struct amdgpu_reset_domain *domain, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c index 3e316b013fd9..d3558c34d406 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c @@ -405,9 +405,6 @@ bool amdgpu_ring_soft_recovery(struct amdgpu_ring *ring, unsigned int vmid, { ktime_t deadline = ktime_add_us(ktime_get(), 10000); - if (!(ring->adev->amdgpu_reset_level_mask & AMDGPU_RESET_LEVEL_SOFT_RECOVERY)) - return false; - if (amdgpu_sriov_vf(ring->adev) || !ring->funcs->soft_recovery || !fence) return false; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_rlc.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_rlc.c index 6373bfb47d55..012b72d00e04 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_rlc.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_rlc.c @@ -272,3 +272,275 @@ void amdgpu_gfx_rlc_fini(struct amdgpu_device *adev) &adev->gfx.rlc.cp_table_gpu_addr, (void **)&adev->gfx.rlc.cp_table_ptr); } + +static int amdgpu_gfx_rlc_init_microcode_v2_0(struct amdgpu_device *adev) +{ + const struct common_firmware_header *common_hdr; + const struct rlc_firmware_header_v2_0 *rlc_hdr; + struct amdgpu_firmware_info *info; + unsigned int *tmp; + unsigned int i; + + rlc_hdr = (const struct rlc_firmware_header_v2_0 *)adev->gfx.rlc_fw->data; + + adev->gfx.rlc_fw_version = le32_to_cpu(rlc_hdr->header.ucode_version); + adev->gfx.rlc_feature_version = le32_to_cpu(rlc_hdr->ucode_feature_version); + adev->gfx.rlc.save_and_restore_offset = + le32_to_cpu(rlc_hdr->save_and_restore_offset); + adev->gfx.rlc.clear_state_descriptor_offset = + le32_to_cpu(rlc_hdr->clear_state_descriptor_offset); + adev->gfx.rlc.avail_scratch_ram_locations = + le32_to_cpu(rlc_hdr->avail_scratch_ram_locations); + adev->gfx.rlc.reg_restore_list_size = + le32_to_cpu(rlc_hdr->reg_restore_list_size); + adev->gfx.rlc.reg_list_format_start = + le32_to_cpu(rlc_hdr->reg_list_format_start); + adev->gfx.rlc.reg_list_format_separate_start = + le32_to_cpu(rlc_hdr->reg_list_format_separate_start); + adev->gfx.rlc.starting_offsets_start = + le32_to_cpu(rlc_hdr->starting_offsets_start); + adev->gfx.rlc.reg_list_format_size_bytes = + le32_to_cpu(rlc_hdr->reg_list_format_size_bytes); + adev->gfx.rlc.reg_list_size_bytes = + le32_to_cpu(rlc_hdr->reg_list_size_bytes); + adev->gfx.rlc.register_list_format = + kmalloc(adev->gfx.rlc.reg_list_format_size_bytes + + adev->gfx.rlc.reg_list_size_bytes, GFP_KERNEL); + if (!adev->gfx.rlc.register_list_format) { + dev_err(adev->dev, "failed to allocate memory for rlc register_list_format\n"); + return -ENOMEM; + } + + tmp = (unsigned int *)((uintptr_t)rlc_hdr + + le32_to_cpu(rlc_hdr->reg_list_format_array_offset_bytes)); + for (i = 0 ; i < (rlc_hdr->reg_list_format_size_bytes >> 2); i++) + adev->gfx.rlc.register_list_format[i] = le32_to_cpu(tmp[i]); + + adev->gfx.rlc.register_restore = adev->gfx.rlc.register_list_format + i; + + tmp = (unsigned int *)((uintptr_t)rlc_hdr + + le32_to_cpu(rlc_hdr->reg_list_array_offset_bytes)); + for (i = 0 ; i < (rlc_hdr->reg_list_size_bytes >> 2); i++) + adev->gfx.rlc.register_restore[i] = le32_to_cpu(tmp[i]); + + if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) { + info = &adev->firmware.ucode[AMDGPU_UCODE_ID_RLC_G]; + info->ucode_id = AMDGPU_UCODE_ID_RLC_G; + info->fw = adev->gfx.rlc_fw; + if (info->fw) { + common_hdr = (const struct common_firmware_header *)info->fw->data; + adev->firmware.fw_size += + ALIGN(le32_to_cpu(common_hdr->ucode_size_bytes), PAGE_SIZE); + } + } + + return 0; +} + +static void amdgpu_gfx_rlc_init_microcode_v2_1(struct amdgpu_device *adev) +{ + const struct rlc_firmware_header_v2_1 *rlc_hdr; + struct amdgpu_firmware_info *info; + + rlc_hdr = (const struct rlc_firmware_header_v2_1 *)adev->gfx.rlc_fw->data; + adev->gfx.rlc_srlc_fw_version = le32_to_cpu(rlc_hdr->save_restore_list_cntl_ucode_ver); + adev->gfx.rlc_srlc_feature_version = le32_to_cpu(rlc_hdr->save_restore_list_cntl_feature_ver); + adev->gfx.rlc.save_restore_list_cntl_size_bytes = le32_to_cpu(rlc_hdr->save_restore_list_cntl_size_bytes); + adev->gfx.rlc.save_restore_list_cntl = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->save_restore_list_cntl_offset_bytes); + adev->gfx.rlc_srlg_fw_version = le32_to_cpu(rlc_hdr->save_restore_list_gpm_ucode_ver); + adev->gfx.rlc_srlg_feature_version = le32_to_cpu(rlc_hdr->save_restore_list_gpm_feature_ver); + adev->gfx.rlc.save_restore_list_gpm_size_bytes = le32_to_cpu(rlc_hdr->save_restore_list_gpm_size_bytes); + adev->gfx.rlc.save_restore_list_gpm = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->save_restore_list_gpm_offset_bytes); + adev->gfx.rlc_srls_fw_version = le32_to_cpu(rlc_hdr->save_restore_list_srm_ucode_ver); + adev->gfx.rlc_srls_feature_version = le32_to_cpu(rlc_hdr->save_restore_list_srm_feature_ver); + adev->gfx.rlc.save_restore_list_srm_size_bytes = le32_to_cpu(rlc_hdr->save_restore_list_srm_size_bytes); + adev->gfx.rlc.save_restore_list_srm = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->save_restore_list_srm_offset_bytes); + adev->gfx.rlc.reg_list_format_direct_reg_list_length = + le32_to_cpu(rlc_hdr->reg_list_format_direct_reg_list_length); + + if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) { + if (adev->gfx.rlc.save_restore_list_cntl_size_bytes) { + info = &adev->firmware.ucode[AMDGPU_UCODE_ID_RLC_RESTORE_LIST_CNTL]; + info->ucode_id = AMDGPU_UCODE_ID_RLC_RESTORE_LIST_CNTL; + info->fw = adev->gfx.rlc_fw; + adev->firmware.fw_size += + ALIGN(adev->gfx.rlc.save_restore_list_cntl_size_bytes, PAGE_SIZE); + } + + if (adev->gfx.rlc.save_restore_list_gpm_size_bytes) { + info = &adev->firmware.ucode[AMDGPU_UCODE_ID_RLC_RESTORE_LIST_GPM_MEM]; + info->ucode_id = AMDGPU_UCODE_ID_RLC_RESTORE_LIST_GPM_MEM; + info->fw = adev->gfx.rlc_fw; + adev->firmware.fw_size += + ALIGN(adev->gfx.rlc.save_restore_list_gpm_size_bytes, PAGE_SIZE); + } + + if (adev->gfx.rlc.save_restore_list_srm_size_bytes) { + info = &adev->firmware.ucode[AMDGPU_UCODE_ID_RLC_RESTORE_LIST_SRM_MEM]; + info->ucode_id = AMDGPU_UCODE_ID_RLC_RESTORE_LIST_SRM_MEM; + info->fw = adev->gfx.rlc_fw; + adev->firmware.fw_size += + ALIGN(adev->gfx.rlc.save_restore_list_srm_size_bytes, PAGE_SIZE); + } + } +} + +static void amdgpu_gfx_rlc_init_microcode_v2_2(struct amdgpu_device *adev) +{ + const struct rlc_firmware_header_v2_2 *rlc_hdr; + struct amdgpu_firmware_info *info; + + rlc_hdr = (const struct rlc_firmware_header_v2_2 *)adev->gfx.rlc_fw->data; + adev->gfx.rlc.rlc_iram_ucode_size_bytes = le32_to_cpu(rlc_hdr->rlc_iram_ucode_size_bytes); + adev->gfx.rlc.rlc_iram_ucode = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->rlc_iram_ucode_offset_bytes); + adev->gfx.rlc.rlc_dram_ucode_size_bytes = le32_to_cpu(rlc_hdr->rlc_dram_ucode_size_bytes); + adev->gfx.rlc.rlc_dram_ucode = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->rlc_dram_ucode_offset_bytes); + + if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) { + if (adev->gfx.rlc.rlc_iram_ucode_size_bytes) { + info = &adev->firmware.ucode[AMDGPU_UCODE_ID_RLC_IRAM]; + info->ucode_id = AMDGPU_UCODE_ID_RLC_IRAM; + info->fw = adev->gfx.rlc_fw; + adev->firmware.fw_size += + ALIGN(adev->gfx.rlc.rlc_iram_ucode_size_bytes, PAGE_SIZE); + } + + if (adev->gfx.rlc.rlc_dram_ucode_size_bytes) { + info = &adev->firmware.ucode[AMDGPU_UCODE_ID_RLC_DRAM]; + info->ucode_id = AMDGPU_UCODE_ID_RLC_DRAM; + info->fw = adev->gfx.rlc_fw; + adev->firmware.fw_size += + ALIGN(adev->gfx.rlc.rlc_dram_ucode_size_bytes, PAGE_SIZE); + } + } +} + +static void amdgpu_gfx_rlc_init_microcode_v2_3(struct amdgpu_device *adev) +{ + const struct rlc_firmware_header_v2_3 *rlc_hdr; + struct amdgpu_firmware_info *info; + + rlc_hdr = (const struct rlc_firmware_header_v2_3 *)adev->gfx.rlc_fw->data; + adev->gfx.rlcp_ucode_version = le32_to_cpu(rlc_hdr->rlcp_ucode_version); + adev->gfx.rlcp_ucode_feature_version = le32_to_cpu(rlc_hdr->rlcp_ucode_feature_version); + adev->gfx.rlc.rlcp_ucode_size_bytes = le32_to_cpu(rlc_hdr->rlcp_ucode_size_bytes); + adev->gfx.rlc.rlcp_ucode = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->rlcp_ucode_offset_bytes); + + adev->gfx.rlcv_ucode_version = le32_to_cpu(rlc_hdr->rlcv_ucode_version); + adev->gfx.rlcv_ucode_feature_version = le32_to_cpu(rlc_hdr->rlcv_ucode_feature_version); + adev->gfx.rlc.rlcv_ucode_size_bytes = le32_to_cpu(rlc_hdr->rlcv_ucode_size_bytes); + adev->gfx.rlc.rlcv_ucode = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->rlcv_ucode_offset_bytes); + + if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) { + if (adev->gfx.rlc.rlcp_ucode_size_bytes) { + info = &adev->firmware.ucode[AMDGPU_UCODE_ID_RLC_P]; + info->ucode_id = AMDGPU_UCODE_ID_RLC_P; + info->fw = adev->gfx.rlc_fw; + adev->firmware.fw_size += + ALIGN(adev->gfx.rlc.rlcp_ucode_size_bytes, PAGE_SIZE); + } + + if (adev->gfx.rlc.rlcv_ucode_size_bytes) { + info = &adev->firmware.ucode[AMDGPU_UCODE_ID_RLC_V]; + info->ucode_id = AMDGPU_UCODE_ID_RLC_V; + info->fw = adev->gfx.rlc_fw; + adev->firmware.fw_size += + ALIGN(adev->gfx.rlc.rlcv_ucode_size_bytes, PAGE_SIZE); + } + } +} + +static void amdgpu_gfx_rlc_init_microcode_v2_4(struct amdgpu_device *adev) +{ + const struct rlc_firmware_header_v2_4 *rlc_hdr; + struct amdgpu_firmware_info *info; + + rlc_hdr = (const struct rlc_firmware_header_v2_4 *)adev->gfx.rlc_fw->data; + adev->gfx.rlc.global_tap_delays_ucode_size_bytes = le32_to_cpu(rlc_hdr->global_tap_delays_ucode_size_bytes); + adev->gfx.rlc.global_tap_delays_ucode = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->global_tap_delays_ucode_offset_bytes); + adev->gfx.rlc.se0_tap_delays_ucode_size_bytes = le32_to_cpu(rlc_hdr->se0_tap_delays_ucode_size_bytes); + adev->gfx.rlc.se0_tap_delays_ucode = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->se0_tap_delays_ucode_offset_bytes); + adev->gfx.rlc.se1_tap_delays_ucode_size_bytes = le32_to_cpu(rlc_hdr->se1_tap_delays_ucode_size_bytes); + adev->gfx.rlc.se1_tap_delays_ucode = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->se1_tap_delays_ucode_offset_bytes); + adev->gfx.rlc.se2_tap_delays_ucode_size_bytes = le32_to_cpu(rlc_hdr->se2_tap_delays_ucode_size_bytes); + adev->gfx.rlc.se2_tap_delays_ucode = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->se2_tap_delays_ucode_offset_bytes); + adev->gfx.rlc.se3_tap_delays_ucode_size_bytes = le32_to_cpu(rlc_hdr->se3_tap_delays_ucode_size_bytes); + adev->gfx.rlc.se3_tap_delays_ucode = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->se3_tap_delays_ucode_offset_bytes); + + if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) { + if (adev->gfx.rlc.global_tap_delays_ucode_size_bytes) { + info = &adev->firmware.ucode[AMDGPU_UCODE_ID_GLOBAL_TAP_DELAYS]; + info->ucode_id = AMDGPU_UCODE_ID_GLOBAL_TAP_DELAYS; + info->fw = adev->gfx.rlc_fw; + adev->firmware.fw_size += + ALIGN(adev->gfx.rlc.global_tap_delays_ucode_size_bytes, PAGE_SIZE); + } + + if (adev->gfx.rlc.se0_tap_delays_ucode_size_bytes) { + info = &adev->firmware.ucode[AMDGPU_UCODE_ID_SE0_TAP_DELAYS]; + info->ucode_id = AMDGPU_UCODE_ID_SE0_TAP_DELAYS; + info->fw = adev->gfx.rlc_fw; + adev->firmware.fw_size += + ALIGN(adev->gfx.rlc.se0_tap_delays_ucode_size_bytes, PAGE_SIZE); + } + + if (adev->gfx.rlc.se1_tap_delays_ucode_size_bytes) { + info = &adev->firmware.ucode[AMDGPU_UCODE_ID_SE1_TAP_DELAYS]; + info->ucode_id = AMDGPU_UCODE_ID_SE1_TAP_DELAYS; + info->fw = adev->gfx.rlc_fw; + adev->firmware.fw_size += + ALIGN(adev->gfx.rlc.se1_tap_delays_ucode_size_bytes, PAGE_SIZE); + } + + if (adev->gfx.rlc.se2_tap_delays_ucode_size_bytes) { + info = &adev->firmware.ucode[AMDGPU_UCODE_ID_SE2_TAP_DELAYS]; + info->ucode_id = AMDGPU_UCODE_ID_SE2_TAP_DELAYS; + info->fw = adev->gfx.rlc_fw; + adev->firmware.fw_size += + ALIGN(adev->gfx.rlc.se2_tap_delays_ucode_size_bytes, PAGE_SIZE); + } + + if (adev->gfx.rlc.se3_tap_delays_ucode_size_bytes) { + info = &adev->firmware.ucode[AMDGPU_UCODE_ID_SE3_TAP_DELAYS]; + info->ucode_id = AMDGPU_UCODE_ID_SE3_TAP_DELAYS; + info->fw = adev->gfx.rlc_fw; + adev->firmware.fw_size += + ALIGN(adev->gfx.rlc.se3_tap_delays_ucode_size_bytes, PAGE_SIZE); + } + } +} + +int amdgpu_gfx_rlc_init_microcode(struct amdgpu_device *adev, + uint16_t version_major, + uint16_t version_minor) +{ + int err; + + if (version_major < 2) { + /* only support rlc_hdr v2.x and onwards */ + dev_err(adev->dev, "unsupported rlc fw hdr\n"); + return -EINVAL; + } + + /* is_rlc_v2_1 is still used in APU code path */ + if (version_major == 2 && version_minor == 1) + adev->gfx.rlc.is_rlc_v2_1 = true; + + if (version_minor >= 0) { + err = amdgpu_gfx_rlc_init_microcode_v2_0(adev); + if (err) { + dev_err(adev->dev, "fail to init rlc v2_0 microcode\n"); + return err; + } + } + if (version_minor >= 1) + amdgpu_gfx_rlc_init_microcode_v2_1(adev); + if (version_minor >= 2) + amdgpu_gfx_rlc_init_microcode_v2_2(adev); + if (version_minor == 3) + amdgpu_gfx_rlc_init_microcode_v2_3(adev); + if (version_minor == 4) + amdgpu_gfx_rlc_init_microcode_v2_4(adev); + + return 0; +} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_rlc.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_rlc.h index 03ac36b2c2cf..23f060db9255 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_rlc.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_rlc.h @@ -267,5 +267,7 @@ int amdgpu_gfx_rlc_init_csb(struct amdgpu_device *adev); int amdgpu_gfx_rlc_init_cpt(struct amdgpu_device *adev); void amdgpu_gfx_rlc_setup_cp_table(struct amdgpu_device *adev); void amdgpu_gfx_rlc_fini(struct amdgpu_device *adev); - +int amdgpu_gfx_rlc_init_microcode(struct amdgpu_device *adev, + uint16_t version_major, + uint16_t version_minor); #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.c index 42c1f050542f..ea5278f094c0 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.c @@ -21,6 +21,7 @@ * */ +#include <linux/firmware.h> #include "amdgpu.h" #include "amdgpu_sdma.h" #include "amdgpu_ras.h" @@ -150,3 +151,158 @@ int amdgpu_sdma_process_ecc_irq(struct amdgpu_device *adev, amdgpu_ras_interrupt_dispatch(adev, &ih_data); return 0; } + +static int amdgpu_sdma_init_inst_ctx(struct amdgpu_sdma_instance *sdma_inst) +{ + int err = 0; + uint16_t version_major; + const struct common_firmware_header *header = NULL; + const struct sdma_firmware_header_v1_0 *hdr; + const struct sdma_firmware_header_v2_0 *hdr_v2; + + err = amdgpu_ucode_validate(sdma_inst->fw); + if (err) + return err; + + header = (const struct common_firmware_header *) + sdma_inst->fw->data; + version_major = le16_to_cpu(header->header_version_major); + + switch (version_major) { + case 1: + hdr = (const struct sdma_firmware_header_v1_0 *)sdma_inst->fw->data; + sdma_inst->fw_version = le32_to_cpu(hdr->header.ucode_version); + sdma_inst->feature_version = le32_to_cpu(hdr->ucode_feature_version); + break; + case 2: + hdr_v2 = (const struct sdma_firmware_header_v2_0 *)sdma_inst->fw->data; + sdma_inst->fw_version = le32_to_cpu(hdr_v2->header.ucode_version); + sdma_inst->feature_version = le32_to_cpu(hdr_v2->ucode_feature_version); + break; + default: + return -EINVAL; + } + + if (sdma_inst->feature_version >= 20) + sdma_inst->burst_nop = true; + + return 0; +} + +void amdgpu_sdma_destroy_inst_ctx(struct amdgpu_device *adev, + bool duplicate) +{ + int i; + + for (i = 0; i < adev->sdma.num_instances; i++) { + release_firmware(adev->sdma.instance[i].fw); + if (duplicate) + break; + } + + memset((void *)adev->sdma.instance, 0, + sizeof(struct amdgpu_sdma_instance) * AMDGPU_MAX_SDMA_INSTANCES); +} + +int amdgpu_sdma_init_microcode(struct amdgpu_device *adev, + char *fw_name, u32 instance, + bool duplicate) +{ + struct amdgpu_firmware_info *info = NULL; + const struct common_firmware_header *header = NULL; + int err = 0, i; + const struct sdma_firmware_header_v2_0 *sdma_hdr; + uint16_t version_major; + + err = request_firmware(&adev->sdma.instance[instance].fw, fw_name, adev->dev); + if (err) + goto out; + + header = (const struct common_firmware_header *) + adev->sdma.instance[instance].fw->data; + version_major = le16_to_cpu(header->header_version_major); + + if ((duplicate && instance) || (!duplicate && version_major > 1)) { + err = -EINVAL; + goto out; + } + + err = amdgpu_sdma_init_inst_ctx(&adev->sdma.instance[instance]); + if (err) + goto out; + + if (duplicate) { + for (i = 1; i < adev->sdma.num_instances; i++) + memcpy((void *)&adev->sdma.instance[i], + (void *)&adev->sdma.instance[0], + sizeof(struct amdgpu_sdma_instance)); + } + + if (amdgpu_sriov_vf(adev)) + return 0; + + DRM_DEBUG("psp_load == '%s'\n", + adev->firmware.load_type == AMDGPU_FW_LOAD_PSP ? "true" : "false"); + + if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) { + switch (version_major) { + case 1: + for (i = 0; i < adev->sdma.num_instances; i++) { + if (!duplicate && (instance != i)) + continue; + else { + info = &adev->firmware.ucode[AMDGPU_UCODE_ID_SDMA0 + i]; + info->ucode_id = AMDGPU_UCODE_ID_SDMA0 + i; + info->fw = adev->sdma.instance[i].fw; + adev->firmware.fw_size += + ALIGN(le32_to_cpu(header->ucode_size_bytes), PAGE_SIZE); + } + } + break; + case 2: + sdma_hdr = (const struct sdma_firmware_header_v2_0 *) + adev->sdma.instance[0].fw->data; + info = &adev->firmware.ucode[AMDGPU_UCODE_ID_SDMA_UCODE_TH0]; + info->ucode_id = AMDGPU_UCODE_ID_SDMA_UCODE_TH0; + info->fw = adev->sdma.instance[0].fw; + adev->firmware.fw_size += + ALIGN(le32_to_cpu(sdma_hdr->ctx_ucode_size_bytes), PAGE_SIZE); + info = &adev->firmware.ucode[AMDGPU_UCODE_ID_SDMA_UCODE_TH1]; + info->ucode_id = AMDGPU_UCODE_ID_SDMA_UCODE_TH1; + info->fw = adev->sdma.instance[0].fw; + adev->firmware.fw_size += + ALIGN(le32_to_cpu(sdma_hdr->ctl_ucode_size_bytes), PAGE_SIZE); + break; + default: + err = -EINVAL; + } + } + +out: + if (err) { + DRM_ERROR("SDMA: Failed to init firmware \"%s\"\n", fw_name); + amdgpu_sdma_destroy_inst_ctx(adev, duplicate); + } + return err; +} + +void amdgpu_sdma_unset_buffer_funcs_helper(struct amdgpu_device *adev) +{ + struct amdgpu_ring *sdma; + int i; + + for (i = 0; i < adev->sdma.num_instances; i++) { + if (adev->sdma.has_page_queue) { + sdma = &adev->sdma.instance[i].page; + if (adev->mman.buffer_funcs_ring == sdma) { + amdgpu_ttm_set_buffer_funcs_status(adev, false); + break; + } + } + sdma = &adev->sdma.instance[i].ring; + if (adev->mman.buffer_funcs_ring == sdma) { + amdgpu_ttm_set_buffer_funcs_status(adev, false); + break; + } + } +} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.h index 53ac3ebae8d6..7d99205c2e01 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.h @@ -124,4 +124,10 @@ int amdgpu_sdma_process_ras_data_cb(struct amdgpu_device *adev, int amdgpu_sdma_process_ecc_irq(struct amdgpu_device *adev, struct amdgpu_irq_src *source, struct amdgpu_iv_entry *entry); +int amdgpu_sdma_init_microcode(struct amdgpu_device *adev, + char *fw_name, u32 instance, bool duplicate); +void amdgpu_sdma_destroy_inst_ctx(struct amdgpu_device *adev, + bool duplicate); +void amdgpu_sdma_unset_buffer_funcs_helper(struct amdgpu_device *adev); + #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h index 5e6ddc7e101c..677ad2016976 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h @@ -127,7 +127,7 @@ TRACE_EVENT(amdgpu_bo_create, TP_fast_assign( __entry->bo = bo; - __entry->pages = bo->tbo.resource->num_pages; + __entry->pages = PFN_UP(bo->tbo.resource->size); __entry->type = bo->tbo.resource->mem_type; __entry->prefer = bo->preferred_domains; __entry->allow = bo->allowed_domains; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c index b1c455329023..7d647993fd7e 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c @@ -381,7 +381,7 @@ static int amdgpu_move_blit(struct ttm_buffer_object *bo, dst.offset = 0; r = amdgpu_ttm_copy_mem_to_mem(adev, &src, &dst, - new_mem->num_pages << PAGE_SHIFT, + new_mem->size, amdgpu_bo_encrypted(abo), bo->base.resv, &fence); if (r) @@ -424,8 +424,9 @@ error: static bool amdgpu_mem_visible(struct amdgpu_device *adev, struct ttm_resource *mem) { - uint64_t mem_size = (u64)mem->num_pages << PAGE_SHIFT; + u64 mem_size = (u64)mem->size; struct amdgpu_res_cursor cursor; + u64 end; if (mem->mem_type == TTM_PL_SYSTEM || mem->mem_type == TTM_PL_TT) @@ -434,12 +435,21 @@ static bool amdgpu_mem_visible(struct amdgpu_device *adev, return false; amdgpu_res_first(mem, 0, mem_size, &cursor); + end = cursor.start + cursor.size; + while (cursor.remaining) { + amdgpu_res_next(&cursor, cursor.size); - /* ttm_resource_ioremap only supports contiguous memory */ - if (cursor.size != mem_size) - return false; + if (!cursor.remaining) + break; + + /* ttm_resource_ioremap only supports contiguous memory */ + if (end != cursor.start) + return false; + + end = cursor.start + cursor.size; + } - return cursor.start + cursor.size <= adev->gmc.visible_vram_size; + return end <= adev->gmc.visible_vram_size; } /* @@ -561,7 +571,7 @@ static int amdgpu_ttm_io_mem_reserve(struct ttm_device *bdev, struct ttm_resource *mem) { struct amdgpu_device *adev = amdgpu_ttm_adev(bdev); - size_t bus_size = (size_t)mem->num_pages << PAGE_SHIFT; + size_t bus_size = (size_t)mem->size; switch (mem->mem_type) { case TTM_PL_SYSTEM: diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.h index 3629d8f292ef..e46439274f3a 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.h @@ -51,6 +51,9 @@ struct amdgpu_umc_ras { struct amdgpu_ras_block_object ras_block; void (*err_cnt_init)(struct amdgpu_device *adev); bool (*query_ras_poison_mode)(struct amdgpu_device *adev); + void (*convert_ras_error_address)(struct amdgpu_device *adev, + struct ras_err_data *err_data, uint64_t err_addr, + uint32_t ch_inst, uint32_t umc_inst); void (*ecc_info_query_ras_error_count)(struct amdgpu_device *adev, void *ras_error_status); void (*ecc_info_query_ras_error_address)(struct amdgpu_device *adev, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c index f36e4f08db6d..0b52af415b28 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c @@ -191,7 +191,7 @@ int amdgpu_vcn_sw_init(struct amdgpu_device *adev) fw_name = FIRMWARE_VCN4_0_2; if ((adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) && (adev->pg_flags & AMD_PG_SUPPORT_VCN_DPG)) - adev->vcn.indirect_sram = false; + adev->vcn.indirect_sram = true; break; case IP_VERSION(4, 0, 4): fw_name = FIRMWARE_VCN4_0_4; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h index 80b7a6cfd026..253ea6b159df 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h @@ -161,7 +161,8 @@ #define AMDGPU_VCN_SW_RING_FLAG (1 << 9) #define AMDGPU_VCN_FW_LOGGING_FLAG (1 << 10) #define AMDGPU_VCN_SMU_VERSION_INFO_FLAG (1 << 11) -#define AMDGPU_VCN_VF_RB_SETUP_FLAG (1 << 12) +#define AMDGPU_VCN_SMU_DPM_INTERFACE_FLAG (1 << 11) +#define AMDGPU_VCN_VF_RB_SETUP_FLAG (1 << 14) #define AMDGPU_VCN_IB_FLAG_DECODE_BUFFER 0x00000001 #define AMDGPU_VCN_CMD_FLAG_MSG_BUFFER 0x00000001 @@ -171,6 +172,9 @@ #define VCN_CODEC_DISABLE_MASK_HEVC (1 << 2) #define VCN_CODEC_DISABLE_MASK_H264 (1 << 3) +#define AMDGPU_VCN_SMU_DPM_INTERFACE_DGPU (0) +#define AMDGPU_VCN_SMU_DPM_INTERFACE_APU (1) + enum fw_queue_mode { FW_QUEUE_RING_RESET = 1, FW_QUEUE_DPG_HOLD_OFF = 2, @@ -335,7 +339,9 @@ struct amdgpu_vcn4_fw_shared { struct amdgpu_fw_shared_unified_queue_struct sq; uint8_t pad1[8]; struct amdgpu_fw_shared_fw_logging fw_log; + uint8_t pad2[20]; struct amdgpu_fw_shared_rb_setup rb_setup; + struct amdgpu_fw_shared_smu_interface_info smu_dpm_interface; }; struct amdgpu_vcn_fwlog { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c index e4af40b9a8aa..9c765b04aae3 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c @@ -726,6 +726,12 @@ void amdgpu_detect_virtualization(struct amdgpu_device *adev) adev->virt.caps |= AMDGPU_PASSTHROUGH_MODE; } + if (amdgpu_sriov_vf(adev) && adev->asic_type == CHIP_SIENNA_CICHLID) + /* VF MMIO access (except mailbox range) from CPU + * will be blocked during sriov runtime + */ + adev->virt.caps |= AMDGPU_VF_MMIO_ACCESS_PROTECT; + /* we have the ability to check now */ if (amdgpu_sriov_vf(adev)) { switch (adev->asic_type) { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h index d94c31e68a14..49c4347d154c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h @@ -31,6 +31,7 @@ #define AMDGPU_SRIOV_CAPS_IS_VF (1 << 2) /* this GPU is a virtual function */ #define AMDGPU_PASSTHROUGH_MODE (1 << 3) /* thw whole GPU is pass through for VM */ #define AMDGPU_SRIOV_CAPS_RUNTIME (1 << 4) /* is out of full access mode */ +#define AMDGPU_VF_MMIO_ACCESS_PROTECT (1 << 5) /* MMIO write access is not allowed in sriov runtime */ /* flags for indirect register access path supported by rlcg for sriov */ #define AMDGPU_RLCG_GC_WRITE_LEGACY (0x8 << 28) @@ -297,6 +298,9 @@ struct amdgpu_video_codec_info; #define amdgpu_passthrough(adev) \ ((adev)->virt.caps & AMDGPU_PASSTHROUGH_MODE) +#define amdgpu_sriov_vf_mmio_access_protection(adev) \ +((adev)->virt.caps & AMDGPU_VF_MMIO_ACCESS_PROTECT) + static inline bool is_virtual_machine(void) { #if defined(CONFIG_X86) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vkms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vkms.c index f4b5301ea2a0..09dec2561adf 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vkms.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vkms.c @@ -498,8 +498,6 @@ static int amdgpu_vkms_sw_init(void *handle) adev_to_drm(adev)->mode_config.preferred_depth = 24; adev_to_drm(adev)->mode_config.prefer_shadow = 1; - adev_to_drm(adev)->mode_config.fb_base = adev->gmc.aper_base; - r = amdgpu_display_modeset_create_props(adev); if (r) return r; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index 83b0c5d86e48..2291aa14d888 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -2338,7 +2338,11 @@ void amdgpu_vm_manager_init(struct amdgpu_device *adev) */ #ifdef CONFIG_X86_64 if (amdgpu_vm_update_mode == -1) { - if (amdgpu_gmc_vram_full_visible(&adev->gmc)) + /* For asic with VF MMIO access protection + * avoid using CPU for VM table updates + */ + if (amdgpu_gmc_vram_full_visible(&adev->gmc) && + !amdgpu_sriov_vf_mmio_access_protection(adev)) adev->vm_manager.vm_update_mode = AMDGPU_VM_USE_CPU_FOR_COMPUTE; else diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_sdma.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_sdma.c index 2b0669c464f6..69e105fa41f6 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_sdma.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_sdma.c @@ -116,8 +116,15 @@ static int amdgpu_vm_sdma_commit(struct amdgpu_vm_update_params *p, DMA_RESV_USAGE_BOOKKEEP); } - if (fence && !p->immediate) + if (fence && !p->immediate) { + /* + * Most hw generations now have a separate queue for page table + * updates, but when the queue is shared with userspace we need + * the extra CPU round trip to correctly flush the TLB. + */ + set_bit(DRM_SCHED_FENCE_DONT_PIPELINE, &f->flags); swap(*fence, f); + } dma_fence_put(f); return 0; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c index 73a517bcf5c1..18c1a173d187 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c @@ -439,7 +439,7 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man, /* Allocate blocks in desired range */ vres->flags |= DRM_BUDDY_RANGE_ALLOCATION; - remaining_size = (u64)vres->base.num_pages << PAGE_SHIFT; + remaining_size = (u64)vres->base.size; mutex_lock(&mgr->lock); while (remaining_size) { @@ -498,7 +498,7 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man, LIST_HEAD(temp); trim_list = &vres->blocks; - original_size = (u64)vres->base.num_pages << PAGE_SHIFT; + original_size = (u64)vres->base.size; /* * If size value is rounded up to min_block_size, trim the last @@ -533,8 +533,8 @@ static int amdgpu_vram_mgr_new(struct ttm_resource_manager *man, amdgpu_vram_mgr_block_size(block); start >>= PAGE_SHIFT; - if (start > vres->base.num_pages) - start -= vres->base.num_pages; + if (start > PFN_UP(vres->base.size)) + start -= PFN_UP(vres->base.size); else start = 0; vres->base.start = max(vres->base.start, start); diff --git a/drivers/gpu/drm/amd/amdgpu/cik_sdma.c b/drivers/gpu/drm/amd/amdgpu/cik_sdma.c index 5647f13b98d4..cbca9866645c 100644 --- a/drivers/gpu/drm/amd/amdgpu/cik_sdma.c +++ b/drivers/gpu/drm/amd/amdgpu/cik_sdma.c @@ -309,14 +309,10 @@ static void cik_sdma_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 seq */ static void cik_sdma_gfx_stop(struct amdgpu_device *adev) { - struct amdgpu_ring *sdma0 = &adev->sdma.instance[0].ring; - struct amdgpu_ring *sdma1 = &adev->sdma.instance[1].ring; u32 rb_cntl; int i; - if ((adev->mman.buffer_funcs_ring == sdma0) || - (adev->mman.buffer_funcs_ring == sdma1)) - amdgpu_ttm_set_buffer_funcs_status(adev, false); + amdgpu_sdma_unset_buffer_funcs_helper(adev); for (i = 0; i < adev->sdma.num_instances; i++) { rb_cntl = RREG32(mmSDMA0_GFX_RB_CNTL + sdma_offsets[i]); diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c index 288fce7dc0ed..05051d5d2ec3 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c @@ -2800,8 +2800,6 @@ static int dce_v10_0_sw_init(void *handle) adev_to_drm(adev)->mode_config.fb_modifiers_not_supported = true; - adev_to_drm(adev)->mode_config.fb_base = adev->gmc.aper_base; - r = amdgpu_display_modeset_create_props(adev); if (r) return r; diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c index cbe5250b31cb..c928bc9eb202 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c @@ -2918,8 +2918,6 @@ static int dce_v11_0_sw_init(void *handle) adev_to_drm(adev)->mode_config.fb_modifiers_not_supported = true; - adev_to_drm(adev)->mode_config.fb_base = adev->gmc.aper_base; - r = amdgpu_display_modeset_create_props(adev); if (r) return r; diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c index b1c44fab074f..62315fd5a05f 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c @@ -2675,7 +2675,6 @@ static int dce_v6_0_sw_init(void *handle) adev_to_drm(adev)->mode_config.preferred_depth = 24; adev_to_drm(adev)->mode_config.prefer_shadow = 1; adev_to_drm(adev)->mode_config.fb_modifiers_not_supported = true; - adev_to_drm(adev)->mode_config.fb_base = adev->gmc.aper_base; r = amdgpu_display_modeset_create_props(adev); if (r) diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c index a22b45c92792..87d5e4c21cb3 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c @@ -2701,8 +2701,6 @@ static int dce_v8_0_sw_init(void *handle) adev_to_drm(adev)->mode_config.fb_modifiers_not_supported = true; - adev_to_drm(adev)->mode_config.fb_base = adev->gmc.aper_base; - r = amdgpu_display_modeset_create_props(adev); if (r) return r; diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c index e4dde41f2f68..af94ac580d3e 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c @@ -3943,56 +3943,6 @@ static void gfx_v10_0_check_fw_write_wait(struct amdgpu_device *adev) DRM_WARN_ONCE("CP firmware version too old, please update!"); } - -static void gfx_v10_0_init_rlc_ext_microcode(struct amdgpu_device *adev) -{ - const struct rlc_firmware_header_v2_1 *rlc_hdr; - - rlc_hdr = (const struct rlc_firmware_header_v2_1 *)adev->gfx.rlc_fw->data; - adev->gfx.rlc_srlc_fw_version = le32_to_cpu(rlc_hdr->save_restore_list_cntl_ucode_ver); - adev->gfx.rlc_srlc_feature_version = le32_to_cpu(rlc_hdr->save_restore_list_cntl_feature_ver); - adev->gfx.rlc.save_restore_list_cntl_size_bytes = le32_to_cpu(rlc_hdr->save_restore_list_cntl_size_bytes); - adev->gfx.rlc.save_restore_list_cntl = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->save_restore_list_cntl_offset_bytes); - adev->gfx.rlc_srlg_fw_version = le32_to_cpu(rlc_hdr->save_restore_list_gpm_ucode_ver); - adev->gfx.rlc_srlg_feature_version = le32_to_cpu(rlc_hdr->save_restore_list_gpm_feature_ver); - adev->gfx.rlc.save_restore_list_gpm_size_bytes = le32_to_cpu(rlc_hdr->save_restore_list_gpm_size_bytes); - adev->gfx.rlc.save_restore_list_gpm = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->save_restore_list_gpm_offset_bytes); - adev->gfx.rlc_srls_fw_version = le32_to_cpu(rlc_hdr->save_restore_list_srm_ucode_ver); - adev->gfx.rlc_srls_feature_version = le32_to_cpu(rlc_hdr->save_restore_list_srm_feature_ver); - adev->gfx.rlc.save_restore_list_srm_size_bytes = le32_to_cpu(rlc_hdr->save_restore_list_srm_size_bytes); - adev->gfx.rlc.save_restore_list_srm = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->save_restore_list_srm_offset_bytes); - adev->gfx.rlc.reg_list_format_direct_reg_list_length = - le32_to_cpu(rlc_hdr->reg_list_format_direct_reg_list_length); -} - -static void gfx_v10_0_init_rlc_iram_dram_microcode(struct amdgpu_device *adev) -{ - const struct rlc_firmware_header_v2_2 *rlc_hdr; - - rlc_hdr = (const struct rlc_firmware_header_v2_2 *)adev->gfx.rlc_fw->data; - adev->gfx.rlc.rlc_iram_ucode_size_bytes = le32_to_cpu(rlc_hdr->rlc_iram_ucode_size_bytes); - adev->gfx.rlc.rlc_iram_ucode = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->rlc_iram_ucode_offset_bytes); - adev->gfx.rlc.rlc_dram_ucode_size_bytes = le32_to_cpu(rlc_hdr->rlc_dram_ucode_size_bytes); - adev->gfx.rlc.rlc_dram_ucode = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->rlc_dram_ucode_offset_bytes); -} - -static void gfx_v10_0_init_tap_delays_microcode(struct amdgpu_device *adev) -{ - const struct rlc_firmware_header_v2_4 *rlc_hdr; - - rlc_hdr = (const struct rlc_firmware_header_v2_4 *)adev->gfx.rlc_fw->data; - adev->gfx.rlc.global_tap_delays_ucode_size_bytes = le32_to_cpu(rlc_hdr->global_tap_delays_ucode_size_bytes); - adev->gfx.rlc.global_tap_delays_ucode = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->global_tap_delays_ucode_offset_bytes); - adev->gfx.rlc.se0_tap_delays_ucode_size_bytes = le32_to_cpu(rlc_hdr->se0_tap_delays_ucode_size_bytes); - adev->gfx.rlc.se0_tap_delays_ucode = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->se0_tap_delays_ucode_offset_bytes); - adev->gfx.rlc.se1_tap_delays_ucode_size_bytes = le32_to_cpu(rlc_hdr->se1_tap_delays_ucode_size_bytes); - adev->gfx.rlc.se1_tap_delays_ucode = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->se1_tap_delays_ucode_offset_bytes); - adev->gfx.rlc.se2_tap_delays_ucode_size_bytes = le32_to_cpu(rlc_hdr->se2_tap_delays_ucode_size_bytes); - adev->gfx.rlc.se2_tap_delays_ucode = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->se2_tap_delays_ucode_offset_bytes); - adev->gfx.rlc.se3_tap_delays_ucode_size_bytes = le32_to_cpu(rlc_hdr->se3_tap_delays_ucode_size_bytes); - adev->gfx.rlc.se3_tap_delays_ucode = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->se3_tap_delays_ucode_offset_bytes); -} - static bool gfx_v10_0_navi10_gfxoff_should_enable(struct amdgpu_device *adev) { bool ret = false; @@ -4028,12 +3978,7 @@ static int gfx_v10_0_init_microcode(struct amdgpu_device *adev) char fw_name[40]; char *wks = ""; int err; - struct amdgpu_firmware_info *info = NULL; - const struct common_firmware_header *header = NULL; - const struct gfx_firmware_header_v1_0 *cp_hdr; const struct rlc_firmware_header_v2_0 *rlc_hdr; - unsigned int *tmp = NULL; - unsigned int i = 0; uint16_t version_major; uint16_t version_minor; @@ -4091,9 +4036,7 @@ static int gfx_v10_0_init_microcode(struct amdgpu_device *adev) err = amdgpu_ucode_validate(adev->gfx.pfp_fw); if (err) goto out; - cp_hdr = (const struct gfx_firmware_header_v1_0 *)adev->gfx.pfp_fw->data; - adev->gfx.pfp_fw_version = le32_to_cpu(cp_hdr->header.ucode_version); - adev->gfx.pfp_feature_version = le32_to_cpu(cp_hdr->ucode_feature_version); + amdgpu_gfx_cp_init_microcode(adev, AMDGPU_UCODE_ID_CP_PFP); snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_me%s.bin", chip_name, wks); err = request_firmware(&adev->gfx.me_fw, fw_name, adev->dev); @@ -4102,9 +4045,7 @@ static int gfx_v10_0_init_microcode(struct amdgpu_device *adev) err = amdgpu_ucode_validate(adev->gfx.me_fw); if (err) goto out; - cp_hdr = (const struct gfx_firmware_header_v1_0 *)adev->gfx.me_fw->data; - adev->gfx.me_fw_version = le32_to_cpu(cp_hdr->header.ucode_version); - adev->gfx.me_feature_version = le32_to_cpu(cp_hdr->ucode_feature_version); + amdgpu_gfx_cp_init_microcode(adev, AMDGPU_UCODE_ID_CP_ME); snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_ce%s.bin", chip_name, wks); err = request_firmware(&adev->gfx.ce_fw, fw_name, adev->dev); @@ -4113,69 +4054,27 @@ static int gfx_v10_0_init_microcode(struct amdgpu_device *adev) err = amdgpu_ucode_validate(adev->gfx.ce_fw); if (err) goto out; - cp_hdr = (const struct gfx_firmware_header_v1_0 *)adev->gfx.ce_fw->data; - adev->gfx.ce_fw_version = le32_to_cpu(cp_hdr->header.ucode_version); - adev->gfx.ce_feature_version = le32_to_cpu(cp_hdr->ucode_feature_version); + amdgpu_gfx_cp_init_microcode(adev, AMDGPU_UCODE_ID_CP_CE); if (!amdgpu_sriov_vf(adev)) { snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_rlc.bin", chip_name); err = request_firmware(&adev->gfx.rlc_fw, fw_name, adev->dev); if (err) goto out; + /* don't check this. There are apparently firmwares in the wild with + * incorrect size in the header + */ err = amdgpu_ucode_validate(adev->gfx.rlc_fw); + if (err) + dev_dbg(adev->dev, + "gfx10: amdgpu_ucode_validate() failed \"%s\"\n", + fw_name); rlc_hdr = (const struct rlc_firmware_header_v2_0 *)adev->gfx.rlc_fw->data; version_major = le16_to_cpu(rlc_hdr->header.header_version_major); version_minor = le16_to_cpu(rlc_hdr->header.header_version_minor); - - adev->gfx.rlc_fw_version = le32_to_cpu(rlc_hdr->header.ucode_version); - adev->gfx.rlc_feature_version = le32_to_cpu(rlc_hdr->ucode_feature_version); - adev->gfx.rlc.save_and_restore_offset = - le32_to_cpu(rlc_hdr->save_and_restore_offset); - adev->gfx.rlc.clear_state_descriptor_offset = - le32_to_cpu(rlc_hdr->clear_state_descriptor_offset); - adev->gfx.rlc.avail_scratch_ram_locations = - le32_to_cpu(rlc_hdr->avail_scratch_ram_locations); - adev->gfx.rlc.reg_restore_list_size = - le32_to_cpu(rlc_hdr->reg_restore_list_size); - adev->gfx.rlc.reg_list_format_start = - le32_to_cpu(rlc_hdr->reg_list_format_start); - adev->gfx.rlc.reg_list_format_separate_start = - le32_to_cpu(rlc_hdr->reg_list_format_separate_start); - adev->gfx.rlc.starting_offsets_start = - le32_to_cpu(rlc_hdr->starting_offsets_start); - adev->gfx.rlc.reg_list_format_size_bytes = - le32_to_cpu(rlc_hdr->reg_list_format_size_bytes); - adev->gfx.rlc.reg_list_size_bytes = - le32_to_cpu(rlc_hdr->reg_list_size_bytes); - adev->gfx.rlc.register_list_format = - kmalloc(adev->gfx.rlc.reg_list_format_size_bytes + - adev->gfx.rlc.reg_list_size_bytes, GFP_KERNEL); - if (!adev->gfx.rlc.register_list_format) { - err = -ENOMEM; + err = amdgpu_gfx_rlc_init_microcode(adev, version_major, version_minor); + if (err) goto out; - } - - tmp = (unsigned int *)((uintptr_t)rlc_hdr + - le32_to_cpu(rlc_hdr->reg_list_format_array_offset_bytes)); - for (i = 0 ; i < (rlc_hdr->reg_list_format_size_bytes >> 2); i++) - adev->gfx.rlc.register_list_format[i] = le32_to_cpu(tmp[i]); - - adev->gfx.rlc.register_restore = adev->gfx.rlc.register_list_format + i; - - tmp = (unsigned int *)((uintptr_t)rlc_hdr + - le32_to_cpu(rlc_hdr->reg_list_array_offset_bytes)); - for (i = 0 ; i < (rlc_hdr->reg_list_size_bytes >> 2); i++) - adev->gfx.rlc.register_restore[i] = le32_to_cpu(tmp[i]); - - if (version_major == 2) { - if (version_minor >= 1) - gfx_v10_0_init_rlc_ext_microcode(adev); - if (version_minor >= 2) - gfx_v10_0_init_rlc_iram_dram_microcode(adev); - if (version_minor == 4) { - gfx_v10_0_init_tap_delays_microcode(adev); - } - } } snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_mec%s.bin", chip_name, wks); @@ -4185,9 +4084,8 @@ static int gfx_v10_0_init_microcode(struct amdgpu_device *adev) err = amdgpu_ucode_validate(adev->gfx.mec_fw); if (err) goto out; - cp_hdr = (const struct gfx_firmware_header_v1_0 *)adev->gfx.mec_fw->data; - adev->gfx.mec_fw_version = le32_to_cpu(cp_hdr->header.ucode_version); - adev->gfx.mec_feature_version = le32_to_cpu(cp_hdr->ucode_feature_version); + amdgpu_gfx_cp_init_microcode(adev, AMDGPU_UCODE_ID_CP_MEC1); + amdgpu_gfx_cp_init_microcode(adev, AMDGPU_UCODE_ID_CP_MEC1_JT); snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_mec2%s.bin", chip_name, wks); err = request_firmware(&adev->gfx.mec2_fw, fw_name, adev->dev); @@ -4195,164 +4093,18 @@ static int gfx_v10_0_init_microcode(struct amdgpu_device *adev) err = amdgpu_ucode_validate(adev->gfx.mec2_fw); if (err) goto out; - cp_hdr = (const struct gfx_firmware_header_v1_0 *) - adev->gfx.mec2_fw->data; - adev->gfx.mec2_fw_version = - le32_to_cpu(cp_hdr->header.ucode_version); - adev->gfx.mec2_feature_version = - le32_to_cpu(cp_hdr->ucode_feature_version); + amdgpu_gfx_cp_init_microcode(adev, AMDGPU_UCODE_ID_CP_MEC2); + amdgpu_gfx_cp_init_microcode(adev, AMDGPU_UCODE_ID_CP_MEC2_JT); } else { err = 0; adev->gfx.mec2_fw = NULL; } - if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) { - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_CP_PFP]; - info->ucode_id = AMDGPU_UCODE_ID_CP_PFP; - info->fw = adev->gfx.pfp_fw; - header = (const struct common_firmware_header *)info->fw->data; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(header->ucode_size_bytes), PAGE_SIZE); - - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_CP_ME]; - info->ucode_id = AMDGPU_UCODE_ID_CP_ME; - info->fw = adev->gfx.me_fw; - header = (const struct common_firmware_header *)info->fw->data; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(header->ucode_size_bytes), PAGE_SIZE); - - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_CP_CE]; - info->ucode_id = AMDGPU_UCODE_ID_CP_CE; - info->fw = adev->gfx.ce_fw; - header = (const struct common_firmware_header *)info->fw->data; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(header->ucode_size_bytes), PAGE_SIZE); - - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_RLC_G]; - info->ucode_id = AMDGPU_UCODE_ID_RLC_G; - info->fw = adev->gfx.rlc_fw; - if (info->fw) { - header = (const struct common_firmware_header *)info->fw->data; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(header->ucode_size_bytes), PAGE_SIZE); - } - if (adev->gfx.rlc.save_restore_list_cntl_size_bytes && - adev->gfx.rlc.save_restore_list_gpm_size_bytes && - adev->gfx.rlc.save_restore_list_srm_size_bytes) { - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_RLC_RESTORE_LIST_CNTL]; - info->ucode_id = AMDGPU_UCODE_ID_RLC_RESTORE_LIST_CNTL; - info->fw = adev->gfx.rlc_fw; - adev->firmware.fw_size += - ALIGN(adev->gfx.rlc.save_restore_list_cntl_size_bytes, PAGE_SIZE); - - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_RLC_RESTORE_LIST_GPM_MEM]; - info->ucode_id = AMDGPU_UCODE_ID_RLC_RESTORE_LIST_GPM_MEM; - info->fw = adev->gfx.rlc_fw; - adev->firmware.fw_size += - ALIGN(adev->gfx.rlc.save_restore_list_gpm_size_bytes, PAGE_SIZE); - - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_RLC_RESTORE_LIST_SRM_MEM]; - info->ucode_id = AMDGPU_UCODE_ID_RLC_RESTORE_LIST_SRM_MEM; - info->fw = adev->gfx.rlc_fw; - adev->firmware.fw_size += - ALIGN(adev->gfx.rlc.save_restore_list_srm_size_bytes, PAGE_SIZE); - - if (adev->gfx.rlc.rlc_iram_ucode_size_bytes && - adev->gfx.rlc.rlc_dram_ucode_size_bytes) { - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_RLC_IRAM]; - info->ucode_id = AMDGPU_UCODE_ID_RLC_IRAM; - info->fw = adev->gfx.rlc_fw; - adev->firmware.fw_size += - ALIGN(adev->gfx.rlc.rlc_iram_ucode_size_bytes, PAGE_SIZE); - - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_RLC_DRAM]; - info->ucode_id = AMDGPU_UCODE_ID_RLC_DRAM; - info->fw = adev->gfx.rlc_fw; - adev->firmware.fw_size += - ALIGN(adev->gfx.rlc.rlc_dram_ucode_size_bytes, PAGE_SIZE); - } - - } - - if (adev->gfx.rlc.global_tap_delays_ucode_size_bytes) { - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_GLOBAL_TAP_DELAYS]; - info->ucode_id = AMDGPU_UCODE_ID_GLOBAL_TAP_DELAYS; - info->fw = adev->gfx.rlc_fw; - adev->firmware.fw_size += - ALIGN(adev->gfx.rlc.global_tap_delays_ucode_size_bytes, PAGE_SIZE); - } - - if (adev->gfx.rlc.se0_tap_delays_ucode_size_bytes) { - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_SE0_TAP_DELAYS]; - info->ucode_id = AMDGPU_UCODE_ID_SE0_TAP_DELAYS; - info->fw = adev->gfx.rlc_fw; - adev->firmware.fw_size += - ALIGN(adev->gfx.rlc.se0_tap_delays_ucode_size_bytes, PAGE_SIZE); - } - - if (adev->gfx.rlc.se1_tap_delays_ucode_size_bytes) { - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_SE1_TAP_DELAYS]; - info->ucode_id = AMDGPU_UCODE_ID_SE1_TAP_DELAYS; - info->fw = adev->gfx.rlc_fw; - adev->firmware.fw_size += - ALIGN(adev->gfx.rlc.se1_tap_delays_ucode_size_bytes, PAGE_SIZE); - } - - if (adev->gfx.rlc.se2_tap_delays_ucode_size_bytes) { - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_SE2_TAP_DELAYS]; - info->ucode_id = AMDGPU_UCODE_ID_SE2_TAP_DELAYS; - info->fw = adev->gfx.rlc_fw; - adev->firmware.fw_size += - ALIGN(adev->gfx.rlc.se2_tap_delays_ucode_size_bytes, PAGE_SIZE); - } - - if (adev->gfx.rlc.se3_tap_delays_ucode_size_bytes) { - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_SE3_TAP_DELAYS]; - info->ucode_id = AMDGPU_UCODE_ID_SE3_TAP_DELAYS; - info->fw = adev->gfx.rlc_fw; - adev->firmware.fw_size += - ALIGN(adev->gfx.rlc.se3_tap_delays_ucode_size_bytes, PAGE_SIZE); - } - - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_CP_MEC1]; - info->ucode_id = AMDGPU_UCODE_ID_CP_MEC1; - info->fw = adev->gfx.mec_fw; - header = (const struct common_firmware_header *)info->fw->data; - cp_hdr = (const struct gfx_firmware_header_v1_0 *)info->fw->data; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(header->ucode_size_bytes) - - le32_to_cpu(cp_hdr->jt_size) * 4, PAGE_SIZE); - - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_CP_MEC1_JT]; - info->ucode_id = AMDGPU_UCODE_ID_CP_MEC1_JT; - info->fw = adev->gfx.mec_fw; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(cp_hdr->jt_size) * 4, PAGE_SIZE); - - if (adev->gfx.mec2_fw) { - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_CP_MEC2]; - info->ucode_id = AMDGPU_UCODE_ID_CP_MEC2; - info->fw = adev->gfx.mec2_fw; - header = (const struct common_firmware_header *)info->fw->data; - cp_hdr = (const struct gfx_firmware_header_v1_0 *)info->fw->data; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(header->ucode_size_bytes) - - le32_to_cpu(cp_hdr->jt_size) * 4, - PAGE_SIZE); - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_CP_MEC2_JT]; - info->ucode_id = AMDGPU_UCODE_ID_CP_MEC2_JT; - info->fw = adev->gfx.mec2_fw; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(cp_hdr->jt_size) * 4, - PAGE_SIZE); - } - } - gfx_v10_0_check_fw_write_wait(adev); out: if (err) { dev_err(adev->dev, - "gfx10: Failed to load firmware \"%s\"\n", + "gfx10: Failed to init firmware \"%s\"\n", fw_name); release_firmware(adev->gfx.pfp_fw); adev->gfx.pfp_fw = NULL; diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c index ce8c792cef1a..671ca5a0f208 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c @@ -438,66 +438,12 @@ static void gfx_v11_0_free_microcode(struct amdgpu_device *adev) kfree(adev->gfx.rlc.register_list_format); } -static void gfx_v11_0_init_rlc_ext_microcode(struct amdgpu_device *adev) -{ - const struct rlc_firmware_header_v2_1 *rlc_hdr; - - rlc_hdr = (const struct rlc_firmware_header_v2_1 *)adev->gfx.rlc_fw->data; - adev->gfx.rlc_srlc_fw_version = le32_to_cpu(rlc_hdr->save_restore_list_cntl_ucode_ver); - adev->gfx.rlc_srlc_feature_version = le32_to_cpu(rlc_hdr->save_restore_list_cntl_feature_ver); - adev->gfx.rlc.save_restore_list_cntl_size_bytes = le32_to_cpu(rlc_hdr->save_restore_list_cntl_size_bytes); - adev->gfx.rlc.save_restore_list_cntl = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->save_restore_list_cntl_offset_bytes); - adev->gfx.rlc_srlg_fw_version = le32_to_cpu(rlc_hdr->save_restore_list_gpm_ucode_ver); - adev->gfx.rlc_srlg_feature_version = le32_to_cpu(rlc_hdr->save_restore_list_gpm_feature_ver); - adev->gfx.rlc.save_restore_list_gpm_size_bytes = le32_to_cpu(rlc_hdr->save_restore_list_gpm_size_bytes); - adev->gfx.rlc.save_restore_list_gpm = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->save_restore_list_gpm_offset_bytes); - adev->gfx.rlc_srls_fw_version = le32_to_cpu(rlc_hdr->save_restore_list_srm_ucode_ver); - adev->gfx.rlc_srls_feature_version = le32_to_cpu(rlc_hdr->save_restore_list_srm_feature_ver); - adev->gfx.rlc.save_restore_list_srm_size_bytes = le32_to_cpu(rlc_hdr->save_restore_list_srm_size_bytes); - adev->gfx.rlc.save_restore_list_srm = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->save_restore_list_srm_offset_bytes); - adev->gfx.rlc.reg_list_format_direct_reg_list_length = - le32_to_cpu(rlc_hdr->reg_list_format_direct_reg_list_length); -} - -static void gfx_v11_0_init_rlc_iram_dram_microcode(struct amdgpu_device *adev) -{ - const struct rlc_firmware_header_v2_2 *rlc_hdr; - - rlc_hdr = (const struct rlc_firmware_header_v2_2 *)adev->gfx.rlc_fw->data; - adev->gfx.rlc.rlc_iram_ucode_size_bytes = le32_to_cpu(rlc_hdr->rlc_iram_ucode_size_bytes); - adev->gfx.rlc.rlc_iram_ucode = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->rlc_iram_ucode_offset_bytes); - adev->gfx.rlc.rlc_dram_ucode_size_bytes = le32_to_cpu(rlc_hdr->rlc_dram_ucode_size_bytes); - adev->gfx.rlc.rlc_dram_ucode = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->rlc_dram_ucode_offset_bytes); -} - -static void gfx_v11_0_init_rlcp_rlcv_microcode(struct amdgpu_device *adev) -{ - const struct rlc_firmware_header_v2_3 *rlc_hdr; - - rlc_hdr = (const struct rlc_firmware_header_v2_3 *)adev->gfx.rlc_fw->data; - adev->gfx.rlcp_ucode_version = le32_to_cpu(rlc_hdr->rlcp_ucode_version); - adev->gfx.rlcp_ucode_feature_version = le32_to_cpu(rlc_hdr->rlcp_ucode_feature_version); - adev->gfx.rlc.rlcp_ucode_size_bytes = le32_to_cpu(rlc_hdr->rlcp_ucode_size_bytes); - adev->gfx.rlc.rlcp_ucode = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->rlcp_ucode_offset_bytes); - - adev->gfx.rlcv_ucode_version = le32_to_cpu(rlc_hdr->rlcv_ucode_version); - adev->gfx.rlcv_ucode_feature_version = le32_to_cpu(rlc_hdr->rlcv_ucode_feature_version); - adev->gfx.rlc.rlcv_ucode_size_bytes = le32_to_cpu(rlc_hdr->rlcv_ucode_size_bytes); - adev->gfx.rlc.rlcv_ucode = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->rlcv_ucode_offset_bytes); -} - static int gfx_v11_0_init_microcode(struct amdgpu_device *adev) { char fw_name[40]; char ucode_prefix[30]; int err; - struct amdgpu_firmware_info *info = NULL; - const struct common_firmware_header *header = NULL; - const struct gfx_firmware_header_v1_0 *cp_hdr; - const struct gfx_firmware_header_v2_0 *cp_hdr_v2_0; const struct rlc_firmware_header_v2_0 *rlc_hdr; - unsigned int *tmp = NULL; - unsigned int i = 0; uint16_t version_major; uint16_t version_minor; @@ -518,14 +464,11 @@ static int gfx_v11_0_init_microcode(struct amdgpu_device *adev) adev->gfx.pfp_fw->data, 2, 0); if (adev->gfx.rs64_enable) { dev_info(adev->dev, "CP RS64 enable\n"); - cp_hdr_v2_0 = (const struct gfx_firmware_header_v2_0 *)adev->gfx.pfp_fw->data; - adev->gfx.pfp_fw_version = le32_to_cpu(cp_hdr_v2_0->header.ucode_version); - adev->gfx.pfp_feature_version = le32_to_cpu(cp_hdr_v2_0->ucode_feature_version); - + amdgpu_gfx_cp_init_microcode(adev, AMDGPU_UCODE_ID_CP_RS64_PFP); + amdgpu_gfx_cp_init_microcode(adev, AMDGPU_UCODE_ID_CP_RS64_PFP_P0_STACK); + amdgpu_gfx_cp_init_microcode(adev, AMDGPU_UCODE_ID_CP_RS64_PFP_P1_STACK); } else { - cp_hdr = (const struct gfx_firmware_header_v1_0 *)adev->gfx.pfp_fw->data; - adev->gfx.pfp_fw_version = le32_to_cpu(cp_hdr->header.ucode_version); - adev->gfx.pfp_feature_version = le32_to_cpu(cp_hdr->ucode_feature_version); + amdgpu_gfx_cp_init_microcode(adev, AMDGPU_UCODE_ID_CP_PFP); } snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_me.bin", ucode_prefix); @@ -536,14 +479,11 @@ static int gfx_v11_0_init_microcode(struct amdgpu_device *adev) if (err) goto out; if (adev->gfx.rs64_enable) { - cp_hdr_v2_0 = (const struct gfx_firmware_header_v2_0 *)adev->gfx.me_fw->data; - adev->gfx.me_fw_version = le32_to_cpu(cp_hdr_v2_0->header.ucode_version); - adev->gfx.me_feature_version = le32_to_cpu(cp_hdr_v2_0->ucode_feature_version); - + amdgpu_gfx_cp_init_microcode(adev, AMDGPU_UCODE_ID_CP_RS64_ME); + amdgpu_gfx_cp_init_microcode(adev, AMDGPU_UCODE_ID_CP_RS64_ME_P0_STACK); + amdgpu_gfx_cp_init_microcode(adev, AMDGPU_UCODE_ID_CP_RS64_ME_P1_STACK); } else { - cp_hdr = (const struct gfx_firmware_header_v1_0 *)adev->gfx.me_fw->data; - adev->gfx.me_fw_version = le32_to_cpu(cp_hdr->header.ucode_version); - adev->gfx.me_feature_version = le32_to_cpu(cp_hdr->ucode_feature_version); + amdgpu_gfx_cp_init_microcode(adev, AMDGPU_UCODE_ID_CP_ME); } if (!amdgpu_sriov_vf(adev)) { @@ -552,58 +492,14 @@ static int gfx_v11_0_init_microcode(struct amdgpu_device *adev) if (err) goto out; err = amdgpu_ucode_validate(adev->gfx.rlc_fw); + if (err) + goto out; rlc_hdr = (const struct rlc_firmware_header_v2_0 *)adev->gfx.rlc_fw->data; version_major = le16_to_cpu(rlc_hdr->header.header_version_major); version_minor = le16_to_cpu(rlc_hdr->header.header_version_minor); - - adev->gfx.rlc_fw_version = le32_to_cpu(rlc_hdr->header.ucode_version); - adev->gfx.rlc_feature_version = le32_to_cpu(rlc_hdr->ucode_feature_version); - adev->gfx.rlc.save_and_restore_offset = - le32_to_cpu(rlc_hdr->save_and_restore_offset); - adev->gfx.rlc.clear_state_descriptor_offset = - le32_to_cpu(rlc_hdr->clear_state_descriptor_offset); - adev->gfx.rlc.avail_scratch_ram_locations = - le32_to_cpu(rlc_hdr->avail_scratch_ram_locations); - adev->gfx.rlc.reg_restore_list_size = - le32_to_cpu(rlc_hdr->reg_restore_list_size); - adev->gfx.rlc.reg_list_format_start = - le32_to_cpu(rlc_hdr->reg_list_format_start); - adev->gfx.rlc.reg_list_format_separate_start = - le32_to_cpu(rlc_hdr->reg_list_format_separate_start); - adev->gfx.rlc.starting_offsets_start = - le32_to_cpu(rlc_hdr->starting_offsets_start); - adev->gfx.rlc.reg_list_format_size_bytes = - le32_to_cpu(rlc_hdr->reg_list_format_size_bytes); - adev->gfx.rlc.reg_list_size_bytes = - le32_to_cpu(rlc_hdr->reg_list_size_bytes); - adev->gfx.rlc.register_list_format = - kmalloc(adev->gfx.rlc.reg_list_format_size_bytes + - adev->gfx.rlc.reg_list_size_bytes, GFP_KERNEL); - if (!adev->gfx.rlc.register_list_format) { - err = -ENOMEM; + err = amdgpu_gfx_rlc_init_microcode(adev, version_major, version_minor); + if (err) goto out; - } - - tmp = (unsigned int *)((uintptr_t)rlc_hdr + - le32_to_cpu(rlc_hdr->reg_list_format_array_offset_bytes)); - for (i = 0 ; i < (rlc_hdr->reg_list_format_size_bytes >> 2); i++) - adev->gfx.rlc.register_list_format[i] = le32_to_cpu(tmp[i]); - - adev->gfx.rlc.register_restore = adev->gfx.rlc.register_list_format + i; - - tmp = (unsigned int *)((uintptr_t)rlc_hdr + - le32_to_cpu(rlc_hdr->reg_list_array_offset_bytes)); - for (i = 0 ; i < (rlc_hdr->reg_list_size_bytes >> 2); i++) - adev->gfx.rlc.register_restore[i] = le32_to_cpu(tmp[i]); - - if (version_major == 2) { - if (version_minor >= 1) - gfx_v11_0_init_rlc_ext_microcode(adev); - if (version_minor >= 2) - gfx_v11_0_init_rlc_iram_dram_microcode(adev); - if (version_minor == 3) - gfx_v11_0_init_rlcp_rlcv_microcode(adev); - } } snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_mec.bin", ucode_prefix); @@ -614,190 +510,23 @@ static int gfx_v11_0_init_microcode(struct amdgpu_device *adev) if (err) goto out; if (adev->gfx.rs64_enable) { - cp_hdr_v2_0 = (const struct gfx_firmware_header_v2_0 *)adev->gfx.mec_fw->data; - adev->gfx.mec_fw_version = le32_to_cpu(cp_hdr_v2_0->header.ucode_version); - adev->gfx.mec_feature_version = le32_to_cpu(cp_hdr_v2_0->ucode_feature_version); - + amdgpu_gfx_cp_init_microcode(adev, AMDGPU_UCODE_ID_CP_RS64_MEC); + amdgpu_gfx_cp_init_microcode(adev, AMDGPU_UCODE_ID_CP_RS64_MEC_P0_STACK); + amdgpu_gfx_cp_init_microcode(adev, AMDGPU_UCODE_ID_CP_RS64_MEC_P1_STACK); + amdgpu_gfx_cp_init_microcode(adev, AMDGPU_UCODE_ID_CP_RS64_MEC_P2_STACK); + amdgpu_gfx_cp_init_microcode(adev, AMDGPU_UCODE_ID_CP_RS64_MEC_P3_STACK); } else { - cp_hdr = (const struct gfx_firmware_header_v1_0 *)adev->gfx.mec_fw->data; - adev->gfx.mec_fw_version = le32_to_cpu(cp_hdr->header.ucode_version); - adev->gfx.mec_feature_version = le32_to_cpu(cp_hdr->ucode_feature_version); + amdgpu_gfx_cp_init_microcode(adev, AMDGPU_UCODE_ID_CP_MEC1); + amdgpu_gfx_cp_init_microcode(adev, AMDGPU_UCODE_ID_CP_MEC1_JT); } /* only one MEC for gfx 11.0.0. */ adev->gfx.mec2_fw = NULL; - if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) { - if (adev->gfx.rs64_enable) { - cp_hdr_v2_0 = (const struct gfx_firmware_header_v2_0 *)adev->gfx.pfp_fw->data; - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_CP_RS64_PFP]; - info->ucode_id = AMDGPU_UCODE_ID_CP_RS64_PFP; - info->fw = adev->gfx.pfp_fw; - header = (const struct common_firmware_header *)info->fw->data; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(cp_hdr_v2_0->ucode_size_bytes), PAGE_SIZE); - - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_CP_RS64_PFP_P0_STACK]; - info->ucode_id = AMDGPU_UCODE_ID_CP_RS64_PFP_P0_STACK; - info->fw = adev->gfx.pfp_fw; - header = (const struct common_firmware_header *)info->fw->data; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(cp_hdr_v2_0->data_size_bytes), PAGE_SIZE); - - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_CP_RS64_PFP_P1_STACK]; - info->ucode_id = AMDGPU_UCODE_ID_CP_RS64_PFP_P1_STACK; - info->fw = adev->gfx.pfp_fw; - header = (const struct common_firmware_header *)info->fw->data; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(cp_hdr_v2_0->data_size_bytes), PAGE_SIZE); - - cp_hdr_v2_0 = (const struct gfx_firmware_header_v2_0 *)adev->gfx.me_fw->data; - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_CP_RS64_ME]; - info->ucode_id = AMDGPU_UCODE_ID_CP_RS64_ME; - info->fw = adev->gfx.me_fw; - header = (const struct common_firmware_header *)info->fw->data; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(cp_hdr_v2_0->ucode_size_bytes), PAGE_SIZE); - - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_CP_RS64_ME_P0_STACK]; - info->ucode_id = AMDGPU_UCODE_ID_CP_RS64_ME_P0_STACK; - info->fw = adev->gfx.me_fw; - header = (const struct common_firmware_header *)info->fw->data; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(cp_hdr_v2_0->data_size_bytes), PAGE_SIZE); - - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_CP_RS64_ME_P1_STACK]; - info->ucode_id = AMDGPU_UCODE_ID_CP_RS64_ME_P1_STACK; - info->fw = adev->gfx.me_fw; - header = (const struct common_firmware_header *)info->fw->data; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(cp_hdr_v2_0->data_size_bytes), PAGE_SIZE); - - cp_hdr_v2_0 = (const struct gfx_firmware_header_v2_0 *)adev->gfx.mec_fw->data; - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_CP_RS64_MEC]; - info->ucode_id = AMDGPU_UCODE_ID_CP_RS64_MEC; - info->fw = adev->gfx.mec_fw; - header = (const struct common_firmware_header *)info->fw->data; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(cp_hdr_v2_0->ucode_size_bytes), PAGE_SIZE); - - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_CP_RS64_MEC_P0_STACK]; - info->ucode_id = AMDGPU_UCODE_ID_CP_RS64_MEC_P0_STACK; - info->fw = adev->gfx.mec_fw; - header = (const struct common_firmware_header *)info->fw->data; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(cp_hdr_v2_0->data_size_bytes), PAGE_SIZE); - - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_CP_RS64_MEC_P1_STACK]; - info->ucode_id = AMDGPU_UCODE_ID_CP_RS64_MEC_P1_STACK; - info->fw = adev->gfx.mec_fw; - header = (const struct common_firmware_header *)info->fw->data; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(cp_hdr_v2_0->data_size_bytes), PAGE_SIZE); - - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_CP_RS64_MEC_P2_STACK]; - info->ucode_id = AMDGPU_UCODE_ID_CP_RS64_MEC_P2_STACK; - info->fw = adev->gfx.mec_fw; - header = (const struct common_firmware_header *)info->fw->data; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(cp_hdr_v2_0->data_size_bytes), PAGE_SIZE); - - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_CP_RS64_MEC_P3_STACK]; - info->ucode_id = AMDGPU_UCODE_ID_CP_RS64_MEC_P3_STACK; - info->fw = adev->gfx.mec_fw; - header = (const struct common_firmware_header *)info->fw->data; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(cp_hdr_v2_0->data_size_bytes), PAGE_SIZE); - } else { - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_CP_PFP]; - info->ucode_id = AMDGPU_UCODE_ID_CP_PFP; - info->fw = adev->gfx.pfp_fw; - header = (const struct common_firmware_header *)info->fw->data; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(header->ucode_size_bytes), PAGE_SIZE); - - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_CP_ME]; - info->ucode_id = AMDGPU_UCODE_ID_CP_ME; - info->fw = adev->gfx.me_fw; - header = (const struct common_firmware_header *)info->fw->data; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(header->ucode_size_bytes), PAGE_SIZE); - - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_CP_MEC1]; - info->ucode_id = AMDGPU_UCODE_ID_CP_MEC1; - info->fw = adev->gfx.mec_fw; - header = (const struct common_firmware_header *)info->fw->data; - cp_hdr = (const struct gfx_firmware_header_v1_0 *)info->fw->data; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(header->ucode_size_bytes) - - le32_to_cpu(cp_hdr->jt_size) * 4, PAGE_SIZE); - - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_CP_MEC1_JT]; - info->ucode_id = AMDGPU_UCODE_ID_CP_MEC1_JT; - info->fw = adev->gfx.mec_fw; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(cp_hdr->jt_size) * 4, PAGE_SIZE); - } - - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_RLC_G]; - info->ucode_id = AMDGPU_UCODE_ID_RLC_G; - info->fw = adev->gfx.rlc_fw; - if (info->fw) { - header = (const struct common_firmware_header *)info->fw->data; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(header->ucode_size_bytes), PAGE_SIZE); - } - if (adev->gfx.rlc.save_restore_list_gpm_size_bytes && - adev->gfx.rlc.save_restore_list_srm_size_bytes) { - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_RLC_RESTORE_LIST_GPM_MEM]; - info->ucode_id = AMDGPU_UCODE_ID_RLC_RESTORE_LIST_GPM_MEM; - info->fw = adev->gfx.rlc_fw; - adev->firmware.fw_size += - ALIGN(adev->gfx.rlc.save_restore_list_gpm_size_bytes, PAGE_SIZE); - - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_RLC_RESTORE_LIST_SRM_MEM]; - info->ucode_id = AMDGPU_UCODE_ID_RLC_RESTORE_LIST_SRM_MEM; - info->fw = adev->gfx.rlc_fw; - adev->firmware.fw_size += - ALIGN(adev->gfx.rlc.save_restore_list_srm_size_bytes, PAGE_SIZE); - } - - if (adev->gfx.rlc.rlc_iram_ucode_size_bytes && - adev->gfx.rlc.rlc_dram_ucode_size_bytes) { - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_RLC_IRAM]; - info->ucode_id = AMDGPU_UCODE_ID_RLC_IRAM; - info->fw = adev->gfx.rlc_fw; - adev->firmware.fw_size += - ALIGN(adev->gfx.rlc.rlc_iram_ucode_size_bytes, PAGE_SIZE); - - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_RLC_DRAM]; - info->ucode_id = AMDGPU_UCODE_ID_RLC_DRAM; - info->fw = adev->gfx.rlc_fw; - adev->firmware.fw_size += - ALIGN(adev->gfx.rlc.rlc_dram_ucode_size_bytes, PAGE_SIZE); - } - - if (adev->gfx.rlc.rlcp_ucode_size_bytes) { - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_RLC_P]; - info->ucode_id = AMDGPU_UCODE_ID_RLC_P; - info->fw = adev->gfx.rlc_fw; - adev->firmware.fw_size += - ALIGN(adev->gfx.rlc.rlcp_ucode_size_bytes, PAGE_SIZE); - } - - if (adev->gfx.rlc.rlcv_ucode_size_bytes) { - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_RLC_V]; - info->ucode_id = AMDGPU_UCODE_ID_RLC_V; - info->fw = adev->gfx.rlc_fw; - adev->firmware.fw_size += - ALIGN(adev->gfx.rlc.rlcv_ucode_size_bytes, PAGE_SIZE); - } - } - out: if (err) { dev_err(adev->dev, - "gfx11: Failed to load firmware \"%s\"\n", + "gfx11: Failed to init firmware \"%s\"\n", fw_name); release_firmware(adev->gfx.pfp_fw); adev->gfx.pfp_fw = NULL; @@ -1842,7 +1571,7 @@ static void gfx_v11_0_init_compute_vmid(struct amdgpu_device *adev) WREG32_SOC15(GC, 0, regSH_MEM_BASES, sh_mem_bases); /* Enable trap for each kfd vmid. */ - data = RREG32(SOC15_REG_OFFSET(GC, 0, regSPI_GDBG_PER_VMID_CNTL)); + data = RREG32_SOC15(GC, 0, regSPI_GDBG_PER_VMID_CNTL); data = REG_SET_FIELD(data, SPI_GDBG_PER_VMID_CNTL, TRAP_EN, 1); } soc21_grbm_select(adev, 0, 0, 0, 0); @@ -5245,6 +4974,8 @@ static void gfx_v11_0_update_spm_vmid(struct amdgpu_device *adev, unsigned vmid) { u32 reg, data; + amdgpu_gfx_off_ctrl(adev, false); + reg = SOC15_REG_OFFSET(GC, 0, regRLC_SPM_MC_CNTL); if (amdgpu_sriov_is_pp_one_vf(adev)) data = RREG32_NO_KIQ(reg); @@ -5258,6 +4989,8 @@ static void gfx_v11_0_update_spm_vmid(struct amdgpu_device *adev, unsigned vmid) WREG32_SOC15_NO_KIQ(GC, 0, regRLC_SPM_MC_CNTL, data); else WREG32_SOC15(GC, 0, regRLC_SPM_MC_CNTL, data); + + amdgpu_gfx_off_ctrl(adev, true); } static const struct amdgpu_rlc_funcs gfx_v11_0_rlc_funcs = { @@ -5343,6 +5076,7 @@ static int gfx_v11_0_set_clockgating_state(void *handle, case IP_VERSION(11, 0, 0): case IP_VERSION(11, 0, 1): case IP_VERSION(11, 0, 2): + case IP_VERSION(11, 0, 3): gfx_v11_0_update_gfx_clock_gating(adev, state == AMD_CG_STATE_GATE); break; diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c index 1d6d3a852a0b..0320be4a5fc6 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c @@ -1091,27 +1091,6 @@ static void gfx_v9_0_free_microcode(struct amdgpu_device *adev) kfree(adev->gfx.rlc.register_list_format); } -static void gfx_v9_0_init_rlc_ext_microcode(struct amdgpu_device *adev) -{ - const struct rlc_firmware_header_v2_1 *rlc_hdr; - - rlc_hdr = (const struct rlc_firmware_header_v2_1 *)adev->gfx.rlc_fw->data; - adev->gfx.rlc_srlc_fw_version = le32_to_cpu(rlc_hdr->save_restore_list_cntl_ucode_ver); - adev->gfx.rlc_srlc_feature_version = le32_to_cpu(rlc_hdr->save_restore_list_cntl_feature_ver); - adev->gfx.rlc.save_restore_list_cntl_size_bytes = le32_to_cpu(rlc_hdr->save_restore_list_cntl_size_bytes); - adev->gfx.rlc.save_restore_list_cntl = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->save_restore_list_cntl_offset_bytes); - adev->gfx.rlc_srlg_fw_version = le32_to_cpu(rlc_hdr->save_restore_list_gpm_ucode_ver); - adev->gfx.rlc_srlg_feature_version = le32_to_cpu(rlc_hdr->save_restore_list_gpm_feature_ver); - adev->gfx.rlc.save_restore_list_gpm_size_bytes = le32_to_cpu(rlc_hdr->save_restore_list_gpm_size_bytes); - adev->gfx.rlc.save_restore_list_gpm = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->save_restore_list_gpm_offset_bytes); - adev->gfx.rlc_srls_fw_version = le32_to_cpu(rlc_hdr->save_restore_list_srm_ucode_ver); - adev->gfx.rlc_srls_feature_version = le32_to_cpu(rlc_hdr->save_restore_list_srm_feature_ver); - adev->gfx.rlc.save_restore_list_srm_size_bytes = le32_to_cpu(rlc_hdr->save_restore_list_srm_size_bytes); - adev->gfx.rlc.save_restore_list_srm = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->save_restore_list_srm_offset_bytes); - adev->gfx.rlc.reg_list_format_direct_reg_list_length = - le32_to_cpu(rlc_hdr->reg_list_format_direct_reg_list_length); -} - static void gfx_v9_0_check_fw_write_wait(struct amdgpu_device *adev) { adev->gfx.me_fw_write_wait = false; @@ -1273,9 +1252,6 @@ static int gfx_v9_0_init_cp_gfx_microcode(struct amdgpu_device *adev, { char fw_name[30]; int err; - struct amdgpu_firmware_info *info = NULL; - const struct common_firmware_header *header = NULL; - const struct gfx_firmware_header_v1_0 *cp_hdr; snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_pfp.bin", chip_name); err = request_firmware(&adev->gfx.pfp_fw, fw_name, adev->dev); @@ -1284,9 +1260,7 @@ static int gfx_v9_0_init_cp_gfx_microcode(struct amdgpu_device *adev, err = amdgpu_ucode_validate(adev->gfx.pfp_fw); if (err) goto out; - cp_hdr = (const struct gfx_firmware_header_v1_0 *)adev->gfx.pfp_fw->data; - adev->gfx.pfp_fw_version = le32_to_cpu(cp_hdr->header.ucode_version); - adev->gfx.pfp_feature_version = le32_to_cpu(cp_hdr->ucode_feature_version); + amdgpu_gfx_cp_init_microcode(adev, AMDGPU_UCODE_ID_CP_PFP); snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_me.bin", chip_name); err = request_firmware(&adev->gfx.me_fw, fw_name, adev->dev); @@ -1295,9 +1269,7 @@ static int gfx_v9_0_init_cp_gfx_microcode(struct amdgpu_device *adev, err = amdgpu_ucode_validate(adev->gfx.me_fw); if (err) goto out; - cp_hdr = (const struct gfx_firmware_header_v1_0 *)adev->gfx.me_fw->data; - adev->gfx.me_fw_version = le32_to_cpu(cp_hdr->header.ucode_version); - adev->gfx.me_feature_version = le32_to_cpu(cp_hdr->ucode_feature_version); + amdgpu_gfx_cp_init_microcode(adev, AMDGPU_UCODE_ID_CP_ME); snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_ce.bin", chip_name); err = request_firmware(&adev->gfx.ce_fw, fw_name, adev->dev); @@ -1306,37 +1278,12 @@ static int gfx_v9_0_init_cp_gfx_microcode(struct amdgpu_device *adev, err = amdgpu_ucode_validate(adev->gfx.ce_fw); if (err) goto out; - cp_hdr = (const struct gfx_firmware_header_v1_0 *)adev->gfx.ce_fw->data; - adev->gfx.ce_fw_version = le32_to_cpu(cp_hdr->header.ucode_version); - adev->gfx.ce_feature_version = le32_to_cpu(cp_hdr->ucode_feature_version); - - if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) { - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_CP_PFP]; - info->ucode_id = AMDGPU_UCODE_ID_CP_PFP; - info->fw = adev->gfx.pfp_fw; - header = (const struct common_firmware_header *)info->fw->data; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(header->ucode_size_bytes), PAGE_SIZE); - - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_CP_ME]; - info->ucode_id = AMDGPU_UCODE_ID_CP_ME; - info->fw = adev->gfx.me_fw; - header = (const struct common_firmware_header *)info->fw->data; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(header->ucode_size_bytes), PAGE_SIZE); - - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_CP_CE]; - info->ucode_id = AMDGPU_UCODE_ID_CP_CE; - info->fw = adev->gfx.ce_fw; - header = (const struct common_firmware_header *)info->fw->data; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(header->ucode_size_bytes), PAGE_SIZE); - } + amdgpu_gfx_cp_init_microcode(adev, AMDGPU_UCODE_ID_CP_CE); out: if (err) { dev_err(adev->dev, - "gfx9: Failed to load firmware \"%s\"\n", + "gfx9: Failed to init firmware \"%s\"\n", fw_name); release_firmware(adev->gfx.pfp_fw); adev->gfx.pfp_fw = NULL; @@ -1353,11 +1300,7 @@ static int gfx_v9_0_init_rlc_microcode(struct amdgpu_device *adev, { char fw_name[30]; int err; - struct amdgpu_firmware_info *info = NULL; - const struct common_firmware_header *header = NULL; const struct rlc_firmware_header_v2_0 *rlc_hdr; - unsigned int *tmp = NULL; - unsigned int i = 0; uint16_t version_major; uint16_t version_minor; uint32_t smu_version; @@ -1386,92 +1329,17 @@ static int gfx_v9_0_init_rlc_microcode(struct amdgpu_device *adev, if (err) goto out; err = amdgpu_ucode_validate(adev->gfx.rlc_fw); + if (err) + goto out; rlc_hdr = (const struct rlc_firmware_header_v2_0 *)adev->gfx.rlc_fw->data; version_major = le16_to_cpu(rlc_hdr->header.header_version_major); version_minor = le16_to_cpu(rlc_hdr->header.header_version_minor); - if (version_major == 2 && version_minor == 1) - adev->gfx.rlc.is_rlc_v2_1 = true; - - adev->gfx.rlc_fw_version = le32_to_cpu(rlc_hdr->header.ucode_version); - adev->gfx.rlc_feature_version = le32_to_cpu(rlc_hdr->ucode_feature_version); - adev->gfx.rlc.save_and_restore_offset = - le32_to_cpu(rlc_hdr->save_and_restore_offset); - adev->gfx.rlc.clear_state_descriptor_offset = - le32_to_cpu(rlc_hdr->clear_state_descriptor_offset); - adev->gfx.rlc.avail_scratch_ram_locations = - le32_to_cpu(rlc_hdr->avail_scratch_ram_locations); - adev->gfx.rlc.reg_restore_list_size = - le32_to_cpu(rlc_hdr->reg_restore_list_size); - adev->gfx.rlc.reg_list_format_start = - le32_to_cpu(rlc_hdr->reg_list_format_start); - adev->gfx.rlc.reg_list_format_separate_start = - le32_to_cpu(rlc_hdr->reg_list_format_separate_start); - adev->gfx.rlc.starting_offsets_start = - le32_to_cpu(rlc_hdr->starting_offsets_start); - adev->gfx.rlc.reg_list_format_size_bytes = - le32_to_cpu(rlc_hdr->reg_list_format_size_bytes); - adev->gfx.rlc.reg_list_size_bytes = - le32_to_cpu(rlc_hdr->reg_list_size_bytes); - adev->gfx.rlc.register_list_format = - kmalloc(adev->gfx.rlc.reg_list_format_size_bytes + - adev->gfx.rlc.reg_list_size_bytes, GFP_KERNEL); - if (!adev->gfx.rlc.register_list_format) { - err = -ENOMEM; - goto out; - } - - tmp = (unsigned int *)((uintptr_t)rlc_hdr + - le32_to_cpu(rlc_hdr->reg_list_format_array_offset_bytes)); - for (i = 0 ; i < (adev->gfx.rlc.reg_list_format_size_bytes >> 2); i++) - adev->gfx.rlc.register_list_format[i] = le32_to_cpu(tmp[i]); - - adev->gfx.rlc.register_restore = adev->gfx.rlc.register_list_format + i; - - tmp = (unsigned int *)((uintptr_t)rlc_hdr + - le32_to_cpu(rlc_hdr->reg_list_array_offset_bytes)); - for (i = 0 ; i < (adev->gfx.rlc.reg_list_size_bytes >> 2); i++) - adev->gfx.rlc.register_restore[i] = le32_to_cpu(tmp[i]); - - if (adev->gfx.rlc.is_rlc_v2_1) - gfx_v9_0_init_rlc_ext_microcode(adev); - - if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) { - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_RLC_G]; - info->ucode_id = AMDGPU_UCODE_ID_RLC_G; - info->fw = adev->gfx.rlc_fw; - header = (const struct common_firmware_header *)info->fw->data; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(header->ucode_size_bytes), PAGE_SIZE); - - if (adev->gfx.rlc.is_rlc_v2_1 && - adev->gfx.rlc.save_restore_list_cntl_size_bytes && - adev->gfx.rlc.save_restore_list_gpm_size_bytes && - adev->gfx.rlc.save_restore_list_srm_size_bytes) { - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_RLC_RESTORE_LIST_CNTL]; - info->ucode_id = AMDGPU_UCODE_ID_RLC_RESTORE_LIST_CNTL; - info->fw = adev->gfx.rlc_fw; - adev->firmware.fw_size += - ALIGN(adev->gfx.rlc.save_restore_list_cntl_size_bytes, PAGE_SIZE); - - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_RLC_RESTORE_LIST_GPM_MEM]; - info->ucode_id = AMDGPU_UCODE_ID_RLC_RESTORE_LIST_GPM_MEM; - info->fw = adev->gfx.rlc_fw; - adev->firmware.fw_size += - ALIGN(adev->gfx.rlc.save_restore_list_gpm_size_bytes, PAGE_SIZE); - - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_RLC_RESTORE_LIST_SRM_MEM]; - info->ucode_id = AMDGPU_UCODE_ID_RLC_RESTORE_LIST_SRM_MEM; - info->fw = adev->gfx.rlc_fw; - adev->firmware.fw_size += - ALIGN(adev->gfx.rlc.save_restore_list_srm_size_bytes, PAGE_SIZE); - } - } - + err = amdgpu_gfx_rlc_init_microcode(adev, version_major, version_minor); out: if (err) { dev_err(adev->dev, - "gfx9: Failed to load firmware \"%s\"\n", + "gfx9: Failed to init firmware \"%s\"\n", fw_name); release_firmware(adev->gfx.rlc_fw); adev->gfx.rlc_fw = NULL; @@ -1494,9 +1362,6 @@ static int gfx_v9_0_init_cp_compute_microcode(struct amdgpu_device *adev, { char fw_name[30]; int err; - struct amdgpu_firmware_info *info = NULL; - const struct common_firmware_header *header = NULL; - const struct gfx_firmware_header_v1_0 *cp_hdr; if (amdgpu_sriov_vf(adev) && (adev->asic_type == CHIP_ALDEBARAN)) snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_sjt_mec.bin", chip_name); @@ -1509,10 +1374,8 @@ static int gfx_v9_0_init_cp_compute_microcode(struct amdgpu_device *adev, err = amdgpu_ucode_validate(adev->gfx.mec_fw); if (err) goto out; - cp_hdr = (const struct gfx_firmware_header_v1_0 *)adev->gfx.mec_fw->data; - adev->gfx.mec_fw_version = le32_to_cpu(cp_hdr->header.ucode_version); - adev->gfx.mec_feature_version = le32_to_cpu(cp_hdr->ucode_feature_version); - + amdgpu_gfx_cp_init_microcode(adev, AMDGPU_UCODE_ID_CP_MEC1); + amdgpu_gfx_cp_init_microcode(adev, AMDGPU_UCODE_ID_CP_MEC1_JT); if (gfx_v9_0_load_mec2_fw_bin_support(adev)) { if (amdgpu_sriov_vf(adev) && (adev->asic_type == CHIP_ALDEBARAN)) @@ -1525,12 +1388,8 @@ static int gfx_v9_0_init_cp_compute_microcode(struct amdgpu_device *adev, err = amdgpu_ucode_validate(adev->gfx.mec2_fw); if (err) goto out; - cp_hdr = (const struct gfx_firmware_header_v1_0 *) - adev->gfx.mec2_fw->data; - adev->gfx.mec2_fw_version = - le32_to_cpu(cp_hdr->header.ucode_version); - adev->gfx.mec2_feature_version = - le32_to_cpu(cp_hdr->ucode_feature_version); + amdgpu_gfx_cp_init_microcode(adev, AMDGPU_UCODE_ID_CP_MEC2); + amdgpu_gfx_cp_init_microcode(adev, AMDGPU_UCODE_ID_CP_MEC2_JT); } else { err = 0; adev->gfx.mec2_fw = NULL; @@ -1540,49 +1399,12 @@ static int gfx_v9_0_init_cp_compute_microcode(struct amdgpu_device *adev, adev->gfx.mec2_feature_version = adev->gfx.mec_feature_version; } - if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) { - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_CP_MEC1]; - info->ucode_id = AMDGPU_UCODE_ID_CP_MEC1; - info->fw = adev->gfx.mec_fw; - header = (const struct common_firmware_header *)info->fw->data; - cp_hdr = (const struct gfx_firmware_header_v1_0 *)info->fw->data; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(header->ucode_size_bytes) - le32_to_cpu(cp_hdr->jt_size) * 4, PAGE_SIZE); - - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_CP_MEC1_JT]; - info->ucode_id = AMDGPU_UCODE_ID_CP_MEC1_JT; - info->fw = adev->gfx.mec_fw; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(cp_hdr->jt_size) * 4, PAGE_SIZE); - - if (adev->gfx.mec2_fw) { - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_CP_MEC2]; - info->ucode_id = AMDGPU_UCODE_ID_CP_MEC2; - info->fw = adev->gfx.mec2_fw; - header = (const struct common_firmware_header *)info->fw->data; - cp_hdr = (const struct gfx_firmware_header_v1_0 *)info->fw->data; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(header->ucode_size_bytes) - le32_to_cpu(cp_hdr->jt_size) * 4, PAGE_SIZE); - - /* TODO: Determine if MEC2 JT FW loading can be removed - for all GFX V9 asic and above */ - if (gfx_v9_0_load_mec2_fw_bin_support(adev)) { - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_CP_MEC2_JT]; - info->ucode_id = AMDGPU_UCODE_ID_CP_MEC2_JT; - info->fw = adev->gfx.mec2_fw; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(cp_hdr->jt_size) * 4, - PAGE_SIZE); - } - } - } - out: gfx_v9_0_check_if_need_gfxoff(adev); gfx_v9_0_check_fw_write_wait(adev); if (err) { dev_err(adev->dev, - "gfx9: Failed to load firmware \"%s\"\n", + "gfx9: Failed to init firmware \"%s\"\n", fw_name); release_firmware(adev->gfx.mec_fw); adev->gfx.mec_fw = NULL; @@ -5607,7 +5429,7 @@ static void gfx_v9_0_ring_emit_patch_cond_exec(struct amdgpu_ring *ring, unsigne BUG_ON(offset > ring->buf_mask); BUG_ON(ring->ring[offset] != 0x55aa55aa); - cur = (ring->wptr & ring->buf_mask) - 1; + cur = (ring->wptr - 1) & ring->buf_mask; if (likely(cur > offset)) ring->ring[offset] = cur - offset; else diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c index 846ccb6cf07d..66dfb574cc7d 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c @@ -186,6 +186,10 @@ static void gmc_v11_0_flush_vm_hub(struct amdgpu_device *adev, uint32_t vmid, /* Use register 17 for GART */ const unsigned eng = 17; unsigned int i; + unsigned char hub_ip = 0; + + hub_ip = (vmhub == AMDGPU_GFXHUB_0) ? + GC_HWIP : MMHUB_HWIP; spin_lock(&adev->gmc.invalidate_lock); /* @@ -199,8 +203,8 @@ static void gmc_v11_0_flush_vm_hub(struct amdgpu_device *adev, uint32_t vmid, if (use_semaphore) { for (i = 0; i < adev->usec_timeout; i++) { /* a read return value of 1 means semaphore acuqire */ - tmp = RREG32_NO_KIQ(hub->vm_inv_eng0_sem + - hub->eng_distance * eng); + tmp = RREG32_RLC_NO_KIQ(hub->vm_inv_eng0_sem + + hub->eng_distance * eng, hub_ip); if (tmp & 0x1) break; udelay(1); @@ -210,12 +214,12 @@ static void gmc_v11_0_flush_vm_hub(struct amdgpu_device *adev, uint32_t vmid, DRM_ERROR("Timeout waiting for sem acquire in VM flush!\n"); } - WREG32_NO_KIQ(hub->vm_inv_eng0_req + hub->eng_distance * eng, inv_req); + WREG32_RLC_NO_KIQ(hub->vm_inv_eng0_req + hub->eng_distance * eng, inv_req, hub_ip); /* Wait for ACK with a delay.*/ for (i = 0; i < adev->usec_timeout; i++) { - tmp = RREG32_NO_KIQ(hub->vm_inv_eng0_ack + - hub->eng_distance * eng); + tmp = RREG32_RLC_NO_KIQ(hub->vm_inv_eng0_ack + + hub->eng_distance * eng, hub_ip); tmp &= 1 << vmid; if (tmp) break; @@ -229,8 +233,8 @@ static void gmc_v11_0_flush_vm_hub(struct amdgpu_device *adev, uint32_t vmid, * add semaphore release after invalidation, * write with 0 means semaphore release */ - WREG32_NO_KIQ(hub->vm_inv_eng0_sem + - hub->eng_distance * eng, 0); + WREG32_RLC_NO_KIQ(hub->vm_inv_eng0_sem + + hub->eng_distance * eng, 0, hub_ip); /* Issue additional private vm invalidation to MMHUB */ if ((vmhub != AMDGPU_GFXHUB_0) && diff --git a/drivers/gpu/drm/amd/amdgpu/mes_v11_0.c b/drivers/gpu/drm/amd/amdgpu/mes_v11_0.c index b64cd46a159a..fef7d020bc5f 100644 --- a/drivers/gpu/drm/amd/amdgpu/mes_v11_0.c +++ b/drivers/gpu/drm/amd/amdgpu/mes_v11_0.c @@ -187,6 +187,19 @@ static int mes_v11_0_add_hw_queue(struct amdgpu_mes *mes, 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; + mes_add_queue_pkt.gds_size = input->queue_size; + + if (!(((adev->mes.sched_version & AMDGPU_MES_VERSION_MASK) >= 4) && + (adev->ip_versions[GC_HWIP][0] >= IP_VERSION(11, 0, 0)) && + (adev->ip_versions[GC_HWIP][0] <= IP_VERSION(11, 0, 3)))) + 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; + mes_add_queue_pkt.gds_size = input->queue_size; + return mes_v11_0_submit_pkt_and_poll_completion(mes, &mes_add_queue_pkt, sizeof(mes_add_queue_pkt), offsetof(union MESAPI__ADD_QUEUE, api_status)); @@ -1143,6 +1156,42 @@ static int mes_v11_0_sw_fini(void *handle) return 0; } +static void mes_v11_0_kiq_dequeue_sched(struct amdgpu_device *adev) +{ + uint32_t data; + int i; + + mutex_lock(&adev->srbm_mutex); + soc21_grbm_select(adev, 3, AMDGPU_MES_SCHED_PIPE, 0, 0); + + /* disable the queue if it's active */ + if (RREG32_SOC15(GC, 0, regCP_HQD_ACTIVE) & 1) { + WREG32_SOC15(GC, 0, regCP_HQD_DEQUEUE_REQUEST, 1); + for (i = 0; i < adev->usec_timeout; i++) { + if (!(RREG32_SOC15(GC, 0, regCP_HQD_ACTIVE) & 1)) + break; + udelay(1); + } + } + data = RREG32_SOC15(GC, 0, regCP_HQD_PQ_DOORBELL_CONTROL); + data = REG_SET_FIELD(data, CP_HQD_PQ_DOORBELL_CONTROL, + DOORBELL_EN, 0); + data = REG_SET_FIELD(data, CP_HQD_PQ_DOORBELL_CONTROL, + DOORBELL_HIT, 1); + WREG32_SOC15(GC, 0, regCP_HQD_PQ_DOORBELL_CONTROL, data); + + WREG32_SOC15(GC, 0, regCP_HQD_PQ_DOORBELL_CONTROL, 0); + + WREG32_SOC15(GC, 0, regCP_HQD_PQ_WPTR_LO, 0); + WREG32_SOC15(GC, 0, regCP_HQD_PQ_WPTR_HI, 0); + WREG32_SOC15(GC, 0, regCP_HQD_PQ_RPTR, 0); + + soc21_grbm_select(adev, 0, 0, 0, 0); + mutex_unlock(&adev->srbm_mutex); + + adev->mes.ring.sched.ready = false; +} + static void mes_v11_0_kiq_setting(struct amdgpu_ring *ring) { uint32_t tmp; @@ -1194,6 +1243,9 @@ failure: static int mes_v11_0_kiq_hw_fini(struct amdgpu_device *adev) { + if (adev->mes.ring.sched.ready) + mes_v11_0_kiq_dequeue_sched(adev); + mes_v11_0_enable(adev, false); return 0; } @@ -1249,9 +1301,6 @@ failure: static int mes_v11_0_hw_fini(void *handle) { - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - adev->mes.ring.sched.ready = false; return 0; } @@ -1283,7 +1332,8 @@ static int mes_v11_0_late_init(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; - if (!amdgpu_in_reset(adev)) + if (!amdgpu_in_reset(adev) && + (adev->ip_versions[GC_HWIP][0] != IP_VERSION(11, 0, 3))) amdgpu_mes_self_test(adev); return 0; diff --git a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c index a2f04b249132..12906ba74462 100644 --- a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c +++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c @@ -290,7 +290,6 @@ flr_done: reset_context.method = AMD_RESET_METHOD_NONE; reset_context.reset_req_dev = adev; clear_bit(AMDGPU_NEED_FULL_RESET, &reset_context.flags); - clear_bit(AMDGPU_SKIP_MODE2_RESET, &reset_context.flags); amdgpu_device_gpu_recover(adev, NULL, &reset_context); } diff --git a/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c b/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c index a977f0027928..e07757eea7ad 100644 --- a/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c +++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c @@ -317,7 +317,6 @@ flr_done: reset_context.method = AMD_RESET_METHOD_NONE; reset_context.reset_req_dev = adev; clear_bit(AMDGPU_NEED_FULL_RESET, &reset_context.flags); - clear_bit(AMDGPU_SKIP_MODE2_RESET, &reset_context.flags); amdgpu_device_gpu_recover(adev, NULL, &reset_context); } diff --git a/drivers/gpu/drm/amd/amdgpu/mxgpu_vi.c b/drivers/gpu/drm/amd/amdgpu/mxgpu_vi.c index fd14fa9b9cd7..288c414babdf 100644 --- a/drivers/gpu/drm/amd/amdgpu/mxgpu_vi.c +++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_vi.c @@ -529,7 +529,6 @@ static void xgpu_vi_mailbox_flr_work(struct work_struct *work) reset_context.method = AMD_RESET_METHOD_NONE; reset_context.reset_req_dev = adev; clear_bit(AMDGPU_NEED_FULL_RESET, &reset_context.flags); - clear_bit(AMDGPU_SKIP_MODE2_RESET, &reset_context.flags); amdgpu_device_gpu_recover(adev, NULL, &reset_context); } diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c b/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c index 6bdffdc1c0b9..c52d246a1d96 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c @@ -342,14 +342,10 @@ static void sdma_v2_4_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 se */ static void sdma_v2_4_gfx_stop(struct amdgpu_device *adev) { - struct amdgpu_ring *sdma0 = &adev->sdma.instance[0].ring; - struct amdgpu_ring *sdma1 = &adev->sdma.instance[1].ring; u32 rb_cntl, ib_cntl; int i; - if ((adev->mman.buffer_funcs_ring == sdma0) || - (adev->mman.buffer_funcs_ring == sdma1)) - amdgpu_ttm_set_buffer_funcs_status(adev, false); + amdgpu_sdma_unset_buffer_funcs_helper(adev); for (i = 0; i < adev->sdma.num_instances; i++) { rb_cntl = RREG32(mmSDMA0_GFX_RB_CNTL + sdma_offsets[i]); diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c index 2584fa3cb13e..486d9b5c1b9e 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c @@ -516,14 +516,10 @@ static void sdma_v3_0_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 se */ static void sdma_v3_0_gfx_stop(struct amdgpu_device *adev) { - struct amdgpu_ring *sdma0 = &adev->sdma.instance[0].ring; - struct amdgpu_ring *sdma1 = &adev->sdma.instance[1].ring; u32 rb_cntl, ib_cntl; int i; - if ((adev->mman.buffer_funcs_ring == sdma0) || - (adev->mman.buffer_funcs_ring == sdma1)) - amdgpu_ttm_set_buffer_funcs_status(adev, false); + amdgpu_sdma_unset_buffer_funcs_helper(adev); for (i = 0; i < adev->sdma.num_instances; i++) { rb_cntl = RREG32(mmSDMA0_GFX_RB_CNTL + sdma_offsets[i]); diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c index 7fe8bf3417db..1122bd4eae98 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c @@ -561,44 +561,6 @@ static void sdma_v4_0_setup_ulv(struct amdgpu_device *adev) } } -static int sdma_v4_0_init_inst_ctx(struct amdgpu_sdma_instance *sdma_inst) -{ - int err = 0; - const struct sdma_firmware_header_v1_0 *hdr; - - err = amdgpu_ucode_validate(sdma_inst->fw); - if (err) - return err; - - hdr = (const struct sdma_firmware_header_v1_0 *)sdma_inst->fw->data; - sdma_inst->fw_version = le32_to_cpu(hdr->header.ucode_version); - sdma_inst->feature_version = le32_to_cpu(hdr->ucode_feature_version); - - if (sdma_inst->feature_version >= 20) - sdma_inst->burst_nop = true; - - return 0; -} - -static void sdma_v4_0_destroy_inst_ctx(struct amdgpu_device *adev) -{ - int i; - - for (i = 0; i < adev->sdma.num_instances; i++) { - release_firmware(adev->sdma.instance[i].fw); - adev->sdma.instance[i].fw = NULL; - - /* arcturus shares the same FW memory across - all SDMA isntances */ - if (adev->ip_versions[SDMA0_HWIP][0] == IP_VERSION(4, 2, 2) || - adev->ip_versions[SDMA0_HWIP][0] == IP_VERSION(4, 4, 0)) - break; - } - - memset((void *)adev->sdma.instance, 0, - sizeof(struct amdgpu_sdma_instance) * AMDGPU_MAX_SDMA_INSTANCES); -} - /** * sdma_v4_0_init_microcode - load ucode images from disk * @@ -615,9 +577,7 @@ static int sdma_v4_0_init_microcode(struct amdgpu_device *adev) { const char *chip_name; char fw_name[30]; - int err = 0, i; - struct amdgpu_firmware_info *info = NULL; - const struct common_firmware_header *header = NULL; + int ret, i; DRM_DEBUG("\n"); @@ -656,58 +616,25 @@ static int sdma_v4_0_init_microcode(struct amdgpu_device *adev) BUG(); } - snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_sdma.bin", chip_name); - - err = request_firmware(&adev->sdma.instance[0].fw, fw_name, adev->dev); - if (err) - goto out; - - err = sdma_v4_0_init_inst_ctx(&adev->sdma.instance[0]); - if (err) - goto out; - - for (i = 1; i < adev->sdma.num_instances; i++) { + for (i = 0; i < adev->sdma.num_instances; i++) { + if (i == 0) + snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_sdma.bin", chip_name); + else + snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_sdma%d.bin", chip_name, i); if (adev->ip_versions[SDMA0_HWIP][0] == IP_VERSION(4, 2, 2) || adev->ip_versions[SDMA0_HWIP][0] == IP_VERSION(4, 4, 0)) { /* Acturus & Aldebaran will leverage the same FW memory for every SDMA instance */ - memcpy((void *)&adev->sdma.instance[i], - (void *)&adev->sdma.instance[0], - sizeof(struct amdgpu_sdma_instance)); - } - else { - snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_sdma%d.bin", chip_name, i); - - err = request_firmware(&adev->sdma.instance[i].fw, fw_name, adev->dev); - if (err) - goto out; - - err = sdma_v4_0_init_inst_ctx(&adev->sdma.instance[i]); - if (err) - goto out; - } - } - - DRM_DEBUG("psp_load == '%s'\n", - adev->firmware.load_type == AMDGPU_FW_LOAD_PSP ? "true" : "false"); - - if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) { - for (i = 0; i < adev->sdma.num_instances; i++) { - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_SDMA0 + i]; - info->ucode_id = AMDGPU_UCODE_ID_SDMA0 + i; - info->fw = adev->sdma.instance[i].fw; - header = (const struct common_firmware_header *)info->fw->data; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(header->ucode_size_bytes), PAGE_SIZE); + ret = amdgpu_sdma_init_microcode(adev, fw_name, 0, true); + break; + } else { + ret = amdgpu_sdma_init_microcode(adev, fw_name, i, false); + if (ret) + return ret; } } -out: - if (err) { - DRM_ERROR("sdma_v4_0: Failed to load firmware \"%s\"\n", fw_name); - sdma_v4_0_destroy_inst_ctx(adev); - } - return err; + return ret; } /** @@ -988,18 +915,12 @@ static void sdma_v4_0_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 se */ static void sdma_v4_0_gfx_stop(struct amdgpu_device *adev) { - struct amdgpu_ring *sdma[AMDGPU_MAX_SDMA_INSTANCES]; u32 rb_cntl, ib_cntl; - int i, unset = 0; - - for (i = 0; i < adev->sdma.num_instances; i++) { - sdma[i] = &adev->sdma.instance[i].ring; + int i; - if ((adev->mman.buffer_funcs_ring == sdma[i]) && unset != 1) { - amdgpu_ttm_set_buffer_funcs_status(adev, false); - unset = 1; - } + amdgpu_sdma_unset_buffer_funcs_helper(adev); + for (i = 0; i < adev->sdma.num_instances; i++) { rb_cntl = RREG32_SDMA(i, mmSDMA0_GFX_RB_CNTL); rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_GFX_RB_CNTL, RB_ENABLE, 0); WREG32_SDMA(i, mmSDMA0_GFX_RB_CNTL, rb_cntl); @@ -1030,20 +951,12 @@ static void sdma_v4_0_rlc_stop(struct amdgpu_device *adev) */ static void sdma_v4_0_page_stop(struct amdgpu_device *adev) { - struct amdgpu_ring *sdma[AMDGPU_MAX_SDMA_INSTANCES]; u32 rb_cntl, ib_cntl; int i; - bool unset = false; - for (i = 0; i < adev->sdma.num_instances; i++) { - sdma[i] = &adev->sdma.instance[i].page; - - if ((adev->mman.buffer_funcs_ring == sdma[i]) && - (!unset)) { - amdgpu_ttm_set_buffer_funcs_status(adev, false); - unset = true; - } + amdgpu_sdma_unset_buffer_funcs_helper(adev); + for (i = 0; i < adev->sdma.num_instances; i++) { rb_cntl = RREG32_SDMA(i, mmSDMA0_PAGE_RB_CNTL); rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_PAGE_RB_CNTL, RB_ENABLE, 0); @@ -1504,11 +1417,6 @@ static int sdma_v4_0_start(struct amdgpu_device *adev) WREG32_SDMA(i, mmSDMA0_CNTL, temp); if (!amdgpu_sriov_vf(adev)) { - ring = &adev->sdma.instance[i].ring; - adev->nbio.funcs->sdma_doorbell_range(adev, i, - ring->use_doorbell, ring->doorbell_index, - adev->doorbell_index.sdma_doorbell_range); - /* unhalt engine */ temp = RREG32_SDMA(i, mmSDMA0_F32_CNTL); temp = REG_SET_FIELD(temp, SDMA0_F32_CNTL, HALT, 0); @@ -2000,7 +1908,11 @@ static int sdma_v4_0_sw_fini(void *handle) amdgpu_ring_fini(&adev->sdma.instance[i].page); } - sdma_v4_0_destroy_inst_ctx(adev); + if (adev->ip_versions[SDMA0_HWIP][0] == IP_VERSION(4, 2, 0) || + adev->ip_versions[SDMA0_HWIP][0] == IP_VERSION(4, 4, 0)) + amdgpu_sdma_destroy_inst_ctx(adev, true); + else + amdgpu_sdma_destroy_inst_ctx(adev, false); return 0; } @@ -2023,8 +1935,11 @@ static int sdma_v4_0_hw_fini(void *handle) struct amdgpu_device *adev = (struct amdgpu_device *)handle; int i; - if (amdgpu_sriov_vf(adev)) + if (amdgpu_sriov_vf(adev)) { + /* disable the scheduler for SDMA */ + amdgpu_sdma_unset_buffer_funcs_helper(adev); return 0; + } for (i = 0; i < adev->sdma.num_instances; i++) { amdgpu_irq_put(adev, &adev->sdma.ecc_irq, diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c index a019ac92edb7..d4d9f196db83 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c @@ -240,10 +240,7 @@ static int sdma_v5_0_init_microcode(struct amdgpu_device *adev) { const char *chip_name; char fw_name[40]; - int err = 0, i; - struct amdgpu_firmware_info *info = NULL; - const struct common_firmware_header *header = NULL; - const struct sdma_firmware_header_v1_0 *hdr; + int ret, i; if (amdgpu_sriov_vf(adev) && (adev->ip_versions[SDMA0_HWIP][0] == IP_VERSION(5, 0, 5))) return 0; @@ -272,38 +269,12 @@ static int sdma_v5_0_init_microcode(struct amdgpu_device *adev) snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_sdma.bin", chip_name); else snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_sdma1.bin", chip_name); - err = request_firmware(&adev->sdma.instance[i].fw, fw_name, adev->dev); - if (err) - goto out; - err = amdgpu_ucode_validate(adev->sdma.instance[i].fw); - if (err) - goto out; - hdr = (const struct sdma_firmware_header_v1_0 *)adev->sdma.instance[i].fw->data; - adev->sdma.instance[i].fw_version = le32_to_cpu(hdr->header.ucode_version); - adev->sdma.instance[i].feature_version = le32_to_cpu(hdr->ucode_feature_version); - if (adev->sdma.instance[i].feature_version >= 20) - adev->sdma.instance[i].burst_nop = true; - DRM_DEBUG("psp_load == '%s'\n", - adev->firmware.load_type == AMDGPU_FW_LOAD_PSP ? "true" : "false"); - - if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) { - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_SDMA0 + i]; - info->ucode_id = AMDGPU_UCODE_ID_SDMA0 + i; - info->fw = adev->sdma.instance[i].fw; - header = (const struct common_firmware_header *)info->fw->data; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(header->ucode_size_bytes), PAGE_SIZE); - } + ret = amdgpu_sdma_init_microcode(adev, fw_name, i, false); + if (ret) + return ret; } -out: - if (err) { - DRM_ERROR("sdma_v5_0: Failed to load firmware \"%s\"\n", fw_name); - for (i = 0; i < adev->sdma.num_instances; i++) { - release_firmware(adev->sdma.instance[i].fw); - adev->sdma.instance[i].fw = NULL; - } - } - return err; + + return ret; } static unsigned sdma_v5_0_ring_init_cond_exec(struct amdgpu_ring *ring) @@ -613,14 +584,10 @@ static void sdma_v5_0_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 se */ static void sdma_v5_0_gfx_stop(struct amdgpu_device *adev) { - struct amdgpu_ring *sdma0 = &adev->sdma.instance[0].ring; - struct amdgpu_ring *sdma1 = &adev->sdma.instance[1].ring; u32 rb_cntl, ib_cntl; int i; - if ((adev->mman.buffer_funcs_ring == sdma0) || - (adev->mman.buffer_funcs_ring == sdma1)) - amdgpu_ttm_set_buffer_funcs_status(adev, false); + amdgpu_sdma_unset_buffer_funcs_helper(adev); for (i = 0; i < adev->sdma.num_instances; i++) { rb_cntl = RREG32_SOC15_IP(GC, sdma_v5_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_CNTL)); @@ -1465,12 +1432,10 @@ static int sdma_v5_0_sw_fini(void *handle) struct amdgpu_device *adev = (struct amdgpu_device *)handle; int i; - for (i = 0; i < adev->sdma.num_instances; i++) { - release_firmware(adev->sdma.instance[i].fw); - adev->sdma.instance[i].fw = NULL; - + for (i = 0; i < adev->sdma.num_instances; i++) amdgpu_ring_fini(&adev->sdma.instance[i].ring); - } + + amdgpu_sdma_destroy_inst_ctx(adev, false); return 0; } @@ -1491,8 +1456,11 @@ static int sdma_v5_0_hw_fini(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; - if (amdgpu_sriov_vf(adev)) + if (amdgpu_sriov_vf(adev)) { + /* disable the scheduler for SDMA */ + amdgpu_sdma_unset_buffer_funcs_helper(adev); return 0; + } sdma_v5_0_ctx_switch_enable(adev, false); sdma_v5_0_enable(adev, false); diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c b/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c index 95689ef4be10..809eca54fc61 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c @@ -89,33 +89,6 @@ static u32 sdma_v5_2_get_reg_offset(struct amdgpu_device *adev, u32 instance, u3 return base + internal_offset; } -static int sdma_v5_2_init_inst_ctx(struct amdgpu_sdma_instance *sdma_inst) -{ - int err = 0; - const struct sdma_firmware_header_v1_0 *hdr; - - err = amdgpu_ucode_validate(sdma_inst->fw); - if (err) - return err; - - hdr = (const struct sdma_firmware_header_v1_0 *)sdma_inst->fw->data; - sdma_inst->fw_version = le32_to_cpu(hdr->header.ucode_version); - sdma_inst->feature_version = le32_to_cpu(hdr->ucode_feature_version); - - if (sdma_inst->feature_version >= 20) - sdma_inst->burst_nop = true; - - return 0; -} - -static void sdma_v5_2_destroy_inst_ctx(struct amdgpu_device *adev) -{ - release_firmware(adev->sdma.instance[0].fw); - - memset((void *)adev->sdma.instance, 0, - sizeof(struct amdgpu_sdma_instance) * AMDGPU_MAX_SDMA_INSTANCES); -} - /** * sdma_v5_2_init_microcode - load ucode images from disk * @@ -132,9 +105,6 @@ static int sdma_v5_2_init_microcode(struct amdgpu_device *adev) { const char *chip_name; char fw_name[40]; - int err = 0, i; - struct amdgpu_firmware_info *info = NULL; - const struct common_firmware_header *header = NULL; DRM_DEBUG("\n"); @@ -169,42 +139,7 @@ static int sdma_v5_2_init_microcode(struct amdgpu_device *adev) snprintf(fw_name, sizeof(fw_name), "amdgpu/%s.bin", chip_name); - err = request_firmware(&adev->sdma.instance[0].fw, fw_name, adev->dev); - if (err) - goto out; - - err = sdma_v5_2_init_inst_ctx(&adev->sdma.instance[0]); - if (err) - goto out; - - for (i = 1; i < adev->sdma.num_instances; i++) - memcpy((void *)&adev->sdma.instance[i], - (void *)&adev->sdma.instance[0], - sizeof(struct amdgpu_sdma_instance)); - - if (amdgpu_sriov_vf(adev) && (adev->ip_versions[SDMA0_HWIP][0] == IP_VERSION(5, 2, 0))) - return 0; - - DRM_DEBUG("psp_load == '%s'\n", - adev->firmware.load_type == AMDGPU_FW_LOAD_PSP ? "true" : "false"); - - if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) { - for (i = 0; i < adev->sdma.num_instances; i++) { - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_SDMA0 + i]; - info->ucode_id = AMDGPU_UCODE_ID_SDMA0 + i; - info->fw = adev->sdma.instance[i].fw; - header = (const struct common_firmware_header *)info->fw->data; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(header->ucode_size_bytes), PAGE_SIZE); - } - } - -out: - if (err) { - DRM_ERROR("sdma_v5_2: Failed to load firmware \"%s\"\n", fw_name); - sdma_v5_2_destroy_inst_ctx(adev); - } - return err; + return amdgpu_sdma_init_microcode(adev, fw_name, 0, true); } static unsigned sdma_v5_2_ring_init_cond_exec(struct amdgpu_ring *ring) @@ -479,18 +414,10 @@ static void sdma_v5_2_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 se */ static void sdma_v5_2_gfx_stop(struct amdgpu_device *adev) { - struct amdgpu_ring *sdma0 = &adev->sdma.instance[0].ring; - struct amdgpu_ring *sdma1 = &adev->sdma.instance[1].ring; - struct amdgpu_ring *sdma2 = &adev->sdma.instance[2].ring; - struct amdgpu_ring *sdma3 = &adev->sdma.instance[3].ring; u32 rb_cntl, ib_cntl; int i; - if ((adev->mman.buffer_funcs_ring == sdma0) || - (adev->mman.buffer_funcs_ring == sdma1) || - (adev->mman.buffer_funcs_ring == sdma2) || - (adev->mman.buffer_funcs_ring == sdma3)) - amdgpu_ttm_set_buffer_funcs_status(adev, false); + amdgpu_sdma_unset_buffer_funcs_helper(adev); for (i = 0; i < adev->sdma.num_instances; i++) { rb_cntl = RREG32_SOC15_IP(GC, sdma_v5_2_get_reg_offset(adev, i, mmSDMA0_GFX_RB_CNTL)); @@ -1406,7 +1333,7 @@ static int sdma_v5_2_sw_fini(void *handle) for (i = 0; i < adev->sdma.num_instances; i++) amdgpu_ring_fini(&adev->sdma.instance[i].ring); - sdma_v5_2_destroy_inst_ctx(adev); + amdgpu_sdma_destroy_inst_ctx(adev, true); return 0; } @@ -1422,8 +1349,11 @@ static int sdma_v5_2_hw_fini(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; - if (amdgpu_sriov_vf(adev)) + if (amdgpu_sriov_vf(adev)) { + /* disable the scheduler for SDMA */ + amdgpu_sdma_unset_buffer_funcs_helper(adev); return 0; + } sdma_v5_2_ctx_switch_enable(adev, false); sdma_v5_2_enable(adev, false); diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v6_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v6_0.c index 3a0b14be1a57..da3beb0bf2fa 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v6_0.c @@ -78,33 +78,6 @@ static u32 sdma_v6_0_get_reg_offset(struct amdgpu_device *adev, u32 instance, u3 return base + internal_offset; } -static int sdma_v6_0_init_inst_ctx(struct amdgpu_sdma_instance *sdma_inst) -{ - int err = 0; - const struct sdma_firmware_header_v2_0 *hdr; - - err = amdgpu_ucode_validate(sdma_inst->fw); - if (err) - return err; - - hdr = (const struct sdma_firmware_header_v2_0 *)sdma_inst->fw->data; - sdma_inst->fw_version = le32_to_cpu(hdr->header.ucode_version); - sdma_inst->feature_version = le32_to_cpu(hdr->ucode_feature_version); - - if (sdma_inst->feature_version >= 20) - sdma_inst->burst_nop = true; - - return 0; -} - -static void sdma_v6_0_destroy_inst_ctx(struct amdgpu_device *adev) -{ - release_firmware(adev->sdma.instance[0].fw); - - memset((void*)adev->sdma.instance, 0, - sizeof(struct amdgpu_sdma_instance) * AMDGPU_MAX_SDMA_INSTANCES); -} - /** * sdma_v6_0_init_microcode - load ucode images from disk * @@ -114,16 +87,10 @@ static void sdma_v6_0_destroy_inst_ctx(struct amdgpu_device *adev) * the driver (not loaded into hw). * Returns 0 on success, error on failure. */ - -// emulation only, won't work on real chip -// sdma 6.0.0 real chip need to use PSP to load firmware static int sdma_v6_0_init_microcode(struct amdgpu_device *adev) { char fw_name[30]; char ucode_prefix[30]; - int err = 0, i; - struct amdgpu_firmware_info *info = NULL; - const struct sdma_firmware_header_v2_0 *sdma_hdr; DRM_DEBUG("\n"); @@ -131,43 +98,7 @@ static int sdma_v6_0_init_microcode(struct amdgpu_device *adev) snprintf(fw_name, sizeof(fw_name), "amdgpu/%s.bin", ucode_prefix); - err = request_firmware(&adev->sdma.instance[0].fw, fw_name, adev->dev); - if (err) - goto out; - - err = sdma_v6_0_init_inst_ctx(&adev->sdma.instance[0]); - if (err) - goto out; - - for (i = 1; i < adev->sdma.num_instances; i++) { - memcpy((void*)&adev->sdma.instance[i], - (void*)&adev->sdma.instance[0], - sizeof(struct amdgpu_sdma_instance)); - } - - DRM_DEBUG("psp_load == '%s'\n", - adev->firmware.load_type == AMDGPU_FW_LOAD_PSP ? "true" : "false"); - - if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) { - sdma_hdr = (const struct sdma_firmware_header_v2_0 *)adev->sdma.instance[0].fw->data; - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_SDMA_UCODE_TH0]; - info->ucode_id = AMDGPU_UCODE_ID_SDMA_UCODE_TH0; - info->fw = adev->sdma.instance[0].fw; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(sdma_hdr->ctx_ucode_size_bytes), PAGE_SIZE); - info = &adev->firmware.ucode[AMDGPU_UCODE_ID_SDMA_UCODE_TH1]; - info->ucode_id = AMDGPU_UCODE_ID_SDMA_UCODE_TH1; - info->fw = adev->sdma.instance[0].fw; - adev->firmware.fw_size += - ALIGN(le32_to_cpu(sdma_hdr->ctl_ucode_size_bytes), PAGE_SIZE); - } - -out: - if (err) { - DRM_ERROR("sdma_v6_0: Failed to load firmware \"%s\"\n", fw_name); - sdma_v6_0_destroy_inst_ctx(adev); - } - return err; + return amdgpu_sdma_init_microcode(adev, fw_name, 0, true); } static unsigned sdma_v6_0_ring_init_cond_exec(struct amdgpu_ring *ring) @@ -467,14 +398,10 @@ static void sdma_v6_0_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 se */ static void sdma_v6_0_gfx_stop(struct amdgpu_device *adev) { - struct amdgpu_ring *sdma0 = &adev->sdma.instance[0].ring; - struct amdgpu_ring *sdma1 = &adev->sdma.instance[1].ring; u32 rb_cntl, ib_cntl; int i; - if ((adev->mman.buffer_funcs_ring == sdma0) || - (adev->mman.buffer_funcs_ring == sdma1)) - amdgpu_ttm_set_buffer_funcs_status(adev, false); + amdgpu_sdma_unset_buffer_funcs_helper(adev); for (i = 0; i < adev->sdma.num_instances; i++) { rb_cntl = RREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_RB_CNTL)); @@ -484,9 +411,6 @@ static void sdma_v6_0_gfx_stop(struct amdgpu_device *adev) ib_cntl = REG_SET_FIELD(ib_cntl, SDMA0_QUEUE0_IB_CNTL, IB_ENABLE, 0); WREG32_SOC15_IP(GC, sdma_v6_0_get_reg_offset(adev, i, regSDMA0_QUEUE0_IB_CNTL), ib_cntl); } - - sdma0->sched.ready = false; - sdma1->sched.ready = false; } /** @@ -915,7 +839,8 @@ static int sdma_v6_0_mqd_init(struct amdgpu_device *adev, void *mqd, m->sdmax_rlcx_rb_cntl = order_base_2(prop->queue_size / 4) << SDMA0_QUEUE0_RB_CNTL__RB_SIZE__SHIFT | 1 << SDMA0_QUEUE0_RB_CNTL__RPTR_WRITEBACK_ENABLE__SHIFT | - 4 << SDMA0_QUEUE0_RB_CNTL__RPTR_WRITEBACK_TIMER__SHIFT; + 4 << SDMA0_QUEUE0_RB_CNTL__RPTR_WRITEBACK_TIMER__SHIFT | + 1 << SDMA0_QUEUE0_RB_CNTL__F32_WPTR_POLL_ENABLE__SHIFT; m->sdmax_rlcx_rb_base = lower_32_bits(prop->hqd_base_gpu_addr >> 8); m->sdmax_rlcx_rb_base_hi = upper_32_bits(prop->hqd_base_gpu_addr >> 8); @@ -1370,7 +1295,7 @@ static int sdma_v6_0_sw_fini(void *handle) for (i = 0; i < adev->sdma.num_instances; i++) amdgpu_ring_fini(&adev->sdma.instance[i].ring); - sdma_v6_0_destroy_inst_ctx(adev); + amdgpu_sdma_destroy_inst_ctx(adev, true); return 0; } @@ -1386,8 +1311,11 @@ static int sdma_v6_0_hw_fini(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; - if (amdgpu_sriov_vf(adev)) + if (amdgpu_sriov_vf(adev)) { + /* disable the scheduler for SDMA */ + amdgpu_sdma_unset_buffer_funcs_helper(adev); return 0; + } sdma_v6_0_ctx_switch_enable(adev, false); sdma_v6_0_enable(adev, false); diff --git a/drivers/gpu/drm/amd/amdgpu/si_dma.c b/drivers/gpu/drm/amd/amdgpu/si_dma.c index f675111ace20..4d5e718540aa 100644 --- a/drivers/gpu/drm/amd/amdgpu/si_dma.c +++ b/drivers/gpu/drm/amd/amdgpu/si_dma.c @@ -116,15 +116,14 @@ static void si_dma_stop(struct amdgpu_device *adev) u32 rb_cntl; unsigned i; + amdgpu_sdma_unset_buffer_funcs_helper(adev); + for (i = 0; i < adev->sdma.num_instances; i++) { ring = &adev->sdma.instance[i].ring; /* dma0 */ rb_cntl = RREG32(DMA_RB_CNTL + sdma_offsets[i]); rb_cntl &= ~DMA_RB_ENABLE; WREG32(DMA_RB_CNTL + sdma_offsets[i], rb_cntl); - - if (adev->mman.buffer_funcs_ring == ring) - amdgpu_ttm_set_buffer_funcs_status(adev, false); } } diff --git a/drivers/gpu/drm/amd/amdgpu/sienna_cichlid.c b/drivers/gpu/drm/amd/amdgpu/sienna_cichlid.c index 7aa570c1ce4a..81a6d5b94987 100644 --- a/drivers/gpu/drm/amd/amdgpu/sienna_cichlid.c +++ b/drivers/gpu/drm/amd/amdgpu/sienna_cichlid.c @@ -31,12 +31,23 @@ #include "amdgpu_psp.h" #include "amdgpu_xgmi.h" +static bool sienna_cichlid_is_mode2_default(struct amdgpu_reset_control *reset_ctl) +{ +#if 0 + struct amdgpu_device *adev = (struct amdgpu_device *)reset_ctl->handle; + + if (adev->ip_versions[MP1_HWIP][0] == IP_VERSION(11, 0, 7) && + adev->pm.fw_version >= 0x3a5500 && !amdgpu_sriov_vf(adev)) + return true; +#endif + return false; +} + static struct amdgpu_reset_handler * sienna_cichlid_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, @@ -44,15 +55,13 @@ sienna_cichlid_get_reset_handler(struct amdgpu_reset_control *reset_ctl, if (handler->reset_method == reset_context->method) return handler; } - } else { - list_for_each_entry(handler, &reset_ctl->reset_handlers, + } + + if (sienna_cichlid_is_mode2_default(reset_ctl)) { + list_for_each_entry (handler, &reset_ctl->reset_handlers, handler_list) { - if (handler->reset_method == AMD_RESET_METHOD_MODE2 && - adev->pm.fw_version >= 0x3a5500 && - !amdgpu_sriov_vf(adev)) { - reset_context->method = AMD_RESET_METHOD_MODE2; + if (handler->reset_method == AMD_RESET_METHOD_MODE2) return handler; - } } } diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c index 183024d7c184..e3b2b6b4f1a6 100644 --- a/drivers/gpu/drm/amd/amdgpu/soc15.c +++ b/drivers/gpu/drm/amd/amdgpu/soc15.c @@ -1211,6 +1211,20 @@ static int soc15_common_sw_fini(void *handle) return 0; } +static void soc15_sdma_doorbell_range_init(struct amdgpu_device *adev) +{ + int i; + + /* sdma doorbell range is programed by hypervisor */ + if (!amdgpu_sriov_vf(adev)) { + for (i = 0; i < adev->sdma.num_instances; i++) { + adev->nbio.funcs->sdma_doorbell_range(adev, i, + true, adev->doorbell_index.sdma_engine[i] << 1, + adev->doorbell_index.sdma_doorbell_range); + } + } +} + static int soc15_common_hw_init(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; @@ -1230,6 +1244,13 @@ static int soc15_common_hw_init(void *handle) /* enable the doorbell aperture */ soc15_enable_doorbell_aperture(adev, true); + /* HW doorbell routing policy: doorbell writing not + * in SDMA/IH/MM/ACV range will be routed to CP. So + * we need to init SDMA doorbell range prior + * to CP ip block init and ring test. IH already + * happens before CP. + */ + soc15_sdma_doorbell_range_init(adev); return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/soc21.c b/drivers/gpu/drm/amd/amdgpu/soc21.c index 5f0d6983714a..e08044008186 100644 --- a/drivers/gpu/drm/amd/amdgpu/soc21.c +++ b/drivers/gpu/drm/amd/amdgpu/soc21.c @@ -423,6 +423,7 @@ static bool soc21_need_full_reset(struct amdgpu_device *adev) case IP_VERSION(11, 0, 0): return amdgpu_ras_is_supported(adev, AMDGPU_RAS_BLOCK__UMC); case IP_VERSION(11, 0, 2): + case IP_VERSION(11, 0, 3): return false; default: return true; @@ -629,12 +630,18 @@ static int soc21_common_early_init(void *handle) 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 + 0x1; break; case IP_VERSION(11, 0, 3): adev->cg_flags = AMD_CG_SUPPORT_VCN_MGCG | - AMD_CG_SUPPORT_JPEG_MGCG; + AMD_CG_SUPPORT_JPEG_MGCG | + AMD_CG_SUPPORT_GFX_CGCG | + AMD_CG_SUPPORT_GFX_CGLS | + AMD_CG_SUPPORT_REPEATER_FGCG | + AMD_CG_SUPPORT_GFX_MGCG; 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_v6_1.c b/drivers/gpu/drm/amd/amdgpu/umc_v6_1.c index 939cb203f7ad..f17d297b594b 100644 --- a/drivers/gpu/drm/amd/amdgpu/umc_v6_1.c +++ b/drivers/gpu/drm/amd/amdgpu/umc_v6_1.c @@ -327,10 +327,9 @@ static void umc_v6_1_query_error_address(struct amdgpu_device *adev, return; } - /* calculate error address if ue/ce error is detected */ + /* 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, UECC) == 1 || - REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, CECC) == 1)) { + REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, UECC) == 1) { err_addr = RREG64_PCIE((mc_umc_addrt0 + umc_reg_offset) * 4); /* the lowest lsb bits should be ignored */ @@ -343,10 +342,7 @@ static void umc_v6_1_query_error_address(struct amdgpu_device *adev, ADDR_OF_256B_BLOCK(channel_index) | OFFSET_IN_256B_BLOCK(err_addr); - /* we only save ue error information currently, ce is skipped */ - if (REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, UECC) - == 1) - amdgpu_umc_fill_error_record(err_data, err_addr, + amdgpu_umc_fill_error_record(err_data, err_addr, retired_page, channel_index, umc_inst); } diff --git a/drivers/gpu/drm/amd/amdgpu/umc_v6_7.c b/drivers/gpu/drm/amd/amdgpu/umc_v6_7.c index bf7524f16b66..5d5d031c9e7d 100644 --- a/drivers/gpu/drm/amd/amdgpu/umc_v6_7.c +++ b/drivers/gpu/drm/amd/amdgpu/umc_v6_7.c @@ -187,20 +187,51 @@ static void umc_v6_7_ecc_info_query_ras_error_count(struct amdgpu_device *adev, } } +static void umc_v6_7_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 channel_index; + uint64_t soc_pa, retired_page, column; + + channel_index = + adev->umc.channel_idx_tbl[umc_inst * adev->umc.channel_inst_num + ch_inst]; + /* translate umc channel address to soc pa, 3 parts are included */ + soc_pa = ADDR_OF_8KB_BLOCK(err_addr) | + ADDR_OF_256B_BLOCK(channel_index) | + OFFSET_IN_256B_BLOCK(err_addr); + + /* The umc channel bits are not original values, they are hashed */ + SET_CHANNEL_HASH(channel_index, soc_pa); + + /* clear [C4 C3 C2] in soc physical address */ + soc_pa &= ~(0x7ULL << UMC_V6_7_PA_C2_BIT); + + /* loop for all possibilities of [C4 C3 C2] */ + for (column = 0; column < UMC_V6_7_NA_MAP_PA_NUM; column++) { + retired_page = soc_pa | (column << UMC_V6_7_PA_C2_BIT); + dev_info(adev->dev, "Error Address(PA): 0x%llx\n", retired_page); + amdgpu_umc_fill_error_record(err_data, err_addr, + retired_page, channel_index, umc_inst); + + /* shift R14 bit */ + retired_page ^= (0x1ULL << UMC_V6_7_PA_R14_BIT); + dev_info(adev->dev, "Error Address(PA): 0x%llx\n", retired_page); + amdgpu_umc_fill_error_record(err_data, err_addr, + retired_page, channel_index, umc_inst); + } +} + static void umc_v6_7_ecc_info_query_error_address(struct amdgpu_device *adev, struct ras_err_data *err_data, uint32_t ch_inst, uint32_t umc_inst) { - uint64_t mc_umc_status, err_addr, soc_pa, retired_page, column; - uint32_t channel_index; + uint64_t mc_umc_status, err_addr; uint32_t eccinfo_table_idx; struct amdgpu_ras *ras = amdgpu_ras_get_context(adev); eccinfo_table_idx = umc_inst * adev->umc.channel_inst_num + ch_inst; - channel_index = - adev->umc.channel_idx_tbl[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) @@ -209,42 +240,15 @@ static void umc_v6_7_ecc_info_query_error_address(struct amdgpu_device *adev, if (!err_data->err_addr) return; - /* calculate error address if ue/ce error is detected */ + /* 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, UECC) == 1 || - REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, CECC) == 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); - /* translate umc channel address to soc pa, 3 parts are included */ - soc_pa = ADDR_OF_8KB_BLOCK(err_addr) | - ADDR_OF_256B_BLOCK(channel_index) | - OFFSET_IN_256B_BLOCK(err_addr); - - /* The umc channel bits are not original values, they are hashed */ - SET_CHANNEL_HASH(channel_index, soc_pa); - - /* clear [C4 C3 C2] in soc physical address */ - soc_pa &= ~(0x7ULL << UMC_V6_7_PA_C2_BIT); - - /* we only save ue error information currently, ce is skipped */ - if (REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, UECC) - == 1) { - /* loop for all possibilities of [C4 C3 C2] */ - for (column = 0; column < UMC_V6_7_NA_MAP_PA_NUM; column++) { - retired_page = soc_pa | (column << UMC_V6_7_PA_C2_BIT); - dev_info(adev->dev, "Error Address(PA): 0x%llx\n", retired_page); - amdgpu_umc_fill_error_record(err_data, err_addr, - retired_page, channel_index, umc_inst); - - /* shift R14 bit */ - retired_page ^= (0x1ULL << UMC_V6_7_PA_R14_BIT); - dev_info(adev->dev, "Error Address(PA): 0x%llx\n", retired_page); - amdgpu_umc_fill_error_record(err_data, err_addr, - retired_page, channel_index, umc_inst); - } - } + umc_v6_7_convert_error_address(adev, err_data, err_addr, + ch_inst, umc_inst); } } @@ -452,14 +456,11 @@ static void umc_v6_7_query_ras_error_count(struct amdgpu_device *adev, static void umc_v6_7_query_error_address(struct amdgpu_device *adev, struct ras_err_data *err_data, - uint32_t umc_reg_offset, - uint32_t ch_inst, + uint32_t umc_reg_offset, uint32_t ch_inst, uint32_t umc_inst) { uint32_t mc_umc_status_addr; - uint32_t channel_index; - uint64_t mc_umc_status, mc_umc_addrt0; - uint64_t err_addr, soc_pa, retired_page, column; + uint64_t mc_umc_status = 0, mc_umc_addrt0, err_addr; mc_umc_status_addr = SOC15_REG_OFFSET(UMC, 0, regMCA_UMC_UMC0_MCUMC_STATUST0); @@ -477,45 +478,15 @@ static void umc_v6_7_query_error_address(struct amdgpu_device *adev, return; } - channel_index = - adev->umc.channel_idx_tbl[umc_inst * adev->umc.channel_inst_num + ch_inst]; - - /* calculate error address if ue/ce error is detected */ + /* 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, UECC) == 1 || - REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, CECC) == 1)) { - + REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, UECC) == 1) { err_addr = RREG64_PCIE((mc_umc_addrt0 + umc_reg_offset) * 4); - err_addr = REG_GET_FIELD(err_addr, MCA_UMC_UMC0_MCUMC_ADDRT0, ErrorAddr); + err_addr = + REG_GET_FIELD(err_addr, MCA_UMC_UMC0_MCUMC_ADDRT0, ErrorAddr); - /* translate umc channel address to soc pa, 3 parts are included */ - soc_pa = ADDR_OF_8KB_BLOCK(err_addr) | - ADDR_OF_256B_BLOCK(channel_index) | - OFFSET_IN_256B_BLOCK(err_addr); - - /* The umc channel bits are not original values, they are hashed */ - SET_CHANNEL_HASH(channel_index, soc_pa); - - /* clear [C4 C3 C2] in soc physical address */ - soc_pa &= ~(0x7ULL << UMC_V6_7_PA_C2_BIT); - - /* we only save ue error information currently, ce is skipped */ - if (REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, UECC) - == 1) { - /* loop for all possibilities of [C4 C3 C2] */ - for (column = 0; column < UMC_V6_7_NA_MAP_PA_NUM; column++) { - retired_page = soc_pa | (column << UMC_V6_7_PA_C2_BIT); - dev_info(adev->dev, "Error Address(PA): 0x%llx\n", retired_page); - amdgpu_umc_fill_error_record(err_data, err_addr, - retired_page, channel_index, umc_inst); - - /* shift R14 bit */ - retired_page ^= (0x1ULL << UMC_V6_7_PA_R14_BIT); - dev_info(adev->dev, "Error Address(PA): 0x%llx\n", retired_page); - amdgpu_umc_fill_error_record(err_data, err_addr, - retired_page, channel_index, umc_inst); - } - } + umc_v6_7_convert_error_address(adev, err_data, err_addr, + ch_inst, umc_inst); } /* clear umc status */ @@ -540,8 +511,7 @@ static void umc_v6_7_query_ras_error_address(struct amdgpu_device *adev, ch_inst); umc_v6_7_query_error_address(adev, err_data, - umc_reg_offset, - ch_inst, + umc_reg_offset, ch_inst, umc_inst); } } @@ -583,4 +553,5 @@ struct amdgpu_umc_ras umc_v6_7_ras = { .query_ras_poison_mode = umc_v6_7_query_ras_poison_mode, .ecc_info_query_ras_error_count = umc_v6_7_ecc_info_query_ras_error_count, .ecc_info_query_ras_error_address = umc_v6_7_ecc_info_query_ras_error_address, + .convert_ras_error_address = umc_v6_7_convert_error_address, }; diff --git a/drivers/gpu/drm/amd/amdgpu/umc_v8_10.c b/drivers/gpu/drm/amd/amdgpu/umc_v8_10.c index a8cbda81828d..91235df54e22 100644 --- a/drivers/gpu/drm/amd/amdgpu/umc_v8_10.c +++ b/drivers/gpu/drm/amd/amdgpu/umc_v8_10.c @@ -208,7 +208,10 @@ 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; - uint32_t channel_index; + 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; mc_umc_status_addr = SOC15_REG_OFFSET(UMC, 0, regMCA_UMC_UMC0_MCUMC_STATUST0); @@ -229,13 +232,10 @@ static void umc_v8_10_query_error_address(struct amdgpu_device *adev, umc_inst * adev->umc.channel_inst_num + ch_inst]; - /* calculate error address if ue/ce error is detected */ + /* 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 || - REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, CECC) == 1)) { - uint32_t addr_lsb; - uint64_t mc_umc_addrt0; + REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, UECC) == 1) { mc_umc_addrt0 = SOC15_REG_OFFSET(UMC, 0, regMCA_UMC_UMC0_MCUMC_ADDRT0); err_addr = RREG64_PCIE((mc_umc_addrt0 + umc_reg_offset) * 4); @@ -243,32 +243,24 @@ static void umc_v8_10_query_error_address(struct amdgpu_device *adev, /* 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); - - /* we only save ue error information currently, ce is skipped */ - if (REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, UECC) == 1) { - uint64_t na_err_addr_base = err_addr & ~(0x3ULL << UMC_V8_10_NA_C5_BIT); - uint64_t na_err_addr, retired_page_addr; - uint32_t col = 0; - int ret = 0; - - /* 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); + 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); } } @@ -338,6 +330,31 @@ static void umc_v8_10_err_cnt_init(struct amdgpu_device *adev) } } +static uint32_t umc_v8_10_query_ras_poison_mode_per_channel( + struct amdgpu_device *adev, + uint32_t umc_reg_offset) +{ + uint32_t ecc_ctrl_addr, ecc_ctrl; + + ecc_ctrl_addr = + SOC15_REG_OFFSET(UMC, 0, regUMCCH0_0_GeccCtrl); + ecc_ctrl = RREG32_PCIE((ecc_ctrl_addr + + umc_reg_offset) * 4); + + return REG_GET_FIELD(ecc_ctrl, UMCCH0_0_GeccCtrl, UCFatalEn); +} + +static bool umc_v8_10_query_ras_poison_mode(struct amdgpu_device *adev) +{ + uint32_t umc_reg_offset = 0; + + /* Enabling fatal error in umc node0 instance0 channel0 will be + * considered as fatal error mode + */ + umc_reg_offset = get_umc_v8_10_reg_offset(adev, 0, 0, 0); + return !umc_v8_10_query_ras_poison_mode_per_channel(adev, umc_reg_offset); +} + 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, @@ -348,4 +365,5 @@ struct amdgpu_umc_ras umc_v8_10_ras = { .hw_ops = &umc_v8_10_ras_hw_ops, }, .err_cnt_init = umc_v8_10_err_cnt_init, + .query_ras_poison_mode = umc_v8_10_query_ras_poison_mode, }; diff --git a/drivers/gpu/drm/amd/amdgpu/umc_v8_7.c b/drivers/gpu/drm/amd/amdgpu/umc_v8_7.c index f35253e0eaa6..b717fdaa46e4 100644 --- a/drivers/gpu/drm/amd/amdgpu/umc_v8_7.c +++ b/drivers/gpu/drm/amd/amdgpu/umc_v8_7.c @@ -108,20 +108,35 @@ static void umc_v8_7_ecc_info_query_ras_error_count(struct amdgpu_device *adev, } } +static void umc_v8_7_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) +{ + uint64_t retired_page; + uint32_t channel_index; + + channel_index = + adev->umc.channel_idx_tbl[umc_inst * adev->umc.channel_inst_num + ch_inst]; + + /* translate umc channel address to soc pa, 3 parts are included */ + retired_page = ADDR_OF_4KB_BLOCK(err_addr) | + ADDR_OF_256B_BLOCK(channel_index) | + OFFSET_IN_256B_BLOCK(err_addr); + + amdgpu_umc_fill_error_record(err_data, err_addr, + retired_page, channel_index, umc_inst); +} + static void umc_v8_7_ecc_info_query_error_address(struct amdgpu_device *adev, struct ras_err_data *err_data, uint32_t ch_inst, uint32_t umc_inst) { - uint64_t mc_umc_status, err_addr, retired_page; - uint32_t channel_index; + uint64_t mc_umc_status, err_addr; uint32_t eccinfo_table_idx; struct amdgpu_ras *ras = amdgpu_ras_get_context(adev); eccinfo_table_idx = umc_inst * adev->umc.channel_inst_num + ch_inst; - channel_index = - adev->umc.channel_idx_tbl[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) @@ -130,24 +145,15 @@ static void umc_v8_7_ecc_info_query_error_address(struct amdgpu_device *adev, if (!err_data->err_addr) return; - /* calculate error address if ue/ce error is detected */ + /* 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, UECC) == 1 || - REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, CECC) == 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); - /* translate umc channel address to soc pa, 3 parts are included */ - retired_page = ADDR_OF_4KB_BLOCK(err_addr) | - ADDR_OF_256B_BLOCK(channel_index) | - OFFSET_IN_256B_BLOCK(err_addr); - - /* we only save ue error information currently, ce is skipped */ - if (REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, UECC) - == 1) - amdgpu_umc_fill_error_record(err_data, err_addr, - retired_page, channel_index, umc_inst); + umc_v8_7_convert_error_address(adev, err_data, err_addr, + ch_inst, umc_inst); } } @@ -324,14 +330,12 @@ static void umc_v8_7_query_error_address(struct amdgpu_device *adev, uint32_t umc_inst) { uint32_t lsb, mc_umc_status_addr; - uint64_t mc_umc_status, err_addr, retired_page, mc_umc_addrt0; - uint32_t channel_index = adev->umc.channel_idx_tbl[umc_inst * adev->umc.channel_inst_num + ch_inst]; + uint64_t mc_umc_status, err_addr, mc_umc_addrt0; mc_umc_status_addr = SOC15_REG_OFFSET(UMC, 0, mmMCA_UMC_UMC0_MCUMC_STATUST0); mc_umc_addrt0 = SOC15_REG_OFFSET(UMC, 0, mmMCA_UMC_UMC0_MCUMC_ADDRT0); - mc_umc_status = RREG64_PCIE((mc_umc_status_addr + umc_reg_offset) * 4); if (mc_umc_status == 0) @@ -343,10 +347,9 @@ static void umc_v8_7_query_error_address(struct amdgpu_device *adev, return; } - /* calculate error address if ue/ce error is detected */ + /* 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, UECC) == 1 || - REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, CECC) == 1)) { + REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, UECC) == 1) { err_addr = RREG64_PCIE((mc_umc_addrt0 + umc_reg_offset) * 4); /* the lowest lsb bits should be ignored */ @@ -354,16 +357,8 @@ static void umc_v8_7_query_error_address(struct amdgpu_device *adev, err_addr = REG_GET_FIELD(err_addr, MCA_UMC_UMC0_MCUMC_ADDRT0, ErrorAddr); err_addr &= ~((0x1ULL << lsb) - 1); - /* translate umc channel address to soc pa, 3 parts are included */ - retired_page = ADDR_OF_4KB_BLOCK(err_addr) | - ADDR_OF_256B_BLOCK(channel_index) | - OFFSET_IN_256B_BLOCK(err_addr); - - /* we only save ue error information currently, ce is skipped */ - if (REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, UECC) - == 1) - amdgpu_umc_fill_error_record(err_data, err_addr, - retired_page, channel_index, umc_inst); + umc_v8_7_convert_error_address(adev, err_data, err_addr, + ch_inst, umc_inst); } /* clear umc status */ diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v4_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v4_0.c index b6f73b87c47e..897a5ce9c9da 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v4_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v4_0.c @@ -150,6 +150,10 @@ static int vcn_v4_0_sw_init(void *handle) fw_shared->present_flag_0 = cpu_to_le32(AMDGPU_FW_SHARED_FLAG_0_UNIFIED_QUEUE); fw_shared->sq.is_enabled = 1; + fw_shared->present_flag_0 |= cpu_to_le32(AMDGPU_VCN_SMU_DPM_INTERFACE_FLAG); + fw_shared->smu_dpm_interface.smu_interface_type = (adev->flags & AMD_IS_APU) ? + AMDGPU_VCN_SMU_DPM_INTERFACE_APU : AMDGPU_VCN_SMU_DPM_INTERFACE_DGPU; + if (amdgpu_sriov_vf(adev)) fw_shared->present_flag_0 |= cpu_to_le32(AMDGPU_VCN_VF_RB_SETUP_FLAG); diff --git a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler.h b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler.h index 60a81649cf12..c7118843db05 100644 --- a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler.h +++ b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler.h @@ -742,7 +742,7 @@ static const uint32_t cwsr_trap_nv1x_hex[] = { 0xbf88fffe, 0x877aff7f, 0x04000000, 0x8f7a857a, 0x886d7a6d, 0xb97b02dc, - 0x8f7b997b, 0xb97a2a05, + 0x8f7b997b, 0xb97a3a05, 0x807a817a, 0xbf0d997b, 0xbf850002, 0x8f7a897a, 0xbf820001, 0x8f7a8a7a, @@ -819,7 +819,7 @@ static const uint32_t cwsr_trap_nv1x_hex[] = { 0xbefe037c, 0xbefc0370, 0xf4611c7a, 0xf8000000, 0x80708470, 0xbefc037e, - 0xb9702a05, 0x80708170, + 0xb9703a05, 0x80708170, 0xbf0d9973, 0xbf850002, 0x8f708970, 0xbf820001, 0x8f708a70, 0xb97a1e06, @@ -1069,7 +1069,7 @@ static const uint32_t cwsr_trap_nv1x_hex[] = { 0xb9f9f816, 0x876f7bff, 0xfffff800, 0x906f8b6f, 0xb9efa2c3, 0xb9f3f801, - 0xb96e2a05, 0x806e816e, + 0xb96e3a05, 0x806e816e, 0xbf0d9972, 0xbf850002, 0x8f6e896e, 0xbf820001, 0x8f6e8a6e, 0xb96f1e06, @@ -2114,7 +2114,7 @@ static const uint32_t cwsr_trap_gfx10_hex[] = { 0x007a0000, 0x7e000280, 0xbefe037a, 0xbeff037b, 0xb97b02dc, 0x8f7b997b, - 0xb97a2a05, 0x807a817a, + 0xb97a3a05, 0x807a817a, 0xbf0d997b, 0xbf850002, 0x8f7a897a, 0xbf820001, 0x8f7a8a7a, 0xb97b1e06, @@ -2157,7 +2157,7 @@ static const uint32_t cwsr_trap_gfx10_hex[] = { 0x01000000, 0xe0704100, 0x705d0100, 0xe0704200, 0x705d0200, 0xe0704300, - 0x705d0300, 0xb9702a05, + 0x705d0300, 0xb9703a05, 0x80708170, 0xbf0d9973, 0xbf850002, 0x8f708970, 0xbf820001, 0x8f708a70, @@ -2189,7 +2189,7 @@ static const uint32_t cwsr_trap_gfx10_hex[] = { 0xbefe03ff, 0x0000ffff, 0xbeff0380, 0xe0704000, 0x705d0200, 0xbefe03c1, - 0xb9702a05, 0x80708170, + 0xb9703a05, 0x80708170, 0xbf0d9973, 0xbf850002, 0x8f708970, 0xbf820001, 0x8f708a70, 0xb97a1e06, @@ -2475,7 +2475,7 @@ static const uint32_t cwsr_trap_gfx10_hex[] = { 0xb9ef4803, 0x876f7bff, 0xfffff800, 0x906f8b6f, 0xb9efa2c3, 0xb9f3f801, - 0xb96e2a05, 0x806e816e, + 0xb96e3a05, 0x806e816e, 0xbf0d9972, 0xbf850002, 0x8f6e896e, 0xbf820001, 0x8f6e8a6e, 0xb96f1e06, @@ -2494,438 +2494,441 @@ static const uint32_t cwsr_trap_gfx10_hex[] = { 0xbf9f0000, 0xbf9f0000, 0xbf9f0000, 0x00000000, }; - static const uint32_t cwsr_trap_gfx11_hex[] = { - 0xbfa00001, 0xbfa0021b, + 0xbfa00001, 0xbfa0021e, 0xb0804006, 0xb8f8f802, - 0x91788678, 0xb8fbf803, - 0x8b6eff78, 0x00002000, - 0xbfa10009, 0x8b6eff6d, - 0x00ff0000, 0xbfa2001e, - 0x8b6eff7b, 0x00000400, - 0xbfa20041, 0xbf830010, - 0xb8fbf803, 0xbfa0fffa, - 0x8b6eff7b, 0x00000900, - 0xbfa20015, 0x8b6eff7b, - 0x000071ff, 0xbfa10008, - 0x8b6fff7b, 0x00007080, - 0xbfa10001, 0xbeee1287, - 0xb8eff801, 0x846e8c6e, - 0x8b6e6f6e, 0xbfa2000a, + 0x9178ff78, 0x00020006, + 0xb8fbf803, 0xbf0d9f6d, + 0xbfa20006, 0x8b6eff78, + 0x00002000, 0xbfa10009, 0x8b6eff6d, 0x00ff0000, - 0xbfa20007, 0xb8eef801, - 0x8b6eff6e, 0x00000800, - 0xbfa20003, 0x8b6eff7b, - 0x00000400, 0xbfa20026, - 0xbefa4d82, 0xbf89fc07, - 0x84fa887a, 0xf4005bbd, - 0xf8000010, 0xbf89fc07, - 0x846e976e, 0x9177ff77, - 0x00800000, 0x8c776e77, - 0xf4045bbd, 0xf8000000, - 0xbf89fc07, 0xf4045ebd, - 0xf8000008, 0xbf89fc07, - 0x8bee6e6e, 0xbfa10001, - 0xbe80486e, 0x8b6eff6d, - 0x01ff0000, 0xbfa20005, - 0x8c78ff78, 0x00002000, - 0x80ec886c, 0x82ed806d, - 0xbfa00005, 0x8b6eff6d, - 0x01000000, 0xbfa20002, - 0x806c846c, 0x826d806d, - 0x8b6dff6d, 0x0000ffff, - 0x8bfe7e7e, 0x8bea6a6a, - 0xb978f802, 0xbe804a6c, - 0x8b6dff6d, 0x0000ffff, - 0xbefa0080, 0xb97a0283, - 0xbeee007e, 0xbeef007f, - 0xbefe0180, 0xbefe4d84, - 0xbf89fc07, 0x8b7aff7f, - 0x04000000, 0x847a857a, - 0x8c6d7a6d, 0xbefa007e, - 0x8b7bff7f, 0x0000ffff, - 0xbefe00c1, 0xbeff00c1, - 0xdca6c000, 0x007a0000, - 0x7e000280, 0xbefe007a, - 0xbeff007b, 0xb8fb02dc, - 0x847b997b, 0xb8fa3b05, - 0x807a817a, 0xbf0d997b, - 0xbfa20002, 0x847a897a, - 0xbfa00001, 0x847a8a7a, - 0xb8fb1e06, 0x847b8a7b, - 0x807a7b7a, 0x8b7bff7f, - 0x0000ffff, 0x807aff7a, - 0x00000200, 0x807a7e7a, - 0x827b807b, 0xd7610000, - 0x00010870, 0xd7610000, - 0x00010a71, 0xd7610000, - 0x00010c72, 0xd7610000, - 0x00010e73, 0xd7610000, - 0x00011074, 0xd7610000, - 0x00011275, 0xd7610000, - 0x00011476, 0xd7610000, - 0x00011677, 0xd7610000, - 0x00011a79, 0xd7610000, - 0x00011c7e, 0xd7610000, - 0x00011e7f, 0xbefe00ff, - 0x00003fff, 0xbeff0080, - 0xdca6c040, 0x007a0000, - 0xd760007a, 0x00011d00, - 0xd760007b, 0x00011f00, + 0xbfa2001e, 0x8b6eff7b, + 0x00000400, 0xbfa20041, + 0xbf830010, 0xb8fbf803, + 0xbfa0fffa, 0x8b6eff7b, + 0x00000900, 0xbfa20015, + 0x8b6eff7b, 0x000071ff, + 0xbfa10008, 0x8b6fff7b, + 0x00007080, 0xbfa10001, + 0xbeee1287, 0xb8eff801, + 0x846e8c6e, 0x8b6e6f6e, + 0xbfa2000a, 0x8b6eff6d, + 0x00ff0000, 0xbfa20007, + 0xb8eef801, 0x8b6eff6e, + 0x00000800, 0xbfa20003, + 0x8b6eff7b, 0x00000400, + 0xbfa20026, 0xbefa4d82, + 0xbf89fc07, 0x84fa887a, + 0xf4005bbd, 0xf8000010, + 0xbf89fc07, 0x846e976e, + 0x9177ff77, 0x00800000, + 0x8c776e77, 0xf4045bbd, + 0xf8000000, 0xbf89fc07, + 0xf4045ebd, 0xf8000008, + 0xbf89fc07, 0x8bee6e6e, + 0xbfa10001, 0xbe80486e, + 0x8b6eff6d, 0x01ff0000, + 0xbfa20005, 0x8c78ff78, + 0x00002000, 0x80ec886c, + 0x82ed806d, 0xbfa00005, + 0x8b6eff6d, 0x01000000, + 0xbfa20002, 0x806c846c, + 0x826d806d, 0x8b6dff6d, + 0x0000ffff, 0x8bfe7e7e, + 0x8bea6a6a, 0xb978f802, + 0xbe804a6c, 0x8b6dff6d, + 0x0000ffff, 0xbefa0080, + 0xb97a0283, 0xbeee007e, + 0xbeef007f, 0xbefe0180, + 0xbefe4d84, 0xbf89fc07, + 0x8b7aff7f, 0x04000000, + 0x847a857a, 0x8c6d7a6d, + 0xbefa007e, 0x8b7bff7f, + 0x0000ffff, 0xbefe00c1, + 0xbeff00c1, 0xdca6c000, + 0x007a0000, 0x7e000280, 0xbefe007a, 0xbeff007b, - 0xbef4007e, 0x8b75ff7f, - 0x0000ffff, 0x8c75ff75, - 0x00040000, 0xbef60080, - 0xbef700ff, 0x10807fac, - 0xbef1007d, 0xbef00080, - 0xb8f302dc, 0x84739973, - 0xbefe00c1, 0x857d9973, - 0x8b7d817d, 0xbf06817d, - 0xbfa20002, 0xbeff0080, - 0xbfa00002, 0xbeff00c1, - 0xbfa00009, 0xbef600ff, - 0x01000000, 0xe0685080, - 0x701d0100, 0xe0685100, - 0x701d0200, 0xe0685180, - 0x701d0300, 0xbfa00008, + 0xb8fb02dc, 0x847b997b, + 0xb8fa3b05, 0x807a817a, + 0xbf0d997b, 0xbfa20002, + 0x847a897a, 0xbfa00001, + 0x847a8a7a, 0xb8fb1e06, + 0x847b8a7b, 0x807a7b7a, + 0x8b7bff7f, 0x0000ffff, + 0x807aff7a, 0x00000200, + 0x807a7e7a, 0x827b807b, + 0xd7610000, 0x00010870, + 0xd7610000, 0x00010a71, + 0xd7610000, 0x00010c72, + 0xd7610000, 0x00010e73, + 0xd7610000, 0x00011074, + 0xd7610000, 0x00011275, + 0xd7610000, 0x00011476, + 0xd7610000, 0x00011677, + 0xd7610000, 0x00011a79, + 0xd7610000, 0x00011c7e, + 0xd7610000, 0x00011e7f, + 0xbefe00ff, 0x00003fff, + 0xbeff0080, 0xdca6c040, + 0x007a0000, 0xd760007a, + 0x00011d00, 0xd760007b, + 0x00011f00, 0xbefe007a, + 0xbeff007b, 0xbef4007e, + 0x8b75ff7f, 0x0000ffff, + 0x8c75ff75, 0x00040000, + 0xbef60080, 0xbef700ff, + 0x10807fac, 0xbef1007d, + 0xbef00080, 0xb8f302dc, + 0x84739973, 0xbefe00c1, + 0x857d9973, 0x8b7d817d, + 0xbf06817d, 0xbfa20002, + 0xbeff0080, 0xbfa00002, + 0xbeff00c1, 0xbfa00009, 0xbef600ff, 0x01000000, - 0xe0685100, 0x701d0100, - 0xe0685200, 0x701d0200, - 0xe0685300, 0x701d0300, + 0xe0685080, 0x701d0100, + 0xe0685100, 0x701d0200, + 0xe0685180, 0x701d0300, + 0xbfa00008, 0xbef600ff, + 0x01000000, 0xe0685100, + 0x701d0100, 0xe0685200, + 0x701d0200, 0xe0685300, + 0x701d0300, 0xb8f03b05, + 0x80708170, 0xbf0d9973, + 0xbfa20002, 0x84708970, + 0xbfa00001, 0x84708a70, + 0xb8fa1e06, 0x847a8a7a, + 0x80707a70, 0x8070ff70, + 0x00000200, 0xbef600ff, + 0x01000000, 0x7e000280, + 0x7e020280, 0x7e040280, + 0xbefd0080, 0xd7610002, + 0x0000fa71, 0x807d817d, + 0xd7610002, 0x0000fa6c, + 0x807d817d, 0x917aff6d, + 0x80000000, 0xd7610002, + 0x0000fa7a, 0x807d817d, + 0xd7610002, 0x0000fa6e, + 0x807d817d, 0xd7610002, + 0x0000fa6f, 0x807d817d, + 0xd7610002, 0x0000fa78, + 0x807d817d, 0xb8faf803, + 0xd7610002, 0x0000fa7a, + 0x807d817d, 0xd7610002, + 0x0000fa7b, 0x807d817d, + 0xb8f1f801, 0xd7610002, + 0x0000fa71, 0x807d817d, + 0xb8f1f814, 0xd7610002, + 0x0000fa71, 0x807d817d, + 0xb8f1f815, 0xd7610002, + 0x0000fa71, 0x807d817d, + 0xbefe00ff, 0x0000ffff, + 0xbeff0080, 0xe0685000, + 0x701d0200, 0xbefe00c1, 0xb8f03b05, 0x80708170, 0xbf0d9973, 0xbfa20002, 0x84708970, 0xbfa00001, 0x84708a70, 0xb8fa1e06, 0x847a8a7a, 0x80707a70, - 0x8070ff70, 0x00000200, 0xbef600ff, 0x01000000, - 0x7e000280, 0x7e020280, - 0x7e040280, 0xbefd0080, - 0xd7610002, 0x0000fa71, - 0x807d817d, 0xd7610002, - 0x0000fa6c, 0x807d817d, - 0x917aff6d, 0x80000000, - 0xd7610002, 0x0000fa7a, - 0x807d817d, 0xd7610002, - 0x0000fa6e, 0x807d817d, - 0xd7610002, 0x0000fa6f, - 0x807d817d, 0xd7610002, - 0x0000fa78, 0x807d817d, - 0xb8faf803, 0xd7610002, - 0x0000fa7a, 0x807d817d, - 0xd7610002, 0x0000fa7b, - 0x807d817d, 0xb8f1f801, - 0xd7610002, 0x0000fa71, - 0x807d817d, 0xb8f1f814, - 0xd7610002, 0x0000fa71, - 0x807d817d, 0xb8f1f815, - 0xd7610002, 0x0000fa71, - 0x807d817d, 0xbefe00ff, - 0x0000ffff, 0xbeff0080, - 0xe0685000, 0x701d0200, - 0xbefe00c1, 0xb8f03b05, - 0x80708170, 0xbf0d9973, - 0xbfa20002, 0x84708970, - 0xbfa00001, 0x84708a70, - 0xb8fa1e06, 0x847a8a7a, - 0x80707a70, 0xbef600ff, - 0x01000000, 0xbef90080, - 0xbefd0080, 0xbf800000, - 0xbe804100, 0xbe824102, - 0xbe844104, 0xbe864106, - 0xbe884108, 0xbe8a410a, - 0xbe8c410c, 0xbe8e410e, - 0xd7610002, 0x0000f200, - 0x80798179, 0xd7610002, - 0x0000f201, 0x80798179, - 0xd7610002, 0x0000f202, - 0x80798179, 0xd7610002, - 0x0000f203, 0x80798179, - 0xd7610002, 0x0000f204, + 0xbef90080, 0xbefd0080, + 0xbf800000, 0xbe804100, + 0xbe824102, 0xbe844104, + 0xbe864106, 0xbe884108, + 0xbe8a410a, 0xbe8c410c, + 0xbe8e410e, 0xd7610002, + 0x0000f200, 0x80798179, + 0xd7610002, 0x0000f201, 0x80798179, 0xd7610002, - 0x0000f205, 0x80798179, - 0xd7610002, 0x0000f206, + 0x0000f202, 0x80798179, + 0xd7610002, 0x0000f203, 0x80798179, 0xd7610002, - 0x0000f207, 0x80798179, - 0xd7610002, 0x0000f208, + 0x0000f204, 0x80798179, + 0xd7610002, 0x0000f205, 0x80798179, 0xd7610002, - 0x0000f209, 0x80798179, - 0xd7610002, 0x0000f20a, + 0x0000f206, 0x80798179, + 0xd7610002, 0x0000f207, 0x80798179, 0xd7610002, - 0x0000f20b, 0x80798179, - 0xd7610002, 0x0000f20c, + 0x0000f208, 0x80798179, + 0xd7610002, 0x0000f209, 0x80798179, 0xd7610002, - 0x0000f20d, 0x80798179, - 0xd7610002, 0x0000f20e, + 0x0000f20a, 0x80798179, + 0xd7610002, 0x0000f20b, 0x80798179, 0xd7610002, - 0x0000f20f, 0x80798179, - 0xbf06a079, 0xbfa10006, - 0xe0685000, 0x701d0200, - 0x8070ff70, 0x00000080, - 0xbef90080, 0x7e040280, - 0x807d907d, 0xbf0aff7d, - 0x00000060, 0xbfa2ffbc, - 0xbe804100, 0xbe824102, - 0xbe844104, 0xbe864106, - 0xbe884108, 0xbe8a410a, - 0xd7610002, 0x0000f200, + 0x0000f20c, 0x80798179, + 0xd7610002, 0x0000f20d, 0x80798179, 0xd7610002, - 0x0000f201, 0x80798179, - 0xd7610002, 0x0000f202, + 0x0000f20e, 0x80798179, + 0xd7610002, 0x0000f20f, + 0x80798179, 0xbf06a079, + 0xbfa10006, 0xe0685000, + 0x701d0200, 0x8070ff70, + 0x00000080, 0xbef90080, + 0x7e040280, 0x807d907d, + 0xbf0aff7d, 0x00000060, + 0xbfa2ffbc, 0xbe804100, + 0xbe824102, 0xbe844104, + 0xbe864106, 0xbe884108, + 0xbe8a410a, 0xd7610002, + 0x0000f200, 0x80798179, + 0xd7610002, 0x0000f201, 0x80798179, 0xd7610002, - 0x0000f203, 0x80798179, - 0xd7610002, 0x0000f204, + 0x0000f202, 0x80798179, + 0xd7610002, 0x0000f203, 0x80798179, 0xd7610002, - 0x0000f205, 0x80798179, - 0xd7610002, 0x0000f206, + 0x0000f204, 0x80798179, + 0xd7610002, 0x0000f205, 0x80798179, 0xd7610002, - 0x0000f207, 0x80798179, - 0xd7610002, 0x0000f208, + 0x0000f206, 0x80798179, + 0xd7610002, 0x0000f207, 0x80798179, 0xd7610002, - 0x0000f209, 0x80798179, - 0xd7610002, 0x0000f20a, + 0x0000f208, 0x80798179, + 0xd7610002, 0x0000f209, 0x80798179, 0xd7610002, - 0x0000f20b, 0x80798179, - 0xe0685000, 0x701d0200, + 0x0000f20a, 0x80798179, + 0xd7610002, 0x0000f20b, + 0x80798179, 0xe0685000, + 0x701d0200, 0xbefe00c1, + 0x857d9973, 0x8b7d817d, + 0xbf06817d, 0xbfa20002, + 0xbeff0080, 0xbfa00001, + 0xbeff00c1, 0xb8fb4306, + 0x8b7bc17b, 0xbfa10044, + 0xbfbd0000, 0x8b7aff6d, + 0x80000000, 0xbfa10040, + 0x847b867b, 0x847b827b, + 0xbef6007b, 0xb8f03b05, + 0x80708170, 0xbf0d9973, + 0xbfa20002, 0x84708970, + 0xbfa00001, 0x84708a70, + 0xb8fa1e06, 0x847a8a7a, + 0x80707a70, 0x8070ff70, + 0x00000200, 0x8070ff70, + 0x00000080, 0xbef600ff, + 0x01000000, 0xd71f0000, + 0x000100c1, 0xd7200000, + 0x000200c1, 0x16000084, + 0x857d9973, 0x8b7d817d, + 0xbf06817d, 0xbefd0080, + 0xbfa20012, 0xbe8300ff, + 0x00000080, 0xbf800000, + 0xbf800000, 0xbf800000, + 0xd8d80000, 0x01000000, + 0xbf890000, 0xe0685000, + 0x701d0100, 0x807d037d, + 0x80700370, 0xd5250000, + 0x0001ff00, 0x00000080, + 0xbf0a7b7d, 0xbfa2fff4, + 0xbfa00011, 0xbe8300ff, + 0x00000100, 0xbf800000, + 0xbf800000, 0xbf800000, + 0xd8d80000, 0x01000000, + 0xbf890000, 0xe0685000, + 0x701d0100, 0x807d037d, + 0x80700370, 0xd5250000, + 0x0001ff00, 0x00000100, + 0xbf0a7b7d, 0xbfa2fff4, 0xbefe00c1, 0x857d9973, 0x8b7d817d, 0xbf06817d, - 0xbfa20002, 0xbeff0080, - 0xbfa00001, 0xbeff00c1, - 0xb8fb4306, 0x8b7bc17b, - 0xbfa10044, 0xbfbd0000, - 0x8b7aff6d, 0x80000000, - 0xbfa10040, 0x847b867b, - 0x847b827b, 0xbef6007b, - 0xb8f03b05, 0x80708170, - 0xbf0d9973, 0xbfa20002, - 0x84708970, 0xbfa00001, - 0x84708a70, 0xb8fa1e06, - 0x847a8a7a, 0x80707a70, - 0x8070ff70, 0x00000200, - 0x8070ff70, 0x00000080, - 0xbef600ff, 0x01000000, - 0xd71f0000, 0x000100c1, - 0xd7200000, 0x000200c1, - 0x16000084, 0x857d9973, + 0xbfa20004, 0xbef000ff, + 0x00000200, 0xbeff0080, + 0xbfa00003, 0xbef000ff, + 0x00000400, 0xbeff00c1, + 0xb8fb3b05, 0x807b817b, + 0x847b827b, 0x857d9973, 0x8b7d817d, 0xbf06817d, - 0xbefd0080, 0xbfa20012, - 0xbe8300ff, 0x00000080, - 0xbf800000, 0xbf800000, - 0xbf800000, 0xd8d80000, - 0x01000000, 0xbf890000, - 0xe0685000, 0x701d0100, - 0x807d037d, 0x80700370, - 0xd5250000, 0x0001ff00, - 0x00000080, 0xbf0a7b7d, - 0xbfa2fff4, 0xbfa00011, - 0xbe8300ff, 0x00000100, - 0xbf800000, 0xbf800000, - 0xbf800000, 0xd8d80000, - 0x01000000, 0xbf890000, - 0xe0685000, 0x701d0100, - 0x807d037d, 0x80700370, - 0xd5250000, 0x0001ff00, - 0x00000100, 0xbf0a7b7d, - 0xbfa2fff4, 0xbefe00c1, - 0x857d9973, 0x8b7d817d, - 0xbf06817d, 0xbfa20004, - 0xbef000ff, 0x00000200, - 0xbeff0080, 0xbfa00003, - 0xbef000ff, 0x00000400, - 0xbeff00c1, 0xb8fb3b05, - 0x807b817b, 0x847b827b, - 0x857d9973, 0x8b7d817d, - 0xbf06817d, 0xbfa20017, + 0xbfa20017, 0xbef600ff, + 0x01000000, 0xbefd0084, + 0xbf0a7b7d, 0xbfa10037, + 0x7e008700, 0x7e028701, + 0x7e048702, 0x7e068703, + 0xe0685000, 0x701d0000, + 0xe0685080, 0x701d0100, + 0xe0685100, 0x701d0200, + 0xe0685180, 0x701d0300, + 0x807d847d, 0x8070ff70, + 0x00000200, 0xbf0a7b7d, + 0xbfa2ffef, 0xbfa00025, 0xbef600ff, 0x01000000, 0xbefd0084, 0xbf0a7b7d, - 0xbfa10037, 0x7e008700, + 0xbfa10011, 0x7e008700, 0x7e028701, 0x7e048702, 0x7e068703, 0xe0685000, - 0x701d0000, 0xe0685080, - 0x701d0100, 0xe0685100, - 0x701d0200, 0xe0685180, + 0x701d0000, 0xe0685100, + 0x701d0100, 0xe0685200, + 0x701d0200, 0xe0685300, 0x701d0300, 0x807d847d, - 0x8070ff70, 0x00000200, + 0x8070ff70, 0x00000400, 0xbf0a7b7d, 0xbfa2ffef, - 0xbfa00025, 0xbef600ff, - 0x01000000, 0xbefd0084, - 0xbf0a7b7d, 0xbfa10011, - 0x7e008700, 0x7e028701, - 0x7e048702, 0x7e068703, + 0xb8fb1e06, 0x8b7bc17b, + 0xbfa1000c, 0x847b837b, + 0x807b7d7b, 0xbefe00c1, + 0xbeff0080, 0x7e008700, 0xe0685000, 0x701d0000, - 0xe0685100, 0x701d0100, - 0xe0685200, 0x701d0200, - 0xe0685300, 0x701d0300, - 0x807d847d, 0x8070ff70, - 0x00000400, 0xbf0a7b7d, - 0xbfa2ffef, 0xb8fb1e06, - 0x8b7bc17b, 0xbfa1000c, - 0x847b837b, 0x807b7d7b, - 0xbefe00c1, 0xbeff0080, - 0x7e008700, 0xe0685000, - 0x701d0000, 0x807d817d, - 0x8070ff70, 0x00000080, - 0xbf0a7b7d, 0xbfa2fff8, - 0xbfa00141, 0xbef4007e, - 0x8b75ff7f, 0x0000ffff, - 0x8c75ff75, 0x00040000, - 0xbef60080, 0xbef700ff, - 0x10807fac, 0xb8f202dc, - 0x84729972, 0x8b6eff7f, - 0x04000000, 0xbfa1003a, + 0x807d817d, 0x8070ff70, + 0x00000080, 0xbf0a7b7d, + 0xbfa2fff8, 0xbfa00146, + 0xbef4007e, 0x8b75ff7f, + 0x0000ffff, 0x8c75ff75, + 0x00040000, 0xbef60080, + 0xbef700ff, 0x10807fac, + 0xb8f202dc, 0x84729972, + 0x8b6eff7f, 0x04000000, + 0xbfa1003a, 0xbefe00c1, + 0x857d9972, 0x8b7d817d, + 0xbf06817d, 0xbfa20002, + 0xbeff0080, 0xbfa00001, + 0xbeff00c1, 0xb8ef4306, + 0x8b6fc16f, 0xbfa1002f, + 0x846f866f, 0x846f826f, + 0xbef6006f, 0xb8f83b05, + 0x80788178, 0xbf0d9972, + 0xbfa20002, 0x84788978, + 0xbfa00001, 0x84788a78, + 0xb8ee1e06, 0x846e8a6e, + 0x80786e78, 0x8078ff78, + 0x00000200, 0x8078ff78, + 0x00000080, 0xbef600ff, + 0x01000000, 0x857d9972, + 0x8b7d817d, 0xbf06817d, + 0xbefd0080, 0xbfa2000c, + 0xe0500000, 0x781d0000, + 0xbf8903f7, 0xdac00000, + 0x00000000, 0x807dff7d, + 0x00000080, 0x8078ff78, + 0x00000080, 0xbf0a6f7d, + 0xbfa2fff5, 0xbfa0000b, + 0xe0500000, 0x781d0000, + 0xbf8903f7, 0xdac00000, + 0x00000000, 0x807dff7d, + 0x00000100, 0x8078ff78, + 0x00000100, 0xbf0a6f7d, + 0xbfa2fff5, 0xbef80080, 0xbefe00c1, 0x857d9972, 0x8b7d817d, 0xbf06817d, 0xbfa20002, 0xbeff0080, 0xbfa00001, 0xbeff00c1, - 0xb8ef4306, 0x8b6fc16f, - 0xbfa1002f, 0x846f866f, - 0x846f826f, 0xbef6006f, - 0xb8f83b05, 0x80788178, - 0xbf0d9972, 0xbfa20002, - 0x84788978, 0xbfa00001, - 0x84788a78, 0xb8ee1e06, - 0x846e8a6e, 0x80786e78, + 0xb8ef3b05, 0x806f816f, + 0x846f826f, 0x857d9972, + 0x8b7d817d, 0xbf06817d, + 0xbfa20024, 0xbef600ff, + 0x01000000, 0xbeee0078, 0x8078ff78, 0x00000200, - 0x8078ff78, 0x00000080, - 0xbef600ff, 0x01000000, - 0x857d9972, 0x8b7d817d, - 0xbf06817d, 0xbefd0080, - 0xbfa2000c, 0xe0500000, - 0x781d0000, 0xbf8903f7, - 0xdac00000, 0x00000000, - 0x807dff7d, 0x00000080, - 0x8078ff78, 0x00000080, - 0xbf0a6f7d, 0xbfa2fff5, - 0xbfa0000b, 0xe0500000, - 0x781d0000, 0xbf8903f7, - 0xdac00000, 0x00000000, - 0x807dff7d, 0x00000100, - 0x8078ff78, 0x00000100, - 0xbf0a6f7d, 0xbfa2fff5, - 0xbef80080, 0xbefe00c1, - 0x857d9972, 0x8b7d817d, - 0xbf06817d, 0xbfa20002, - 0xbeff0080, 0xbfa00001, - 0xbeff00c1, 0xb8ef3b05, - 0x806f816f, 0x846f826f, - 0x857d9972, 0x8b7d817d, - 0xbf06817d, 0xbfa20024, - 0xbef600ff, 0x01000000, - 0xbeee0078, 0x8078ff78, - 0x00000200, 0xbefd0084, - 0xbf0a6f7d, 0xbfa10050, + 0xbefd0084, 0xbf0a6f7d, + 0xbfa10050, 0xe0505000, + 0x781d0000, 0xe0505080, + 0x781d0100, 0xe0505100, + 0x781d0200, 0xe0505180, + 0x781d0300, 0xbf8903f7, + 0x7e008500, 0x7e028501, + 0x7e048502, 0x7e068503, + 0x807d847d, 0x8078ff78, + 0x00000200, 0xbf0a6f7d, + 0xbfa2ffee, 0xe0505000, + 0x6e1d0000, 0xe0505080, + 0x6e1d0100, 0xe0505100, + 0x6e1d0200, 0xe0505180, + 0x6e1d0300, 0xbf8903f7, + 0xbfa00034, 0xbef600ff, + 0x01000000, 0xbeee0078, + 0x8078ff78, 0x00000400, + 0xbefd0084, 0xbf0a6f7d, + 0xbfa10012, 0xe0505000, + 0x781d0000, 0xe0505100, + 0x781d0100, 0xe0505200, + 0x781d0200, 0xe0505300, + 0x781d0300, 0xbf8903f7, + 0x7e008500, 0x7e028501, + 0x7e048502, 0x7e068503, + 0x807d847d, 0x8078ff78, + 0x00000400, 0xbf0a6f7d, + 0xbfa2ffee, 0xb8ef1e06, + 0x8b6fc16f, 0xbfa1000e, + 0x846f836f, 0x806f7d6f, + 0xbefe00c1, 0xbeff0080, 0xe0505000, 0x781d0000, - 0xe0505080, 0x781d0100, - 0xe0505100, 0x781d0200, - 0xe0505180, 0x781d0300, 0xbf8903f7, 0x7e008500, - 0x7e028501, 0x7e048502, - 0x7e068503, 0x807d847d, - 0x8078ff78, 0x00000200, - 0xbf0a6f7d, 0xbfa2ffee, + 0x807d817d, 0x8078ff78, + 0x00000080, 0xbf0a6f7d, + 0xbfa2fff7, 0xbeff00c1, 0xe0505000, 0x6e1d0000, - 0xe0505080, 0x6e1d0100, - 0xe0505100, 0x6e1d0200, - 0xe0505180, 0x6e1d0300, - 0xbf8903f7, 0xbfa00034, - 0xbef600ff, 0x01000000, - 0xbeee0078, 0x8078ff78, - 0x00000400, 0xbefd0084, - 0xbf0a6f7d, 0xbfa10012, - 0xe0505000, 0x781d0000, - 0xe0505100, 0x781d0100, - 0xe0505200, 0x781d0200, - 0xe0505300, 0x781d0300, - 0xbf8903f7, 0x7e008500, - 0x7e028501, 0x7e048502, - 0x7e068503, 0x807d847d, - 0x8078ff78, 0x00000400, - 0xbf0a6f7d, 0xbfa2ffee, - 0xb8ef1e06, 0x8b6fc16f, - 0xbfa1000e, 0x846f836f, - 0x806f7d6f, 0xbefe00c1, - 0xbeff0080, 0xe0505000, - 0x781d0000, 0xbf8903f7, - 0x7e008500, 0x807d817d, - 0x8078ff78, 0x00000080, - 0xbf0a6f7d, 0xbfa2fff7, - 0xbeff00c1, 0xe0505000, - 0x6e1d0000, 0xe0505100, - 0x6e1d0100, 0xe0505200, - 0x6e1d0200, 0xe0505300, - 0x6e1d0300, 0xbf8903f7, + 0xe0505100, 0x6e1d0100, + 0xe0505200, 0x6e1d0200, + 0xe0505300, 0x6e1d0300, + 0xbf8903f7, 0xb8f83b05, + 0x80788178, 0xbf0d9972, + 0xbfa20002, 0x84788978, + 0xbfa00001, 0x84788a78, + 0xb8ee1e06, 0x846e8a6e, + 0x80786e78, 0x8078ff78, + 0x00000200, 0x80f8ff78, + 0x00000050, 0xbef600ff, + 0x01000000, 0xbefd00ff, + 0x0000006c, 0x80f89078, + 0xf428403a, 0xf0000000, + 0xbf89fc07, 0x80fd847d, + 0xbf800000, 0xbe804300, + 0xbe824302, 0x80f8a078, + 0xf42c403a, 0xf0000000, + 0xbf89fc07, 0x80fd887d, + 0xbf800000, 0xbe804300, + 0xbe824302, 0xbe844304, + 0xbe864306, 0x80f8c078, + 0xf430403a, 0xf0000000, + 0xbf89fc07, 0x80fd907d, + 0xbf800000, 0xbe804300, + 0xbe824302, 0xbe844304, + 0xbe864306, 0xbe884308, + 0xbe8a430a, 0xbe8c430c, + 0xbe8e430e, 0xbf06807d, + 0xbfa1fff0, 0xb980f801, + 0x00000000, 0xbfbd0000, 0xb8f83b05, 0x80788178, 0xbf0d9972, 0xbfa20002, 0x84788978, 0xbfa00001, 0x84788a78, 0xb8ee1e06, 0x846e8a6e, 0x80786e78, 0x8078ff78, 0x00000200, - 0x80f8ff78, 0x00000050, 0xbef600ff, 0x01000000, - 0xbefd00ff, 0x0000006c, - 0x80f89078, 0xf428403a, - 0xf0000000, 0xbf89fc07, - 0x80fd847d, 0xbf800000, - 0xbe804300, 0xbe824302, - 0x80f8a078, 0xf42c403a, - 0xf0000000, 0xbf89fc07, - 0x80fd887d, 0xbf800000, - 0xbe804300, 0xbe824302, - 0xbe844304, 0xbe864306, - 0x80f8c078, 0xf430403a, - 0xf0000000, 0xbf89fc07, - 0x80fd907d, 0xbf800000, - 0xbe804300, 0xbe824302, - 0xbe844304, 0xbe864306, - 0xbe884308, 0xbe8a430a, - 0xbe8c430c, 0xbe8e430e, - 0xbf06807d, 0xbfa1fff0, - 0xb980f801, 0x00000000, - 0xbfbd0000, 0xb8f83b05, - 0x80788178, 0xbf0d9972, - 0xbfa20002, 0x84788978, - 0xbfa00001, 0x84788a78, - 0xb8ee1e06, 0x846e8a6e, - 0x80786e78, 0x8078ff78, - 0x00000200, 0xbef600ff, - 0x01000000, 0xf4205bfa, + 0xf4205bfa, 0xf0000000, + 0x80788478, 0xf4205b3a, 0xf0000000, 0x80788478, - 0xf4205b3a, 0xf0000000, - 0x80788478, 0xf4205b7a, + 0xf4205b7a, 0xf0000000, + 0x80788478, 0xf4205c3a, 0xf0000000, 0x80788478, - 0xf4205c3a, 0xf0000000, - 0x80788478, 0xf4205c7a, + 0xf4205c7a, 0xf0000000, + 0x80788478, 0xf4205eba, 0xf0000000, 0x80788478, - 0xf4205eba, 0xf0000000, - 0x80788478, 0xf4205efa, + 0xf4205efa, 0xf0000000, + 0x80788478, 0xf4205e7a, 0xf0000000, 0x80788478, - 0xf4205e7a, 0xf0000000, - 0x80788478, 0xf4205cfa, + 0xf4205cfa, 0xf0000000, + 0x80788478, 0xf4205bba, 0xf0000000, 0x80788478, + 0xbf89fc07, 0xb96ef814, 0xf4205bba, 0xf0000000, 0x80788478, 0xbf89fc07, - 0xb96ef814, 0xf4205bba, - 0xf0000000, 0x80788478, - 0xbf89fc07, 0xb96ef815, - 0xbefd006f, 0xbefe0070, - 0xbeff0071, 0x8b6f7bff, - 0x000003ff, 0xb96f4803, - 0x8b6f7bff, 0xfffff800, - 0x856f8b6f, 0xb96fa2c3, - 0xb973f801, 0xb8ee3b05, - 0x806e816e, 0xbf0d9972, - 0xbfa20002, 0x846e896e, - 0xbfa00001, 0x846e8a6e, - 0xb8ef1e06, 0x846f8a6f, - 0x806e6f6e, 0x806eff6e, - 0x00000200, 0x806e746e, - 0x826f8075, 0x8b6fff6f, - 0x0000ffff, 0xf4085c37, - 0xf8000050, 0xf4085d37, - 0xf8000060, 0xf4005e77, - 0xf8000074, 0xbf89fc07, - 0x8b6dff6d, 0x0000ffff, - 0x8bfe7e7e, 0x8bea6a6a, + 0xb96ef815, 0xbefd006f, + 0xbefe0070, 0xbeff0071, + 0x8b6f7bff, 0x000003ff, + 0xb96f4803, 0x8b6f7bff, + 0xfffff800, 0x856f8b6f, + 0xb96fa2c3, 0xb973f801, + 0xb8ee3b05, 0x806e816e, + 0xbf0d9972, 0xbfa20002, + 0x846e896e, 0xbfa00001, + 0x846e8a6e, 0xb8ef1e06, + 0x846f8a6f, 0x806e6f6e, + 0x806eff6e, 0x00000200, + 0x806e746e, 0x826f8075, + 0x8b6fff6f, 0x0000ffff, + 0xf4085c37, 0xf8000050, + 0xf4085d37, 0xf8000060, + 0xf4005e77, 0xf8000074, + 0xbf89fc07, 0x8b6dff6d, + 0x0000ffff, 0x8bfe7e7e, + 0x8bea6a6a, 0xb8eef802, + 0xbf0d866e, 0xbfa20002, + 0xb97af802, 0xbe80486c, 0xb97af802, 0xbe804a6c, 0xbfb00000, 0xbf9f0000, 0xbf9f0000, 0xbf9f0000, diff --git a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx10.asm b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx10.asm index 250ab007399b..0f81670f6f9c 100644 --- a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx10.asm +++ b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx10.asm @@ -43,12 +43,14 @@ #define HAVE_XNACK (ASIC_FAMILY < CHIP_SIENNA_CICHLID) #define HAVE_SENDMSG_RTN (ASIC_FAMILY >= CHIP_PLUM_BONITO) #define HAVE_BUFFER_LDS_LOAD (ASIC_FAMILY < CHIP_PLUM_BONITO) +#define SW_SA_TRAP (ASIC_FAMILY >= CHIP_PLUM_BONITO) var SINGLE_STEP_MISSED_WORKAROUND = 1 //workaround for lost MODE.DEBUG_EN exception when SAVECTX raised var SQ_WAVE_STATUS_SPI_PRIO_MASK = 0x00000006 var SQ_WAVE_STATUS_HALT_MASK = 0x2000 var SQ_WAVE_STATUS_ECC_ERR_MASK = 0x20000 +var SQ_WAVE_STATUS_TRAP_EN_SHIFT = 6 var SQ_WAVE_LDS_ALLOC_LDS_SIZE_SHIFT = 12 var SQ_WAVE_LDS_ALLOC_LDS_SIZE_SIZE = 9 @@ -183,6 +185,13 @@ L_SKIP_RESTORE: s_getreg_b32 s_save_trapsts, hwreg(HW_REG_TRAPSTS) +#if SW_SA_TRAP + // If ttmp1[31] is set then trap may occur early. + // Spin wait until SAVECTX exception is raised. + s_bitcmp1_b32 s_save_pc_hi, 31 + s_cbranch_scc1 L_CHECK_SAVE +#endif + s_and_b32 ttmp2, s_save_status, SQ_WAVE_STATUS_HALT_MASK s_cbranch_scc0 L_NOT_HALTED @@ -1061,8 +1070,20 @@ L_RESTORE_HWREG: s_and_b32 s_restore_pc_hi, s_restore_pc_hi, 0x0000ffff //pc[47:32] //Do it here in order not to affect STATUS s_and_b64 exec, exec, exec // Restore STATUS.EXECZ, not writable by s_setreg_b32 s_and_b64 vcc, vcc, vcc // Restore STATUS.VCCZ, not writable by s_setreg_b32 + +#if SW_SA_TRAP + // If traps are enabled then return to the shader with PRIV=0. + // Otherwise retain PRIV=1 for subsequent context save requests. + s_getreg_b32 s_restore_tmp, hwreg(HW_REG_STATUS) + s_bitcmp1_b32 s_restore_tmp, SQ_WAVE_STATUS_TRAP_EN_SHIFT + s_cbranch_scc1 L_RETURN_WITHOUT_PRIV + s_setreg_b32 hwreg(HW_REG_STATUS), s_restore_status // SCC is included, which is changed by previous salu + s_setpc_b64 [s_restore_pc_lo, s_restore_pc_hi] +L_RETURN_WITHOUT_PRIV: +#endif + s_setreg_b32 hwreg(HW_REG_STATUS), s_restore_status // SCC is included, which is changed by previous salu s_rfe_b64 s_restore_pc_lo //Return to the main shader program and resume execution L_END_PGM: diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c index 56f7307c21d2..5feaba6a77de 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c @@ -1584,6 +1584,8 @@ static int kfd_ioctl_smi_events(struct file *filep, return kfd_smi_event_open(pdd->dev, &args->anon_fd); } +#if IS_ENABLED(CONFIG_HSA_AMD_SVM) + static int kfd_ioctl_set_xnack_mode(struct file *filep, struct kfd_process *p, void *data) { @@ -1594,22 +1596,29 @@ static int kfd_ioctl_set_xnack_mode(struct file *filep, if (args->xnack_enabled >= 0) { if (!list_empty(&p->pqm.queues)) { pr_debug("Process has user queues running\n"); - mutex_unlock(&p->mutex); - return -EBUSY; + r = -EBUSY; + goto out_unlock; } - if (args->xnack_enabled && !kfd_process_xnack_mode(p, true)) + + if (p->xnack_enabled == args->xnack_enabled) + goto out_unlock; + + if (args->xnack_enabled && !kfd_process_xnack_mode(p, true)) { r = -EPERM; - else - p->xnack_enabled = args->xnack_enabled; + goto out_unlock; + } + + r = svm_range_switch_xnack_reserve_mem(p, args->xnack_enabled); } else { args->xnack_enabled = p->xnack_enabled; } + +out_unlock: mutex_unlock(&p->mutex); return r; } -#if IS_ENABLED(CONFIG_HSA_AMD_SVM) static int kfd_ioctl_svm(struct file *filep, struct kfd_process *p, void *data) { struct kfd_ioctl_svm_args *args = data; @@ -1629,6 +1638,11 @@ static int kfd_ioctl_svm(struct file *filep, struct kfd_process *p, void *data) return r; } #else +static int kfd_ioctl_set_xnack_mode(struct file *filep, + struct kfd_process *p, void *data) +{ + return -EPERM; +} static int kfd_ioctl_svm(struct file *filep, struct kfd_process *p, void *data) { return -EPERM; diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c index e83725a28106..ecb4c3abc629 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c @@ -205,6 +205,8 @@ static int add_queue_mes(struct device_queue_manager *dqm, struct queue *q, } queue_input.is_kfd_process = 1; + queue_input.is_aql_queue = (q->properties.format == KFD_QUEUE_FORMAT_AQL); + queue_input.queue_size = q->properties.queue_size >> 2; queue_input.paging = false; queue_input.tba_addr = qpd->tba_addr; @@ -1240,6 +1242,24 @@ static void init_interrupts(struct device_queue_manager *dqm) dqm->dev->kfd2kgd->init_interrupts(dqm->dev->adev, i); } +static void init_sdma_bitmaps(struct device_queue_manager *dqm) +{ + unsigned int num_sdma_queues = + min_t(unsigned int, sizeof(dqm->sdma_bitmap)*8, + get_num_sdma_queues(dqm)); + unsigned int num_xgmi_sdma_queues = + min_t(unsigned int, sizeof(dqm->xgmi_sdma_bitmap)*8, + get_num_xgmi_sdma_queues(dqm)); + + if (num_sdma_queues) + dqm->sdma_bitmap = GENMASK_ULL(num_sdma_queues-1, 0); + if (num_xgmi_sdma_queues) + dqm->xgmi_sdma_bitmap = GENMASK_ULL(num_xgmi_sdma_queues-1, 0); + + dqm->sdma_bitmap &= ~get_reserved_sdma_queues_bitmap(dqm); + pr_info("sdma_bitmap: %llx\n", dqm->sdma_bitmap); +} + static int initialize_nocpsch(struct device_queue_manager *dqm) { int pipe, queue; @@ -1268,11 +1288,7 @@ static int initialize_nocpsch(struct device_queue_manager *dqm) memset(dqm->vmid_pasid, 0, sizeof(dqm->vmid_pasid)); - dqm->sdma_bitmap = ~0ULL >> (64 - get_num_sdma_queues(dqm)); - dqm->sdma_bitmap &= ~(get_reserved_sdma_queues_bitmap(dqm)); - pr_info("sdma_bitmap: %llx\n", dqm->sdma_bitmap); - - dqm->xgmi_sdma_bitmap = ~0ULL >> (64 - get_num_xgmi_sdma_queues(dqm)); + init_sdma_bitmaps(dqm); return 0; } @@ -1450,9 +1466,6 @@ static int set_sched_resources(struct device_queue_manager *dqm) static int initialize_cpsch(struct device_queue_manager *dqm) { - uint64_t num_sdma_queues; - uint64_t num_xgmi_sdma_queues; - pr_debug("num of pipes: %d\n", get_pipes_per_mec(dqm)); mutex_init(&dqm->lock_hidden); @@ -1461,24 +1474,10 @@ static int initialize_cpsch(struct device_queue_manager *dqm) dqm->active_cp_queue_count = 0; dqm->gws_queue_count = 0; dqm->active_runlist = false; - - num_sdma_queues = get_num_sdma_queues(dqm); - if (num_sdma_queues >= BITS_PER_TYPE(dqm->sdma_bitmap)) - dqm->sdma_bitmap = ULLONG_MAX; - else - dqm->sdma_bitmap = (BIT_ULL(num_sdma_queues) - 1); - - dqm->sdma_bitmap &= ~(get_reserved_sdma_queues_bitmap(dqm)); - pr_info("sdma_bitmap: %llx\n", dqm->sdma_bitmap); - - num_xgmi_sdma_queues = get_num_xgmi_sdma_queues(dqm); - if (num_xgmi_sdma_queues >= BITS_PER_TYPE(dqm->xgmi_sdma_bitmap)) - dqm->xgmi_sdma_bitmap = ULLONG_MAX; - else - dqm->xgmi_sdma_bitmap = (BIT_ULL(num_xgmi_sdma_queues) - 1); - INIT_WORK(&dqm->hw_exception_work, kfd_process_hw_exception); + init_sdma_bitmaps(dqm); + return 0; } diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v11.c b/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v11.c index a6fcbeeb7428..0d53f6067422 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v11.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v11.c @@ -350,11 +350,11 @@ static void event_interrupt_wq_v11(struct kfd_dev *dev, print_sq_intr_info_inst(context_id0, context_id1); sq_int_priv = REG_GET_FIELD(context_id0, SQ_INTERRUPT_WORD_WAVE_CTXID0, PRIV); - if (sq_int_priv /*&& (kfd_set_dbg_ev_from_interrupt(dev, pasid, + /*if (sq_int_priv && (kfd_set_dbg_ev_from_interrupt(dev, pasid, KFD_CTXID0_DOORBELL_ID(context_id0), KFD_CTXID0_TRAP_CODE(context_id0), - NULL, 0))*/) - return; + NULL, 0))) + return;*/ break; case SQ_INTERRUPT_WORD_ENCODING_ERROR: print_sq_intr_info_error(context_id0, context_id1); diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c b/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c index c70c026c9a93..2797029bd500 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c @@ -223,7 +223,7 @@ svm_migrate_get_vram_page(struct svm_range *prange, unsigned long pfn) page = pfn_to_page(pfn); svm_range_bo_ref(prange->svm_bo); page->zone_device_data = prange->svm_bo; - lock_page(page); + zone_device_page_init(page); } static void @@ -410,7 +410,7 @@ svm_migrate_vma_to_vram(struct amdgpu_device *adev, struct svm_range *prange, uint64_t npages = (end - start) >> PAGE_SHIFT; struct kfd_process_device *pdd; struct dma_fence *mfence = NULL; - struct migrate_vma migrate; + struct migrate_vma migrate = { 0 }; unsigned long cpages = 0; dma_addr_t *scratch; void *buf; @@ -666,7 +666,7 @@ out_oom: static long svm_migrate_vma_to_ram(struct amdgpu_device *adev, struct svm_range *prange, struct vm_area_struct *vma, uint64_t start, uint64_t end, - uint32_t trigger) + uint32_t trigger, struct page *fault_page) { struct kfd_process *p = container_of(prange->svms, struct kfd_process, svms); uint64_t npages = (end - start) >> PAGE_SHIFT; @@ -674,7 +674,7 @@ svm_migrate_vma_to_ram(struct amdgpu_device *adev, struct svm_range *prange, unsigned long cpages = 0; struct kfd_process_device *pdd; struct dma_fence *mfence = NULL; - struct migrate_vma migrate; + struct migrate_vma migrate = { 0 }; dma_addr_t *scratch; void *buf; int r = -ENOMEM; @@ -697,6 +697,7 @@ svm_migrate_vma_to_ram(struct amdgpu_device *adev, struct svm_range *prange, migrate.src = buf; migrate.dst = migrate.src + npages; + migrate.fault_page = fault_page; scratch = (dma_addr_t *)(migrate.dst + npages); kfd_smi_event_migration_start(adev->kfd.dev, p->lead_thread->pid, @@ -764,7 +765,7 @@ out: * 0 - OK, otherwise error code */ int svm_migrate_vram_to_ram(struct svm_range *prange, struct mm_struct *mm, - uint32_t trigger) + uint32_t trigger, struct page *fault_page) { struct amdgpu_device *adev; struct vm_area_struct *vma; @@ -805,7 +806,8 @@ int svm_migrate_vram_to_ram(struct svm_range *prange, struct mm_struct *mm, } next = min(vma->vm_end, end); - r = svm_migrate_vma_to_ram(adev, prange, vma, addr, next, trigger); + r = svm_migrate_vma_to_ram(adev, prange, vma, addr, next, trigger, + fault_page); if (r < 0) { pr_debug("failed %ld to migrate prange %p\n", r, prange); break; @@ -849,7 +851,7 @@ svm_migrate_vram_to_vram(struct svm_range *prange, uint32_t best_loc, pr_debug("from gpu 0x%x to gpu 0x%x\n", prange->actual_loc, best_loc); do { - r = svm_migrate_vram_to_ram(prange, mm, trigger); + r = svm_migrate_vram_to_ram(prange, mm, trigger, NULL); if (r) return r; } while (prange->actual_loc && --retries); @@ -950,7 +952,8 @@ static vm_fault_t svm_migrate_to_ram(struct vm_fault *vmf) } r = svm_migrate_vram_to_ram(prange, vmf->vma->vm_mm, - KFD_MIGRATE_TRIGGER_PAGEFAULT_CPU); + KFD_MIGRATE_TRIGGER_PAGEFAULT_CPU, + vmf->page); if (r) pr_debug("failed %d migrate svms 0x%p range 0x%p [0x%lx 0x%lx]\n", r, prange->svms, prange, prange->start, prange->last); diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_migrate.h b/drivers/gpu/drm/amd/amdkfd/kfd_migrate.h index b3f0754b32fa..a5d7e6d22264 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_migrate.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_migrate.h @@ -43,7 +43,7 @@ enum MIGRATION_COPY_DIR { int svm_migrate_to_vram(struct svm_range *prange, uint32_t best_loc, struct mm_struct *mm, uint32_t trigger); int svm_migrate_vram_to_ram(struct svm_range *prange, struct mm_struct *mm, - uint32_t trigger); + uint32_t trigger, struct page *fault_page); unsigned long svm_migrate_addr_to_pfn(struct amdgpu_device *adev, unsigned long addr); diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v11.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v11.c index d982c154537e..4f6390f3236e 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v11.