diff options
Diffstat (limited to 'drivers/gpu/drm')
196 files changed, 35484 insertions, 2710 deletions
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_acp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_acp.c index f4c474a95875..71efcf38f11b 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_acp.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_acp.c @@ -57,6 +57,10 @@ #define ACP_I2S_COMP2_CAP_REG_OFFSET 0xa8 #define ACP_I2S_COMP1_PLAY_REG_OFFSET 0x6c #define ACP_I2S_COMP2_PLAY_REG_OFFSET 0x68 +#define ACP_BT_PLAY_REGS_START 0x14970 +#define ACP_BT_PLAY_REGS_END 0x14a24 +#define ACP_BT_COMP1_REG_OFFSET 0xac +#define ACP_BT_COMP2_REG_OFFSET 0xa8 #define mmACP_PGFSM_RETAIN_REG 0x51c9 #define mmACP_PGFSM_CONFIG_REG 0x51ca @@ -77,7 +81,7 @@ #define ACP_SOFT_RESET_DONE_TIME_OUT_VALUE 0x000000FF #define ACP_TIMEOUT_LOOP 0x000000FF -#define ACP_DEVS 3 +#define ACP_DEVS 4 #define ACP_SRC_ID 162 enum { @@ -316,14 +320,13 @@ static int acp_hw_init(void *handle) if (adev->acp.acp_cell == NULL) return -ENOMEM; - adev->acp.acp_res = kcalloc(4, sizeof(struct resource), GFP_KERNEL); - + adev->acp.acp_res = kcalloc(5, sizeof(struct resource), GFP_KERNEL); if (adev->acp.acp_res == NULL) { kfree(adev->acp.acp_cell); return -ENOMEM; } - i2s_pdata = kcalloc(2, sizeof(struct i2s_platform_data), GFP_KERNEL); + i2s_pdata = kcalloc(3, sizeof(struct i2s_platform_data), GFP_KERNEL); if (i2s_pdata == NULL) { kfree(adev->acp.acp_res); kfree(adev->acp.acp_cell); @@ -358,6 +361,20 @@ static int acp_hw_init(void *handle) i2s_pdata[1].i2s_reg_comp1 = ACP_I2S_COMP1_CAP_REG_OFFSET; i2s_pdata[1].i2s_reg_comp2 = ACP_I2S_COMP2_CAP_REG_OFFSET; + i2s_pdata[2].quirks = DW_I2S_QUIRK_COMP_REG_OFFSET; + switch (adev->asic_type) { + case CHIP_STONEY: + i2s_pdata[2].quirks |= DW_I2S_QUIRK_16BIT_IDX_OVERRIDE; + break; + default: + break; + } + + i2s_pdata[2].cap = DWC_I2S_PLAY | DWC_I2S_RECORD; + i2s_pdata[2].snd_rates = SNDRV_PCM_RATE_8000_96000; + i2s_pdata[2].i2s_reg_comp1 = ACP_BT_COMP1_REG_OFFSET; + i2s_pdata[2].i2s_reg_comp2 = ACP_BT_COMP2_REG_OFFSET; + adev->acp.acp_res[0].name = "acp2x_dma"; adev->acp.acp_res[0].flags = IORESOURCE_MEM; adev->acp.acp_res[0].start = acp_base; @@ -373,13 +390,18 @@ static int acp_hw_init(void *handle) adev->acp.acp_res[2].start = acp_base + ACP_I2S_CAP_REGS_START; adev->acp.acp_res[2].end = acp_base + ACP_I2S_CAP_REGS_END; - adev->acp.acp_res[3].name = "acp2x_dma_irq"; - adev->acp.acp_res[3].flags = IORESOURCE_IRQ; - adev->acp.acp_res[3].start = amdgpu_irq_create_mapping(adev, 162); - adev->acp.acp_res[3].end = adev->acp.acp_res[3].start; + adev->acp.acp_res[3].name = "acp2x_dw_bt_i2s_play_cap"; + adev->acp.acp_res[3].flags = IORESOURCE_MEM; + adev->acp.acp_res[3].start = acp_base + ACP_BT_PLAY_REGS_START; + adev->acp.acp_res[3].end = acp_base + ACP_BT_PLAY_REGS_END; + + adev->acp.acp_res[4].name = "acp2x_dma_irq"; + adev->acp.acp_res[4].flags = IORESOURCE_IRQ; + adev->acp.acp_res[4].start = amdgpu_irq_create_mapping(adev, 162); + adev->acp.acp_res[4].end = adev->acp.acp_res[4].start; adev->acp.acp_cell[0].name = "acp_audio_dma"; - adev->acp.acp_cell[0].num_resources = 4; + adev->acp.acp_cell[0].num_resources = 5; adev->acp.acp_cell[0].resources = &adev->acp.acp_res[0]; adev->acp.acp_cell[0].platform_data = &adev->asic_type; adev->acp.acp_cell[0].pdata_size = sizeof(adev->asic_type); @@ -396,6 +418,12 @@ static int acp_hw_init(void *handle) adev->acp.acp_cell[2].platform_data = &i2s_pdata[1]; adev->acp.acp_cell[2].pdata_size = sizeof(struct i2s_platform_data); + adev->acp.acp_cell[3].name = "designware-i2s"; + adev->acp.acp_cell[3].num_resources = 1; + adev->acp.acp_cell[3].resources = &adev->acp.acp_res[3]; + adev->acp.acp_cell[3].platform_data = &i2s_pdata[2]; + adev->acp.acp_cell[3].pdata_size = sizeof(struct i2s_platform_data); + r = mfd_add_hotplug_devices(adev->acp.parent, adev->acp.acp_cell, ACP_DEVS); if (r) @@ -451,7 +479,6 @@ static int acp_hw_init(void *handle) val = cgs_read_register(adev->acp.cgs_device, mmACP_SOFT_RESET); val &= ~ACP_SOFT_RESET__SoftResetAud_MASK; cgs_write_register(adev->acp.cgs_device, mmACP_SOFT_RESET, val); - return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c index e3ed08dca7b7..f8bbbb3a9504 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c @@ -243,6 +243,33 @@ int amdgpu_amdkfd_resume(struct amdgpu_device *adev) return r; } +int amdgpu_amdkfd_pre_reset(struct amdgpu_device *adev) +{ + int r = 0; + + if (adev->kfd) + r = kgd2kfd->pre_reset(adev->kfd); + + return r; +} + +int amdgpu_amdkfd_post_reset(struct amdgpu_device *adev) +{ + int r = 0; + + if (adev->kfd) + r = kgd2kfd->post_reset(adev->kfd); + + return r; +} + +void amdgpu_amdkfd_gpu_reset(struct kgd_dev *kgd) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)kgd; + + amdgpu_device_gpu_recover(adev, NULL, false); +} + int alloc_gtt_mem(struct kgd_dev *kgd, size_t size, void **mem_obj, uint64_t *gpu_addr, void **cpu_ptr) @@ -461,6 +488,14 @@ err: return ret; } +void amdgpu_amdkfd_set_compute_idle(struct kgd_dev *kgd, bool idle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)kgd; + + amdgpu_dpm_switch_power_profile(adev, + PP_SMC_POWER_PROFILE_COMPUTE, !idle); +} + bool amdgpu_amdkfd_is_kfd_vmid(struct amdgpu_device *adev, u32 vmid) { if (adev->kfd) { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h index a8418a3f4e9d..2f379c183ed2 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h @@ -119,6 +119,7 @@ int amdgpu_amdkfd_evict_userptr(struct kgd_mem *mem, struct mm_struct *mm); int amdgpu_amdkfd_submit_ib(struct kgd_dev *kgd, enum kgd_engine_type engine, uint32_t vmid, uint64_t gpu_addr, uint32_t *ib_cmd, uint32_t ib_len); +void amdgpu_amdkfd_set_compute_idle(struct kgd_dev *kgd, bool idle); struct kfd2kgd_calls *amdgpu_amdkfd_gfx_7_get_functions(void); struct kfd2kgd_calls *amdgpu_amdkfd_gfx_8_0_get_functions(void); @@ -126,6 +127,12 @@ struct kfd2kgd_calls *amdgpu_amdkfd_gfx_9_0_get_functions(void); bool amdgpu_amdkfd_is_kfd_vmid(struct amdgpu_device *adev, u32 vmid); +int amdgpu_amdkfd_pre_reset(struct amdgpu_device *adev); + +int amdgpu_amdkfd_post_reset(struct amdgpu_device *adev); + +void amdgpu_amdkfd_gpu_reset(struct kgd_dev *kgd); + /* Shared API */ int alloc_gtt_mem(struct kgd_dev *kgd, size_t size, void **mem_obj, uint64_t *gpu_addr, @@ -183,6 +190,9 @@ int amdgpu_amdkfd_gpuvm_map_gtt_bo_to_kernel(struct kgd_dev *kgd, int amdgpu_amdkfd_gpuvm_restore_process_bos(void *process_info, struct dma_fence **ef); +int amdgpu_amdkfd_gpuvm_get_vm_fault_info(struct kgd_dev *kgd, + struct kfd_vm_fault_info *info); + void amdgpu_amdkfd_gpuvm_init_mem_limits(void); void amdgpu_amdkfd_unreserve_system_memory_limit(struct amdgpu_bo *bo); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c index ea79908dac4c..ea3f698aef5e 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c @@ -145,6 +145,7 @@ static void set_vm_context_page_table_base(struct kgd_dev *kgd, uint32_t vmid, uint32_t page_table_base); static int invalidate_tlbs(struct kgd_dev *kgd, uint16_t pasid); static int invalidate_tlbs_vmid(struct kgd_dev *kgd, uint16_t vmid); +static uint32_t read_vmid_from_vmfault_reg(struct kgd_dev *kgd); /* Because of REG_GET_FIELD() being used, we put this function in the * asic specific file. @@ -216,6 +217,10 @@ static const struct kfd2kgd_calls kfd2kgd = { .invalidate_tlbs = invalidate_tlbs, .invalidate_tlbs_vmid = invalidate_tlbs_vmid, .submit_ib = amdgpu_amdkfd_submit_ib, + .get_vm_fault_info = amdgpu_amdkfd_gpuvm_get_vm_fault_info, + .read_vmid_from_vmfault_reg = read_vmid_from_vmfault_reg, + .gpu_recover = amdgpu_amdkfd_gpu_reset, + .set_compute_idle = amdgpu_amdkfd_set_compute_idle }; struct kfd2kgd_calls *amdgpu_amdkfd_gfx_7_get_functions(void) @@ -571,6 +576,9 @@ static int kgd_hqd_destroy(struct kgd_dev *kgd, void *mqd, unsigned long flags, end_jiffies; int retry; + if (adev->in_gpu_reset) + return -EIO; + acquire_queue(kgd, pipe_id, queue_id); WREG32(mmCP_HQD_PQ_DOORBELL_CONTROL, 0); @@ -882,6 +890,9 @@ static int invalidate_tlbs(struct kgd_dev *kgd, uint16_t pasid) int vmid; unsigned int tmp; + if (adev->in_gpu_reset) + return -EIO; + for (vmid = 0; vmid < 16; vmid++) { if (!amdgpu_amdkfd_is_kfd_vmid(adev, vmid)) continue; @@ -911,3 +922,19 @@ static int invalidate_tlbs_vmid(struct kgd_dev *kgd, uint16_t vmid) RREG32(mmVM_INVALIDATE_RESPONSE); return 0; } + + /** + * read_vmid_from_vmfault_reg - read vmid from register + * + * adev: amdgpu_device pointer + * @vmid: vmid pointer + * read vmid from register (CIK). + */ +static uint32_t read_vmid_from_vmfault_reg(struct kgd_dev *kgd) +{ + struct amdgpu_device *adev = get_amdgpu_device(kgd); + + uint32_t status = RREG32(mmVM_CONTEXT1_PROTECTION_FAULT_STATUS); + + return REG_GET_FIELD(status, VM_CONTEXT1_PROTECTION_FAULT_STATUS, VMID); +} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c index 19dd665e7307..f6e53e9352bd 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c @@ -176,6 +176,9 @@ static const struct kfd2kgd_calls kfd2kgd = { .invalidate_tlbs = invalidate_tlbs, .invalidate_tlbs_vmid = invalidate_tlbs_vmid, .submit_ib = amdgpu_amdkfd_submit_ib, + .get_vm_fault_info = amdgpu_amdkfd_gpuvm_get_vm_fault_info, + .gpu_recover = amdgpu_amdkfd_gpu_reset, + .set_compute_idle = amdgpu_amdkfd_set_compute_idle }; struct kfd2kgd_calls *amdgpu_amdkfd_gfx_8_0_get_functions(void) @@ -568,6 +571,9 @@ static int kgd_hqd_destroy(struct kgd_dev *kgd, void *mqd, int retry; struct vi_mqd *m = get_mqd(mqd); + if (adev->in_gpu_reset) + return -EIO; + acquire_queue(kgd, pipe_id, queue_id); if (m->cp_hqd_vmid == 0) @@ -844,6 +850,9 @@ static int invalidate_tlbs(struct kgd_dev *kgd, uint16_t pasid) int vmid; unsigned int tmp; + if (adev->in_gpu_reset) + return -EIO; + for (vmid = 0; vmid < 16; vmid++) { if (!amdgpu_amdkfd_is_kfd_vmid(adev, vmid)) continue; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c index 1db60aa5b7f0..8efedfcb9dfc 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c @@ -213,6 +213,8 @@ static const struct kfd2kgd_calls kfd2kgd = { .invalidate_tlbs = invalidate_tlbs, .invalidate_tlbs_vmid = invalidate_tlbs_vmid, .submit_ib = amdgpu_amdkfd_submit_ib, + .gpu_recover = amdgpu_amdkfd_gpu_reset, + .set_compute_idle = amdgpu_amdkfd_set_compute_idle }; struct kfd2kgd_calls *amdgpu_amdkfd_gfx_9_0_get_functions(void) @@ -679,6 +681,9 @@ static int kgd_hqd_destroy(struct kgd_dev *kgd, void *mqd, uint32_t temp; struct v9_mqd *m = get_mqd(mqd); + if (adev->in_gpu_reset) + return -EIO; + acquire_queue(kgd, pipe_id, queue_id); if (m->cp_hqd_vmid == 0) @@ -866,6 +871,9 @@ static int invalidate_tlbs(struct kgd_dev *kgd, uint16_t pasid) int vmid; struct amdgpu_ring *ring = &adev->gfx.kiq.ring; + if (adev->in_gpu_reset) + return -EIO; + if (ring->ready) return invalidate_tlbs_with_kiq(adev, pasid); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c index fa38a960ce00..8a707d8bbb1c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c @@ -1621,6 +1621,20 @@ bo_reserve_failed: return ret; } +int amdgpu_amdkfd_gpuvm_get_vm_fault_info(struct kgd_dev *kgd, + struct kfd_vm_fault_info *mem) +{ + struct amdgpu_device *adev; + + adev = (struct amdgpu_device *)kgd; + if (atomic_read(&adev->gmc.vm_fault_info_updated) == 1) { + *mem = *adev->gmc.vm_fault_info; + mb(); + atomic_set(&adev->gmc.vm_fault_info_updated, 0); + } + return 0; +} + /* Evict a userptr BO by stopping the queues if necessary * * Runs in MMU notifier, may be in RECLAIM_FS context. This means it diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c index b33f1680c9a3..a028661d9e20 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c @@ -575,6 +575,7 @@ static const struct amdgpu_px_quirk amdgpu_px_quirk_list[] = { { 0x1002, 0x6900, 0x1002, 0x0124, AMDGPU_PX_QUIRK_FORCE_ATPX }, { 0x1002, 0x6900, 0x1028, 0x0812, AMDGPU_PX_QUIRK_FORCE_ATPX }, { 0x1002, 0x6900, 0x1028, 0x0813, AMDGPU_PX_QUIRK_FORCE_ATPX }, + { 0x1002, 0x6900, 0x1025, 0x125A, AMDGPU_PX_QUIRK_FORCE_ATPX }, { 0, 0, 0, 0, 0 }, }; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index ec53d8f96d06..e839470880d7 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -2843,6 +2843,9 @@ int amdgpu_device_resume(struct drm_device *dev, bool resume, bool fbcon) if (r) return r; + /* Make sure IB tests flushed */ + flush_delayed_work(&adev->late_init_work); + /* blat the mode back in */ if (fbcon) { if (!amdgpu_device_has_dc_support(adev)) { @@ -3307,6 +3310,9 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev, atomic_inc(&adev->gpu_reset_counter); adev->in_gpu_reset = 1; + /* Block kfd */ + amdgpu_amdkfd_pre_reset(adev); + /* block TTM */ resched = ttm_bo_lock_delayed_workqueue(&adev->mman.bdev); @@ -3322,7 +3328,7 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev, if (job && job->base.sched == &ring->sched) continue; - drm_sched_hw_job_reset(&ring->sched, &job->base); + drm_sched_hw_job_reset(&ring->sched, job ? &job->base : NULL); /* after all hw jobs are reset, hw fence is meaningless, so force_completion */ amdgpu_fence_driver_force_completion(ring); @@ -3363,6 +3369,8 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev, dev_info(adev->dev, "GPU reset(%d) succeeded!\n",atomic_read(&adev->gpu_reset_counter)); } + /*unlock kfd */ + amdgpu_amdkfd_post_reset(adev); amdgpu_vf_error_trans_all(adev); adev->in_gpu_reset = 0; mutex_unlock(&adev->lock_reset); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h index 6cb4948233cb..bb5a47a45790 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h @@ -105,6 +105,8 @@ struct amdgpu_gmc { /* protects concurrent invalidation */ spinlock_t invalidate_lock; bool translate_further; + struct kfd_vm_fault_info *vm_fault_info; + atomic_t vm_fault_info_updated; const struct amdgpu_gmc_funcs *gmc_funcs; }; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index 015613b4f98b..ece0ac703e27 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -495,11 +495,12 @@ static int amdgpu_vm_alloc_levels(struct amdgpu_device *adev, eaddr = eaddr & ((1 << shift) - 1); flags = AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS; + if (vm->root.base.bo->shadow) + flags |= AMDGPU_GEM_CREATE_SHADOW; if (vm->use_cpu_for_update) flags |= AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED; else - flags |= (AMDGPU_GEM_CREATE_NO_CPU_ACCESS | - AMDGPU_GEM_CREATE_SHADOW); + flags |= AMDGPU_GEM_CREATE_NO_CPU_ACCESS; /* walk over the address space and allocate the page tables */ for (pt_idx = from; pt_idx <= to; ++pt_idx) { @@ -2616,7 +2617,7 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm, flags = AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS; if (vm->use_cpu_for_update) flags |= AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED; - else + else if (vm_context != AMDGPU_VM_CONTEXT_COMPUTE) flags |= AMDGPU_GEM_CREATE_SHADOW; size = amdgpu_vm_bo_size(adev, adev->vm_manager.root_level); @@ -2691,8 +2692,7 @@ error_free_sched_entity: * - pasid (old PASID is released, because compute manages its own PASIDs) * * Reinitializes the page directory to reflect the changed ATS - * setting. May leave behind an unused shadow BO for the page - * directory when switching from SDMA updates to CPU updates. + * setting. * * Returns: * 0 for success, -errno for errors. @@ -2742,6 +2742,9 @@ int amdgpu_vm_make_compute(struct amdgpu_device *adev, struct amdgpu_vm *vm) vm->pasid = 0; } + /* Free the shadow bo for compute VM */ + amdgpu_bo_unref(&vm->root.base.bo->shadow); + error: amdgpu_bo_unreserve(vm->root.base.bo); return r; diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c index 10920f0bd85f..36dc367c4b45 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c @@ -28,6 +28,7 @@ #include "cik.h" #include "gmc_v7_0.h" #include "amdgpu_ucode.h" +#include "amdgpu_amdkfd.h" #include "bif/bif_4_1_d.h" #include "bif/bif_4_1_sh_mask.h" @@ -1078,6 +1079,12 @@ static int gmc_v7_0_sw_init(void *handle) adev->vm_manager.vram_base_offset = 0; } + adev->gmc.vm_fault_info = kmalloc(sizeof(struct kfd_vm_fault_info), + GFP_KERNEL); + if (!adev->gmc.vm_fault_info) + return -ENOMEM; + atomic_set(&adev->gmc.vm_fault_info_updated, 0); + return 0; } @@ -1087,6 +1094,7 @@ static int gmc_v7_0_sw_fini(void *handle) amdgpu_gem_force_release(adev); amdgpu_vm_manager_fini(adev); + kfree(adev->gmc.vm_fault_info); gmc_v7_0_gart_fini(adev); amdgpu_bo_fini(adev); release_firmware(adev->gmc.fw); @@ -1276,7 +1284,7 @@ static int gmc_v7_0_process_interrupt(struct amdgpu_device *adev, struct amdgpu_irq_src *source, struct amdgpu_iv_entry *entry) { - u32 addr, status, mc_client; + u32 addr, status, mc_client, vmid; addr = RREG32(mmVM_CONTEXT1_PROTECTION_FAULT_ADDR); status = RREG32(mmVM_CONTEXT1_PROTECTION_FAULT_STATUS); @@ -1301,6 +1309,29 @@ static int gmc_v7_0_process_interrupt(struct amdgpu_device *adev, entry->pasid); } + vmid = REG_GET_FIELD(status, VM_CONTEXT1_PROTECTION_FAULT_STATUS, + VMID); + if (amdgpu_amdkfd_is_kfd_vmid(adev, vmid) + && !atomic_read(&adev->gmc.vm_fault_info_updated)) { + struct kfd_vm_fault_info *info = adev->gmc.vm_fault_info; + u32 protections = REG_GET_FIELD(status, + VM_CONTEXT1_PROTECTION_FAULT_STATUS, + PROTECTIONS); + + info->vmid = vmid; + info->mc_id = REG_GET_FIELD(status, + VM_CONTEXT1_PROTECTION_FAULT_STATUS, + MEMORY_CLIENT_ID); + info->status = status; + info->page_addr = addr; + info->prot_valid = protections & 0x7 ? true : false; + info->prot_read = protections & 0x8 ? true : false; + info->prot_write = protections & 0x10 ? true : false; + info->prot_exec = protections & 0x20 ? true : false; + mb(); + atomic_set(&adev->gmc.vm_fault_info_updated, 1); + } + return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c index 75f3ffb2891e..70fc97b59b4f 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c @@ -26,6 +26,7 @@ #include "amdgpu.h" #include "gmc_v8_0.h" #include "amdgpu_ucode.h" +#include "amdgpu_amdkfd.h" #include "gmc/gmc_8_1_d.h" #include "gmc/gmc_8_1_sh_mask.h" @@ -1182,6 +1183,12 @@ static int gmc_v8_0_sw_init(void *handle) adev->vm_manager.vram_base_offset = 0; } + adev->gmc.vm_fault_info = kmalloc(sizeof(struct kfd_vm_fault_info), + GFP_KERNEL); + if (!adev->gmc.vm_fault_info) + return -ENOMEM; + atomic_set(&adev->gmc.vm_fault_info_updated, 0); + return 0; } @@ -1191,6 +1198,7 @@ static int gmc_v8_0_sw_fini(void *handle) amdgpu_gem_force_release(adev); amdgpu_vm_manager_fini(adev); + kfree(adev->gmc.vm_fault_info); gmc_v8_0_gart_fini(adev); amdgpu_bo_fini(adev); release_firmware(adev->gmc.fw); @@ -1426,7 +1434,7 @@ static int gmc_v8_0_process_interrupt(struct amdgpu_device *adev, struct amdgpu_irq_src *source, struct amdgpu_iv_entry *entry) { - u32 addr, status, mc_client; + u32 addr, status, mc_client, vmid; if (amdgpu_sriov_vf(adev)) { dev_err(adev->dev, "GPU fault detected: %d 0x%08x\n", @@ -1463,6 +1471,29 @@ static int gmc_v8_0_process_interrupt(struct amdgpu_device *adev, entry->pasid); } + vmid = REG_GET_FIELD(status, VM_CONTEXT1_PROTECTION_FAULT_STATUS, + VMID); + if (amdgpu_amdkfd_is_kfd_vmid(adev, vmid) + && !atomic_read(&adev->gmc.vm_fault_info_updated)) { + struct kfd_vm_fault_info *info = adev->gmc.vm_fault_info; + u32 protections = REG_GET_FIELD(status, + VM_CONTEXT1_PROTECTION_FAULT_STATUS, + PROTECTIONS); + + info->vmid = vmid; + info->mc_id = REG_GET_FIELD(status, + VM_CONTEXT1_PROTECTION_FAULT_STATUS, + MEMORY_CLIENT_ID); + info->status = status; + info->page_addr = addr; + info->prot_valid = protections & 0x7 ? true : false; + info->prot_read = protections & 0x8 ? true : false; + info->prot_write = protections & 0x10 ? true : false; + info->prot_exec = protections & 0x20 ? true : false; + mb(); + atomic_set(&adev->gmc.vm_fault_info_updated, 1); + } + return 0; } diff --git a/drivers/gpu/drm/amd/amdkfd/cik_event_interrupt.c b/drivers/gpu/drm/amd/amdkfd/cik_event_interrupt.c index 49df6c791cfc..5d2475d5392c 100644 --- a/drivers/gpu/drm/amd/amdkfd/cik_event_interrupt.c +++ b/drivers/gpu/drm/amd/amdkfd/cik_event_interrupt.c @@ -25,12 +25,39 @@ #include "cik_int.h" static bool cik_event_interrupt_isr(struct kfd_dev *dev, - const uint32_t *ih_ring_entry) + const uint32_t *ih_ring_entry, + uint32_t *patched_ihre, + bool *patched_flag) { const struct cik_ih_ring_entry *ihre = (const struct cik_ih_ring_entry *)ih_ring_entry; + const struct kfd2kgd_calls *f2g = dev->kfd2kgd; unsigned int vmid, pasid; + /* This workaround is due to HW/FW limitation on Hawaii that + * VMID and PASID are not written into ih_ring_entry + */ + if ((ihre->source_id == CIK_INTSRC_GFX_PAGE_INV_FAULT || + ihre->source_id == CIK_INTSRC_GFX_MEM_PROT_FAULT) && + dev->device_info->asic_family == CHIP_HAWAII) { + struct