diff options
author | Dave Airlie <airlied@redhat.com> | 2022-10-04 09:42:23 +1000 |
---|---|---|
committer | Dave Airlie <airlied@redhat.com> | 2022-10-04 09:42:24 +1000 |
commit | 65898687cf7392c372ea8d04a88617e2cb794465 (patch) | |
tree | b524feb316f5a12f6f799b2fa14f3678966a9111 | |
parent | 4ae9f874dc1d662ce7bfdb8144903608bcc3706b (diff) | |
parent | b292cafe2dd02d96a07147e4b160927e8399d5cc (diff) | |
download | linux-65898687cf7392c372ea8d04a88617e2cb794465.tar.bz2 |
Merge tag 'amd-drm-next-6.1-2022-09-30' of https://gitlab.freedesktop.org/agd5f/linux into drm-next
amd-drm-next-6.1-2022-09-30:
amdgpu:
- RLC FW code cleanup
- RLC fixes for GC 11.x
- SMU 13.x fixes
- CP FW code cleanup
- SDMA FW code cleanup
- GC 11.x fixes
- DCN 3.2.x fixes
- DCN 3.1.4 fixes
- Misc fixes
- RAS fixes
- SR-IOV fixes
- VCN 4.x fixes
amdkfd:
- GC 11.x fixes
- Xnack fixes
- UBSAN warning fix
Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20220930162012.5823-1-alexander.deucher@amd.com
82 files changed, 2116 insertions, 1987 deletions
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..9e98f3866edc 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c @@ -756,11 +756,7 @@ 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_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index c04ea7f1e819..ab8f970b2849 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -3154,7 +3154,8 @@ 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 +4070,20 @@ 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 +4107,9 @@ 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 +4128,12 @@ 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 +4148,13 @@ 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; @@ -5576,9 +5601,9 @@ 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 && 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_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_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_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_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c index e55f106621ef..ccebd8e2a2d8 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c @@ -2720,7 +2720,8 @@ 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 +2834,8 @@ 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 +2874,22 @@ 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 address 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, 0, ch_inst, umc_inst, m->addr); if (amdgpu_bad_page_threshold != 0) { amdgpu_ras_add_bad_pages(adev, err_data.err_addr, @@ -2898,6 +2897,7 @@ 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_reset.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_reset.h index dc43fcb93eac..f5318fedf2f0 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_reset.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_reset.h @@ -113,7 +113,8 @@ 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_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..3949b7e3907f 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 +151,135 @@ 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)) + return -EINVAL; + + 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: + return -EINVAL; + } + } + +out: + if (err) { + DRM_ERROR("SDMA: Failed to init firmware \"%s\"\n", fw_name); + amdgpu_sdma_destroy_inst_ctx(adev, duplicate); + } + return err; +} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.h index 53ac3ebae8d6..d2d88279fefb 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_sdma.h @@ -124,4 +124,8 @@ 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); #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.h index 3629d8f292ef..2fb4951a6433 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.h @@ -22,6 +22,8 @@ #define __AMDGPU_UMC_H__ #include "amdgpu_ras.h" +#define UMC_INVALID_ADDR 0x1ULL + /* * (addr / 256) * 4096, the higher 26 bits in ErrorAddr * is the index of 4KB block @@ -51,6 +53,10 @@ 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, + uint32_t umc_reg_offset, uint32_t ch_inst, + uint32_t umc_inst, uint64_t mca_addr); 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/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..251109723ab6 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; @@ -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 = { 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/mes_v11_0.c b/drivers/gpu/drm/amd/amdgpu/mes_v11_0.c index b64cd46a159a..b48684db2832 100644 --- a/drivers/gpu/drm/amd/amdgpu/mes_v11_0.c +++ b/drivers/gpu/drm/amd/amdgpu/mes_v11_0.c @@ -185,7 +185,15 @@ static int mes_v11_0_add_hw_queue(struct amdgpu_mes *mes, mes_add_queue_pkt.trap_handler_addr = input->tba_addr; mes_add_queue_pkt.tma_addr = input->tma_addr; mes_add_queue_pkt.is_kfd_process = input->is_kfd_process; - mes_add_queue_pkt.trap_en = 1; + + 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), diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c index 7fe8bf3417db..7241a9fb0121 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; } /** @@ -2000,7 +1927,11 @@ 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; } diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c index a019ac92edb7..c05c3eebde4c 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); - } - } -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; - } + ret = amdgpu_sdma_init_microcode(adev, fw_name, i, false); + if (ret) + return ret; } - return err; + + return ret; } static unsigned sdma_v5_0_ring_init_cond_exec(struct amdgpu_ring *ring) @@ -1465,12 +1436,10 @@ 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; } diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c b/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c index 95689ef4be10..f136fec7b4f4 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) @@ -1406,7 +1341,7 @@ 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; } diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v6_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v6_0.c index 3a0b14be1a57..db51230163c5 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) @@ -1370,7 +1301,7 @@ 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; } diff --git a/drivers/gpu/drm/amd/amdgpu/soc21.c b/drivers/gpu/drm/amd/amdgpu/soc21.c index 5f0d6983714a..16b757664a35 100644 --- a/drivers/gpu/drm/amd/amdgpu/soc21.c +++ b/drivers/gpu/drm/amd/amdgpu/soc21.c @@ -629,6 +629,7 @@ 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_DPG | AMD_PG_SUPPORT_JPEG; adev->external_rev_id = adev->rev_id + 0x1; break; diff --git a/drivers/gpu/drm/amd/amdgpu/umc_v6_7.c b/drivers/gpu/drm/amd/amdgpu/umc_v6_7.c index bf7524f16b66..a0d19b768346 100644 --- a/drivers/gpu/drm/amd/amdgpu/umc_v6_7.c +++ b/drivers/gpu/drm/amd/amdgpu/umc_v6_7.c @@ -452,41 +452,47 @@ 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_inst) + uint32_t umc_reg_offset, uint32_t ch_inst, + uint32_t umc_inst, uint64_t mca_addr) { uint32_t mc_umc_status_addr; uint32_t channel_index; - uint64_t mc_umc_status, mc_umc_addrt0; + uint64_t mc_umc_status = 0, mc_umc_addrt0; uint64_t err_addr, soc_pa, retired_page, column; - mc_umc_status_addr = - SOC15_REG_OFFSET(UMC, 0, regMCA_UMC_UMC0_MCUMC_STATUST0); - mc_umc_addrt0 = - SOC15_REG_OFFSET(UMC, 0, regMCA_UMC_UMC0_MCUMC_ADDRT0); + if (mca_addr == UMC_INVALID_ADDR) { + mc_umc_status_addr = + SOC15_REG_OFFSET(UMC, 0, regMCA_UMC_UMC0_MCUMC_STATUST0); + mc_umc_addrt0 = + SOC15_REG_OFFSET(UMC, 0, regMCA_UMC_UMC0_MCUMC_ADDRT0); - mc_umc_status = RREG64_PCIE((mc_umc_status_addr + umc_reg_offset) * 4); + mc_umc_status = RREG64_PCIE((mc_umc_status_addr + umc_reg_offset) * 4); - if (mc_umc_status == 0) - return; + if (mc_umc_status == 0) + return; - if (!err_data->err_addr) { - /* clear umc status */ - WREG64_PCIE((mc_umc_status_addr + umc_reg_offset) * 4, 0x0ULL); - return; + if (!err_data->err_addr) { + /* clear umc status */ + WREG64_PCIE((mc_umc_status_addr + umc_reg_offset) * 4, 0x0ULL); + 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 */ - if (REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, Val) == 1 && + 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)) { - - err_addr = RREG64_PCIE((mc_umc_addrt0 + umc_reg_offset) * 4); - err_addr = REG_GET_FIELD(err_addr, MCA_UMC_UMC0_MCUMC_ADDRT0, ErrorAddr); + REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, CECC) == 1)) || + mca_addr != UMC_INVALID_ADDR) { + if (mca_addr == UMC_INVALID_ADDR) { + err_addr = RREG64_PCIE((mc_umc_addrt0 + umc_reg_offset) * 4); + err_addr = + REG_GET_FIELD(err_addr, MCA_UMC_UMC0_MCUMC_ADDRT0, ErrorAddr); + } else { + err_addr = mca_addr; + } /* translate umc channel address to soc pa, 3 parts are included */ soc_pa = ADDR_OF_8KB_BLOCK(err_addr) | @@ -501,7 +507,8 @@ static void umc_v6_7_query_error_address(struct amdgpu_device *adev, /* we only save ue error information currently, ce is skipped */ if (REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, UECC) - == 1) { + == 1 || + mca_addr != UMC_INVALID_ADDR) { /* 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); @@ -519,7 +526,8 @@ static void umc_v6_7_query_error_address(struct amdgpu_device *adev, } /* clear umc status */ - WREG64_PCIE((mc_umc_status_addr + umc_reg_offset) * 4, 0x0ULL); + if (mca_addr == UMC_INVALID_ADDR) + WREG64_PCIE((mc_umc_status_addr + umc_reg_offset) * 4, 0x0ULL); } static void umc_v6_7_query_ras_error_address(struct amdgpu_device *adev, @@ -540,9 +548,8 @@ 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_inst); + umc_reg_offset, ch_inst, + umc_inst, UMC_INVALID_ADDR); } } @@ -583,4 +590,5 @@ 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_query_error_address, }; 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_mqd_manager_v11.