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v11.c @@ -126,6 +126,10 @@ static void init_mqd(struct mqd_manager *mm, void **mqd, m->compute_static_thread_mgmt_se1 = 0xFFFFFFFF; m->compute_static_thread_mgmt_se2 = 0xFFFFFFFF; m->compute_static_thread_mgmt_se3 = 0xFFFFFFFF; + m->compute_static_thread_mgmt_se4 = 0xFFFFFFFF; + m->compute_static_thread_mgmt_se5 = 0xFFFFFFFF; + m->compute_static_thread_mgmt_se6 = 0xFFFFFFFF; + m->compute_static_thread_mgmt_se7 = 0xFFFFFFFF; m->cp_hqd_persistent_state = CP_HQD_PERSISTENT_STATE__PRELOAD_REQ_MASK | 0x55 << CP_HQD_PERSISTENT_STATE__PRELOAD_SIZE__SHIFT; @@ -329,7 +333,8 @@ static void update_mqd_sdma(struct mqd_manager *mm, void *mqd, << SDMA0_QUEUE0_RB_CNTL__RB_SIZE__SHIFT | q->vmid << SDMA0_QUEUE0_RB_CNTL__RB_VMID__SHIFT | 1 << SDMA0_QUEUE0_RB_CNTL__RPTR_WRITEBACK_ENABLE__SHIFT | - 6 << SDMA0_QUEUE0_RB_CNTL__RPTR_WRITEBACK_TIMER__SHIFT; + 6 << SDMA0_QUEUE0_RB_CNTL__RPTR_WRITEBACK_TIMER__SHIFT | + 1 << SDMA0_QUEUE0_RB_CNTL__F32_WPTR_POLL_ENABLE__SHIFT; m->sdmax_rlcx_rb_base = lower_32_bits(q->queue_address >> 8); m->sdmax_rlcx_rb_base_hi = upper_32_bits(q->queue_address >> 8); diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c index cf5b4005534c..64fdf63093a0 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c @@ -278,7 +278,7 @@ static void svm_range_free(struct svm_range *prange, bool update_mem_usage) svm_range_free_dma_mappings(prange); if (update_mem_usage && !p->xnack_enabled) { - pr_debug("unreserve mem limit: %lld\n", size); + pr_debug("unreserve prange 0x%p size: 0x%llx\n", prange, size); amdgpu_amdkfd_unreserve_mem_limit(NULL, size, KFD_IOC_ALLOC_MEM_FLAGS_USERPTR); } @@ -2913,13 +2913,15 @@ retry_write_locked: */ if (prange->actual_loc) r = svm_migrate_vram_to_ram(prange, mm, - KFD_MIGRATE_TRIGGER_PAGEFAULT_GPU); + KFD_MIGRATE_TRIGGER_PAGEFAULT_GPU, + NULL); else r = 0; } } else { r = svm_migrate_vram_to_ram(prange, mm, - KFD_MIGRATE_TRIGGER_PAGEFAULT_GPU); + KFD_MIGRATE_TRIGGER_PAGEFAULT_GPU, + NULL); } if (r) { pr_debug("failed %d to migrate svms %p [0x%lx 0x%lx]\n", @@ -2956,6 +2958,64 @@ out: return r; } +int +svm_range_switch_xnack_reserve_mem(struct kfd_process *p, bool xnack_enabled) +{ + struct svm_range *prange, *pchild; + uint64_t reserved_size = 0; + uint64_t size; + int r = 0; + + pr_debug("switching xnack from %d to %d\n", p->xnack_enabled, xnack_enabled); + + mutex_lock(&p->svms.lock); + + list_for_each_entry(prange, &p->svms.list, list) { + svm_range_lock(prange); + list_for_each_entry(pchild, &prange->child_list, child_list) { + size = (pchild->last - pchild->start + 1) << PAGE_SHIFT; + if (xnack_enabled) { + amdgpu_amdkfd_unreserve_mem_limit(NULL, size, + KFD_IOC_ALLOC_MEM_FLAGS_USERPTR); + } else { + r = amdgpu_amdkfd_reserve_mem_limit(NULL, size, + KFD_IOC_ALLOC_MEM_FLAGS_USERPTR); + if (r) + goto out_unlock; + reserved_size += size; + } + } + + size = (prange->last - prange->start + 1) << PAGE_SHIFT; + if (xnack_enabled) { + amdgpu_amdkfd_unreserve_mem_limit(NULL, size, + KFD_IOC_ALLOC_MEM_FLAGS_USERPTR); + } else { + r = amdgpu_amdkfd_reserve_mem_limit(NULL, size, + KFD_IOC_ALLOC_MEM_FLAGS_USERPTR); + if (r) + goto out_unlock; + reserved_size += size; + } +out_unlock: + svm_range_unlock(prange); + if (r) + break; + } + + if (r) + amdgpu_amdkfd_unreserve_mem_limit(NULL, reserved_size, + KFD_IOC_ALLOC_MEM_FLAGS_USERPTR); + else + /* Change xnack mode must be inside svms lock, to avoid race with + * svm_range_deferred_list_work unreserve memory in parallel. + */ + p->xnack_enabled = xnack_enabled; + + mutex_unlock(&p->svms.lock); + return r; +} + void svm_range_list_fini(struct kfd_process *p) { struct svm_range *prange; @@ -3220,7 +3280,8 @@ svm_range_trigger_migration(struct mm_struct *mm, struct svm_range *prange, return 0; if (!best_loc) { - r = svm_migrate_vram_to_ram(prange, mm, KFD_MIGRATE_TRIGGER_PREFETCH); + r = svm_migrate_vram_to_ram(prange, mm, + KFD_MIGRATE_TRIGGER_PREFETCH, NULL); *migrated = !r; return r; } @@ -3281,7 +3342,7 @@ static void svm_range_evict_svm_bo_worker(struct work_struct *work) mutex_lock(&prange->migrate_mutex); do { r = svm_migrate_vram_to_ram(prange, mm, - KFD_MIGRATE_TRIGGER_TTM_EVICTION); + KFD_MIGRATE_TRIGGER_TTM_EVICTION, NULL); } while (!r && prange->actual_loc && --retries); if (!r && prange->actual_loc) diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_svm.h b/drivers/gpu/drm/amd/amdkfd/kfd_svm.h index 012c53729516..7a33b93f9df6 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_svm.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_svm.h @@ -203,6 +203,7 @@ void svm_range_list_lock_and_flush_work(struct svm_range_list *svms, struct mm_s void svm_range_bo_unref_async(struct svm_range_bo *svm_bo); void svm_range_set_max_pages(struct amdgpu_device *adev); +int svm_range_switch_xnack_reserve_mem(struct kfd_process *p, bool xnack_enabled); #else diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c index 4c73727e0b7d..3c072754738d 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -1110,7 +1110,8 @@ static int dm_dmub_hw_init(struct amdgpu_device *adev) hw_params.fb[i] = &fb_info->fb[i]; switch (adev->ip_versions[DCE_HWIP][0]) { - case IP_VERSION(3, 1, 3): /* Only for this asic hw internal rev B0 */ + case IP_VERSION(3, 1, 3): + case IP_VERSION(3, 1, 4): hw_params.dpia_supported = true; hw_params.disable_dpia = adev->dm.dc->debug.dpia_debug.bits.disable_dpia; break; @@ -3815,8 +3816,6 @@ static int amdgpu_dm_mode_config_init(struct amdgpu_device *adev) /* indicates support for immediate flip */ adev_to_drm(adev)->mode_config.async_page_flip = true; - adev_to_drm(adev)->mode_config.fb_base = adev->gmc.aper_base; - state = kzalloc(sizeof(*state), GFP_KERNEL); if (!state) return -ENOMEM; @@ -6109,7 +6108,6 @@ static void create_eml_sink(struct amdgpu_dm_connector *aconnector) aconnector->base.name); aconnector->base.force = DRM_FORCE_OFF; - aconnector->base.override_edid = false; return; } @@ -6144,8 +6142,6 @@ static void handle_edid_mgmt(struct amdgpu_dm_connector *aconnector) link->verified_link_cap.link_rate = LINK_RATE_HIGH2; } - - aconnector->base.override_edid = true; create_eml_sink(aconnector); } @@ -7478,15 +7474,15 @@ static void amdgpu_dm_handle_vrr_transition(struct dm_crtc_state *old_state, * We also need vupdate irq for the actual core vblank handling * at end of vblank. */ - dm_set_vupdate_irq(new_state->base.crtc, true); - drm_crtc_vblank_get(new_state->base.crtc); + WARN_ON(dm_set_vupdate_irq(new_state->base.crtc, true) != 0); + WARN_ON(drm_crtc_vblank_get(new_state->base.crtc) != 0); DRM_DEBUG_DRIVER("%s: crtc=%u VRR off->on: Get vblank ref\n", __func__, new_state->base.crtc->base.id); } else if (old_vrr_active && !new_vrr_active) { /* Transition VRR active -> inactive: * Allow vblank irq disable again for fixed refresh rate. */ - dm_set_vupdate_irq(new_state->base.crtc, false); + WARN_ON(dm_set_vupdate_irq(new_state->base.crtc, false) != 0); drm_crtc_vblank_put(new_state->base.crtc); DRM_DEBUG_DRIVER("%s: crtc=%u VRR on->off: Drop vblank ref\n", __func__, new_state->base.crtc->base.id); @@ -8242,23 +8238,6 @@ static void amdgpu_dm_atomic_commit_tail(struct drm_atomic_state *state) mutex_unlock(&dm->dc_lock); } - /* Count number of newly disabled CRTCs for dropping PM refs later. */ - for_each_oldnew_crtc_in_state(state, crtc, old_crtc_state, - new_crtc_state, i) { - if (old_crtc_state->active && !new_crtc_state->active) - crtc_disable_count++; - - dm_new_crtc_state = to_dm_crtc_state(new_crtc_state); - dm_old_crtc_state = to_dm_crtc_state(old_crtc_state); - - /* For freesync config update on crtc state and params for irq */ - update_stream_irq_parameters(dm, dm_new_crtc_state); - - /* Handle vrr on->off / off->on transitions */ - amdgpu_dm_handle_vrr_transition(dm_old_crtc_state, - dm_new_crtc_state); - } - /** * Enable interrupts for CRTCs that are newly enabled or went through * a modeset. It was intentionally deferred until after the front end @@ -8268,16 +8247,29 @@ static void amdgpu_dm_atomic_commit_tail(struct drm_atomic_state *state) for_each_oldnew_crtc_in_state(state, crtc, old_crtc_state, new_crtc_state, i) { struct amdgpu_crtc *acrtc = to_amdgpu_crtc(crtc); #ifdef CONFIG_DEBUG_FS - bool configure_crc = false; enum amdgpu_dm_pipe_crc_source cur_crc_src; #if defined(CONFIG_DRM_AMD_SECURE_DISPLAY) - struct crc_rd_work *crc_rd_wrk = dm->crc_rd_wrk; + struct crc_rd_work *crc_rd_wrk; +#endif +#endif + /* Count number of newly disabled CRTCs for dropping PM refs later. */ + if (old_crtc_state->active && !new_crtc_state->active) + crtc_disable_count++; + + dm_new_crtc_state = to_dm_crtc_state(new_crtc_state); + dm_old_crtc_state = to_dm_crtc_state(old_crtc_state); + + /* For freesync config update on crtc state and params for irq */ + update_stream_irq_parameters(dm, dm_new_crtc_state); + +#ifdef CONFIG_DEBUG_FS +#if defined(CONFIG_DRM_AMD_SECURE_DISPLAY) + crc_rd_wrk = dm->crc_rd_wrk; #endif spin_lock_irqsave(&adev_to_drm(adev)->event_lock, flags); cur_crc_src = acrtc->dm_irq_params.crc_src; spin_unlock_irqrestore(&adev_to_drm(adev)->event_lock, flags); #endif - dm_new_crtc_state = to_dm_crtc_state(new_crtc_state); if (new_crtc_state->active && (!old_crtc_state->active || @@ -8285,16 +8277,19 @@ static void amdgpu_dm_atomic_commit_tail(struct drm_atomic_state *state) dc_stream_retain(dm_new_crtc_state->stream); acrtc->dm_irq_params.stream = dm_new_crtc_state->stream; manage_dm_interrupts(adev, acrtc, true); + } + /* Handle vrr on->off / off->on transitions */ + amdgpu_dm_handle_vrr_transition(dm_old_crtc_state, dm_new_crtc_state); #ifdef CONFIG_DEBUG_FS + if (new_crtc_state->active && + (!old_crtc_state->active || + drm_atomic_crtc_needs_modeset(new_crtc_state))) { /** * Frontend may have changed so reapply the CRC capture * settings for the stream. */ - dm_new_crtc_state = to_dm_crtc_state(new_crtc_state); - if (amdgpu_dm_is_valid_crc_source(cur_crc_src)) { - configure_crc = true; #if defined(CONFIG_DRM_AMD_SECURE_DISPLAY) if (amdgpu_dm_crc_window_is_activated(crtc)) { spin_lock_irqsave(&adev_to_drm(adev)->event_lock, flags); @@ -8306,14 +8301,12 @@ static void amdgpu_dm_atomic_commit_tail(struct drm_atomic_state *state) spin_unlock_irqrestore(&adev_to_drm(adev)->event_lock, flags); } #endif - } - - if (configure_crc) if (amdgpu_dm_crtc_configure_crc_source( crtc, dm_new_crtc_state, cur_crc_src)) DRM_DEBUG_DRIVER("Failed to configure crc source"); -#endif + } } +#endif } for_each_new_crtc_in_state(state, crtc, new_crtc_state, j) @@ -9392,10 +9385,6 @@ static int amdgpu_dm_atomic_check(struct drm_device *dev, } } } - if (!pre_validate_dsc(state, &dm_state, vars)) { - ret = -EINVAL; - goto fail; - } } #endif for_each_oldnew_crtc_in_state(state, crtc, old_crtc_state, new_crtc_state, i) { @@ -9529,6 +9518,15 @@ static int amdgpu_dm_atomic_check(struct drm_device *dev, } } +#if defined(CONFIG_DRM_AMD_DC_DCN) + if (dc_resource_is_dsc_encoding_supported(dc)) { + if (!pre_validate_dsc(state, &dm_state, vars)) { + ret = -EINVAL; + goto fail; + } + } +#endif + /* Run this here since we want to validate the streams we created */ ret = drm_atomic_helper_check_planes(dev, state); if (ret) { diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.c index c8da18e45b0e..26291db0a3cf 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.c @@ -60,11 +60,15 @@ static bool link_supports_psrsu(struct dc_link *link) */ void amdgpu_dm_set_psr_caps(struct dc_link *link) { - if (!(link->connector_signal & SIGNAL_TYPE_EDP)) + if (!(link->connector_signal & SIGNAL_TYPE_EDP)) { + link->psr_settings.psr_feature_enabled = false; return; + } - if (link->type == dc_connection_none) + if (link->type == dc_connection_none) { + link->psr_settings.psr_feature_enabled = false; return; + } if (link->dpcd_caps.psr_info.psr_version == 0) { link->psr_settings.psr_version = DC_PSR_VERSION_UNSUPPORTED; @@ -170,7 +174,13 @@ bool amdgpu_dm_psr_enable(struct dc_stream_state *stream) &stream, 1, ¶ms); - power_opt |= psr_power_opt_z10_static_screen; + /* + * Only enable static-screen optimizations for PSR1. For PSR SU, this + * causes vstartup interrupt issues, used by amdgpu_dm to send vblank + * events. + */ + if (link->psr_settings.psr_version < DC_PSR_VERSION_SU_1) + power_opt |= psr_power_opt_z10_static_screen; return dc_link_set_psr_allow_active(link, &psr_enable, false, false, &power_opt); } diff --git a/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c b/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c index 5d70f9901d13..ee0456b5e14e 100644 --- a/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c +++ b/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c @@ -24,6 +24,7 @@ */ #include "dm_services.h" +#include "core_types.h" #include "ObjectID.h" #include "atomfirmware.h" @@ -50,13 +51,6 @@ #define LAST_RECORD_TYPE 0xff #define SMU9_SYSPLL0_ID 0 -struct i2c_id_config_access { - uint8_t bfI2C_LineMux:4; - uint8_t bfHW_EngineID:3; - uint8_t bfHW_Capable:1; - uint8_t ucAccess; -}; - static enum bp_result get_gpio_i2c_info(struct bios_parser *bp, struct atom_i2c_record *record, struct graphics_object_i2c_info *info); @@ -1374,7 +1368,7 @@ static enum bp_result bios_parser_get_lttpr_interop( default: break; } - + DC_LOG_BIOS("DCE_INFO_CAPS_VBIOS_LTTPR_TRANSPARENT_ENABLE: %d tbl_revision.major = %d tbl_revision.minor = %d\n", *dce_caps, tbl_revision.major, tbl_revision.minor); return result; } @@ -1390,6 +1384,7 @@ static enum bp_result bios_parser_get_lttpr_caps( if (!DATA_TABLES(dce_info)) return BP_RESULT_UNSUPPORTED; + *dce_caps = 0; header = GET_IMAGE(struct atom_common_table_header, DATA_TABLES(dce_info)); get_atom_data_table_revision(header, &tbl_revision); @@ -1423,7 +1418,11 @@ static enum bp_result bios_parser_get_lttpr_caps( default: break; } - + DC_LOG_BIOS("DCE_INFO_CAPS_LTTPR_SUPPORT_ENABLE: %d tbl_revision.major = %d tbl_revision.minor = %d\n", *dce_caps, tbl_revision.major, tbl_revision.minor); + if (dcb->ctx->dc->config.force_bios_enable_lttpr && *dce_caps == 0) { + *dce_caps = 1; + DC_LOG_BIOS("DCE_INFO_CAPS_VBIOS_LTTPR_TRANSPARENT_ENABLE: forced enabled"); + } return result; } @@ -2994,13 +2993,22 @@ static enum bp_result construct_integrated_info( info->ext_disp_conn_info.path[i].ext_encoder_obj_id.id, info->ext_disp_conn_info.path[i].caps ); + if (info->ext_disp_conn_info.path[i].caps & EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN) + DC_LOG_BIOS("BIOS EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN on path %d\n", i); + else if (bp->base.ctx->dc->config.force_bios_fixed_vs) { + info->ext_disp_conn_info.path[i].caps |= EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN; + DC_LOG_BIOS("driver forced EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN on path %d\n", i); + } } - // Log the Checksum and Voltage Swing DC_LOG_BIOS("Integrated info table CHECKSUM: %d\n" "Integrated info table FIX_DP_VOLTAGE_SWING: %d\n", info->ext_disp_conn_info.checksum, info->ext_disp_conn_info.fixdpvoltageswing); + if (bp->base.ctx->dc->config.force_bios_fixed_vs && info->ext_disp_conn_info.fixdpvoltageswing == 0) { + info->ext_disp_conn_info.fixdpvoltageswing = bp->base.ctx->dc->config.force_bios_fixed_vs & 0xF; + DC_LOG_BIOS("driver forced fixdpvoltageswing = %d\n", info->ext_disp_conn_info.fixdpvoltageswing); + } } /* Sort voltage table from low to high*/ for (i = 1; i < NUMBER_OF_DISP_CLK_VOLTAGE; ++i) { diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn20/dcn20_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn20/dcn20_clk_mgr.c index 0d30d1d9d67e..650f3b4b562e 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn20/dcn20_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn20/dcn20_clk_mgr.c @@ -179,7 +179,7 @@ void dcn20_update_clocks_update_dentist(struct clk_mgr_internal *clk_mgr, struct } else if (dispclk_wdivider == 127 && current_dispclk_wdivider != 127) { REG_UPDATE(DENTIST_DISPCLK_CNTL, DENTIST_DISPCLK_WDIVIDER, 126); - REG_WAIT(DENTIST_DISPCLK_CNTL, DENTIST_DISPCLK_CHG_DONE, 1, 50, 100); + REG_WAIT(DENTIST_DISPCLK_CNTL, DENTIST_DISPCLK_CHG_DONE, 1, 50, 2000); for (i = 0; i < clk_mgr->base.ctx->dc->res_pool->pipe_count; i++) { struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i]; struct dccg *dccg = clk_mgr->base.ctx->dc->res_pool->dccg; @@ -206,7 +206,7 @@ void dcn20_update_clocks_update_dentist(struct clk_mgr_internal *clk_mgr, struct REG_UPDATE(DENTIST_DISPCLK_CNTL, DENTIST_DISPCLK_WDIVIDER, dispclk_wdivider); - REG_WAIT(DENTIST_DISPCLK_CNTL, DENTIST_DISPCLK_CHG_DONE, 1, 50, 1000); + REG_WAIT(DENTIST_DISPCLK_CNTL, DENTIST_DISPCLK_CHG_DONE, 1, 50, 2000); REG_UPDATE(DENTIST_DISPCLK_CNTL, DENTIST_DPPCLK_WDIVIDER, dppclk_wdivider); REG_WAIT(DENTIST_DISPCLK_CNTL, DENTIST_DPPCLK_CHG_DONE, 1, 5, 100); diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn314/dcn314_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn314/dcn314_clk_mgr.c index 193a0f3de18d..1131c6d73f6c 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn314/dcn314_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn314/dcn314_clk_mgr.c @@ -137,11 +137,20 @@ static void dcn314_disable_otg_wa(struct clk_mgr *clk_mgr_base, struct dc_state if (pipe->top_pipe || pipe->prev_odm_pipe) continue; if (pipe->stream && (pipe->stream->dpms_off || dc_is_virtual_signal(pipe->stream->signal))) { + struct stream_encoder *stream_enc = pipe->stream_res.stream_enc; + if (disable) { + if (stream_enc && stream_enc->funcs->disable_fifo) + pipe->stream_res.stream_enc->funcs->disable_fifo(stream_enc); + pipe->stream_res.tg->funcs->immediate_disable_crtc(pipe->stream_res.tg); reset_sync_context_for_pipe(dc, context, i); - } else + } else { pipe->stream_res.tg->funcs->enable_crtc(pipe->stream_res.tg); + + if (stream_enc && stream_enc->funcs->enable_fifo) + pipe->stream_res.stream_enc->funcs->enable_fifo(stream_enc); + } } } } diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn314/dcn314_smu.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn314/dcn314_smu.c index 897105d1c111..ef0795b14a1f 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn314/dcn314_smu.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn314/dcn314_smu.c @@ -339,29 +339,24 @@ void dcn314_smu_set_zstate_support(struct clk_mgr_internal *clk_mgr, enum dcn_zs if (!clk_mgr->smu_present) return; - if (!clk_mgr->base.ctx->dc->debug.enable_z9_disable_interface && - (support == DCN_ZSTATE_SUPPORT_ALLOW_Z10_ONLY)) - support = DCN_ZSTATE_SUPPORT_DISALLOW; - - // Arg[15:0] = 8/9/0 for Z8/Z9/disallow -> existing bits // Arg[16] = Disallow Z9 -> new bit switch (support) { case DCN_ZSTATE_SUPPORT_ALLOW: msg_id = VBIOSSMC_MSG_AllowZstatesEntry; - param = 9; + param = (1 << 10) | (1 << 9) | (1 << 8); break; case DCN_ZSTATE_SUPPORT_DISALLOW: msg_id = VBIOSSMC_MSG_AllowZstatesEntry; - param = 8; + param = 0; break; case DCN_ZSTATE_SUPPORT_ALLOW_Z10_ONLY: msg_id = VBIOSSMC_MSG_AllowZstatesEntry; - param = 0x00010008; + param = (1 << 10); break; default: //DCN_ZSTATE_SUPPORT_UNKNOWN diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn32/dcn32_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn32/dcn32_clk_mgr.c index c6785969eb1a..1c612ccf1944 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn32/dcn32_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn32/dcn32_clk_mgr.c @@ -156,12 +156,14 @@ void dcn32_init_clocks(struct clk_mgr *clk_mgr_base) { struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base); unsigned int num_levels; + struct clk_limit_num_entries *num_entries_per_clk = &clk_mgr_base->bw_params->clk_table.num_entries_per_clk; memset(&(clk_mgr_base->clks), 0, sizeof(struct dc_clocks)); clk_mgr_base->clks.p_state_change_support = true; clk_mgr_base->clks.prev_p_state_change_support = true; clk_mgr_base->clks.fclk_prev_p_state_change_support = true; clk_mgr->smu_present = false; + clk_mgr->dpm_present = false; if (!clk_mgr_base->bw_params) return; @@ -178,22 +180,29 @@ void dcn32_init_clocks(struct clk_mgr *clk_mgr_base) /* DCFCLK */ dcn32_init_single_clock(clk_mgr, PPCLK_DCFCLK, &clk_mgr_base->bw_params->clk_table.entries[0].dcfclk_mhz, - &num_levels); + &num_entries_per_clk->num_dcfclk_levels); /* SOCCLK */ dcn32_init_single_clock(clk_mgr, PPCLK_SOCCLK, &clk_mgr_base->bw_params->clk_table.entries[0].socclk_mhz, - &num_levels); + &num_entries_per_clk->num_socclk_levels); + /* DTBCLK */ if (!clk_mgr->base.ctx->dc->debug.disable_dtb_ref_clk_switch) dcn32_init_single_clock(clk_mgr, PPCLK_DTBCLK, &clk_mgr_base->bw_params->clk_table.entries[0].dtbclk_mhz, - &num_levels); + &num_entries_per_clk->num_dtbclk_levels); /* DISPCLK */ dcn32_init_single_clock(clk_mgr, PPCLK_DISPCLK, &clk_mgr_base->bw_params->clk_table.entries[0].dispclk_mhz, - &num_levels); + &num_entries_per_clk->num_dispclk_levels); + num_levels = num_entries_per_clk->num_dispclk_levels; + + if (num_entries_per_clk->num_dcfclk_levels && + num_entries_per_clk->num_dtbclk_levels && + num_entries_per_clk->num_dispclk_levels) + clk_mgr->dpm_present = true; if (clk_mgr_base->ctx->dc->debug.min_disp_clk_khz) { unsigned int i; @@ -325,6 +334,21 @@ static void dcn32_update_clocks(struct clk_mgr *clk_mgr_base, if (enter_display_off == safe_to_lower) dcn30_smu_set_num_of_displays(clk_mgr, display_count); + clk_mgr_base->clks.fclk_prev_p_state_change_support = clk_mgr_base->clks.fclk_p_state_change_support; + + total_plane_count = clk_mgr_helper_get_active_plane_cnt(dc, context); + fclk_p_state_change_support = new_clocks->fclk_p_state_change_support || (total_plane_count == 0); + + if (should_update_pstate_support(safe_to_lower, fclk_p_state_change_support, clk_mgr_base->clks.fclk_p_state_change_support)) { + clk_mgr_base->clks.fclk_p_state_change_support = fclk_p_state_change_support; + + /* To enable FCLK P-state switching, send FCLK_PSTATE_SUPPORTED message to PMFW */ + if (clk_mgr_base->ctx->dce_version != DCN_VERSION_3_21 && clk_mgr_base->clks.fclk_p_state_change_support) { + /* Handle the code for sending a message to PMFW that FCLK P-state change is supported */ + dcn32_smu_send_fclk_pstate_message(clk_mgr, FCLK_PSTATE_SUPPORTED); + } + } + if (dc->debug.force_min_dcfclk_mhz > 0) new_clocks->dcfclk_khz = (new_clocks->dcfclk_khz > (dc->debug.force_min_dcfclk_mhz * 1000)) ? new_clocks->dcfclk_khz : (dc->debug.force_min_dcfclk_mhz * 1000); @@ -344,7 +368,6 @@ static void dcn32_update_clocks(struct clk_mgr *clk_mgr_base, clk_mgr_base->clks.socclk_khz = new_clocks->socclk_khz; clk_mgr_base->clks.prev_p_state_change_support = clk_mgr_base->clks.p_state_change_support; - clk_mgr_base->clks.fclk_prev_p_state_change_support = clk_mgr_base->clks.fclk_p_state_change_support; clk_mgr_base->clks.prev_num_ways = clk_mgr_base->clks.num_ways; if (clk_mgr_base->clks.num_ways != new_clocks->num_ways && @@ -353,27 +376,25 @@ static void dcn32_update_clocks(struct clk_mgr *clk_mgr_base, dcn32_smu_send_cab_for_uclk_message(clk_mgr, clk_mgr_base->clks.num_ways); } - total_plane_count = clk_mgr_helper_get_active_plane_cnt(dc, context); + p_state_change_support = new_clocks->p_state_change_support || (total_plane_count == 0); - fclk_p_state_change_support = new_clocks->fclk_p_state_change_support || (total_plane_count == 0); if (should_update_pstate_support(safe_to_lower, p_state_change_support, clk_mgr_base->clks.p_state_change_support)) { clk_mgr_base->clks.p_state_change_support = p_state_change_support; /* to disable P-State switching, set UCLK min = max */ if (!clk_mgr_base->clks.p_state_change_support) dcn32_smu_set_hard_min_by_freq(clk_mgr, PPCLK_UCLK, - clk_mgr_base->bw_params->clk_table.entries[clk_mgr_base->bw_params->clk_table.num_entries - 1].memclk_mhz); + clk_mgr_base->bw_params->clk_table.entries[clk_mgr_base->bw_params->clk_table.num_entries_per_clk.num_memclk_levels - 1].memclk_mhz); } - if (should_update_pstate_support(safe_to_lower, fclk_p_state_change_support, clk_mgr_base->clks.fclk_p_state_change_support) && - clk_mgr_base->ctx->dce_version != DCN_VERSION_3_21) { - clk_mgr_base->clks.fclk_p_state_change_support = fclk_p_state_change_support; + /* Always update saved value, even if new value not set due to P-State switching unsupported. Also check safe_to_lower for FCLK */ + if (safe_to_lower && (clk_mgr_base->clks.fclk_p_state_change_support != clk_mgr_base->clks.fclk_prev_p_state_change_support)) { + update_fclk = true; + } - /* To disable FCLK P-state switching, send FCLK_PSTATE_NOTSUPPORTED message to PMFW */ - if (clk_mgr_base->ctx->dce_version != DCN_VERSION_3_21 && !clk_mgr_base->clks.fclk_p_state_change_support) { - /* Handle code for sending a message to PMFW that FCLK P-state change is not supported */ - dcn32_smu_send_fclk_pstate_message(clk_mgr, FCLK_PSTATE_NOTSUPPORTED); - } + if (clk_mgr_base->ctx->dce_version != DCN_VERSION_3_21 && !clk_mgr_base->clks.fclk_p_state_change_support && update_fclk) { + /* Handle code for sending a message to PMFW that FCLK P-state change is not supported */ + dcn32_smu_send_fclk_pstate_message(clk_mgr, FCLK_PSTATE_NOTSUPPORTED); } /* Always update saved value, even if new value not set due to P-State switching unsupported */ @@ -382,21 +403,11 @@ static void dcn32_update_clocks(struct clk_mgr *clk_mgr_base, update_uclk = true; } - /* Always update saved value, even if new value not set due to P-State switching unsupported. Also check safe_to_lower for FCLK */ - if (safe_to_lower && (clk_mgr_base->clks.fclk_p_state_change_support != clk_mgr_base->clks.fclk_prev_p_state_change_support)) { - update_fclk = true; - } - /* set UCLK to requested value if P-State switching is supported, or to re-enable P-State switching */ if (clk_mgr_base->clks.p_state_change_support && (update_uclk || !clk_mgr_base->clks.prev_p_state_change_support)) dcn32_smu_set_hard_min_by_freq(clk_mgr, PPCLK_UCLK, khz_to_mhz_ceil(clk_mgr_base->clks.dramclk_khz)); - if (clk_mgr_base->ctx->dce_version != DCN_VERSION_3_21 && clk_mgr_base->clks.fclk_p_state_change_support && update_fclk) { - /* Handle the code for sending a message to PMFW that FCLK P-state change is supported */ - dcn32_smu_send_fclk_pstate_message(clk_mgr, FCLK_PSTATE_SUPPORTED); - } - if (clk_mgr_base->clks.num_ways != new_clocks->num_ways && clk_mgr_base->clks.num_ways > new_clocks->num_ways) { clk_mgr_base->clks.num_ways = new_clocks->num_ways; @@ -624,7 +635,7 @@ static void dcn32_set_hard_min_memclk(struct clk_mgr *clk_mgr_base, bool current khz_to_mhz_ceil(clk_mgr_base->clks.dramclk_khz)); else dcn32_smu_set_hard_min_by_freq(clk_mgr, PPCLK_UCLK, - clk_mgr_base->bw_params->clk_table.entries[clk_mgr_base->bw_params->clk_table.num_entries - 1].memclk_mhz); + clk_mgr_base->bw_params->clk_table.entries[clk_mgr_base->bw_params->clk_table.num_entries_per_clk.num_memclk_levels - 1].memclk_mhz); } else { dcn32_smu_set_hard_min_by_freq(clk_mgr, PPCLK_UCLK, clk_mgr_base->bw_params->clk_table.entries[0].memclk_mhz); @@ -640,24 +651,42 @@ static void dcn32_set_hard_max_memclk(struct clk_mgr *clk_mgr_base) return; dcn30_smu_set_hard_max_by_freq(clk_mgr, PPCLK_UCLK, - clk_mgr_base->bw_params->clk_table.entries[clk_mgr_base->bw_params->clk_table.num_entries - 1].memclk_mhz); + clk_mgr_base->bw_params->clk_table.entries[clk_mgr_base->bw_params->clk_table.num_entries_per_clk.num_memclk_levels - 1].memclk_mhz); } /* Get current memclk states, update bounding box */ static void dcn32_get_memclk_states_from_smu(struct clk_mgr *clk_mgr_base) { struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base); + struct clk_limit_num_entries *num_entries_per_clk = &clk_mgr_base->bw_params->clk_table.num_entries_per_clk; unsigned int num_levels; if (!clk_mgr->smu_present) return; - /* Refresh memclk states */ + /* Refresh memclk and fclk states */ dcn32_init_single_clock(clk_mgr, PPCLK_UCLK, &clk_mgr_base->bw_params->clk_table.entries[0].memclk_mhz, - &num_levels); + &num_entries_per_clk->num_memclk_levels); + + dcn32_init_single_clock(clk_mgr, PPCLK_FCLK, + &clk_mgr_base->bw_params->clk_table.entries[0].fclk_mhz, + &num_entries_per_clk->num_fclk_levels); + + if (num_entries_per_clk->num_memclk_levels >= num_entries_per_clk->num_fclk_levels) { + num_levels = num_entries_per_clk->num_memclk_levels; + } else { + num_levels = num_entries_per_clk->num_fclk_levels; + } + clk_mgr_base->bw_params->clk_table.num_entries = num_levels ? num_levels : 1; + if (clk_mgr->dpm_present && !num_levels) + clk_mgr->dpm_present = false; + + if (!clk_mgr->dpm_present) + dcn32_patch_dpm_table(clk_mgr_base->bw_params); + DC_FP_START(); /* Refresh bounding box */ clk_mgr_base->ctx->dc->res_pool->funcs->update_bw_bounding_box( diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c index 19eb960d75d8..997ab031f816 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc.c @@ -1201,7 +1201,7 @@ static void wait_for_no_pipes_pending(struct dc *dc, struct dc_state *context) int count = 0; struct pipe_ctx *pipe = &context->res_ctx.pipe_ctx[i]; - if (!pipe->plane_state) + if (!pipe->plane_state || pipe->stream->mall_stream_config.type == SUBVP_PHANTOM) continue; /* Timeout 100 ms */ @@ -1734,10 +1734,20 @@ static enum dc_status dc_commit_state_no_check(struct dc *dc, struct dc_state *c int i, k, l; struct dc_stream_state *dc_streams[MAX_STREAMS] = {0}; struct dc_state *old_state; + bool subvp_prev_use = false; dc_z10_restore(dc); dc_allow_idle_optimizations(dc, false); + for (i = 0; i < dc->res_pool->pipe_count; i++) { + struct pipe_ctx *old_pipe = &dc->current_state->res_ctx.pipe_ctx[i]; + + /* Check old context for SubVP */ + subvp_prev_use |= (old_pipe->stream && old_pipe->stream->mall_stream_config.type == SUBVP_PHANTOM); + if (subvp_prev_use) + break; + } + for (i = 0; i < context->stream_count; i++) dc_streams[i] = context->streams[i]; @@ -1750,6 +1760,9 @@ static enum dc_status dc_commit_state_no_check(struct dc *dc, struct dc_state *c context->stream_count == 0) dc->hwss.prepare_bandwidth(dc, context); + if (dc->debug.enable_double_buffered_dsc_pg_support) + dc->hwss.update_dsc_pg(dc, context, false); + disable_dangling_plane(dc, context); /* re-program planes for existing stream, in case we need to * free up plane resource for later use @@ -1774,6 +1787,9 @@ static enum dc_status dc_commit_state_no_check(struct dc *dc, struct dc_state *c dc->hwss.wait_for_mpcc_disconnect(dc, dc->res_pool, pipe); } + if (dc->hwss.subvp_pipe_control_lock) + dc->hwss.subvp_pipe_control_lock(dc, context, true, true, NULL, subvp_prev_use); + result = dc->hwss.apply_ctx_to_hw(dc, context); if (result != DC_OK) { @@ -1791,6 +1807,12 @@ static enum dc_status dc_commit_state_no_check(struct dc *dc, struct dc_state *c dc->hwss.interdependent_update_lock(dc, context, false); dc->hwss.post_unlock_program_front_end(dc, context); } + + if (dc->hwss.commit_subvp_config) + dc->hwss.commit_subvp_config(dc, context); + if (dc->hwss.subvp_pipe_control_lock) + dc->hwss.subvp_pipe_control_lock(dc, context, false, true, NULL, subvp_prev_use); + for (i = 0; i < context->stream_count; i++) { const struct dc_link *link = context->streams[i]->link; @@ -1840,6 +1862,9 @@ static enum dc_status dc_commit_state_no_check(struct dc *dc, struct dc_state *c dc->hwss.optimize_bandwidth(dc, context); } + if (dc->debug.enable_double_buffered_dsc_pg_support) + dc->hwss.update_dsc_pg(dc, context, true); + if (dc->ctx->dce_version >= DCE_VERSION_MAX) TRACE_DCN_CLOCK_STATE(&context->bw_ctx.bw.dcn.clk); else @@ -2003,6 +2028,9 @@ void dc_post_update_surfaces_to_stream(struct dc *dc) dc->hwss.optimize_bandwidth(dc, context); + if (dc->debug.enable_double_buffered_dsc_pg_support) + dc->hwss.update_dsc_pg(dc, context, true); + dc->optimized_required = false; dc->wm_optimized_required = false; } @@ -2322,9 +2350,13 @@ static enum surface_update_type det_surface_update(const struct dc *dc, type = get_scaling_info_update_type(u); elevate_update_type(&overall_type, type); - if (u->flip_addr) + if (u->flip_addr) { update_flags->bits.addr_update = 1; - + if (u->flip_addr->address.tmz_surface != u->surface->address.tmz_surface) { + update_flags->bits.tmz_changed = 1; + elevate_update_type(&overall_type, UPDATE_TYPE_FULL); + } + } if (u->in_transfer_func) update_flags->bits.in_transfer_func_change = 1; @@ -2759,11 +2791,8 @@ static void copy_stream_update_to_stream(struct dc *dc, if (update->abm_level) stream->abm_level = *update->abm_level; - if (update->periodic_interrupt0) - stream->periodic_interrupt0 = *update->periodic_interrupt0; - - if (update->periodic_interrupt1) - stream->periodic_interrupt1 = *update->periodic_interrupt1; + if (update->periodic_interrupt) + stream->periodic_interrupt = *update->periodic_interrupt; if (update->gamut_remap) stream->gamut_remap_matrix = *update->gamut_remap; @@ -2917,6 +2946,12 @@ static bool update_planes_and_stream_state(struct dc *dc, dc_resource_state_copy_construct( dc->current_state, context); + /* For each full update, remove all existing phantom pipes first. + * Ensures that we have enough pipes for newly added MPO planes + */ + if (dc->res_pool->funcs->remove_phantom_pipes) + dc->res_pool->funcs->remove_phantom_pipes(dc, context); + /*remove old surfaces from context */ if (!dc_rem_all_planes_for_stream(dc, stream, context)) { @@ -2983,13 +3018,8 @@ static void commit_planes_do_stream_update(struct dc *dc, if (!pipe_ctx->top_pipe && !pipe_ctx->prev_odm_pipe && pipe_ctx->stream == stream) { - if (stream_update->periodic_interrupt0 && - dc->hwss.setup_periodic_interrupt) - dc->hwss.setup_periodic_interrupt(dc, pipe_ctx, VLINE0); - - if (stream_update->periodic_interrupt1 && - dc->hwss.setup_periodic_interrupt) - dc->hwss.setup_periodic_interrupt(dc, pipe_ctx, VLINE1); + if (stream_update->periodic_interrupt && dc->hwss.setup_periodic_interrupt) + dc->hwss.setup_periodic_interrupt(dc, pipe_ctx); if ((stream_update->hdr_static_metadata && !stream->use_dynamic_meta) || stream_update->vrr_infopacket || @@ -3198,6 +3228,9 @@ static void commit_planes_for_stream(struct dc *dc, if (get_seamless_boot_stream_count(context) == 0) dc->hwss.prepare_bandwidth(dc, context); + if (dc->debug.enable_double_buffered_dsc_pg_support) + dc->hwss.update_dsc_pg(dc, context, false); + context_clock_trace(dc, context); } @@ -3326,8 +3359,14 @@ static void commit_planes_for_stream(struct dc *dc, /* Since phantom pipe programming is moved to post_unlock_program_front_end, * move the SubVP lock to after the phantom pipes have been setup */ - if (dc->hwss.subvp_pipe_control_lock) - dc->hwss.subvp_pipe_control_lock(dc, context, false, should_lock_all_pipes, NULL, subvp_prev_use); + if (should_lock_all_pipes && dc->hwss.interdependent_update_lock) { + if (dc->hwss.subvp_pipe_control_lock) + dc->hwss.subvp_pipe_control_lock(dc, context, false, should_lock_all_pipes, NULL, subvp_prev_use); + } else { + if (dc->hwss.subvp_pipe_control_lock) + dc->hwss.subvp_pipe_control_lock(dc, context, false, should_lock_all_pipes, NULL, subvp_prev_use); + } + return; } @@ -3487,6 +3526,9 @@ static void commit_planes_for_stream(struct dc *dc, if (update_type != UPDATE_TYPE_FAST) dc->hwss.post_unlock_program_front_end(dc, context); + if (update_type != UPDATE_TYPE_FAST) + if (dc->hwss.commit_subvp_config) + dc->hwss.commit_subvp_config(dc, context); if (update_type != UPDATE_TYPE_FAST) if (dc->hwss.commit_subvp_config) @@ -3521,11 +3563,91 @@ static void commit_planes_for_stream(struct dc *dc, } } +/* Determines if the incoming context requires a applying transition state with unnecessary + * pipe splitting and ODM disabled, due to hardware limitations. In a case where + * the OPP associated with an MPCC might change due to plane additions, this function + * returns true. + */ +static bool could_mpcc_tree_change_for_active_pipes(struct dc *dc, + struct dc_stream_state *stream, + int surface_count, + bool *is_plane_addition) +{ + + struct dc_stream_status *cur_stream_status = stream_get_status(dc->current_state, stream); + bool force_minimal_pipe_splitting = false; + uint32_t i; + + *is_plane_addition = false; + + if (cur_stream_status && + dc->current_state->stream_count > 0 && + dc->debug.pipe_split_policy != MPC_SPLIT_AVOID) { + /* determine if minimal transition is required due to MPC*/ + if (surface_count > 0) { + if (cur_stream_status->plane_count > surface_count) { + force_minimal_pipe_splitting = true; + } else if (cur_stream_status->plane_count < surface_count) { + force_minimal_pipe_splitting = true; + *is_plane_addition = true; + } + } + } + + if (cur_stream_status && + dc->current_state->stream_count == 1 && + dc->debug.enable_single_display_2to1_odm_policy) { + /* determine if minimal transition is required due to dynamic ODM*/ + if (surface_count > 0) { + if (cur_stream_status->plane_count > 2 && cur_stream_status->plane_count > surface_count) { + force_minimal_pipe_splitting = true; + } else if (surface_count > 2 && cur_stream_status->plane_count < surface_count) { + force_minimal_pipe_splitting = true; + *is_plane_addition = true; + } + } + } + + /* For SubVP pipe split case when adding MPO video + * we need to add a minimal transition. In this case + * there will be 2 streams (1 main stream, 1 phantom + * stream). + */ + if (cur_stream_status && + dc->current_state->stream_count == 2 && + stream->mall_stream_config.type == SUBVP_MAIN) { + bool is_pipe_split = false; + + for (i = 0; i < dc->res_pool->pipe_count; i++) { + if (dc->current_state->res_ctx.pipe_ctx[i].stream == stream && + (dc->current_state->res_ctx.pipe_ctx[i].bottom_pipe || + dc->current_state->res_ctx.pipe_ctx[i].next_odm_pipe)) { + is_pipe_split = true; + break; + } + } + + /* determine if minimal transition is required due to SubVP*/ + if (surface_count > 0 && is_pipe_split) { + if (cur_stream_status->plane_count > surface_count) { + force_minimal_pipe_splitting = true; + } else if (cur_stream_status->plane_count < surface_count) { + force_minimal_pipe_splitting = true; + *is_plane_addition = true; + } + } + } + + return force_minimal_pipe_splitting; +} + static bool commit_minimal_transition_state(struct dc *dc, struct dc_state *transition_base_context) { struct dc_state *transition_context = dc_create_state(dc); - enum pipe_split_policy tmp_policy; + enum pipe_split_policy tmp_mpc_policy; + bool temp_dynamic_odm_policy; + bool temp_subvp_policy; enum dc_status ret = DC_ERROR_UNEXPECTED; unsigned int i, j; @@ -3533,10 +3655,16 @@ static bool commit_minimal_transition_state(struct dc *dc, return false; if (!dc->config.is_vmin_only_asic) { - tmp_policy = dc->debug.pipe_split_policy; + tmp_mpc_policy = dc->debug.pipe_split_policy; dc->debug.pipe_split_policy = MPC_SPLIT_AVOID; } + temp_dynamic_odm_policy = dc->debug.enable_single_display_2to1_odm_policy; + dc->debug.enable_single_display_2to1_odm_policy = false; + + temp_subvp_policy = dc->debug.force_disable_subvp; + dc->debug.force_disable_subvp = true; + dc_resource_state_copy_construct(transition_base_context, transition_context); //commit minimal state @@ -3557,20 +3685,23 @@ static bool commit_minimal_transition_state(struct dc *dc, ret = dc_commit_state_no_check(dc, transition_context); } - //always release as dc_commit_state_no_check retains in good case + /*always release as dc_commit_state_no_check retains in good case*/ dc_release_state(transition_context); - //restore previous pipe split policy + /*restore previous pipe split and odm policy*/ if (!dc->config.is_vmin_only_asic) - dc->debug.pipe_split_policy = tmp_policy; + dc->debug.pipe_split_policy = tmp_mpc_policy; + + dc->debug.enable_single_display_2to1_odm_policy = temp_dynamic_odm_policy; + dc->debug.force_disable_subvp = temp_subvp_policy; if (ret != DC_OK) { - //this should never happen + /*this should never happen*/ BREAK_TO_DEBUGGER(); return false; } - //force full surface update + /*force full surface update*/ for (i = 0; i < dc->current_state->stream_count; i++) { for (j = 0; j < dc->current_state->stream_status[i].plane_count; j++) { dc->current_state->stream_status[i].plane_states[j]->update_flags.raw = 0xFFFFFFFF; @@ -3593,24 +3724,14 @@ bool dc_update_planes_and_stream(struct dc *dc, * cause underflow. Apply stream configuration with minimal pipe * split first to avoid unsupported transitions for active pipes. */ - bool force_minimal_pipe_splitting = false; - bool is_plane_addition = false; + bool force_minimal_pipe_splitting; + bool is_plane_addition; - struct dc_stream_status *cur_stream_status = stream_get_status(dc->current_state, stream); - - if (cur_stream_status && - dc->current_state->stream_count > 0 && - dc->debug.pipe_split_policy != MPC_SPLIT_AVOID) { - /* determine if minimal transition is required */ - if (surface_count > 0) { - if (cur_stream_status->plane_count > surface_count) { - force_minimal_pipe_splitting = true; - } else if (cur_stream_status->plane_count < surface_count) { - force_minimal_pipe_splitting = true; - is_plane_addition = true; - } - } - } + force_minimal_pipe_splitting = could_mpcc_tree_change_for_active_pipes( + dc, + stream, + surface_count, + &is_plane_addition); /* on plane addition, minimal state is the current one */ if (force_minimal_pipe_splitting && is_plane_addition && @@ -3627,7 +3748,7 @@ bool dc_update_planes_and_stream(struct dc *dc, &context)) return false; - /* on plane addition, minimal state is the new one */ + /* on plane removal, minimal state is the new one */ if (force_minimal_pipe_splitting && !is_plane_addition) { if (!commit_minimal_transition_state(dc, context)) { dc_release_state(context); @@ -4014,7 +4135,7 @@ struct dc_sink *dc_link_add_remote_sink( * Treat device as no EDID device if EDID * parsing fails */ - if (edid_status != EDID_OK) { + if (edid_status != EDID_OK && edid_status != EDID_PARTIAL_VALID) { dc_sink->dc_edid.length = 0; dm_error("Bad EDID, status%d!\n", edid_status); } @@ -4536,6 +4657,37 @@ enum dc_status dc_process_dmub_set_mst_slots(const struct dc *dc, } /** + ***************************************************************************** + * Function: dc_process_dmub_dpia_hpd_int_enable + * + * @brief + * Submits dpia hpd int enable command to dmub via inbox message + * + * @param + * [in] dc: dc structure + * [in] hpd_int_enable: 1 for hpd int enable, 0 to disable + * + * @return + * None + ***************************************************************************** + */ +void dc_process_dmub_dpia_hpd_int_enable(const struct dc *dc, + uint32_t hpd_int_enable) +{ + union dmub_rb_cmd cmd = {0}; + struct dc_dmub_srv *dmub_srv = dc->ctx->dmub_srv; + + cmd.dpia_hpd_int_enable.header.type = DMUB_CMD__DPIA_HPD_INT_ENABLE; + cmd.dpia_hpd_int_enable.enable = hpd_int_enable; + + dc_dmub_srv_cmd_queue(dmub_srv, &cmd); + dc_dmub_srv_cmd_execute(dmub_srv); + dc_dmub_srv_wait_idle(dmub_srv); + + DC_LOG_DEBUG("%s: hpd_int_enable(%d)\n", __func__, hpd_int_enable); +} + +/** * dc_disable_accelerated_mode - disable accelerated mode * @dc: dc structure */ diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link.c b/drivers/gpu/drm/amd/display/dc/core/dc_link.c index 3d19fb92333b..d7b1ace6328a 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_link.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_link.c @@ -1307,7 +1307,10 @@ static bool detect_link_and_local_sink(struct dc_link *link, } if (link->connector_signal == SIGNAL_TYPE_EDP) { - // Init dc_panel_config + /* Init dc_panel_config by HW config */ + if (dc_ctx->dc->res_pool->funcs->get_panel_config_defaults) + dc_ctx->dc->res_pool->funcs->get_panel_config_defaults(&link->panel_config); + /* Pickup base DM settings */ dm_helpers_init_panel_settings(dc_ctx, &link->panel_config, sink); // Override dc_panel_config if system has specific settings dm_helpers_override_panel_settings(dc_ctx, &link->panel_config); @@ -3143,7 +3146,7 @@ bool dc_link_set_psr_allow_active(struct dc_link *link, const bool *allow_active if (!dc_get_edp_link_panel_inst(dc, link, &panel_inst)) return false; - if (allow_active && link->type == dc_connection_none) { + if ((allow_active != NULL) && (*allow_active == true) && (link->type == dc_connection_none)) { // Don't enter PSR if panel is not connected return false; } @@ -3375,8 +3378,8 @@ bool dc_link_setup_psr(struct dc_link *link, case FAMILY_YELLOW_CARP: case AMDGPU_FAMILY_GC_10_3_6: case AMDGPU_FAMILY_GC_11_0_1: - if(!dc->debug.disable_z10) - psr_context->psr_level.bits.SKIP_CRTC_DISABLE = false; + if (dc->debug.disable_z10) + psr_context->psr_level.bits.SKIP_CRTC_DISABLE = true; break; default: psr_context->psr_level.bits.SKIP_CRTC_DISABLE = true; diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c index d01d2eeed813..651231387043 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c @@ -35,6 +35,8 @@ #include "dc_link_ddc.h" #include "dce/dce_aux.h" #include "dmub/inc/dmub_cmd.h" +#include "link_dpcd.h" +#include "include/dal_asic_id.h" #define DC_LOGGER_INIT(logger) @@ -683,6 +685,21 @@ bool dc_link_aux_try_to_configure_timeout(struct ddc_service *ddc, bool result = false; struct ddc *ddc_pin = ddc->ddc_pin; + if ((ddc->link->chip_caps & EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN) && + !ddc->link->dc->debug.disable_fixed_vs_aux_timeout_wa && + ASICREV_IS_YELLOW_CARP(ddc->ctx->asic_id.hw_internal_rev)) { + /* Fixed VS workaround for AUX timeout */ + const uint32_t fixed_vs_address = 0xF004F; + const uint8_t fixed_vs_data[4] = {0x1, 0x22, 0x63, 0xc}; + + core_link_write_dpcd(ddc->link, + fixed_vs_address, + fixed_vs_data, + sizeof(fixed_vs_data)); + + timeout = 3072; + } + /* Do not try to access nonexistent DDC pin. */ if (ddc->link->ep_type != DISPLAY_ENDPOINT_PHY) return true; @@ -691,6 +708,7 @@ bool dc_link_aux_try_to_configure_timeout(struct ddc_service *ddc, ddc->ctx->dc->res_pool->engines[ddc_pin->pin_data->en]->funcs->configure_timeout(ddc, timeout); result = true; } + return result; } diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c index 9eab6f48de1a..1254d38f1778 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c @@ -526,9 +526,9 @@ uint8_t dc_dp_initialize_scrambling_data_symbols( return disable_scrabled_data_symbols; } -static inline bool is_repeater(struct dc_link *link, uint32_t offset) +static inline bool is_repeater(const struct link_training_settings *lt_settings, uint32_t offset) { - return (link->lttpr_mode == LTTPR_MODE_NON_TRANSPARENT) && (offset != 0); + return (lt_settings->lttpr_mode == LTTPR_MODE_NON_TRANSPARENT) && (offset != 0); } static void dpcd_set_lt_pattern_and_lane_settings( @@ -545,7 +545,7 @@ static void dpcd_set_lt_pattern_and_lane_settings( bool edp_workaround = false; /* TODO link_prop.INTERNAL */ dpcd_base_lt_offset = DP_TRAINING_PATTERN_SET; - if (is_repeater(link, offset)) + if (is_repeater(lt_settings, offset)) dpcd_base_lt_offset = DP_TRAINING_PATTERN_SET_PHY_REPEATER1 + ((DP_REPEATER_CONFIGURATION_AND_STATUS_SIZE) * (offset - 1)); @@ -561,7 +561,7 @@ static void dpcd_set_lt_pattern_and_lane_settings( dpcd_lt_buffer[DP_TRAINING_PATTERN_SET - DP_TRAINING_PATTERN_SET] = dpcd_pattern.raw; - if (is_repeater(link, offset)) { + if (is_repeater(lt_settings, offset)) { DC_LOG_HW_LINK_TRAINING("%s\n LTTPR Repeater ID: %d\n 0x%X pattern = %x\n", __func__, offset, @@ -584,7 +584,7 @@ static void dpcd_set_lt_pattern_and_lane_settings( lt_settings->dpcd_lane_settings, size_in_bytes); - if (is_repeater(link, offset)) { + if (is_repeater(lt_settings, offset)) { if (dp_get_link_encoding_format(<_settings->link_settings) == DP_128b_132b_ENCODING) DC_LOG_HW_LINK_TRAINING("%s:\n LTTPR Repeater ID: %d\n" @@ -873,7 +873,7 @@ enum dc_status dp_get_lane_status_and_lane_adjust( uint32_t lane; enum dc_status status; - if (is_repeater(link, offset)) { + if (is_repeater(link_training_setting, offset)) { lane01_status_address = DP_LANE0_1_STATUS_PHY_REPEATER1 + ((DP_REPEATER_CONFIGURATION_AND_STATUS_SIZE) * (offset - 1)); @@ -906,7 +906,7 @@ enum dc_status dp_get_lane_status_and_lane_adjust( ln_align->raw = dpcd_buf[2]; - if (is_repeater(link, offset)) { + if (is_repeater(link_training_setting, offset)) { DC_LOG_HW_LINK_TRAINING("%s:\n LTTPR Repeater ID: %d\n" " 0x%X Lane01Status = %x\n 0x%X Lane23Status = %x\n ", __func__, @@ -944,6 +944,23 @@ enum dc_status dp_get_lane_status_and_lane_adjust( return status; } +static enum dc_status dpcd_128b_132b_set_lane_settings( + struct dc_link *link, + const struct link_training_settings *link_training_setting) +{ + enum dc_status status = core_link_write_dpcd(link, + DP_TRAINING_LANE0_SET, + (uint8_t *)(link_training_setting->dpcd_lane_settings), + sizeof(link_training_setting->dpcd_lane_settings)); + + DC_LOG_HW_LINK_TRAINING("%s:\n 0x%X TX_FFE_PRESET_VALUE = %x\n", + __func__, + DP_TRAINING_LANE0_SET, + link_training_setting->dpcd_lane_settings[0].tx_ffe.PRESET_VALUE); + return status; +} + + enum dc_status dpcd_set_lane_settings( struct dc_link *link, const struct link_training_settings *link_training_setting, @@ -954,7 +971,7 @@ enum dc_status dpcd_set_lane_settings( lane0_set_address = DP_TRAINING_LANE0_SET; - if (is_repeater(link, offset)) + if (is_repeater(link_training_setting, offset)) lane0_set_address = DP_TRAINING_LANE0_SET_PHY_REPEATER1 + ((DP_REPEATER_CONFIGURATION_AND_STATUS_SIZE) * (offset - 1)); @@ -963,17 +980,7 @@ enum dc_status dpcd_set_lane_settings( (uint8_t *)(link_training_setting->dpcd_lane_settings), link_training_setting->link_settings.lane_count); - if (is_repeater(link, offset)) { - if (dp_get_link_encoding_format(&link_training_setting->link_settings) == - DP_128b_132b_ENCODING) - DC_LOG_HW_LINK_TRAINING("%s:\n LTTPR Repeater ID: %d\n" - " 0x%X TX_FFE_PRESET_VALUE = %x\n", - __func__, - offset, - lane0_set_address, - link_training_setting->dpcd_lane_settings[0].tx_ffe.PRESET_VALUE); - else if (dp_get_link_encoding_format(&link_training_setting->link_settings) == - DP_8b_10b_ENCODING) + if (is_repeater(link_training_setting, offset)) { DC_LOG_HW_LINK_TRAINING("%s\n LTTPR Repeater ID: %d\n" " 0x%X VS set = %x PE set = %x max VS Reached = %x max PE Reached = %x\n", __func__, @@ -985,14 +992,6 @@ enum dc_status dpcd_set_lane_settings( link_training_setting->dpcd_lane_settings[0].bits.MAX_PRE_EMPHASIS_REACHED); } else { - if (dp_get_link_encoding_format(&link_training_setting->link_settings) == - DP_128b_132b_ENCODING) - DC_LOG_HW_LINK_TRAINING("%s:\n 0x%X TX_FFE_PRESET_VALUE = %x\n", - __func__, - lane0_set_address, - link_training_setting->dpcd_lane_settings[0].tx_ffe.PRESET_VALUE); - else if (dp_get_link_encoding_format(&link_training_setting->link_settings) == - DP_8b_10b_ENCODING) DC_LOG_HW_LINK_TRAINING("%s\n 0x%X VS set = %x PE set = %x max VS Reached = %x max PE Reached = %x\n", __func__, lane0_set_address, @@ -1172,7 +1171,7 @@ static enum link_training_result perform_channel_equalization_sequence( /* Note: also check that TPS4 is a supported feature*/ tr_pattern = lt_settings->pattern_for_eq; - if (is_repeater(link, offset) && dp_get_link_encoding_format(<_settings->link_settings) == DP_8b_10b_ENCODING) + if (is_repeater(lt_settings, offset) && dp_get_link_encoding_format(<_settings->link_settings) == DP_8b_10b_ENCODING) tr_pattern = DP_TRAINING_PATTERN_SEQUENCE_4; dp_set_hw_training_pattern(link, link_res, tr_pattern, offset); @@ -1198,7 +1197,7 @@ static enum link_training_result perform_channel_equalization_sequence( /* 3. wait for receiver to lock-on*/ wait_time_microsec = lt_settings->eq_pattern_time; - if (is_repeater(link, offset)) + if (is_repeater(lt_settings, offset)) wait_time_microsec = dp_translate_training_aux_read_interval( link->dpcd_caps.lttpr_caps.aux_rd_interval[offset - 1]); @@ -1469,7 +1468,6 @@ static inline void decide_8b_10b_training_settings( */ lt_settings->link_settings.link_spread = link->dp_ss_off ? LINK_SPREAD_DISABLED : LINK_SPREAD_05_DOWNSPREAD_30KHZ; - lt_settings->lttpr_mode = link->lttpr_mode; lt_settings->cr_pattern_time = get_cr_training_aux_rd_interval(link, link_setting); lt_settings->eq_pattern_time = get_eq_training_aux_rd_interval(link, link_setting); lt_settings->pattern_for_cr = decide_cr_training_pattern(link_setting); @@ -1478,6 +1476,7 @@ static inline void decide_8b_10b_training_settings( lt_settings->should_set_fec_ready = true; lt_settings->disallow_per_lane_settings = true; lt_settings->always_match_dpcd_with_hw_lane_settings = true; + lt_settings->lttpr_mode = dp_decide_8b_10b_lttpr_mode(link); dp_hw_to_dpcd_lane_settings(lt_settings, lt_settings->hw_lane_settings, lt_settings->dpcd_lane_settings); } @@ -1501,9 +1500,8 @@ static inline void decide_128b_132b_training_settings(struct dc_link *link, lt_settings->cds_pattern_time = 2500; lt_settings->cds_wait_time_limit = (dp_convert_to_count( link->dpcd_caps.lttpr_caps.phy_repeater_cnt) + 1) * 20000; - lt_settings->lttpr_mode = dp_convert_to_count(link->dpcd_caps.lttpr_caps.phy_repeater_cnt) ? - LTTPR_MODE_NON_TRANSPARENT : LTTPR_MODE_TRANSPARENT; lt_settings->disallow_per_lane_settings = true; + lt_settings->lttpr_mode = dp_decide_128b_132b_lttpr_mode(link); dp_hw_to_dpcd_lane_settings(lt_settings, lt_settings->hw_lane_settings, lt_settings->dpcd_lane_settings); } @@ -1543,7 +1541,7 @@ static void override_training_settings( lt_settings->ffe_preset = overrides->ffe_preset; /* Override HW lane settings with BIOS forced values if present */ if (link->chip_caps & EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN && - link->lttpr_mode == LTTPR_MODE_TRANSPARENT) { + lt_settings->lttpr_mode == LTTPR_MODE_TRANSPARENT) { lt_settings->voltage_swing = &link->bios_forced_drive_settings.VOLTAGE_SWING; lt_settings->pre_emphasis = &link->bios_forced_drive_settings.PRE_EMPHASIS; lt_settings->always_match_dpcd_with_hw_lane_settings = false; @@ -1584,6 +1582,15 @@ static void override_training_settings( if (link->preferred_training_settings.fec_enable != NULL) lt_settings->should_set_fec_ready = *link->preferred_training_settings.fec_enable; + + #if defined(CONFIG_DRM_AMD_DC_DCN) + /* Check DP tunnel LTTPR mode debug option. */ + if (link->ep_type == DISPLAY_ENDPOINT_USB4_DPIA && link->dc->debug.dpia_debug.bits.force_non_lttpr) + lt_settings->lttpr_mode = LTTPR_MODE_NON_LTTPR; + +#endif + dp_get_lttpr_mode_override(link, <_settings->lttpr_mode); + } uint8_t dp_convert_to_count(uint8_t lttpr_repeater_count) @@ -1649,7 +1656,7 @@ static enum dc_status configure_lttpr_mode_non_transparent( link->dpcd_caps.lttpr_caps.mode = repeater_mode; } - if (link->lttpr_mode == LTTPR_MODE_NON_TRANSPARENT) { + if (lt_settings->lttpr_mode == LTTPR_MODE_NON_TRANSPARENT) { DC_LOG_HW_LINK_TRAINING("%s\n Set LTTPR to Non Transparent Mode\n", __func__); @@ -2015,7 +2022,7 @@ static enum link_training_result dp_perform_128b_132b_channel_eq_done_sequence( result = DP_128b_132b_LT_FAILED; } else { dp_set_hw_lane_settings(link, link_res, lt_settings, DPRX); - dpcd_set_lane_settings(link, lt_settings, DPRX); + dpcd_128b_132b_set_lane_settings(link, lt_settings); } loop_count++; } @@ -2099,7 +2106,7 @@ static enum link_training_result dp_perform_8b_10b_link_training( /* 1. set link rate, lane count and spread. */ dpcd_set_link_settings(link, lt_settings); - if (link->lttpr_mode == LTTPR_MODE_NON_TRANSPARENT) { + if (lt_settings->lttpr_mode == LTTPR_MODE_NON_TRANSPARENT) { /* 2. perform link training (set link training done * to false is done as well) @@ -2216,7 +2223,7 @@ static enum link_training_result perform_fixed_vs_pe_nontransparent_training_seq link->vendor_specific_lttpr_link_rate_wa = target_rate; - if (link->lttpr_mode == LTTPR_MODE_NON_TRANSPARENT) { + if (lt_settings->lttpr_mode == LTTPR_MODE_NON_TRANSPARENT) { /* 2. perform link training (set link training done * to false is done as well) @@ -2288,7 +2295,7 @@ static enum link_training_result dp_perform_fixed_vs_pe_training_sequence( ASSERT(dp_get_link_encoding_format(<_settings->link_settings) == DP_8b_10b_ENCODING); - if (link->lttpr_mode == LTTPR_MODE_NON_TRANSPARENT) { + if (lt_settings->lttpr_mode == LTTPR_MODE_NON_TRANSPARENT) { status = perform_fixed_vs_pe_nontransparent_training_sequence(link, link_res, lt_settings); return status; } @@ -2635,6 +2642,7 @@ enum link_training_result dc_link_dp_perform_link_training( link, link_settings, <_settings); + override_training_settings( link, &link->preferred_training_settings, @@ -2652,7 +2660,7 @@ enum link_training_result dc_link_dp_perform_link_training( * Per DP specs starting from here, DPTX device shall not issue * Non-LT AUX transactions inside training mode. */ - if (link->chip_caps & EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN) + if (link->chip_caps & EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN && encoding == DP_8b_10b_ENCODING) status = dp_perform_fixed_vs_pe_training_sequence(link, link_res, <_settings); else if (encoding == DP_8b_10b_ENCODING) status = dp_perform_8b_10b_link_training(link, link_res, <_settings); @@ -3086,7 +3094,7 @@ struct dc_link_settings dp_get_max_link_cap(struct dc_link *link) * account for lttpr repeaters cap * notes: repeaters do not snoop in the DPRX Capabilities addresses (3.6.3). */ - if (link->lttpr_mode != LTTPR_MODE_NON_LTTPR) { + if (dp_is_lttpr_present(link)) { if (link->dpcd_caps.lttpr_caps.max_lane_count < max_link_cap.lane_count) max_link_cap.lane_count = link->dpcd_caps.lttpr_caps.max_lane_count; lttpr_max_link_rate = get_lttpr_max_link_rate(link); @@ -3240,7 +3248,7 @@ static bool dp_verify_link_cap( cur_link_settings = max_link_settings; /* Grant extended timeout request */ - if ((link->lttpr_mode == LTTPR_MODE_NON_TRANSPARENT) && (link->dpcd_caps.lttpr_caps.max_ext_timeout > 0)) { + if (dp_is_lttpr_present(link) && link->dpcd_caps.lttpr_caps.max_ext_timeout > 0) { uint8_t grant = link->dpcd_caps.lttpr_caps.max_ext_timeout & 0x80; core_link_write_dpcd(link, DP_PHY_REPEATER_EXTENDED_WAIT_TIMEOUT, &grant, sizeof(grant)); @@ -4101,8 +4109,13 @@ static void dp_test_send_phy_test_pattern(struct dc_link *link) &dpcd_lane_adjustment[0].raw, sizeof(dpcd_lane_adjustment)); + /* prepare link training settings */ + link_training_settings.link_settings = link->cur_link_settings; + + link_training_settings.lttpr_mode = dp_decide_lttpr_mode(link, &link->cur_link_settings); + if ((link->chip_caps & EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN) && - link->lttpr_mode == LTTPR_MODE_TRANSPARENT) + link_training_settings.lttpr_mode == LTTPR_MODE_TRANSPARENT) dp_fixed_vs_pe_read_lane_adjust( link, link_training_settings.dpcd_lane_settings); @@ -4209,9 +4222,6 @@ static void dp_test_send_phy_test_pattern(struct dc_link *link) test_pattern_size); } - /* prepare link training settings */ - link_training_settings.link_settings = link->cur_link_settings; - for (lane = 0; lane < (unsigned int)(link->cur_link_settings.lane_count); lane++) { @@ -5021,133 +5031,151 @@ static bool dpcd_read_sink_ext_caps(struct dc_link *link) return true; } -/* Logic to determine LTTPR mode */ -static void determine_lttpr_mode(struct dc_link *link) +bool dp_retrieve_lttpr_cap(struct dc_link *link) { - bool allow_lttpr_non_transparent_mode = 0; - bool vbios_lttpr_enable = link->dc->caps.vbios_lttpr_enable; + uint8_t lttpr_dpcd_data[8]; + enum dc_status status = DC_ERROR_UNEXPECTED; + bool is_lttpr_present = false; + + /* Logic to determine LTTPR support*/ bool vbios_lttpr_interop = link->dc->caps.vbios_lttpr_aware; - if (link->ctx->dc->debug.lttpr_mode_override != 0) { - link->lttpr_mode = link->ctx->dc->debug.lttpr_mode_override; - return; - } + if (!vbios_lttpr_interop || !link->dc->caps.extended_aux_timeout_support) + return false; - if ((link->dc->config.allow_lttpr_non_transparent_mode.bits.DP2_0 && - link->dpcd_caps.channel_coding_cap.bits.DP_128b_132b_SUPPORTED)) { - allow_lttpr_non_transparent_mode = 1; - } else if (link->dc->config.allow_lttpr_non_transparent_mode.bits.DP1_4A && - !link->dpcd_caps.channel_coding_cap.bits.DP_128b_132b_SUPPORTED) { - allow_lttpr_non_transparent_mode = 1; + /* By reading LTTPR capability, RX assumes that we will enable + * LTTPR extended aux timeout if LTTPR is present. + */ + status = core_link_read_dpcd(link, + DP_LT_TUNABLE_PHY_REPEATER_FIELD_DATA_STRUCTURE_REV, + lttpr_dpcd_data, + sizeof(lttpr_dpcd_data)); + + link->dpcd_caps.lttpr_caps.revision.raw = + lttpr_dpcd_data[DP_LT_TUNABLE_PHY_REPEATER_FIELD_DATA_STRUCTURE_REV - + DP_LT_TUNABLE_PHY_REPEATER_FIELD_DATA_STRUCTURE_REV]; + + link->dpcd_caps.lttpr_caps.max_link_rate = + lttpr_dpcd_data[DP_MAX_LINK_RATE_PHY_REPEATER - + DP_LT_TUNABLE_PHY_REPEATER_FIELD_DATA_STRUCTURE_REV]; + + link->dpcd_caps.lttpr_caps.phy_repeater_cnt = + lttpr_dpcd_data[DP_PHY_REPEATER_CNT - + DP_LT_TUNABLE_PHY_REPEATER_FIELD_DATA_STRUCTURE_REV]; + + link->dpcd_caps.lttpr_caps.max_lane_count = + lttpr_dpcd_data[DP_MAX_LANE_COUNT_PHY_REPEATER - + DP_LT_TUNABLE_PHY_REPEATER_FIELD_DATA_STRUCTURE_REV]; + + link->dpcd_caps.lttpr_caps.mode = + lttpr_dpcd_data[DP_PHY_REPEATER_MODE - + DP_LT_TUNABLE_PHY_REPEATER_FIELD_DATA_STRUCTURE_REV]; + + link->dpcd_caps.lttpr_caps.max_ext_timeout = + lttpr_dpcd_data[DP_PHY_REPEATER_EXTENDED_WAIT_TIMEOUT - + DP_LT_TUNABLE_PHY_REPEATER_FIELD_DATA_STRUCTURE_REV]; + link->dpcd_caps.lttpr_caps.main_link_channel_coding.raw = + lttpr_dpcd_data[DP_MAIN_LINK_CHANNEL_CODING_PHY_REPEATER - + DP_LT_TUNABLE_PHY_REPEATER_FIELD_DATA_STRUCTURE_REV]; + + link->dpcd_caps.lttpr_caps.supported_128b_132b_rates.raw = + lttpr_dpcd_data[DP_PHY_REPEATER_128B132B_RATES - + DP_LT_TUNABLE_PHY_REPEATER_FIELD_DATA_STRUCTURE_REV]; + + /* If this chip cap is set, at least one retimer must exist in the chain + * Override count to 1 if we receive a known bad count (0 or an invalid value) + */ + if (link->chip_caps & EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN && + (dp_convert_to_count(link->dpcd_caps.lttpr_caps.phy_repeater_cnt) == 0)) { + ASSERT(0); + link->dpcd_caps.lttpr_caps.phy_repeater_cnt = 0x80; + DC_LOG_DC("lttpr_caps forced phy_repeater_cnt = %d\n", link->dpcd_caps.lttpr_caps.phy_repeater_cnt); } - link->lttpr_mode = LTTPR_MODE_NON_LTTPR; - if (vbios_lttpr_enable && vbios_lttpr_interop) - link->lttpr_mode = LTTPR_MODE_NON_TRANSPARENT; - else if (!vbios_lttpr_enable && vbios_lttpr_interop) { - if (allow_lttpr_non_transparent_mode) - link->lttpr_mode = LTTPR_MODE_NON_TRANSPARENT; - else - link->lttpr_mode = LTTPR_MODE_TRANSPARENT; - } else if (!vbios_lttpr_enable && !vbios_lttpr_interop) { - if (!allow_lttpr_non_transparent_mode || !link->dc->caps.extended_aux_timeout_support) - link->lttpr_mode = LTTPR_MODE_NON_LTTPR; - else - link->lttpr_mode = LTTPR_MODE_NON_TRANSPARENT; - } + /* Attempt to train in LTTPR transparent mode if repeater count exceeds 8. */ + is_lttpr_present = dp_is_lttpr_present(link); -#if defined(CONFIG_DRM_AMD_DC_DCN) - /* Check DP tunnel LTTPR mode debug option. */ - if (link->ep_type == DISPLAY_ENDPOINT_USB4_DPIA && - link->dc->debug.dpia_debug.bits.force_non_lttpr) - link->lttpr_mode = LTTPR_MODE_NON_LTTPR; -#endif + if (is_lttpr_present) + CONN_DATA_DETECT(link, lttpr_dpcd_data, sizeof(lttpr_dpcd_data), "LTTPR Caps: "); + + DC_LOG_DC("is_lttpr_present = %d\n", is_lttpr_present); + return is_lttpr_present; } -bool dp_retrieve_lttpr_cap(struct dc_link *link) +bool dp_is_lttpr_present(struct dc_link *link) { - uint8_t lttpr_dpcd_data[8]; - enum dc_status status = DC_ERROR_UNEXPECTED; - bool is_lttpr_present = false; + return (dp_convert_to_count(link->dpcd_caps.lttpr_caps.phy_repeater_cnt) != 0 && + link->dpcd_caps.lttpr_caps.max_lane_count > 0 && + link->dpcd_caps.lttpr_caps.max_lane_count <= 4 && + link->dpcd_caps.lttpr_caps.revision.raw >= 0x14); +} - memset(lttpr_dpcd_data, '\0', sizeof(lttpr_dpcd_data)); +enum lttpr_mode dp_decide_lttpr_mode(struct dc_link *link, struct dc_link_settings *link_setting) +{ + enum dp_link_encoding encoding = dp_get_link_encoding_format(link_setting); - /* Logic to determine LTTPR mode*/ - determine_lttpr_mode(link); + if (encoding == DP_8b_10b_ENCODING) + return dp_decide_8b_10b_lttpr_mode(link); + else if (encoding == DP_128b_132b_ENCODING) + return dp_decide_128b_132b_lttpr_mode(link); - if (link->lttpr_mode == LTTPR_MODE_NON_TRANSPARENT || link->lttpr_mode == LTTPR_MODE_TRANSPARENT) { - if ((link->chip_caps & EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN) && - !link->dc->debug.disable_fixed_vs_aux_timeout_wa) { - /* Fixed VS workaround for AUX timeout */ - const uint32_t fixed_vs_address = 0xF004F; - const uint8_t fixed_vs_data[4] = {0x1, 0x22, 0x63, 0xc}; + ASSERT(0); + return LTTPR_MODE_NON_LTTPR; +} - core_link_write_dpcd( - link, - fixed_vs_address, - fixed_vs_data, - sizeof(fixed_vs_data)); - } +void dp_get_lttpr_mode_override(struct dc_link *link, enum lttpr_mode *override) +{ + if (!dp_is_lttpr_present(link)) + return; - /* By reading LTTPR capability, RX assumes that we will enable - * LTTPR extended aux timeout if LTTPR is present. - */ - status = core_link_read_dpcd( - link, - DP_LT_TUNABLE_PHY_REPEATER_FIELD_DATA_STRUCTURE_REV, - lttpr_dpcd_data, - sizeof(lttpr_dpcd_data)); - - link->dpcd_caps.lttpr_caps.revision.raw = - lttpr_dpcd_data[DP_LT_TUNABLE_PHY_REPEATER_FIELD_DATA_STRUCTURE_REV - - DP_LT_TUNABLE_PHY_REPEATER_FIELD_DATA_STRUCTURE_REV]; - - link->dpcd_caps.lttpr_caps.max_link_rate = - lttpr_dpcd_data[DP_MAX_LINK_RATE_PHY_REPEATER - - DP_LT_TUNABLE_PHY_REPEATER_FIELD_DATA_STRUCTURE_REV]; - - link->dpcd_caps.lttpr_caps.phy_repeater_cnt = - lttpr_dpcd_data[DP_PHY_REPEATER_CNT - - DP_LT_TUNABLE_PHY_REPEATER_FIELD_DATA_STRUCTURE_REV]; - - link->dpcd_caps.lttpr_caps.max_lane_count = - lttpr_dpcd_data[DP_MAX_LANE_COUNT_PHY_REPEATER - - DP_LT_TUNABLE_PHY_REPEATER_FIELD_DATA_STRUCTURE_REV]; - - link->dpcd_caps.lttpr_caps.mode = - lttpr_dpcd_data[DP_PHY_REPEATER_MODE - - DP_LT_TUNABLE_PHY_REPEATER_FIELD_DATA_STRUCTURE_REV]; - - link->dpcd_caps.lttpr_caps.max_ext_timeout = - lttpr_dpcd_data[DP_PHY_REPEATER_EXTENDED_WAIT_TIMEOUT - - DP_LT_TUNABLE_PHY_REPEATER_FIELD_DATA_STRUCTURE_REV]; - link->dpcd_caps.lttpr_caps.main_link_channel_coding.raw = - lttpr_dpcd_data[DP_MAIN_LINK_CHANNEL_CODING_PHY_REPEATER - - DP_LT_TUNABLE_PHY_REPEATER_FIELD_DATA_STRUCTURE_REV]; - - link->dpcd_caps.lttpr_caps.supported_128b_132b_rates.raw = - lttpr_dpcd_data[DP_PHY_REPEATER_128B132B_RATES - - DP_LT_TUNABLE_PHY_REPEATER_FIELD_DATA_STRUCTURE_REV]; - - /* If this chip cap is set, at least one retimer must exist in the chain - * Override count to 1 if we receive a known bad count (0 or an invalid value) */ - if (link->chip_caps & EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN && - (dp_convert_to_count(link->dpcd_caps.lttpr_caps.phy_repeater_cnt) == 0)) { - ASSERT(0); - link->dpcd_caps.lttpr_caps.phy_repeater_cnt = 0x80; + if (link->dc->debug.lttpr_mode_override == LTTPR_MODE_TRANSPARENT) { + *override = LTTPR_MODE_TRANSPARENT; + } else if (link->dc->debug.lttpr_mode_override == LTTPR_MODE_NON_TRANSPARENT) { + *override = LTTPR_MODE_NON_TRANSPARENT; + } else if (link->dc->debug.lttpr_mode_override == LTTPR_MODE_NON_LTTPR) { + *override = LTTPR_MODE_NON_LTTPR; + } + DC_LOG_DC("lttpr_mode_override chose LTTPR_MODE = %d\n", (uint8_t)(*override)); +} + +enum lttpr_mode dp_decide_8b_10b_lttpr_mode(struct dc_link *link) +{ + bool is_lttpr_present = dp_is_lttpr_present(link); + bool vbios_lttpr_force_non_transparent = link->dc->caps.vbios_lttpr_enable; + bool vbios_lttpr_aware = link->dc->caps.vbios_lttpr_aware; + + if (!is_lttpr_present) + return LTTPR_MODE_NON_LTTPR; + + if (vbios_lttpr_aware) { + if (vbios_lttpr_force_non_transparent) { + DC_LOG_DC("chose LTTPR_MODE_NON_TRANSPARENT due to VBIOS DCE_INFO_CAPS_LTTPR_SUPPORT_ENABLE set to 1.\n"); + return LTTPR_MODE_NON_TRANSPARENT; + } else { + DC_LOG_DC("chose LTTPR_MODE_NON_TRANSPARENT by default due to VBIOS not set DCE_INFO_CAPS_LTTPR_SUPPORT_ENABLE set to 1.\n"); + return LTTPR_MODE_TRANSPARENT; } + } - /* Attempt to train in LTTPR transparent mode if repeater count exceeds 8. */ - is_lttpr_present = (link->dpcd_caps.lttpr_caps.max_lane_count > 0 && - link->dpcd_caps.lttpr_caps.max_lane_count <= 4 && - link->dpcd_caps.lttpr_caps.revision.raw >= 0x14); - if (is_lttpr_present) { - CONN_DATA_DETECT(link, lttpr_dpcd_data, sizeof(lttpr_dpcd_data), "LTTPR Caps: "); - configure_lttpr_mode_transparent(link); - } else - link->lttpr_mode = LTTPR_MODE_NON_LTTPR; + if (link->dc->config.allow_lttpr_non_transparent_mode.bits.DP1_4A && + link->dc->caps.extended_aux_timeout_support) { + DC_LOG_DC("chose LTTPR_MODE_NON_TRANSPARENT by default and dc->config.allow_lttpr_non_transparent_mode.bits.DP1_4A set to 1.\n"); + return LTTPR_MODE_NON_TRANSPARENT; } - return is_lttpr_present; + + DC_LOG_DC("chose LTTPR_MODE_NON_LTTPR.\n"); + return LTTPR_MODE_NON_LTTPR; +} + +enum lttpr_mode dp_decide_128b_132b_lttpr_mode(struct dc_link *link) +{ + enum lttpr_mode mode = LTTPR_MODE_NON_LTTPR; + + if (dp_is_lttpr_present(link)) + mode = LTTPR_MODE_NON_TRANSPARENT; + + DC_LOG_DC("128b_132b chose LTTPR_MODE %d.\n", mode); + return mode; } static bool get_usbc_cable_id(struct dc_link *link, union dp_cable_id *cable_id) @@ -5165,9 +5193,10 @@ static bool get_usbc_cable_id(struct dc_link *link, union dp_cable_id *cable_id) cmd.cable_id.data.input.phy_inst = resource_transmitter_to_phy_idx( link->dc, link->link_enc->transmitter); if (dc_dmub_srv_cmd_with_reply_data(link->ctx->dmub_srv, &cmd) && - cmd.cable_id.header.ret_status == 1) + cmd.cable_id.header.ret_status == 1) { cable_id->raw = cmd.cable_id.data.output_raw; - + DC_LOG_DC("usbc_cable_id = %d.\n", cable_id->raw); + } return cmd.cable_id.header.ret_status == 1; } @@ -5209,13 +5238,17 @@ static enum dc_status wa_try_to_wake_dprx(struct dc_link *link, uint64_t timeout uint64_t current_ts = 0; uint64_t time_taken_ms = 0; enum dc_connection_type type = dc_connection_none; + bool lttpr_present; + bool vbios_lttpr_interop = link->dc->caps.vbios_lttpr_aware; - determine_lttpr_mode(link); + lttpr_present = dp_is_lttpr_present(link) || + (!vbios_lttpr_interop || !link->dc->caps.extended_aux_timeout_support); + DC_LOG_DC("lttpr_present = %d.\n", lttpr_present ? 1 : 0); /* Issue an AUX read to test DPRX responsiveness. If LTTPR is supported the first read is expected to * be to determine LTTPR capabilities. Otherwise trying to read power state should be an innocuous AUX read. */ - if (link->lttpr_mode == LTTPR_MODE_NON_TRANSPARENT || link->lttpr_mode == LTTPR_MODE_TRANSPARENT) + if (lttpr_present) status = core_link_read_dpcd( link, DP_LT_TUNABLE_PHY_REPEATER_FIELD_DATA_STRUCTURE_REV, @@ -5345,6 +5378,10 @@ static bool retrieve_link_cap(struct dc_link *link) } is_lttpr_present = dp_retrieve_lttpr_cap(link); + + if (is_lttpr_present) + configure_lttpr_mode_transparent(link); + /* Read DP tunneling information. */ status = dpcd_get_tunneling_device_data(link); @@ -5774,7 +5811,7 @@ void detect_edp_sink_caps(struct dc_link *link) * Per VESA eDP spec, "The DPCD revision for eDP v1.4 is 13h" */ if (link->dpcd_caps.dpcd_rev.raw >= DPCD_REV_13 && - (link->dc->debug.optimize_edp_link_rate || + (link->panel_config.ilr.optimize_edp_link_rate || link->reported_link_cap.link_rate == LINK_RATE_UNKNOWN)) { // Read DPCD 00010h - 0001Fh 16 bytes at one shot core_link_read_dpcd(link, DP_SUPPORTED_LINK_RATES, @@ -6096,7 +6133,7 @@ bool dc_link_dp_set_test_pattern( /* Set DPCD Lane Settings before running test pattern */ if (p_link_settings != NULL) { if ((link->chip_caps & EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN) && - link->lttpr_mode == LTTPR_MODE_TRANSPARENT) { + p_link_settings->lttpr_mode == LTTPR_MODE_TRANSPARENT) { dp_fixed_vs_pe_set_retimer_lane_settings( link, p_link_settings->dpcd_lane_settings, @@ -6723,7 +6760,7 @@ bool is_edp_ilr_optimization_required(struct dc_link *link, struct dc_crtc_timin ASSERT(link || crtc_timing); // invalid input if (link->dpcd_caps.edp_supported_link_rates_count == 0 || - !link->dc->debug.optimize_edp_link_rate) + !link->panel_config.ilr.optimize_edp_link_rate) return false; @@ -7218,7 +7255,7 @@ void dp_set_hw_lane_settings( { const struct link_hwss *link_hwss = get_link_hwss(link, link_res); - if ((link->lttpr_mode == LTTPR_MODE_NON_TRANSPARENT) && !is_immediate_downstream(link, offset)) + if ((link_settings->lttpr_mode == LTTPR_MODE_NON_TRANSPARENT) && !is_immediate_downstream(link, offset)) return; if (link_hwss->ext.set_dp_lane_settings) diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_dpia.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_dpia.c index 468e39589ed8..74e36b34d3f7 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_link_dpia.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_dpia.c @@ -115,12 +115,14 @@ static enum link_training_result dpia_configure_link( DC_LOG_HW_LINK_TRAINING("%s\n DPIA(%d) configuring\n - LTTPR mode(%d)\n", __func__, link->link_id.enum_id - ENUM_ID_1, - link->lttpr_mode); + lt_settings->lttpr_mode); dp_decide_training_settings(link, link_setting, lt_settings); + dp_get_lttpr_mode_override(link, <_settings->lttpr_mode); + status = dpcd_configure_channel_coding(link, lt_settings); if (status != DC_OK && link->is_hpd_pending) return LINK_TRAINING_ABORT; @@ -178,7 +180,7 @@ static uint8_t dpia_build_set_config_data(enum dpia_set_config_type type, switch (type) { case DPIA_SET_CFG_SET_LINK: - data.set_link.mode = link->lttpr_mode == LTTPR_MODE_NON_TRANSPARENT ? 1 : 0; + data.set_link.mode = lt_settings->lttpr_mode == LTTPR_MODE_NON_TRANSPARENT ? 1 : 0; break; case DPIA_SET_CFG_SET_PHY_TEST_MODE: break; @@ -553,7 +555,7 @@ static enum link_training_result dpia_training_cr_phase( { enum link_training_result result = LINK_TRAINING_CR_FAIL_LANE0; - if (link->lttpr_mode == LTTPR_MODE_NON_TRANSPARENT) + if (lt_settings->lttpr_mode == LTTPR_MODE_NON_TRANSPARENT) result = dpia_training_cr_non_transparent(link, link_res, lt_settings, hop); else result = dpia_training_cr_transparent(link, link_res, lt_settings); @@ -830,7 +832,7 @@ static enum link_training_result dpia_training_eq_phase( { enum link_training_result result; - if (link->lttpr_mode == LTTPR_MODE_NON_TRANSPARENT) + if (lt_settings->lttpr_mode == LTTPR_MODE_NON_TRANSPARENT) result = dpia_training_eq_non_transparent(link, link_res, lt_settings, hop); else result = dpia_training_eq_transparent(link, link_res, lt_settings); @@ -870,13 +872,14 @@ static enum dc_status dpcd_clear_lt_pattern(struct dc_link *link, uint32_t hop) * @param hop The Hop in display path. DPRX = 0. */ static enum link_training_result dpia_training_end(struct dc_link *link, + struct link_training_settings *lt_settings, uint32_t hop) { enum link_training_result result = LINK_TRAINING_SUCCESS; uint8_t repeater_cnt = 0; /* Number of hops/repeaters in display path. */ enum dc_status status; - if (link->lttpr_mode == LTTPR_MODE_NON_TRANSPARENT) { + if (lt_settings->lttpr_mode == LTTPR_MODE_NON_TRANSPARENT) { repeater_cnt = dp_convert_to_count(link->dpcd_caps.lttpr_caps.phy_repeater_cnt); if (hop == repeater_cnt) { /* DPTX-to-DPIA */ @@ -916,7 +919,7 @@ static enum link_training_result dpia_training_end(struct dc_link *link, link->link_id.enum_id - ENUM_ID_1, hop, result, - link->lttpr_mode); + lt_settings->lttpr_mode); return result; } @@ -928,7 +931,9 @@ static enum link_training_result dpia_training_end(struct dc_link *link, * @param link DPIA link being trained. * @param hop The Hop in display path. DPRX = 0. */ -static void dpia_training_abort(struct dc_link *link, uint32_t hop) +static void dpia_training_abort(struct dc_link *link, + struct link_training_settings *lt_settings, + uint32_t hop) { uint8_t data = 0; uint32_t dpcd_tps_offset = DP_TRAINING_PATTERN_SET; @@ -936,7 +941,7 @@ static void dpia_training_abort(struct dc_link *link, uint32_t hop) DC_LOG_HW_LINK_TRAINING("%s\n DPIA(%d) aborting\n - LTTPR mode(%d)\n - HPD(%d)\n", __func__, link->link_id.enum_id - ENUM_ID_1, - link->lttpr_mode, + lt_settings->lttpr_mode, link->is_hpd_pending); /* Abandon clean-up if sink unplugged. */ @@ -964,12 +969,16 @@ enum link_training_result dc_link_dpia_perform_link_training( uint8_t repeater_cnt = 0; /* Number of hops/repeaters in display path. */ int8_t repeater_id; /* Current hop. */ + struct dc_link_settings link_settings = *link_setting; // non-const copy to pass in + + lt_settings.lttpr_mode = dp_decide_lttpr_mode(link, &link_settings); + /* Configure link as prescribed in link_setting and set LTTPR mode. */ result = dpia_configure_link(link, link_res, link_setting, <_settings); if (result != LINK_TRAINING_SUCCESS) return result; - if (link->lttpr_mode == LTTPR_MODE_NON_TRANSPARENT) + if (lt_settings.lttpr_mode == LTTPR_MODE_NON_TRANSPARENT) repeater_cnt = dp_convert_to_count(link->dpcd_caps.lttpr_caps.phy_repeater_cnt); /* Train each hop in turn starting with the one closest to DPTX. @@ -987,7 +996,7 @@ enum link_training_result dc_link_dpia_perform_link_training( break; /* Stop training hop. */ - result = dpia_training_end(link, repeater_id); + result = dpia_training_end(link, <_settings, repeater_id); if (result != LINK_TRAINING_SUCCESS) break; } @@ -1001,9 +1010,9 @@ enum link_training_result dc_link_dpia_perform_link_training( msleep(5); result = dp_check_link_loss_status(link, <_settings); } else if (result == LINK_TRAINING_ABORT) { - dpia_training_abort(link, repeater_id); + dpia_training_abort(link, <_settings, repeater_id); } else { - dpia_training_end(link, repeater_id); + dpia_training_end(link, <_settings, repeater_id); } return result; } diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c index 8ee0d946bb2f..fd8db482e56f 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c @@ -1747,7 +1747,6 @@ bool dc_remove_plane_from_context( for (i = 0; i < stream_status->plane_count; i++) { if (stream_status->plane_states[i] == plane_state) { - dc_plane_state_release(stream_status->plane_states[i]); break; } @@ -3683,4 +3682,56 @@ bool is_h_timing_divisible_by_2(struct dc_stream_state *stream) (stream->timing.h_sync_width % 2 == 0); } return divisible; +} + +bool dc_resource_acquire_secondary_pipe_for_mpc_odm( + const struct dc *dc, + struct dc_state *state, + struct pipe_ctx *pri_pipe, + struct pipe_ctx *sec_pipe, + bool odm) +{ + int pipe_idx = sec_pipe->pipe_idx; + struct pipe_ctx *sec_top, *sec_bottom, *sec_next, *sec_prev; + const struct resource_pool *pool = dc->res_pool; + + sec_top = sec_pipe->top_pipe; + sec_bottom = sec_pipe->bottom_pipe; + sec_next = sec_pipe->next_odm_pipe; + sec_prev = sec_pipe->prev_odm_pipe; + + *sec_pipe = *pri_pipe; + + sec_pipe->top_pipe = sec_top; + sec_pipe->bottom_pipe = sec_bottom; + sec_pipe->next_odm_pipe = sec_next; + sec_pipe->prev_odm_pipe = sec_prev; + + sec_pipe->pipe_idx = pipe_idx; + sec_pipe->plane_res.mi = pool->mis[pipe_idx]; + sec_pipe->plane_res.hubp = pool->hubps[pipe_idx]; + sec_pipe->plane_res.ipp = pool->ipps[pipe_idx]; + sec_pipe->plane_res.xfm = pool->transforms[pipe_idx]; + sec_pipe->plane_res.dpp = pool->dpps[pipe_idx]; + sec_pipe->plane_res.mpcc_inst = pool->dpps[pipe_idx]->inst; + sec_pipe->stream_res.dsc = NULL; + if (odm) { + if (!sec_pipe->top_pipe) + sec_pipe->stream_res.opp = pool->opps[pipe_idx]; + else + sec_pipe->stream_res.opp = sec_pipe->top_pipe->stream_res.opp; + if (sec_pipe->stream->timing.flags.DSC == 1) { +#if defined(CONFIG_DRM_AMD_DC_DCN) + dcn20_acquire_dsc(dc, &state->res_ctx, &sec_pipe->stream_res.dsc, pipe_idx); +#endif + ASSERT(sec_pipe->stream_res.dsc); + if (sec_pipe->stream_res.dsc == NULL) + return false; + } +#if defined(CONFIG_DRM_AMD_DC_DCN) + dcn20_build_mapped_resource(dc, state, sec_pipe->stream); +#endif + } + + return true; }
\ No newline at end of file diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_stream.c b/drivers/gpu/drm/amd/display/dc/core/dc_stream.c index 0c85ab5933b4..38d71b5c1f2d 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_stream.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_stream.c @@ -30,6 +30,7 @@ #include "resource.h" #include "ipp.h" #include "timing_generator.h" +#include "dc_dmub_srv.h" #define DC_LOGGER dc->ctx->logger @@ -275,6 +276,8 @@ static void program_cursor_attributes( } dc->hwss.set_cursor_attribute(pipe_ctx); + + dc_send_update_cursor_info_to_dmu(pipe_ctx, i); if (dc->hwss.set_cursor_sdr_white_level) dc->hwss.set_cursor_sdr_white_level(pipe_ctx); } @@ -381,6 +384,8 @@ static void program_cursor_position( } dc->hwss.set_cursor_position(pipe_ctx); + + dc_send_update_cursor_info_to_dmu(pipe_ctx, i); } if (pipe_to_program) @@ -521,7 +526,7 @@ bool dc_stream_remove_writeback(struct dc *dc, /* remove writeback info for disabled writeback pipes from stream */ for (i = 0, j = 0; i < stream->num_wb_info; i++) { if (stream->writeback_info[i].wb_enabled) { - if (i != j) + if (j < i) /* trim the array */ stream->writeback_info[j] = stream->writeback_info[i]; j++; diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h index 659c71778a64..bfc5474c0f4c 100644 --- a/drivers/gpu/drm/amd/display/dc/dc.h +++ b/drivers/gpu/drm/amd/display/dc/dc.h @@ -47,7 +47,7 @@ struct aux_payload; struct set_config_cmd_payload; struct dmub_notification; -#define DC_VER "3.2.204" +#define DC_VER "3.2.207" #define MAX_SURFACES 3 #define MAX_PLANES 6 @@ -406,6 +406,9 @@ struct dc_config { bool ignore_dpref_ss; bool enable_mipi_converter_optimization; bool use_default_clock_table; + bool force_bios_enable_lttpr; + uint8_t force_bios_fixed_vs; + }; enum visual_confirm { @@ -818,7 +821,6 @@ struct dc_debug_options { /* Enable dmub aux for legacy ddc */ bool enable_dmub_aux_for_legacy_ddc; bool disable_fams; - bool optimize_edp_link_rate; /* eDP ILR */ /* FEC/PSR1 sequence enable delay in 100us */ uint8_t fec_enable_delay_in100us; bool enable_driver_sequence_debug; @@ -832,6 +834,7 @@ struct dc_debug_options { bool force_subvp_mclk_switch; bool allow_sw_cursor_fallback; unsigned int force_subvp_num_ways; + unsigned int force_mall_ss_num_ways; bool alloc_extra_way_for_cursor; bool force_usr_allow; /* uses value at boot and disables switch */ @@ -846,6 +849,7 @@ struct dc_debug_options { bool use_legacy_soc_bb_mechanism; bool exit_idle_opt_for_cursor_updates; bool enable_single_display_2to1_odm_policy; + bool enable_double_buffered_dsc_pg_support; bool enable_dp_dig_pixel_rate_div_policy; enum lttpr_mode lttpr_mode_override; }; @@ -1117,6 +1121,7 @@ union surface_update_flags { uint32_t clock_change:1; uint32_t stereo_format_change:1; uint32_t lut_3d:1; + uint32_t tmz_changed:1; uint32_t full_update:1; } bits; @@ -1186,6 +1191,8 @@ struct dc_plane_state { enum dc_irq_source irq_source; struct kref refcount; struct tg_color visual_confirm_color; + + bool is_statically_allocated; }; struct dc_plane_info { @@ -1605,6 +1612,9 @@ enum dc_status dc_process_dmub_set_mst_slots(const struct dc *dc, uint8_t mst_alloc_slots, uint8_t *mst_slots_in_use); +void dc_process_dmub_dpia_hpd_int_enable(const struct dc *dc, + uint32_t hpd_int_enable); + /******************************************************************************* * DSC Interfaces ******************************************************************************/ diff --git a/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c b/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c index 89d7d3fd3321..0541e87e4f38 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c +++ b/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c @@ -30,6 +30,7 @@ #include "dc_hw_types.h" #include "core_types.h" #include "../basics/conversion.h" +#include "cursor_reg_cache.h" #define CTX dc_dmub_srv->ctx #define DC_LOGGER CTX->logger @@ -780,7 +781,7 @@ void dc_dmub_setup_subvp_dmub_command(struct dc *dc, // Store the original watermark value for this SubVP config so we can lower it when the // MCLK switch starts wm_val_refclk = context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.pstate_change_ns * - dc->res_pool->ref_clocks.dchub_ref_clock_inKhz / 1000 / 1000; + (dc->res_pool->ref_clocks.dchub_ref_clock_inKhz / 1000) / 1000; cmd.fw_assisted_mclk_switch_v2.config_data.watermark_a_cache = wm_val_refclk < 0xFFFF ? wm_val_refclk : 0xFFFF; } @@ -880,3 +881,147 @@ void dc_dmub_srv_log_diagnostic_data(struct dc_dmub_srv *dc_dmub_srv) diag_data.is_cw0_enabled, diag_data.is_cw6_enabled); } + +static bool dc_dmub_should_update_cursor_data(struct pipe_ctx *pipe_ctx) +{ + if (pipe_ctx->plane_state != NULL) { + if (pipe_ctx->plane_state->address.type == PLN_ADDR_TYPE_VIDEO_PROGRESSIVE) + return false; + } + + if ((pipe_ctx->stream->link->psr_settings.psr_version == DC_PSR_VERSION_SU_1 || + pipe_ctx->stream->link->psr_settings.psr_version == DC_PSR_VERSION_1) && + pipe_ctx->stream->ctx->dce_version >= DCN_VERSION_3_1) + return true; + + return false; +} + +static void dc_build_cursor_update_payload0( + struct pipe_ctx *pipe_ctx, uint8_t p_idx, + struct dmub_cmd_update_cursor_payload0 *payload) +{ + struct hubp *hubp = pipe_ctx->plane_res.hubp; + unsigned int panel_inst = 0; + + if (!dc_get_edp_link_panel_inst(hubp->ctx->dc, + pipe_ctx->stream->link, &panel_inst)) + return; + + /* Payload: Cursor Rect is built from position & attribute + * x & y are obtained from postion + */ + payload->cursor_rect.x = hubp->cur_rect.x; + payload->cursor_rect.y = hubp->cur_rect.y; + /* w & h are obtained from attribute */ + payload->cursor_rect.width = hubp->cur_rect.w; + payload->cursor_rect.height = hubp->cur_rect.h; + + payload->enable = hubp->pos.cur_ctl.bits.cur_enable; + payload->pipe_idx = p_idx; + payload->cmd_version = DMUB_CMD_PSR_CONTROL_VERSION_1; + payload->panel_inst = panel_inst; +} + +static void dc_send_cmd_to_dmu(struct dc_dmub_srv *dmub_srv, + union dmub_rb_cmd *cmd) +{ + dc_dmub_srv_cmd_queue(dmub_srv, cmd); + dc_dmub_srv_cmd_execute(dmub_srv); + dc_dmub_srv_wait_idle(dmub_srv); +} + +static void dc_build_cursor_position_update_payload0( + struct dmub_cmd_update_cursor_payload0 *pl, const uint8_t p_idx, + const struct hubp *hubp, const struct dpp *dpp) +{ + /* Hubp */ + pl->position_cfg.pHubp.cur_ctl.raw = hubp->pos.cur_ctl.raw; + pl->position_cfg.pHubp.position.raw = hubp->pos.position.raw; + pl->position_cfg.pHubp.hot_spot.raw = hubp->pos.hot_spot.raw; + pl->position_cfg.pHubp.dst_offset.raw = hubp->pos.dst_offset.raw; + + /* dpp */ + pl->position_cfg.pDpp.cur0_ctl.raw = dpp->pos.cur0_ctl.raw; + pl->position_cfg.pipe_idx = p_idx; +} + +static void dc_build_cursor_attribute_update_payload1( + struct dmub_cursor_attributes_cfg *pl_A, const uint8_t p_idx, + const struct hubp *hubp, const struct dpp *dpp) +{ + /* Hubp */ + pl_A->aHubp.SURFACE_ADDR_HIGH = hubp->att.SURFACE_ADDR_HIGH; + pl_A->aHubp.SURFACE_ADDR = hubp->att.SURFACE_ADDR; + pl_A->aHubp.cur_ctl.raw = hubp->att.cur_ctl.raw; + pl_A->aHubp.size.raw = hubp->att.size.raw; + pl_A->aHubp.settings.raw = hubp->att.settings.raw; + + /* dpp */ + pl_A->aDpp.cur0_ctl.raw = dpp->att.cur0_ctl.raw; +} + +/** + * *************************************************************************************** + * dc_send_update_cursor_info_to_dmu: Populate the DMCUB Cursor update info command + * + * This function would store the cursor related information and pass it into dmub + * + * @param [in] pCtx: pipe context + * @param [in] pipe_idx: pipe index + * + * @return: void + * + * *************************************************************************************** + */ + +void dc_send_update_cursor_info_to_dmu( + struct pipe_ctx *pCtx, uint8_t pipe_idx) +{ + union dmub_rb_cmd cmd = { 0 }; + union dmub_cmd_update_cursor_info_data *update_cursor_info = + &cmd.update_cursor_info.update_cursor_info_data; + + if (!dc_dmub_should_update_cursor_data(pCtx)) + return; + /* + * Since we use multi_cmd_pending for dmub command, the 2nd command is + * only assigned to store cursor attributes info. + * 1st command can view as 2 parts, 1st is for PSR/Replay data, the other + * is to store cursor position info. + * + * Command heaer type must be the same type if using multi_cmd_pending. + * Besides, while process 2nd command in DMU, the sub type is useless. + * So it's meanless to pass the sub type header with different type. + */ + + { + /* Build Payload#0 Header */ + cmd.update_cursor_info.header.type = DMUB_CMD__UPDATE_CURSOR_INFO; + cmd.update_cursor_info.header.payload_bytes = + sizeof(cmd.update_cursor_info.update_cursor_info_data); + cmd.update_cursor_info.header.multi_cmd_pending = 1; /* To combine multi dmu cmd, 1st cmd */ + + /* Prepare Payload */ + dc_build_cursor_update_payload0(pCtx, pipe_idx, &update_cursor_info->payload0); + + dc_build_cursor_position_update_payload0(&update_cursor_info->payload0, pipe_idx, + pCtx->plane_res.hubp, pCtx->plane_res.dpp); + /* Send update_curosr_info to queue */ + dc_dmub_srv_cmd_queue(pCtx->stream->ctx->dmub_srv, &cmd); + } + { + /* Build Payload#1 Header */ + memset(update_cursor_info, 0, sizeof(union dmub_cmd_update_cursor_info_data)); + cmd.update_cursor_info.header.type = DMUB_CMD__UPDATE_CURSOR_INFO; + cmd.update_cursor_info.header.payload_bytes = sizeof(struct cursor_attributes_cfg); + cmd.update_cursor_info.header.multi_cmd_pending = 0; /* Indicate it's the last command. */ + + dc_build_cursor_attribute_update_payload1( + &cmd.update_cursor_info.update_cursor_info_data.payload1.attribute_cfg, + pipe_idx, pCtx->plane_res.hubp, pCtx->plane_res.dpp); + + /* Combine 2nd cmds update_curosr_info to DMU */ + dc_send_cmd_to_dmu(pCtx->stream->ctx->dmub_srv, &cmd); + } +} diff --git a/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.h b/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.h index 7e438345b1a8..d34f5563df2e 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.h +++ b/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.h @@ -88,4 +88,5 @@ bool dc_dmub_srv_get_diagnostic_data(struct dc_dmub_srv *dc_dmub_srv, struct dmu void dc_dmub_setup_subvp_dmub_command(struct dc *dc, struct dc_state *context, bool enable); void dc_dmub_srv_log_diagnostic_data(struct dc_dmub_srv *dc_dmub_srv); +void dc_send_update_cursor_info_to_dmu(struct pipe_ctx *pCtx, uint8_t pipe_idx); #endif /* _DMUB_DC_SRV_H_ */ diff --git a/drivers/gpu/drm/amd/display/dc/dc_link.h b/drivers/gpu/drm/amd/display/dc/dc_link.h index d890a03910c7..caf0c7af2d0b 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_link.h +++ b/drivers/gpu/drm/amd/display/dc/dc_link.h @@ -127,11 +127,21 @@ struct dc_panel_config { unsigned int extra_t12_ms; unsigned int extra_post_OUI_ms; } pps; + // ABM + struct varib { + unsigned int varibright_feature_enable; + unsigned int def_varibright_level; + unsigned int abm_config_setting; + } varib; // edp DSC struct dsc { bool disable_dsc_edp; unsigned int force_dsc_edp_policy; } dsc; + /* eDP ILR */ + struct ilr { + bool optimize_edp_link_rate; /* eDP ILR */ + } ilr; }; /* * A link contains one or more sinks and their connected status. @@ -151,7 +161,6 @@ struct dc_link { bool link_state_valid; bool aux_access_disabled; bool sync_lt_in_progress; - enum lttpr_mode lttpr_mode; bool is_internal_display; /* TODO: Rename. Flag an endpoint as having a programmable mapping to a diff --git a/drivers/gpu/drm/amd/display/dc/dc_stream.h b/drivers/gpu/drm/amd/display/dc/dc_stream.h index 9fcf9dc5bce4..9e6025c98db9 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_stream.h +++ b/drivers/gpu/drm/amd/display/dc/dc_stream.h @@ -212,8 +212,7 @@ struct dc_stream_state { /* DMCU info */ unsigned int abm_level; - struct periodic_interrupt_config periodic_interrupt0; - struct periodic_interrupt_config periodic_interrupt1; + struct periodic_interrupt_config periodic_interrupt; /* from core_stream struct */ struct dc_context *ctx; @@ -281,8 +280,7 @@ struct dc_stream_update { struct dc_info_packet *hdr_static_metadata; unsigned int *abm_level; - struct periodic_interrupt_config *periodic_interrupt0; - struct periodic_interrupt_config *periodic_interrupt1; + struct periodic_interrupt_config *periodic_interrupt; struct dc_info_packet *vrr_infopacket; struct dc_info_packet *vsc_infopacket; diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_aux.c b/drivers/gpu/drm/amd/display/dc/dce/dce_aux.c index 32782ef9ef77..140297c8ff55 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_aux.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_aux.c @@ -942,10 +942,6 @@ bool dce_aux_transfer_with_retries(struct ddc_service *ddc, case AUX_RET_ERROR_ENGINE_ACQUIRE: case AUX_RET_ERROR_UNKNOWN: default: - DC_TRACE_LEVEL_MESSAGE(DAL_TRACE_LEVEL_INFORMATION, - LOG_FLAG_I2cAux_DceAux, - "dce_aux_transfer_with_retries: Failure: operation_result=%d", - (int)operation_result); goto fail; } } @@ -953,14 +949,11 @@ bool dce_aux_transfer_with_retries(struct ddc_service *ddc, fail: DC_TRACE_LEVEL_MESSAGE(DAL_TRACE_LEVEL_ERROR, LOG_FLAG_Error_I2cAux, - "dce_aux_transfer_with_retries: FAILURE"); + "%s: Failure: operation_result=%d", + __func__, + (int)operation_result); if (!payload_reply) payload->reply = NULL; - DC_TRACE_LEVEL_MESSAGE(DAL_TRACE_LEVEL_ERROR, - WPP_BIT_FLAG_DC_ERROR, - "AUX transaction failed. Result: %d", - operation_result); - return false; } diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c index 2341982ee0a5..d260eaa1509e 100644 --- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c +++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c @@ -2178,8 +2178,7 @@ static void dce110_setup_audio_dto( continue; if (pipe_ctx->stream->signal != SIGNAL_TYPE_HDMI_TYPE_A) continue; - if (pipe_ctx->stream_res.audio != NULL && - pipe_ctx->stream_res.audio->enabled == false) { + if (pipe_ctx->stream_res.audio != NULL) { struct audio_output audio_output; build_audio_output(context, pipe_ctx, &audio_output); @@ -2219,8 +2218,7 @@ static void dce110_setup_audio_dto( if (!dc_is_dp_signal(pipe_ctx->stream->signal)) continue; - if (pipe_ctx->stream_res.audio != NULL && - pipe_ctx->stream_res.audio->enabled == false) { + if (pipe_ctx->stream_res.audio != NULL) { struct audio_output audio_output; build_audio_output(context, pipe_ctx, &audio_output); diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c index 897f412f539e..b9765b3899e1 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c @@ -469,6 +469,7 @@ void dpp1_set_cursor_position( REG_UPDATE(CURSOR0_CONTROL, CUR0_ENABLE, cur_en); + dpp_base->pos.cur0_ctl.bits.cur0_enable = cur_en; } void dpp1_cnv_set_optional_cursor_attributes( diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c index 287fdecc0b10..11e4c4e46947 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c @@ -2244,6 +2244,9 @@ void dcn10_enable_timing_synchronization( DC_SYNC_INFO("Setting up OTG reset trigger\n"); for (i = 1; i < group_size; i++) { + if (grouped_pipes[i]->stream && grouped_pipes[i]->stream->mall_stream_config.type == SUBVP_PHANTOM) + continue; + opp = grouped_pipes[i]->stream_res.opp; tg = grouped_pipes[i]->stream_res.tg; tg->funcs->get_otg_active_size(tg, &width, &height); @@ -2254,13 +2257,21 @@ void dcn10_enable_timing_synchronization( for (i = 0; i < group_size; i++) { if (grouped_pipes[i]->stream == NULL) continue; + + if (grouped_pipes[i]->stream && grouped_pipes[i]->stream->mall_stream_config.type == SUBVP_PHANTOM) + continue; + grouped_pipes[i]->stream->vblank_synchronized = false; } - for (i = 1; i < group_size; i++) + for (i = 1; i < group_size; i++) { + if (grouped_pipes[i]->stream && grouped_pipes[i]->stream->mall_stream_config.type == SUBVP_PHANTOM) + continue; + grouped_pipes[i]->stream_res.tg->funcs->enable_reset_trigger( grouped_pipes[i]->stream_res.tg, grouped_pipes[0]->stream_res.tg->inst); + } DC_SYNC_INFO("Waiting for trigger\n"); @@ -2268,12 +2279,21 @@ void dcn10_enable_timing_synchronization( * synchronized. Look at last pipe programmed to reset. */ - wait_for_reset_trigger_to_occur(dc_ctx, grouped_pipes[1]->stream_res.tg); - for (i = 1; i < group_size; i++) + if (grouped_pipes[1]->stream && grouped_pipes[1]->stream->mall_stream_config.type != SUBVP_PHANTOM) + wait_for_reset_trigger_to_occur(dc_ctx, grouped_pipes[1]->stream_res.tg); + + for (i = 1; i < group_size; i++) { + if (grouped_pipes[i]->stream && grouped_pipes[i]->stream->mall_stream_config.type == SUBVP_PHANTOM) + continue; + grouped_pipes[i]->stream_res.tg->funcs->disable_reset_trigger( grouped_pipes[i]->stream_res.tg); + } for (i = 1; i < group_size; i++) { + if (grouped_pipes[i]->stream && grouped_pipes[i]->stream->mall_stream_config.type == SUBVP_PHANTOM) + continue; + opp = grouped_pipes[i]->stream_res.opp; tg = grouped_pipes[i]->stream_res.tg; tg->funcs->get_otg_active_size(tg, &width, &height); @@ -3005,6 +3025,7 @@ void dcn10_prepare_bandwidth( { struct dce_hwseq *hws = dc->hwseq; struct hubbub *hubbub = dc->res_pool->hubbub; + int min_fclk_khz, min_dcfclk_khz, socclk_khz; if (dc->debug.sanity_checks) hws->funcs.verify_allow_pstate_change_high(dc); @@ -3027,8 +3048,11 @@ void dcn10_prepare_bandwidth( if (dc->debug.pplib_wm_report_mode == WM_REPORT_OVERRIDE) { DC_FP_START(); - dcn_bw_notify_pplib_of_wm_ranges(dc); + dcn_get_soc_clks( + dc, &min_fclk_khz, &min_dcfclk_khz, &socclk_khz); DC_FP_END(); + dcn_bw_notify_pplib_of_wm_ranges( + dc, min_fclk_khz, min_dcfclk_khz, socclk_khz); } if (dc->debug.sanity_checks) @@ -3041,6 +3065,7 @@ void dcn10_optimize_bandwidth( { struct dce_hwseq *hws = dc->hwseq; struct hubbub *hubbub = dc->res_pool->hubbub; + int min_fclk_khz, min_dcfclk_khz, socclk_khz; if (dc->debug.sanity_checks) hws->funcs.verify_allow_pstate_change_high(dc); @@ -3064,8 +3089,11 @@ void dcn10_optimize_bandwidth( if (dc->debug.pplib_wm_report_mode == WM_REPORT_OVERRIDE) { DC_FP_START(); - dcn_bw_notify_pplib_of_wm_ranges(dc); + dcn_get_soc_clks( + dc, &min_fclk_khz, &min_dcfclk_khz, &socclk_khz); DC_FP_END(); + dcn_bw_notify_pplib_of_wm_ranges( + dc, min_fclk_khz, min_dcfclk_khz, socclk_khz); } if (dc->debug.sanity_checks) @@ -3344,127 +3372,6 @@ static bool dcn10_can_pipe_disable_cursor(struct pipe_ctx *pipe_ctx) return false; } -static bool dcn10_dmub_should_update_cursor_data( - struct pipe_ctx *pipe_ctx, - struct dc_debug_options *debug) -{ - if (pipe_ctx->plane_state->address.type == PLN_ADDR_TYPE_VIDEO_PROGRESSIVE) - return false; - - if (dcn10_can_pipe_disable_cursor(pipe_ctx)) - return false; - - if ((pipe_ctx->stream->link->psr_settings.psr_version == DC_PSR_VERSION_SU_1 || pipe_ctx->stream->link->psr_settings.psr_version == DC_PSR_VERSION_1) - && pipe_ctx->stream->ctx->dce_version >= DCN_VERSION_3_1) - return true; - - return false; -} - -static void dcn10_dmub_update_cursor_data( - struct pipe_ctx *pipe_ctx, - struct hubp *hubp, - const struct dc_cursor_mi_param *param, - const struct dc_cursor_position *cur_pos, - const struct dc_cursor_attributes *cur_attr) -{ - union dmub_rb_cmd cmd; - struct dmub_cmd_update_cursor_info_data *update_cursor_info; - const struct dc_cursor_position *pos; - const struct dc_cursor_attributes *attr; - int src_x_offset = 0; - int src_y_offset = 0; - int x_hotspot = 0; - int cursor_height = 0; - int cursor_width = 0; - uint32_t cur_en = 0; - unsigned int panel_inst = 0; - - struct dc_debug_options *debug = &hubp->ctx->dc->debug; - - if (!dcn10_dmub_should_update_cursor_data(pipe_ctx, debug)) - return; - /** - * if cur_pos == NULL means the caller is from cursor_set_attribute - * then driver use previous cursor position data - * if cur_attr == NULL means the caller is from cursor_set_position - * then driver use previous cursor attribute - * if cur_pos or cur_attr is not NULL then update it - */ - if (cur_pos != NULL) - pos = cur_pos; - else - pos = &hubp->curs_pos; - - if (cur_attr != NULL) - attr = cur_attr; - else - attr = &hubp->curs_attr; - - if (!dc_get_edp_link_panel_inst(hubp->ctx->dc, pipe_ctx->stream->link, &panel_inst)) - return; - - src_x_offset = pos->x - pos->x_hotspot - param->viewport.x; - src_y_offset = pos->y - pos->y_hotspot - param->viewport.y; - x_hotspot = pos->x_hotspot; - cursor_height = (int)attr->height; - cursor_width = (int)attr->width; - cur_en = pos->enable ? 1:0; - - // Rotated cursor width/height and hotspots tweaks for offset calculation - if (param->rotation == ROTATION_ANGLE_90 || param->rotation == ROTATION_ANGLE_270) { - swap(cursor_height, cursor_width); - if (param->rotation == ROTATION_ANGLE_90) { - src_x_offset = pos->x - pos->y_hotspot - param->viewport.x; - src_y_offset = pos->y - pos->x_hotspot - param->viewport.y; - } - } else if (param->rotation == ROTATION_ANGLE_180) { - src_x_offset = pos->x - param->viewport.x; - src_y_offset = pos->y - param->viewport.y; - } - - if (param->mirror) { - x_hotspot = param->viewport.width - x_hotspot; - src_x_offset = param->viewport.x + param->viewport.width - src_x_offset; - } - - if (src_x_offset >= (int)param->viewport.width) - cur_en = 0; /* not visible beyond right edge*/ - - if (src_x_offset + cursor_width <= 0) - cur_en = 0; /* not visible beyond left edge*/ - - if (src_y_offset >= (int)param->viewport.height) - cur_en = 0; /* not visible beyond bottom edge*/ - - if (src_y_offset + cursor_height <= 0) - cur_en = 0; /* not visible beyond top edge*/ - - // Cursor bitmaps have different hotspot values - // There's a possibility that the above logic returns a negative value, so we clamp them to 0 - if (src_x_offset < 0) - src_x_offset = 0; - if (src_y_offset < 0) - src_y_offset = 0; - - memset(&cmd, 0x0, sizeof(cmd)); - cmd.update_cursor_info.header.type = DMUB_CMD__UPDATE_CURSOR_INFO; - cmd.update_cursor_info.header.payload_bytes = - sizeof(cmd.update_cursor_info.update_cursor_info_data); - update_cursor_info = &cmd.update_cursor_info.update_cursor_info_data; - update_cursor_info->cursor_rect.x = src_x_offset + param->viewport.x; - update_cursor_info->cursor_rect.y = src_y_offset + param->viewport.y; - update_cursor_info->cursor_rect.width = attr->width; - update_cursor_info->cursor_rect.height = attr->height; - update_cursor_info->enable = cur_en; - update_cursor_info->pipe_idx = pipe_ctx->pipe_idx; - update_cursor_info->cmd_version = DMUB_CMD_PSR_CONTROL_VERSION_1; - update_cursor_info->panel_inst = panel_inst; - dc_dmub_srv_cmd_queue(pipe_ctx->stream->ctx->dmub_srv, &cmd); - dc_dmub_srv_cmd_execute(pipe_ctx->stream->ctx->dmub_srv); - dc_dmub_srv_wait_idle(pipe_ctx->stream->ctx->dmub_srv); -} - void dcn10_set_cursor_position(struct pipe_ctx *pipe_ctx) { struct dc_cursor_position pos_cpy = pipe_ctx->stream->cursor_position; @@ -3699,7 +3606,6 @@ void dcn10_set_cursor_position(struct pipe_ctx *pipe_ctx) pipe_ctx->plane_res.scl_data.viewport.height - pos_cpy.y; } - dcn10_dmub_update_cursor_data(pipe_ctx, hubp, ¶m, &pos_cpy, NULL); hubp->funcs->set_cursor_position(hubp, &pos_cpy, ¶m); dpp->funcs->set_cursor_position(dpp, &pos_cpy, ¶m, hubp->curs_attr.width, hubp->curs_attr.height); } @@ -3707,25 +3613,6 @@ void dcn10_set_cursor_position(struct pipe_ctx *pipe_ctx) void dcn10_set_cursor_attribute(struct pipe_ctx *pipe_ctx) { struct dc_cursor_attributes *attributes = &pipe_ctx->stream->cursor_attributes; - struct dc_cursor_mi_param param = { 0 }; - - /** - * If enter PSR without cursor attribute update - * the cursor attribute of dmub_restore_plane - * are initial value. call dmub to exit PSR and - * restore plane then update cursor attribute to - * avoid override with initial value - */ - if (pipe_ctx->plane_state != NULL) { - param.pixel_clk_khz = pipe_ctx->stream->timing.pix_clk_100hz / 10; - param.ref_clk_khz = pipe_ctx->stream->ctx->dc->res_pool->ref_clocks.dchub_ref_clock_inKhz; - param.viewport = pipe_ctx->plane_res.scl_data.viewport; - param.h_scale_ratio = pipe_ctx->plane_res.scl_data.ratios.horz; - param.v_scale_ratio = pipe_ctx->plane_res.scl_data.ratios.vert; - param.rotation = pipe_ctx->plane_state->rotation; - param.mirror = pipe_ctx->plane_state->horizontal_mirror; - dcn10_dmub_update_cursor_data(pipe_ctx, pipe_ctx->plane_res.hubp, ¶m, NULL, attributes); - } pipe_ctx->plane_res.hubp->funcs->set_cursor_attributes( pipe_ctx->plane_res.hubp, attributes); @@ -3810,81 +3697,56 @@ void dcn10_calc_vupdate_position( uint32_t *start_line, uint32_t *end_line) { - const struct dc_crtc_timing *dc_crtc_timing = &pipe_ctx->stream->timing; - int vline_int_offset_from_vupdate = - pipe_ctx->stream->periodic_interrupt0.lines_offset; - int vupdate_offset_from_vsync = dc->hwss.get_vupdate_offset_from_vsync(pipe_ctx); - int start_position; + const struct dc_crtc_timing *timing = &pipe_ctx->stream->timing; + int vupdate_pos = dc->hwss.get_vupdate_offset_from_vsync(pipe_ctx); - if (vline_int_offset_from_vupdate > 0) - vline_int_offset_from_vupdate--; - else if (vline_int_offset_from_vupdate < 0) - vline_int_offset_from_vupdate++; - - start_position = vline_int_offset_from_vupdate + vupdate_offset_from_vsync; - - if (start_position >= 0) - *start_line = start_position; + if (vupdate_pos >= 0) + *start_line = vupdate_pos - ((vupdate_pos / timing->v_total) * timing->v_total); else - *start_line = dc_crtc_timing->v_total + start_position - 1; - - *end_line = *start_line + 2; - - if (*end_line >= dc_crtc_timing->v_total) - *end_line = 2; + *start_line = vupdate_pos + ((-vupdate_pos / timing->v_total) + 1) * timing->v_total - 1; + *end_line = (*start_line + 2) % timing->v_total; } static void dcn10_cal_vline_position( struct dc *dc, struct pipe_ctx *pipe_ctx, - enum vline_select vline, uint32_t *start_line, uint32_t *end_line) { - enum vertical_interrupt_ref_point ref_point = INVALID_POINT; + const struct dc_crtc_timing *timing = &pipe_ctx->stream->timing; + int vline_pos = pipe_ctx->stream->periodic_interrupt.lines_offset; - if (vline == VLINE0) - ref_point = pipe_ctx->stream->periodic_interrupt0.ref_point; - else if (vline == VLINE1) - ref_point = pipe_ctx->stream->periodic_interrupt1.ref_point; + if (pipe_ctx->stream->periodic_interrupt.ref_point == START_V_UPDATE) { + if (vline_pos > 0) + vline_pos--; + else if (vline_pos < 0) + vline_pos++; - switch (ref_point) { - case START_V_UPDATE: - dcn10_calc_vupdate_position( - dc, - pipe_ctx, - start_line, - end_line); - break; - case START_V_SYNC: - // Suppose to do nothing because vsync is 0; - break; - default: + vline_pos += dc->hwss.get_vupdate_offset_from_vsync(pipe_ctx); + if (vline_pos >= 0) + *start_line = vline_pos - ((vline_pos / timing->v_total) * timing->v_total); + else + *start_line = vline_pos + ((-vline_pos / timing->v_total) + 1) * timing->v_total - 1; + *end_line = (*start_line + 2) % timing->v_total; + } else if (pipe_ctx->stream->periodic_interrupt.ref_point == START_V_SYNC) { + // vsync is line 0 so start_line is just the requested line offset + *start_line = vline_pos; + *end_line = (*start_line + 2) % timing->v_total; + } else ASSERT(0); - break; - } } void dcn10_setup_periodic_interrupt( struct dc *dc, - struct pipe_ctx *pipe_ctx, - enum vline_select vline) + struct pipe_ctx *pipe_ctx) { struct timing_generator *tg = pipe_ctx->stream_res.tg; + uint32_t start_line = 0; + uint32_t end_line = 0; - if (vline == VLINE0) { - uint32_t start_line = 0; - uint32_t end_line = 0; - - dcn10_cal_vline_position(dc, pipe_ctx, vline, &start_line, &end_line); + dcn10_cal_vline_position(dc, pipe_ctx, &start_line, &end_line); - tg->funcs->setup_vertical_interrupt0(tg, start_line, end_line); - - } else if (vline == VLINE1) { - pipe_ctx->stream_res.tg->funcs->setup_vertical_interrupt1( - tg, - pipe_ctx->stream->periodic_interrupt1.lines_offset); - } + tg->funcs->setup_vertical_interrupt0(tg, start_line, end_line); } void dcn10_setup_vupdate_interrupt(struct dc *dc, struct pipe_ctx *pipe_ctx) diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.h index 9ae07c77fdc0..0ef7bf7ddb75 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.h +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.h @@ -175,8 +175,7 @@ void dcn10_set_cursor_attribute(struct pipe_ctx *pipe_ctx); void dcn10_set_cursor_sdr_white_level(struct pipe_ctx *pipe_ctx); void dcn10_setup_periodic_interrupt( struct dc *dc, - struct pipe_ctx *pipe_ctx, - enum vline_select vline); + struct pipe_ctx *pipe_ctx); enum dc_status dcn10_set_clock(struct dc *dc, enum dc_clock_type clock_type, uint32_t clk_khz, diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c index ea7739255119..33d780218790 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c @@ -207,10 +207,7 @@ void optc1_program_timing( /* In case of V_TOTAL_CONTROL is on, make sure OTG_V_TOTAL_MAX and * OTG_V_TOTAL_MIN are equal to V_TOTAL. */ - REG_SET(OTG_V_TOTAL_MAX, 0, - OTG_V_TOTAL_MAX, v_total); - REG_SET(OTG_V_TOTAL_MIN, 0, - OTG_V_TOTAL_MIN, v_total); + optc->funcs->set_vtotal_min_max(optc, v_total, v_total); /* v_sync_start = 0, v_sync_end = v_sync_width */ v_sync_end = patched_crtc_timing.v_sync_width; @@ -649,13 +646,6 @@ uint32_t optc1_get_vblank_counter(struct timing_generator *optc) void optc1_lock(struct timing_generator *optc) { struct optc *optc1 = DCN10TG_FROM_TG(optc); - uint32_t regval = 0; - - regval = REG_READ(OTG_CONTROL); - - /* otg is not running, do not need to be locked */ - if ((regval & 0x1) == 0x0) - return; REG_SET(OTG_GLOBAL_CONTROL0, 0, OTG_MASTER_UPDATE_LOCK_SEL, optc->inst); @@ -663,12 +653,10 @@ void optc1_lock(struct timing_generator *optc) OTG_MASTER_UPDATE_LOCK, 1); /* Should be fast, status does not update on maximus */ - if (optc->ctx->dce_environment != DCE_ENV_FPGA_MAXIMUS) { - + if (optc->ctx->dce_environment != DCE_ENV_FPGA_MAXIMUS) REG_WAIT(OTG_MASTER_UPDATE_LOCK, UPDATE_LOCK_STATUS, 1, 1, 10); - } } void optc1_unlock(struct timing_generator *optc) @@ -679,16 +667,6 @@ void optc1_unlock(struct timing_generator *optc) OTG_MASTER_UPDATE_LOCK, 0); } -bool optc1_is_locked(struct timing_generator *optc) -{ - struct optc *optc1 = DCN10TG_FROM_TG(optc); - uint32_t locked; - - REG_GET(OTG_MASTER_UPDATE_LOCK, UPDATE_LOCK_STATUS, &locked); - - return (locked == 1); -} - void optc1_get_position(struct timing_generator *optc, struct crtc_position *position) { @@ -941,11 +919,7 @@ void optc1_set_drr( } - REG_SET(OTG_V_TOTAL_MAX, 0, - OTG_V_TOTAL_MAX, params->vertical_total_max - 1); - - REG_SET(OTG_V_TOTAL_MIN, 0, - OTG_V_TOTAL_MIN, params->vertical_total_min - 1); + optc->funcs->set_vtotal_min_max(optc, params->vertical_total_min - 1, params->vertical_total_max - 1); REG_UPDATE_5(OTG_V_TOTAL_CONTROL, OTG_V_TOTAL_MIN_SEL, 1, @@ -964,11 +938,7 @@ void optc1_set_drr( OTG_V_TOTAL_MAX_SEL, 0, OTG_FORCE_LOCK_ON_EVENT, 0); - REG_SET(OTG_V_TOTAL_MIN, 0, - OTG_V_TOTAL_MIN, 0); - - REG_SET(OTG_V_TOTAL_MAX, 0, - OTG_V_TOTAL_MAX, 0); + optc->funcs->set_vtotal_min_max(optc, 0, 0); } } @@ -1583,11 +1553,11 @@ static const struct timing_generator_funcs dcn10_tg_funcs = { .enable_crtc_reset = optc1_enable_crtc_reset, .disable_reset_trigger = optc1_disable_reset_trigger, .lock = optc1_lock, - .is_locked = optc1_is_locked, .unlock = optc1_unlock, .enable_optc_clock = optc1_enable_optc_clock, .set_drr = optc1_set_drr, .get_last_used_drr_vtotal = NULL, + .set_vtotal_min_max = optc1_set_vtotal_min_max, .set_static_screen_control = optc1_set_static_screen_control, .set_test_pattern = optc1_set_test_pattern, .program_stereo = optc1_program_stereo, diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h index 6323ca6dc3b3..88ac5f6f4c96 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h @@ -654,7 +654,6 @@ void optc1_set_blank(struct timing_generator *optc, bool enable_blanking); bool optc1_is_blanked(struct timing_generator *optc); -bool optc1_is_locked(struct timing_generator *optc); void optc1_program_blank_color( struct timing_generator *optc, diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c index 831080b9eb87..56d30baf12df 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c @@ -1336,6 +1336,21 @@ static noinline void dcn10_resource_construct_fp( } } +static bool verify_clock_values(struct dm_pp_clock_levels_with_voltage *clks) +{ + int i; + + if (clks->num_levels == 0) + return false; + + for (i = 0; i < clks->num_levels; i++) + /* Ensure that the result is sane */ + if (clks->data[i].clocks_in_khz == 0) + return false; + + return true; +} + static bool dcn10_resource_construct( uint8_t num_virtual_links, struct dc *dc, @@ -1345,6 +1360,9 @@ static bool dcn10_resource_construct( int j; struct dc_context *ctx = dc->ctx; uint32_t pipe_fuses = read_pipe_fuses(ctx); + struct dm_pp_clock_levels_with_voltage fclks = {0}, dcfclks = {0}; + int min_fclk_khz, min_dcfclk_khz, socclk_khz; + bool res; ctx->dc_bios->regs = &bios_regs; @@ -1523,15 +1541,53 @@ static bool dcn10_resource_construct( && pool->base.pp_smu->rv_funcs.set_pme_wa_enable != NULL) dc->debug.az_endpoint_mute_only = false; - DC_FP_START(); - if (!dc->debug.disable_pplib_clock_request) - dcn_bw_update_from_pplib(dc); + + if (!dc->debug.disable_pplib_clock_request) { + /* + * TODO: This is not the proper way to obtain + * fabric_and_dram_bandwidth, should be min(fclk, memclk). + */ + res = dm_pp_get_clock_levels_by_type_with_voltage( + ctx, DM_PP_CLOCK_TYPE_FCLK, &fclks); + + DC_FP_START(); + + if (res) + res = verify_clock_values(&fclks); + + if (res) + dcn_bw_update_from_pplib_fclks(dc, &fclks); + else + BREAK_TO_DEBUGGER(); + + DC_FP_END(); + + res = dm_pp_get_clock_levels_by_type_with_voltage( + ctx, DM_PP_CLOCK_TYPE_DCFCLK, &dcfclks); + + DC_FP_START(); + + if (res) + res = verify_clock_values(&dcfclks); + + if (res) + dcn_bw_update_from_pplib_dcfclks(dc, &dcfclks); + else + BREAK_TO_DEBUGGER(); + + DC_FP_END(); + } + dcn_bw_sync_calcs_and_dml(dc); if (!dc->debug.disable_pplib_wm_range) { dc->res_pool = &pool->base; - dcn_bw_notify_pplib_of_wm_ranges(dc); + DC_FP_START(); + dcn_get_soc_clks( + dc, &min_fclk_khz, &min_dcfclk_khz, &socclk_khz); + DC_FP_END(); + dcn_bw_notify_pplib_of_wm_ranges( + dc, min_fclk_khz, min_dcfclk_khz, socclk_khz); } - DC_FP_END(); { struct irq_service_init_data init_data; diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubp.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubp.c index b1ec0e6f7f58..4996d2810edb 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubp.c +++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubp.c @@ -617,6 +617,17 @@ void hubp2_cursor_set_attributes( CURSOR0_DST_Y_OFFSET, 0, /* used to shift the cursor chunk request deadline */ CURSOR0_CHUNK_HDL_ADJUST, 3); + + hubp->att.SURFACE_ADDR_HIGH = attr->address.high_part; + hubp->att.SURFACE_ADDR = attr->address.low_part; + hubp->att.size.bits.width = attr->width; + hubp->att.size.bits.height = attr->height; + hubp->att.cur_ctl.bits.mode = attr->color_format; + hubp->att.cur_ctl.bits.pitch = hw_pitch; + hubp->att.cur_ctl.bits.line_per_chunk = lpc; + hubp->att.cur_ctl.bits.cur_2x_magnify = attr->attribute_flags.bits.ENABLE_MAGNIFICATION; + hubp->att.settings.bits.dst_y_offset = 0; + hubp->att.settings.bits.chunk_hdl_adjust = 3; } void hubp2_dmdata_set_attributes( @@ -1033,6 +1044,25 @@ void hubp2_cursor_set_position( REG_SET(CURSOR_DST_OFFSET, 0, CURSOR_DST_X_OFFSET, dst_x_offset); /* TODO Handle surface pixel formats other than 4:4:4 */ + /* Cursor Position Register Config */ + hubp->pos.cur_ctl.bits.cur_enable = cur_en; + hubp->pos.position.bits.x_pos = pos->x; + hubp->pos.position.bits.y_pos = pos->y; + hubp->pos.hot_spot.bits.x_hot = x_hotspot; + hubp->pos.hot_spot.bits.y_hot = y_hotspot; + hubp->pos.dst_offset.bits.dst_x_offset = dst_x_offset; + /* Cursor Rectangle Cache + * Cursor bitmaps have different hotspot values + * There's a possibility that the above logic returns a negative value, + * so we clamp them to 0 + */ + if (src_x_offset < 0) + src_x_offset = 0; + if (src_y_offset < 0) + src_y_offset = 0; + /* Save necessary cursor info x, y position. w, h is saved in attribute func. */ + hubp->cur_rect.x = src_x_offset + param->viewport.x; + hubp->cur_rect.y = src_y_offset + param->viewport.y; } void hubp2_clk_cntl(struct hubp *hubp, bool enable) diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c index e1d271fe9e64..d732b6f031a1 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c @@ -1862,24 +1862,6 @@ void dcn20_post_unlock_program_front_end( for (i = 0; i < dc->res_pool->pipe_count; i++) { struct pipe_ctx *pipe = &context->res_ctx.pipe_ctx[i]; - struct pipe_ctx *mpcc_pipe; - - if (pipe->vtp_locked) { - dc->hwseq->funcs.wait_for_blank_complete(pipe->stream_res.opp); - pipe->plane_res.hubp->funcs->set_blank(pipe->plane_res.hubp, true); - pipe->vtp_locked = false; - - for (mpcc_pipe = pipe->bottom_pipe; mpcc_pipe; mpcc_pipe = mpcc_pipe->bottom_pipe) - mpcc_pipe->plane_res.hubp->funcs->set_blank(mpcc_pipe->plane_res.hubp, true); - - for (i = 0; i < dc->res_pool->pipe_count; i++) - if (context->res_ctx.pipe_ctx[i].update_flags.bits.disable) - dc->hwss.disable_plane(dc, &dc->current_state->res_ctx.pipe_ctx[i]); - } - } - - for (i = 0; i < dc->res_pool->pipe_count; i++) { - struct pipe_ctx *pipe = &context->res_ctx.pipe_ctx[i]; struct pipe_ctx *old_pipe = &dc->current_state->res_ctx.pipe_ctx[i]; /* If an active, non-phantom pipe is being transitioned into a phantom @@ -2018,6 +2000,10 @@ void dcn20_optimize_bandwidth( context->bw_ctx.bw.dcn.clk.dramclk_khz <= dc->clk_mgr->bw_params->dc_mode_softmax_memclk * 1000) dc->clk_mgr->funcs->set_max_memclk(dc->clk_mgr, dc->clk_mgr->bw_params->dc_mode_softmax_memclk); + /* increase compbuf size */ + if (hubbub->funcs->program_compbuf_size) + hubbub->funcs->program_compbuf_size(hubbub, context->bw_ctx.bw.dcn.compbuf_size_kb, true); + dc->clk_mgr->funcs->update_clocks( dc->clk_mgr, context, @@ -2033,9 +2019,6 @@ void dcn20_optimize_bandwidth( pipe_ctx->dlg_regs.optimized_min_dst_y_next_start); } } - /* increase compbuf size */ - if (hubbub->funcs->program_compbuf_size) - hubbub->funcs->program_compbuf_size(hubbub, context->bw_ctx.bw.dcn.compbuf_size_kb, true); } bool dcn20_update_bandwidth( diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_optc.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_optc.c index 0340fdd3f5fb..a08c335b7383 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_optc.c +++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_optc.c @@ -529,6 +529,7 @@ static struct timing_generator_funcs dcn20_tg_funcs = { .enable_optc_clock = optc1_enable_optc_clock, .set_drr = optc1_set_drr, .get_last_used_drr_vtotal = optc2_get_last_used_drr_vtotal, + .set_vtotal_min_max = optc1_set_vtotal_min_max, .set_static_screen_control = optc1_set_static_screen_control, .program_stereo = optc1_program_stereo, .is_stereo_left_eye = optc1_is_stereo_left_eye, diff --git a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hubbub.c b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hubbub.c index 5752271f22df..c5e200d09038 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hubbub.c +++ b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hubbub.c @@ -67,15 +67,9 @@ static uint32_t convert_and_clamp( void dcn21_dchvm_init(struct hubbub *hubbub) { struct dcn20_hubbub *hubbub1 = TO_DCN20_HUBBUB(hubbub); - uint32_t riommu_active, prefetch_done; + uint32_t riommu_active; int i; - REG_GET(DCHVM_RIOMMU_STAT0, HOSTVM_PREFETCH_DONE, &prefetch_done); - - if (prefetch_done) { - hubbub->riommu_active = true; - return; - } //Init DCHVM block REG_UPDATE(DCHVM_CTRL0, HOSTVM_INIT_REQ, 1); diff --git a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_resource.c b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_resource.c index 7cb35bb1c0f1..887081472c0d 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_resource.c @@ -657,7 +657,6 @@ static const struct dc_debug_options debug_defaults_drv = { .usbc_combo_phy_reset_wa = true, .dmub_command_table = true, .use_max_lb = true, - .optimize_edp_link_rate = true }; static const struct dc_debug_options debug_defaults_diags = { @@ -677,6 +676,12 @@ static const struct dc_debug_options debug_defaults_diags = { .use_max_lb = true }; +static const struct dc_panel_config panel_config_defaults = { + .ilr = { + .optimize_edp_link_rate = true, + }, +}; + enum dcn20_clk_src_array_id { DCN20_CLK_SRC_PLL0, DCN20_CLK_SRC_PLL1, @@ -1367,6 +1372,11 @@ static struct panel_cntl *dcn21_panel_cntl_create(const struct panel_cntl_init_d return &panel_cntl->base; } +static void dcn21_get_panel_config_defaults(struct dc_panel_config *panel_config) +{ + *panel_config = panel_config_defaults; +} + #define CTX ctx #define REG(reg_name) \ @@ -1408,6 +1418,7 @@ static const struct resource_funcs dcn21_res_pool_funcs = { .set_mcif_arb_params = dcn20_set_mcif_arb_params, .find_first_free_match_stream_enc_for_link = dcn10_find_first_free_match_stream_enc_for_link, .update_bw_bounding_box = dcn21_update_bw_bounding_box, + .get_panel_config_defaults = dcn21_get_panel_config_defaults, }; static bool dcn21_resource_construct( diff --git a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dpp.c b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dpp.c index 4a668d6563df..e5b7ef7422b8 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dpp.c +++ b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_dpp.c @@ -372,6 +372,10 @@ void dpp3_set_cursor_attributes( REG_UPDATE(CURSOR0_COLOR1, CUR0_COLOR1, 0xFFFFFFFF); } + + dpp_base->att.cur0_ctl.bits.expansion_mode = 0; + dpp_base->att.cur0_ctl.bits.cur0_rom_en = cur_rom_en; + dpp_base->att.cur0_ctl.bits.mode = color_format; } diff --git a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_optc.c b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_optc.c index 1782b9c26cf4..892d3c4d01a1 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_optc.c +++ b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_optc.c @@ -319,13 +319,13 @@ static struct timing_generator_funcs dcn30_tg_funcs = { .enable_crtc_reset = optc1_enable_crtc_reset, .disable_reset_trigger = optc1_disable_reset_trigger, .lock = optc3_lock, - .is_locked = optc1_is_locked, .unlock = optc1_unlock, .lock_doublebuffer_enable = optc3_lock_doublebuffer_enable, .lock_doublebuffer_disable = optc3_lock_doublebuffer_disable, .enable_optc_clock = optc1_enable_optc_clock, .set_drr = optc1_set_drr, .get_last_used_drr_vtotal = optc2_get_last_used_drr_vtotal, + .set_vtotal_min_max = optc3_set_vtotal_min_max, .set_static_screen_control = optc1_set_static_screen_control, .program_stereo = optc1_program_stereo, .is_stereo_left_eye = optc1_is_stereo_left_eye, @@ -366,4 +366,3 @@ void dcn30_timing_generator_init(struct optc *optc1) optc1->min_h_sync_width = 4; optc1->min_v_sync_width = 1; } - diff --git a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_resource.c b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_resource.c index f6f3878c99b8..020f512e9690 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_resource.c @@ -724,7 +724,8 @@ static const struct dc_debug_options debug_defaults_drv = { .dwb_fi_phase = -1, // -1 = disable, .dmub_command_table = true, .disable_psr = false, - .use_max_lb = true + .use_max_lb = true, + .exit_idle_opt_for_cursor_updates = true }; static const struct dc_debug_options debug_defaults_diags = { @@ -1654,6 +1655,9 @@ noinline bool dcn30_internal_validate_bw( if (!pipes) return false; + context->bw_ctx.dml.vba.maxMpcComb = 0; + context->bw_ctx.dml.vba.VoltageLevel = 0; + context->bw_ctx.dml.vba.DRAMClockChangeSupport[0][0] = dm_dram_clock_change_vactive; dc->res_pool->funcs->update_soc_for_wm_a(dc, context); pipe_cnt = dc->res_pool->funcs->populate_dml_pipes(dc, context, pipes, fast_validate); @@ -1872,6 +1876,7 @@ noinline bool dcn30_internal_validate_bw( if (repopulate_pipes) pipe_cnt = dc->res_pool->funcs->populate_dml_pipes(dc, context, pipes, fast_validate); + context->bw_ctx.dml.vba.VoltageLevel = vlevel; *vlevel_out = vlevel; *pipe_cnt_out = pipe_cnt; diff --git a/drivers/gpu/drm/amd/display/dc/dcn301/dcn301_resource.c b/drivers/gpu/drm/amd/display/dc/dcn301/dcn301_resource.c index 0c2b15a0f3a7..f04595b750ab 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn301/dcn301_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn301/dcn301_resource.c @@ -700,6 +700,7 @@ static const struct dc_debug_options debug_defaults_drv = { .dwb_fi_phase = -1, // -1 = disable .dmub_command_table = true, .use_max_lb = false, + .exit_idle_opt_for_cursor_updates = true }; static const struct dc_debug_options debug_defaults_diags = { @@ -851,7 +852,7 @@ static struct hubbub *dcn301_hubbub_create(struct dc_context *ctx) vmid->masks = &vmid_masks; } - hubbub3->num_vmid = res_cap_dcn301.num_vmid; + hubbub3->num_vmid = res_cap_dcn301.num_vmid; return &hubbub3->base; } diff --git a/drivers/gpu/drm/amd/display/dc/dcn302/dcn302_resource.c b/drivers/gpu/drm/amd/display/dc/dcn302/dcn302_resource.c index 4fab537e822f..b925b6ddde5a 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn302/dcn302_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn302/dcn302_resource.c @@ -93,7 +93,8 @@ static const struct dc_debug_options debug_defaults_drv = { .underflow_assert_delay_us = 0xFFFFFFFF, .dwb_fi_phase = -1, // -1 = disable, .dmub_command_table = true, - .use_max_lb = true + .use_max_lb = true, + .exit_idle_opt_for_cursor_updates = true }; static const struct dc_debug_options debug_defaults_diags = { diff --git a/drivers/gpu/drm/amd/display/dc/dcn303/dcn303_resource.c b/drivers/gpu/drm/amd/display/dc/dcn303/dcn303_resource.c index d97076648acb..527d5c902878 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn303/dcn303_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn303/dcn303_resource.c @@ -77,6 +77,7 @@ static const struct dc_debug_options debug_defaults_drv = { .underflow_assert_delay_us = 0xFFFFFFFF, .dwb_fi_phase = -1, // -1 = disable, .dmub_command_table = true, + .exit_idle_opt_for_cursor_updates = true, .disable_idle_power_optimizations = false, }; diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hpo_dp_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hpo_dp_stream_encoder.c index 23621ff08c90..814f401db3b3 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hpo_dp_stream_encoder.c +++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hpo_dp_stream_encoder.c @@ -150,9 +150,9 @@ static void dcn31_hpo_dp_stream_enc_dp_blank( * 10us*5000=50ms. This covers 41.7ms of minimum 24 Hz mode + * a little more because we may not trust delay accuracy. */ - //REG_WAIT(DP_SYM32_ENC_VID_STREAM_CONTROL, - // VID_STREAM_STATUS, 0, - // 10, 5000); + REG_WAIT(DP_SYM32_ENC_VID_STREAM_CONTROL, + VID_STREAM_STATUS, 0, + 10, 5000); /* Disable SDP tranmission */ REG_UPDATE(DP_SYM32_ENC_SDP_CONTROL, @@ -197,7 +197,7 @@ static void dcn31_hpo_dp_stream_enc_set_stream_attribute( uint32_t h_back_porch; uint32_t h_width; uint32_t v_height; - unsigned long long v_freq; + uint64_t v_freq; uint8_t misc0 = 0; uint8_t misc1 = 0; uint8_t hsp; @@ -360,7 +360,7 @@ static void dcn31_hpo_dp_stream_enc_set_stream_attribute( v_height = hw_crtc_timing.v_border_top + hw_crtc_timing.v_addressable + hw_crtc_timing.v_border_bottom; hsp = hw_crtc_timing.flags.HSYNC_POSITIVE_POLARITY ? 0 : 0x80; vsp = hw_crtc_timing.flags.VSYNC_POSITIVE_POLARITY ? 0 : 0x80; - v_freq = hw_crtc_timing.pix_clk_100hz * 100; + v_freq = (uint64_t)hw_crtc_timing.pix_clk_100hz * 100; /* MSA Packet Mapping to 32-bit Link Symbols - DP2 spec, section 2.7.4.1 * @@ -436,32 +436,28 @@ static void dcn31_hpo_dp_stream_enc_update_dp_info_packets( { struct dcn31_hpo_dp_stream_encoder *enc3 = DCN3_1_HPO_DP_STREAM_ENC_FROM_HPO_STREAM_ENC(enc); uint32_t dmdata_packet_enabled = 0; - bool sdp_stream_enable = false; - if (info_frame->vsc.valid) { + if (info_frame->vsc.valid) enc->vpg->funcs->update_generic_info_packet( enc->vpg, 0, /* packetIndex */ &info_frame->vsc, true); - sdp_stream_enable = true; - } - if (info_frame->spd.valid) { + + if (info_frame->spd.valid) enc->vpg->funcs->update_generic_info_packet( enc->vpg, 2, /* packetIndex */ &info_frame->spd, true); - sdp_stream_enable = true; - } - if (info_frame->hdrsmd.valid) { + + if (info_frame->hdrsmd.valid) enc->vpg->funcs->update_generic_info_packet( enc->vpg, 3, /* packetIndex */ &info_frame->hdrsmd, true); - sdp_stream_enable = true; - } + /* enable/disable transmission of packet(s). * If enabled, packet transmission begins on the next frame */ diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_optc.c b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_optc.c index 2f7404a97479..63a677c8ee27 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_optc.c +++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_optc.c @@ -201,7 +201,6 @@ void optc31_set_drr( // Setup manual flow control for EOF via TRIG_A optc->funcs->setup_manual_trigger(optc); - } else { REG_UPDATE_4(OTG_V_TOTAL_CONTROL, OTG_SET_V_TOTAL_MIN_MASK, 0, @@ -260,7 +259,6 @@ static struct timing_generator_funcs dcn31_tg_funcs = { .enable_crtc_reset = optc1_enable_crtc_reset, .disable_reset_trigger = optc1_disable_reset_trigger, .lock = optc3_lock, - .is_locked = optc1_is_locked, .unlock = optc1_unlock, .lock_doublebuffer_enable = optc3_lock_doublebuffer_enable, .lock_doublebuffer_disable = optc3_lock_doublebuffer_disable, diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.c b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.c index 8c1a6fb36306..fddc21a5a04c 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.c @@ -888,9 +888,8 @@ static const struct dc_debug_options debug_defaults_drv = { } }, .disable_z10 = true, - .optimize_edp_link_rate = true, .enable_z9_disable_interface = true, /* Allow support for the PMFW interface for disable Z9*/ - .dml_hostvm_override = DML_HOSTVM_NO_OVERRIDE, + .dml_hostvm_override = DML_HOSTVM_OVERRIDE_FALSE, }; static const struct dc_debug_options debug_defaults_diags = { @@ -911,6 +910,12 @@ static const struct dc_debug_options debug_defaults_diags = { .use_max_lb = true }; +static const struct dc_panel_config panel_config_defaults = { + .ilr = { + .optimize_edp_link_rate = true, + }, +}; + static void dcn31_dpp_destroy(struct dpp **dpp) { kfree(TO_DCN20_DPP(*dpp)); @@ -1803,6 +1808,11 @@ validate_out: return out; } +static void dcn31_get_panel_config_defaults(struct dc_panel_config *panel_config) +{ + *panel_config = panel_config_defaults; +} + static struct dc_cap_funcs cap_funcs = { .get_dcc_compression_cap = dcn20_get_dcc_compression_cap }; @@ -1829,6 +1839,7 @@ static struct resource_funcs dcn31_res_pool_funcs = { .release_post_bldn_3dlut = dcn30_release_post_bldn_3dlut, .update_bw_bounding_box = dcn31_update_bw_bounding_box, .patch_unknown_plane_state = dcn20_patch_unknown_plane_state, + .get_panel_config_defaults = dcn31_get_panel_config_defaults, }; static struct clock_source *dcn30_clock_source_create( diff --git a/drivers/gpu/drm/amd/display/dc/dcn314/dcn314_dccg.c b/drivers/gpu/drm/amd/display/dc/dcn314/dcn314_dccg.c index 36630d532c18..1bd7e0f327d8 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn314/dcn314_dccg.c +++ b/drivers/gpu/drm/amd/display/dc/dcn314/dcn314_dccg.c @@ -45,6 +45,48 @@ #define DC_LOGGER \ dccg->ctx->logger +static void dccg314_get_pixel_rate_div( + struct dccg *dccg, + uint32_t otg_inst, + enum pixel_rate_div *k1, + enum pixel_rate_div *k2) +{ + struct dcn_dccg *dccg_dcn = TO_DCN_DCCG(dccg); + uint32_t val_k1 = PIXEL_RATE_DIV_NA, val_k2 = PIXEL_RATE_DIV_NA; + + *k1 = PIXEL_RATE_DIV_NA; + *k2 = PIXEL_RATE_DIV_NA; + + switch (otg_inst) { + case 0: + REG_GET_2(OTG_PIXEL_RATE_DIV, + OTG0_PIXEL_RATE_DIVK1, &val_k1, + OTG0_PIXEL_RATE_DIVK2, &val_k2); + break; + case 1: + REG_GET_2(OTG_PIXEL_RATE_DIV, + OTG1_PIXEL_RATE_DIVK1, &val_k1, + OTG1_PIXEL_RATE_DIVK2, &val_k2); + break; + case 2: + REG_GET_2(OTG_PIXEL_RATE_DIV, + OTG2_PIXEL_RATE_DIVK1, &val_k1, + OTG2_PIXEL_RATE_DIVK2, &val_k2); + break; + case 3: + REG_GET_2(OTG_PIXEL_RATE_DIV, + OTG3_PIXEL_RATE_DIVK1, &val_k1, + OTG3_PIXEL_RATE_DIVK2, &val_k2); + break; + default: + BREAK_TO_DEBUGGER(); + return; + } + + *k1 = (enum pixel_rate_div)val_k1; + *k2 = (enum pixel_rate_div)val_k2; +} + static void dccg314_set_pixel_rate_div( struct dccg *dccg, uint32_t otg_inst, @@ -52,6 +94,11 @@ static void dccg314_set_pixel_rate_div( enum pixel_rate_div k2) { struct dcn_dccg *dccg_dcn = TO_DCN_DCCG(dccg); + enum pixel_rate_div cur_k1 = PIXEL_RATE_DIV_NA, cur_k2 = PIXEL_RATE_DIV_NA; + + dccg314_get_pixel_rate_div(dccg, otg_inst, &cur_k1, &cur_k2); + if (k1 == PIXEL_RATE_DIV_NA || k2 == PIXEL_RATE_DIV_NA || (k1 == cur_k1 && k2 == cur_k2)) + return; switch (otg_inst) { case 0: diff --git a/drivers/gpu/drm/amd/display/dc/dcn314/dcn314_dio_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dcn314/dcn314_dio_stream_encoder.c index 914c5da737ed..7e773bf7b895 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn314/dcn314_dio_stream_encoder.c +++ b/drivers/gpu/drm/amd/display/dc/dcn314/dcn314_dio_stream_encoder.c @@ -56,7 +56,8 @@ static void enc314_enable_fifo(struct stream_encoder *enc) /* TODO: Confirm if we need to wait for DIG_SYMCLK_FE_ON */ REG_WAIT(DIG_FE_CNTL, DIG_SYMCLK_FE_ON, 1, 10, 5000); - REG_UPDATE_2(DIG_FIFO_CTRL0, DIG_FIFO_RESET, 1, DIG_FIFO_READ_START_LEVEL, 0x7); + REG_UPDATE(DIG_FIFO_CTRL0, DIG_FIFO_READ_START_LEVEL, 0x7); + REG_UPDATE(DIG_FIFO_CTRL0, DIG_FIFO_RESET, 1); REG_WAIT(DIG_FIFO_CTRL0, DIG_FIFO_RESET_DONE, 1, 10, 5000); REG_UPDATE(DIG_FIFO_CTRL0, DIG_FIFO_RESET, 0); REG_WAIT(DIG_FIFO_CTRL0, DIG_FIFO_RESET_DONE, 0, 10, 5000); @@ -261,6 +262,16 @@ static bool is_two_pixels_per_containter(const struct dc_crtc_timing *timing) return two_pix; } +static void enc314_stream_encoder_dp_blank( + struct dc_link *link, + struct stream_encoder *enc) +{ + /* New to DCN314 - disable the FIFO before VID stream disable. */ + enc314_disable_fifo(enc); + + enc1_stream_encoder_dp_blank(link, enc); +} + static void enc314_stream_encoder_dp_unblank( struct dc_link *link, struct stream_encoder *enc, @@ -316,15 +327,11 @@ static void enc314_stream_encoder_dp_unblank( /* switch DP encoder to CRTC data, but reset it the fifo first. It may happen * that it overflows during mode transition, and sometimes doesn't recover. */ - REG_UPDATE(DIG_FIFO_CTRL0, DIG_FIFO_READ_START_LEVEL, 0x7); REG_UPDATE(DP_STEER_FIFO, DP_STEER_FIFO_RESET, 1); udelay(10); REG_UPDATE(DP_STEER_FIFO, DP_STEER_FIFO_RESET, 0); - /* DIG Resync FIFO now needs to be explicitly enabled. */ - enc314_enable_fifo(enc); - /* wait 100us for DIG/DP logic to prime * (i.e. a few video lines) */ @@ -340,6 +347,12 @@ static void enc314_stream_encoder_dp_unblank( REG_UPDATE(DP_VID_STREAM_CNTL, DP_VID_STREAM_ENABLE, true); + /* + * DIG Resync FIFO now needs to be explicitly enabled. + * This should come after DP_VID_STREAM_ENABLE per HW docs. + */ + enc314_enable_fifo(enc); + dp_source_sequence_trace(link, DPCD_SOURCE_SEQ_AFTER_ENABLE_DP_VID_STREAM); } @@ -408,7 +421,7 @@ static const struct stream_encoder_funcs dcn314_str_enc_funcs = { .stop_dp_info_packets = enc1_stream_encoder_stop_dp_info_packets, .dp_blank = - enc1_stream_encoder_dp_blank, + enc314_stream_encoder_dp_blank, .dp_unblank = enc314_stream_encoder_dp_unblank, .audio_mute_control = enc3_audio_mute_control, diff --git a/drivers/gpu/drm/amd/display/dc/dcn314/dcn314_resource.c b/drivers/gpu/drm/amd/display/dc/dcn314/dcn314_resource.c index 24ec71cbd3e3..d0ad72caead2 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn314/dcn314_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn314/dcn314_resource.c @@ -881,7 +881,8 @@ static const struct dc_plane_cap plane_cap = { }; static const struct dc_debug_options debug_defaults_drv = { - .disable_z10 = true, /*hw not support it*/ + .disable_z10 = false, + .enable_z9_disable_interface = true, .disable_dmcu = true, .force_abm_enable = false, .timing_trace = false, @@ -914,7 +915,6 @@ static const struct dc_debug_options debug_defaults_drv = { .afmt = true, } }, - .optimize_edp_link_rate = true, .seamless_boot_odm_combine = true }; @@ -936,6 +936,12 @@ static const struct dc_debug_options debug_defaults_diags = { .use_max_lb = true }; +static const struct dc_panel_config panel_config_defaults = { + .ilr = { + .optimize_edp_link_rate = true, + }, +}; + static void dcn31_dpp_destroy(struct dpp **dpp) { kfree(TO_DCN20_DPP(*dpp)); @@ -1675,6 +1681,11 @@ static void dcn314_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *b DC_FP_END(); } +static void dcn314_get_panel_config_defaults(struct dc_panel_config *panel_config) +{ + *panel_config = panel_config_defaults; +} + static struct resource_funcs dcn314_res_pool_funcs = { .destroy = dcn314_destroy_resource_pool, .link_enc_create = dcn31_link_encoder_create, @@ -1697,6 +1708,7 @@ static struct resource_funcs dcn314_res_pool_funcs = { .release_post_bldn_3dlut = dcn30_release_post_bldn_3dlut, .update_bw_bounding_box = dcn314_update_bw_bounding_box, .patch_unknown_plane_state = dcn20_patch_unknown_plane_state, + .get_panel_config_defaults = dcn314_get_panel_config_defaults, }; static struct clock_source *dcn30_clock_source_create( diff --git a/drivers/gpu/drm/amd/display/dc/dcn315/dcn315_resource.c b/drivers/gpu/drm/amd/display/dc/dcn315/dcn315_resource.c index eebb42c9ddd6..58746c437554 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn315/dcn315_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn315/dcn315_resource.c @@ -885,7 +885,6 @@ static const struct dc_debug_options debug_defaults_drv = { .afmt = true, } }, - .optimize_edp_link_rate = true, .psr_power_use_phy_fsm = 0, }; @@ -907,6 +906,12 @@ static const struct dc_debug_options debug_defaults_diags = { .use_max_lb = true }; +static const struct dc_panel_config panel_config_defaults = { + .ilr = { + .optimize_edp_link_rate = true, + }, +}; + static void dcn31_dpp_destroy(struct dpp **dpp) { kfree(TO_DCN20_DPP(*dpp)); @@ -1708,6 +1713,11 @@ static int dcn315_populate_dml_pipes_from_context( return pipe_cnt; } +static void dcn315_get_panel_config_defaults(struct dc_panel_config *panel_config) +{ + *panel_config = panel_config_defaults; +} + static struct dc_cap_funcs cap_funcs = { .get_dcc_compression_cap = dcn20_get_dcc_compression_cap }; @@ -1721,7 +1731,7 @@ static struct resource_funcs dcn315_res_pool_funcs = { .panel_cntl_create = dcn31_panel_cntl_create, .validate_bandwidth = dcn31_validate_bandwidth, .calculate_wm_and_dlg = dcn31_calculate_wm_and_dlg, - .update_soc_for_wm_a = dcn31_update_soc_for_wm_a, + .update_soc_for_wm_a = dcn315_update_soc_for_wm_a, .populate_dml_pipes = dcn315_populate_dml_pipes_from_context, .acquire_idle_pipe_for_layer = dcn20_acquire_idle_pipe_for_layer, .add_stream_to_ctx = dcn30_add_stream_to_ctx, @@ -1734,6 +1744,7 @@ static struct resource_funcs dcn315_res_pool_funcs = { .release_post_bldn_3dlut = dcn30_release_post_bldn_3dlut, .update_bw_bounding_box = dcn315_update_bw_bounding_box, .patch_unknown_plane_state = dcn20_patch_unknown_plane_state, + .get_panel_config_defaults = dcn315_get_panel_config_defaults, }; static bool dcn315_resource_construct( diff --git a/drivers/gpu/drm/amd/display/dc/dcn316/dcn316_resource.c b/drivers/gpu/drm/amd/display/dc/dcn316/dcn316_resource.c index f4b52a35ad84..6b40a11ac83a 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn316/dcn316_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn316/dcn316_resource.c @@ -885,7 +885,6 @@ static const struct dc_debug_options debug_defaults_drv = { .afmt = true, } }, - .optimize_edp_link_rate = true, }; static const struct dc_debug_options debug_defaults_diags = { @@ -906,6 +905,12 @@ static const struct dc_debug_options debug_defaults_diags = { .use_max_lb = true }; +static const struct dc_panel_config panel_config_defaults = { + .ilr = { + .optimize_edp_link_rate = true, + }, +}; + static void dcn31_dpp_destroy(struct dpp **dpp) { kfree(TO_DCN20_DPP(*dpp)); @@ -1710,6 +1715,11 @@ static int dcn316_populate_dml_pipes_from_context( return pipe_cnt; } +static void dcn316_get_panel_config_defaults(struct dc_panel_config *panel_config) +{ + *panel_config = panel_config_defaults; +} + static struct dc_cap_funcs cap_funcs = { .get_dcc_compression_cap = dcn20_get_dcc_compression_cap }; @@ -1736,6 +1746,7 @@ static struct resource_funcs dcn316_res_pool_funcs = { .release_post_bldn_3dlut = dcn30_release_post_bldn_3dlut, .update_bw_bounding_box = dcn316_update_bw_bounding_box, .patch_unknown_plane_state = dcn20_patch_unknown_plane_state, + .get_panel_config_defaults = dcn316_get_panel_config_defaults, }; static bool dcn316_resource_construct( diff --git a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dccg.c b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dccg.c index 26eb04ea472c..e4daed44ef5f 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dccg.c +++ b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dccg.c @@ -42,6 +42,48 @@ #define DC_LOGGER \ dccg->ctx->logger +static void dccg32_get_pixel_rate_div( + struct dccg *dccg, + uint32_t otg_inst, + enum pixel_rate_div *k1, + enum pixel_rate_div *k2) +{ + struct dcn_dccg *dccg_dcn = TO_DCN_DCCG(dccg); + uint32_t val_k1 = PIXEL_RATE_DIV_NA, val_k2 = PIXEL_RATE_DIV_NA; + + *k1 = PIXEL_RATE_DIV_NA; + *k2 = PIXEL_RATE_DIV_NA; + + switch (otg_inst) { + case 0: + REG_GET_2(OTG_PIXEL_RATE_DIV, + OTG0_PIXEL_RATE_DIVK1, &val_k1, + OTG0_PIXEL_RATE_DIVK2, &val_k2); + break; + case 1: + REG_GET_2(OTG_PIXEL_RATE_DIV, + OTG1_PIXEL_RATE_DIVK1, &val_k1, + OTG1_PIXEL_RATE_DIVK2, &val_k2); + break; + case 2: + REG_GET_2(OTG_PIXEL_RATE_DIV, + OTG2_PIXEL_RATE_DIVK1, &val_k1, + OTG2_PIXEL_RATE_DIVK2, &val_k2); + break; + case 3: + REG_GET_2(OTG_PIXEL_RATE_DIV, + OTG3_PIXEL_RATE_DIVK1, &val_k1, + OTG3_PIXEL_RATE_DIVK2, &val_k2); + break; + default: + BREAK_TO_DEBUGGER(); + return; + } + + *k1 = (enum pixel_rate_div)val_k1; + *k2 = (enum pixel_rate_div)val_k2; +} + static void dccg32_set_pixel_rate_div( struct dccg *dccg, uint32_t otg_inst, @@ -50,6 +92,17 @@ static void dccg32_set_pixel_rate_div( { struct dcn_dccg *dccg_dcn = TO_DCN_DCCG(dccg); + enum pixel_rate_div cur_k1 = PIXEL_RATE_DIV_NA, cur_k2 = PIXEL_RATE_DIV_NA; + + // Don't program 0xF into the register field. Not valid since + // K1 / K2 field is only 1 / 2 bits wide + if (k1 == PIXEL_RATE_DIV_NA || k2 == PIXEL_RATE_DIV_NA) + return; + + dccg32_get_pixel_rate_div(dccg, otg_inst, &cur_k1, &cur_k2); + if (k1 == cur_k1 && k2 == cur_k2) + return; + switch (otg_inst) { case 0: REG_UPDATE_2(OTG_PIXEL_RATE_DIV, diff --git a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dio_link_encoder.c b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dio_link_encoder.c index fdae6aa89908..076969d928af 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dio_link_encoder.c +++ b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dio_link_encoder.c @@ -150,12 +150,6 @@ static void dcn32_link_encoder_get_max_link_cap(struct link_encoder *enc, } -void enc32_set_dig_output_mode(struct link_encoder *enc, uint8_t pix_per_container) -{ - struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc); - REG_UPDATE(DIG_FIFO_CTRL0, DIG_FIFO_OUTPUT_PIXEL_MODE, pix_per_container); -} - static const struct link_encoder_funcs dcn32_link_enc_funcs = { .read_state = link_enc2_read_state, .validate_output_with_stream = @@ -186,7 +180,6 @@ static const struct link_encoder_funcs dcn32_link_enc_funcs = { .is_in_alt_mode = dcn32_link_encoder_is_in_alt_mode, .get_max_link_cap = dcn32_link_encoder_get_max_link_cap, .set_dio_phy_mux = dcn31_link_encoder_set_dio_phy_mux, - .set_dig_output_mode = enc32_set_dig_output_mode, }; void dcn32_link_encoder_construct( diff --git a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dio_link_encoder.h b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dio_link_encoder.h index 749a1e8cb811..bbcfce06bec0 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dio_link_encoder.h +++ b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dio_link_encoder.h @@ -53,8 +53,4 @@ void dcn32_link_encoder_enable_dp_output( const struct dc_link_settings *link_settings, enum clock_source_id clock_source); -void enc32_set_dig_output_mode( - struct link_encoder *enc, - uint8_t pix_per_container); - #endif /* __DC_LINK_ENCODER__DCN32_H__ */ diff --git a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dio_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dio_stream_encoder.c index 0e9dce414641..d19fc93dbc75 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dio_stream_encoder.c +++ b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dio_stream_encoder.c @@ -243,6 +243,39 @@ static bool is_two_pixels_per_containter(const struct dc_crtc_timing *timing) return two_pix; } +static bool is_h_timing_divisible_by_2(const struct dc_crtc_timing *timing) +{ + /* math borrowed from function of same name in inc/resource + * checks if h_timing is divisible by 2 + */ + + bool divisible = false; + uint16_t h_blank_start = 0; + uint16_t h_blank_end = 0; + + if (timing) { + h_blank_start = timing->h_total - timing->h_front_porch; + h_blank_end = h_blank_start - timing->h_addressable; + + /* HTOTAL, Hblank start/end, and Hsync start/end all must be + * divisible by 2 in order for the horizontal timing params + * to be considered divisible by 2. Hsync start is always 0. + */ + divisible = (timing->h_total % 2 == 0) && + (h_blank_start % 2 == 0) && + (h_blank_end % 2 == 0) && + (timing->h_sync_width % 2 == 0); + } + return divisible; +} + +static bool is_dp_dig_pixel_rate_div_policy(struct dc *dc, const struct dc_crtc_timing *timing) +{ + /* should be functionally the same as dcn32_is_dp_dig_pixel_rate_div_policy for DP encoders*/ + return is_h_timing_divisible_by_2(timing) && + dc->debug.enable_dp_dig_pixel_rate_div_policy; +} + static void enc32_stream_encoder_dp_unblank( struct dc_link *link, struct stream_encoder *enc, @@ -259,7 +292,7 @@ static void enc32_stream_encoder_dp_unblank( /* YCbCr 4:2:0 : Computed VID_M will be 2X the input rate */ if (is_two_pixels_per_containter(¶m->timing) || param->opp_cnt > 1 - || dc->debug.enable_dp_dig_pixel_rate_div_policy) { + || is_dp_dig_pixel_rate_div_policy(dc, ¶m->timing)) { /*this logic should be the same in get_pixel_clock_parameters() */ n_multiply = 1; } @@ -355,7 +388,7 @@ static void enc32_dp_set_dsc_config(struct stream_encoder *enc, { struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc); - REG_UPDATE(DP_DSC_CNTL, DP_DSC_MODE, dsc_mode); + REG_UPDATE(DP_DSC_CNTL, DP_DSC_MODE, dsc_mode == OPTC_DSC_DISABLED ? 0 : 1); } /* this function read dsc related register fields to be logged later in dcn10_log_hw_state @@ -378,24 +411,6 @@ static void enc32_read_state(struct stream_encoder *enc, struct enc_state *s) } } -static void enc32_stream_encoder_reset_fifo(struct stream_encoder *enc) -{ - struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc); - uint32_t fifo_enabled; - - REG_GET(DIG_FIFO_CTRL0, DIG_FIFO_ENABLE, &fifo_enabled); - - if (fifo_enabled == 0) { - /* reset DIG resync FIFO */ - REG_UPDATE(DIG_FIFO_CTRL0, DIG_FIFO_RESET, 1); - /* TODO: fix timeout when wait for DIG_FIFO_RESET_DONE */ - //REG_WAIT(DIG_FIFO_CTRL0, DIG_FIFO_RESET_DONE, 1, 1, 100); - udelay(1); - REG_UPDATE(DIG_FIFO_CTRL0, DIG_FIFO_RESET, 0); - REG_WAIT(DIG_FIFO_CTRL0, DIG_FIFO_RESET_DONE, 0, 1, 100); - } -} - static void enc32_set_dig_input_mode(struct stream_encoder *enc, unsigned int pix_per_container) { struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc); @@ -425,8 +440,6 @@ static const struct stream_encoder_funcs dcn32_str_enc_funcs = { enc3_stream_encoder_update_dp_info_packets, .stop_dp_info_packets = enc1_stream_encoder_stop_dp_info_packets, - .reset_fifo = - enc32_stream_encoder_reset_fifo, .dp_blank = enc1_stream_encoder_dp_blank, .dp_unblank = diff --git a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dio_stream_encoder.h b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dio_stream_encoder.h index 250d9a341cf6..ecd041a446d2 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dio_stream_encoder.h +++ b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dio_stream_encoder.h @@ -71,7 +71,9 @@ SRI(DP_MSE_RATE_UPDATE, DP, id), \ SRI(DP_PIXEL_FORMAT, DP, id), \ SRI(DP_SEC_CNTL, DP, id), \ + SRI(DP_SEC_CNTL1, DP, id), \ SRI(DP_SEC_CNTL2, DP, id), \ + SRI(DP_SEC_CNTL5, DP, id), \ SRI(DP_SEC_CNTL6, DP, id), \ SRI(DP_STEER_FIFO, DP, id), \ SRI(DP_VID_M, DP, id), \ @@ -93,7 +95,7 @@ SRI(DIG_FIFO_CTRL0, DIG, id) -#define SE_COMMON_MASK_SH_LIST_DCN32_BASE(mask_sh)\ +#define SE_COMMON_MASK_SH_LIST_DCN32(mask_sh)\ SE_SF(DP0_DP_PIXEL_FORMAT, DP_PIXEL_ENCODING, mask_sh),\ SE_SF(DP0_DP_PIXEL_FORMAT, DP_COMPONENT_DEPTH, mask_sh),\ SE_SF(DP0_DP_PIXEL_FORMAT, DP_PIXEL_PER_CYCLE_PROCESSING_MODE, mask_sh),\ @@ -106,6 +108,7 @@ SE_SF(DIG0_HDMI_VBI_PACKET_CONTROL, HDMI_GC_CONT, mask_sh),\ SE_SF(DIG0_HDMI_VBI_PACKET_CONTROL, HDMI_GC_SEND, mask_sh),\ SE_SF(DIG0_HDMI_VBI_PACKET_CONTROL, HDMI_NULL_SEND, mask_sh),\ + SE_SF(DIG0_HDMI_VBI_PACKET_CONTROL, HDMI_ACP_SEND, mask_sh),\ SE_SF(DIG0_HDMI_INFOFRAME_CONTROL0, HDMI_AUDIO_INFO_SEND, mask_sh),\ SE_SF(DIG0_HDMI_INFOFRAME_CONTROL1, HDMI_AUDIO_INFO_LINE, mask_sh),\ SE_SF(DIG0_HDMI_GC, HDMI_GC_AVMUTE, mask_sh),\ @@ -244,15 +247,6 @@ SE_SF(DIG0_DIG_FIFO_CTRL0, DIG_FIFO_RESET_DONE, mask_sh),\ SE_SF(DIG0_DIG_FIFO_CTRL0, DIG_FIFO_OUTPUT_PIXEL_MODE, mask_sh) -#if defined(CONFIG_DRM_AMD_DC_HDCP) -#define SE_COMMON_MASK_SH_LIST_DCN32(mask_sh)\ - SE_COMMON_MASK_SH_LIST_DCN32_BASE(mask_sh),\ - SE_SF(DIG0_HDMI_VBI_PACKET_CONTROL, HDMI_ACP_SEND, mask_sh) -#else -#define SE_COMMON_MASK_SH_LIST_DCN32(mask_sh)\ - SE_COMMON_MASK_SH_LIST_DCN32_BASE(mask_sh) -#endif - void dcn32_dio_stream_encoder_construct( struct dcn10_stream_encoder *enc1, struct dc_context *ctx, diff --git a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_hpo_dp_link_encoder.h b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_hpo_dp_link_encoder.h index 9db1323e1933..176b1537d2a1 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_hpo_dp_link_encoder.h +++ b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_hpo_dp_link_encoder.h @@ -47,6 +47,7 @@ SE_SF(DP_DPHY_SYM320_DP_DPHY_SYM32_TP_CONFIG, TP_PRBS_SEL1, mask_sh),\ SE_SF(DP_DPHY_SYM320_DP_DPHY_SYM32_TP_CONFIG, TP_PRBS_SEL2, mask_sh),\ SE_SF(DP_DPHY_SYM320_DP_DPHY_SYM32_TP_CONFIG, TP_PRBS_SEL3, mask_sh),\ + SE_SF(DP_DPHY_SYM320_DP_DPHY_SYM32_TP_SQ_PULSE, TP_SQ_PULSE_WIDTH, mask_sh),\ SE_SF(DP_DPHY_SYM320_DP_DPHY_SYM32_SAT_VC0, SAT_STREAM_SOURCE, mask_sh),\ SE_SF(DP_DPHY_SYM320_DP_DPHY_SYM32_SAT_VC0, SAT_SLOT_COUNT, mask_sh),\ SE_SF(DP_DPHY_SYM320_DP_DPHY_SYM32_VC_RATE_CNTL0, STREAM_VC_RATE_X, mask_sh),\ diff --git a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_hubbub.c b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_hubbub.c index 0b70247a5d36..9fbb72369c10 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_hubbub.c +++ b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_hubbub.c @@ -98,9 +98,13 @@ void dcn32_program_det_size(struct hubbub *hubbub, int hubp_inst, unsigned int d default: break; } - /* Should never be hit, if it is we have an erroneous hw config*/ - ASSERT(hubbub2->det0_size + hubbub2->det1_size + hubbub2->det2_size - + hubbub2->det3_size + hubbub2->compbuf_size_segments <= hubbub2->crb_size_segs); + if (hubbub2->det0_size + hubbub2->det1_size + hubbub2->det2_size + + hubbub2->det3_size + hubbub2->compbuf_size_segments > hubbub2->crb_size_segs) { + /* This may happen during seamless transition from ODM 2:1 to ODM4:1 */ + DC_LOG_WARNING("CRB Config Warning: DET size (%d,%d,%d,%d) + Compbuf size (%d) > CRB segments (%d)\n", + hubbub2->det0_size, hubbub2->det1_size, hubbub2->det2_size, hubbub2->det3_size, + hubbub2->compbuf_size_segments, hubbub2->crb_size_segs); + } } static void dcn32_program_compbuf_size(struct hubbub *hubbub, unsigned int compbuf_size_kb, bool safe_to_increase) @@ -932,6 +936,7 @@ static const struct hubbub_funcs hubbub32_funcs = { .program_watermarks = hubbub32_program_watermarks, .allow_self_refresh_control = hubbub1_allow_self_refresh_control, .is_allow_self_refresh_enabled = hubbub1_is_allow_self_refresh_enabled, + .verify_allow_pstate_change_high = hubbub1_verify_allow_pstate_change_high, .force_wm_propagate_to_pipes = hubbub32_force_wm_propagate_to_pipes, .force_pstate_change_control = hubbub3_force_pstate_change_control, .init_watermarks = hubbub32_init_watermarks, diff --git a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_hubp.c b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_hubp.c index 2038cbda33f7..ac1c6458dd55 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_hubp.c +++ b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_hubp.c @@ -79,6 +79,8 @@ void hubp32_phantom_hubp_post_enable(struct hubp *hubp) uint32_t reg_val; struct dcn20_hubp *hubp2 = TO_DCN20_HUBP(hubp); + /* For phantom pipe enable, disable GSL */ + REG_UPDATE(DCSURF_FLIP_CONTROL2, SURFACE_GSL_ENABLE, 0); REG_UPDATE(DCHUBP_CNTL, HUBP_BLANK_EN, 1); reg_val = REG_READ(DCHUBP_CNTL); if (reg_val) { @@ -179,12 +181,12 @@ static struct hubp_funcs dcn32_hubp_funcs = { .hubp_init = hubp3_init, .set_unbounded_requesting = hubp31_set_unbounded_requesting, .hubp_soft_reset = hubp31_soft_reset, + .hubp_set_flip_int = hubp1_set_flip_int, .hubp_in_blank = hubp1_in_blank, .hubp_update_force_pstate_disallow = hubp32_update_force_pstate_disallow, .phantom_hubp_post_enable = hubp32_phantom_hubp_post_enable, .hubp_update_mall_sel = hubp32_update_mall_sel, - .hubp_prepare_subvp_buffering = hubp32_prepare_subvp_buffering, - .hubp_set_flip_int = hubp1_set_flip_int + .hubp_prepare_subvp_buffering = hubp32_prepare_subvp_buffering }; bool hubp32_construct( diff --git a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_hwseq.c b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_hwseq.c index 7544ac8e33ac..cf5bd9713f54 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_hwseq.c @@ -49,6 +49,7 @@ #include "dcn20/dcn20_optc.h" #include "dmub_subvp_state.h" #include "dce/dmub_hw_lock_mgr.h" +#include "dcn32_resource.h" #include "dc_link_dp.h" #include "dmub/inc/dmub_subvp_state.h" @@ -198,42 +199,6 @@ static bool dcn32_check_no_memory_request_for_cab(struct dc *dc) return false; } -/* This function takes in the start address and surface size to be cached in CAB - * and calculates the total number of cache lines required to store the surface. - * The number of cache lines used for each surface is calculated independently of - * one another. For example, if there is a primary surface(1), meta surface(2), and - * cursor(3), this function should be called 3 times to calculate the number of cache - * lines used for each of those surfaces. - */ -static uint32_t dcn32_cache_lines_for_surface(struct dc *dc, uint32_t surface_size, uint64_t start_address) -{ - uint32_t lines_used = 1; - uint32_t num_cached_bytes = 0; - uint32_t remaining_size = 0; - uint32_t cache_line_size = dc->caps.cache_line_size; - uint32_t remainder = 0; - - /* 1. Calculate surface size minus the number of bytes stored - * in the first cache line (all bytes in first cache line might - * not be fully used). - */ - div_u64_rem(start_address, cache_line_size, &remainder); - num_cached_bytes = cache_line_size - remainder; - remaining_size = surface_size - num_cached_bytes; - - /* 2. Calculate number of cache lines that will be fully used with - * the remaining number of bytes to be stored. - */ - lines_used += (remaining_size / cache_line_size); - - /* 3. Check if we need an extra line due to the remaining size not being - * a multiple of CACHE_LINE_SIZE. - */ - if (remaining_size % cache_line_size > 0) - lines_used++; - - return lines_used; -} /* This function loops through every surface that needs to be cached in CAB for SS, * and calculates the total number of ways required to store all surfaces (primary, @@ -241,96 +206,115 @@ static uint32_t dcn32_cache_lines_for_surface(struct dc *dc, uint32_t surface_si */ static uint32_t dcn32_calculate_cab_allocation(struct dc *dc, struct dc_state *ctx) { - uint8_t i, j; + int i, j; struct dc_stream_state *stream = NULL; struct dc_plane_state *plane = NULL; - uint32_t surface_size = 0; uint32_t cursor_size = 0; - uint32_t cache_lines_used = 0; uint32_t total_lines = 0; uint32_t lines_per_way = 0; - uint32_t num_ways = 0; - uint32_t prev_addr_low = 0; + uint8_t num_ways = 0; + uint8_t bytes_per_pixel = 0; + uint8_t cursor_bpp = 0; + uint16_t mblk_width = 0; + uint16_t mblk_height = 0; + uint16_t mall_alloc_width_blk_aligned = 0; + uint16_t mall_alloc_height_blk_aligned = 0; + uint16_t num_mblks = 0; + uint32_t bytes_in_mall = 0; + uint32_t cache_lines_used = 0; + uint32_t cache_lines_per_plane = 0; - for (i = 0; i < ctx->stream_count; i++) { - stream = ctx->streams[i]; + for (i = 0; i < dc->res_pool->pipe_count; i++) { + struct pipe_ctx *pipe = &dc->current_state->res_ctx.pipe_ctx[i]; - // Don't include PSR surface in the total surface size for CAB allocation - if (stream->link->psr_settings.psr_version != DC_PSR_VERSION_UNSUPPORTED) + if (!pipe->stream || !pipe->plane_state || + pipe->stream->link->psr_settings.psr_version != DC_PSR_VERSION_UNSUPPORTED || + pipe->stream->mall_stream_config.type == SUBVP_PHANTOM) continue; - if (ctx->stream_status[i].plane_count == 0) - continue; + bytes_per_pixel = pipe->plane_state->format >= SURFACE_PIXEL_FORMAT_GRPH_ARGB16161616 ? 8 : 4; + mblk_width = DCN3_2_MBLK_WIDTH; + mblk_height = bytes_per_pixel == 4 ? DCN3_2_MBLK_HEIGHT_4BPE : DCN3_2_MBLK_HEIGHT_8BPE; - // For each stream, loop through each plane to calculate the number of cache - // lines required to store the surface in CAB - for (j = 0; j < ctx->stream_status[i].plane_count; j++) { - plane = ctx->stream_status[i].plane_states[j]; + /* full_vp_width_blk_aligned = FLOOR(vp_x_start + full_vp_width + blk_width - 1, blk_width) - + * FLOOR(vp_x_start, blk_width) + * + * mall_alloc_width_blk_aligned_l/c = full_vp_width_blk_aligned_l/c + */ + mall_alloc_width_blk_aligned = ((pipe->plane_res.scl_data.viewport.x + + pipe->plane_res.scl_data.viewport.width + mblk_width - 1) / mblk_width * mblk_width) - + (pipe->plane_res.scl_data.viewport.x / mblk_width * mblk_width); + + /* full_vp_height_blk_aligned = FLOOR(vp_y_start + full_vp_height + blk_height - 1, blk_height) - + * FLOOR(vp_y_start, blk_height) + * + * mall_alloc_height_blk_aligned_l/c = full_vp_height_blk_aligned_l/c + */ + mall_alloc_height_blk_aligned = ((pipe->plane_res.scl_data.viewport.y + + pipe->plane_res.scl_data.viewport.height + mblk_height - 1) / mblk_height * mblk_height) - + (pipe->plane_res.scl_data.viewport.y / mblk_height * mblk_height); - // Calculate total surface size - if (prev_addr_low != plane->address.grph.addr.u.low_part) { - /* if plane address are different from prev FB, then userspace allocated separate FBs*/ - surface_size += plane->plane_size.surface_pitch * - plane->plane_size.surface_size.height * - (plane->format >= SURFACE_PIXEL_FORMAT_GRPH_ARGB16161616 ? 