cik_ih_ring_entry *tmp_ihre = + (struct cik_ih_ring_entry *)patched_ihre; + + *patched_flag = true; + *tmp_ihre = *ihre; + + vmid = f2g->read_vmid_from_vmfault_reg(dev->kgd); + pasid = f2g->get_atc_vmid_pasid_mapping_pasid(dev->kgd, vmid); + + tmp_ihre->ring_id &= 0x000000ff; + tmp_ihre->ring_id |= vmid << 8; + tmp_ihre->ring_id |= pasid << 16; + + return (pasid != 0) && + vmid >= dev->vm_info.first_vmid_kfd && + vmid <= dev->vm_info.last_vmid_kfd; + } + /* Only handle interrupts from KFD VMIDs */ vmid = (ihre->ring_id & 0x0000ff00) >> 8; if (vmid < dev->vm_info.first_vmid_kfd || @@ -48,18 +75,19 @@ static bool cik_event_interrupt_isr(struct kfd_dev *dev, return ihre->source_id == CIK_INTSRC_CP_END_OF_PIPE || ihre->source_id == CIK_INTSRC_SDMA_TRAP || ihre->source_id == CIK_INTSRC_SQ_INTERRUPT_MSG || - ihre->source_id == CIK_INTSRC_CP_BAD_OPCODE; + ihre->source_id == CIK_INTSRC_CP_BAD_OPCODE || + ihre->source_id == CIK_INTSRC_GFX_PAGE_INV_FAULT || + ihre->source_id == CIK_INTSRC_GFX_MEM_PROT_FAULT; } static void cik_event_interrupt_wq(struct kfd_dev *dev, const uint32_t *ih_ring_entry) { - unsigned int pasid; const struct cik_ih_ring_entry *ihre = (const struct cik_ih_ring_entry *)ih_ring_entry; uint32_t context_id = ihre->data & 0xfffffff; - - pasid = (ihre->ring_id & 0xffff0000) >> 16; + unsigned int vmid = (ihre->ring_id & 0x0000ff00) >> 8; + unsigned int pasid = (ihre->ring_id & 0xffff0000) >> 16; if (pasid == 0) return; @@ -72,6 +100,22 @@ static void cik_event_interrupt_wq(struct kfd_dev *dev, kfd_signal_event_interrupt(pasid, context_id & 0xff, 8); else if (ihre->source_id == CIK_INTSRC_CP_BAD_OPCODE) kfd_signal_hw_exception_event(pasid); + else if (ihre->source_id == CIK_INTSRC_GFX_PAGE_INV_FAULT || + ihre->source_id == CIK_INTSRC_GFX_MEM_PROT_FAULT) { + struct kfd_vm_fault_info info; + + kfd_process_vm_fault(dev->dqm, pasid); + + memset(&info, 0, sizeof(info)); + dev->kfd2kgd->get_vm_fault_info(dev->kgd, &info); + if (!info.page_addr && !info.status) + return; + + if (info.vmid == vmid) + kfd_signal_vm_fault_event(dev, pasid, &info); + else + kfd_signal_vm_fault_event(dev, pasid, NULL); + } } const struct kfd_event_interrupt_class event_interrupt_class_cik = { diff --git a/drivers/gpu/drm/amd/amdkfd/cik_int.h b/drivers/gpu/drm/amd/amdkfd/cik_int.h index 109298b9d507..76f8677a7926 100644 --- a/drivers/gpu/drm/amd/amdkfd/cik_int.h +++ b/drivers/gpu/drm/amd/amdkfd/cik_int.h @@ -20,8 +20,8 @@ * OTHER DEALINGS IN THE SOFTWARE. */ -#ifndef HSA_RADEON_CIK_INT_H_INCLUDED -#define HSA_RADEON_CIK_INT_H_INCLUDED +#ifndef CIK_INT_H_INCLUDED +#define CIK_INT_H_INCLUDED #include <linux/types.h> @@ -34,9 +34,10 @@ struct cik_ih_ring_entry { #define CIK_INTSRC_CP_END_OF_PIPE 0xB5 #define CIK_INTSRC_CP_BAD_OPCODE 0xB7 -#define CIK_INTSRC_DEQUEUE_COMPLETE 0xC6 #define CIK_INTSRC_SDMA_TRAP 0xE0 #define CIK_INTSRC_SQ_INTERRUPT_MSG 0xEF +#define CIK_INTSRC_GFX_PAGE_INV_FAULT 0x92 +#define CIK_INTSRC_GFX_MEM_PROT_FAULT 0x93 #endif diff --git a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler.h b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler.h index f68aef02fc1f..3621efbd5759 100644 --- a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler.h +++ b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler.h @@ -21,18 +21,21 @@ */ static const uint32_t cwsr_trap_gfx8_hex[] = { - 0xbf820001, 0xbf820125, + 0xbf820001, 0xbf82012b, 0xb8f4f802, 0x89748674, 0xb8f5f803, 0x8675ff75, - 0x00000400, 0xbf850011, + 0x00000400, 0xbf850017, 0xc00a1e37, 0x00000000, 0xbf8c007f, 0x87777978, - 0xbf840002, 0xb974f802, - 0xbe801d78, 0xb8f5f803, - 0x8675ff75, 0x000001ff, - 0xbf850002, 0x80708470, - 0x82718071, 0x8671ff71, - 0x0000ffff, 0xb974f802, + 0xbf840005, 0x8f728374, + 0xb972e0c2, 0xbf800002, + 0xb9740002, 0xbe801d78, + 0xb8f5f803, 0x8675ff75, + 0x000001ff, 0xbf850002, + 0x80708470, 0x82718071, + 0x8671ff71, 0x0000ffff, + 0x8f728374, 0xb972e0c2, + 0xbf800002, 0xb9740002, 0xbe801f70, 0xb8f5f803, 0x8675ff75, 0x00000100, 0xbf840006, 0xbefa0080, @@ -168,7 +171,7 @@ static const uint32_t cwsr_trap_gfx8_hex[] = { 0x807c847c, 0x806eff6e, 0x00000400, 0xbf0a757c, 0xbf85ffef, 0xbf9c0000, - 0xbf8200ca, 0xbef8007e, + 0xbf8200cd, 0xbef8007e, 0x8679ff7f, 0x0000ffff, 0x8779ff79, 0x00040000, 0xbefa0080, 0xbefb00ff, @@ -268,16 +271,18 @@ static const uint32_t cwsr_trap_gfx8_hex[] = { 0x8f739773, 0xb976f807, 0x8671ff71, 0x0000ffff, 0x86fe7e7e, 0x86ea6a6a, - 0xb974f802, 0xbf8a0000, - 0x95807370, 0xbf810000, + 0x8f768374, 0xb976e0c2, + 0xbf800002, 0xb9740002, + 0xbf8a0000, 0x95807370, + 0xbf810000, 0x00000000, }; static const uint32_t cwsr_trap_gfx9_hex[] = { - 0xbf820001, 0xbf82015a, + 0xbf820001, 0xbf82015d, 0xb8f8f802, 0x89788678, 0xb8f1f803, 0x866eff71, - 0x00000400, 0xbf850034, + 0x00000400, 0xbf850037, 0x866eff71, 0x00000800, 0xbf850003, 0x866eff71, 0x00000100, 0xbf840008, @@ -303,258 +308,261 @@ static const uint32_t cwsr_trap_gfx9_hex[] = { 0x8f6e8b77, 0x866eff6e, 0x001f8000, 0xb96ef807, 0x86fe7e7e, 0x86ea6a6a, - 0xb978f802, 0xbe801f6c, - 0x866dff6d, 0x0000ffff, - 0xbef00080, 0xb9700283, - 0xb8f02407, 0x8e709c70, - 0x876d706d, 0xb8f003c7, - 0x8e709b70, 0x876d706d, - 0xb8f0f807, 0x8670ff70, - 0x00007fff, 0xb970f807, - 0xbeee007e, 0xbeef007f, - 0xbefe0180, 0xbf900004, - 0x87708478, 0xb970f802, - 0xbf8e0002, 0xbf88fffe, - 0xb8f02a05, 0x80708170, - 0x8e708a70, 0xb8f11605, - 0x80718171, 0x8e718671, - 0x80707170, 0x80707e70, - 0x8271807f, 0x8671ff71, - 0x0000ffff, 0xc0471cb8, - 0x00000040, 0xbf8cc07f, - 0xc04b1d38, 0x00000048, - 0xbf8cc07f, 0xc0431e78, - 0x00000058, 0xbf8cc07f, - 0xc0471eb8, 0x0000005c, - 0xbf8cc07f, 0xbef4007e, - 0x8675ff7f, 0x0000ffff, - 0x8775ff75, 0x00040000, - 0xbef60080, 0xbef700ff, - 0x00807fac, 0x8670ff7f, - 0x08000000, 0x8f708370, - 0x87777077, 0x8670ff7f, - 0x70000000, 0x8f708170, - 0x87777077, 0xbefb007c, - 0xbefa0080, 0xb8fa2a05, - 0x807a817a, 0x8e7a8a7a, - 0xb8f01605, 0x80708170, - 0x8e708670, 0x807a707a, - 0xbef60084, 0xbef600ff, - 0x01000000, 0xbefe007c, - 0xbefc007a, 0xc0611efa, - 0x0000007c, 0xbf8cc07f, - 0x807a847a, 0xbefc007e, + 0x8f6e8378, 0xb96ee0c2, + 0xbf800002, 0xb9780002, + 0xbe801f6c, 0x866dff6d, + 0x0000ffff, 0xbef00080, + 0xb9700283, 0xb8f02407, + 0x8e709c70, 0x876d706d, + 0xb8f003c7, 0x8e709b70, + 0x876d706d, 0xb8f0f807, + 0x8670ff70, 0x00007fff, + 0xb970f807, 0xbeee007e, + 0xbeef007f, 0xbefe0180, + 0xbf900004, 0x87708478, + 0xb970f802, 0xbf8e0002, + 0xbf88fffe, 0xb8f02a05, + 0x80708170, 0x8e708a70, + 0xb8f11605, 0x80718171, + 0x8e718671, 0x80707170, + 0x80707e70, 0x8271807f, + 0x8671ff71, 0x0000ffff, + 0xc0471cb8, 0x00000040, + 0xbf8cc07f, 0xc04b1d38, + 0x00000048, 0xbf8cc07f, + 0xc0431e78, 0x00000058, + 0xbf8cc07f, 0xc0471eb8, + 0x0000005c, 0xbf8cc07f, + 0xbef4007e, 0x8675ff7f, + 0x0000ffff, 0x8775ff75, + 0x00040000, 0xbef60080, + 0xbef700ff, 0x00807fac, + 0x8670ff7f, 0x08000000, + 0x8f708370, 0x87777077, + 0x8670ff7f, 0x70000000, + 0x8f708170, 0x87777077, + 0xbefb007c, 0xbefa0080, + 0xb8fa2a05, 0x807a817a, + 0x8e7a8a7a, 0xb8f01605, + 0x80708170, 0x8e708670, + 0x807a707a, 0xbef60084, + 0xbef600ff, 0x01000000, 0xbefe007c, 0xbefc007a, - 0xc0611b3a, 0x0000007c, + 0xc0611efa, 0x0000007c, 0xbf8cc07f, 0x807a847a, 0xbefc007e, 0xbefe007c, - 0xbefc007a, 0xc0611b7a, + 0xbefc007a, 0xc0611b3a, 0x0000007c, 0xbf8cc07f, 0x807a847a, 0xbefc007e, 0xbefe007c, 0xbefc007a, - 0xc0611bba, 0x0000007c, + 0xc0611b7a, 0x0000007c, 0xbf8cc07f, 0x807a847a, 0xbefc007e, 0xbefe007c, - 0xbefc007a, 0xc0611bfa, + 0xbefc007a, 0xc0611bba, 0x0000007c, 0xbf8cc07f, 0x807a847a, 0xbefc007e, 0xbefe007c, 0xbefc007a, - 0xc0611e3a, 0x0000007c, - 0xbf8cc07f, 0x807a847a, - 0xbefc007e, 0xb8f1f803, - 0xbefe007c, 0xbefc007a, - 0xc0611c7a, 0x0000007c, + 0xc0611bfa, 0x0000007c, 0xbf8cc07f, 0x807a847a, 0xbefc007e, 0xbefe007c, - 0xbefc007a, 0xc0611a3a, + 0xbefc007a, 0xc0611e3a, + 0x0000007c, 0xbf8cc07f, + 0x807a847a, 0xbefc007e, + 0xb8f1f803, 0xbefe007c, + 0xbefc007a, 0xc0611c7a, 0x0000007c, 0xbf8cc07f, 0x807a847a, 0xbefc007e, 0xbefe007c, 0xbefc007a, - 0xc0611a7a, 0x0000007c, - 0xbf8cc07f, 0x807a847a, - 0xbefc007e, 0xb8fbf801, - 0xbefe007c, 0xbefc007a, - 0xc0611efa, 0x0000007c, + 0xc0611a3a, 0x0000007c, 0xbf8cc07f, 0x807a847a, - 0xbefc007e, 0x8670ff7f, - 0x04000000, 0xbeef0080, - 0x876f6f70, 0xb8fa2a05, + 0xbefc007e, 0xbefe007c, + 0xbefc007a, 0xc0611a7a, + 0x0000007c, 0xbf8cc07f, + 0x807a847a, 0xbefc007e, + 0xb8fbf801, 0xbefe007c, + 0xbefc007a, 0xc0611efa, + 0x0000007c, 0xbf8cc07f, + 0x807a847a, 0xbefc007e, + 0x8670ff7f, 0x04000000, + 0xbeef0080, 0x876f6f70, + 0xb8fa2a05, 0x807a817a, + 0x8e7a8a7a, 0xb8f11605, + 0x80718171, 0x8e718471, + 0x8e768271, 0xbef600ff, + 0x01000000, 0xbef20174, + 0x80747a74, 0x82758075, + 0xbefc0080, 0xbf800000, + 0xbe802b00, 0xbe822b02, + 0xbe842b04, 0xbe862b06, + 0xbe882b08, 0xbe8a2b0a, + 0xbe8c2b0c, 0xbe8e2b0e, + 0xc06b003a, 0x00000000, + 0xbf8cc07f, 0xc06b013a, + 0x00000010, 0xbf8cc07f, + 0xc06b023a, 0x00000020, + 0xbf8cc07f, 0xc06b033a, + 0x00000030, 0xbf8cc07f, + 0x8074c074, 0x82758075, + 0x807c907c, 0xbf0a717c, + 0xbf85ffe7, 0xbef40172, + 0xbefa0080, 0xbefe00c1, + 0xbeff00c1, 0xbee80080, + 0xbee90080, 0xbef600ff, + 0x01000000, 0xe0724000, + 0x7a1d0000, 0xe0724100, + 0x7a1d0100, 0xe0724200, + 0x7a1d0200, 0xe0724300, + 0x7a1d0300, 0xbefe00c1, + 0xbeff00c1, 0xb8f14306, + 0x8671c171, 0xbf84002c, + 0xbf8a0000, 0x8670ff6f, + 0x04000000, 0xbf840028, + 0x8e718671, 0x8e718271, + 0xbef60071, 0xb8fa2a05, 0x807a817a, 0x8e7a8a7a, - 0xb8f11605, 0x80718171, - 0x8e718471, 0x8e768271, + 0xb8f01605, 0x80708170, + 0x8e708670, 0x807a707a, + 0x807aff7a, 0x00000080, 0xbef600ff, 0x01000000, - 0xbef20174, 0x80747a74, - 0x82758075, 0xbefc0080, - 0xbf800000, 0xbe802b00, - 0xbe822b02, 0xbe842b04, - 0xbe862b06, 0xbe882b08, - 0xbe8a2b0a, 0xbe8c2b0c, - 0xbe8e2b0e, 0xc06b003a, - 0x00000000, 0xbf8cc07f, - 0xc06b013a, 0x00000010, - 0xbf8cc07f, 0xc06b023a, - 0x00000020, 0xbf8cc07f, - 0xc06b033a, 0x00000030, - 0xbf8cc07f, 0x8074c074, - 0x82758075, 0x807c907c, - 0xbf0a717c, 0xbf85ffe7, - 0xbef40172, 0xbefa0080, + 0xbefc0080, 0xd28c0002, + 0x000100c1, 0xd28d0003, + 0x000204c1, 0xd1060002, + 0x00011103, 0x7e0602ff, + 0x00000200, 0xbefc00ff, + 0x00010000, 0xbe800077, + 0x8677ff77, 0xff7fffff, + 0x8777ff77, 0x00058000, + 0xd8ec0000, 0x00000002, + 0xbf8cc07f, 0xe0765000, + 0x7a1d0002, 0x68040702, + 0xd0c9006a, 0x0000e302, + 0xbf87fff7, 0xbef70000, + 0xbefa00ff, 0x00000400, 0xbefe00c1, 0xbeff00c1, - 0xbee80080, 0xbee90080, + 0xb8f12a05, 0x80718171, + 0x8e718271, 0x8e768871, 0xbef600ff, 0x01000000, + 0xbefc0084, 0xbf0a717c, + 0xbf840015, 0xbf11017c, + 0x8071ff71, 0x00001000, + 0x7e000300, 0x7e020301, + 0x7e040302, 0x7e060303, 0xe0724000, 0x7a1d0000, 0xe0724100, 0x7a1d0100, 0xe0724200, 0x7a1d0200, 0xe0724300, 0x7a1d0300, + 0x807c847c, 0x807aff7a, + 0x00000400, 0xbf0a717c, + 0xbf85ffef, 0xbf9c0000, + 0xbf8200dc, 0xbef4007e, + 0x8675ff7f, 0x0000ffff, + 0x8775ff75, 0x00040000, + 0xbef60080, 0xbef700ff, + 0x00807fac, 0x866eff7f, + 0x08000000, 0x8f6e836e, + 0x87776e77, 0x866eff7f, + 0x70000000, 0x8f6e816e, + 0x87776e77, 0x866eff7f, + 0x04000000, 0xbf84001e, 0xbefe00c1, 0xbeff00c1, - 0xb8f14306, 0x8671c171, - 0xbf84002c, 0xbf8a0000, - 0x8670ff6f, 0x04000000, - 0xbf840028, 0x8e718671, - 0x8e718271, 0xbef60071, - 0xb8fa2a05, 0x807a817a, - 0x8e7a8a7a, 0xb8f01605, - 0x80708170, 0x8e708670, - 0x807a707a, 0x807aff7a, + 0xb8ef4306, 0x866fc16f, + 0xbf840019, 0x8e6f866f, + 0x8e6f826f, 0xbef6006f, + 0xb8f82a05, 0x80788178, + 0x8e788a78, 0xb8ee1605, + 0x806e816e, 0x8e6e866e, + 0x80786e78, 0x8078ff78, 0x00000080, 0xbef600ff, 0x01000000, 0xbefc0080, - 0xd28c0002, 0x000100c1, - 0xd28d0003, 0x000204c1, - 0xd1060002, 0x00011103, - 0x7e0602ff, 0x00000200, - 0xbefc00ff, 0x00010000, - 0xbe800077, 0x8677ff77, - 0xff7fffff, 0x8777ff77, - 0x00058000, 0xd8ec0000, - 0x00000002, 0xbf8cc07f, - 0xe0765000, 0x7a1d0002, - 0x68040702, 0xd0c9006a, - 0x0000e302, 0xbf87fff7, - 0xbef70000, 0xbefa00ff, - 0x00000400, 0xbefe00c1, - 0xbeff00c1, 0xb8f12a05, - 0x80718171, 0x8e718271, - 0x8e768871, 0xbef600ff, - 0x01000000, 0xbefc0084, - 0xbf0a717c, 0xbf840015, - 0xbf11017c, 0x8071ff71, - 0x00001000, 0x7e000300, + 0xe0510000, 0x781d0000, + 0xe0510100, 0x781d0000, + 0x807cff7c, 0x00000200, + 0x8078ff78, 0x00000200, + 0xbf0a6f7c, 0xbf85fff6, + 0xbef80080, 0xbefe00c1, + 0xbeff00c1, 0xb8ef2a05, + 0x806f816f, 0x8e6f826f, + 0x8e76886f, 0xbef600ff, + 0x01000000, 0xbeee0078, + 0x8078ff78, 0x00000400, + 0xbefc0084, 0xbf11087c, + 0x806fff6f, 0x00008000, + 0xe0524000, 0x781d0000, + 0xe0524100, 0x781d0100, + 0xe0524200, 0x781d0200, + 0xe0524300, 0x781d0300, + 0xbf8c0f70, 0x7e000300, 0x7e020301, 0x7e040302, - 0x7e060303, 0xe0724000, - 0x7a1d0000, 0xe0724100, - 0x7a1d0100, 0xe0724200, - 0x7a1d0200, 0xe0724300, - 0x7a1d0300, 0x807c847c, - 0x807aff7a, 0x00000400, - 0xbf0a717c, 0xbf85ffef, - 0xbf9c0000, 0xbf8200d9, - 0xbef4007e, 0x8675ff7f, - 0x0000ffff, 0x8775ff75, - 0x00040000, 0xbef60080, - 0xbef700ff, 0x00807fac, - 0x866eff7f, 0x08000000, - 0x8f6e836e, 0x87776e77, - 0x866eff7f, 0x70000000, - 0x8f6e816e, 0x87776e77, - 0x866eff7f, 0x04000000, - 0xbf84001e, 0xbefe00c1, - 0xbeff00c1, 0xb8ef4306, - 0x866fc16f, 0xbf840019, - 0x8e6f866f, 0x8e6f826f, - 0xbef6006f, 0xb8f82a05, + 0x7e060303, 0x807c847c, + 0x8078ff78, 0x00000400, + 0xbf0a6f7c, 0xbf85ffee, + 0xbf9c0000, 0xe0524000, + 0x6e1d0000, 0xe0524100, + 0x6e1d0100, 0xe0524200, + 0x6e1d0200, 0xe0524300, + 0x6e1d0300, 0xb8f82a05, 0x80788178, 0x8e788a78, 0xb8ee1605, 0x806e816e, 0x8e6e866e, 0x80786e78, - 0x8078ff78, 0x00000080, - 0xbef600ff, 0x01000000, - 0xbefc0080, 0xe0510000, - 0x781d0000, 0xe0510100, - 0x781d0000, 0x807cff7c, - 0x00000200, 0x8078ff78, - 0x00000200, 0xbf0a6f7c, - 0xbf85fff6, 0xbef80080, - 0xbefe00c1, 0xbeff00c1, - 0xb8ef2a05, 0x806f816f, - 0x8e6f826f, 0x8e76886f, - 0xbef600ff, 0x01000000, - 0xbeee0078, 0x8078ff78, - 0x00000400, 0xbefc0084, - 0xbf11087c, 0x806fff6f, - 0x00008000, 0xe0524000, - 0x781d0000, 0xe0524100, - 0x781d0100, 0xe0524200, - 0x781d0200, 0xe0524300, - 0x781d0300, 0xbf8c0f70, - 0x7e000300, 0x7e020301, - 0x7e040302, 0x7e060303, - 0x807c847c, 0x8078ff78, - 0x00000400, 0xbf0a6f7c, - 0xbf85ffee, 0xbf9c0000, - 0xe0524000, 0x6e1d0000, - 0xe0524100, 0x6e1d0100, - 0xe0524200, 0x6e1d0200, - 0xe0524300, 0x6e1d0300, + 0x80f8c078, 0xb8ef1605, + 0x806f816f, 0x8e6f846f, + 0x8e76826f, 0xbef600ff, + 0x01000000, 0xbefc006f, + 0xc031003a, 0x00000078, + 0x80f8c078, 0xbf8cc07f, + 0x80fc907c, 0xbf800000, + 0xbe802d00, 0xbe822d02, + 0xbe842d04, 0xbe862d06, + 0xbe882d08, 0xbe8a2d0a, + 0xbe8c2d0c, 0xbe8e2d0e, + 0xbf06807c, 0xbf84fff0, 0xb8f82a05, 0x80788178, 0x8e788a78, 0xb8ee1605, 0x806e816e, 0x8e6e866e, - 0x80786e78, 0x80f8c078, - 0xb8ef1605, 0x806f816f, - 0x8e6f846f, 0x8e76826f, + 0x80786e78, 0xbef60084, 0xbef600ff, 0x01000000, - 0xbefc006f, 0xc031003a, - 0x00000078, 0x80f8c078, - 0xbf8cc07f, 0x80fc907c, - 0xbf800000, 0xbe802d00, - 0xbe822d02, 0xbe842d04, - 0xbe862d06, 0xbe882d08, - 0xbe8a2d0a, 0xbe8c2d0c, - 0xbe8e2d0e, 0xbf06807c, - 0xbf84fff0, 0xb8f82a05, - 0x80788178, 0x8e788a78, - 0xb8ee1605, 0x806e816e, - 0x8e6e866e, 0x80786e78, - 0xbef60084, 0xbef600ff, - 0x01000000, 0xc0211bfa, + 0xc0211bfa, 0x00000078, + 0x80788478, 0xc0211b3a, 0x00000078, 0x80788478, - 0xc0211b3a, 0x00000078, - 0x80788478, 0xc0211b7a, + 0xc0211b7a, 0x00000078, + 0x80788478, 0xc0211eba, 0x00000078, 0x80788478, - 0xc0211eba, 0x00000078, - 0x80788478, 0xc0211efa, + 0xc0211efa, 0x00000078, + 0x80788478, 0xc0211c3a, 0x00000078, 0x80788478, - 0xc0211c3a, 0x00000078, - 0x80788478, 0xc0211c7a, + 0xc0211c7a, 0x00000078, + 0x80788478, 0xc0211a3a, 0x00000078, 0x80788478, - 0xc0211a3a, 0x00000078, - 0x80788478, 0xc0211a7a, + 0xc0211a7a, 0x00000078, + 0x80788478, 0xc0211cfa, 0x00000078, 0x80788478, - 0xc0211cfa, 0x00000078, - 0x80788478, 0xbf8cc07f, - 0xbefc006f, 0xbefe007a, - 0xbeff007b, 0x866f71ff, - 0x000003ff, 0xb96f4803, - 0x866f71ff, 0xfffff800, - 0x8f6f8b6f, 0xb96fa2c3, - 0xb973f801, 0xb8ee2a05, - 0x806e816e, 0x8e6e8a6e, - 0xb8ef1605, 0x806f816f, - 0x8e6f866f, 0x806e6f6e, - 0x806e746e, 0x826f8075, - 0x866fff6f, 0x0000ffff, - 0xc0071cb7, 0x00000040, - 0xc00b1d37, 0x00000048, - 0xc0031e77, 0x00000058, - 0xc0071eb7, 0x0000005c, - 0xbf8cc07f, 0x866fff6d, - 0xf0000000, 0x8f6f9c6f, - 0x8e6f906f, 0xbeee0080, - 0x876e6f6e, 0x866fff6d, - 0x08000000, 0x8f6f9b6f, - 0x8e6f8f6f, 0x876e6f6e, - 0x866fff70, 0x00800000, - 0x8f6f976f, 0xb96ef807, - 0x866dff6d, 0x0000ffff, - 0x86fe7e7e, 0x86ea6a6a, - 0xb970f802, 0xbf8a0000, + 0xbf8cc07f, 0xbefc006f, + 0xbefe007a, 0xbeff007b, + 0x866f71ff, 0x000003ff, + 0xb96f4803, 0x866f71ff, + 0xfffff800, 0x8f6f8b6f, + 0xb96fa2c3, 0xb973f801, + 0xb8ee2a05, 0x806e816e, + 0x8e6e8a6e, 0xb8ef1605, + 0x806f816f, 0x8e6f866f, + 0x806e6f6e, 0x806e746e, + 0x826f8075, 0x866fff6f, + 0x0000ffff, 0xc0071cb7, + 0x00000040, 0xc00b1d37, + 0x00000048, 0xc0031e77, + 0x00000058, 0xc0071eb7, + 0x0000005c, 0xbf8cc07f, + 0x866fff6d, 0xf0000000, + 0x8f6f9c6f, 0x8e6f906f, + 0xbeee0080, 0x876e6f6e, + 0x866fff6d, 0x08000000, + 0x8f6f9b6f, 0x8e6f8f6f, + 0x876e6f6e, 0x866fff70, + 0x00800000, 0x8f6f976f, + 0xb96ef807, 0x866dff6d, + 0x0000ffff, 0x86fe7e7e, + 0x86ea6a6a, 0x8f6e8370, + 0xb96ee0c2, 0xbf800002, + 0xb9700002, 0xbf8a0000, 0x95806f6c, 0xbf810000, }; diff --git a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx8.asm b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx8.asm index a2a04bb64096..abe1a5da29fb 100644 --- a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx8.asm +++ b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx8.asm @@ -103,6 +103,10 @@ var SQ_WAVE_STATUS_INST_ATC_SHIFT = 23 var SQ_WAVE_STATUS_INST_ATC_MASK = 0x00800000 var SQ_WAVE_STATUS_SPI_PRIO_SHIFT = 1 var SQ_WAVE_STATUS_SPI_PRIO_MASK = 0x00000006 +var SQ_WAVE_STATUS_PRE_SPI_PRIO_SHIFT = 0 +var SQ_WAVE_STATUS_PRE_SPI_PRIO_SIZE = 1 +var SQ_WAVE_STATUS_POST_SPI_PRIO_SHIFT = 3 +var SQ_WAVE_STATUS_POST_SPI_PRIO_SIZE = 29 var SQ_WAVE_LDS_ALLOC_LDS_SIZE_SHIFT = 12 var SQ_WAVE_LDS_ALLOC_LDS_SIZE_SIZE = 9 @@ -251,7 +255,7 @@ if (!EMU_RUN_HACK) s_waitcnt lgkmcnt(0) s_or_b32 ttmp7, ttmp8, ttmp9 s_cbranch_scc0 L_NO_NEXT_TRAP //next level trap handler not been set - s_setreg_b32 hwreg(HW_REG_STATUS), s_save_status //restore HW status(SCC) + set_status_without_spi_prio(s_save_status, ttmp2) //restore HW status(SCC) s_setpc_b64 [ttmp8,ttmp9] //jump to next level trap handler L_NO_NEXT_TRAP: @@ -262,7 +266,7 @@ L_NO_NEXT_TRAP: s_addc_u32 ttmp1, ttmp1, 0 L_EXCP_CASE: s_and_b32 ttmp1, ttmp1, 0xFFFF - s_setreg_b32 hwreg(HW_REG_STATUS), s_save_status //restore HW status(SCC) + set_status_without_spi_prio(s_save_status, ttmp2) //restore HW status(SCC) s_rfe_b64 [ttmp0, ttmp1] end // ********* End handling of non-CWSR traps ******************* @@ -1053,7 +1057,7 @@ end 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 - s_setreg_b32 hwreg(HW_REG_STATUS), s_restore_status // SCC is included, which is changed by previous salu + set_status_without_spi_prio(s_restore_status, s_restore_tmp) // SCC is included, which is changed by previous salu s_barrier //barrier to ensure the readiness of LDS before access attempts from any other wave in the same TG //FIXME not performance-optimal at this time @@ -1134,3 +1138,11 @@ end function get_hwreg_size_bytes return 128 //HWREG size 128 bytes end + +function set_status_without_spi_prio(status, tmp) + // Do not restore STATUS.SPI_PRIO since scheduler may have raised it. + s_lshr_b32 tmp, status, SQ_WAVE_STATUS_POST_SPI_PRIO_SHIFT + s_setreg_b32 hwreg(HW_REG_STATUS, SQ_WAVE_STATUS_POST_SPI_PRIO_SHIFT, SQ_WAVE_STATUS_POST_SPI_PRIO_SIZE), tmp + s_nop 0x2 // avoid S_SETREG => S_SETREG hazard + s_setreg_b32 hwreg(HW_REG_STATUS, SQ_WAVE_STATUS_PRE_SPI_PRIO_SHIFT, SQ_WAVE_STATUS_PRE_SPI_PRIO_SIZE), status +end diff --git a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx9.asm b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx9.asm index 998be96be736..0bb9c577b3a2 100644 --- a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx9.asm +++ b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx9.asm @@ -103,6 +103,10 @@ var SQ_WAVE_STATUS_INST_ATC_MASK = 0x00800000 var SQ_WAVE_STATUS_SPI_PRIO_SHIFT = 1 var SQ_WAVE_STATUS_SPI_PRIO_MASK = 0x00000006 var SQ_WAVE_STATUS_HALT_MASK = 0x2000 +var SQ_WAVE_STATUS_PRE_SPI_PRIO_SHIFT = 0 +var SQ_WAVE_STATUS_PRE_SPI_PRIO_SIZE = 1 +var SQ_WAVE_STATUS_POST_SPI_PRIO_SHIFT = 3 +var SQ_WAVE_STATUS_POST_SPI_PRIO_SIZE = 29 var SQ_WAVE_LDS_ALLOC_LDS_SIZE_SHIFT = 12 var SQ_WAVE_LDS_ALLOC_LDS_SIZE_SIZE = 9 @@ -317,7 +321,7 @@ L_EXCP_CASE: // Restore SQ_WAVE_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 - s_setreg_b32 hwreg(HW_REG_STATUS), s_save_status + set_status_without_spi_prio(s_save_status, ttmp2) s_rfe_b64 [ttmp0, ttmp1] end @@ -1120,7 +1124,7 @@ end 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 - s_setreg_b32 hwreg(HW_REG_STATUS), s_restore_status // SCC is included, which is changed by previous salu + set_status_without_spi_prio(s_restore_status, s_restore_tmp) // SCC is included, which is changed by previous salu s_barrier //barrier to ensure the readiness of LDS before access attempts from any other wave in the same TG //FIXME not performance-optimal at this time @@ -1212,3 +1216,11 @@ function ack_sqc_store_workaround s_waitcnt lgkmcnt(0) end end + +function set_status_without_spi_prio(status, tmp) + // Do not restore STATUS.SPI_PRIO since scheduler may have raised it. + s_lshr_b32 tmp, status, SQ_WAVE_STATUS_POST_SPI_PRIO_SHIFT + s_setreg_b32 hwreg(HW_REG_STATUS, SQ_WAVE_STATUS_POST_SPI_PRIO_SHIFT, SQ_WAVE_STATUS_POST_SPI_PRIO_SIZE), tmp + s_nop 0x2 // avoid S_SETREG => S_SETREG hazard + s_setreg_b32 hwreg(HW_REG_STATUS, SQ_WAVE_STATUS_PRE_SPI_PRIO_SHIFT, SQ_WAVE_STATUS_PRE_SPI_PRIO_SIZE), status +end diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c index f64c5551cdba..297b36c26a05 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c @@ -122,6 +122,9 @@ static int kfd_open(struct inode *inode, struct file *filep) if (IS_ERR(process)) return PTR_ERR(process); + if (kfd_is_locked()) + return -EAGAIN; + dev_dbg(kfd_device, "process %d opened, compat mode (32 bit) - %d\n", process->pasid, process->is_32bit_user_mode); @@ -389,6 +392,61 @@ static int kfd_ioctl_update_queue(struct file *filp, struct kfd_process *p, return retval; } +static int kfd_ioctl_set_cu_mask(struct file *filp, struct kfd_process *p, + void *data) +{ + int retval; + const int max_num_cus = 1024; + struct kfd_ioctl_set_cu_mask_args *args = data; + struct queue_properties properties; + uint32_t __user *cu_mask_ptr = (uint32_t __user *)args->cu_mask_ptr; + size_t cu_mask_size = sizeof(uint32_t) * (args->num_cu_mask / 32); + + if ((args->num_cu_mask % 32) != 0) { + pr_debug("num_cu_mask 0x%x must be a multiple of 32", + args->num_cu_mask); + return -EINVAL; + } + + properties.cu_mask_count = args->num_cu_mask; + if (properties.cu_mask_count == 0) { + pr_debug("CU mask cannot be 0"); + return -EINVAL; + } + + /* To prevent an unreasonably large CU mask size, set an arbitrary + * limit of max_num_cus bits. We can then just drop any CU mask bits + * past max_num_cus bits and just use the first max_num_cus bits. + */ + if (properties.cu_mask_count > max_num_cus) { + pr_debug("CU mask cannot be greater than 1024 bits"); + properties.cu_mask_count = max_num_cus; + cu_mask_size = sizeof(uint32_t) * (max_num_cus/32); + } + + properties.cu_mask = kzalloc(cu_mask_size, GFP_KERNEL); + if (!properties.cu_mask) + return -ENOMEM; + + retval = copy_from_user(properties.cu_mask, cu_mask_ptr, cu_mask_size); + if (retval) { + pr_debug("Could not copy CU mask from userspace"); + kfree(properties.cu_mask); + return -EFAULT; + } + + mutex_lock(&p->mutex); + + retval = pqm_set_cu_mask(&p->pqm, args->queue_id, &properties); + + mutex_unlock(&p->mutex); + + if (retval) + kfree(properties.cu_mask); + + return retval; +} + static int kfd_ioctl_set_memory_policy(struct file *filep, struct kfd_process *p, void *data) { @@ -754,7 +812,6 @@ static int kfd_ioctl_get_clock_counters(struct file *filep, { struct kfd_ioctl_get_clock_counters_args *args = data; struct kfd_dev *dev; - struct timespec64 time; dev = kfd_device_by_id(args->gpu_id); if (dev) @@ -766,11 +823,8 @@ static int kfd_ioctl_get_clock_counters(struct file *filep, args->gpu_clock_counter = 0; /* No access to rdtsc. Using raw monotonic time */ - getrawmonotonic64(&time); - args->cpu_clock_counter = (uint64_t)timespec64_to_ns(&time); - - get_monotonic_boottime64(&time); - args->system_clock_counter = (uint64_t)timespec64_to_ns(&time); + args->cpu_clock_counter = ktime_get_raw_ns(); + args->system_clock_counter = ktime_get_boot_ns(); /* Since the counter is in nano-seconds we use 1GHz frequency */ args->system_clock_freq = 1000000000; @@ -1558,6 +1612,9 @@ static const struct amdkfd_ioctl_desc amdkfd_ioctls[] = { AMDKFD_IOCTL_DEF(AMDKFD_IOC_UNMAP_MEMORY_FROM_GPU, kfd_ioctl_unmap_memory_from_gpu, 0), + AMDKFD_IOCTL_DEF(AMDKFD_IOC_SET_CU_MASK, + kfd_ioctl_set_cu_mask, 0), + }; #define AMDKFD_CORE_IOCTL_COUNT ARRAY_SIZE(amdkfd_ioctls) diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_crat.c b/drivers/gpu/drm/amd/amdkfd/kfd_crat.c index 296b3f230280..ee4996029a86 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_crat.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_crat.c @@ -189,6 +189,21 @@ static int kfd_parse_subtype_cu(struct crat_subtype_computeunit *cu, return 0; } +static struct kfd_mem_properties * +find_subtype_mem(uint32_t heap_type, uint32_t flags, uint32_t width, + struct kfd_topology_device *dev) +{ + struct kfd_mem_properties *props; + + list_for_each_entry(props, &dev->mem_props, list) { + if (props->heap_type == heap_type + && props->flags == flags + && props->width == width) + return props; + } + + return NULL; +} /* kfd_parse_subtype_mem - parse memory subtypes and attach it to correct * topology device present in the device_list */ @@ -197,36 +212,56 @@ static int kfd_parse_subtype_mem(struct crat_subtype_memory *mem, { struct kfd_mem_properties *props; struct kfd_topology_device *dev; + uint32_t heap_type; + uint64_t size_in_bytes; + uint32_t flags = 0; + uint32_t width; pr_debug("Found memory entry in CRAT table with proximity_domain=%d\n", mem->proximity_domain); list_for_each_entry(dev, device_list, list) { if (mem->proximity_domain == dev->proximity_domain) { - props = kfd_alloc_struct(props); - if (!props) - return -ENOMEM; - /* We're on GPU node */ if (dev->node_props.cpu_cores_count == 0) { /* APU */ if (mem->visibility_type == 0) - props->heap_type = + heap_type = HSA_MEM_HEAP_TYPE_FB_PRIVATE; /* dGPU */ else - props->heap_type = mem->visibility_type; + heap_type = mem->visibility_type; } else - props->heap_type = HSA_MEM_HEAP_TYPE_SYSTEM; + heap_type = HSA_MEM_HEAP_TYPE_SYSTEM; if (mem->flags & CRAT_MEM_FLAGS_HOT_PLUGGABLE) - props->flags |= HSA_MEM_FLAGS_HOT_PLUGGABLE; + flags |= HSA_MEM_FLAGS_HOT_PLUGGABLE; if (mem->flags & CRAT_MEM_FLAGS_NON_VOLATILE) - props->flags |= HSA_MEM_FLAGS_NON_VOLATILE; + flags |= HSA_MEM_FLAGS_NON_VOLATILE; - props->size_in_bytes = + size_in_bytes = ((uint64_t)mem->length_high << 32) + mem->length_low; - props->width = mem->width; + width = mem->width; + + /* Multiple banks of the same type are aggregated into + * one. User mode doesn't care about multiple physical + * memory segments. It's managed as a single virtual + * heap for user mode. + */ + props = find_subtype_mem(heap_type, flags, width, dev); + if (props) { + props->size_in_bytes += size_in_bytes; + break; + } + + props = kfd_alloc_struct(props); + if (!props) + return -ENOMEM; + + props->heap_type = heap_type; + props->flags = flags; + props->size_in_bytes = size_in_bytes; + props->width = width; dev->node_props.mem_banks_count++; list_add_tail(&props->list, &dev->mem_props); diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_dbgdev.c b/drivers/gpu/drm/amd/amdkfd/kfd_dbgdev.c index afb26f205d29..a3441b0e385b 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_dbgdev.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_dbgdev.c @@ -38,7 +38,6 @@ #include "kfd_dbgmgr.h" #include "kfd_dbgdev.h" #include "kfd_device_queue_manager.h" -#include "../../radeon/cik_reg.h" static void dbgdev_address_watch_disable_nodiq(struct kfd_dev *dev) { diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_dbgdev.h b/drivers/gpu/drm/amd/amdkfd/kfd_dbgdev.h index 03424c20920c..0619c777b47e 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_dbgdev.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_dbgdev.h @@ -60,6 +60,9 @@ enum { SH_REG_SIZE = SH_REG_END - SH_REG_BASE }; +/* SQ_CMD definitions */ +#define SQ_CMD 0x8DEC + enum SQ_IND_CMD_CMD { SQ_IND_CMD_CMD_NULL = 0x00000000, SQ_IND_CMD_CMD_HALT = 0x00000001, @@ -190,4 +193,38 @@ union ULARGE_INTEGER { void kfd_dbgdev_init(struct kfd_dbgdev *pdbgdev, struct kfd_dev *pdev, enum DBGDEV_TYPE type); +union TCP_WATCH_CNTL_BITS { + struct { + uint32_t mask:24; + uint32_t vmid:4; + uint32_t atc:1; + uint32_t mode:2; + uint32_t valid:1; + } bitfields, bits; + uint32_t u32All; + signed int i32All; + float f32All; +}; + +enum { + ADDRESS_WATCH_REG_CNTL_ATC_BIT = 0x10000000UL, + ADDRESS_WATCH_REG_CNTL_DEFAULT_MASK = 0x00FFFFFF, + ADDRESS_WATCH_REG_ADDLOW_MASK_EXTENSION = 0x03000000, + /* extend the mask to 26 bits in order to match the low address field */ + ADDRESS_WATCH_REG_ADDLOW_SHIFT = 6, + ADDRESS_WATCH_REG_ADDHIGH_MASK = 0xFFFF +}; + +enum { + MAX_TRAPID = 8, /* 3 bits in the bitfield. */ + MAX_WATCH_ADDRESSES = 4 +}; + +enum { + ADDRESS_WATCH_REG_ADDR_HI = 0, + ADDRESS_WATCH_REG_ADDR_LO, + ADDRESS_WATCH_REG_CNTL, + ADDRESS_WATCH_REG_MAX +}; + #endif /* KFD_DBGDEV_H_ */ diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_debugfs.c b/drivers/gpu/drm/amd/amdkfd/kfd_debugfs.c index 4bd6ebfaf425..ab37d36d9cd6 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_debugfs.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_debugfs.c @@ -21,6 +21,8 @@ */ #include <linux/debugfs.h> +#include <linux/uaccess.h> + #include "kfd_priv.h" static struct dentry *debugfs_root; @@ -32,6 +34,38 @@ static int kfd_debugfs_open(struct inode *inode, struct file *file) return single_open(file, show, NULL); } +static ssize_t kfd_debugfs_hang_hws_write(struct file *file, + const char __user *user_buf, size_t size, loff_t *ppos) +{ + struct kfd_dev *dev; + char tmp[16]; + uint32_t gpu_id; + int ret = -EINVAL; + + memset(tmp, 0, 16); + if (size >= 16) { + pr_err("Invalid input for gpu id.\n"); + goto out; + } + if (copy_from_user(tmp, user_buf, size)) { + ret = -EFAULT; + goto out; + } + if (kstrtoint(tmp, 10, &gpu_id)) { + pr_err("Invalid input for gpu id.\n"); + goto out; + } + dev = kfd_device_by_id(gpu_id); + if (dev) { + kfd_debugfs_hang_hws(dev); + ret = size; + } else + pr_err("Cannot find device %d.\n", gpu_id); + +out: + return ret; +} + static const struct file_operations kfd_debugfs_fops = { .owner = THIS_MODULE, .open = kfd_debugfs_open, @@ -40,6 +74,15 @@ static const struct file_operations kfd_debugfs_fops = { .release = single_release, }; +static const struct file_operations kfd_debugfs_hang_hws_fops = { + .owner = THIS_MODULE, + .open = kfd_debugfs_open, + .read = seq_read, + .write = kfd_debugfs_hang_hws_write, + .llseek = seq_lseek, + .release = single_release, +}; + void kfd_debugfs_init(void) { struct dentry *ent; @@ -65,6 +108,11 @@ void kfd_debugfs_init(void) ent = debugfs_create_file("rls", S_IFREG | 0444, debugfs_root, kfd_debugfs_rls_by_device, &kfd_debugfs_fops); + + ent = debugfs_create_file("hang_hws", S_IFREG | 0644, debugfs_root, + NULL, + &kfd_debugfs_hang_hws_fops); + if (!ent) pr_warn("Failed to create rls in kfd debugfs\n"); } diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device.c b/drivers/gpu/drm/amd/amdkfd/kfd_device.c index 7ee6cec2c060..1b048715ab8a 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_device.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device.c @@ -30,7 +30,13 @@ #include "kfd_iommu.h" #define MQD_SIZE_ALIGNED 768 -static atomic_t kfd_device_suspended = ATOMIC_INIT(0); + +/* + * kfd_locked is used to lock the kfd driver during suspend or reset + * once locked, kfd driver will stop any further GPU execution. + * create process (open) will return -EAGAIN. + */ +static atomic_t kfd_locked = ATOMIC_INIT(0); #ifdef KFD_SUPPORT_IOMMU_V2 static const struct kfd_device_info kaveri_device_info = { @@ -46,6 +52,7 @@ static const struct kfd_device_info kaveri_device_info = { .supports_cwsr = false, .needs_iommu_device = true, .needs_pci_atomics = false, + .num_sdma_engines = 2, }; static const struct kfd_device_info carrizo_device_info = { @@ -61,6 +68,22 @@ static const struct kfd_device_info carrizo_device_info = { .supports_cwsr = true, .needs_iommu_device = true, .needs_pci_atomics = false, + .num_sdma_engines = 2, +}; + +static const struct kfd_device_info raven_device_info = { + .asic_family = CHIP_RAVEN, + .max_pasid_bits = 16, + .max_no_of_hqd = 24, + .doorbell_size = 8, + .ih_ring_entry_size = 8 * sizeof(uint32_t), + .event_interrupt_class = &event_interrupt_class_v9, + .num_of_watch_points = 4, + .mqd_size_aligned = MQD_SIZE_ALIGNED, + .supports_cwsr = true, + .needs_iommu_device = true, + .needs_pci_atomics = true, + .num_sdma_engines = 1, }; #endif @@ -77,6 +100,7 @@ static const struct kfd_device_info hawaii_device_info = { .supports_cwsr = false, .needs_iommu_device = false, .needs_pci_atomics = false, + .num_sdma_engines = 2, }; static const struct kfd_device_info tonga_device_info = { @@ -91,6 +115,7 @@ static const struct kfd_device_info tonga_device_info = { .supports_cwsr = false, .needs_iommu_device = false, .needs_pci_atomics = true, + .num_sdma_engines = 2, }; static const struct kfd_device_info tonga_vf_device_info = { @@ -105,6 +130,7 @@ static const struct kfd_device_info tonga_vf_device_info = { .supports_cwsr = false, .needs_iommu_device = false, .needs_pci_atomics = false, + .num_sdma_engines = 2, }; static const struct kfd_device_info fiji_device_info = { @@ -119,6 +145,7 @@ static const struct kfd_device_info fiji_device_info = { .supports_cwsr = true, .needs_iommu_device = false, .needs_pci_atomics = true, + .num_sdma_engines = 2, }; static const struct kfd_device_info fiji_vf_device_info = { @@ -133,6 +160,7 @@ static const struct kfd_device_info fiji_vf_device_info = { .supports_cwsr = true, .needs_iommu_device = false, .needs_pci_atomics = false, + .num_sdma_engines = 2, }; @@ -148,6 +176,7 @@ static const struct kfd_device_info polaris10_device_info = { .supports_cwsr = true, .needs_iommu_device = false, .needs_pci_atomics = true, + .num_sdma_engines = 2, }; static const struct kfd_device_info polaris10_vf_device_info = { @@ -162,6 +191,7 @@ static const struct kfd_device_info polaris10_vf_device_info = { .supports_cwsr = true, .needs_iommu_device = false, .needs_pci_atomics = false, + .num_sdma_engines = 2, }; static const struct kfd_device_info polaris11_device_info = { @@ -176,6 +206,7 @@ static const struct kfd_device_info polaris11_device_info = { .supports_cwsr = true, .needs_iommu_device = false, .needs_pci_atomics = true, + .num_sdma_engines = 2, }; static const struct kfd_device_info vega10_device_info = { @@ -190,6 +221,7 @@ static const struct kfd_device_info vega10_device_info = { .supports_cwsr = true, .needs_iommu_device = false, .needs_pci_atomics = false, + .num_sdma_engines = 2, }; static const struct kfd_device_info vega10_vf_device_info = { @@ -204,6 +236,7 @@ static const struct kfd_device_info vega10_vf_device_info = { .supports_cwsr = true, .needs_iommu_device = false, .needs_pci_atomics = false, + .num_sdma_engines = 2, }; @@ -241,6 +274,7 @@ static const struct kfd_deviceid supported_devices[] = { { 0x9875, &carrizo_device_info }, /* Carrizo */ { 0x9876, &carrizo_device_info }, /* Carrizo */ { 0x9877, &carrizo_device_info }, /* Carrizo */ + { 0x15DD, &raven_device_info }, /* Raven */ #endif { 0x67A0, &hawaii_device_info }, /* Hawaii */ { 0x67A1, &hawaii_device_info }, /* Hawaii */ @@ -514,13 +548,54 @@ void kgd2kfd_device_exit(struct kfd_dev *kfd) kfree(kfd); } +int kgd2kfd_pre_reset(struct kfd_dev *kfd) +{ + if (!kfd->init_complete) + return 0; + kgd2kfd_suspend(kfd); + + /* hold dqm->lock to prevent further execution*/ + dqm_lock(kfd->dqm); + + kfd_signal_reset_event(kfd); + return 0; +} + +/* + * Fix me. KFD won't be able to resume existing process for now. + * We will keep all existing process in a evicted state and + * wait the process to be terminated. + */ + +int kgd2kfd_post_reset(struct kfd_dev *kfd) +{ + int ret, count; + + if (!kfd->init_complete) + return 0; + + dqm_unlock(kfd->dqm); + + ret = kfd_resume(kfd); + if (ret) + return ret; + count = atomic_dec_return(&kfd_locked); + WARN_ONCE(count != 0, "KFD reset ref. error"); + return 0; +} + +bool kfd_is_locked(void) +{ + return (atomic_read(&kfd_locked) > 0); +} + void kgd2kfd_suspend(struct kfd_dev *kfd) { if (!kfd->init_complete) return; /* For first KFD device suspend all the KFD processes */ - if (atomic_inc_return(&kfd_device_suspended) == 1) + if (atomic_inc_return(&kfd_locked) == 1) kfd_suspend_all_processes(); kfd->dqm->ops.stop(kfd->dqm); @@ -539,7 +614,7 @@ int kgd2kfd_resume(struct kfd_dev *kfd) if (ret) return ret; - count = atomic_dec_return(&kfd_device_suspended); + count = atomic_dec_return(&kfd_locked); WARN_ONCE(count < 0, "KFD suspend / resume ref. error"); if (count == 0) ret = kfd_resume_all_processes(); @@ -577,14 +652,24 @@ dqm_start_error: /* This is called directly from KGD at ISR. */ void kgd2kfd_interrupt(struct kfd_dev *kfd, const void *ih_ring_entry) { + uint32_t patched_ihre[KFD_MAX_RING_ENTRY_SIZE]; + bool is_patched = false; + if (!kfd->init_complete) return; + if (kfd->device_info->ih_ring_entry_size > sizeof(patched_ihre)) { + dev_err_once(kfd_device, "Ring entry too small\n"); + return; + } + spin_lock(&kfd->interrupt_lock); if (kfd->interrupts_active - && interrupt_is_wanted(kfd, ih_ring_entry) - && enqueue_ih_ring_entry(kfd, ih_ring_entry)) + && interrupt_is_wanted(kfd, ih_ring_entry, + patched_ihre, &is_patched) + && enqueue_ih_ring_entry(kfd, + is_patched ? patched_ihre : ih_ring_entry)) queue_work(kfd->ih_wq, &kfd->interrupt_work); spin_unlock(&kfd->interrupt_lock); @@ -739,8 +824,8 @@ int kfd_gtt_sa_allocate(struct kfd_dev *kfd, unsigned int size, if (size > kfd->gtt_sa_num_of_chunks * kfd->gtt_sa_chunk_size) return -ENOMEM; - *mem_obj = kzalloc(sizeof(struct kfd_mem_obj), GFP_NOIO); - if ((*mem_obj) == NULL) + *mem_obj = kzalloc(sizeof(struct kfd_mem_obj), GFP_KERNEL); + if (!