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v11.c index d982c154537e..26b53b6d673e 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; diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c index cf5b4005534c..f5913ba22174 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); } @@ -2956,6 +2956,64 @@ 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; 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_psr.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.c index c8da18e45b0e..8ca10ab3dfc1 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 @@ -170,7 +170,13 @@ 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..53b077b40d72 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" @@ -1374,7 +1375,7 @@ 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 +1391,7 @@ 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 +1425,11 @@ 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 +3000,22 @@ 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/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/dcn32/dcn32_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn32/dcn32_clk_mgr.c index c6785969eb1a..f0f3f66629cc 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; + unsigned int num_dcfclk_levels, num_dtbclk_levels, num_dispclk_levels; 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; @@ -179,6 +181,7 @@ void dcn32_init_clocks(struct clk_mgr *clk_mgr_base) dcn32_init_single_clock(clk_mgr, PPCLK_DCFCLK, &clk_mgr_base->bw_params->clk_table.entries[0].dcfclk_mhz, &num_levels); + num_dcfclk_levels = num_levels; /* SOCCLK */ dcn32_init_single_clock(clk_mgr, PPCLK_SOCCLK, @@ -189,11 +192,16 @@ void dcn32_init_clocks(struct clk_mgr *clk_mgr_base) dcn32_init_single_clock(clk_mgr, PPCLK_DTBCLK, &clk_mgr_base->bw_params->clk_table.entries[0].dtbclk_mhz, &num_levels); + num_dtbclk_levels = num_levels; /* DISPCLK */ dcn32_init_single_clock(clk_mgr, PPCLK_DISPCLK, &clk_mgr_base->bw_params->clk_table.entries[0].dispclk_mhz, &num_levels); + num_dispclk_levels = num_levels; + + if (num_dcfclk_levels && num_dtbclk_levels && num_dispclk_levels) + clk_mgr->dpm_present = true; if (clk_mgr_base->ctx->dc->debug.min_disp_clk_khz) { unsigned int i; @@ -658,6 +666,12 @@ static void dcn32_get_memclk_states_from_smu(struct clk_mgr *clk_mgr_base) &num_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..258ba5a872b1 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 */ @@ -1750,6 +1750,9 @@ 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 @@ -1840,6 +1843,9 @@ 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 +2009,9 @@ 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 +2331,13 @@ 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 +2772,8 @@ 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; @@ -2983,13 +2993,8 @@ 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 +3203,9 @@ 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); } @@ -3521,11 +3529,59 @@ 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; + + *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; + } + } + } + + 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; enum dc_status ret = DC_ERROR_UNEXPECTED; unsigned int i, j; @@ -3533,10 +3589,13 @@ 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; + dc_resource_state_copy_construct(transition_base_context, transition_context); //commit minimal state @@ -3557,20 +3616,22 @@ 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; 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 +3654,14 @@ 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; - - struct dc_stream_status *cur_stream_status = stream_get_status(dc->current_state, stream); + bool force_minimal_pipe_splitting; + bool is_plane_addition; - 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 +3678,7 @@ 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 +4065,7 @@ 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); } 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..c57df45e83ff 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__, @@ -954,7 +954,7 @@ 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)); @@ -963,7 +963,7 @@ 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 (is_repeater(link_training_setting, 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" @@ -1172,7 +1172,7 @@ 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 +1198,7 @@ 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 +1469,6 @@ 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 +1477,7 @@ 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 +1501,8 @@ 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 +1542,7 @@ 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 +1583,15 @@ 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 +1657,7 @@ 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__); @@ -2099,7 +2107,7 @@ 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 +2224,7 @@ 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 +2296,7 @@ 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 +2643,7 @@ enum link_training_result dc_link_dp_perform_link_training( link, link_settings, <_settings); + override_training_settings( link, &link->preferred_training_settings, @@ -2652,7 +2661,7 @@ 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 +3095,7 @@ 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 +3249,7 @@ 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 +4110,13 @@ 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 +4223,6 @@ 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 +5032,136 @@ 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; } - 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: "); + + 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; + } +} - /* 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; +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) + return LTTPR_MODE_NON_TRANSPARENT; + else + return LTTPR_MODE_TRANSPARENT; } - return is_lttpr_present; + + if (link->dc->config.allow_lttpr_non_transparent_mode.bits.DP1_4A && + link->dc->caps.extended_aux_timeout_support) + return LTTPR_MODE_NON_TRANSPARENT; + + return LTTPR_MODE_NON_LTTPR; +} + +enum lttpr_mode dp_decide_128b_132b_lttpr_mode(struct dc_link *link) +{ + return dp_is_lttpr_present(link) ? LTTPR_MODE_NON_TRANSPARENT : LTTPR_MODE_NON_LTTPR; } static bool get_usbc_cable_id(struct dc_link *link, union dp_cable_id *cable_id) @@ -5209,13 +5223,16 @@ 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); /* 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 +5362,10 @@ 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); @@ -6096,7 +6117,7 @@ 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, @@ -7218,7 +7239,7 @@ 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_stream.c b/drivers/gpu/drm/amd/display/dc/core/dc_stream.c index 0c85ab5933b4..ae13887756bf 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 @@ -519,7 +520,7 @@ 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++) { + for (i = 0, j = 0; i < stream->num_wb_info && j < MAX_DWB_PIPES; i++) { if (stream->writeback_info[i].wb_enabled) { if (i != j) /* trim the array */ diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h index 659c71778a64..2ecf36e6329b 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 "3.2.205" #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 { @@ -832,6 +835,7 @@ 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 +850,7 @@ 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 +1122,7 @@ 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; diff --git a/drivers/gpu/drm/amd/display/dc/dc_link.h b/drivers/gpu/drm/amd/display/dc/dc_link.h index d890a03910c7..bf5f9e2773bc 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_link.h +++ b/drivers/gpu/drm/amd/display/dc/dc_link.h @@ -127,6 +127,12 @@ 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; @@ -151,7 +157,6 @@ 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/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_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c index 287fdecc0b10..72521749c01d 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 @@ -3812,7 +3812,7 @@ void dcn10_calc_vupdate_position( { 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; + pipe_ctx->stream->periodic_interrupt.lines_offset; int vupdate_offset_from_vsync = dc->hwss.get_vupdate_offset_from_vsync(pipe_ctx); int start_position; @@ -3837,18 +3837,10 @@ void dcn10_calc_vupdate_position( 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; - - 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; - - switch (ref_point) { + switch (pipe_ctx->stream->periodic_interrupt.ref_point) { case START_V_UPDATE: dcn10_calc_vupdate_position( dc, @@ -3857,7 +3849,9 @@ static void dcn10_cal_vline_position( end_line); break; case START_V_SYNC: - // Suppose to do nothing because vsync is 0; + // vsync is line 0 so start_line is just the requested line offset + *start_line = pipe_ctx->stream->periodic_interrupt.lines_offset; + *end_line = *start_line + 2; break; default: ASSERT(0); @@ -3867,24 +3861,15 @@ static void dcn10_cal_vline_position( 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, &start_line, &end_line); - dcn10_cal_vline_position(dc, pipe_ctx, vline, &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/dcn30/dcn30_resource.c