8 : 4); + num_mblks = ((mall_alloc_width_blk_aligned + mblk_width - 1) / mblk_width) * + ((mall_alloc_height_blk_aligned + mblk_height - 1) / mblk_height); - prev_addr_low = plane->address.grph.addr.u.low_part; - } else { - /* We have the same fb for all the planes. - * Xorg always creates one giant fb that holds all surfaces, - * so allocating it once is sufficient. - * */ - continue; - } - // Convert surface size + starting address to number of cache lines required - // (alignment accounted for) - cache_lines_used += dcn32_cache_lines_for_surface(dc, surface_size, - plane->address.grph.addr.quad_part); - - if (plane->address.grph.meta_addr.quad_part) { - // Meta surface - cache_lines_used += dcn32_cache_lines_for_surface(dc, surface_size, - plane->address.grph.meta_addr.quad_part); - } - } + /* For DCC: + * meta_num_mblk = CEILING(full_mblk_width_ub_l*full_mblk_height_ub_l*Bpe/256/mblk_bytes, 1) + */ + if (pipe->plane_state->dcc.enable) + num_mblks += (mall_alloc_width_blk_aligned * mall_alloc_width_blk_aligned * bytes_per_pixel + + (256 * DCN3_2_MALL_MBLK_SIZE_BYTES) - 1) / (256 * DCN3_2_MALL_MBLK_SIZE_BYTES); - // Include cursor size for CAB allocation - for (j = 0; j < dc->res_pool->pipe_count; j++) { - struct pipe_ctx *pipe = &ctx->res_ctx.pipe_ctx[j]; - struct hubp *hubp = pipe->plane_res.hubp; + bytes_in_mall = num_mblks * DCN3_2_MALL_MBLK_SIZE_BYTES; - if (pipe->stream && pipe->plane_state && hubp) - /* Find the cursor plane and use the exact size instead of - * using the max for calculation - */ - if (hubp->curs_attr.width > 0) { - // Round cursor width to next multiple of 64 - cursor_size = (((hubp->curs_attr.width + 63) / 64) * 64) * hubp->curs_attr.height; - break; - } - } + /* (cache lines used is total bytes / cache_line size. Add +2 for worst case alignment + * (MALL is 64-byte aligned) + */ + cache_lines_per_plane = bytes_in_mall / dc->caps.cache_line_size + 2; + cache_lines_used += cache_lines_per_plane; + } - switch (stream->cursor_attributes.color_format) { - case CURSOR_MODE_MONO: - cursor_size /= 2; - break; - case CURSOR_MODE_COLOR_1BIT_AND: - case CURSOR_MODE_COLOR_PRE_MULTIPLIED_ALPHA: - case CURSOR_MODE_COLOR_UN_PRE_MULTIPLIED_ALPHA: - cursor_size *= 4; - break; + // Include cursor size for CAB allocation + for (j = 0; j < dc->res_pool->pipe_count; j++) { + struct pipe_ctx *pipe = &ctx->res_ctx.pipe_ctx[j]; + struct hubp *hubp = pipe->plane_res.hubp; - case CURSOR_MODE_COLOR_64BIT_FP_PRE_MULTIPLIED: - case CURSOR_MODE_COLOR_64BIT_FP_UN_PRE_MULTIPLIED: - cursor_size *= 8; - break; - } + if (pipe->stream && pipe->plane_state && hubp) + /* Find the cursor plane and use the exact size instead of + using the max for calculation */ - if (stream->cursor_position.enable && !dc->debug.alloc_extra_way_for_cursor && - cursor_size > 16384) { - cache_lines_used += dcn32_cache_lines_for_surface(dc, cursor_size, - plane->address.grph.cursor_cache_addr.quad_part); - } + if (hubp->curs_attr.width > 0) { + // Round cursor width to next multiple of 64 + cursor_size = (((hubp->curs_attr.width + 63) / 64) * 64) * hubp->curs_attr.height; + + switch (pipe->stream->cursor_attributes.color_format) { + case CURSOR_MODE_MONO: + cursor_size /= 2; + cursor_bpp = 4; + break; + case CURSOR_MODE_COLOR_1BIT_AND: + case CURSOR_MODE_COLOR_PRE_MULTIPLIED_ALPHA: + case CURSOR_MODE_COLOR_UN_PRE_MULTIPLIED_ALPHA: + cursor_size *= 4; + cursor_bpp = 4; + break; + + case CURSOR_MODE_COLOR_64BIT_FP_PRE_MULTIPLIED: + case CURSOR_MODE_COLOR_64BIT_FP_UN_PRE_MULTIPLIED: + cursor_size *= 8; + cursor_bpp = 8; + break; + } + + if (pipe->stream->cursor_position.enable && !dc->debug.alloc_extra_way_for_cursor && + cursor_size > 16384) { + /* cursor_num_mblk = CEILING(num_cursors*cursor_width*cursor_width*cursor_Bpe/mblk_bytes, 1) + */ + cache_lines_used += (((hubp->curs_attr.width * hubp->curs_attr.height * cursor_bpp + + DCN3_2_MALL_MBLK_SIZE_BYTES - 1) / DCN3_2_MALL_MBLK_SIZE_BYTES) * + DCN3_2_MALL_MBLK_SIZE_BYTES) / dc->caps.cache_line_size + 2; + } + break; + } } // Convert number of cache lines required to number of ways @@ -360,7 +344,9 @@ static uint32_t dcn32_calculate_cab_allocation(struct dc *dc, struct dc_state *c } } } - + if (dc->debug.force_mall_ss_num_ways > 0) { + num_ways = dc->debug.force_mall_ss_num_ways; + } return num_ways; } @@ -369,7 +355,7 @@ bool dcn32_apply_idle_power_optimizations(struct dc *dc, bool enable) union dmub_rb_cmd cmd; uint8_t ways, i; int j; - bool stereo_in_use = false; + bool mall_ss_unsupported = false; struct dc_plane_state *plane = NULL; if (!dc->ctx->dmub_srv) @@ -400,22 +386,23 @@ bool dcn32_apply_idle_power_optimizations(struct dc *dc, bool enable) */ ways = dcn32_calculate_cab_allocation(dc, dc->current_state); - /* MALL not supported with Stereo3D. If any plane is using stereo, - * don't try to enter MALL. + /* MALL not supported with Stereo3D or TMZ surface. If any plane is using stereo, + * or TMZ surface, don't try to enter MALL. */ for (i = 0; i < dc->current_state->stream_count; i++) { for (j = 0; j < dc->current_state->stream_status[i].plane_count; j++) { plane = dc->current_state->stream_status[i].plane_states[j]; - if (plane->address.type == PLN_ADDR_TYPE_GRPH_STEREO) { - stereo_in_use = true; + if (plane->address.type == PLN_ADDR_TYPE_GRPH_STEREO || + plane->address.tmz_surface) { + mall_ss_unsupported = true; break; } } - if (stereo_in_use) + if (mall_ss_unsupported) break; } - if (ways <= dc->caps.cache_num_ways && !stereo_in_use) { + if (ways <= dc->caps.cache_num_ways && !mall_ss_unsupported) { memset(&cmd, 0, sizeof(cmd)); cmd.cab.header.type = DMUB_CMD__CAB_FOR_SS; cmd.cab.header.sub_type = DMUB_CMD__CAB_DCN_SS_FIT_IN_CAB; @@ -642,10 +629,9 @@ bool dcn32_set_input_transfer_func(struct dc *dc, params = &dpp_base->degamma_params; } - result = dpp_base->funcs->dpp_program_gamcor_lut(dpp_base, params); + dpp_base->funcs->dpp_program_gamcor_lut(dpp_base, params); - if (result && - pipe_ctx->stream_res.opp && + if (pipe_ctx->stream_res.opp && pipe_ctx->stream_res.opp->ctx && hws->funcs.set_mcm_luts) result = hws->funcs.set_mcm_luts(pipe_ctx, plane_state); @@ -773,7 +759,8 @@ void dcn32_update_mall_sel(struct dc *dc, struct dc_state *context) hubp->funcs->hubp_update_mall_sel(hubp, num_ways <= dc->caps.cache_num_ways && pipe->stream->link->psr_settings.psr_version == DC_PSR_VERSION_UNSUPPORTED && - pipe->plane_state->address.type != PLN_ADDR_TYPE_GRPH_STEREO ? 2 : 0, + pipe->plane_state->address.type != PLN_ADDR_TYPE_GRPH_STEREO && + !pipe->plane_state->address.tmz_surface ? 2 : 0, cache_cursor); } } @@ -1002,6 +989,10 @@ void dcn32_init_hw(struct dc *dc) dc_dmub_srv_query_caps_cmd(dc->ctx->dmub_srv->dmub); dc->caps.dmub_caps.psr = dc->ctx->dmub_srv->dmub->feature_caps.psr; } + + /* Enable support for ODM and windowed MPO if policy flag is set */ + if (dc->debug.enable_single_display_2to1_odm_policy) + dc->config.enable_windowed_mpo_odm = true; } static int calc_mpc_flow_ctrl_cnt(const struct dc_stream_state *stream, @@ -1156,23 +1147,25 @@ void dcn32_update_odm(struct dc *dc, struct dc_state *context, struct pipe_ctx * true); } - // Don't program pixel clock after link is already enabled -/* if (false == pipe_ctx->clock_source->funcs->program_pix_clk( - pipe_ctx->clock_source, - &pipe_ctx->stream_res.pix_clk_params, - &pipe_ctx->pll_settings)) { - BREAK_TO_DEBUGGER(); - }*/ + if (pipe_ctx->stream_res.dsc) { + struct pipe_ctx *current_pipe_ctx = &dc->current_state->res_ctx.pipe_ctx[pipe_ctx->pipe_idx]; - if (pipe_ctx->stream_res.dsc) update_dsc_on_stream(pipe_ctx, pipe_ctx->stream->timing.flags.DSC); + + /* Check if no longer using pipe for ODM, then need to disconnect DSC for that pipe */ + if (!pipe_ctx->next_odm_pipe && current_pipe_ctx->next_odm_pipe && + current_pipe_ctx->next_odm_pipe->stream_res.dsc) { + struct display_stream_compressor *dsc = current_pipe_ctx->next_odm_pipe->stream_res.dsc; + /* disconnect DSC block from stream */ + dsc->funcs->dsc_disconnect(dsc); + } + } } unsigned int dcn32_calculate_dccg_k1_k2_values(struct pipe_ctx *pipe_ctx, unsigned int *k1_div, unsigned int *k2_div) { struct dc_stream_state *stream = pipe_ctx->stream; unsigned int odm_combine_factor = 0; - struct dc *dc = pipe_ctx->stream->ctx->dc; bool two_pix_per_container = false; // For phantom pipes, use the same programming as the main pipes @@ -1200,7 +1193,7 @@ unsigned int dcn32_calculate_dccg_k1_k2_values(struct pipe_ctx *pipe_ctx, unsign } else { *k1_div = PIXEL_RATE_DIV_BY_1; *k2_div = PIXEL_RATE_DIV_BY_4; - if ((odm_combine_factor == 2) || dc->debug.enable_dp_dig_pixel_rate_div_policy) + if ((odm_combine_factor == 2) || dcn32_is_dp_dig_pixel_rate_div_policy(pipe_ctx)) *k2_div = PIXEL_RATE_DIV_BY_2; } } @@ -1237,7 +1230,6 @@ void dcn32_unblank_stream(struct pipe_ctx *pipe_ctx, struct dc_link *link = stream->link; struct dce_hwseq *hws = link->dc->hwseq; struct pipe_ctx *odm_pipe; - struct dc *dc = pipe_ctx->stream->ctx->dc; uint32_t pix_per_cycle = 1; params.opp_cnt = 1; @@ -1256,7 +1248,7 @@ void dcn32_unblank_stream(struct pipe_ctx *pipe_ctx, pipe_ctx->stream_res.tg->inst); } else if (dc_is_dp_signal(pipe_ctx->stream->signal)) { if (optc2_is_two_pixels_per_containter(&stream->timing) || params.opp_cnt > 1 - || dc->debug.enable_dp_dig_pixel_rate_div_policy) { + || dcn32_is_dp_dig_pixel_rate_div_policy(pipe_ctx)) { params.timing.pix_clk_100hz /= 2; pix_per_cycle = 2; } @@ -1273,6 +1265,9 @@ bool dcn32_is_dp_dig_pixel_rate_div_policy(struct pipe_ctx *pipe_ctx) { struct dc *dc = pipe_ctx->stream->ctx->dc; + if (!is_h_timing_divisible_by_2(pipe_ctx->stream)) + return false; + if (dc_is_dp_signal(pipe_ctx->stream->signal) && !is_dp_128b_132b_signal(pipe_ctx) && dc->debug.enable_dp_dig_pixel_rate_div_policy) return true; @@ -1303,7 +1298,7 @@ static void apply_symclk_on_tx_off_wa(struct dc_link *link) if (link->phy_state.symclk_ref_cnts.otg > 0) { for (i = 0; i < MAX_PIPES; i++) { pipe_ctx = &dc->current_state->res_ctx.pipe_ctx[i]; - if (pipe_ctx->stream->link == link && pipe_ctx->top_pipe == NULL) { + if (pipe_ctx->stream && pipe_ctx->stream->link == link && pipe_ctx->top_pipe == NULL) { pipe_ctx->clock_source->funcs->program_pix_clk( pipe_ctx->clock_source, &pipe_ctx->stream_res.pix_clk_params, @@ -1375,3 +1370,58 @@ void dcn32_update_phantom_vp_position(struct dc *dc, } } } + +bool dcn32_dsc_pg_status( + struct dce_hwseq *hws, + unsigned int dsc_inst) +{ + uint32_t pwr_status = 0; + + switch (dsc_inst) { + case 0: /* DSC0 */ + REG_GET(DOMAIN16_PG_STATUS, + DOMAIN_PGFSM_PWR_STATUS, &pwr_status); + break; + case 1: /* DSC1 */ + + REG_GET(DOMAIN17_PG_STATUS, + DOMAIN_PGFSM_PWR_STATUS, &pwr_status); + break; + case 2: /* DSC2 */ + REG_GET(DOMAIN18_PG_STATUS, + DOMAIN_PGFSM_PWR_STATUS, &pwr_status); + break; + case 3: /* DSC3 */ + REG_GET(DOMAIN19_PG_STATUS, + DOMAIN_PGFSM_PWR_STATUS, &pwr_status); + break; + default: + BREAK_TO_DEBUGGER(); + break; + } + + return pwr_status == 0; +} + +void dcn32_update_dsc_pg(struct dc *dc, + struct dc_state *context, + bool safe_to_disable) +{ + struct dce_hwseq *hws = dc->hwseq; + int i; + + for (i = 0; i < dc->res_pool->res_cap->num_dsc; i++) { + struct display_stream_compressor *dsc = dc->res_pool->dscs[i]; + bool is_dsc_ungated = hws->funcs.dsc_pg_status(hws, dsc->inst); + + if (context->res_ctx.is_dsc_acquired[i]) { + if (!is_dsc_ungated) { + hws->funcs.dsc_pg_control(hws, dsc->inst, true); + } + } else if (safe_to_disable) { + if (is_dsc_ungated) { + hws->funcs.dsc_pg_control(hws, dsc->inst, false); + } + } + } +} diff --git a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_hwseq.h b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_hwseq.h index 86c018b2f132..ac3657a5b9ea 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_hwseq.h +++ b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_hwseq.h @@ -92,4 +92,12 @@ void dcn32_update_phantom_vp_position(struct dc *dc, struct dc_state *context, struct pipe_ctx *phantom_pipe); +bool dcn32_dsc_pg_status( + struct dce_hwseq *hws, + unsigned int dsc_inst); + +void dcn32_update_dsc_pg(struct dc *dc, + struct dc_state *context, + bool safe_to_disable); + #endif /* __DC_HWSS_DCN32_H__ */ diff --git a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_init.c b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_init.c index c66dfe14909a..45a949ba6f3f 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_init.c +++ b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_init.c @@ -109,6 +109,7 @@ static const struct hw_sequencer_funcs dcn32_funcs = { .subvp_pipe_control_lock = dcn32_subvp_pipe_control_lock, .update_visual_confirm_color = dcn20_update_visual_confirm_color, .update_phantom_vp_position = dcn32_update_phantom_vp_position, + .update_dsc_pg = dcn32_update_dsc_pg, }; static const struct hwseq_private_funcs dcn32_private_funcs = { @@ -138,6 +139,7 @@ static const struct hwseq_private_funcs dcn32_private_funcs = { .program_all_writeback_pipes_in_tree = dcn30_program_all_writeback_pipes_in_tree, .update_odm = dcn32_update_odm, .dsc_pg_control = dcn32_dsc_pg_control, + .dsc_pg_status = dcn32_dsc_pg_status, .set_hdr_multiplier = dcn10_set_hdr_multiplier, .verify_allow_pstate_change_high = dcn10_verify_allow_pstate_change_high, .wait_for_blank_complete = dcn20_wait_for_blank_complete, diff --git a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_optc.c b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_optc.c index ec3989d37086..2b33eeb213e2 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_optc.c +++ b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_optc.c @@ -151,7 +151,7 @@ static bool optc32_disable_crtc(struct timing_generator *optc) /* CRTC disabled, so disable clock. */ REG_WAIT(OTG_CLOCK_CONTROL, OTG_BUSY, 0, - 1, 100000); + 1, 150000); return true; } diff --git a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_resource.c b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_resource.c index f8c6afacf633..a88dd7b3d1c1 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_resource.c @@ -718,6 +718,9 @@ static const struct dc_debug_options debug_defaults_drv = { .force_disable_subvp = false, .exit_idle_opt_for_cursor_updates = true, .enable_single_display_2to1_odm_policy = true, + + /* Must match enable_single_display_2to1_odm_policy to support dynamic ODM transitions*/ + .enable_double_buffered_dsc_pg_support = true, .enable_dp_dig_pixel_rate_div_policy = 1, .allow_sw_cursor_fallback = false, .alloc_extra_way_for_cursor = true, @@ -1677,6 +1680,8 @@ static void dcn32_enable_phantom_plane(struct dc *dc, phantom_plane->clip_rect.y = 0; phantom_plane->clip_rect.height = phantom_stream->timing.v_addressable; + phantom_plane->is_phantom = true; + dc_add_plane_to_context(dc, phantom_stream, phantom_plane, context); curr_pipe = curr_pipe->bottom_pipe; @@ -1746,6 +1751,10 @@ bool dcn32_remove_phantom_pipes(struct dc *dc, struct dc_state *context) pipe->stream->mall_stream_config.type = SUBVP_NONE; pipe->stream->mall_stream_config.paired_stream = NULL; } + + if (pipe->plane_state) { + pipe->plane_state->is_phantom = false; + } } return removed_pipe; } @@ -1795,14 +1804,39 @@ bool dcn32_validate_bandwidth(struct dc *dc, int vlevel = 0; int pipe_cnt = 0; display_e2e_pipe_params_st *pipes = kzalloc(dc->res_pool->pipe_count * sizeof(display_e2e_pipe_params_st), GFP_KERNEL); + struct mall_temp_config mall_temp_config; + + /* To handle Freesync properly, setting FreeSync DML parameters + * to its default state for the first stage of validation + */ + context->bw_ctx.bw.dcn.clk.fw_based_mclk_switching = false; + context->bw_ctx.dml.soc.dram_clock_change_requirement_final = true; + DC_LOGGER_INIT(dc->ctx->logger); + /* For fast validation, there are situations where a shallow copy of + * of the dc->current_state is created for the validation. In this case + * we want to save and restore the mall config because we always + * teardown subvp at the beginning of validation (and don't attempt + * to add it back if it's fast validation). If we don't restore the + * subvp config in cases of fast validation + shallow copy of the + * dc->current_state, the dc->current_state will have a partially + * removed subvp state when we did not intend to remove it. + */ + if (fast_validate) { + memset(&mall_temp_config, 0, sizeof(mall_temp_config)); + dcn32_save_mall_state(dc, context, &mall_temp_config); + } + BW_VAL_TRACE_COUNT(); DC_FP_START(); out = dcn32_internal_validate_bw(dc, context, pipes, &pipe_cnt, &vlevel, fast_validate); DC_FP_END(); + if (fast_validate) + dcn32_restore_mall_state(dc, context, &mall_temp_config); + if (pipe_cnt == 0) goto validate_out; @@ -1846,7 +1880,7 @@ int dcn32_populate_dml_pipes_from_context( struct resource_context *res_ctx = &context->res_ctx; struct pipe_ctx *pipe; bool subvp_in_use = false; - int plane_count = 0; + uint8_t is_pipe_split_expected[MAX_PIPES] = {0}; struct dc_crtc_timing *timing; dcn20_populate_dml_pipes_from_context(dc, context, pipes, fast_validate); @@ -1865,12 +1899,13 @@ int dcn32_populate_dml_pipes_from_context( timing = &pipe->stream->timing; pipes[pipe_cnt].pipe.dest.odm_combine_policy = dm_odm_combine_policy_dal; - if (context->stream_count == 1 && !dc_is_hdmi_signal(res_ctx->pipe_ctx[i].stream->signal) && - is_h_timing_divisible_by_2(res_ctx->pipe_ctx[i].stream)) { - if (dc->debug.enable_single_display_2to1_odm_policy) { - if (!((plane_count > 2) && pipe->top_pipe)) - pipes[pipe_cnt].pipe.dest.odm_combine_policy = dm_odm_combine_policy_2to1; - } + if (context->stream_count == 1 && + context->stream_status[0].plane_count <= 1 && + !dc_is_hdmi_signal(res_ctx->pipe_ctx[i].stream->signal) && + is_h_timing_divisible_by_2(res_ctx->pipe_ctx[i].stream) && + pipe->stream->timing.pix_clk_100hz * 100 > DCN3_2_VMIN_DISPCLK_HZ && + dc->debug.enable_single_display_2to1_odm_policy) { + pipes[pipe_cnt].pipe.dest.odm_combine_policy = dm_odm_combine_policy_2to1; } pipe_cnt++; } @@ -1927,12 +1962,10 @@ int dcn32_populate_dml_pipes_from_context( } } - /* Calculate the number of planes we have so we can determine - * whether to apply ODM 2to1 policy or not - */ - if (pipe->stream && !pipe->prev_odm_pipe && - (!pipe->top_pipe || pipe->top_pipe->plane_state != pipe->plane_state)) - ++plane_count; + DC_FP_START(); + is_pipe_split_expected[i] = dcn32_predict_pipe_split(context, &pipes[pipe_cnt]); + DC_FP_END(); + pipe_cnt++; } diff --git a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_resource.h b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_resource.h index 40e9211eec1e..f76120e67c16 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_resource.h +++ b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_resource.h @@ -37,6 +37,7 @@ #define DCN3_2_MBLK_WIDTH 128 #define DCN3_2_MBLK_HEIGHT_4BPE 128 #define DCN3_2_MBLK_HEIGHT_8BPE 64 +#define DCN3_2_VMIN_DISPCLK_HZ 717000000 #define TO_DCN32_RES_POOL(pool)\ container_of(pool, struct dcn32_resource_pool, base) @@ -44,6 +45,17 @@ extern struct _vcs_dpi_ip_params_st dcn3_2_ip; extern struct _vcs_dpi_soc_bounding_box_st dcn3_2_soc; +/* Temp struct used to save and restore MALL config + * during validation. + * + * TODO: Move MALL config into dc_state instead of stream struct + * to avoid needing to save/restore. + */ +struct mall_temp_config { + struct mall_stream_config mall_stream_config[MAX_PIPES]; + bool is_phantom_plane[MAX_PIPES]; +}; + struct dcn32_resource_pool { struct resource_pool base; }; @@ -107,6 +119,8 @@ bool dcn32_subvp_in_use(struct dc *dc, bool dcn32_mpo_in_use(struct dc_state *context); +bool dcn32_any_surfaces_rotated(struct dc *dc, struct dc_state *context); + struct pipe_ctx *dcn32_acquire_idle_pipe_for_head_pipe_in_layer( struct dc_state *state, const struct resource_pool *pool, @@ -119,6 +133,15 @@ void dcn32_determine_det_override(struct dc *dc, void dcn32_set_det_allocations(struct dc *dc, struct dc_state *context, display_e2e_pipe_params_st *pipes); + +void dcn32_save_mall_state(struct dc *dc, + struct dc_state *context, + struct mall_temp_config *temp_config); + +void dcn32_restore_mall_state(struct dc *dc, + struct dc_state *context, + struct mall_temp_config *temp_config); + /* definitions for run time init of reg offsets */ /* CLK SRC */ diff --git a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_resource_helpers.c b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_resource_helpers.c index 46ba6eee69ea..d51d0c40ae5b 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_resource_helpers.c +++ b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_resource_helpers.c @@ -233,6 +233,23 @@ bool dcn32_mpo_in_use(struct dc_state *context) return false; } + +bool dcn32_any_surfaces_rotated(struct dc *dc, struct dc_state *context) +{ + uint32_t i; + + for (i = 0; i < dc->res_pool->pipe_count; i++) { + struct pipe_ctx *pipe = &context->res_ctx.pipe_ctx[i]; + + if (!pipe->stream) + continue; + + if (pipe->plane_state && pipe->plane_state->rotation != ROTATION_ANGLE_0) + return true; + } + return false; +} + /** * ******************************************************************************************* * dcn32_determine_det_override: Determine DET allocation for each pipe @@ -278,7 +295,7 @@ void dcn32_determine_det_override(struct dc *dc, } } - if (context->stream_count > 0) { + if (stream_count > 0) { stream_segments = 18 / stream_count; for (i = 0; i < context->stream_count; i++) { if (context->streams[i]->mall_stream_config.type == SUBVP_PHANTOM) @@ -363,3 +380,74 @@ void dcn32_set_det_allocations(struct dc *dc, struct dc_state *context, } else dcn32_determine_det_override(dc, context, pipes); } + +/** + * ******************************************************************************************* + * dcn32_save_mall_state: Save MALL (SubVP) state for fast validation cases + * + * This function saves the MALL (SubVP) case for fast validation cases. For fast validation, + * there are situations where a shallow copy of the dc->current_state is created for the + * validation. In this case we want to save and restore the mall config because we always + * teardown subvp at the beginning of validation (and don't attempt to add it back if it's + * fast validation). If we don't restore the subvp config in cases of fast validation + + * shallow copy of the dc->current_state, the dc->current_state will have a partially + * removed subvp state when we did not intend to remove it. + * + * NOTE: This function ONLY works if the streams are not moved to a different pipe in the + * validation. We don't expect this to happen in fast_validation=1 cases. + * + * @param [in]: dc: Current DC state + * @param [in]: context: New DC state to be programmed + * @param [out]: temp_config: struct used to cache the existing MALL state + * + * @return: void + * + * ******************************************************************************************* + */ +void dcn32_save_mall_state(struct dc *dc, + struct dc_state *context, + struct mall_temp_config *temp_config) +{ + uint32_t i; + + for (i = 0; i < dc->res_pool->pipe_count; i++) { + struct pipe_ctx *pipe = &context->res_ctx.pipe_ctx[i]; + + if (pipe->stream) + temp_config->mall_stream_config[i] = pipe->stream->mall_stream_config; + + if (pipe->plane_state) + temp_config->is_phantom_plane[i] = pipe->plane_state->is_phantom; + } +} + +/** + * ******************************************************************************************* + * dcn32_restore_mall_state: Restore MALL (SubVP) state for fast validation cases + * + * Restore the MALL state based on the previously saved state from dcn32_save_mall_state + * + * @param [in]: dc: Current DC state + * @param [in/out]: context: New DC state to be programmed, restore MALL state into here + * @param [in]: temp_config: struct that has the cached MALL state + * + * @return: void + * + * ******************************************************************************************* + */ +void dcn32_restore_mall_state(struct dc *dc, + struct dc_state *context, + struct mall_temp_config *temp_config) +{ + uint32_t i; + + for (i = 0; i < dc->res_pool->pipe_count; i++) { + struct pipe_ctx *pipe = &context->res_ctx.pipe_ctx[i]; + + if (pipe->stream) + pipe->stream->mall_stream_config = temp_config->mall_stream_config[i]; + + if (pipe->plane_state) + pipe->plane_state->is_phantom = temp_config->is_phantom_plane[i]; + } +} diff --git a/drivers/gpu/drm/amd/display/dc/dcn321/dcn321_dio_link_encoder.c b/drivers/gpu/drm/amd/display/dc/dcn321/dcn321_dio_link_encoder.c index 49682a31ecbd..fa9b6603cfd3 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn321/dcn321_dio_link_encoder.c +++ b/drivers/gpu/drm/amd/display/dc/dcn321/dcn321_dio_link_encoder.c @@ -91,7 +91,6 @@ static const struct link_encoder_funcs dcn321_link_enc_funcs = { .is_in_alt_mode = dcn20_link_encoder_is_in_alt_mode, .get_max_link_cap = dcn20_link_encoder_get_max_link_cap, .set_dio_phy_mux = dcn31_link_encoder_set_dio_phy_mux, - .set_dig_output_mode = enc32_set_dig_output_mode, }; void dcn321_link_encoder_construct( diff --git a/drivers/gpu/drm/amd/display/dc/dcn321/dcn321_resource.c b/drivers/gpu/drm/amd/display/dc/dcn321/dcn321_resource.c index c772ef962194..61087f2385a9 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn321/dcn321_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn321/dcn321_resource.c @@ -94,8 +94,6 @@ #include "dcn20/dcn20_vmid.h" #define DC_LOGGER_INIT(logger) -#define fixed16_to_double(x) (((double)x) / ((double) (1 << 16))) -#define fixed16_to_double_to_cpu(x) fixed16_to_double(le32_to_cpu(x)) enum dcn321_clk_src_array_id { DCN321_CLK_SRC_PLL0, @@ -718,6 +716,9 @@ static const struct dc_debug_options debug_defaults_drv = { .force_disable_subvp = false, .exit_idle_opt_for_cursor_updates = true, .enable_single_display_2to1_odm_policy = true, + + /*must match enable_single_display_2to1_odm_policy to support dynamic ODM transitions*/ + .enable_double_buffered_dsc_pg_support = true, .enable_dp_dig_pixel_rate_div_policy = 1, .allow_sw_cursor_fallback = false, .alloc_extra_way_for_cursor = true, @@ -1603,7 +1604,7 @@ static struct resource_funcs dcn321_res_pool_funcs = { .validate_bandwidth = dcn32_validate_bandwidth, .calculate_wm_and_dlg = dcn32_calculate_wm_and_dlg, .populate_dml_pipes = dcn32_populate_dml_pipes_from_context, - .acquire_idle_pipe_for_layer = dcn20_acquire_idle_pipe_for_layer, + .acquire_idle_pipe_for_head_pipe_in_layer = dcn32_acquire_idle_pipe_for_head_pipe_in_layer, .add_stream_to_ctx = dcn30_add_stream_to_ctx, .add_dsc_to_stream_resource = dcn20_add_dsc_to_stream_resource, .remove_stream_from_ctx = dcn20_remove_stream_from_ctx, @@ -1653,7 +1654,7 @@ static bool dcn321_resource_construct( #undef REG_STRUCT #define REG_STRUCT dccg_regs - dccg_regs_init(); + dccg_regs_init(); ctx->dc_bios->regs = &bios_regs; diff --git a/drivers/gpu/drm/amd/display/dc/dml/Makefile b/drivers/gpu/drm/amd/display/dc/dml/Makefile index cb81ed2fbd53..ca7d24000621 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/Makefile +++ b/drivers/gpu/drm/amd/display/dc/dml/Makefile @@ -34,7 +34,7 @@ dml_ccflags := -mhard-float -maltivec endif ifdef CONFIG_CC_IS_GCC -ifeq ($(call cc-ifversion, -lt, 0701, y), y) +ifneq ($(call gcc-min-version, 70100),y) IS_OLD_GCC = 1 endif endif @@ -77,7 +77,7 @@ CFLAGS_$(AMDDALPATH)/dc/dml/dcn30/dcn30_fpu.o := $(dml_ccflags) CFLAGS_$(AMDDALPATH)/dc/dml/dcn32/dcn32_fpu.o := $(dml_ccflags) CFLAGS_$(AMDDALPATH)/dc/dml/dcn32/display_mode_vba_32.o := $(dml_ccflags) $(frame_warn_flag) CFLAGS_$(AMDDALPATH)/dc/dml/dcn32/display_rq_dlg_calc_32.o := $(dml_ccflags) -CFLAGS_$(AMDDALPATH)/dc/dml/dcn32/display_mode_vba_util_32.o := $(dml_ccflags) +CFLAGS_$(AMDDALPATH)/dc/dml/dcn32/display_mode_vba_util_32.o := $(dml_ccflags) $(frame_warn_flag) CFLAGS_$(AMDDALPATH)/dc/dml/dcn321/dcn321_fpu.o := $(dml_ccflags) CFLAGS_$(AMDDALPATH)/dc/dml/dcn31/dcn31_fpu.o := $(dml_ccflags) CFLAGS_$(AMDDALPATH)/dc/dml/dcn301/dcn301_fpu.o := $(dml_ccflags) diff --git a/drivers/gpu/drm/amd/display/dc/dml/calcs/dcn_calcs.c b/drivers/gpu/drm/amd/display/dc/dml/calcs/dcn_calcs.c index d46adc849d2a..e73f089c84bb 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/calcs/dcn_calcs.c +++ b/drivers/gpu/drm/amd/display/dc/dml/calcs/dcn_calcs.c @@ -1444,81 +1444,67 @@ unsigned int dcn_find_dcfclk_suits_all( return dcf_clk; } -static bool verify_clock_values(struct dm_pp_clock_levels_with_voltage *clks) +void dcn_bw_update_from_pplib_fclks( + struct dc *dc, + struct dm_pp_clock_levels_with_voltage *fclks) { - int i; - - if (clks->num_levels == 0) - return false; - - for (i = 0; i < clks->num_levels; i++) - /* Ensure that the result is sane */ - if (clks->data[i].clocks_in_khz == 0) - return false; + unsigned vmin0p65_idx, vmid0p72_idx, vnom0p8_idx, vmax0p9_idx; - return true; + ASSERT(fclks->num_levels); + + vmin0p65_idx = 0; + vmid0p72_idx = fclks->num_levels - + (fclks->num_levels > 2 ? 3 : (fclks->num_levels > 1 ? 2 : 1)); + vnom0p8_idx = fclks->num_levels - (fclks->num_levels > 1 ? 2 : 1); + vmax0p9_idx = fclks->num_levels - 1; + + dc->dcn_soc->fabric_and_dram_bandwidth_vmin0p65 = + 32 * (fclks->data[vmin0p65_idx].clocks_in_khz / 1000.0) / 1000.0; + dc->dcn_soc->fabric_and_dram_bandwidth_vmid0p72 = + dc->dcn_soc->number_of_channels * + (fclks->data[vmid0p72_idx].clocks_in_khz / 1000.0) + * ddr4_dram_factor_single_Channel / 1000.0; + dc->dcn_soc->fabric_and_dram_bandwidth_vnom0p8 = + dc->dcn_soc->number_of_channels * + (fclks->data[vnom0p8_idx].clocks_in_khz / 1000.0) + * ddr4_dram_factor_single_Channel / 1000.0; + dc->dcn_soc->fabric_and_dram_bandwidth_vmax0p9 = + dc->dcn_soc->number_of_channels * + (fclks->data[vmax0p9_idx].clocks_in_khz / 1000.0) + * ddr4_dram_factor_single_Channel / 1000.0; } -void dcn_bw_update_from_pplib(struct dc *dc) +void dcn_bw_update_from_pplib_dcfclks( + struct dc *dc, + struct dm_pp_clock_levels_with_voltage *dcfclks) { - struct dc_context *ctx = dc->ctx; - struct dm_pp_clock_levels_with_voltage fclks = {0}, dcfclks = {0}; - bool res; - unsigned vmin0p65_idx, vmid0p72_idx, vnom0p8_idx, vmax0p9_idx; - - /* TODO: This is not the proper way to obtain fabric_and_dram_bandwidth, should be min(fclk, memclk) */ - res = dm_pp_get_clock_levels_by_type_with_voltage( - ctx, DM_PP_CLOCK_TYPE_FCLK, &fclks); - - if (res) - res = verify_clock_values(&fclks); - - if (res) { - ASSERT(fclks.num_levels); - - vmin0p65_idx = 0; - vmid0p72_idx = fclks.num_levels - - (fclks.num_levels > 2 ? 3 : (fclks.num_levels > 1 ? 2 : 1)); - vnom0p8_idx = fclks.num_levels - (fclks.num_levels > 1 ? 2 : 1); - vmax0p9_idx = fclks.num_levels - 1; - - dc->dcn_soc->fabric_and_dram_bandwidth_vmin0p65 = - 32 * (fclks.data[vmin0p65_idx].clocks_in_khz / 1000.0) / 1000.0; - dc->dcn_soc->fabric_and_dram_bandwidth_vmid0p72 = - dc->dcn_soc->number_of_channels * - (fclks.data[vmid0p72_idx].clocks_in_khz / 1000.0) - * ddr4_dram_factor_single_Channel / 1000.0; - dc->dcn_soc->fabric_and_dram_bandwidth_vnom0p8 = - dc->dcn_soc->number_of_channels * - (fclks.data[vnom0p8_idx].clocks_in_khz / 1000.0) - * ddr4_dram_factor_single_Channel / 1000.0; - dc->dcn_soc->fabric_and_dram_bandwidth_vmax0p9 = - dc->dcn_soc->number_of_channels * - (fclks.data[vmax0p9_idx].clocks_in_khz / 1000.0) - * ddr4_dram_factor_single_Channel / 1000.0; - } else - BREAK_TO_DEBUGGER(); - - res = dm_pp_get_clock_levels_by_type_with_voltage( - ctx, DM_PP_CLOCK_TYPE_DCFCLK, &dcfclks); - - if (res) - res = verify_clock_values(&dcfclks); + if (dcfclks->num_levels >= 3) { + dc->dcn_soc->dcfclkv_min0p65 = dcfclks->data[0].clocks_in_khz / 1000.0; + dc->dcn_soc->dcfclkv_mid0p72 = dcfclks->data[dcfclks->num_levels - 3].clocks_in_khz / 1000.0; + dc->dcn_soc->dcfclkv_nom0p8 = dcfclks->data[dcfclks->num_levels - 2].clocks_in_khz / 1000.0; + dc->dcn_soc->dcfclkv_max0p9 = dcfclks->data[dcfclks->num_levels - 1].clocks_in_khz / 1000.0; + } +} - if (res && dcfclks.num_levels >= 3) { - dc->dcn_soc->dcfclkv_min0p65 = dcfclks.data[0].clocks_in_khz / 1000.0; - dc->dcn_soc->dcfclkv_mid0p72 = dcfclks.data[dcfclks.num_levels - 3].clocks_in_khz / 1000.0; - dc->dcn_soc->dcfclkv_nom0p8 = dcfclks.data[dcfclks.num_levels - 2].clocks_in_khz / 1000.0; - dc->dcn_soc->dcfclkv_max0p9 = dcfclks.data[dcfclks.num_levels - 1].clocks_in_khz / 1000.0; - } else - BREAK_TO_DEBUGGER(); +void dcn_get_soc_clks( + struct dc *dc, + int *min_fclk_khz, + int *min_dcfclk_khz, + int *socclk_khz) +{ + *min_fclk_khz = dc->dcn_soc->fabric_and_dram_bandwidth_vmin0p65 * 1000000 / 32; + *min_dcfclk_khz = dc->dcn_soc->dcfclkv_min0p65 * 1000; + *socclk_khz = dc->dcn_soc->socclk * 1000; } -void dcn_bw_notify_pplib_of_wm_ranges(struct dc *dc) +void dcn_bw_notify_pplib_of_wm_ranges( + struct dc *dc, + int min_fclk_khz, + int min_dcfclk_khz, + int socclk_khz) { struct pp_smu_funcs_rv *pp = NULL; struct pp_smu_wm_range_sets ranges = {0}; - int min_fclk_khz, min_dcfclk_khz, socclk_khz; const int overdrive = 5000000; /* 5 GHz to cover Overdrive */ if (dc->res_pool->pp_smu) @@ -1526,10 +1512,6 @@ void dcn_bw_notify_pplib_of_wm_ranges(struct dc *dc) if (!pp || !pp->set_wm_ranges) return; - min_fclk_khz = dc->dcn_soc->fabric_and_dram_bandwidth_vmin0p65 * 1000000 / 32; - min_dcfclk_khz = dc->dcn_soc->dcfclkv_min0p65 * 1000; - socclk_khz = dc->dcn_soc->socclk * 1000; - /* Now notify PPLib/SMU about which Watermarks sets they should select * depending on DPM state they are in. And update BW MGR GFX Engine and * Memory clock member variables for Watermarks calculations for each diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.c index b6e99eefe869..7dd0845d1bd9 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.c @@ -292,6 +292,7 @@ static struct _vcs_dpi_soc_bounding_box_st dcn3_15_soc = { .urgent_latency_adjustment_fabric_clock_component_us = 0, .urgent_latency_adjustment_fabric_clock_reference_mhz = 0, .num_chans = 4, + .dummy_pstate_latency_us = 10.0 }; struct _vcs_dpi_ip_params_st dcn3_16_ip = { @@ -459,13 +460,30 @@ void dcn31_update_soc_for_wm_a(struct dc *dc, struct dc_state *context) } } +void dcn315_update_soc_for_wm_a(struct dc *dc, struct dc_state *context) +{ + dc_assert_fp_enabled(); + + if (dc->clk_mgr->bw_params->wm_table.entries[WM_A].valid) { + /* For 315 pstate change is only supported if possible in vactive */ + if (context->bw_ctx.dml.vba.DRAMClockChangeSupport[context->bw_ctx.dml.vba.VoltageLevel][context->bw_ctx.dml.vba.maxMpcComb] != dm_dram_clock_change_vactive) + context->bw_ctx.dml.soc.dram_clock_change_latency_us = context->bw_ctx.dml.soc.dummy_pstate_latency_us; + else + context->bw_ctx.dml.soc.dram_clock_change_latency_us = dc->clk_mgr->bw_params->wm_table.entries[WM_A].pstate_latency_us; + context->bw_ctx.dml.soc.sr_enter_plus_exit_time_us = + dc->clk_mgr->bw_params->wm_table.entries[WM_A].sr_enter_plus_exit_time_us; + context->bw_ctx.dml.soc.sr_exit_time_us = + dc->clk_mgr->bw_params->wm_table.entries[WM_A].sr_exit_time_us; + } +} + void dcn31_calculate_wm_and_dlg_fp( struct dc *dc, struct dc_state *context, display_e2e_pipe_params_st *pipes, int pipe_cnt, int vlevel) { - int i, pipe_idx; + int i, pipe_idx, active_dpp_count = 0; double dcfclk = context->bw_ctx.dml.vba.DCFCLKState[vlevel][context->bw_ctx.dml.vba.maxMpcComb]; dc_assert_fp_enabled(); @@ -486,72 +504,6 @@ void dcn31_calculate_wm_and_dlg_fp( pipes[0].clks_cfg.dcfclk_mhz = dcfclk; pipes[0].clks_cfg.socclk_mhz = context->bw_ctx.dml.soc.clock_limits[vlevel].socclk_mhz; -#if 0 // TODO - /* Set B: - * TODO - */ - if (dc->clk_mgr->bw_params->wm_table.nv_entries[WM_B].valid) { - if (vlevel == 0) { - pipes[0].clks_cfg.voltage = 1; - pipes[0].clks_cfg.dcfclk_mhz = context->bw_ctx.dml.soc.clock_limits[0].dcfclk_mhz; - } - context->bw_ctx.dml.soc.dram_clock_change_latency_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_B].dml_input.pstate_latency_us; - context->bw_ctx.dml.soc.sr_enter_plus_exit_time_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_B].dml_input.sr_enter_plus_exit_time_us; - context->bw_ctx.dml.soc.sr_exit_time_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_B].dml_input.sr_exit_time_us; - } - context->bw_ctx.bw.dcn.watermarks.b.urgent_ns = get_wm_urgent(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.b.cstate_pstate.cstate_enter_plus_exit_ns = get_wm_stutter_enter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.b.cstate_pstate.cstate_exit_ns = get_wm_stutter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.b.cstate_pstate.cstate_enter_plus_exit_z8_ns = get_wm_z8_stutter_enter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.b.cstate_pstate.cstate_exit_z8_ns = get_wm_z8_stutter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.b.cstate_pstate.pstate_change_ns = get_wm_dram_clock_change(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.b.pte_meta_urgent_ns = get_wm_memory_trip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.b.frac_urg_bw_nom = get_fraction_of_urgent_bandwidth(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.b.frac_urg_bw_flip = get_fraction_of_urgent_bandwidth_imm_flip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.b.urgent_latency_ns = get_urgent_latency(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - - pipes[0].clks_cfg.voltage = vlevel; - pipes[0].clks_cfg.dcfclk_mhz = dcfclk; - - /* Set C: - * TODO - */ - if (dc->clk_mgr->bw_params->wm_table.nv_entries[WM_C].valid) { - context->bw_ctx.dml.soc.dram_clock_change_latency_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_C].dml_input.pstate_latency_us; - context->bw_ctx.dml.soc.sr_enter_plus_exit_time_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_C].dml_input.sr_enter_plus_exit_time_us; - context->bw_ctx.dml.soc.sr_exit_time_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_C].dml_input.sr_exit_time_us; - } - context->bw_ctx.bw.dcn.watermarks.c.urgent_ns = get_wm_urgent(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.c.cstate_pstate.cstate_enter_plus_exit_ns = get_wm_stutter_enter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.c.cstate_pstate.cstate_exit_ns = get_wm_stutter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.c.cstate_pstate.cstate_enter_plus_exit_z8_ns = get_wm_z8_stutter_enter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.c.cstate_pstate.cstate_exit_z8_ns = get_wm_z8_stutter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.c.cstate_pstate.pstate_change_ns = get_wm_dram_clock_change(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.c.pte_meta_urgent_ns = get_wm_memory_trip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.c.frac_urg_bw_nom = get_fraction_of_urgent_bandwidth(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.c.frac_urg_bw_flip = get_fraction_of_urgent_bandwidth_imm_flip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.c.urgent_latency_ns = get_urgent_latency(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - - /* Set D: - * TODO - */ - if (dc->clk_mgr->bw_params->wm_table.nv_entries[WM_D].valid) { - context->bw_ctx.dml.soc.dram_clock_change_latency_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_D].dml_input.pstate_latency_us; - context->bw_ctx.dml.soc.sr_enter_plus_exit_time_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_D].dml_input.sr_enter_plus_exit_time_us; - context->bw_ctx.dml.soc.sr_exit_time_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_D].dml_input.sr_exit_time_us; - } - context->bw_ctx.bw.dcn.watermarks.d.urgent_ns = get_wm_urgent(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.d.cstate_pstate.cstate_enter_plus_exit_ns = get_wm_stutter_enter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.d.cstate_pstate.cstate_exit_ns = get_wm_stutter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.d.cstate_pstate.pstate_change_ns = get_wm_dram_clock_change(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.d.cstate_pstate.cstate_enter_plus_exit_z8_ns = get_wm_z8_stutter_enter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.d.cstate_pstate.cstate_exit_z8_ns = get_wm_z8_stutter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.d.pte_meta_urgent_ns = get_wm_memory_trip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.d.frac_urg_bw_nom = get_fraction_of_urgent_bandwidth(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.d.frac_urg_bw_flip = get_fraction_of_urgent_bandwidth_imm_flip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.d.urgent_latency_ns = get_urgent_latency(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; -#endif - /* Set A: * All clocks min required * @@ -568,16 +520,17 @@ void dcn31_calculate_wm_and_dlg_fp( context->bw_ctx.bw.dcn.watermarks.a.frac_urg_bw_nom = get_fraction_of_urgent_bandwidth(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; context->bw_ctx.bw.dcn.watermarks.a.frac_urg_bw_flip = get_fraction_of_urgent_bandwidth_imm_flip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; context->bw_ctx.bw.dcn.watermarks.a.urgent_latency_ns = get_urgent_latency(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - /* TODO: remove: */ context->bw_ctx.bw.dcn.watermarks.b = context->bw_ctx.bw.dcn.watermarks.a; context->bw_ctx.bw.dcn.watermarks.c = context->bw_ctx.bw.dcn.watermarks.a; context->bw_ctx.bw.dcn.watermarks.d = context->bw_ctx.bw.dcn.watermarks.a; - /* end remove*/ for (i = 0, pipe_idx = 0; i < dc->res_pool->pipe_count; i++) { if (!context->res_ctx.pipe_ctx[i].stream) continue; + if (context->res_ctx.pipe_ctx[i].plane_state) + active_dpp_count++; + pipes[pipe_idx].clks_cfg.dispclk_mhz = get_dispclk_calculated(&context->bw_ctx.dml, pipes, pipe_cnt); pipes[pipe_idx].clks_cfg.dppclk_mhz = get_dppclk_calculated(&context->bw_ctx.dml, pipes, pipe_cnt, pipe_idx); @@ -594,6 +547,9 @@ void dcn31_calculate_wm_and_dlg_fp( } dcn20_calculate_dlg_params(dc, context, pipes, pipe_cnt, vlevel); + /* For 31x apu pstate change is only supported if possible in vactive or if there are no active dpps */ + context->bw_ctx.bw.dcn.clk.p_state_change_support = + context->bw_ctx.dml.vba.DRAMClockChangeSupport[vlevel][context->bw_ctx.dml.vba.maxMpcComb] == dm_dram_clock_change_vactive || !active_dpp_count; } void dcn31_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_params) @@ -739,7 +695,7 @@ void dcn315_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_param } if (!IS_FPGA_MAXIMUS_DC(dc->ctx->dce_environment)) - dml_init_instance(&dc->dml, &dcn3_15_soc, &dcn3_15_ip, DML_PROJECT_DCN31); + dml_init_instance(&dc->dml, &dcn3_15_soc, &dcn3_15_ip, DML_PROJECT_DCN315); else dml_init_instance(&dc->dml, &dcn3_15_soc, &dcn3_15_ip, DML_PROJECT_DCN31_FPGA); } diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.h b/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.h index 4372f17b55d4..fd58b2561ec9 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.h +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.h @@ -35,6 +35,7 @@ void dcn31_zero_pipe_dcc_fraction(display_e2e_pipe_params_st *pipes, int pipe_cnt); void dcn31_update_soc_for_wm_a(struct dc *dc, struct dc_state *context); +void dcn315_update_soc_for_wm_a(struct dc *dc, struct dc_state *context); void dcn31_calculate_wm_and_dlg_fp( struct dc *dc, struct dc_state *context, diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn31/display_mode_vba_31.c b/drivers/gpu/drm/amd/display/dc/dml/dcn31/display_mode_vba_31.c index 8dfe639b6508..b612edb14417 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn31/display_mode_vba_31.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn31/display_mode_vba_31.c @@ -43,6 +43,8 @@ #define BPP_BLENDED_PIPE 0xffffffff #define DCN31_MAX_DSC_IMAGE_WIDTH 5184 #define DCN31_MAX_FMT_420_BUFFER_WIDTH 4096 +#define DCN3_15_MIN_COMPBUF_SIZE_KB 128 +#define DCN3_15_MAX_DET_SIZE 384 // For DML-C changes that hasn't been propagated to VBA yet //#define __DML_VBA_ALLOW_DELTA__ @@ -3775,6 +3777,17 @@ static noinline void CalculatePrefetchSchedulePerPlane( &v->VReadyOffsetPix[k]); } +static void PatchDETBufferSizeInKByte(unsigned int NumberOfActivePlanes, int NoOfDPPThisState[], unsigned int config_return_buffer_size_in_kbytes, unsigned int *DETBufferSizeInKByte) +{ + int i, total_pipes = 0; + for (i = 0; i < NumberOfActivePlanes; i++) + total_pipes += NoOfDPPThisState[i]; + *DETBufferSizeInKByte = ((config_return_buffer_size_in_kbytes - DCN3_15_MIN_COMPBUF_SIZE_KB) / 64 / total_pipes) * 64; + if (*DETBufferSizeInKByte > DCN3_15_MAX_DET_SIZE) + *DETBufferSizeInKByte = DCN3_15_MAX_DET_SIZE; +} + + void dml31_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_lib) { struct vba_vars_st *v = &mode_lib->vba; @@ -4533,6 +4546,8 @@ void dml31_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l v->ODMCombineEnableThisState[k] = v->ODMCombineEnablePerState[i][k]; } + if (v->NumberOfActivePlanes > 1 && mode_lib->project == DML_PROJECT_DCN315) + PatchDETBufferSizeInKByte(v->NumberOfActivePlanes, v->NoOfDPPThisState, v->ip.config_return_buffer_size_in_kbytes, &v->DETBufferSizeInKByte[0]); CalculateSwathAndDETConfiguration( false, v->NumberOfActivePlanes, diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c index fe0770038a90..819de0f11012 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c @@ -121,8 +121,8 @@ struct _vcs_dpi_soc_bounding_box_st dcn3_2_soc = { }, }, .num_states = 1, - .sr_exit_time_us = 20.16, - .sr_enter_plus_exit_time_us = 27.13, + .sr_exit_time_us = 42.97, + .sr_enter_plus_exit_time_us = 49.94, .sr_exit_z8_time_us = 285.0, .sr_enter_plus_exit_z8_time_us = 320, .writeback_latency_us = 12.0, @@ -243,7 +243,7 @@ void dcn32_build_wm_range_table_fpu(struct clk_mgr_internal *clk_mgr) clk_mgr->base.bw_params->wm_table.nv_entries[WM_D].pmfw_breakdown.max_uclk = 0xFFFF; } -/** +/* * Finds dummy_latency_index when MCLK switching using firmware based * vblank stretch is enabled. This function will iterate through the * table of dummy pstate latencies until the lowest value that allows @@ -290,15 +290,14 @@ int dcn32_find_dummy_latency_index_for_fw_based_mclk_switch(struct dc *dc, /** * dcn32_helper_populate_phantom_dlg_params - Get DLG params for phantom pipes * and populate pipe_ctx with those params. - * - * This function must be called AFTER the phantom pipes are added to context - * and run through DML (so that the DLG params for the phantom pipes can be - * populated), and BEFORE we program the timing for the phantom pipes. - * * @dc: [in] current dc state * @context: [in] new dc state * @pipes: [in] DML pipe params array * @pipe_cnt: [in] DML pipe count + * + * This function must be called AFTER the phantom pipes are added to context + * and run through DML (so that the DLG params for the phantom pipes can be + * populated), and BEFORE we program the timing for the phantom pipes. */ void dcn32_helper_populate_phantom_dlg_params(struct dc *dc, struct dc_state *context, @@ -331,8 +330,9 @@ void dcn32_helper_populate_phantom_dlg_params(struct dc *dc, } /** - * ******************************************************************************************* - * dcn32_predict_pipe_split: Predict if pipe split will occur for a given DML pipe + * dcn32_predict_pipe_split - Predict if pipe split will occur for a given DML pipe + * @context: [in] New DC state to be programmed + * @pipe_e2e: [in] DML pipe end to end context * * This function takes in a DML pipe (pipe_e2e) and predicts if pipe split is required (both * ODM and MPC). For pipe split, ODM combine is determined by the ODM mode, and MPC combine is @@ -343,12 +343,7 @@ void dcn32_helper_populate_phantom_dlg_params(struct dc *dc, * - MPC combine is only chosen if there is no ODM combine requirements / policy in place, and * MPC is required * - * @param [in]: context: New DC state to be programmed - * @param [in]: pipe_e2e: DML pipe end to end context - * - * @return: Number of splits expected (1 for 2:1 split, 3 for 4:1 split, 0 for no splits). - * - * ******************************************************************************************* + * Return: Number of splits expected (1 for 2:1 split, 3 for 4:1 split, 0 for no splits). */ uint8_t dcn32_predict_pipe_split(struct dc_state *context, display_e2e_pipe_params_st *pipe_e2e) @@ -504,7 +499,14 @@ void insert_entry_into_table_sorted(struct _vcs_dpi_voltage_scaling_st *table, } /** - * dcn32_set_phantom_stream_timing: Set timing params for the phantom stream + * dcn32_set_phantom_stream_timing - Set timing params for the phantom stream + * @dc: current dc state + * @context: new dc state + * @ref_pipe: Main pipe for the phantom stream + * @phantom_stream: target phantom stream state + * @pipes: DML pipe params + * @pipe_cnt: number of DML pipes + * @dc_pipe_idx: DC pipe index for the main pipe (i.e. ref_pipe) * * Set timing params of the phantom stream based on calculated output from DML. * This function first gets the DML pipe index using the DC pipe index, then @@ -517,13 +519,6 @@ void insert_entry_into_table_sorted(struct _vcs_dpi_voltage_scaling_st *table, * that separately. * * - Set phantom backporch = vstartup of main pipe - * - * @dc: current dc state - * @context: new dc state - * @ref_pipe: Main pipe for the phantom stream - * @pipes: DML pipe params - * @pipe_cnt: number of DML pipes - * @dc_pipe_idx: DC pipe index for the main pipe (i.e. ref_pipe) */ void dcn32_set_phantom_stream_timing(struct dc *dc, struct dc_state *context, @@ -592,16 +587,14 @@ void dcn32_set_phantom_stream_timing(struct dc *dc, } /** - * dcn32_get_num_free_pipes: Calculate number of free pipes + * dcn32_get_num_free_pipes - Calculate number of free pipes + * @dc: current dc state + * @context: new dc state * * This function assumes that a "used" pipe is a pipe that has * both a stream and a plane assigned to it. * - * @dc: current dc state - * @context: new dc state - * - * Return: - * Number of free pipes available in the context + * Return: Number of free pipes available in the context */ static unsigned int dcn32_get_num_free_pipes(struct dc *dc, struct dc_state *context) { @@ -625,7 +618,10 @@ static unsigned int dcn32_get_num_free_pipes(struct dc *dc, struct dc_state *con } /** - * dcn32_assign_subvp_pipe: Function to decide which pipe will use Sub-VP. + * dcn32_assign_subvp_pipe - Function to decide which pipe will use Sub-VP. + * @dc: current dc state + * @context: new dc state + * @index: [out] dc pipe index for the pipe chosen to have phantom pipes assigned * * We enter this function if we are Sub-VP capable (i.e. enough pipes available) * and regular P-State switching (i.e. VACTIVE/VBLANK) is not supported, or if @@ -639,12 +635,7 @@ static unsigned int dcn32_get_num_free_pipes(struct dc *dc, struct dc_state *con * for determining which should be the SubVP pipe (need a way to determine if a pipe / plane doesn't * support MCLK switching naturally [i.e. ACTIVE or VBLANK]). * - * @param dc: current dc state - * @param context: new dc state - * @param index: [out] dc pipe index for the pipe chosen to have phantom pipes assigned - * - * Return: - * True if a valid pipe assignment was found for Sub-VP. Otherwise false. + * Return: True if a valid pipe assignment was found for Sub-VP. Otherwise false. */ static bool dcn32_assign_subvp_pipe(struct dc *dc, struct dc_state *context, @@ -674,9 +665,10 @@ static bool dcn32_assign_subvp_pipe(struct dc *dc, * - Not able to switch in vactive naturally (switching in active means the * DET provides enough buffer to hide the P-State switch latency -- trying * to combine this with SubVP can cause issues with the scheduling). + * - Not TMZ surface */ if (pipe->plane_state && !pipe->top_pipe && - pipe->stream->mall_stream_config.type == SUBVP_NONE && refresh_rate < 120 && + pipe->stream->mall_stream_config.type == SUBVP_NONE && refresh_rate < 120 && !pipe->plane_state->address.tmz_surface && vba->ActiveDRAMClockChangeLatencyMarginPerState[vba->VoltageLevel][vba->maxMpcComb][vba->pipe_plane[pipe_idx]] <= 0) { while (pipe) { num_pipes++; @@ -710,7 +702,9 @@ static bool dcn32_assign_subvp_pipe(struct dc *dc, } /** - * dcn32_enough_pipes_for_subvp: Function to check if there are "enough" pipes for SubVP. + * dcn32_enough_pipes_for_subvp - Function to check if there are "enough" pipes for SubVP. + * @dc: current dc state + * @context: new dc state * * This function returns true if there are enough free pipes * to create the required phantom pipes for any given stream @@ -722,9 +716,6 @@ static bool dcn32_assign_subvp_pipe(struct dc *dc, * pipe which can be used as the phantom pipe for the non pipe * split pipe. * - * @dc: current dc state - * @context: new dc state - * * Return: * True if there are enough free pipes to assign phantom pipes to at least one * stream that does not already have phantom pipes assigned. Otherwise false. @@ -763,7 +754,9 @@ static bool dcn32_enough_pipes_for_subvp(struct dc *dc, struct dc_state *context } /** - * subvp_subvp_schedulable: Determine if SubVP + SubVP config is schedulable + * subvp_subvp_schedulable - Determine if SubVP + SubVP config is schedulable + * @dc: current dc state + * @context: new dc state * * High level algorithm: * 1. Find longest microschedule length (in us) between the two SubVP pipes @@ -771,11 +764,7 @@ static bool dcn32_enough_pipes_for_subvp(struct dc *dc, struct dc_state *context * pipes still allows for the maximum microschedule to fit in the active * region for both pipes. * - * @dc: current dc state - * @context: new dc state - * - * Return: - * bool - True if the SubVP + SubVP config is schedulable, false otherwise + * Return: True if the SubVP + SubVP config is schedulable, false otherwise */ static bool subvp_subvp_schedulable(struct dc *dc, struct dc_state *context) { @@ -835,7 +824,10 @@ static bool subvp_subvp_schedulable(struct dc *dc, struct dc_state *context) } /** - * subvp_drr_schedulable: Determine if SubVP + DRR config is schedulable + * subvp_drr_schedulable - Determine if SubVP + DRR config is schedulable + * @dc: current dc state + * @context: new dc state + * @drr_pipe: DRR pipe_ctx for the SubVP + DRR config * * High level algorithm: * 1. Get timing for SubVP pipe, phantom pipe, and DRR pipe @@ -844,12 +836,7 @@ static bool subvp_subvp_schedulable(struct dc *dc, struct dc_state *context) * 3.If (SubVP Active - Prefetch > Stretched DRR frame + max(MALL region, Stretched DRR frame)) * then report the configuration as supported * - * @dc: current dc state - * @context: new dc state - * @drr_pipe: DRR pipe_ctx for the SubVP + DRR config - * - * Return: - * bool - True if the SubVP + DRR config is schedulable, false otherwise + * Return: True if the SubVP + DRR config is schedulable, false otherwise */ static bool subvp_drr_schedulable(struct dc *dc, struct dc_state *context, struct pipe_ctx *drr_pipe) { @@ -913,7 +900,9 @@ static bool subvp_drr_schedulable(struct dc *dc, struct dc_state *context, struc /** - * subvp_vblank_schedulable: Determine if SubVP + VBLANK config is schedulable + * subvp_vblank_schedulable - Determine if SubVP + VBLANK config is schedulable + * @dc: current dc state + * @context: new dc state * * High level algorithm: * 1. Get timing for SubVP pipe, phantom pipe, and VBLANK pipe @@ -921,11 +910,7 @@ static bool subvp_drr_schedulable(struct dc *dc, struct dc_state *context, struc * then report the configuration as supported * 3. If the VBLANK display is DRR, then take the DRR static schedulability path * - * @dc: current dc state - * @context: new dc state - * - * Return: - * bool - True if the SubVP + VBLANK/DRR config is schedulable, false otherwise + * Return: True if the SubVP + VBLANK/DRR config is schedulable, false otherwise */ static bool subvp_vblank_schedulable(struct dc *dc, struct dc_state *context) { @@ -1002,20 +987,18 @@ static bool subvp_vblank_schedulable(struct dc *dc, struct dc_state *context) } /** - * subvp_validate_static_schedulability: Check which SubVP case is calculated and handle - * static analysis based on the case. + * subvp_validate_static_schedulability - Check which SubVP case is calculated + * and handle static analysis based on the case. + * @dc: current dc state + * @context: new dc state + * @vlevel: Voltage level calculated by DML * * Three cases: * 1. SubVP + SubVP * 2. SubVP + VBLANK (DRR checked internally) * 3. SubVP + VACTIVE (currently unsupported) * - * @dc: current dc state - * @context: new dc state - * @vlevel: Voltage level calculated by DML - * - * Return: - * bool - True if statically schedulable, false otherwise + * Return: True if statically schedulable, false otherwise */ static bool subvp_validate_static_schedulability(struct dc *dc, struct dc_state *context, @@ -1114,12 +1097,15 @@ static void dcn32_full_validate_bw_helper(struct dc *dc, * 5. (Config doesn't support MCLK in VACTIVE/VBLANK || dc->debug.force_subvp_mclk_switch) */ if (!dc->debug.force_disable_subvp && dcn32_all_pipes_have_stream_and_plane(dc, context) && - !dcn32_mpo_in_use(context) && (*vlevel == context->bw_ctx.dml.soc.num_states || + !dcn32_mpo_in_use(context) && !dcn32_any_surfaces_rotated(dc, context) && + (*vlevel == context->bw_ctx.dml.soc.num_states || vba->DRAMClockChangeSupport[*vlevel][vba->maxMpcComb] == dm_dram_clock_change_unsupported || dc->debug.force_subvp_mclk_switch)) { dcn32_merge_pipes_for_subvp(dc, context); - // to re-initialize viewport after the pipe merge + memset(merge, 0, MAX_PIPES * sizeof(bool)); + + /* to re-initialize viewport after the pipe merge */ for (i = 0; i < dc->res_pool->pipe_count; i++) { struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i]; @@ -1588,6 +1574,33 @@ bool dcn32_internal_validate_bw(struct dc *dc, if (pipe->next_odm_pipe) pipe->next_odm_pipe->prev_odm_pipe = pipe->prev_odm_pipe; + /*2:1ODM+MPC Split MPO to Single Pipe + MPC Split MPO*/ + if (pipe->bottom_pipe) { + if (pipe->bottom_pipe->prev_odm_pipe || pipe->bottom_pipe->next_odm_pipe) { + /*MPC split rules will handle this case*/ + pipe->bottom_pipe->top_pipe = NULL; + } else { + /* when merging an ODM pipes, the bottom MPC pipe must now point to + * the previous ODM pipe and its associated stream assets + */ + if (pipe->prev_odm_pipe->bottom_pipe) { + /* 3 plane MPO*/ + pipe->bottom_pipe->top_pipe = pipe->prev_odm_pipe->bottom_pipe; + pipe->prev_odm_pipe->bottom_pipe->bottom_pipe = pipe->bottom_pipe; + } else { + /* 2 plane MPO*/ + pipe->bottom_pipe->top_pipe = pipe->prev_odm_pipe; + pipe->prev_odm_pipe->bottom_pipe = pipe->bottom_pipe; + } + + memcpy(&pipe->bottom_pipe->stream_res, &pipe->bottom_pipe->top_pipe->stream_res, sizeof(struct stream_resource)); + } + } + + if (pipe->top_pipe) { + pipe->top_pipe->bottom_pipe = NULL; + } + pipe->bottom_pipe = NULL; pipe->next_odm_pipe = NULL; pipe->plane_state = NULL; @@ -1756,6 +1769,7 @@ void dcn32_calculate_wm_and_dlg_fpu(struct dc *dc, struct dc_state *context, int i, pipe_idx, vlevel_temp = 0; double dcfclk = dcn3_2_soc.clock_limits[0].dcfclk_mhz; double dcfclk_from_validation = context->bw_ctx.dml.vba.DCFCLKState[vlevel][context->bw_ctx.dml.vba.maxMpcComb]; + double dcfclk_from_fw_based_mclk_switching = dcfclk_from_validation; bool pstate_en = context->bw_ctx.dml.vba.DRAMClockChangeSupport[vlevel][context->bw_ctx.dml.vba.maxMpcComb] != dm_dram_clock_change_unsupported; unsigned int dummy_latency_index = 0; @@ -1791,7 +1805,7 @@ void dcn32_calculate_wm_and_dlg_fpu(struct dc *dc, struct dc_state *context, dc->clk_mgr->bw_params->wm_table.nv_entries[WM_A].dml_input.pstate_latency_us; dcn32_internal_validate_bw(dc, context, pipes, &pipe_cnt, &vlevel, false); maxMpcComb = context->bw_ctx.dml.vba.maxMpcComb; - dcfclk = context->bw_ctx.dml.vba.DCFCLKState[vlevel][context->bw_ctx.dml.vba.maxMpcComb]; + dcfclk_from_fw_based_mclk_switching = context->bw_ctx.dml.vba.DCFCLKState[vlevel][context->bw_ctx.dml.vba.maxMpcComb]; pstate_en = context->bw_ctx.dml.vba.DRAMClockChangeSupport[vlevel][maxMpcComb] != dm_dram_clock_change_unsupported; } @@ -1877,6 +1891,10 @@ void dcn32_calculate_wm_and_dlg_fpu(struct dc *dc, struct dc_state *context, pipes[0].clks_cfg.dcfclk_mhz = dcfclk_from_validation; pipes[0].clks_cfg.socclk_mhz = context->bw_ctx.dml.soc.clock_limits[vlevel].socclk_mhz; + if (context->bw_ctx.bw.dcn.clk.fw_based_mclk_switching) { + pipes[0].clks_cfg.dcfclk_mhz = dcfclk_from_fw_based_mclk_switching; + } + if (dc->clk_mgr->bw_params->wm_table.nv_entries[WM_C].valid) { min_dram_speed_mts = context->bw_ctx.dml.vba.DRAMSpeed; min_dram_speed_mts_margin = 160; @@ -2019,6 +2037,45 @@ static void remove_entry_from_table_at_index(struct _vcs_dpi_voltage_scaling_st memset(&table[--(*num_entries)], 0, sizeof(struct _vcs_dpi_voltage_scaling_st)); } +void dcn32_patch_dpm_table(struct clk_bw_params *bw_params) +{ + int i; + unsigned int max_dcfclk_mhz = 0, max_dispclk_mhz = 0, max_dppclk_mhz = 0, + max_phyclk_mhz = 0, max_dtbclk_mhz = 0, max_fclk_mhz = 0, max_uclk_mhz = 0; + + for (i = 0; i < MAX_NUM_DPM_LVL; i++) { + if (bw_params->clk_table.entries[i].dcfclk_mhz > max_dcfclk_mhz) + max_dcfclk_mhz = bw_params->clk_table.entries[i].dcfclk_mhz; + if (bw_params->clk_table.entries[i].fclk_mhz > max_fclk_mhz) + max_fclk_mhz = bw_params->clk_table.entries[i].fclk_mhz; + if (bw_params->clk_table.entries[i].memclk_mhz > max_uclk_mhz) + max_uclk_mhz = bw_params->clk_table.entries[i].memclk_mhz; + if (bw_params->clk_table.entries[i].dispclk_mhz > max_dispclk_mhz) + max_dispclk_mhz = bw_params->clk_table.entries[i].dispclk_mhz; + if (bw_params->clk_table.entries[i].dppclk_mhz > max_dppclk_mhz) + max_dppclk_mhz = bw_params->clk_table.entries[i].dppclk_mhz; + if (bw_params->clk_table.entries[i].phyclk_mhz > max_phyclk_mhz) + max_phyclk_mhz = bw_params->clk_table.entries[i].phyclk_mhz; + if (bw_params->clk_table.entries[i].dtbclk_mhz > max_dtbclk_mhz) + max_dtbclk_mhz = bw_params->clk_table.entries[i].dtbclk_mhz; + } + + /* Scan through clock values we currently have and if they are 0, + * then populate it with dcn3_2_soc.clock_limits[] value. + * + * Do it for DCFCLK, DISPCLK, DTBCLK and UCLK as any of those being + * 0, will cause it to skip building the clock table. + */ + if (max_dcfclk_mhz == 0) + bw_params->clk_table.entries[0].dcfclk_mhz = dcn3_2_soc.clock_limits[0].dcfclk_mhz; + if (max_dispclk_mhz == 0) + bw_params->clk_table.entries[0].dispclk_mhz = dcn3_2_soc.clock_limits[0].dispclk_mhz; + if (max_dtbclk_mhz == 0) + bw_params->clk_table.entries[0].dtbclk_mhz = dcn3_2_soc.clock_limits[0].dtbclk_mhz; + if (max_uclk_mhz == 0) + bw_params->clk_table.entries[0].memclk_mhz = dcn3_2_soc.clock_limits[0].dram_speed_mts / 16; +} + static int build_synthetic_soc_states(struct clk_bw_params *bw_params, struct _vcs_dpi_voltage_scaling_st *table, unsigned int *num_entries) { @@ -2211,7 +2268,7 @@ static int build_synthetic_soc_states(struct clk_bw_params *bw_params, return 0; } -/** +/* * dcn32_update_bw_bounding_box * * This would override some dcn3_2 ip_or_soc initial parameters hardcoded from diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.h b/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.h index 732ed7fadb8a..3a3dc2ce4c73 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.h +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.h @@ -71,4 +71,6 @@ int dcn32_find_dummy_latency_index_for_fw_based_mclk_switch(struct dc *dc, int pipe_cnt, int vlevel); +void dcn32_patch_dpm_table(struct clk_bw_params *bw_params); + #endif diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_32.c b/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_32.c index 75be1e1ce543..5b91660a6496 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_32.