(*mem_obj)) return -ENOMEM; pr_debug("Allocated mem_obj = %p for size = %d\n", *mem_obj, size); @@ -857,3 +942,26 @@ int kfd_gtt_sa_free(struct kfd_dev *kfd, struct kfd_mem_obj *mem_obj) kfree(mem_obj); return 0; } + +#if defined(CONFIG_DEBUG_FS) + +/* This function will send a package to HIQ to hang the HWS + * which will trigger a GPU reset and bring the HWS back to normal state + */ +int kfd_debugfs_hang_hws(struct kfd_dev *dev) +{ + int r = 0; + + if (dev->dqm->sched_policy != KFD_SCHED_POLICY_HWS) { + pr_err("HWS is not enabled"); + return -EINVAL; + } + + r = pm_debugfs_hang_hws(&dev->dqm->packets); + if (!r) + r = dqm_debugfs_execute_queues(dev->dqm); + + return r; +} + +#endif 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 668ad07ebe1f..ec0d62a16e53 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c @@ -61,6 +61,8 @@ static int create_sdma_queue_nocpsch(struct device_queue_manager *dqm, static void deallocate_sdma_queue(struct device_queue_manager *dqm, unsigned int sdma_queue_id); +static void kfd_process_hw_exception(struct work_struct *work); + static inline enum KFD_MQD_TYPE get_mqd_type_from_queue_type(enum kfd_queue_type type) { @@ -99,6 +101,17 @@ unsigned int get_pipes_per_mec(struct device_queue_manager *dqm) return dqm->dev->shared_resources.num_pipe_per_mec; } +static unsigned int get_num_sdma_engines(struct device_queue_manager *dqm) +{ + return dqm->dev->device_info->num_sdma_engines; +} + +unsigned int get_num_sdma_queues(struct device_queue_manager *dqm) +{ + return dqm->dev->device_info->num_sdma_engines + * KFD_SDMA_QUEUES_PER_ENGINE; +} + void program_sh_mem_settings(struct device_queue_manager *dqm, struct qcm_process_device *qpd) { @@ -240,7 +253,7 @@ static int create_queue_nocpsch(struct device_queue_manager *dqm, print_queue(q); - mutex_lock(&dqm->lock); + dqm_lock(dqm); if (dqm->total_queue_count >= max_num_of_queues_per_device) { pr_warn("Can't create new usermode queue because %d queues were already created\n", @@ -297,7 +310,7 @@ static int create_queue_nocpsch(struct device_queue_manager *dqm, dqm->total_queue_count); out_unlock: - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); return retval; } @@ -346,10 +359,10 @@ static int create_compute_queue_nocpsch(struct device_queue_manager *dqm, struct qcm_process_device *qpd) { int retval; - struct mqd_manager *mqd; + struct mqd_manager *mqd_mgr; - mqd = dqm->ops.get_mqd_manager(dqm, KFD_MQD_TYPE_COMPUTE); - if (!mqd) + mqd_mgr = dqm->ops.get_mqd_manager(dqm, KFD_MQD_TYPE_COMPUTE); + if (!mqd_mgr) return -ENOMEM; retval = allocate_hqd(dqm, q); @@ -360,7 +373,7 @@ static int create_compute_queue_nocpsch(struct device_queue_manager *dqm, if (retval) goto out_deallocate_hqd; - retval = mqd->init_mqd(mqd, &q->mqd, &q->mqd_mem_obj, + retval = mqd_mgr->init_mqd(mqd_mgr, &q->mqd, &q->mqd_mem_obj, &q->gart_mqd_addr, &q->properties); if (retval) goto out_deallocate_doorbell; @@ -374,15 +387,15 @@ static int create_compute_queue_nocpsch(struct device_queue_manager *dqm, if (!q->properties.is_active) return 0; - retval = mqd->load_mqd(mqd, q->mqd, q->pipe, q->queue, &q->properties, - q->process->mm); + retval = mqd_mgr->load_mqd(mqd_mgr, q->mqd, q->pipe, q->queue, + &q->properties, q->process->mm); if (retval) goto out_uninit_mqd; return 0; out_uninit_mqd: - mqd->uninit_mqd(mqd, q->mqd, q->mqd_mem_obj); + mqd_mgr->uninit_mqd(mqd_mgr, q->mqd, q->mqd_mem_obj); out_deallocate_doorbell: deallocate_doorbell(qpd, q); out_deallocate_hqd: @@ -399,11 +412,11 @@ static int destroy_queue_nocpsch_locked(struct device_queue_manager *dqm, struct queue *q) { int retval; - struct mqd_manager *mqd; + struct mqd_manager *mqd_mgr; - mqd = dqm->ops.get_mqd_manager(dqm, + mqd_mgr = dqm->ops.get_mqd_manager(dqm, get_mqd_type_from_queue_type(q->properties.type)); - if (!mqd) + if (!mqd_mgr) return -ENOMEM; if (q->properties.type == KFD_QUEUE_TYPE_COMPUTE) { @@ -420,14 +433,14 @@ static int destroy_queue_nocpsch_locked(struct device_queue_manager *dqm, deallocate_doorbell(qpd, q); - retval = mqd->destroy_mqd(mqd, q->mqd, + retval = mqd_mgr->destroy_mqd(mqd_mgr, q->mqd, KFD_PREEMPT_TYPE_WAVEFRONT_RESET, KFD_UNMAP_LATENCY_MS, q->pipe, q->queue); if (retval == -ETIME) qpd->reset_wavefronts = true; - mqd->uninit_mqd(mqd, q->mqd, q->mqd_mem_obj); + mqd_mgr->uninit_mqd(mqd_mgr, q->mqd, q->mqd_mem_obj); list_del(&q->list); if (list_empty(&qpd->queues_list)) { @@ -457,9 +470,9 @@ static int destroy_queue_nocpsch(struct device_queue_manager *dqm, { int retval; - mutex_lock(&dqm->lock); + dqm_lock(dqm); retval = destroy_queue_nocpsch_locked(dqm, qpd, q); - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); return retval; } @@ -467,19 +480,19 @@ static int destroy_queue_nocpsch(struct device_queue_manager *dqm, static int update_queue(struct device_queue_manager *dqm, struct queue *q) { int retval; - struct mqd_manager *mqd; + struct mqd_manager *mqd_mgr; struct kfd_process_device *pdd; bool prev_active = false; - mutex_lock(&dqm->lock); + dqm_lock(dqm); pdd = kfd_get_process_device_data(q->device, q->process); if (!pdd) { retval = -ENODEV; goto out_unlock; } - mqd = dqm->ops.get_mqd_manager(dqm, + mqd_mgr = dqm->ops.get_mqd_manager(dqm, get_mqd_type_from_queue_type(q->properties.type)); - if (!mqd) { + if (!mqd_mgr) { retval = -ENOMEM; goto out_unlock; } @@ -506,7 +519,7 @@ static int update_queue(struct device_queue_manager *dqm, struct queue *q) } else if (prev_active && (q->properties.type == KFD_QUEUE_TYPE_COMPUTE || q->properties.type == KFD_QUEUE_TYPE_SDMA)) { - retval = mqd->destroy_mqd(mqd, q->mqd, + retval = mqd_mgr->destroy_mqd(mqd_mgr, q->mqd, KFD_PREEMPT_TYPE_WAVEFRONT_DRAIN, KFD_UNMAP_LATENCY_MS, q->pipe, q->queue); if (retval) { @@ -515,7 +528,7 @@ static int update_queue(struct device_queue_manager *dqm, struct queue *q) } } - retval = mqd->update_mqd(mqd, q->mqd, &q->properties); + retval = mqd_mgr->update_mqd(mqd_mgr, q->mqd, &q->properties); /* * check active state vs. the previous state and modify @@ -533,44 +546,44 @@ static int update_queue(struct device_queue_manager *dqm, struct queue *q) else if (q->properties.is_active && (q->properties.type == KFD_QUEUE_TYPE_COMPUTE || q->properties.type == KFD_QUEUE_TYPE_SDMA)) - retval = mqd->load_mqd(mqd, q->mqd, q->pipe, q->queue, + retval = mqd_mgr->load_mqd(mqd_mgr, q->mqd, q->pipe, q->queue, &q->properties, q->process->mm); out_unlock: - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); return retval; } static struct mqd_manager *get_mqd_manager( struct device_queue_manager *dqm, enum KFD_MQD_TYPE type) { - struct mqd_manager *mqd; + struct mqd_manager *mqd_mgr; if (WARN_ON(type >= KFD_MQD_TYPE_MAX)) return NULL; pr_debug("mqd type %d\n", type); - mqd = dqm->mqds[type]; - if (!mqd) { - mqd = mqd_manager_init(type, dqm->dev); - if (!mqd) + mqd_mgr = dqm->mqd_mgrs[type]; + if (!mqd_mgr) { + mqd_mgr = mqd_manager_init(type, dqm->dev); + if (!mqd_mgr) pr_err("mqd manager is NULL"); - dqm->mqds[type] = mqd; + dqm->mqd_mgrs[type] = mqd_mgr; } - return mqd; + return mqd_mgr; } static int evict_process_queues_nocpsch(struct device_queue_manager *dqm, struct qcm_process_device *qpd) { struct queue *q; - struct mqd_manager *mqd; + struct mqd_manager *mqd_mgr; struct kfd_process_device *pdd; int retval = 0; - mutex_lock(&dqm->lock); + dqm_lock(dqm); if (qpd->evicted++ > 0) /* already evicted, do nothing */ goto out; @@ -582,16 +595,16 @@ static int evict_process_queues_nocpsch(struct device_queue_manager *dqm, list_for_each_entry(q, &qpd->queues_list, list) { if (!q->properties.is_active) continue; - mqd = dqm->ops.get_mqd_manager(dqm, + mqd_mgr = dqm->ops.get_mqd_manager(dqm, get_mqd_type_from_queue_type(q->properties.type)); - if (!mqd) { /* should not be here */ + if (!mqd_mgr) { /* should not be here */ pr_err("Cannot evict queue, mqd mgr is NULL\n"); retval = -ENOMEM; goto out; } q->properties.is_evicted = true; q->properties.is_active = false; - retval = mqd->destroy_mqd(mqd, q->mqd, + retval = mqd_mgr->destroy_mqd(mqd_mgr, q->mqd, KFD_PREEMPT_TYPE_WAVEFRONT_DRAIN, KFD_UNMAP_LATENCY_MS, q->pipe, q->queue); if (retval) @@ -600,7 +613,7 @@ static int evict_process_queues_nocpsch(struct device_queue_manager *dqm, } out: - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); return retval; } @@ -611,7 +624,7 @@ static int evict_process_queues_cpsch(struct device_queue_manager *dqm, struct kfd_process_device *pdd; int retval = 0; - mutex_lock(&dqm->lock); + dqm_lock(dqm); if (qpd->evicted++ > 0) /* already evicted, do nothing */ goto out; @@ -633,7 +646,7 @@ static int evict_process_queues_cpsch(struct device_queue_manager *dqm, KFD_UNMAP_QUEUES_FILTER_DYNAMIC_QUEUES, 0); out: - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); return retval; } @@ -641,7 +654,7 @@ static int restore_process_queues_nocpsch(struct device_queue_manager *dqm, struct qcm_process_device *qpd) { struct queue *q; - struct mqd_manager *mqd; + struct mqd_manager *mqd_mgr; struct kfd_process_device *pdd; uint32_t pd_base; int retval = 0; @@ -650,7 +663,7 @@ static int restore_process_queues_nocpsch(struct device_queue_manager *dqm, /* Retrieve PD base */ pd_base = dqm->dev->kfd2kgd->get_process_page_dir(pdd->vm); - mutex_lock(&dqm->lock); + dqm_lock(dqm); if (WARN_ON_ONCE(!qpd->evicted)) /* already restored, do nothing */ goto out; if (qpd->evicted > 1) { /* ref count still > 0, decrement & quit */ @@ -677,16 +690,16 @@ static int restore_process_queues_nocpsch(struct device_queue_manager *dqm, list_for_each_entry(q, &qpd->queues_list, list) { if (!q->properties.is_evicted) continue; - mqd = dqm->ops.get_mqd_manager(dqm, + mqd_mgr = dqm->ops.get_mqd_manager(dqm, get_mqd_type_from_queue_type(q->properties.type)); - if (!mqd) { /* should not be here */ + if (!mqd_mgr) { /* should not be here */ pr_err("Cannot restore queue, mqd mgr is NULL\n"); retval = -ENOMEM; goto out; } q->properties.is_evicted = false; q->properties.is_active = true; - retval = mqd->load_mqd(mqd, q->mqd, q->pipe, + retval = mqd_mgr->load_mqd(mqd_mgr, q->mqd, q->pipe, q->queue, &q->properties, q->process->mm); if (retval) @@ -695,7 +708,7 @@ static int restore_process_queues_nocpsch(struct device_queue_manager *dqm, } qpd->evicted = 0; out: - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); return retval; } @@ -711,7 +724,7 @@ static int restore_process_queues_cpsch(struct device_queue_manager *dqm, /* Retrieve PD base */ pd_base = dqm->dev->kfd2kgd->get_process_page_dir(pdd->vm); - mutex_lock(&dqm->lock); + dqm_lock(dqm); if (WARN_ON_ONCE(!qpd->evicted)) /* already restored, do nothing */ goto out; if (qpd->evicted > 1) { /* ref count still > 0, decrement & quit */ @@ -739,7 +752,7 @@ static int restore_process_queues_cpsch(struct device_queue_manager *dqm, if (!retval) qpd->evicted = 0; out: - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); return retval; } @@ -761,7 +774,7 @@ static int register_process(struct device_queue_manager *dqm, /* Retrieve PD base */ pd_base = dqm->dev->kfd2kgd->get_process_page_dir(pdd->vm); - mutex_lock(&dqm->lock); + dqm_lock(dqm); list_add(&n->list, &dqm->queues); /* Update PD Base in QPD */ @@ -769,9 +782,10 @@ static int register_process(struct device_queue_manager *dqm, retval = dqm->asic_ops.update_qpd(dqm, qpd); - dqm->processes_count++; + if (dqm->processes_count++ == 0) + dqm->dev->kfd2kgd->set_compute_idle(dqm->dev->kgd, false); - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); return retval; } @@ -786,20 +800,22 @@ static int unregister_process(struct device_queue_manager *dqm, list_empty(&qpd->queues_list) ? "empty" : "not empty"); retval = 0; - mutex_lock(&dqm->lock); + dqm_lock(dqm); list_for_each_entry_safe(cur, next, &dqm->queues, list) { if (qpd == cur->qpd) { list_del(&cur->list); kfree(cur); - dqm->processes_count--; + if (--dqm->processes_count == 0) + dqm->dev->kfd2kgd->set_compute_idle( + dqm->dev->kgd, true); goto out; } } /* qpd not found in dqm list */ retval = 1; out: - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); return retval; } @@ -838,7 +854,7 @@ static int initialize_nocpsch(struct device_queue_manager *dqm) if (!dqm->allocated_queues) return -ENOMEM; - mutex_init(&dqm->lock); + mutex_init(&dqm->lock_hidden); INIT_LIST_HEAD(&dqm->queues); dqm->queue_count = dqm->next_pipe_to_allocate = 0; dqm->sdma_queue_count = 0; @@ -853,7 +869,7 @@ static int initialize_nocpsch(struct device_queue_manager *dqm) } dqm->vmid_bitmap = (1 << dqm->dev->vm_info.vmid_num_kfd) - 1; - dqm->sdma_bitmap = (1 << CIK_SDMA_QUEUES) - 1; + dqm->sdma_bitmap = (1 << get_num_sdma_queues(dqm)) - 1; return 0; } @@ -866,8 +882,8 @@ static void uninitialize(struct device_queue_manager *dqm) kfree(dqm->allocated_queues); for (i = 0 ; i < KFD_MQD_TYPE_MAX ; i++) - kfree(dqm->mqds[i]); - mutex_destroy(&dqm->lock); + kfree(dqm->mqd_mgrs[i]); + mutex_destroy(&dqm->lock_hidden); kfd_gtt_sa_free(dqm->dev, dqm->pipeline_mem); } @@ -901,7 +917,7 @@ static int allocate_sdma_queue(struct device_queue_manager *dqm, static void deallocate_sdma_queue(struct device_queue_manager *dqm, unsigned int sdma_queue_id) { - if (sdma_queue_id >= CIK_SDMA_QUEUES) + if (sdma_queue_id >= get_num_sdma_queues(dqm)) return; dqm->sdma_bitmap |= (1 << sdma_queue_id); } @@ -910,19 +926,19 @@ static int create_sdma_queue_nocpsch(struct device_queue_manager *dqm, struct queue *q, struct qcm_process_device *qpd) { - struct mqd_manager *mqd; + struct mqd_manager *mqd_mgr; int retval; - mqd = dqm->ops.get_mqd_manager(dqm, KFD_MQD_TYPE_SDMA); - if (!mqd) + mqd_mgr = dqm->ops.get_mqd_manager(dqm, KFD_MQD_TYPE_SDMA); + if (!mqd_mgr) return -ENOMEM; retval = allocate_sdma_queue(dqm, &q->sdma_id); if (retval) return retval; - q->properties.sdma_queue_id = q->sdma_id / CIK_SDMA_QUEUES_PER_ENGINE; - q->properties.sdma_engine_id = q->sdma_id % CIK_SDMA_QUEUES_PER_ENGINE; + q->properties.sdma_queue_id = q->sdma_id / get_num_sdma_engines(dqm); + q->properties.sdma_engine_id = q->sdma_id % get_num_sdma_engines(dqm); retval = allocate_doorbell(qpd, q); if (retval) @@ -933,19 +949,20 @@ static int create_sdma_queue_nocpsch(struct device_queue_manager *dqm, pr_debug("SDMA engine id: %d\n", q->properties.sdma_engine_id); dqm->asic_ops.init_sdma_vm(dqm, q, qpd); - retval = mqd->init_mqd(mqd, &q->mqd, &q->mqd_mem_obj, + retval = mqd_mgr->init_mqd(mqd_mgr, &q->mqd, &q->mqd_mem_obj, &q->gart_mqd_addr, &q->properties); if (retval) goto out_deallocate_doorbell; - retval = mqd->load_mqd(mqd, q->mqd, 0, 0, &q->properties, NULL); + retval = mqd_mgr->load_mqd(mqd_mgr, q->mqd, 0, 0, &q->properties, + NULL); if (retval) goto out_uninit_mqd; return 0; out_uninit_mqd: - mqd->uninit_mqd(mqd, q->mqd, q->mqd_mem_obj); + mqd_mgr->uninit_mqd(mqd_mgr, q->mqd, q->mqd_mem_obj); out_deallocate_doorbell: deallocate_doorbell(qpd, q); out_deallocate_sdma_queue: @@ -1003,12 +1020,14 @@ static int initialize_cpsch(struct device_queue_manager *dqm) { pr_debug("num of pipes: %d\n", get_pipes_per_mec(dqm)); - mutex_init(&dqm->lock); + mutex_init(&dqm->lock_hidden); INIT_LIST_HEAD(&dqm->queues); dqm->queue_count = dqm->processes_count = 0; dqm->sdma_queue_count = 0; dqm->active_runlist = false; - dqm->sdma_bitmap = (1 << CIK_SDMA_QUEUES) - 1; + dqm->sdma_bitmap = (1 << get_num_sdma_queues(dqm)) - 1; + + INIT_WORK(&dqm->hw_exception_work, kfd_process_hw_exception); return 0; } @@ -1041,9 +1060,11 @@ static int start_cpsch(struct device_queue_manager *dqm) init_interrupts(dqm); - mutex_lock(&dqm->lock); + dqm_lock(dqm); + /* clear hang status when driver try to start the hw scheduler */ + dqm->is_hws_hang = false; execute_queues_cpsch(dqm, KFD_UNMAP_QUEUES_FILTER_DYNAMIC_QUEUES, 0); - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); return 0; fail_allocate_vidmem: @@ -1055,9 +1076,9 @@ fail_packet_manager_init: static int stop_cpsch(struct device_queue_manager *dqm) { - mutex_lock(&dqm->lock); + dqm_lock(dqm); unmap_queues_cpsch(dqm, KFD_UNMAP_QUEUES_FILTER_ALL_QUEUES, 0); - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); kfd_gtt_sa_free(dqm->dev, dqm->fence_mem); pm_uninit(&dqm->packets); @@ -1069,11 +1090,11 @@ static int create_kernel_queue_cpsch(struct device_queue_manager *dqm, struct kernel_queue *kq, struct qcm_process_device *qpd) { - mutex_lock(&dqm->lock); + dqm_lock(dqm); if (dqm->total_queue_count >= max_num_of_queues_per_device) { pr_warn("Can't create new kernel queue because %d queues were already created\n", dqm->total_queue_count); - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); return -EPERM; } @@ -1089,7 +1110,7 @@ static int create_kernel_queue_cpsch(struct device_queue_manager *dqm, dqm->queue_count++; qpd->is_debug = true; execute_queues_cpsch(dqm, KFD_UNMAP_QUEUES_FILTER_DYNAMIC_QUEUES, 0); - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); return 0; } @@ -1098,7 +1119,7 @@ static void destroy_kernel_queue_cpsch(struct device_queue_manager *dqm, struct kernel_queue *kq, struct qcm_process_device *qpd) { - mutex_lock(&dqm->lock); + dqm_lock(dqm); list_del(&kq->list); dqm->queue_count--; qpd->is_debug = false; @@ -1110,18 +1131,18 @@ static void destroy_kernel_queue_cpsch(struct device_queue_manager *dqm, dqm->total_queue_count--; pr_debug("Total of %d queues are accountable so far\n", dqm->total_queue_count); - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); } static int create_queue_cpsch(struct device_queue_manager *dqm, struct queue *q, struct qcm_process_device *qpd) { int retval; - struct mqd_manager *mqd; + struct mqd_manager *mqd_mgr; retval = 0; - mutex_lock(&dqm->lock); + dqm_lock(dqm); if (dqm->total_queue_count >= max_num_of_queues_per_device) { pr_warn("Can't create new usermode queue because %d queues were already created\n", @@ -1135,19 +1156,19 @@ static int create_queue_cpsch(struct device_queue_manager *dqm, struct queue *q, if (retval) goto out_unlock; q->properties.sdma_queue_id = - q->sdma_id / CIK_SDMA_QUEUES_PER_ENGINE; + q->sdma_id / get_num_sdma_engines(dqm); q->properties.sdma_engine_id = - q->sdma_id % CIK_SDMA_QUEUES_PER_ENGINE; + q->sdma_id % get_num_sdma_engines(dqm); } retval = allocate_doorbell(qpd, q); if (retval) goto out_deallocate_sdma_queue; - mqd = dqm->ops.get_mqd_manager(dqm, + mqd_mgr = dqm->ops.get_mqd_manager(dqm, get_mqd_type_from_queue_type(q->properties.type)); - if (!mqd) { + if (!mqd_mgr) { retval = -ENOMEM; goto out_deallocate_doorbell; } @@ -1164,7 +1185,7 @@ static int create_queue_cpsch(struct device_queue_manager *dqm, struct queue *q, q->properties.tba_addr = qpd->tba_addr; q->properties.tma_addr = qpd->tma_addr; - retval = mqd->init_mqd(mqd, &q->mqd, &q->mqd_mem_obj, + retval = mqd_mgr->init_mqd(mqd_mgr, &q->mqd, &q->mqd_mem_obj, &q->gart_mqd_addr, &q->properties); if (retval) goto out_deallocate_doorbell; @@ -1188,7 +1209,7 @@ static int create_queue_cpsch(struct device_queue_manager *dqm, struct queue *q, pr_debug("Total of %d queues are accountable so far\n", dqm->total_queue_count); - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); return retval; out_deallocate_doorbell: @@ -1197,7 +1218,8 @@ out_deallocate_sdma_queue: if (q->properties.type == KFD_QUEUE_TYPE_SDMA) deallocate_sdma_queue(dqm, q->sdma_id); out_unlock: - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); + return retval; } @@ -1210,6 +1232,13 @@ int amdkfd_fence_wait_timeout(unsigned int *fence_addr, while (*fence_addr != fence_value) { if (time_after(jiffies, end_jiffies)) { pr_err("qcm fence wait loop timeout expired\n"); + /* In HWS case, this is used to halt the driver thread + * in order not to mess up CP states before doing + * scandumps for FW debugging. + */ + while (halt_if_hws_hang) + schedule(); + return -ETIME; } schedule(); @@ -1254,6 +1283,8 @@ static int unmap_queues_cpsch(struct device_queue_manager *dqm, { int retval = 0; + if (dqm->is_hws_hang) + return -EIO; if (!dqm->active_runlist) return retval; @@ -1292,9 +1323,13 @@ static int execute_queues_cpsch(struct device_queue_manager *dqm, { int retval; + if (dqm->is_hws_hang) + return -EIO; retval = unmap_queues_cpsch(dqm, filter, filter_param); if (retval) { pr_err("The cp might be in an unrecoverable state due to an unsuccessful queues preemption\n"); + dqm->is_hws_hang = true; + schedule_work(&dqm->hw_exception_work); return retval; } @@ -1306,7 +1341,7 @@ static int destroy_queue_cpsch(struct device_queue_manager *dqm, struct queue *q) { int retval; - struct mqd_manager *mqd; + struct mqd_manager *mqd_mgr; bool preempt_all_queues; preempt_all_queues = false; @@ -1314,7 +1349,7 @@ static int destroy_queue_cpsch(struct device_queue_manager *dqm, retval = 0; /* remove queue from list to prevent rescheduling after preemption */ - mutex_lock(&dqm->lock); + dqm_lock(dqm); if (qpd->is_debug) { /* @@ -1326,9 +1361,9 @@ static int destroy_queue_cpsch(struct device_queue_manager *dqm, } - mqd = dqm->ops.get_mqd_manager(dqm, + mqd_mgr = dqm->ops.get_mqd_manager(dqm, get_mqd_type_from_queue_type(q->properties.type)); - if (!mqd) { + if (!mqd_mgr) { retval = -ENOMEM; goto failed; } @@ -1350,7 +1385,7 @@ static int destroy_queue_cpsch(struct device_queue_manager *dqm, qpd->reset_wavefronts = true; } - mqd->uninit_mqd(mqd, q->mqd, q->mqd_mem_obj); + mqd_mgr->uninit_mqd(mqd_mgr, q->mqd, q->mqd_mem_obj); /* * Unconditionally decrement this counter, regardless of the queue's @@ -1360,14 +1395,14 @@ static int destroy_queue_cpsch(struct device_queue_manager *dqm, pr_debug("Total of %d queues are accountable so far\n", dqm->total_queue_count); - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); return retval; failed: failed_try_destroy_debugged_queue: - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); return retval; } @@ -1391,7 +1426,7 @@ static bool set_cache_memory_policy(struct device_queue_manager *dqm, if (!dqm->asic_ops.set_cache_memory_policy) return retval; - mutex_lock(&dqm->lock); + dqm_lock(dqm); if (alternate_aperture_size == 0) { /* base > limit disables APE1 */ @@ -1437,7 +1472,7 @@ static bool set_cache_memory_policy(struct device_queue_manager *dqm, qpd->sh_mem_ape1_limit); out: - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); return retval; } @@ -1468,7 +1503,7 @@ static int process_termination_nocpsch(struct device_queue_manager *dqm, struct device_process_node *cur, *next_dpn; int retval = 0; - mutex_lock(&dqm->lock); + dqm_lock(dqm); /* Clear all user mode queues */ list_for_each_entry_safe(q, next, &qpd->queues_list, list) { @@ -1489,7 +1524,7 @@ static int process_termination_nocpsch(struct device_queue_manager *dqm, } } - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); return retval; } @@ -1500,14 +1535,14 @@ static int process_termination_cpsch(struct device_queue_manager *dqm, int retval; struct queue *q, *next; struct kernel_queue *kq, *kq_next; - struct mqd_manager *mqd; + struct mqd_manager *mqd_mgr; struct device_process_node *cur, *next_dpn; enum kfd_unmap_queues_filter filter = KFD_UNMAP_QUEUES_FILTER_DYNAMIC_QUEUES; retval = 0; - mutex_lock(&dqm->lock); + dqm_lock(dqm); /* Clean all kernel queues */ list_for_each_entry_safe(kq, kq_next, &qpd->priv_queue_list, list) { @@ -1542,7 +1577,7 @@ static int process_termination_cpsch(struct device_queue_manager *dqm, } retval = execute_queues_cpsch(dqm, filter, 0); - if (retval || qpd->reset_wavefronts) { + if ((!dqm->is_hws_hang) && (retval || qpd->reset_wavefronts)) { pr_warn("Resetting wave fronts (cpsch) on dev %p\n", dqm->dev); dbgdev_wave_reset_wavefronts(dqm->dev, qpd->pqm->process); qpd->reset_wavefronts = false; @@ -1550,19 +1585,19 @@ static int process_termination_cpsch(struct device_queue_manager *dqm, /* lastly, free mqd resources */ list_for_each_entry_safe(q, next, &qpd->queues_list, list) { - mqd = dqm->ops.get_mqd_manager(dqm, + mqd_mgr = dqm->ops.get_mqd_manager(dqm, get_mqd_type_from_queue_type(q->properties.type)); - if (!mqd) { + if (!mqd_mgr) { retval = -ENOMEM; goto out; } list_del(&q->list); qpd->queue_count--; - mqd->uninit_mqd(mqd, q->mqd, q->mqd_mem_obj); + mqd_mgr->uninit_mqd(mqd_mgr, q->mqd, q->mqd_mem_obj); } out: - mutex_unlock(&dqm->lock); + dqm_unlock(dqm); return retval; } @@ -1683,6 +1718,30 @@ void device_queue_manager_uninit(struct device_queue_manager *dqm) kfree(dqm); } +int kfd_process_vm_fault(struct device_queue_manager *dqm, + unsigned int pasid) +{ + struct kfd_process_device *pdd; + struct kfd_process *p = kfd_lookup_process_by_pasid(pasid); + int ret = 0; + + if (!p) + return -EINVAL; + pdd = kfd_get_process_device_data(dqm->dev, p); + if (pdd) + ret = dqm->ops.evict_process_queues(dqm, &pdd->qpd); + kfd_unref_process(p); + + return ret; +} + +static void kfd_process_hw_exception(struct work_struct *work) +{ + struct device_queue_manager *dqm = container_of(work, + struct device_queue_manager, hw_exception_work); + dqm->dev->kfd2kgd->gpu_recover(dqm->dev->kgd); +} + #if defined(CONFIG_DEBUG_FS) static void seq_reg_dump(struct seq_file *m, @@ -1746,8 +1805,8 @@ int dqm_debugfs_hqds(struct seq_file *m, void *data) } } - for (pipe = 0; pipe < CIK_SDMA_ENGINE_NUM; pipe++) { - for (queue = 0; queue < CIK_SDMA_QUEUES_PER_ENGINE; queue++) { + for (pipe = 0; pipe < get_num_sdma_engines(dqm); pipe++) { + for (queue = 0; queue < KFD_SDMA_QUEUES_PER_ENGINE; queue++) { r = dqm->dev->kfd2kgd->hqd_sdma_dump( dqm->dev->kgd, pipe, queue, &dump, &n_regs); if (r) @@ -1764,4 +1823,16 @@ int dqm_debugfs_hqds(struct seq_file *m, void *data) return r; } +int dqm_debugfs_execute_queues(struct device_queue_manager *dqm) +{ + int r = 0; + + dqm_lock(dqm); + dqm->active_runlist = true; + r = execute_queues_cpsch(dqm, KFD_UNMAP_QUEUES_FILTER_ALL_QUEUES, 0); + dqm_unlock(dqm); + + return r; +} + #endif diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h index 59a6b1956932..00da3169a004 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h @@ -26,15 +26,14 @@ #include <linux/rwsem.h> #include <linux/list.h> +#include <linux/mutex.h> +#include <linux/sched/mm.h> #include "kfd_priv.h" #include "kfd_mqd_manager.h" #define KFD_UNMAP_LATENCY_MS (4000) #define QUEUE_PREEMPT_DEFAULT_TIMEOUT_MS (2 * KFD_UNMAP_LATENCY_MS + 1000) - -#define CIK_SDMA_QUEUES (4) -#define CIK_SDMA_QUEUES_PER_ENGINE (2) -#define CIK_SDMA_ENGINE_NUM (2) +#define KFD_SDMA_QUEUES_PER_ENGINE (2) struct device_process_node { struct qcm_process_device *qpd; @@ -170,11 +169,12 @@ struct device_queue_manager { struct device_queue_manager_ops ops; struct device_queue_manager_asic_ops asic_ops; - struct mqd_manager *mqds[KFD_MQD_TYPE_MAX]; + struct mqd_manager *mqd_mgrs[KFD_MQD_TYPE_MAX]; struct packet_manager packets; struct kfd_dev *dev; - struct mutex lock; + struct mutex lock_hidden; /* use dqm_lock/unlock(dqm) */ struct list_head queues; + unsigned int saved_flags; unsigned int processes_count; unsigned int queue_count; unsigned int sdma_queue_count; @@ -190,6 +190,10 @@ struct device_queue_manager { struct kfd_mem_obj *fence_mem; bool active_runlist; int sched_policy; + + /* hw exception */ + bool is_hws_hang; + struct work_struct hw_exception_work; }; void device_queue_manager_init_cik( @@ -207,6 +211,7 @@ void program_sh_mem_settings(struct device_queue_manager *dqm, unsigned int get_queues_num(struct device_queue_manager *dqm); unsigned int get_queues_per_pipe(struct device_queue_manager *dqm); unsigned int get_pipes_per_mec(struct device_queue_manager *dqm); +unsigned int get_num_sdma_queues(struct device_queue_manager *dqm); static inline unsigned int get_sh_mem_bases_32(struct kfd_process_device *pdd) { @@ -219,4 +224,19 @@ get_sh_mem_bases_nybble_64(struct kfd_process_device *pdd) return (pdd->lds_base >> 60) & 0x0E; } +/* The DQM lock can be taken in MMU notifiers. Make sure no reclaim-FS + * happens while holding this lock anywhere to prevent deadlocks when + * an MMU notifier runs in reclaim-FS context. + */ +static inline void dqm_lock(struct device_queue_manager *dqm) +{ + mutex_lock(&dqm->lock_hidden); + dqm->saved_flags = memalloc_nofs_save(); +} +static inline void dqm_unlock(struct device_queue_manager *dqm) +{ + memalloc_nofs_restore(dqm->saved_flags); + mutex_unlock(&dqm->lock_hidden); +} + #endif /* KFD_DEVICE_QUEUE_MANAGER_H_ */ diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager_v9.c b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager_v9.c index 79e5bcf6367c..417515332c35 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager_v9.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager_v9.c @@ -60,7 +60,7 @@ static int update_qpd_v9(struct device_queue_manager *dqm, qpd->sh_mem_config = SH_MEM_ALIGNMENT_MODE_UNALIGNED << SH_MEM_CONFIG__ALIGNMENT_MODE__SHIFT; - if (vega10_noretry && + if (noretry && !dqm->dev->device_info->needs_iommu_device) qpd->sh_mem_config |= 1 << SH_MEM_CONFIG__RETRY_DISABLE__SHIFT; diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c b/drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c index c3744d89352c..ebe79bf00145 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c @@ -188,9 +188,9 @@ void __iomem *kfd_get_kernel_doorbell(struct kfd_dev *kfd, *doorbell_off = kfd->doorbell_id_offset + inx; pr_debug("Get kernel queue doorbell\n" - " doorbell offset == 0x%08X\n" - " kernel address == %p\n", - *doorbell_off, (kfd->doorbell_kernel_ptr + inx)); + " doorbell offset == 0x%08X\n" + " doorbell index == 0x%x\n", + *doorbell_off, inx); return kfd->doorbell_kernel_ptr + inx; } @@ -199,7 +199,8 @@ void kfd_release_kernel_doorbell(struct kfd_dev *kfd, u32 __iomem *db_addr) { unsigned int inx; - inx = (unsigned int)(db_addr - kfd->doorbell_kernel_ptr); + inx = (unsigned int)(db_addr - kfd->doorbell_kernel_ptr) + * sizeof(u32) / kfd->device_info->doorbell_size; mutex_lock(&kfd->doorbell_mutex); __clear_bit(inx, kfd->doorbell_available_index); diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_events.c b/drivers/gpu/drm/amd/amdkfd/kfd_events.c index 5562e94e786a..e9f0e0a1b41c 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_events.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_events.c @@ -850,6 +850,13 @@ static void lookup_events_by_type_and_signal(struct kfd_process *p, ev->memory_exception_data = *ev_data; } + if (type == KFD_EVENT_TYPE_MEMORY) { + dev_warn(kfd_device, + "Sending SIGSEGV to HSA Process with PID %d ", + p->lead_thread->pid); + send_sig(SIGSEGV, p->lead_thread, 0); + } + /* Send SIGTERM no event of type "type" has been found*/ if (send_signal) { if (send_sigterm) { @@ -904,34 +911,41 @@ void kfd_signal_iommu_event(struct kfd_dev *dev, unsigned int pasid, memory_exception_data.failure.NotPresent = 1; memory_exception_data.failure.NoExecute = 0; memory_exception_data.failure.ReadOnly = 0; - if (vma) { - if (vma->vm_start > address) { - memory_exception_data.failure.NotPresent = 1; - memory_exception_data.failure.NoExecute = 0; + if (vma && address >= vma->vm_start) { + memory_exception_data.failure.NotPresent = 0; + + if (is_write_requested && !(vma->vm_flags & VM_WRITE)) + memory_exception_data.failure.ReadOnly = 1; + else memory_exception_data.failure.ReadOnly = 0; - } else { - memory_exception_data.failure.NotPresent = 0; - if (is_write_requested && !(vma->vm_flags & VM_WRITE)) - memory_exception_data.failure.ReadOnly = 1; - else - memory_exception_data.failure.ReadOnly = 0; - if (is_execute_requested && !(vma->vm_flags & VM_EXEC)) - memory_exception_data.failure.NoExecute = 1; - else - memory_exception_data.failure.NoExecute = 0; - } + + if (is_execute_requested && !(vma->vm_flags & VM_EXEC)) + memory_exception_data.failure.NoExecute = 1; + else + memory_exception_data.failure.NoExecute = 0; } up_read(&mm->mmap_sem); mmput(mm); - mutex_lock(&p->event_mutex); + pr_debug("notpresent %d, noexecute %d, readonly %d\n", + memory_exception_data.failure.NotPresent, + memory_exception_data.failure.NoExecute, + memory_exception_data.failure.ReadOnly); - /* Lookup events by type and signal them */ - lookup_events_by_type_and_signal(p, KFD_EVENT_TYPE_MEMORY, - &memory_exception_data); + /* Workaround on Raven to not kill the process when memory is freed + * before IOMMU is able to finish processing all the excessive PPRs + */ + if (dev->device_info->asic_family != CHIP_RAVEN) { + mutex_lock(&p->event_mutex); + + /* Lookup events by type and signal them */ + lookup_events_by_type_and_signal(p, KFD_EVENT_TYPE_MEMORY, + &memory_exception_data); + + mutex_unlock(&p->event_mutex); + } - mutex_unlock(&p->event_mutex); kfd_unref_process(p); } #endif /* KFD_SUPPORT_IOMMU_V2 */ @@ -956,3 +970,67 @@ void kfd_signal_hw_exception_event(unsigned int pasid) mutex_unlock(&p->event_mutex); kfd_unref_process(p); } + +void kfd_signal_vm_fault_event(struct kfd_dev *dev, unsigned int pasid, + struct kfd_vm_fault_info *info) +{ + struct kfd_event *ev; + uint32_t id; + struct kfd_process *p = kfd_lookup_process_by_pasid(pasid); + struct kfd_hsa_memory_exception_data memory_exception_data; + + if (!p) + return; /* Presumably process exited. */ + memset(&memory_exception_data, 0, sizeof(memory_exception_data)); + memory_exception_data.gpu_id = dev->id; + memory_exception_data.failure.imprecise = 1; + /* Set failure reason */ + if (info) { + memory_exception_data.va = (info->page_addr) << PAGE_SHIFT; + memory_exception_data.failure.NotPresent = + info->prot_valid ? 1 : 0; + memory_exception_data.failure.NoExecute = + info->prot_exec ? 1 : 0; + memory_exception_data.failure.ReadOnly = + info->prot_write ? 1 : 0; + memory_exception_data.failure.imprecise = 0; + } + mutex_lock(&p->event_mutex); + + id = KFD_FIRST_NONSIGNAL_EVENT_ID; + idr_for_each_entry_continue(&p->event_idr, ev, id) + if (ev->type == KFD_EVENT_TYPE_MEMORY) { + ev->memory_exception_data = memory_exception_data; + set_event(ev); + } + + mutex_unlock(&p->event_mutex); + kfd_unref_process(p); +} + +void kfd_signal_reset_event(struct kfd_dev *dev) +{ + struct kfd_hsa_hw_exception_data hw_exception_data; + struct kfd_process *p; + struct kfd_event *ev; + unsigned int temp; + uint32_t id, idx; + + /* Whole gpu reset caused by GPU hang and memory is lost */ + memset(&hw_exception_data, 0, sizeof(hw_exception_data)); + hw_exception_data.gpu_id = dev->id; + hw_exception_data.memory_lost = 1; + + idx = srcu_read_lock(&kfd_processes_srcu); + hash_for_each_rcu(kfd_processes_table, temp, p, kfd_processes) { + mutex_lock(&p->event_mutex); + id = KFD_FIRST_NONSIGNAL_EVENT_ID; + idr_for_each_entry_continue(&p->event_idr, ev, id) + if (ev->type == KFD_EVENT_TYPE_HW_EXCEPTION) { + ev->hw_exception_data = hw_exception_data; + set_event(ev); + } + mutex_unlock(&p->event_mutex); + } + srcu_read_unlock(&kfd_processes_srcu, idx); +} diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_events.h b/drivers/gpu/drm/amd/amdkfd/kfd_events.h index abca5bfebbff..c7ac6c73af86 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_events.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_events.h @@ -66,6 +66,7 @@ struct kfd_event { /* type specific data */ union { struct kfd_hsa_memory_exception_data memory_exception_data; + struct kfd_hsa_hw_exception_data hw_exception_data; }; }; diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c b/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c index 37029baa3346..f836897bbf58 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c @@ -26,7 +26,9 @@ static bool event_interrupt_isr_v9(struct kfd_dev *dev, - const uint32_t *ih_ring_entry) + const uint32_t *ih_ring_entry, + uint32_t *patched_ihre, + bool *patched_flag) { uint16_t source_id, client_id, pasid, vmid; const uint32_t *data = ih_ring_entry; @@ -57,7 +59,9 @@ static bool event_interrupt_isr_v9(struct kfd_dev *dev, return source_id == SOC15_INTSRC_CP_END_OF_PIPE || source_id == SOC15_INTSRC_SDMA_TRAP || source_id == SOC15_INTSRC_SQ_INTERRUPT_MSG || - source_id == SOC15_INTSRC_CP_BAD_OPCODE; + source_id == SOC15_INTSRC_CP_BAD_OPCODE || + client_id == SOC15_IH_CLIENTID_VMC || + client_id == SOC15_IH_CLIENTID_UTCL2; } static void event_interrupt_wq_v9(struct kfd_dev *dev, @@ -82,7 +86,19 @@ static void event_interrupt_wq_v9(struct kfd_dev *dev, kfd_signal_hw_exception_event(pasid); else if (client_id == SOC15_IH_CLIENTID_VMC || client_id == SOC15_IH_CLIENTID_UTCL2) { - /* TODO */ + struct kfd_vm_fault_info info = {0}; + uint16_t ring_id = SOC15_RING_ID_FROM_IH_ENTRY(ih_ring_entry); + + info.vmid = vmid; + info.mc_id = client_id; + info.page_addr = ih_ring_entry[4] | + (uint64_t)(ih_ring_entry[5] & 0xf) << 32; + info.prot_valid = ring_id & 0x08; + info.prot_read = ring_id & 0x10; + info.prot_write = ring_id & 0x20; + + kfd_process_vm_fault(dev->dqm, pasid); + kfd_signal_vm_fault_event(dev, pasid, &info); } } diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_interrupt.c b/drivers/gpu/drm/amd/amdkfd/kfd_interrupt.c index db6d9336b80d..c56ac47cd318 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_interrupt.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_interrupt.c @@ -151,13 +151,15 @@ static void interrupt_wq(struct work_struct *work) ih_ring_entry); } -bool interrupt_is_wanted(struct kfd_dev *dev, const uint32_t *ih_ring_entry) +bool interrupt_is_wanted(struct kfd_dev *dev, + const uint32_t *ih_ring_entry, + uint32_t *patched_ihre, bool *flag) { /* integer and bitwise OR so there is no boolean short-circuiting */ unsigned int wanted = 0; wanted |= dev->device_info->event_interrupt_class->interrupt_isr(dev, - ih_ring_entry); + ih_ring_entry, patched_ihre, flag); return wanted != 0; } diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_iommu.c b/drivers/gpu/drm/amd/amdkfd/kfd_iommu.c index c71817963eea..7a61f38c09e6 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_iommu.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_iommu.c @@ -190,7 +190,7 @@ static int iommu_invalid_ppr_cb(struct pci_dev *pdev, int pasid, { struct kfd_dev *dev; - dev_warn(kfd_device, + dev_warn_ratelimited(kfd_device, "Invalid PPR device %x:%x.%x pasid %d address 0x%lX flags 0x%X", PCI_BUS_NUM(pdev->devfn), PCI_SLOT(pdev->devfn), diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c index 476951d8c91c..9f84b4d9fb88 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c @@ -59,7 +59,7 @@ static bool initialize(struct kernel_queue *kq, struct kfd_dev *dev, switch (type) { case KFD_QUEUE_TYPE_DIQ: case KFD_QUEUE_TYPE_HIQ: - kq->mqd = dev->dqm->ops.get_mqd_manager(dev->dqm, + kq->mqd_mgr = dev->dqm->ops.get_mqd_manager(dev->dqm, KFD_MQD_TYPE_HIQ); break; default: @@ -67,7 +67,7 @@ static bool initialize(struct kernel_queue *kq, struct kfd_dev *dev, return false; } - if (!kq->mqd) + if (!kq->mqd_mgr) return false; prop.doorbell_ptr = kfd_get_kernel_doorbell(dev, &prop.doorbell_off); @@ -123,6 +123,7 @@ static bool initialize(struct kernel_queue *kq, struct kfd_dev *dev, prop.write_ptr = (uint32_t *) kq->wptr_gpu_addr; prop.eop_ring_buffer_address = kq->eop_gpu_addr; prop.eop_ring_buffer_size = PAGE_SIZE; + prop.cu_mask = NULL; if (init_queue(&kq->queue, &prop) != 0) goto err_init_queue; @@ -130,7 +131,7 @@ static bool initialize(struct kernel_queue *kq, struct kfd_dev *dev, kq->queue->device = dev; kq->queue->process = kfd_get_process(current); - retval = kq->mqd->init_mqd(kq->mqd, &kq->queue->mqd, + retval = kq->mqd_mgr->init_mqd(kq->mqd_mgr, &kq->queue->mqd, &kq->queue->mqd_mem_obj, &kq->queue->gart_mqd_addr, &kq->queue->properties); @@ -142,9 +143,9 @@ static bool initialize(struct kernel_queue *kq, struct kfd_dev *dev, pr_debug("Assigning hiq to hqd\n"); kq->queue->pipe = KFD_CIK_HIQ_PIPE; kq->queue->queue = KFD_CIK_HIQ_QUEUE; - kq->mqd->load_mqd(kq->mqd, kq->queue->mqd, kq->queue->pipe, - kq->queue->queue, &kq->queue->properties, - NULL); + kq->mqd_mgr->load_mqd(kq->mqd_mgr, kq->queue->mqd, + kq->queue->pipe, kq->queue->queue, + &kq->queue->properties, NULL); } else { /* allocate fence for DIQ */ @@ -182,7 +183,7 @@ err_get_kernel_doorbell: static void uninitialize(struct kernel_queue *kq) { if (kq->queue->properties.type == KFD_QUEUE_TYPE_HIQ) - kq->mqd->destroy_mqd(kq->mqd, + kq->mqd_mgr->destroy_mqd(kq->mqd_mgr, kq->queue->mqd, KFD_PREEMPT_TYPE_WAVEFRONT_RESET, KFD_UNMAP_LATENCY_MS, @@ -191,7 +192,8 @@ static void uninitialize(struct kernel_queue *kq) else if (kq->queue->properties.type == KFD_QUEUE_TYPE_DIQ) kfd_gtt_sa_free(kq->dev, kq->fence_mem_obj); - kq->mqd->uninit_mqd(kq->mqd, kq->queue->mqd, kq->queue->mqd_mem_obj); + kq->mqd_mgr->uninit_mqd(kq->mqd_mgr, kq->queue->mqd, + kq->queue->mqd_mem_obj); kfd_gtt_sa_free(kq->dev, kq->rptr_mem); kfd_gtt_sa_free(kq->dev, kq->wptr_mem); diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.h b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.h index 97aff2041a5d..a7116a939029 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.h @@ -70,7 +70,7 @@ struct kernel_queue { /* data */ struct kfd_dev *dev; - struct mqd_manager *mqd; + struct mqd_manager *mqd_mgr; struct queue *queue; uint64_t pending_wptr64; uint32_t pending_wptr; diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_module.c b/drivers/gpu/drm/amd/amdkfd/kfd_module.c index 76bf2dc8aec4..6e1f5c7c2d4b 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_module.