b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_resource.c index f6f3878c99b8..3a3b2ac791c7 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 = { 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..559e563d5bc1 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 = { 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..52fb2bf3d578 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, 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..0d2ffb692957 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; } +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/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_hubbub.c b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_hubbub.c index 0b70247a5d36..f6d3da475835 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) 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..a750343ca521 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 +206,116 @@ 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; + uint8_t i; + int 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 +345,9 @@ 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 +356,7 @@ 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 +387,23 @@ 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; @@ -773,7 +761,8 @@ 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); } } @@ -1303,7 +1292,7 @@ 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 +1364,58 @@ 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 ? true : false; +} + +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_resource.c b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_resource.c index f8c6afacf633..05de97ea855f 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, @@ -1846,7 +1849,7 @@ 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 +1868,13 @@ 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 +1931,10 @@ 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..55945cca2260 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) 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..a2a70a1572b7 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 @@ -278,7 +278,7 @@ 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) 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..aed0f689cbbf 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn321/dcn321_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn321/dcn321_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, 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..0571700f53f9 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, @@ -674,9 +674,10 @@ 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++; @@ -1119,7 +1120,9 @@ static void dcn32_full_validate_bw_helper(struct dc *dc, 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 +1591,28 @@ 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 { + 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; + } + } + } + + if (pipe->top_pipe) { + pipe->top_pipe->bottom_pipe = NULL; + } + pipe->bottom_pipe = NULL; pipe->next_odm_pipe = NULL; pipe->plane_state = NULL; @@ -2019,6 +2044,45 @@ 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) { 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/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/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_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/dmub/inc/dmub_cmd.h b/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h index 6b9a529e9f12..5d1aadade8a5 100644 --- a/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h +++ b/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h @@ -760,6 +760,11 @@ enum dmub_cmd_dpia_type { DMUB_CMD__DPIA_MST_ALLOC_SLOTS = 2, }; +enum dmub_cmd_header_sub_type { + DMUB_CMD__SUB_TYPE_GENERAL = 0, + DMUB_CMD__SUB_TYPE_CURSOR_POSITION = 1 +}; + #pragma pack(push, 1) /** 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/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/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/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/swsmu/smu13/smu_v13_0_0_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c index 096327513dd0..1d454485e0d9 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 @@ -239,82 +239,47 @@ 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->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 + if (!(adev->pm.pp_feature & PP_SOCCLK_DPM_MASK)) + *(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_DPM_SOCCLK_BIT); - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_THROTTLERS_BIT); - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_FAN_CONTROL_BIT); - - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DF_CSTATE_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; } |