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_32.c @@ -733,6 +733,8 @@ static void DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerforman mode_lib->vba.FCLKChangeLatency, v->UrgentLatency, mode_lib->vba.SREnterPlusExitTime); + memset(&v->dummy_vars.DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerformanceCalculation.myPipe, 0, sizeof(DmlPipe)); + v->dummy_vars.DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerformanceCalculation.myPipe.Dppclk = mode_lib->vba.DPPCLK[k]; v->dummy_vars.DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerformanceCalculation.myPipe.Dispclk = mode_lib->vba.DISPCLK; v->dummy_vars.DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerformanceCalculation.myPipe.PixelClock = mode_lib->vba.PixelClock[k]; @@ -2252,9 +2254,8 @@ void dml32_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l for (k = 0; k <= mode_lib->vba.NumberOfActiveSurfaces - 1; k++) { if (!(mode_lib->vba.DSCInputBitPerComponent[k] == 12.0 || mode_lib->vba.DSCInputBitPerComponent[k] == 10.0 - || mode_lib->vba.DSCInputBitPerComponent[k] == 8.0 - || mode_lib->vba.DSCInputBitPerComponent[k] > - mode_lib->vba.MaximumDSCBitsPerComponent)) { + || mode_lib->vba.DSCInputBitPerComponent[k] == 8.0) + || mode_lib->vba.DSCInputBitPerComponent[k] > mode_lib->vba.MaximumDSCBitsPerComponent) { mode_lib->vba.NonsupportedDSCInputBPC = true; } } @@ -2330,16 +2331,15 @@ void dml32_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l if (mode_lib->vba.OutputMultistreamId[k] == k && mode_lib->vba.ForcedOutputLinkBPP[k] == 0) mode_lib->vba.BPPForMultistreamNotIndicated = true; for (j = 0; j < mode_lib->vba.NumberOfActiveSurfaces; ++j) { - if (mode_lib->vba.OutputMultistreamId[k] == j && mode_lib->vba.OutputMultistreamEn[k] + if (mode_lib->vba.OutputMultistreamId[k] == j && mode_lib->vba.ForcedOutputLinkBPP[k] == 0) mode_lib->vba.BPPForMultistreamNotIndicated = true; } } if ((mode_lib->vba.Output[k] == dm_edp || mode_lib->vba.Output[k] == dm_hdmi)) { - if (mode_lib->vba.OutputMultistreamId[k] == k && mode_lib->vba.OutputMultistreamEn[k]) + if (mode_lib->vba.OutputMultistreamEn[k] == true && mode_lib->vba.OutputMultistreamId[k] == k) mode_lib->vba.MultistreamWithHDMIOreDP = true; - for (j = 0; j < mode_lib->vba.NumberOfActiveSurfaces; ++j) { if (mode_lib->vba.OutputMultistreamEn[k] == true && mode_lib->vba.OutputMultistreamId[k] == j) mode_lib->vba.MultistreamWithHDMIOreDP = true; @@ -2478,8 +2478,6 @@ void dml32_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l mode_lib->vba.PixelClock[k], mode_lib->vba.PixelClockBackEnd[k]); } - m = 0; - for (k = 0; k <= mode_lib->vba.NumberOfActiveSurfaces - 1; k++) { for (m = 0; m <= mode_lib->vba.NumberOfActiveSurfaces - 1; m++) { for (j = 0; j <= mode_lib->vba.NumberOfActiveSurfaces - 1; j++) { @@ -2856,8 +2854,6 @@ void dml32_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l } } - m = 0; - //Calculate Return BW for (i = 0; i < (int) v->soc.num_states; ++i) { for (j = 0; j <= 1; ++j) { @@ -3618,11 +3614,10 @@ void dml32_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l mode_lib->vba.ModeIsSupported = mode_lib->vba.ModeSupport[i][0] == true || mode_lib->vba.ModeSupport[i][1] == true; - if (mode_lib->vba.ModeSupport[i][0] == true) { + if (mode_lib->vba.ModeSupport[i][0] == true) MaximumMPCCombine = 0; - } else { + else MaximumMPCCombine = 1; - } } } diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.c b/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.c index f5400eda07a5..4125d3d111d1 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.c +++ b/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.c @@ -114,6 +114,7 @@ void dml_init_instance(struct display_mode_lib *lib, break; case DML_PROJECT_DCN31: case DML_PROJECT_DCN31_FPGA: + case DML_PROJECT_DCN315: lib->funcs = dml31_funcs; break; case DML_PROJECT_DCN314: diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.h b/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.h index b1878a1440e2..3d643d50c3eb 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.h +++ b/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.h @@ -40,6 +40,7 @@ enum dml_project { DML_PROJECT_DCN21, DML_PROJECT_DCN30, DML_PROJECT_DCN31, + DML_PROJECT_DCN315, DML_PROJECT_DCN31_FPGA, DML_PROJECT_DCN314, DML_PROJECT_DCN32, diff --git a/drivers/gpu/drm/amd/display/dc/inc/core_types.h b/drivers/gpu/drm/amd/display/dc/inc/core_types.h index 8919a2092ac5..9498105c98ab 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/core_types.h +++ b/drivers/gpu/drm/amd/display/dc/inc/core_types.h @@ -39,6 +39,8 @@ #include "panel_cntl.h" #define MAX_CLOCK_SOURCES 7 +#define MAX_SVP_PHANTOM_STREAMS 2 +#define MAX_SVP_PHANTOM_PLANES 2 void enable_surface_flip_reporting(struct dc_plane_state *plane_state, uint32_t controller_id); @@ -232,6 +234,7 @@ struct resource_funcs { unsigned int index); bool (*remove_phantom_pipes)(struct dc *dc, struct dc_state *context); + void (*get_panel_config_defaults)(struct dc_panel_config *panel_config); }; struct audio_support{ @@ -438,7 +441,6 @@ struct pipe_ctx { union pipe_update_flags update_flags; struct dwbc *dwbc; struct mcif_wb *mcif_wb; - bool vtp_locked; }; /* Data used for dynamic link encoder assignment. @@ -492,6 +494,8 @@ struct dcn_bw_output { struct dcn_watermark_set watermarks; struct dcn_bw_writeback bw_writeback; int compbuf_size_kb; + unsigned int legacy_svp_drr_stream_index; + bool legacy_svp_drr_stream_index_valid; }; union bw_output { diff --git a/drivers/gpu/drm/amd/display/dc/inc/dc_link_dp.h b/drivers/gpu/drm/amd/display/dc/inc/dc_link_dp.h index 6682d9e181c6..b304d450b038 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/dc_link_dp.h +++ b/drivers/gpu/drm/amd/display/dc/inc/dc_link_dp.h @@ -194,6 +194,11 @@ enum dc_status dpcd_configure_lttpr_mode( enum dp_link_encoding dp_get_link_encoding_format(const struct dc_link_settings *link_settings); bool dp_retrieve_lttpr_cap(struct dc_link *link); +bool dp_is_lttpr_present(struct dc_link *link); +enum lttpr_mode dp_decide_lttpr_mode(struct dc_link *link, struct dc_link_settings *link_setting); +void dp_get_lttpr_mode_override(struct dc_link *link, enum lttpr_mode *override); +enum lttpr_mode dp_decide_8b_10b_lttpr_mode(struct dc_link *link); +enum lttpr_mode dp_decide_128b_132b_lttpr_mode(struct dc_link *link); bool dpcd_write_128b_132b_sst_payload_allocation_table( const struct dc_stream_state *stream, struct dc_link *link, diff --git a/drivers/gpu/drm/amd/display/dc/inc/dcn_calcs.h b/drivers/gpu/drm/amd/display/dc/inc/dcn_calcs.h index 806f3041db14..9e4ddc985240 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/dcn_calcs.h +++ b/drivers/gpu/drm/amd/display/dc/inc/dcn_calcs.h @@ -628,8 +628,23 @@ unsigned int dcn_find_dcfclk_suits_all( const struct dc *dc, struct dc_clocks *clocks); -void dcn_bw_update_from_pplib(struct dc *dc); -void dcn_bw_notify_pplib_of_wm_ranges(struct dc *dc); +void dcn_get_soc_clks( + struct dc *dc, + int *min_fclk_khz, + int *min_dcfclk_khz, + int *socclk_khz); + +void dcn_bw_update_from_pplib_fclks( + struct dc *dc, + struct dm_pp_clock_levels_with_voltage *fclks); +void dcn_bw_update_from_pplib_dcfclks( + struct dc *dc, + struct dm_pp_clock_levels_with_voltage *dcfclks); +void dcn_bw_notify_pplib_of_wm_ranges( + struct dc *dc, + int min_fclk_khz, + int min_dcfclk_khz, + int socclk_khz); void dcn_bw_sync_calcs_and_dml(struct dc *dc); enum source_macro_tile_size swizzle_mode_to_macro_tile_size(enum swizzle_mode_values sw_mode); diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr.h b/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr.h index d9f1b0a4fbd4..591ab1389e3b 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr.h @@ -95,10 +95,23 @@ struct clk_limit_table_entry { unsigned int wck_ratio; }; +struct clk_limit_num_entries { + unsigned int num_dcfclk_levels; + unsigned int num_fclk_levels; + unsigned int num_memclk_levels; + unsigned int num_socclk_levels; + unsigned int num_dtbclk_levels; + unsigned int num_dispclk_levels; + unsigned int num_dppclk_levels; + unsigned int num_phyclk_levels; + unsigned int num_phyclk_d18_levels; +}; + /* This table is contiguous */ struct clk_limit_table { struct clk_limit_table_entry entries[MAX_NUM_DPM_LVL]; - unsigned int num_entries; + struct clk_limit_num_entries num_entries_per_clk; + unsigned int num_entries; /* highest populated dpm level for back compatibility */ }; struct wm_range_table_entry { diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr_internal.h b/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr_internal.h index 68c2ed434d2c..cff5fd55a0ad 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr_internal.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr_internal.h @@ -340,6 +340,8 @@ struct clk_mgr_internal { bool smu_present; void *wm_range_table; long long wm_range_table_addr; + + bool dpm_present; }; struct clk_mgr_internal_funcs { diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/cursor_reg_cache.h b/drivers/gpu/drm/amd/display/dc/inc/hw/cursor_reg_cache.h new file mode 100644 index 000000000000..45645f9fd86c --- /dev/null +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/cursor_reg_cache.h @@ -0,0 +1,99 @@ +/* SPDX-License-Identifier: MIT */ +/* Copyright © 2022 Advanced Micro Devices, Inc. All rights reserved. */ + +#ifndef __DAL_CURSOR_CACHE_H__ +#define __DAL_CURSOR_CACHE_H__ + +union reg_cursor_control_cfg { + struct { + uint32_t cur_enable: 1; + uint32_t reser0: 3; + uint32_t cur_2x_magnify: 1; + uint32_t reser1: 3; + uint32_t mode: 3; + uint32_t reser2: 5; + uint32_t pitch: 2; + uint32_t reser3: 6; + uint32_t line_per_chunk: 5; + uint32_t reser4: 3; + } bits; + uint32_t raw; +}; +struct cursor_position_cache_hubp { + union reg_cursor_control_cfg cur_ctl; + union reg_position_cfg { + struct { + uint32_t x_pos: 16; + uint32_t y_pos: 16; + } bits; + uint32_t raw; + } position; + union reg_hot_spot_cfg { + struct { + uint32_t x_hot: 16; + uint32_t y_hot: 16; + } bits; + uint32_t raw; + } hot_spot; + union reg_dst_offset_cfg { + struct { + uint32_t dst_x_offset: 13; + uint32_t reserved: 19; + } bits; + uint32_t raw; + } dst_offset; +}; + +struct cursor_attribute_cache_hubp { + uint32_t SURFACE_ADDR_HIGH; + uint32_t SURFACE_ADDR; + union reg_cursor_control_cfg cur_ctl; + union reg_cursor_size_cfg { + struct { + uint32_t width: 16; + uint32_t height: 16; + } bits; + uint32_t raw; + } size; + union reg_cursor_settings_cfg { + struct { + uint32_t dst_y_offset: 8; + uint32_t chunk_hdl_adjust: 2; + uint32_t reserved: 22; + } bits; + uint32_t raw; + } settings; +}; + +struct cursor_rect { + uint32_t x; + uint32_t y; + uint32_t w; + uint32_t h; +}; + +union reg_cur0_control_cfg { + struct { + uint32_t cur0_enable: 1; + uint32_t expansion_mode: 1; + uint32_t reser0: 1; + uint32_t cur0_rom_en: 1; + uint32_t mode: 3; + uint32_t reserved: 25; + } bits; + uint32_t raw; +}; +struct cursor_position_cache_dpp { + union reg_cur0_control_cfg cur0_ctl; +}; + +struct cursor_attribute_cache_dpp { + union reg_cur0_control_cfg cur0_ctl; +}; + +struct cursor_attributes_cfg { + struct cursor_attribute_cache_hubp aHubp; + struct cursor_attribute_cache_dpp aDpp; +}; + +#endif diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h b/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h index 3ef7faa92052..dcb80c4747b0 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h @@ -28,6 +28,7 @@ #define __DAL_DPP_H__ #include "transform.h" +#include "cursor_reg_cache.h" union defer_reg_writes { struct { @@ -58,6 +59,9 @@ struct dpp { struct pwl_params shaper_params; bool cm_bypass_mode; + + struct cursor_position_cache_dpp pos; + struct cursor_attribute_cache_dpp att; }; struct dpp_input_csc_matrix { diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h b/drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h index 44c4578193a3..d5ea7545583e 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h @@ -27,6 +27,7 @@ #define __DAL_HUBP_H__ #include "mem_input.h" +#include "cursor_reg_cache.h" #define OPP_ID_INVALID 0xf #define MAX_TTU 0xffffff @@ -65,6 +66,10 @@ struct hubp { struct dc_cursor_attributes curs_attr; struct dc_cursor_position curs_pos; bool power_gated; + + struct cursor_position_cache_hubp pos; + struct cursor_attribute_cache_hubp att; + struct cursor_rect cur_rect; }; struct surface_flip_registers { diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h b/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h index 72eef7a5ed83..25a1df45b264 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h @@ -209,7 +209,6 @@ struct timing_generator_funcs { void (*set_blank)(struct timing_generator *tg, bool enable_blanking); bool (*is_blanked)(struct timing_generator *tg); - bool (*is_locked)(struct timing_generator *tg); void (*set_overscan_blank_color) (struct timing_generator *tg, const struct tg_color *color); void (*set_blank_color)(struct timing_generator *tg, const struct tg_color *color); void (*set_colors)(struct timing_generator *tg, diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h b/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h index 416b700a8d5a..d04b68dad413 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h @@ -32,11 +32,6 @@ #include "inc/hw/link_encoder.h" #include "core_status.h" -enum vline_select { - VLINE0, - VLINE1 -}; - struct pipe_ctx; struct dc_state; struct dc_stream_status; @@ -89,6 +84,7 @@ struct hw_sequencer_funcs { struct pipe_ctx *pipe_ctx, bool enableTripleBuffer); void (*update_pending_status)(struct pipe_ctx *pipe_ctx); void (*power_down)(struct dc *dc); + void (*update_dsc_pg)(struct dc *dc, struct dc_state *context, bool safe_to_disable); /* Pipe Lock Related */ void (*pipe_control_lock)(struct dc *dc, @@ -117,8 +113,7 @@ struct hw_sequencer_funcs { int group_index, int group_size, struct pipe_ctx *grouped_pipes[]); void (*setup_periodic_interrupt)(struct dc *dc, - struct pipe_ctx *pipe_ctx, - enum vline_select vline); + struct pipe_ctx *pipe_ctx); void (*set_drr)(struct pipe_ctx **pipe_ctx, int num_pipes, struct dc_crtc_timing_adjust adjust); void (*set_static_screen_control)(struct pipe_ctx **pipe_ctx, diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer_private.h b/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer_private.h index 1cdea0efe5c1..a4d61bb724b6 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer_private.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer_private.h @@ -124,6 +124,8 @@ struct hwseq_private_funcs { void (*dsc_pg_control)(struct dce_hwseq *hws, unsigned int dsc_inst, bool power_on); + bool (*dsc_pg_status)(struct dce_hwseq *hws, + unsigned int dsc_inst); void (*update_odm)(struct dc *dc, struct dc_state *context, struct pipe_ctx *pipe_ctx); void (*program_all_writeback_pipes_in_tree)(struct dc *dc, diff --git a/drivers/gpu/drm/amd/display/dc/inc/resource.h b/drivers/gpu/drm/amd/display/dc/inc/resource.h index c37d1141febe..5040836f404d 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/resource.h +++ b/drivers/gpu/drm/amd/display/dc/inc/resource.h @@ -230,4 +230,10 @@ const struct link_hwss *get_link_hwss(const struct dc_link *link, bool is_h_timing_divisible_by_2(struct dc_stream_state *stream); +bool dc_resource_acquire_secondary_pipe_for_mpc_odm( + const struct dc *dc, + struct dc_state *state, + struct pipe_ctx *pri_pipe, + struct pipe_ctx *sec_pipe, + bool odm); #endif /* DRIVERS_GPU_DRM_AMD_DC_DEV_DC_INC_RESOURCE_H_ */ diff --git a/drivers/gpu/drm/amd/display/dc/link/link_hwss_hpo_dp.c b/drivers/gpu/drm/amd/display/dc/link/link_hwss_hpo_dp.c index 7d3147175ca2..153a88381f2c 100644 --- a/drivers/gpu/drm/amd/display/dc/link/link_hwss_hpo_dp.c +++ b/drivers/gpu/drm/amd/display/dc/link/link_hwss_hpo_dp.c @@ -111,7 +111,7 @@ static void setup_hpo_dp_stream_encoder(struct pipe_ctx *pipe_ctx) enum phyd32clk_clock_source phyd32clk = get_phyd32clk_src(pipe_ctx->stream->link); dto_params.otg_inst = tg->inst; - dto_params.pixclk_khz = pipe_ctx->stream->phy_pix_clk; + dto_params.pixclk_khz = pipe_ctx->stream->timing.pix_clk_100hz / 10; dto_params.num_odm_segments = get_odm_segment_count(pipe_ctx); dto_params.timing = &pipe_ctx->stream->timing; dto_params.ref_dtbclk_khz = dc->clk_mgr->funcs->get_dtb_ref_clk_frequency(dc->clk_mgr); diff --git a/drivers/gpu/drm/amd/display/dc/virtual/virtual_link_hwss.c b/drivers/gpu/drm/amd/display/dc/virtual/virtual_link_hwss.c index 9522fe0b36c9..4f7f99156897 100644 --- a/drivers/gpu/drm/amd/display/dc/virtual/virtual_link_hwss.c +++ b/drivers/gpu/drm/amd/display/dc/virtual/virtual_link_hwss.c @@ -37,7 +37,7 @@ void virtual_reset_stream_encoder(struct pipe_ctx *pipe_ctx) { } -void virtual_disable_link_output(struct dc_link *link, +static void virtual_disable_link_output(struct dc_link *link, const struct link_resource *link_res, enum signal_type signal) { diff --git a/drivers/gpu/drm/amd/display/dmub/dmub_srv.h b/drivers/gpu/drm/amd/display/dmub/dmub_srv.h index f34c45b19fcb..eb5b7eb292ef 100644 --- a/drivers/gpu/drm/amd/display/dmub/dmub_srv.h +++ b/drivers/gpu/drm/amd/display/dmub/dmub_srv.h @@ -248,6 +248,7 @@ struct dmub_srv_hw_params { bool disable_dpia; bool usb4_cm_version; bool fw_in_system_memory; + bool dpia_hpd_int_enable_supported; }; /** diff --git a/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h b/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h index 6b9a529e9f12..7a8f61517424 100644 --- a/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h +++ b/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h @@ -400,8 +400,9 @@ union dmub_fw_boot_options { uint32_t diag_env: 1; /* 1 if diagnostic environment */ uint32_t gpint_scratch8: 1; /* 1 if GPINT is in scratch8*/ uint32_t usb4_cm_version: 1; /**< 1 CM support */ + uint32_t dpia_hpd_int_enable_supported: 1; /* 1 if dpia hpd int enable supported */ - uint32_t reserved : 17; /**< reserved */ + uint32_t reserved : 16; /**< reserved */ } bits; /**< boot bits */ uint32_t all; /**< 32-bit access to bits */ }; @@ -728,6 +729,12 @@ enum dmub_cmd_type { /** * Command type used for all VBIOS interface commands. */ + + /** + * Command type used to set DPIA HPD interrupt state + */ + DMUB_CMD__DPIA_HPD_INT_ENABLE = 86, + DMUB_CMD__VBIOS = 128, }; @@ -1256,6 +1263,14 @@ struct dmub_rb_cmd_set_mst_alloc_slots { }; /** + * DMUB command structure for DPIA HPD int enable control. + */ +struct dmub_rb_cmd_dpia_hpd_int_enable { + struct dmub_cmd_header header; /* header */ + uint32_t enable; /* dpia hpd interrupt enable */ +}; + +/** * struct dmub_rb_cmd_dpphy_init - DPPHY init. */ struct dmub_rb_cmd_dpphy_init { @@ -2084,7 +2099,99 @@ struct dmub_rb_cmd_update_dirty_rect { /** * Data passed from driver to FW in a DMUB_CMD__UPDATE_CURSOR_INFO command. */ -struct dmub_cmd_update_cursor_info_data { +union dmub_reg_cursor_control_cfg { + struct { + uint32_t cur_enable: 1; + uint32_t reser0: 3; + uint32_t cur_2x_magnify: 1; + uint32_t reser1: 3; + uint32_t mode: 3; + uint32_t reser2: 5; + uint32_t pitch: 2; + uint32_t reser3: 6; + uint32_t line_per_chunk: 5; + uint32_t reser4: 3; + } bits; + uint32_t raw; +}; +struct dmub_cursor_position_cache_hubp { + union dmub_reg_cursor_control_cfg cur_ctl; + union dmub_reg_position_cfg { + struct { + uint32_t cur_x_pos: 16; + uint32_t cur_y_pos: 16; + } bits; + uint32_t raw; + } position; + union dmub_reg_hot_spot_cfg { + struct { + uint32_t hot_x: 16; + uint32_t hot_y: 16; + } bits; + uint32_t raw; + } hot_spot; + union dmub_reg_dst_offset_cfg { + struct { + uint32_t dst_x_offset: 13; + uint32_t reserved: 19; + } bits; + uint32_t raw; + } dst_offset; +}; + +union dmub_reg_cur0_control_cfg { + struct { + uint32_t cur0_enable: 1; + uint32_t expansion_mode: 1; + uint32_t reser0: 1; + uint32_t cur0_rom_en: 1; + uint32_t mode: 3; + uint32_t reserved: 25; + } bits; + uint32_t raw; +}; +struct dmub_cursor_position_cache_dpp { + union dmub_reg_cur0_control_cfg cur0_ctl; +}; +struct dmub_cursor_position_cfg { + struct dmub_cursor_position_cache_hubp pHubp; + struct dmub_cursor_position_cache_dpp pDpp; + uint8_t pipe_idx; + /* + * Padding is required. To be 4 Bytes Aligned. + */ + uint8_t padding[3]; +}; + +struct dmub_cursor_attribute_cache_hubp { + uint32_t SURFACE_ADDR_HIGH; + uint32_t SURFACE_ADDR; + union dmub_reg_cursor_control_cfg cur_ctl; + union dmub_reg_cursor_size_cfg { + struct { + uint32_t width: 16; + uint32_t height: 16; + } bits; + uint32_t raw; + } size; + union dmub_reg_cursor_settings_cfg { + struct { + uint32_t dst_y_offset: 8; + uint32_t chunk_hdl_adjust: 2; + uint32_t reserved: 22; + } bits; + uint32_t raw; + } settings; +}; +struct dmub_cursor_attribute_cache_dpp { + union dmub_reg_cur0_control_cfg cur0_ctl; +}; +struct dmub_cursor_attributes_cfg { + struct dmub_cursor_attribute_cache_hubp aHubp; + struct dmub_cursor_attribute_cache_dpp aDpp; +}; + +struct dmub_cmd_update_cursor_payload0 { /** * Cursor dirty rects. */ @@ -2111,6 +2218,20 @@ struct dmub_cmd_update_cursor_info_data { * Currently the support is only for 0 or 1 */ uint8_t panel_inst; + /** + * Cursor Position Register. + * Registers contains Hubp & Dpp modules + */ + struct dmub_cursor_position_cfg position_cfg; +}; + +struct dmub_cmd_update_cursor_payload1 { + struct dmub_cursor_attributes_cfg attribute_cfg; +}; + +union dmub_cmd_update_cursor_info_data { + struct dmub_cmd_update_cursor_payload0 payload0; + struct dmub_cmd_update_cursor_payload1 payload1; }; /** * Definition of a DMUB_CMD__UPDATE_CURSOR_INFO command. @@ -2123,7 +2244,7 @@ struct dmub_rb_cmd_update_cursor_info { /** * Data passed from driver to FW in a DMUB_CMD__UPDATE_CURSOR_INFO command. */ - struct dmub_cmd_update_cursor_info_data update_cursor_info_data; + union dmub_cmd_update_cursor_info_data update_cursor_info_data; }; /** @@ -2820,11 +2941,7 @@ struct dmub_rb_cmd_get_visual_confirm_color { struct dmub_optc_state { uint32_t v_total_max; uint32_t v_total_min; - uint32_t v_total_mid; - uint32_t v_total_mid_frame_num; uint32_t tg_inst; - uint32_t enable_manual_trigger; - uint32_t clear_force_vsync; }; struct dmub_rb_cmd_drr_update { @@ -3230,6 +3347,10 @@ union dmub_rb_cmd { * Definition of a DMUB_CMD__QUERY_HPD_STATE command. */ struct dmub_rb_cmd_query_hpd_state query_hpd; + /** + * Definition of a DMUB_CMD__DPIA_HPD_INT_ENABLE command. + */ + struct dmub_rb_cmd_dpia_hpd_int_enable dpia_hpd_int_enable; }; /** diff --git a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn31.c b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn31.c index c7bd7e216710..c90b9ee42e12 100644 --- a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn31.c +++ b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn31.c @@ -350,6 +350,7 @@ void dmub_dcn31_enable_dmub_boot_options(struct dmub_srv *dmub, const struct dmu boot_options.bits.dpia_supported = params->dpia_supported; boot_options.bits.enable_dpia = params->disable_dpia ? 0 : 1; boot_options.bits.usb4_cm_version = params->usb4_cm_version; + boot_options.bits.dpia_hpd_int_enable_supported = params->dpia_hpd_int_enable_supported; boot_options.bits.power_optimization = params->power_optimization; boot_options.bits.sel_mux_phy_c_d_phy_f_g = (dmub->asic == DMUB_ASIC_DCN31B) ? 1 : 0; diff --git a/drivers/gpu/drm/amd/display/include/link_service_types.h b/drivers/gpu/drm/amd/display/include/link_service_types.h index d76ab72baf0c..d1e91d31d151 100644 --- a/drivers/gpu/drm/amd/display/include/link_service_types.h +++ b/drivers/gpu/drm/amd/display/include/link_service_types.h @@ -83,6 +83,7 @@ enum link_training_result { }; enum lttpr_mode { + LTTPR_MODE_UNKNOWN, LTTPR_MODE_NON_LTTPR, LTTPR_MODE_TRANSPARENT, LTTPR_MODE_NON_TRANSPARENT, diff --git a/drivers/gpu/drm/amd/display/modules/color/color_gamma.c b/drivers/gpu/drm/amd/display/modules/color/color_gamma.c index 04f7656906ca..447a0ec9cbe2 100644 --- a/drivers/gpu/drm/amd/display/modules/color/color_gamma.c +++ b/drivers/gpu/drm/amd/display/modules/color/color_gamma.c @@ -1692,7 +1692,7 @@ static void apply_degamma_for_user_regamma(struct pwl_float_data_ex *rgb_regamma struct pwl_float_data_ex *rgb = rgb_regamma; const struct hw_x_point *coord_x = coordinates_x; - build_coefficients(&coeff, true); + build_coefficients(&coeff, TRANSFER_FUNCTION_SRGB); i = 0; while (i != hw_points_num + 1) { diff --git a/drivers/gpu/drm/amd/include/asic_reg/mmhub/mmhub_9_4_1_offset.h b/drivers/gpu/drm/amd/include/asic_reg/mmhub/mmhub_9_4_1_offset.h index d8632ccf3494..c488d4a50cf4 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/mmhub/mmhub_9_4_1_offset.h +++ b/drivers/gpu/drm/amd/include/asic_reg/mmhub/mmhub_9_4_1_offset.h @@ -4409,6 +4409,10 @@ #define mmVMSHAREDPF0_MC_VM_XGMI_LFB_SIZE_BASE_IDX 1 #define mmVMSHAREDPF0_MC_VM_CACHEABLE_DRAM_CNTL 0x0af9 #define mmVMSHAREDPF0_MC_VM_CACHEABLE_DRAM_CNTL_BASE_IDX 1 +#define mmMC_VM_XGMI_LFB_CNTL 0x0823 +#define mmMC_VM_XGMI_LFB_CNTL_BASE_IDX 0 +#define mmMC_VM_XGMI_LFB_SIZE 0x0824 +#define mmMC_VM_XGMI_LFB_SIZE_BASE_IDX 0 // addressBlock: mmhub_utcl2_vmsharedvcdec diff --git a/drivers/gpu/drm/amd/include/asic_reg/mmhub/mmhub_9_4_1_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/mmhub/mmhub_9_4_1_sh_mask.h index 111a71b434e2..2969fbf282b7 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/mmhub/mmhub_9_4_1_sh_mask.h +++ b/drivers/gpu/drm/amd/include/asic_reg/mmhub/mmhub_9_4_1_sh_mask.h @@ -26728,6 +26728,14 @@ //VMSHAREDPF0_MC_VM_CACHEABLE_DRAM_CNTL #define VMSHAREDPF0_MC_VM_CACHEABLE_DRAM_CNTL__ENABLE_CACHEABLE_DRAM_ADDRESS_APERTURE__SHIFT 0x0 #define VMSHAREDPF0_MC_VM_CACHEABLE_DRAM_CNTL__ENABLE_CACHEABLE_DRAM_ADDRESS_APERTURE_MASK 0x00000001L +//MC_VM_XGMI_LFB_CNTL +#define MC_VM_XGMI_LFB_CNTL__PF_LFB_REGION__SHIFT 0x0 +#define MC_VM_XGMI_LFB_CNTL__PF_MAX_REGION__SHIFT 0x3 +#define MC_VM_XGMI_LFB_CNTL__PF_LFB_REGION_MASK 0x00000007L +#define MC_VM_XGMI_LFB_CNTL__PF_MAX_REGION_MASK 0x00000038L +//MC_VM_XGMI_LFB_SIZE +#define MC_VM_XGMI_LFB_SIZE__PF_LFB_SIZE__SHIFT 0x0 +#define MC_VM_XGMI_LFB_SIZE__PF_LFB_SIZE_MASK 0x0000FFFFL // addressBlock: mmhub_utcl2_vmsharedvcdec diff --git a/drivers/gpu/drm/amd/include/asic_reg/umc/umc_8_10_0_offset.h b/drivers/gpu/drm/amd/include/asic_reg/umc/umc_8_10_0_offset.h index b798cf5a2c39..38adde3cae5a 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/umc/umc_8_10_0_offset.h +++ b/drivers/gpu/drm/amd/include/asic_reg/umc/umc_8_10_0_offset.h @@ -29,5 +29,7 @@ #define regMCA_UMC_UMC0_MCUMC_STATUST0_BASE_IDX 2 #define regMCA_UMC_UMC0_MCUMC_ADDRT0 0x03c4 #define regMCA_UMC_UMC0_MCUMC_ADDRT0_BASE_IDX 2 +#define regUMCCH0_0_GeccCtrl 0x0053 +#define regUMCCH0_0_GeccCtrl_BASE_IDX 2 #endif diff --git a/drivers/gpu/drm/amd/include/asic_reg/umc/umc_8_10_0_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/umc/umc_8_10_0_sh_mask.h index bd99b431247f..4dbec524f943 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/umc/umc_8_10_0_sh_mask.h +++ b/drivers/gpu/drm/amd/include/asic_reg/umc/umc_8_10_0_sh_mask.h @@ -90,5 +90,8 @@ #define MCA_UMC_UMC0_MCUMC_ADDRT0__ErrorAddr__SHIFT 0x0 #define MCA_UMC_UMC0_MCUMC_ADDRT0__Reserved__SHIFT 0x38 #define MCA_UMC_UMC0_MCUMC_ADDRT0__ErrorAddr_MASK 0x00FFFFFFFFFFFFFFL +//UMCCH0_0_GeccCtrl +#define UMCCH0_0_GeccCtrl__UCFatalEn__SHIFT 0xd +#define UMCCH0_0_GeccCtrl__UCFatalEn_MASK 0x00002000L #endif diff --git a/drivers/gpu/drm/amd/include/kgd_kfd_interface.h b/drivers/gpu/drm/amd/include/kgd_kfd_interface.h index e85364dff4e0..5cb3e8634739 100644 --- a/drivers/gpu/drm/amd/include/kgd_kfd_interface.h +++ b/drivers/gpu/drm/amd/include/kgd_kfd_interface.h @@ -262,8 +262,9 @@ struct kfd2kgd_calls { uint32_t queue_id); int (*hqd_destroy)(struct amdgpu_device *adev, void *mqd, - uint32_t reset_type, unsigned int timeout, - uint32_t pipe_id, uint32_t queue_id); + enum kfd_preempt_type reset_type, + unsigned int timeout, uint32_t pipe_id, + uint32_t queue_id); bool (*hqd_sdma_is_occupied)(struct amdgpu_device *adev, void *mqd); diff --git a/drivers/gpu/drm/amd/include/mes_v11_api_def.h b/drivers/gpu/drm/amd/include/mes_v11_api_def.h index 50bfa513cb35..7e85cdc5bd34 100644 --- a/drivers/gpu/drm/amd/include/mes_v11_api_def.h +++ b/drivers/gpu/drm/amd/include/mes_v11_api_def.h @@ -269,7 +269,8 @@ union MESAPI__ADD_QUEUE { uint32_t map_kiq_utility_queue : 1; uint32_t is_kfd_process : 1; uint32_t trap_en : 1; - uint32_t reserved : 21; + uint32_t is_aql_queue : 1; + uint32_t reserved : 20; }; struct MES_API_STATUS api_status; uint64_t tma_addr; diff --git a/drivers/gpu/drm/amd/pm/amdgpu_pm.c b/drivers/gpu/drm/amd/pm/amdgpu_pm.c index 948cc75376f8..236657eece47 100644 --- a/drivers/gpu/drm/amd/pm/amdgpu_pm.c +++ b/drivers/gpu/drm/amd/pm/amdgpu_pm.c @@ -3362,11 +3362,11 @@ int amdgpu_pm_sysfs_init(struct amdgpu_device *adev) if (adev->pm.sysfs_initialized) return 0; + INIT_LIST_HEAD(&adev->pm.pm_attr_list); + if (adev->pm.dpm_enabled == 0) return 0; - INIT_LIST_HEAD(&adev->pm.pm_attr_list); - adev->pm.int_hwmon_dev = hwmon_device_register_with_groups(adev->dev, DRIVER_NAME, adev, hwmon_groups); diff --git a/drivers/gpu/drm/amd/pm/legacy-dpm/kv_dpm.c b/drivers/gpu/drm/amd/pm/legacy-dpm/kv_dpm.c index 8fd0782a2b20..f5e08b60f66e 100644 --- a/drivers/gpu/drm/amd/pm/legacy-dpm/kv_dpm.c +++ b/drivers/gpu/drm/amd/pm/legacy-dpm/kv_dpm.c @@ -1384,13 +1384,16 @@ static int kv_dpm_enable(struct amdgpu_device *adev) static void kv_dpm_disable(struct amdgpu_device *adev) { struct kv_power_info *pi = kv_get_pi(adev); + int err; amdgpu_irq_put(adev, &adev->pm.dpm.thermal.irq, AMDGPU_THERMAL_IRQ_LOW_TO_HIGH); amdgpu_irq_put(adev, &adev->pm.dpm.thermal.irq, AMDGPU_THERMAL_IRQ_HIGH_TO_LOW); - amdgpu_kv_smc_bapm_enable(adev, false); + err = amdgpu_kv_smc_bapm_enable(adev, false); + if (err) + DRM_ERROR("amdgpu_kv_smc_bapm_enable failed\n"); if (adev->asic_type == CHIP_MULLINS) kv_enable_nb_dpm(adev, false); diff --git a/drivers/gpu/drm/amd/pm/powerplay/amd_powerplay.c b/drivers/gpu/drm/amd/pm/powerplay/amd_powerplay.c index 1eb4e613b27a..ec055858eb95 100644 --- a/drivers/gpu/drm/amd/pm/powerplay/amd_powerplay.c +++ b/drivers/gpu/drm/amd/pm/powerplay/amd_powerplay.c @@ -1485,6 +1485,7 @@ static int pp_get_prv_buffer_details(void *handle, void **addr, size_t *size) { struct pp_hwmgr *hwmgr = handle; struct amdgpu_device *adev = hwmgr->adev; + int err; if (!addr || !size) return -EINVAL; @@ -1492,7 +1493,9 @@ static int pp_get_prv_buffer_details(void *handle, void **addr, size_t *size) *addr = NULL; *size = 0; if (adev->pm.smu_prv_buffer) { - amdgpu_bo_kmap(adev->pm.smu_prv_buffer, addr); + err = amdgpu_bo_kmap(adev->pm.smu_prv_buffer, addr); + if (err) + return err; *size = adev->pm.smu_prv_buffer_size; } diff --git a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/smu7_hwmgr.c b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/smu7_hwmgr.c index e4fcbf8a7eb5..7ef7e81525a3 100644 --- a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/smu7_hwmgr.c +++ b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/smu7_hwmgr.c @@ -3603,7 +3603,7 @@ static int smu7_get_pp_table_entry_callback_func_v1(struct pp_hwmgr *hwmgr, return -EINVAL); PP_ASSERT_WITH_CODE( - (smu7_power_state->performance_level_count <= + (smu7_power_state->performance_level_count < hwmgr->platform_descriptor.hardwareActivityPerformanceLevels), "Performance levels exceeds Driver limit!", return -EINVAL); diff --git a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega10_hwmgr.c b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega10_hwmgr.c index 99bfe5efe171..c8c9fb827bda 100644 --- a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega10_hwmgr.c +++ b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega10_hwmgr.c @@ -3155,7 +3155,7 @@ static int vega10_get_pp_table_entry_callback_func(struct pp_hwmgr *hwmgr, return -1); PP_ASSERT_WITH_CODE( - (vega10_ps->performance_level_count <= + (vega10_ps->performance_level_count < hwmgr->platform_descriptor. hardwareActivityPerformanceLevels), "Performance levels exceeds Driver limit!", diff --git a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega10_thermal.c b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega10_thermal.c index 190af79f3236..dad3e3741a4e 100644 --- a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega10_thermal.c +++ b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega10_thermal.c @@ -67,21 +67,22 @@ int vega10_fan_ctrl_get_fan_speed_info(struct pp_hwmgr *hwmgr, int vega10_fan_ctrl_get_fan_speed_pwm(struct pp_hwmgr *hwmgr, uint32_t *speed) { - struct amdgpu_device *adev = hwmgr->adev; - uint32_t duty100, duty; - uint64_t tmp64; + uint32_t current_rpm; + uint32_t percent = 0; - duty100 = REG_GET_FIELD(RREG32_SOC15(THM, 0, mmCG_FDO_CTRL1), - CG_FDO_CTRL1, FMAX_DUTY100); - duty = REG_GET_FIELD(RREG32_SOC15(THM, 0, mmCG_THERMAL_STATUS), - CG_THERMAL_STATUS, FDO_PWM_DUTY); + if (hwmgr->thermal_controller.fanInfo.bNoFan) + return 0; - if (!duty100) - return -EINVAL; + if (vega10_get_current_rpm(hwmgr, ¤t_rpm)) + return -1; + + if (hwmgr->thermal_controller. + advanceFanControlParameters.usMaxFanRPM != 0) + percent = current_rpm * 255 / + hwmgr->thermal_controller. + advanceFanControlParameters.usMaxFanRPM; - tmp64 = (uint64_t)duty * 255; - do_div(tmp64, duty100); - *speed = MIN((uint32_t)tmp64, 255); + *speed = MIN(percent, 255); return 0; } diff --git a/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c b/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c index 13c5c7f1ecb9..4fe75dd2b329 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c +++ b/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c @@ -1314,8 +1314,8 @@ static int smu_smc_hw_setup(struct smu_context *smu) ret = smu_enable_thermal_alert(smu); if (ret) { - dev_err(adev->dev, "Failed to enable thermal alert!\n"); - return ret; + dev_err(adev->dev, "Failed to enable thermal alert!\n"); + return ret; } ret = smu_notify_display_change(smu); diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu13_driver_if_v13_0_4.h b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu13_driver_if_v13_0_4.h index ae2d337158f3..f77401709d83 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu13_driver_if_v13_0_4.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu13_driver_if_v13_0_4.h @@ -27,7 +27,7 @@ // *** IMPORTANT *** // SMU TEAM: Always increment the interface version if // any structure is changed in this file -#define PMFW_DRIVER_IF_VERSION 5 +#define PMFW_DRIVER_IF_VERSION 7 typedef struct { int32_t value; @@ -163,8 +163,8 @@ typedef struct { uint16_t DclkFrequency; //[MHz] uint16_t MemclkFrequency; //[MHz] uint16_t spare; //[centi] - uint16_t UvdActivity; //[centi] uint16_t GfxActivity; //[centi] + uint16_t UvdActivity; //[centi] uint16_t Voltage[2]; //[mV] indices: VDDCR_VDD, VDDCR_SOC uint16_t Current[2]; //[mA] indices: VDDCR_VDD, VDDCR_SOC @@ -199,6 +199,19 @@ typedef struct { uint16_t DeviceState; uint16_t CurTemp; //[centi-Celsius] uint16_t spare2; + + uint16_t AverageGfxclkFrequency; + uint16_t AverageFclkFrequency; + uint16_t AverageGfxActivity; + uint16_t AverageSocclkFrequency; + uint16_t AverageVclkFrequency; + uint16_t AverageVcnActivity; + uint16_t AverageDRAMReads; //Filtered DF Bandwidth::DRAM Reads + uint16_t AverageDRAMWrites; //Filtered DF Bandwidth::DRAM Writes + uint16_t AverageSocketPower; //Filtered value of CurrentSocketPower + uint16_t AverageCorePower; //Filtered of [sum of CorePower[8]]) + uint16_t AverageCoreC0Residency[8]; //Filtered of [average C0 residency % per core] + uint32_t MetricsCounter; //Counts the # of metrics table parameter reads per update to the metrics table, i.e. if the metrics table update happens every 1 second, this value could be up to 1000 if the smu collected metrics data every cycle, or as low as 0 if the smu was asleep the whole time. Reset to 0 after writing. } SmuMetrics_t; typedef struct { diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v13_0.h b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v13_0.h index 9d62ea2af132..8f72202aea8e 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v13_0.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v13_0.h @@ -28,7 +28,7 @@ #define SMU13_DRIVER_IF_VERSION_INV 0xFFFFFFFF #define SMU13_DRIVER_IF_VERSION_YELLOW_CARP 0x04 #define SMU13_DRIVER_IF_VERSION_ALDE 0x08 -#define SMU13_DRIVER_IF_VERSION_SMU_V13_0_4 0x05 +#define SMU13_DRIVER_IF_VERSION_SMU_V13_0_4 0x07 #define SMU13_DRIVER_IF_VERSION_SMU_V13_0_5 0x04 #define SMU13_DRIVER_IF_VERSION_SMU_V13_0_0 0x30 #define SMU13_DRIVER_IF_VERSION_SMU_V13_0_7 0x2C diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu11/arcturus_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu11/arcturus_ppt.c index 445005571f76..9cd005131f56 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu11/arcturus_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/arcturus_ppt.c @@ -2242,9 +2242,17 @@ static void arcturus_get_unique_id(struct smu_context *smu) static int arcturus_set_df_cstate(struct smu_context *smu, enum pp_df_cstate state) { + struct amdgpu_device *adev = smu->adev; uint32_t smu_version; int ret; + /* + * Arcturus does not need the cstate disablement + * prerequisite for gpu reset. + */ + if (amdgpu_in_reset(adev) || adev->in_suspend) + return 0; + ret = smu_cmn_get_smc_version(smu, NULL, &smu_version); if (ret) { dev_err(smu->adev->dev, "Failed to get smu version!\n"); diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/aldebaran_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/aldebaran_ppt.c index 619aee51b123..d30ec3005ea1 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu13/aldebaran_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/aldebaran_ppt.c @@ -1640,6 +1640,15 @@ static bool aldebaran_is_baco_supported(struct smu_context *smu) static int aldebaran_set_df_cstate(struct smu_context *smu, enum pp_df_cstate state) { + struct amdgpu_device *adev = smu->adev; + + /* + * Aldebaran does not need the cstate disablement + * prerequisite for gpu reset. + */ + if (amdgpu_in_reset(adev) || adev->in_suspend) + return 0; + return smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_DFCstateControl, state, NULL); } diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0.c index 93fffdbab4f0..c4552ade8d44 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0.c @@ -211,7 +211,8 @@ int smu_v13_0_init_pptable_microcode(struct smu_context *smu) return 0; if ((adev->ip_versions[MP1_HWIP][0] == IP_VERSION(13, 0, 7)) || - (adev->ip_versions[MP1_HWIP][0] == IP_VERSION(13, 0, 0))) + (adev->ip_versions[MP1_HWIP][0] == IP_VERSION(13, 0, 0)) || + (adev->ip_versions[MP1_HWIP][0] == IP_VERSION(13, 0, 10))) return 0; /* override pptable_id from driver parameter */ @@ -454,9 +455,6 @@ int smu_v13_0_setup_pptable(struct smu_context *smu) dev_info(adev->dev, "override pptable id %d\n", pptable_id); } else { pptable_id = smu->smu_table.boot_values.pp_table_id; - - if (adev->ip_versions[MP1_HWIP][0] == IP_VERSION(13, 0, 10)) - pptable_id = 6666; } /* force using vbios pptable in sriov mode */ diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c index 096327513dd0..29529328152d 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c @@ -119,6 +119,7 @@ static struct cmn2asic_msg_mapping smu_v13_0_0_message_map[SMU_MSG_MAX_COUNT] = MSG_MAP(NotifyPowerSource, PPSMC_MSG_NotifyPowerSource, 0), MSG_MAP(Mode1Reset, PPSMC_MSG_Mode1Reset, 0), MSG_MAP(PrepareMp1ForUnload, PPSMC_MSG_PrepareMp1ForUnload, 0), + MSG_MAP(DFCstateControl, PPSMC_MSG_SetExternalClientDfCstateAllow, 0), }; static struct cmn2asic_mapping smu_v13_0_0_clk_map[SMU_CLK_COUNT] = { @@ -239,82 +240,47 @@ smu_v13_0_0_get_allowed_feature_mask(struct smu_context *smu, uint32_t *feature_mask, uint32_t num) { struct amdgpu_device *adev = smu->adev; + u32 smu_version; if (num > 2) return -EINVAL; - memset(feature_mask, 0, sizeof(uint32_t) * num); - - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_FW_DATA_READ_BIT); + memset(feature_mask, 0xff, sizeof(uint32_t) * num); - if (adev->pm.pp_feature & PP_SCLK_DPM_MASK) { - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DPM_GFXCLK_BIT); - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_GFX_IMU_BIT); + if (!(adev->pm.pp_feature & PP_SCLK_DPM_MASK)) { + *(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_DPM_GFXCLK_BIT); + *(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_GFX_IMU_BIT); } - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_MM_DPM_BIT); - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DS_VCN_BIT); - - if ((adev->pg_flags & AMD_PG_SUPPORT_ATHUB) && - (adev->pg_flags & AMD_PG_SUPPORT_MMHUB)) - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_ATHUB_MMHUB_PG_BIT); - - if (adev->pm.pp_feature & PP_SOCCLK_DPM_MASK) - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DPM_SOCCLK_BIT); - -#if 0 - if (adev->pm.pp_feature & PP_GFXOFF_MASK) - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_GFXOFF_BIT); -#endif - - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_THROTTLERS_BIT); - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_FAN_CONTROL_BIT); + if (!(adev->pg_flags & AMD_PG_SUPPORT_ATHUB) || + !(adev->pg_flags & AMD_PG_SUPPORT_MMHUB)) + *(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_ATHUB_MMHUB_PG_BIT); - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DF_CSTATE_BIT); + if (!(adev->pm.pp_feature & PP_SOCCLK_DPM_MASK)) + *(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_DPM_SOCCLK_BIT); - if (adev->pm.pp_feature & PP_MCLK_DPM_MASK) { - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DPM_UCLK_BIT); - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_VMEMP_SCALING_BIT); - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_VDDIO_MEM_SCALING_BIT); + /* PMFW 78.58 contains a critical fix for gfxoff feature */ + smu_cmn_get_smc_version(smu, NULL, &smu_version); + if ((smu_version < 0x004e3a00) || + !(adev->pm.pp_feature & PP_GFXOFF_MASK)) + *(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_GFXOFF_BIT); + + if (!(adev->pm.pp_feature & PP_MCLK_DPM_MASK)) { + *(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_DPM_UCLK_BIT); + *(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_VMEMP_SCALING_BIT); + *(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_VDDIO_MEM_SCALING_BIT); } - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_MEM_TEMP_READ_BIT); - - if (adev->pm.pp_feature & PP_SCLK_DEEP_SLEEP_MASK) - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DS_GFXCLK_BIT); - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DS_SOCCLK_BIT); - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_SOC_MPCLK_DS_BIT); - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_BACO_MPCLK_DS_BIT); + if (!(adev->pm.pp_feature & PP_SCLK_DEEP_SLEEP_MASK)) + *(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_DS_GFXCLK_BIT); - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DPM_FCLK_BIT); - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DPM_DCN_BIT); - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DS_DCFCLK_BIT); - - if (adev->pm.pp_feature & PP_PCIE_DPM_MASK) { - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DPM_LINK_BIT); - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DS_LCLK_BIT); + if (!(adev->pm.pp_feature & PP_PCIE_DPM_MASK)) { + *(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_DPM_LINK_BIT); + *(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_DS_LCLK_BIT); } - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_BACO_BIT); - - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DPM_MP0CLK_BIT); - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_FW_DSTATE_BIT); - - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_OUT_OF_BAND_MONITOR_BIT); - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_SOC_CG_BIT); - - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DS_FCLK_BIT); - - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_FW_CTF_BIT); - - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DS_UCLK_BIT); - - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_VR0HOT_BIT); - - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DPM_GFX_POWER_OPTIMIZER_BIT); - - if (adev->pm.pp_feature & PP_ULV_MASK) - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_GFX_ULV_BIT); + if (!(adev->pm.pp_feature & PP_ULV_MASK)) + *(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_GFX_ULV_BIT); return 0; } @@ -1788,6 +1754,15 @@ static int smu_v13_0_0_set_mp1_state(struct smu_context *smu, return ret; } +static int smu_v13_0_0_set_df_cstate(struct smu_context *smu, + enum pp_df_cstate state) +{ + return smu_cmn_send_smc_msg_with_param(smu, + SMU_MSG_DFCstateControl, + state, + NULL); +} + static const struct pptable_funcs smu_v13_0_0_ppt_funcs = { .get_allowed_feature_mask = smu_v13_0_0_get_allowed_feature_mask, .set_default_dpm_table = smu_v13_0_0_set_default_dpm_table, @@ -1857,6 +1832,7 @@ static const struct pptable_funcs smu_v13_0_0_ppt_funcs = { .mode1_reset_is_support = smu_v13_0_0_is_mode1_reset_supported, .mode1_reset = smu_v13_0_mode1_reset, .set_mp1_state = smu_v13_0_0_set_mp1_state, + .set_df_cstate = smu_v13_0_0_set_df_cstate, }; void smu_v13_0_0_set_ppt_funcs(struct smu_context *smu) diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c index c422bf8a09b1..c4102cfb734c 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c @@ -121,6 +121,7 @@ static struct cmn2asic_msg_mapping smu_v13_0_7_message_map[SMU_MSG_MAX_COUNT] = MSG_MAP(Mode1Reset, PPSMC_MSG_Mode1Reset, 0), MSG_MAP(PrepareMp1ForUnload, PPSMC_MSG_PrepareMp1ForUnload, 0), MSG_MAP(SetMGpuFanBoostLimitRpm, PPSMC_MSG_SetMGpuFanBoostLimitRpm, 0), + MSG_MAP(DFCstateControl, PPSMC_MSG_SetExternalClientDfCstateAllow, 0), }; static struct cmn2asic_mapping smu_v13_0_7_clk_map[SMU_CLK_COUNT] = { @@ -1587,6 +1588,16 @@ static bool smu_v13_0_7_is_mode1_reset_supported(struct smu_context *smu) return true; } + +static int smu_v13_0_7_set_df_cstate(struct smu_context *smu, + enum pp_df_cstate state) +{ + return smu_cmn_send_smc_msg_with_param(smu, + SMU_MSG_DFCstateControl, + state, + NULL); +} + static const struct pptable_funcs smu_v13_0_7_ppt_funcs = { .get_allowed_feature_mask = smu_v13_0_7_get_allowed_feature_mask, .set_default_dpm_table = smu_v13_0_7_set_default_dpm_table, @@ -1649,6 +1660,7 @@ static const struct pptable_funcs smu_v13_0_7_ppt_funcs = { .mode1_reset_is_support = smu_v13_0_7_is_mode1_reset_supported, .mode1_reset = smu_v13_0_mode1_reset, .set_mp1_state = smu_v13_0_7_set_mp1_state, + .set_df_cstate = smu_v13_0_7_set_df_cstate, }; void smu_v13_0_7_set_ppt_funcs(struct smu_context *smu) |