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_module.c @@ -47,6 +47,8 @@ static const struct kgd2kfd_calls kgd2kfd = { .resume_mm = kgd2kfd_resume_mm, .schedule_evict_and_restore_process = kgd2kfd_schedule_evict_and_restore_process, + .pre_reset = kgd2kfd_pre_reset, + .post_reset = kgd2kfd_post_reset, }; int sched_policy = KFD_SCHED_POLICY_HWS; @@ -61,7 +63,7 @@ MODULE_PARM_DESC(hws_max_conc_proc, int cwsr_enable = 1; module_param(cwsr_enable, int, 0444); -MODULE_PARM_DESC(cwsr_enable, "CWSR enable (0 = Off, 1 = On (Default))"); +MODULE_PARM_DESC(cwsr_enable, "CWSR enable (0 = off, 1 = on (default))"); int max_num_of_queues_per_device = KFD_MAX_NUM_OF_QUEUES_PER_DEVICE_DEFAULT; module_param(max_num_of_queues_per_device, int, 0444); @@ -83,13 +85,19 @@ module_param(ignore_crat, int, 0444); MODULE_PARM_DESC(ignore_crat, "Ignore CRAT table during KFD initialization (0 = use CRAT (default), 1 = ignore CRAT)"); -int vega10_noretry; -module_param_named(noretry, vega10_noretry, int, 0644); +int noretry; +module_param(noretry, int, 0644); MODULE_PARM_DESC(noretry, - "Set sh_mem_config.retry_disable on Vega10 (0 = retry enabled (default), 1 = retry disabled)"); + "Set sh_mem_config.retry_disable on GFXv9+ dGPUs (0 = retry enabled (default), 1 = retry disabled)"); + +int halt_if_hws_hang; +module_param(halt_if_hws_hang, int, 0644); +MODULE_PARM_DESC(halt_if_hws_hang, "Halt if HWS hang is detected (0 = off (default), 1 = on)"); + static int amdkfd_init_completed; + int kgd2kfd_init(unsigned int interface_version, const struct kgd2kfd_calls **g2f) { diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.c index 4b8eb506642b..3bc25ab84f34 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.c @@ -21,7 +21,7 @@ * */ -#include "kfd_priv.h" +#include "kfd_mqd_manager.h" struct mqd_manager *mqd_manager_init(enum KFD_MQD_TYPE type, struct kfd_dev *dev) @@ -48,3 +48,42 @@ struct mqd_manager *mqd_manager_init(enum KFD_MQD_TYPE type, return NULL; } + +void mqd_symmetrically_map_cu_mask(struct mqd_manager *mm, + const uint32_t *cu_mask, uint32_t cu_mask_count, + uint32_t *se_mask) +{ + struct kfd_cu_info cu_info; + uint32_t cu_per_sh[4] = {0}; + int i, se, cu = 0; + + mm->dev->kfd2kgd->get_cu_info(mm->dev->kgd, &cu_info); + + if (cu_mask_count > cu_info.cu_active_number) + cu_mask_count = cu_info.cu_active_number; + + for (se = 0; se < cu_info.num_shader_engines; se++) + for (i = 0; i < 4; i++) + cu_per_sh[se] += hweight32(cu_info.cu_bitmap[se][i]); + + /* Symmetrically map cu_mask to all SEs: + * cu_mask[0] bit0 -> se_mask[0] bit0; + * cu_mask[0] bit1 -> se_mask[1] bit0; + * ... (if # SE is 4) + * cu_mask[0] bit4 -> se_mask[0] bit1; + * ... + */ + se = 0; + for (i = 0; i < cu_mask_count; i++) { + if (cu_mask[i / 32] & (1 << (i % 32))) + se_mask[se] |= 1 << cu; + + do { + se++; + if (se == cu_info.num_shader_engines) { + se = 0; + cu++; + } + } while (cu >= cu_per_sh[se] && cu < 32); + } +} diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.h b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.h index 8972bcfbf701..4e84052d4e21 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.h @@ -93,4 +93,8 @@ struct mqd_manager { struct kfd_dev *dev; }; +void mqd_symmetrically_map_cu_mask(struct mqd_manager *mm, + const uint32_t *cu_mask, uint32_t cu_mask_count, + uint32_t *se_mask); + #endif /* KFD_MQD_MANAGER_H_ */ diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c index 06eaa218eba6..47243165a082 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c @@ -41,6 +41,31 @@ static inline struct cik_sdma_rlc_registers *get_sdma_mqd(void *mqd) return (struct cik_sdma_rlc_registers *)mqd; } +static void update_cu_mask(struct mqd_manager *mm, void *mqd, + struct queue_properties *q) +{ + struct cik_mqd *m; + uint32_t se_mask[4] = {0}; /* 4 is the max # of SEs */ + + if (q->cu_mask_count == 0) + return; + + mqd_symmetrically_map_cu_mask(mm, + q->cu_mask, q->cu_mask_count, se_mask); + + m = get_mqd(mqd); + m->compute_static_thread_mgmt_se0 = se_mask[0]; + m->compute_static_thread_mgmt_se1 = se_mask[1]; + m->compute_static_thread_mgmt_se2 = se_mask[2]; + m->compute_static_thread_mgmt_se3 = se_mask[3]; + + pr_debug("Update cu mask to %#x %#x %#x %#x\n", + m->compute_static_thread_mgmt_se0, + m->compute_static_thread_mgmt_se1, + m->compute_static_thread_mgmt_se2, + m->compute_static_thread_mgmt_se3); +} + static int init_mqd(struct mqd_manager *mm, void **mqd, struct kfd_mem_obj **mqd_mem_obj, uint64_t *gart_addr, struct queue_properties *q) @@ -196,6 +221,8 @@ static int __update_mqd(struct mqd_manager *mm, void *mqd, if (q->format == KFD_QUEUE_FORMAT_AQL) m->cp_hqd_pq_control |= NO_UPDATE_RPTR; + update_cu_mask(mm, mqd, q); + q->is_active = (q->queue_size > 0 && q->queue_address != 0 && q->queue_percent > 0 && @@ -408,7 +435,7 @@ struct mqd_manager *mqd_manager_init_cik(enum KFD_MQD_TYPE type, if (WARN_ON(type >= KFD_MQD_TYPE_MAX)) return NULL; - mqd = kzalloc(sizeof(*mqd), GFP_NOIO); + mqd = kzalloc(sizeof(*mqd), GFP_KERNEL); if (!mqd) return NULL; diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c index 684054ff02cd..f5fc3675f21e 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c @@ -41,6 +41,31 @@ static inline struct v9_sdma_mqd *get_sdma_mqd(void *mqd) return (struct v9_sdma_mqd *)mqd; } +static void update_cu_mask(struct mqd_manager *mm, void *mqd, + struct queue_properties *q) +{ + struct v9_mqd *m; + uint32_t se_mask[4] = {0}; /* 4 is the max # of SEs */ + + if (q->cu_mask_count == 0) + return; + + mqd_symmetrically_map_cu_mask(mm, + q->cu_mask, q->cu_mask_count, se_mask); + + m = get_mqd(mqd); + m->compute_static_thread_mgmt_se0 = se_mask[0]; + m->compute_static_thread_mgmt_se1 = se_mask[1]; + m->compute_static_thread_mgmt_se2 = se_mask[2]; + m->compute_static_thread_mgmt_se3 = se_mask[3]; + + pr_debug("update cu mask to %#x %#x %#x %#x\n", + m->compute_static_thread_mgmt_se0, + m->compute_static_thread_mgmt_se1, + m->compute_static_thread_mgmt_se2, + m->compute_static_thread_mgmt_se3); +} + static int init_mqd(struct mqd_manager *mm, void **mqd, struct kfd_mem_obj **mqd_mem_obj, uint64_t *gart_addr, struct queue_properties *q) @@ -55,7 +80,7 @@ static int init_mqd(struct mqd_manager *mm, void **mqd, * instead of sub-allocation function. */ if (kfd->cwsr_enabled && (q->type == KFD_QUEUE_TYPE_COMPUTE)) { - *mqd_mem_obj = kzalloc(sizeof(struct kfd_mem_obj), GFP_NOIO); + *mqd_mem_obj = kzalloc(sizeof(struct kfd_mem_obj), GFP_KERNEL); if (!*mqd_mem_obj) return -ENOMEM; retval = kfd->kfd2kgd->init_gtt_mem_allocation(kfd->kgd, @@ -198,6 +223,8 @@ static int update_mqd(struct mqd_manager *mm, void *mqd, if (mm->dev->cwsr_enabled && q->ctx_save_restore_area_address) m->cp_hqd_ctx_save_control = 0; + update_cu_mask(mm, mqd, q); + q->is_active = (q->queue_size > 0 && q->queue_address != 0 && q->queue_percent > 0 && @@ -393,7 +420,7 @@ struct mqd_manager *mqd_manager_init_v9(enum KFD_MQD_TYPE type, if (WARN_ON(type >= KFD_MQD_TYPE_MAX)) return NULL; - mqd = kzalloc(sizeof(*mqd), GFP_NOIO); + mqd = kzalloc(sizeof(*mqd), GFP_KERNEL); if (!mqd) return NULL; diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c index 481307b8b4db..b81fda3754da 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c @@ -43,6 +43,31 @@ static inline struct vi_sdma_mqd *get_sdma_mqd(void *mqd) return (struct vi_sdma_mqd *)mqd; } +static void update_cu_mask(struct mqd_manager *mm, void *mqd, + struct queue_properties *q) +{ + struct vi_mqd *m; + uint32_t se_mask[4] = {0}; /* 4 is the max # of SEs */ + + if (q->cu_mask_count == 0) + return; + + mqd_symmetrically_map_cu_mask(mm, + q->cu_mask, q->cu_mask_count, se_mask); + + m = get_mqd(mqd); + m->compute_static_thread_mgmt_se0 = se_mask[0]; + m->compute_static_thread_mgmt_se1 = se_mask[1]; + m->compute_static_thread_mgmt_se2 = se_mask[2]; + m->compute_static_thread_mgmt_se3 = se_mask[3]; + + pr_debug("Update cu mask to %#x %#x %#x %#x\n", + m->compute_static_thread_mgmt_se0, + m->compute_static_thread_mgmt_se1, + m->compute_static_thread_mgmt_se2, + m->compute_static_thread_mgmt_se3); +} + static int init_mqd(struct mqd_manager *mm, void **mqd, struct kfd_mem_obj **mqd_mem_obj, uint64_t *gart_addr, struct queue_properties *q) @@ -196,6 +221,8 @@ static int __update_mqd(struct mqd_manager *mm, void *mqd, atc_bit << CP_HQD_CTX_SAVE_CONTROL__ATC__SHIFT | mtype << CP_HQD_CTX_SAVE_CONTROL__MTYPE__SHIFT; + update_cu_mask(mm, mqd, q); + q->is_active = (q->queue_size > 0 && q->queue_address != 0 && q->queue_percent > 0 && @@ -394,7 +421,7 @@ struct mqd_manager *mqd_manager_init_vi(enum KFD_MQD_TYPE type, if (WARN_ON(type >= KFD_MQD_TYPE_MAX)) return NULL; - mqd = kzalloc(sizeof(*mqd), GFP_NOIO); + mqd = kzalloc(sizeof(*mqd), GFP_KERNEL); if (!mqd) return NULL; diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c index c317feb43f69..1092631765cb 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c @@ -418,4 +418,30 @@ out: return 0; } +int pm_debugfs_hang_hws(struct packet_manager *pm) +{ + uint32_t *buffer, size; + int r = 0; + + size = pm->pmf->query_status_size; + mutex_lock(&pm->lock); + pm->priv_queue->ops.acquire_packet_buffer(pm->priv_queue, + size / sizeof(uint32_t), (unsigned int **)&buffer); + if (!buffer) { + pr_err("Failed to allocate buffer on kernel queue\n"); + r = -ENOMEM; + goto out; + } + memset(buffer, 0x55, size); + pm->priv_queue->ops.submit_packet(pm->priv_queue); + + pr_info("Submitting %x %x %x %x %x %x %x to HIQ to hang the HWS.", + buffer[0], buffer[1], buffer[2], buffer[3], + buffer[4], buffer[5], buffer[6]); +out: + mutex_unlock(&pm->lock); + return r; +} + + #endif diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h index 5e3990bb4c4b..f971710f1c91 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h @@ -73,7 +73,7 @@ /* * When working with cp scheduler we should assign the HIQ manually or via - * the radeon driver to a fixed hqd slot, here are the fixed HIQ hqd slot + * the amdgpu driver to a fixed hqd slot, here are the fixed HIQ hqd slot * definitions for Kaveri. In Kaveri only the first ME queues participates * in the cp scheduling taking that in mind we set the HIQ slot in the * second ME. @@ -142,7 +142,12 @@ extern int ignore_crat; /* * Set sh_mem_config.retry_disable on Vega10 */ -extern int vega10_noretry; +extern int noretry; + +/* + * Halt if HWS hang is detected + */ +extern int halt_if_hws_hang; /** * enum kfd_sched_policy @@ -180,9 +185,10 @@ enum cache_policy { struct kfd_event_interrupt_class { bool (*interrupt_isr)(struct kfd_dev *dev, - const uint32_t *ih_ring_entry); + const uint32_t *ih_ring_entry, uint32_t *patched_ihre, + bool *patched_flag); void (*interrupt_wq)(struct kfd_dev *dev, - const uint32_t *ih_ring_entry); + const uint32_t *ih_ring_entry); }; struct kfd_device_info { @@ -197,6 +203,7 @@ struct kfd_device_info { bool supports_cwsr; bool needs_iommu_device; bool needs_pci_atomics; + unsigned int num_sdma_engines; }; struct kfd_mem_obj { @@ -415,6 +422,9 @@ struct queue_properties { uint32_t ctl_stack_size; uint64_t tba_addr; uint64_t tma_addr; + /* Relevant for CU */ + uint32_t cu_mask_count; /* Must be a multiple of 32 */ + uint32_t *cu_mask; }; /** @@ -806,12 +816,18 @@ int kfd_interrupt_init(struct kfd_dev *dev); void kfd_interrupt_exit(struct kfd_dev *dev); void kgd2kfd_interrupt(struct kfd_dev *kfd, const void *ih_ring_entry); bool enqueue_ih_ring_entry(struct kfd_dev *kfd, const void *ih_ring_entry); -bool interrupt_is_wanted(struct kfd_dev *dev, const uint32_t *ih_ring_entry); +bool interrupt_is_wanted(struct kfd_dev *dev, + const uint32_t *ih_ring_entry, + uint32_t *patched_ihre, bool *flag); /* Power Management */ void kgd2kfd_suspend(struct kfd_dev *kfd); int kgd2kfd_resume(struct kfd_dev *kfd); +/* GPU reset */ +int kgd2kfd_pre_reset(struct kfd_dev *kfd); +int kgd2kfd_post_reset(struct kfd_dev *kfd); + /* amdkfd Apertures */ int kfd_init_apertures(struct kfd_process *process); @@ -838,6 +854,7 @@ void device_queue_manager_uninit(struct device_queue_manager *dqm); struct kernel_queue *kernel_queue_init(struct kfd_dev *dev, enum kfd_queue_type type); void kernel_queue_uninit(struct kernel_queue *kq); +int kfd_process_vm_fault(struct device_queue_manager *dqm, unsigned int pasid); /* Process Queue Manager */ struct process_queue_node { @@ -858,6 +875,8 @@ int pqm_create_queue(struct process_queue_manager *pqm, int pqm_destroy_queue(struct process_queue_manager *pqm, unsigned int qid); int pqm_update_queue(struct process_queue_manager *pqm, unsigned int qid, struct queue_properties *p); +int pqm_set_cu_mask(struct process_queue_manager *pqm, unsigned int qid, + struct queue_properties *p); struct kernel_queue *pqm_get_kernel_queue(struct process_queue_manager *pqm, unsigned int qid); @@ -964,10 +983,17 @@ int kfd_event_create(struct file *devkfd, struct kfd_process *p, uint64_t *event_page_offset, uint32_t *event_slot_index); int kfd_event_destroy(struct kfd_process *p, uint32_t event_id); +void kfd_signal_vm_fault_event(struct kfd_dev *dev, unsigned int pasid, + struct kfd_vm_fault_info *info); + +void kfd_signal_reset_event(struct kfd_dev *dev); + void kfd_flush_tlb(struct kfd_process_device *pdd); int dbgdev_wave_reset_wavefronts(struct kfd_dev *dev, struct kfd_process *p); +bool kfd_is_locked(void); + /* Debugfs */ #if defined(CONFIG_DEBUG_FS) @@ -980,6 +1006,10 @@ int dqm_debugfs_hqds(struct seq_file *m, void *data); int kfd_debugfs_rls_by_device(struct seq_file *m, void *data); int pm_debugfs_runlist(struct seq_file *m, void *data); +int kfd_debugfs_hang_hws(struct kfd_dev *dev); +int pm_debugfs_hang_hws(struct packet_manager *pm); +int dqm_debugfs_execute_queues(struct device_queue_manager *dqm); + #else static inline void kfd_debugfs_init(void) {} diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process.c b/drivers/gpu/drm/amd/amdkfd/kfd_process.c index 1d80b4f7c681..4694386cc623 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_process.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_process.c @@ -244,6 +244,8 @@ struct kfd_process *kfd_get_process(const struct task_struct *thread) return ERR_PTR(-EINVAL); process = find_process(thread); + if (!process) + return ERR_PTR(-EINVAL); return process; } diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c index d65ce0436b31..c8cad9c078ae 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c @@ -186,8 +186,7 @@ int pqm_create_queue(struct process_queue_manager *pqm, switch (type) { case KFD_QUEUE_TYPE_SDMA: - if (dev->dqm->queue_count >= - CIK_SDMA_QUEUES_PER_ENGINE * CIK_SDMA_ENGINE_NUM) { + if (dev->dqm->queue_count >= get_num_sdma_queues(dev->dqm)) { pr_err("Over-subscription is not allowed for SDMA.\n"); retval = -EPERM; goto err_create_queue; @@ -209,7 +208,7 @@ int pqm_create_queue(struct process_queue_manager *pqm, KFD_SCHED_POLICY_HWS_NO_OVERSUBSCRIPTION) && ((dev->dqm->processes_count >= dev->vm_info.vmid_num_kfd) || (dev->dqm->queue_count >= get_queues_num(dev->dqm)))) { - pr_err("Over-subscription is not allowed in radeon_kfd.sched_policy == 1\n"); + pr_debug("Over-subscription is not allowed when amdkfd.sched_policy == 1\n"); retval = -EPERM; goto err_create_queue; } @@ -326,6 +325,8 @@ int pqm_destroy_queue(struct process_queue_manager *pqm, unsigned int qid) if (retval != -ETIME) goto err_destroy_queue; } + kfree(pqn->q->properties.cu_mask); + pqn->q->properties.cu_mask = NULL; uninit_queue(pqn->q); } @@ -366,6 +367,34 @@ int pqm_update_queue(struct process_queue_manager *pqm, unsigned int qid, return 0; } +int pqm_set_cu_mask(struct process_queue_manager *pqm, unsigned int qid, + struct queue_properties *p) +{ + int retval; + struct process_queue_node *pqn; + + pqn = get_queue_by_qid(pqm, qid); + if (!pqn) { + pr_debug("No queue %d exists for update operation\n", qid); + return -EFAULT; + } + + /* Free the old CU mask memory if it is already allocated, then + * allocate memory for the new CU mask. + */ + kfree(pqn->q->properties.cu_mask); + + pqn->q->properties.cu_mask_count = p->cu_mask_count; + pqn->q->properties.cu_mask = p->cu_mask; + + retval = pqn->q->device->dqm->ops.update_queue(pqn->q->device->dqm, + pqn->q); + if (retval != 0) + return retval; + + return 0; +} + struct kernel_queue *pqm_get_kernel_queue( struct process_queue_manager *pqm, unsigned int qid) @@ -387,7 +416,7 @@ int pqm_debugfs_mqds(struct seq_file *m, void *data) struct process_queue_node *pqn; struct queue *q; enum KFD_MQD_TYPE mqd_type; - struct mqd_manager *mqd_manager; + struct mqd_manager *mqd_mgr; int r = 0; list_for_each_entry(pqn, &pqm->queues, process_queue_list) { @@ -410,11 +439,11 @@ int pqm_debugfs_mqds(struct seq_file *m, void *data) q->properties.type, q->device->id); continue; } - mqd_manager = q->device->dqm->ops.get_mqd_manager( + mqd_mgr = q->device->dqm->ops.get_mqd_manager( q->device->dqm, mqd_type); } else if (pqn->kq) { q = pqn->kq->queue; - mqd_manager = pqn->kq->mqd; + mqd_mgr = pqn->kq->mqd_mgr; switch (q->properties.type) { case KFD_QUEUE_TYPE_DIQ: seq_printf(m, " DIQ on device %x\n", @@ -434,7 +463,7 @@ int pqm_debugfs_mqds(struct seq_file *m, void *data) continue; } - r = mqd_manager->debugfs_show_mqd(m, q->mqd); + r = mqd_mgr->debugfs_show_mqd(m, q->mqd); if (r != 0) break; } diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c index 9f0a217603ad..516795342dd2 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c @@ -74,4 +74,3 @@ bool dm_read_persistent_data(struct dc_context *ctx, /**** power component interfaces ****/ - 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 58ee9aad13fb..160841da72a7 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 @@ -1784,12 +1784,10 @@ static void dp_test_send_link_training(struct dc_link *link) dp_retrain_link_dp_test(link, &link_settings, false); } -/* TODO hbr2 compliance eye output is unstable +/* TODO Raven hbr2 compliance eye output is unstable * (toggling on and off) with debugger break * This caueses intermittent PHY automation failure * Need to look into the root cause */ -static uint8_t force_tps4_for_cp2520 = 1; - static void dp_test_send_phy_test_pattern(struct dc_link *link) { union phy_test_pattern dpcd_test_pattern; @@ -1849,13 +1847,13 @@ static void dp_test_send_phy_test_pattern(struct dc_link *link) break; case PHY_TEST_PATTERN_CP2520_1: /* CP2520 pattern is unstable, temporarily use TPS4 instead */ - test_pattern = (force_tps4_for_cp2520 == 1) ? + test_pattern = (link->dc->caps.force_dp_tps4_for_cp2520 == 1) ? DP_TEST_PATTERN_TRAINING_PATTERN4 : DP_TEST_PATTERN_HBR2_COMPLIANCE_EYE; break; case PHY_TEST_PATTERN_CP2520_2: /* CP2520 pattern is unstable, temporarily use TPS4 instead */ - test_pattern = (force_tps4_for_cp2520 == 1) ? + test_pattern = (link->dc->caps.force_dp_tps4_for_cp2520 == 1) ? DP_TEST_PATTERN_TRAINING_PATTERN4 : DP_TEST_PATTERN_HBR2_COMPLIANCE_EYE; break; diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h index c2d390fba0bf..55bcc3bdc6a3 100644 --- a/drivers/gpu/drm/amd/display/dc/dc.h +++ b/drivers/gpu/drm/amd/display/dc/dc.h @@ -77,6 +77,7 @@ struct dc_caps { bool is_apu; bool dual_link_dvi; bool post_blend_color_processing; + bool force_dp_tps4_for_cp2520; }; struct dc_dcc_surface_param { diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c index 5d169f07d745..6b44ed3697a4 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c @@ -1127,6 +1127,8 @@ static bool construct( dc->caps.max_slave_planes = 1; dc->caps.is_apu = true; dc->caps.post_blend_color_processing = false; + /* Raven DP PHY HBR2 eye diagram pattern is not stable. Use TP4 */ + dc->caps.force_dp_tps4_for_cp2520 = true; if (dc->ctx->dce_environment == DCE_ENV_PRODUCTION_DRV) dc->debug = debug_defaults_drv; diff --git a/drivers/gpu/drm/amd/include/kgd_kfd_interface.h b/drivers/gpu/drm/amd/include/kgd_kfd_interface.h index 5733fbee07f7..14391b06080c 100644 --- a/drivers/gpu/drm/amd/include/kgd_kfd_interface.h +++ b/drivers/gpu/drm/amd/include/kgd_kfd_interface.h @@ -47,6 +47,17 @@ enum kfd_preempt_type { KFD_PREEMPT_TYPE_WAVEFRONT_RESET, }; +struct kfd_vm_fault_info { + uint64_t page_addr; + uint32_t vmid; + uint32_t mc_id; + uint32_t status; + bool prot_valid; + bool prot_read; + bool prot_write; + bool prot_exec; +}; + struct kfd_cu_info { uint32_t num_shader_engines; uint32_t num_shader_arrays_per_engine; @@ -259,6 +270,21 @@ struct tile_config { * IB to the corresponding ring (ring type). The IB is executed with the * specified VMID in a user mode context. * + * @get_vm_fault_info: Return information about a recent VM fault on + * GFXv7 and v8. If multiple VM faults occurred since the last call of + * this function, it will return information about the first of those + * faults. On GFXv9 VM fault information is fully contained in the IH + * packet and this function is not needed. + * + * @read_vmid_from_vmfault_reg: On Hawaii the VMID is not set in the + * IH ring entry. This function allows the KFD ISR to get the VMID + * from the fault status register as early as possible. + * + * @gpu_recover: let kgd reset gpu after kfd detect CPC hang + * + * @set_compute_idle: Indicates that compute is idle on a device. This + * can be used to change power profiles depending on compute activity. + * * This structure contains function pointers to services that the kgd driver * provides to amdkfd driver. * @@ -374,6 +400,14 @@ struct kfd2kgd_calls { int (*submit_ib)(struct kgd_dev *kgd, enum kgd_engine_type engine, uint32_t vmid, uint64_t gpu_addr, uint32_t *ib_cmd, uint32_t ib_len); + + int (*get_vm_fault_info)(struct kgd_dev *kgd, + struct kfd_vm_fault_info *info); + uint32_t (*read_vmid_from_vmfault_reg)(struct kgd_dev *kgd); + + void (*gpu_recover)(struct kgd_dev *kgd); + + void (*set_compute_idle)(struct kgd_dev *kgd, bool idle); }; /** @@ -399,6 +433,10 @@ struct kfd2kgd_calls { * @schedule_evict_and_restore_process: Schedules work queue that will prepare * for safe eviction of KFD BOs that belong to the specified process. * + * @pre_reset: Notifies amdkfd that amdgpu about to reset the gpu + * + * @post_reset: Notify amdkfd that amgpu successfully reseted the gpu + * * This structure contains function callback pointers so the kgd driver * will notify to the amdkfd about certain status changes. * @@ -417,6 +455,8 @@ struct kgd2kfd_calls { int (*resume_mm)(struct mm_struct *mm); int (*schedule_evict_and_restore_process)(struct mm_struct *mm, struct dma_fence *fence); + int (*pre_reset)(struct kfd_dev *kfd); + int (*post_reset)(struct kfd_dev *kfd); }; int kgd2kfd_init(unsigned interface_version, diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_processpptables.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_processpptables.c index f4f366b26fd1..cb3a5b1737c8 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_processpptables.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_processpptables.c @@ -226,6 +226,8 @@ static int append_vbios_pptable(struct pp_hwmgr *hwmgr, PPTable_t *ppsmc_pptable ppsmc_pptable->Vr2_I2C_address = smc_dpm_table.Vr2_I2C_address; + ppsmc_pptable->Vr2_I2C_address = smc_dpm_table.Vr2_I2C_address; + return 0; } diff --git a/drivers/gpu/drm/arm/hdlcd_crtc.c b/drivers/gpu/drm/arm/hdlcd_crtc.c index f3f08cd6e9ef..e4d67b70244d 100644 --- a/drivers/gpu/drm/arm/hdlcd_crtc.c +++ b/drivers/gpu/drm/arm/hdlcd_crtc.c @@ -229,6 +229,8 @@ static const struct drm_crtc_helper_funcs hdlcd_crtc_helper_funcs = { static int hdlcd_plane_atomic_check(struct drm_plane *plane, struct drm_plane_state *state) { + int i; + struct drm_crtc *crtc; struct drm_crtc_state *crtc_state; u32 src_h = state->src_h >> 16; @@ -238,20 +240,17 @@ static int hdlcd_plane_atomic_check(struct drm_plane *plane, return -EINVAL; } - if (!state->fb || !state->crtc) - return 0; - - crtc_state = drm_atomic_get_existing_crtc_state(state->state, - state->crtc); - if (!crtc_state) { - DRM_DEBUG_KMS("Invalid crtc state\n"); - return -EINVAL; + for_each_new_crtc_in_state(state->state, crtc, crtc_state, i) { + /* we cannot disable the plane while the CRTC is active */ + if (!state->fb && crtc_state->active) + return -EINVAL; + return drm_atomic_helper_check_plane_state(state, crtc_state, + DRM_PLANE_HELPER_NO_SCALING, + DRM_PLANE_HELPER_NO_SCALING, + false, true); } - return drm_atomic_helper_check_plane_state(state, crtc_state, - DRM_PLANE_HELPER_NO_SCALING, - DRM_PLANE_HELPER_NO_SCALING, - false, true); + return 0; } static void hdlcd_plane_atomic_update(struct drm_plane *plane, @@ -280,16 +279,10 @@ static const struct drm_plane_helper_funcs hdlcd_plane_helper_funcs = { .atomic_update = hdlcd_plane_atomic_update, }; -static void hdlcd_plane_destroy(struct drm_plane *plane) -{ - drm_plane_helper_disable(plane, NULL); - drm_plane_cleanup(plane); -} - static const struct drm_plane_funcs hdlcd_plane_funcs = { .update_plane = drm_atomic_helper_update_plane, .disable_plane = drm_atomic_helper_disable_plane, - .destroy = hdlcd_plane_destroy, + .destroy = drm_plane_cleanup, .reset = drm_atomic_helper_plane_reset, .atomic_duplicate_state = drm_atomic_helper_plane_duplicate_state, .atomic_destroy_state = drm_atomic_helper_plane_destroy_state, @@ -334,10 +327,8 @@ int hdlcd_setup_crtc(struct drm_device *drm) ret = drm_crtc_init_with_planes(drm, &hdlcd->crtc, primary, NULL, &hdlcd_crtc_funcs, NULL); - if (ret) { - hdlcd_plane_destroy(primary); + if (ret) return ret; - } drm_crtc_helper_add(&hdlcd->crtc, &hdlcd_crtc_helper_funcs); return 0; diff --git a/drivers/gpu/drm/arm/hdlcd_drv.c b/drivers/gpu/drm/arm/hdlcd_drv.c index feaa8bc3d7b7..0ed1cde98cf8 100644 --- a/drivers/gpu/drm/arm/hdlcd_drv.c +++ b/drivers/gpu/drm/arm/hdlcd_drv.c @@ -27,6 +27,7 @@ #include <drm/drm_fb_cma_helper.h> #include <drm/drm_gem_cma_helper.h> #include <drm/drm_gem_framebuffer_helper.h> +#include <drm/drm_modeset_helper.h> #include <drm/drm_of.h> #include "hdlcd_drv.h" @@ -100,16 +101,9 @@ setup_fail: return ret; } -static void hdlcd_fb_output_poll_changed(struct drm_device *drm) -{ - struct hdlcd_drm_private *hdlcd = drm->dev_private; - - drm_fbdev_cma_hotplug_event(hdlcd->fbdev); -} - static const struct drm_mode_config_funcs hdlcd_mode_config_funcs = { .fb_create = drm_gem_fb_create, - .output_poll_changed = hdlcd_fb_output_poll_changed, + .output_poll_changed = drm_fb_helper_output_poll_changed, .atomic_check = drm_atomic_helper_check, .atomic_commit = drm_atomic_helper_commit, }; @@ -124,13 +118,6 @@ static void hdlcd_setup_mode_config(struct drm_device *drm) drm->mode_config.funcs = &hdlcd_mode_config_funcs; } -static void hdlcd_lastclose(struct drm_device *drm) -{ - struct hdlcd_drm_private *hdlcd = drm->dev_private; - - drm_fbdev_cma_restore_mode(hdlcd->fbdev); -} - static irqreturn_t hdlcd_irq(int irq, void *arg) { struct drm_device *drm = arg; @@ -246,7 +233,7 @@ static struct drm_driver hdlcd_driver = { .driver_features = DRIVER_HAVE_IRQ | DRIVER_GEM | DRIVER_MODESET | DRIVER_PRIME | DRIVER_ATOMIC, - .lastclose = hdlcd_lastclose, + .lastclose = drm_fb_helper_lastclose, .irq_handler = hdlcd_irq, .irq_preinstall = hdlcd_irq_preinstall, .irq_postinstall = hdlcd_irq_postinstall, @@ -321,14 +308,9 @@ static int hdlcd_drm_bind(struct device *dev) drm_mode_config_reset(drm); drm_kms_helper_poll_init(drm); - hdlcd->fbdev = drm_fbdev_cma_init(drm, 32, - drm->mode_config.num_connector); - - if (IS_ERR(hdlcd->fbdev)) { - ret = PTR_ERR(hdlcd->fbdev); - hdlcd->fbdev = NULL; + ret = drm_fb_cma_fbdev_init(drm, 32, 0); + if (ret) goto err_fbdev; - } ret = drm_dev_register(drm, 0); if (ret) @@ -337,15 +319,13 @@ static int hdlcd_drm_bind(struct device *dev) return 0; err_register: - if (hdlcd->fbdev) { - drm_fbdev_cma_fini(hdlcd->fbdev); - hdlcd->fbdev = NULL; - } + drm_fb_cma_fbdev_fini(drm); err_fbdev: drm_kms_helper_poll_fini(drm); err_vblank: pm_runtime_disable(drm->dev); err_pm_active: + drm_atomic_helper_shutdown(drm); component_unbind_all(dev, drm); err_unload: of_node_put(hdlcd->crtc.port); @@ -366,23 +346,23 @@ static void hdlcd_drm_unbind(struct device *dev) struct hdlcd_drm_private *hdlcd = drm->dev_private; drm_dev_unregister(drm); - if (hdlcd->fbdev) { - drm_fbdev_cma_fini(hdlcd->fbdev); - hdlcd->fbdev = NULL; - } + drm_fb_cma_fbdev_fini(drm); drm_kms_helper_poll_fini(drm); component_unbind_all(dev, drm); of_node_put(hdlcd->crtc.port); hdlcd->crtc.port = NULL; - pm_runtime_get_sync(drm->dev); + pm_runtime_get_sync(dev); + drm_crtc_vblank_off(&hdlcd->crtc); drm_irq_uninstall(drm); - pm_runtime_put_sync(drm->dev); - pm_runtime_disable(drm->dev); - of_reserved_mem_device_release(drm->dev); + drm_atomic_helper_shutdown(drm); + pm_runtime_put(dev); + if (pm_runtime_enabled(dev)) + pm_runtime_disable(dev); + of_reserved_mem_device_release(dev); drm_mode_config_cleanup(drm); - drm_dev_put(drm); drm->dev_private = NULL; dev_set_drvdata(dev, NULL); + drm_dev_put(drm); } static const struct component_master_ops hdlcd_master_ops = { @@ -427,35 +407,15 @@ MODULE_DEVICE_TABLE(of, hdlcd_of_match); static int __maybe_unused hdlcd_pm_suspend(struct device *dev) { struct drm_device *drm = dev_get_drvdata(dev); - struct hdlcd_drm_private *hdlcd = drm ? drm->dev_private : NULL; - - if (!hdlcd) - return 0; - drm_kms_helper_poll_disable(drm); - drm_fbdev_cma_set_suspend_unlocked(hdlcd->fbdev, 1); - - hdlcd->state = drm_atomic_helper_suspend(drm); - if (IS_ERR(hdlcd->state)) { - drm_fbdev_cma_set_suspend_unlocked(hdlcd->fbdev, 0); - drm_kms_helper_poll_enable(drm); - return PTR_ERR(hdlcd->state); - } - - return 0; + return drm_mode_config_helper_suspend(drm); } static int __maybe_unused hdlcd_pm_resume(struct device *dev) { struct drm_device *drm = dev_get_drvdata(dev); - struct hdlcd_drm_private *hdlcd = drm ? drm->dev_private : NULL; - - if (!hdlcd) - return 0; - drm_atomic_helper_resume(drm, hdlcd->state); - drm_fbdev_cma_set_suspend_unlocked(hdlcd->fbdev, 0); - drm_kms_helper_poll_enable(drm); + drm_mode_config_helper_resume(drm); return 0; } diff --git a/drivers/gpu/drm/arm/hdlcd_drv.h b/drivers/gpu/drm/arm/hdlcd_drv.h index 56f34dfff640..fd438d177b64 100644 --- a/drivers/gpu/drm/arm/hdlcd_drv.h +++ b/drivers/gpu/drm/arm/hdlcd_drv.h @@ -9,10 +9,8 @@ struct hdlcd_drm_private { void __iomem *mmio; struct clk *clk; - struct drm_fbdev_cma *fbdev; struct drm_crtc crtc; struct drm_plane *plane; - struct drm_atomic_state *state; #ifdef CONFIG_DEBUG_FS atomic_t buffer_underrun_count; atomic_t bus_error_count; diff --git a/drivers/gpu/drm/armada/Makefile b/drivers/gpu/drm/armada/Makefile index ecf25cf9f9f5..9bc3c3213724 100644 --- a/drivers/gpu/drm/armada/Makefile +++ b/drivers/gpu/drm/armada/Makefile @@ -1,6 +1,6 @@ # SPDX-License-Identifier: GPL-2.0 armada-y := armada_crtc.o armada_drv.o armada_fb.o armada_fbdev.o \ - armada_gem.o armada_overlay.o armada_trace.o + armada_gem.o armada_overlay.o armada_plane.o armada_trace.o armada-y += armada_510.o armada-$(CONFIG_DEBUG_FS) += armada_debugfs.o diff --git a/drivers/gpu/drm/armada/armada_510.c b/drivers/gpu/drm/armada/armada_510.c index 41a784f5a5e6..2f7c048c5361 100644 --- a/drivers/gpu/drm/armada/armada_510.c +++ b/drivers/gpu/drm/armada/armada_510.c @@ -27,6 +27,10 @@ static int armada510_crtc_init(struct armada_crtc *dcrtc, struct device *dev) /* Lower the watermark so to eliminate jitter at higher bandwidths */ armada_updatel(0x20, (1 << 11) | 0xff, dcrtc->base + LCD_CFG_RDREG4F); + /* Initialise SPU register */ + writel_relaxed(ADV_HWC32ENABLE | ADV_HWC32ARGB | ADV_HWC32BLEND, + dcrtc->base + LCD_SPU_ADV_REG); + return 0; } @@ -75,9 +79,27 @@ static int armada510_crtc_compute_clock(struct armada_crtc *dcrtc, return 0; } +static void armada510_crtc_disable(struct armada_crtc *dcrtc) +{ + if (!IS_ERR(dcrtc->clk)) { + clk_disable_unprepare(dcrtc->clk); + dcrtc->clk = ERR_PTR(-EINVAL); + } +} + +static void armada510_crtc_enable(struct armada_crtc *dcrtc, + const struct drm_display_mode *mode) +{ + if (IS_ERR(dcrtc->clk)) { + dcrtc->clk = dcrtc->extclk[0]; + WARN_ON(clk_prepare_enable(dcrtc->clk)); + } +} + const struct armada_variant armada510_ops = { .has_spu_adv_reg = true, - .spu_adv_reg = ADV_HWC32ENABLE | ADV_HWC32ARGB | ADV_HWC32BLEND, .init = armada510_crtc_init, .compute_clock = armada510_crtc_compute_clock, + .disable = armada510_crtc_disable, + .enable = armada510_crtc_enable, }; diff --git a/drivers/gpu/drm/armada/armada_crtc.c b/drivers/gpu/drm/armada/armada_crtc.c index 03eeee11dd5b..da9360688b55 100644 --- a/drivers/gpu/drm/armada/armada_crtc.c +++ b/drivers/gpu/drm/armada/armada_crtc.c @@ -11,6 +11,7 @@ #include <linux/of_device.h> #include <linux/platform_device.h> #include <drm/drmP.h> +#include <drm/drm_atomic.h> #include <drm/drm_crtc_helper.h> #include <drm/drm_plane_helper.h> #include <drm/drm_atomic_helper.h> @@ -19,33 +20,9 @@ #include "armada_fb.h" #include "armada_gem.h" #include "armada_hw.h" +#include "armada_plane.h" #include "armada_trace.h" -enum csc_mode { - CSC_AUTO = 0, - CSC_YUV_CCIR601 = 1, - CSC_YUV_CCIR709 = 2, - CSC_RGB_COMPUTER = 1, - CSC_RGB_STUDIO = 2, -}; - -static const uint32_t armada_primary_formats[] = { - DRM_FORMAT_UYVY, - DRM_FORMAT_YUYV, - DRM_FORMAT_VYUY, - DRM_FORMAT_YVYU, - DRM_FORMAT_ARGB8888, - DRM_FORMAT_ABGR8888, - DRM_FORMAT_XRGB8888, - DRM_FORMAT_XBGR8888, - DRM_FORMAT_RGB888, - DRM_FORMAT_BGR888, - DRM_FORMAT_ARGB1555, - DRM_FORMAT_ABGR1555, - DRM_FORMAT_RGB565, - DRM_FORMAT_BGR565, -}; - /* * A note about interlacing. Let's consider HDMI 1920x1080i. * The timing parameters we have from X are: @@ -115,15 +92,13 @@ armada_drm_crtc_update_regs(struct armada_crtc *dcrtc, struct armada_regs *regs) } } -#define dpms_blanked(dpms) ((dpms) != DRM_MODE_DPMS_ON) - -static void armada_drm_crtc_update(struct armada_crtc *dcrtc) +static void armada_drm_crtc_update(struct armada_crtc *dcrtc, bool enable) { uint32_t dumb_ctrl; dumb_ctrl = dcrtc->cfg_dumb_ctrl; - if (!dpms_blanked(dcrtc->dpms)) + if (enable) dumb_ctrl |= CFG_DUMB_ENA; /* @@ -132,295 +107,26 @@ static void armada_drm_crtc_update(struct armada_crtc *dcrtc) * force LCD_D[23:0] to output blank color, overriding the GPIO or * SPI usage. So leave it as-is unless in DUMB24_RGB888_0 mode. */ - if (dpms_blanked(dcrtc->dpms) && - (dumb_ctrl & DUMB_MASK) == DUMB24_RGB888_0) { + if (!enable && (dumb_ctrl & DUMB_MASK) == DUMB24_RGB888_0) { dumb_ctrl &= ~DUMB_MASK; dumb_ctrl |= DUMB_BLANK; } - /* - * The documentation doesn't indicate what the normal state of - * the sync signals are. Sebastian Hesselbart kindly probed - * these signals on his board to determine their state. - * - * The non-inverted state of the sync signals is active high. - * Setting these bits makes the appropriate signal active low. - */ - if (dcrtc->crtc.mode.flags & DRM_MODE_FLAG_NCSYNC) - dumb_ctrl |= CFG_INV_CSYNC; - if (dcrtc->crtc.mode.flags & DRM_MODE_FLAG_NHSYNC) - dumb_ctrl |= CFG_INV_HSYNC; - if (dcrtc->crtc.mode.flags & DRM_MODE_FLAG_NVSYNC) - dumb_ctrl |= CFG_INV_VSYNC; - - if (dcrtc->dumb_ctrl != dumb_ctrl) { - dcrtc->dumb_ctrl = dumb_ctrl; - writel_relaxed(dumb_ctrl, dcrtc->base + LCD_SPU_DUMB_CTRL); - } -} - -void armada_drm_plane_calc_addrs(u32 *addrs, struct drm_framebuffer *fb, - int x, int y) -{ - const struct drm_format_info *format = fb->format; - unsigned int num_planes = format->num_planes; - u32 addr = drm_fb_obj(fb)->dev_addr; - int i; - - if (num_planes > 3) - num_planes = 3; - - addrs[0] = addr + fb->offsets[0] + y * fb->pitches[0] + - x * format->cpp[0]; - - y /= format->vsub; - x /= format->hsub; - - for (i = 1; i < num_planes; i++) - addrs[i] = addr + fb->offsets[i] + y * fb->pitches[i] + - x * format->cpp[i]; - for (; i < 3; i++) - addrs[i] = 0; -} - -static unsigned armada_drm_crtc_calc_fb(struct drm_framebuffer *fb, - int x, int y, struct armada_regs *regs, bool interlaced) -{ - unsigned pitch = fb->pitches[0]; - u32 addrs[3], addr_odd, addr_even; - unsigned i = 0; - - DRM_DEBUG_DRIVER("pitch %u x %d y %d bpp %d\n", - pitch, x, y, fb->format->cpp[0] * 8); - - armada_drm_plane_calc_addrs(addrs, fb, x, y); - - addr_odd = addr_even = addrs[0]; - - if (interlaced) { - addr_even += pitch; - pitch *= 2; - } - - /* write offset, base, and pitch */ - armada_reg_queue_set(regs, i, addr_odd, LCD_CFG_GRA_START_ADDR0); - armada_reg_queue_set(regs, i, addr_even, LCD_CFG_GRA_START_ADDR1); - armada_reg_queue_mod(regs, i, pitch, 0xffff, LCD_CFG_GRA_PITCH); - - return i; -} - -static void armada_drm_plane_work_call(struct armada_crtc *dcrtc, - struct armada_plane_work *work, - void (*fn)(struct armada_crtc *, struct armada_plane_work *)) -{ - struct armada_plane *dplane = drm_to_armada_plane(work->plane); - struct drm_pending_vblank_event *event; - struct drm_framebuffer *fb; - - if (fn) - fn(dcrtc, work); - drm_crtc_vblank_put(&dcrtc->crtc); - - event = work->event; - fb = work->old_fb; - if (event || fb) { - struct drm_device *dev = dcrtc->crtc.dev; - unsigned long flags; - - spin_lock_irqsave(&dev->event_lock, flags); - if (event) - drm_crtc_send_vblank_event(&dcrtc->crtc, event); - if (fb) - __armada_drm_queue_unref_work(dev, fb); - spin_unlock_irqrestore(&dev->event_lock, flags); - } - - if (work->need_kfree) - kfree(work); - - wake_up(&dplane->frame_wait); + armada_updatel(dumb_ctrl, + ~(CFG_INV_CSYNC | CFG_INV_HSYNC | CFG_INV_VSYNC), + dcrtc->base + LCD_SPU_DUMB_CTRL); } -static void armada_drm_plane_work_run(struct armada_crtc *dcrtc, - struct drm_plane *plane) -{ - struct armada_plane *dplane = drm_to_armada_plane(plane); - struct armada_plane_work *work = xchg(&dplane->work, NULL); - - /* Handle any pending frame work. */ - if (work) - armada_drm_plane_work_call(dcrtc, work, work->fn); -} - -int armada_drm_plane_work_queue(struct armada_crtc *dcrtc, - struct armada_plane_work *work) -{ - struct armada_plane *plane = drm_to_armada_plane(work->plane); - int ret; - - ret = drm_crtc_vblank_get(&dcrtc->crtc); - if (ret) - return ret; - - ret = cmpxchg(&plane->work, NULL, work) ? -EBUSY : 0; - if (ret) - drm_crtc_vblank_put(&dcrtc->crtc); - - return ret; -} - -int armada_drm_plane_work_wait(struct armada_plane *plane, long timeout) -{ - return wait_event_timeout(plane->frame_wait, !plane->work, timeout); -} - -void armada_drm_plane_work_cancel(struct armada_crtc *dcrtc, - struct armada_plane *dplane) -{ - struct armada_plane_work *work = xchg(&dplane->work, NULL); - - if (work) - armada_drm_plane_work_call(dcrtc, work, work->cancel); -} - -static void armada_drm_crtc_complete_frame_work(struct armada_crtc *dcrtc, - struct armada_plane_work *work) -{ - unsigned long flags; - - spin_lock_irqsave(&dcrtc->irq_lock, flags); - armada_drm_crtc_update_regs(dcrtc, work->regs); - spin_unlock_irqrestore(&dcrtc->irq_lock, flags); -} - -static void armada_drm_crtc_complete_disable_work(struct armada_crtc *dcrtc, - struct armada_plane_work *work) -{ - unsigned long flags; - - if (dcrtc->plane == work->plane) - dcrtc->plane = NULL; - - spin_lock_irqsave(&dcrtc->irq_lock, flags); - armada_drm_crtc_update_regs(dcrtc, work->regs); - spin_unlock_irqrestore(&dcrtc->irq_lock, flags); -} - -static struct armada_plane_work * -armada_drm_crtc_alloc_plane_work(struct drm_plane *plane) -{ - struct armada_plane_work *work; - int i = 0; - - work = kzalloc(sizeof(*work), GFP_KERNEL); - if (!work) - return NULL; - - work->plane = plane; - work->fn = armada_drm_crtc_complete_frame_work; - work->need_kfree = true; - armada_reg_queue_end(work->regs, i); - - return work; -} - -static void armada_drm_crtc_finish_fb(struct armada_crtc *dcrtc, - struct drm_framebuffer *fb, bool force) -{ - struct armada_plane_work *work; - - if (!fb) - return; - - if (force) { - /* Display is disabled, so just drop the old fb */ - drm_framebuffer_put(fb); - return; - } - - work = armada_drm_crtc_alloc_plane_work(dcrtc->crtc.primary); - if (work) { - work->old_fb = fb; - - if (armada_drm_plane_work_queue(dcrtc, work) == 0) - return; - - kfree(work); - } - - /* - * Oops - just drop the reference immediately and hope for - * the best. The worst that will happen is the buffer gets - * reused before it has finished being displayed. - */ - drm_framebuffer_put(fb); -} - -static void armada_drm_vblank_off(struct armada_crtc *dcrtc) -{ - /* - * Tell the DRM core that vblank IRQs aren't going to happen for - * a while. This cleans up any pending vblank events for us. - */ - drm_crtc_vblank_off(&dcrtc->crtc); - armada_drm_plane_work_run(dcrtc, dcrtc->crtc.primary); -} - -/* The mode_config.mutex will be held for this call */ -static void armada_drm_crtc_dpms(struct drm_crtc *crtc, int dpms) -{ - struct armada_crtc *dcrtc = drm_to_armada_crtc(crtc); - - if (dpms_blanked(dcrtc->dpms) != dpms_blanked(dpms)) { - if (dpms_blanked(dpms)) - armada_drm_vblank_off(dcrtc); - else if (!IS_ERR(dcrtc->clk)) - WARN_ON(clk_prepare_enable(dcrtc->clk)); - dcrtc->dpms = dpms; - armada_drm_crtc_update(dcrtc); - if (!dpms_blanked(dpms)) - drm_crtc_vblank_on(&dcrtc->crtc); - else if (!IS_ERR(dcrtc->clk)) - clk_disable_unprepare(dcrtc->clk); - } else if (dcrtc->dpms != dpms) { - dcrtc->dpms = dpms; - } -} - -/* - * Prepare for a mode set. Turn off overlay to ensure that we don't end - * up with the overlay size being bigger than the active screen size. - * We rely upon X refreshing this state after the mode set has completed. - * - * The mode_config.mutex will be held for this call - */ -static void armada_drm_crtc_prepare(struct drm_crtc *crtc) -{ - struct armada_crtc *dcrtc = drm_to_armada_crtc(crtc); - struct drm_plane *plane; - - /* - * If we have an overlay plane associated with this CRTC, disable - * it before the modeset to avoid its coordinates being outside - * the new mode parameters. - */ - plane = dcrtc->plane; - if (plane) { - drm_plane_force_disable(plane); - WARN_ON(!armada_drm_plane_work_wait(drm_to_armada_plane(plane), - HZ)); - } -} - -/* The mode_config.mutex will be held for this call */ -static void armada_drm_crtc_commit(struct drm_crtc *crtc) +static void armada_drm_crtc_queue_state_event(struct drm_crtc *crtc) { struct armada_crtc *dcrtc = drm_to_armada_crtc(crtc); + struct drm_pending_vblank_event *event; - if (dcrtc->dpms != DRM_MODE_DPMS_ON) { - dcrtc->dpms = DRM_MODE_DPMS_ON; - armada_drm_crtc_update(dcrtc); + /* If we have an event, we need vblank events enabled */ + event = xchg(&crtc->state->event, NULL); + if (event) { + WARN_ON(drm_crtc_vblank_get(crtc) != 0); + dcrtc->event = event; } } @@ -465,8 +171,8 @@ static void armada_drm_crtc_enable_irq(struct armada_crtc *dcrtc, u32 mask) static void armada_drm_crtc_irq(struct armada_crtc *dcrtc, u32 stat) { + struct drm_pending_vblank_event *event; void __iomem *base = dcrtc->base; - struct drm_plane *ovl_plane; if (stat & DMA_FF_UNDERFLOW) DRM_ERROR("video underflow on crtc %u\n", dcrtc->num); @@ -476,10 +182,6 @@ static void armada_drm_crtc_irq(struct armada_crtc *dcrtc, u32 stat) if (stat & VSYNC_IRQ) drm_crtc_handle_vblank(&dcrtc->crtc); - ovl_plane = dcrtc->plane; - if (ovl_plane) - armada_drm_plane_work_run(dcrtc, ovl_plane); - spin_lock(&dcrtc->irq_lock); if (stat & GRA_FRAME_IRQ && dcrtc->interlaced) { int i = stat & GRA_FRAME_IRQ0 ? 0 : 1; @@ -495,22 +197,35 @@ static void armada_drm_crtc_irq(struct armada_crtc *dcrtc, u32 stat) writel_relaxed(val, base + LCD_SPU_ADV_REG); } - if (stat & DUMB_FRAMEDONE && dcrtc->cursor_update) { - writel_relaxed(dcrtc->cursor_hw_pos, - base + LCD_SPU_HWC_OVSA_HPXL_VLN); - writel_relaxed(dcrtc->cursor_hw_sz, - base + LCD_SPU_HWC_HPXL_VLN); - armada_updatel(CFG_HWC_ENA, - CFG_HWC_ENA | CFG_HWC_1BITMOD | CFG_HWC_1BITENA, - base + LCD_SPU_DMA_CTRL0); - dcrtc->cursor_update = false; + if (stat & dcrtc->irq_ena & DUMB_FRAMEDONE) { + if (dcrtc->update_pending) { + armada_drm_crtc_update_regs(dcrtc, dcrtc->regs); + dcrtc->update_pending = false; + } + if (dcrtc->cursor_update) { + writel_relaxed(dcrtc->cursor_hw_pos, + base + LCD_SPU_HWC_OVSA_HPXL_VLN); + writel_relaxed(dcrtc->cursor_hw_sz, + base + LCD_SPU_HWC_HPXL_VLN); + armada_updatel(CFG_HWC_ENA, + CFG_HWC_ENA | CFG_HWC_1BITMOD | + CFG_HWC_1BITENA, + base + LCD_SPU_DMA_CTRL0); + dcrtc->cursor_update = false; + } armada_drm_crtc_disable_irq(dcrtc, DUMB_FRAMEDONE_ENA); } - spin_unlock(&dcrtc->irq_lock); - if (stat & GRA_FRAME_IRQ) - armada_drm_plane_work_run(dcrtc, dcrtc->crtc.primary); + if (stat & VSYNC_IRQ && !dcrtc->update_pending) { + event = xchg(&dcrtc->event, NULL); + if (event) { + spin_lock(&dcrtc->crtc.dev->event_lock); + drm_crtc_send_vblank_event(&dcrtc->crtc, event); + spin_unlock(&dcrtc->crtc.dev->event_lock); + drm_crtc_vblank_put(&dcrtc->crtc); + } + } } static irqreturn_t armada_drm_irq(int irq, void *arg) @@ -519,8 +234,9 @@ static irqreturn_t armada_drm_irq(int irq, void *arg) u32 v, stat = readl_relaxed(dcrtc->base + LCD_SPU_IRQ_ISR); /* - * This is rediculous - rather than writing bits to clear, we - * have to set the actual status register value. This is racy. + * Reading the ISR appears to clear bits provided CLEAN_SPU_IRQ_ISR + * is set. Writing has some other effect to acknowledge the IRQ - + * without this, we only get a single IRQ. */ writel_relaxed(0, dcrtc->base + LCD_SPU_IRQ_ISR); @@ -536,107 +252,16 @@ static irqreturn_t armada_drm_irq(int irq, void *arg) return IRQ_NONE; } -static uint32_t armada_drm_crtc_calculate_csc(struct armada_crtc *dcrtc) -{ - struct drm_display_mode *adj = &dcrtc->crtc.mode; - uint32_t val = 0; - - if (dcrtc->csc_yuv_mode == CSC_YUV_CCIR709) - val |= CFG_CSC_YUV_CCIR709; - if (dcrtc->csc_rgb_mode == CSC_RGB_STUDIO) - val |= CFG_CSC_RGB_STUDIO; - - /* - * In auto mode, set the colorimetry, based upon the HDMI spec. - * 1280x720p, 1920x1080p and 1920x1080i use ITU709, others use - * ITU601. It may be more appropriate to set this depending on - * the source - but what if the graphic frame is YUV and the - * video frame is RGB? - */ - if ((adj->hdisplay == 1280 && adj->vdisplay == 720 && - !(adj->flags & DRM_MODE_FLAG_INTERLACE)) || - (adj->hdisplay == 1920 && adj->vdisplay == 1080)) { - if (dcrtc->csc_yuv_mode == CSC_AUTO) - val |= CFG_CSC_YUV_CCIR709; - } - - /* - * We assume we're connected to a TV-like device, so the YUV->RGB - * conversion should produce a limited range. We should set this - * depending on the connectors attached to this CRTC, and what - * kind of device they report being connected. - */ - if (dcrtc->csc_rgb_mode == CSC_AUTO) - val |= CFG_CSC_RGB_STUDIO; - - return val; -} - -static void armada_drm_gra_plane_regs(struct armada_regs *regs, - struct drm_framebuffer *fb, struct armada_plane_state *state, - int x, int y, bool interlaced) -{ - unsigned int i; - u32 ctrl0; - - i = armada_drm_crtc_calc_fb(fb, x, y, regs, interlaced); - armada_reg_queue_set(regs, i, state->dst_yx, LCD_SPU_GRA_OVSA_HPXL_VLN); - armada_reg_queue_set(regs, i, state->src_hw, LCD_SPU_GRA_HPXL_VLN); - armada_reg_queue_set(regs, i, state->dst_hw, LCD_SPU_GZM_HPXL_VLN); - - ctrl0 = state->ctrl0; - if (interlaced) - ctrl0 |= CFG_GRA_FTOGGLE; - - armada_reg_queue_mod(regs, i, ctrl0, CFG_GRAFORMAT | - CFG_GRA_MOD(CFG_SWAPRB | CFG_SWAPUV | - CFG_SWAPYU | CFG_YUV2RGB) | - CFG_PALETTE_ENA | CFG_GRA_FTOGGLE | - CFG_GRA_HSMOOTH | CFG_GRA_ENA, - LCD_SPU_DMA_CTRL0); - armada_reg_queue_end(regs, i); -} - -static void armada_drm_primary_set(struct drm_crtc *crtc, - struct drm_plane *plane, int x, int y) -{ - struct armada_plane_state *state = &drm_to_armada_plane(plane)->state; - struct armada_crtc *dcrtc = drm_to_armada_crtc(crtc); - struct armada_regs regs[8]; - bool interlaced = dcrtc->interlaced; - - armada_drm_gra_plane_regs(regs, plane->fb, state, x, y, interlaced); - armada_drm_crtc_update_regs(dcrtc, regs); -} - /* The mode_config.mutex will be held for this call */ -static int armada_drm_crtc_mode_set(struct drm_crtc *crtc, - struct drm_display_mode *mode, struct drm_display_mode *adj, - int x, int y, struct drm_framebuffer *old_fb) +static void armada_drm_crtc_mode_set_nofb(struct drm_crtc *crtc) { + struct drm_display_mode *adj = &crtc->state->adjusted_mode; struct armada_crtc *dcrtc = drm_to_armada_crtc(crtc); struct armada_regs regs[17]; uint32_t lm, rm, tm, bm, val, sclk; unsigned long flags; unsigned i; - bool interlaced; - - drm_framebuffer_get(crtc->primary->fb); - - interlaced = !!(adj->flags & DRM_MODE_FLAG_INTERLACE); - - val = CFG_GRA_ENA; - val |= CFG_GRA_FMT(drm_fb_to_armada_fb(dcrtc->crtc.primary->fb)->fmt); - val |= CFG_GRA_MOD(drm_fb_to_armada_fb(dcrtc->crtc.primary->fb)->mod); - - if (drm_fb_to_armada_fb(dcrtc->crtc.primary->fb)->fmt > CFG_420) - val |= CFG_PALETTE_ENA; - - drm_to_armada_plane(crtc->primary)->state.ctrl0 = val; - drm_to_armada_plane(crtc->primary)->state.src_hw = - drm_to_armada_plane(crtc->primary)->state.dst_hw = - adj->crtc_vdisplay << 16 | adj->crtc_hdisplay; - drm_to_armada_plane(crtc->primary)->state.dst_yx = 0; + bool interlaced = !!(adj->flags & DRM_MODE_FLAG_INTERLACE); i = 0; rm = adj->crtc_hsync_start - adj->crtc_hdisplay; @@ -644,35 +269,15 @@ static int armada_drm_crtc_mode_set(struct drm_crtc *crtc, bm = adj->crtc_vsync_start - adj->crtc_vdisplay; tm = adj->crtc_vtotal - adj->crtc_vsync_end; - DRM_DEBUG_DRIVER("H: %d %d %d %d lm %d rm %d\n", - adj->crtc_hdisplay, - adj->crtc_hsync_start, - adj->crtc_hsync_end, - adj->crtc_htotal, lm, rm); - DRM_DEBUG_DRIVER("V: %d %d %d %d tm %d bm %d\n", - adj->crtc_vdisplay, - adj->crtc_vsync_start, - adj->crtc_vsync_end, - adj->crtc_vtotal, tm, bm); - - /* Wait for pending flips to complete */ - armada_drm_plane_work_wait(drm_to_armada_plane(dcrtc->crtc.primary), - MAX_SCHEDULE_TIMEOUT); - - drm_crtc_vblank_off(crtc); - - val = dcrtc->dumb_ctrl & ~CFG_DUMB_ENA; - if (val != dcrtc->dumb_ctrl) { - dcrtc->dumb_ctrl = val; - writel_relaxed(val, dcrtc->base + LCD_SPU_DUMB_CTRL); - } - - /* - * If we are blanked, we would have disabled the clock. Re-enable - * it so that compute_clock() does the right thing. - */ - if (!IS_ERR(dcrtc->clk) && dpms_blanked(dcrtc->dpms)) - WARN_ON(clk_prepare_enable(dcrtc->clk)); + DRM_DEBUG_KMS("[CRTC:%d:%s] mode " DRM_MODE_FMT "\n", + crtc->base.id, crtc->name, + adj->base.id, adj->name, adj->vrefresh, adj->clock, + adj->crtc_hdisplay, adj->crtc_hsync_start, + adj->crtc_hsync_end, adj->crtc_htotal, + adj->crtc_vdisplay, adj->crtc_vsync_start, + adj->crtc_vsync_end, adj->crtc_vtotal, + adj->type, adj->flags); + DRM_DEBUG_KMS("lm %d rm %d tm %d bm %d\n", lm, rm, tm, bm); /* Now compute the divider for real */ dcrtc->variant->compute_clock(dcrtc, adj, &sclk); @@ -689,25 +294,20 @@ static int armada_drm_crtc_mode_set(struct drm_crtc *crtc, spin_lock_irqsave(&dcrtc->irq_lock, flags); - /* Ensure graphic fifo is enabled */ - armada_reg_queue_mod(regs, i, 0, CFG_PDWN64x66, LCD_SPU_SRAM_PARA1); - /* Even interlaced/progressive frame */ dcrtc->v[1].spu_v_h_total = adj->crtc_vtotal << 16 | adj->crtc_htotal; dcrtc->v[1].spu_v_porch = tm << 16 | bm; val = adj->crtc_hsync_start; - dcrtc->v[1].spu_adv_reg = val << 20 | val | ADV_VSYNCOFFEN | - dcrtc->variant->spu_adv_reg; + dcrtc->v[1].spu_adv_reg = val << 20 | val | ADV_VSYNCOFFEN; if (interlaced) { /* Odd interlaced frame */ + val -= adj->crtc_htotal / 2; + dcrtc->v[0].spu_adv_reg = val << 20 | val | ADV_VSYNCOFFEN; dcrtc->v[0].spu_v_h_total = dcrtc->v[1].spu_v_h_total + (1 << 16); dcrtc->v[0].spu_v_porch = dcrtc->v[1].spu_v_porch + 1; - val = adj->crtc_hsync_start - adj->crtc_htotal / 2; - dcrtc->v[0].spu_adv_reg = val << 20 | val | ADV_VSYNCOFFEN | - dcrtc->variant->spu_adv_reg; } else { dcrtc->v[0] = dcrtc->v[1]; } @@ -720,77 +320,136 @@ static int armada_drm_crtc_mode_set(struct drm_crtc *crtc, armada_reg_queue_set(regs, i, dcrtc->v[0].spu_v_h_total, LCD_SPUT_V_H_TOTAL); - if (dcrtc->variant->has_spu_adv_reg) { + if (dcrtc->variant->has_spu_adv_reg) armada_reg_queue_mod(regs, i, dcrtc->v[0].spu_adv_reg, ADV_VSYNC_L_OFF | ADV_VSYNC_H_OFF | ADV_VSYNCOFFEN, LCD_SPU_ADV_REG); - } val = adj->flags & DRM_MODE_FLAG_NVSYNC ? CFG_VSYNC_INV : 0; armada_reg_queue_mod(regs, i, val, CFG_VSYNC_INV, LCD_SPU_DMA_CTRL1); - val = dcrtc->spu_iopad_ctrl | armada_drm_crtc_calculate_csc(dcrtc); - armada_reg_queue_set(regs, i, val, LCD_SPU_IOPAD_CONTROL); + /* + * The documentation doesn't indicate what the normal state of + * the sync signals are. Sebastian Hesselbart kindly probed + * these signals on his board to determine their state. + * + * The non-inverted state of the sync signals is active high. + * Setting these bits makes the appropriate signal active low. + */ + val = 0; + if (adj->flags & DRM_MODE_FLAG_NCSYNC) + val |= CFG_INV_CSYNC; + if (adj->flags & DRM_MODE_FLAG_NHSYNC) + val |= CFG_INV_HSYNC; + if (adj->flags & DRM_MODE_FLAG_NVSYNC) + val |= CFG_INV_VSYNC; + armada_reg_queue_mod(regs, i, val, CFG_INV_CSYNC | CFG_INV_HSYNC | + CFG_INV_VSYNC, LCD_SPU_DUMB_CTRL); armada_reg_queue_end(regs, i); armada_drm_crtc_update_regs(dcrtc, regs); - - armada_drm_primary_set(crtc, crtc->primary, x, y); spin_unlock_irqrestore(&dcrtc->irq_lock, flags); +} - armada_drm_crtc_update(dcrtc); +static void armada_drm_crtc_atomic_begin(struct drm_crtc *crtc, + struct drm_crtc_state *old_crtc_state) +{ + struct armada_crtc *dcrtc = drm_to_armada_crtc(crtc); - drm_crtc_vblank_on(crtc); - armada_drm_crtc_finish_fb(dcrtc, old_fb, dpms_blanked(dcrtc->dpms)); + DRM_DEBUG_KMS("[CRTC:%d:%s]\n", crtc->base.id, crtc->name); - return 0; + dcrtc->regs_idx = 0; + dcrtc->regs = dcrtc->atomic_regs; } -/* The mode_config.mutex will be held for this call */ -static int armada_drm_crtc_mode_set_base(struct drm_crtc *crtc, int x, int y, - struct drm_framebuffer *old_fb) +static void armada_drm_crtc_atomic_flush(struct drm_crtc *crtc, + struct drm_crtc_state *old_crtc_state) { struct armada_crtc *dcrtc = drm_to_armada_crtc(crtc); - struct armada_regs regs[4]; - unsigned i; - i = armada_drm_crtc_calc_fb(crtc->primary->fb, crtc->x, crtc->y, regs, - dcrtc->interlaced); - armada_reg_queue_end(regs, i); + DRM_DEBUG_KMS("[CRTC:%d:%s]\n", crtc->base.id, crtc->name); + + armada_reg_queue_end(dcrtc->regs, dcrtc->regs_idx); + + /* + * If we aren't doing a full modeset, then we need to queue + * the event here. + */ + if (!drm_atomic_crtc_needs_modeset(crtc->state)) { + dcrtc->update_pending = true; + armada_drm_crtc_queue_state_event(crtc); + spin_lock_irq(&dcrtc->irq_lock); + armada_drm_crtc_enable_irq(dcrtc, DUMB_FRAMEDONE_ENA); + spin_unlock_irq(&dcrtc->irq_lock); + } else { + spin_lock_irq(&dcrtc->irq_lock); + armada_drm_crtc_update_regs(dcrtc, dcrtc->regs); + spin_unlock_irq(&dcrtc->irq_lock); + } +} - /* Wait for pending flips to complete */ - armada_drm_plane_work_wait(drm_to_armada_plane(dcrtc->crtc.primary), - MAX_SCHEDULE_TIMEOUT); +static void armada_drm_crtc_atomic_disable(struct drm_crtc *crtc, + struct drm_crtc_state *old_state) +{ + struct armada_crtc *dcrtc = drm_to_armada_crtc(crtc); + struct drm_pending_vblank_event *event; - /* Take a reference to the new fb as we're using it */ - drm_framebuffer_get(crtc->primary->fb); + DRM_DEBUG_KMS("[CRTC:%d:%s]\n", crtc->base.id, crtc->name); - /* Update the base in the CRTC */ - armada_drm_crtc_update_regs(dcrtc, regs); + drm_crtc_vblank_off(crtc); + armada_drm_crtc_update(dcrtc, false); - /* Drop our previously held reference */ - armada_drm_crtc_finish_fb(dcrtc, old_fb, dpms_blanked(dcrtc->dpms)); + if (!crtc->state->active) { + /* + * This modeset will be leaving the CRTC disabled, so + * call the backend to disable upstream clocks etc. + */ + if (dcrtc->variant->disable) + dcrtc->variant->disable(dcrtc); - return 0; + /* + * We will not receive any further vblank events. + * Send the flip_done event manually. + */ + event = crtc->state->event; + crtc->state->event = NULL; + if (event) { + spin_lock_irq(&crtc->dev->event_lock); + drm_crtc_send_vblank_event(crtc, event); + spin_unlock_irq(&crtc->dev->event_lock); + } + } } -/* The mode_config.mutex will be held for this call */ -static void armada_drm_crtc_disable(struct drm_crtc *crtc) +static void armada_drm_crtc_atomic_enable(struct drm_crtc *crtc, + struct drm_crtc_state *old_state) { - armada_drm_crtc_dpms(crtc, DRM_MODE_DPMS_OFF); + struct armada_crtc *dcrtc = drm_to_armada_crtc(crtc); + + DRM_DEBUG_KMS("[CRTC:%d:%s]\n", crtc->base.id, crtc->name); + + if (!old_state->active) { + /* + * This modeset is enabling the CRTC after it having + * been disabled. Reverse the call to ->disable in + * the atomic_disable(). + */ + if (dcrtc->variant->enable) + dcrtc->variant->enable(dcrtc, &crtc->state->adjusted_mode); + } + armada_drm_crtc_update(dcrtc, true); + drm_crtc_vblank_on(crtc); - /* Disable our primary plane when we disable the CRTC. */ - crtc->primary->funcs->disable_plane(crtc->primary, NULL); + armada_drm_crtc_queue_state_event(crtc); } static const struct drm_crtc_helper_funcs armada_crtc_helper_funcs = { - .dpms = armada_drm_crtc_dpms, - .prepare = armada_drm_crtc_prepare, - .commit = armada_drm_crtc_commit, .mode_fixup = armada_drm_crtc_mode_fixup, - .mode_set = armada_drm_crtc_mode_set, - .mode_set_base = armada_drm_crtc_mode_set_base, - .disable = armada_drm_crtc_disable, + .mode_set_nofb = armada_drm_crtc_mode_set_nofb, + .atomic_begin = armada_drm_crtc_atomic_begin, + .atomic_flush = armada_drm_crtc_atomic_flush, + .atomic_disable = armada_drm_crtc_atomic_disable, + .atomic_enable = armada_drm_crtc_atomic_enable, }; static void armada_load_cursor_argb(void __iomem *base, uint32_t *pix, @@ -883,7 +542,6 @@ static int armada_drm_crtc_cursor_update(struct armada_crtc *dcrtc, bool reload) if (!dcrtc->cursor_obj || !h || !w) { spin_lock_irq(&dcrtc->irq_lock); - armada_drm_crtc_disable_irq(dcrtc, DUMB_FRAMEDONE_ENA); dcrtc->cursor_update = false; armada_updatel(0, CFG_HWC_ENA, dcrtc->base + LCD_SPU_DMA_CTRL0); spin_unlock_irq(&dcrtc->irq_lock); @@ -907,7 +565,6 @@ static int armada_drm_crtc_cursor_update(struct armada_crtc *dcrtc, bool reload) if (dcrtc->cursor_hw_sz != (h << 16 | w)) { spin_lock_irq(&dcrtc->irq_lock); - armada_drm_crtc_disable_irq(dcrtc, DUMB_FRAMEDONE_ENA); dcrtc->cursor_update = false; armada_updatel(0, CFG_HWC_ENA, dcrtc->base + LCD_SPU_DMA_CTRL0); spin_unlock_irq(&dcrtc->irq_lock); @@ -1015,8 +672,8 @@ static void armada_drm_crtc_destroy(struct drm_crtc *crtc) priv->dcrtc[dcrtc->num] = NULL; drm_crtc_cleanup(&dcrtc->crtc); - if (!IS_ERR(dcrtc->clk)) - clk_disable_unprepare(dcrtc->clk); + if (dcrtc->variant->disable) + dcrtc->variant->disable(dcrtc); writel_relaxed(0, dcrtc->base + LCD_SPU_IRQ_ENA); @@ -1025,361 +682,51 @@ static void armada_drm_crtc_destroy(struct drm_crtc *crtc) kfree(dcrtc); } -/* - * The mode_config lock is held here, to prevent races between this - * and a mode_set. - */ -static int armada_drm_crtc_page_flip(struct drm_crtc *crtc, - struct drm_framebuffer *fb, struct drm_pending_vblank_event *event, uint32_t page_flip_flags, - struct drm_modeset_acquire_ctx *ctx) -{ - struct armada_crtc *dcrtc = drm_to_armada_crtc(crtc); - struct armada_plane_work *work; - unsigned i; - int ret; - - /* We don't support changing the pixel format */ - if (fb->format != crtc->primary->fb->format) - return -EINVAL; - - work = armada_drm_crtc_alloc_plane_work(dcrtc->crtc.primary); - if (!work) - return -ENOMEM; - - work->event = event; - work->old_fb = dcrtc->crtc.primary->fb; - - i = armada_drm_crtc_calc_fb(fb, crtc->x, crtc->y, work->regs, - dcrtc->interlaced); - armada_reg_queue_end(work->regs, i); - - /* - * Ensure that we hold a reference on the new framebuffer. - * This has to match the behaviour in mode_set. - */ - drm_framebuffer_get(fb); - - ret = armada_drm_plane_work_queue(dcrtc, work); - if (ret) { - /* Undo our reference above */ - drm_framebuffer_put(fb); - kfree(work); - return ret; - } - - /* - * Don't take a reference on the new framebuffer; - * drm_mode_page_flip_ioctl() has already grabbed a reference and - * will _not_ drop that reference on successful return from this - * function. Simply mark this new framebuffer as the current one. - */ - dcrtc->crtc.primary->fb = fb; - - /* - * Finally, if the display is blanked, we won't receive an - * interrupt, so complete it now. - */ - if (dpms_blanked(dcrtc->dpms)) - armada_drm_plane_work_run(dcrtc, dcrtc->crtc.primary); - - return 0; -} - -static int -armada_drm_crtc_set_property(struct drm_crtc *crtc, - struct drm_property *property, uint64_t val) -{ - struct armada_private *priv = crtc->dev->dev_private; - struct armada_crtc *dcrtc = drm_to_armada_crtc(crtc); - bool update_csc = false; - - if (property == priv->csc_yuv_prop) { - dcrtc->csc_yuv_mode = val; - update_csc = true; - } else if (property == priv->csc_rgb_prop) { - dcrtc->csc_rgb_mode = val; - update_csc = true; - } - - if (update_csc) { - uint32_t val; - - val = dcrtc->spu_iopad_ctrl | - armada_drm_crtc_calculate_csc(dcrtc); - writel_relaxed(val, dcrtc->base + LCD_SPU_IOPAD_CONTROL); - } - - return 0; -} - /* These are called under the vbl_lock. */ static int armada_drm_crtc_enable_vblank(struct drm_crtc *crtc) { struct armada_crtc *dcrtc = drm_to_armada_crtc(crtc); + unsigned long flags; + spin_lock_irqsave(&dcrtc->irq_lock, flags); armada_drm_crtc_enable_irq(dcrtc, VSYNC_IRQ_ENA); + spin_unlock_irqrestore(&dcrtc->irq_lock, flags); return 0; } static void armada_drm_crtc_disable_vblank(struct drm_crtc *crtc) { struct armada_crtc *dcrtc = drm_to_armada_crtc(crtc); + unsigned long flags; + spin_lock_irqsave(&dcrtc->irq_lock, flags); armada_drm_crtc_disable_irq(dcrtc, VSYNC_IRQ_ENA); + spin_unlock_irqrestore(&dcrtc->irq_lock, flags); } static const struct drm_crtc_funcs armada_crtc_funcs = { + .reset = drm_atomic_helper_crtc_reset, .cursor_set = armada_drm_crtc_cursor_set, .cursor_move = armada_drm_crtc_cursor_move, .destroy = armada_drm_crtc_destroy, - .set_config = drm_crtc_helper_set_config, - .page_flip = armada_drm_crtc_page_flip, - .set_property = armada_drm_crtc_set_property, + .set_config = drm_atomic_helper_set_config, + .page_flip = drm_atomic_helper_page_flip, + .atomic_duplicate_state = drm_atomic_helper_crtc_duplicate_state, + .atomic_destroy_state = drm_atomic_helper_crtc_destroy_state, .enable_vblank = armada_drm_crtc_enable_vblank, .disable_vblank = armada_drm_crtc_disable_vblank, }; -static void armada_drm_primary_update_state(struct drm_plane_state *state, - struct armada_regs *regs) -{ - struct armada_plane *dplane = drm_to_armada_plane(state->plane); - struct armada_crtc *dcrtc = drm_to_armada_crtc(state->crtc); - struct armada_framebuffer *dfb = drm_fb_to_armada_fb(state->fb); - bool was_disabled; - unsigned int idx = 0; - u32 val; - - val = CFG_GRA_FMT(dfb->fmt) | CFG_GRA_MOD(dfb->mod); - if (dfb->fmt > CFG_420) - val |= CFG_PALETTE_ENA; - if (state->visible) - val |= CFG_GRA_ENA; - if (drm_rect_width(&state->src) >> 16 != drm_rect_width(&state->dst)) - val |= CFG_GRA_HSMOOTH; - - was_disabled = !(dplane->state.ctrl0 & CFG_GRA_ENA); - if (was_disabled) - armada_reg_queue_mod(regs, idx, - 0, CFG_PDWN64x66, LCD_SPU_SRAM_PARA1); - - dplane->state.ctrl0 = val; - dplane->state.src_hw = (drm_rect_height(&state->src) & 0xffff0000) | - drm_rect_width(&state->src) >> 16; - dplane->state.dst_hw = drm_rect_height(&state->dst) << 16 | - drm_rect_width(&state->dst); - dplane->state.dst_yx = state->dst.y1 << 16 | state->dst.x1; - - armada_drm_gra_plane_regs(regs + idx, &dfb->fb, &dplane->state, - state->src.x1 >> 16, state->src.y1 >> 16, - dcrtc->interlaced); - - dplane->state.vsync_update = !was_disabled; - dplane->state.changed = true; -} - -static int armada_drm_primary_update(struct drm_plane *plane, - struct drm_crtc *crtc, struct drm_framebuffer *fb, - int crtc_x, int crtc_y, unsigned int crtc_w, unsigned int crtc_h, - uint32_t src_x, uint32_t src_y, uint32_t src_w, uint32_t src_h, - struct drm_modeset_acquire_ctx *ctx) -{ - struct armada_plane *dplane = drm_to_armada_plane(plane); - struct armada_crtc *dcrtc = drm_to_armada_crtc(crtc); - struct armada_plane_work *work; - struct drm_plane_state state = { - .plane = plane, - .crtc = crtc, - .fb = fb, - .src_x = src_x, - .src_y = src_y, - .src_w = src_w, - .src_h = src_h, - .crtc_x = crtc_x, - .crtc_y = crtc_y, - .crtc_w = crtc_w, - .crtc_h = crtc_h, - .rotation = DRM_MODE_ROTATE_0, - }; - struct drm_crtc_state crtc_state = { - .crtc = crtc, - .enable = crtc->enabled, - .mode = crtc->mode, - }; - int ret; - - ret = drm_atomic_helper_check_plane_state(&state, &crtc_state, 0, - INT_MAX, true, false); - if (ret) - return ret; - - work = &dplane->works[dplane->next_work]; - work->fn = armada_drm_crtc_complete_frame_work; - - if (plane->fb != fb) { - /* - * Take a reference on the new framebuffer - we want to - * hold on to it while the hardware is displaying it. - */ - drm_framebuffer_reference(fb); - - work->old_fb = plane->fb; - } else { - work->old_fb = NULL; - } - - armada_drm_primary_update_state(&state, work->regs); - - if (!dplane->state.changed) - return 0; - - /* Wait for pending work to complete */ - if (armada_drm_plane_work_wait(dplane, HZ / 10) == 0) - armada_drm_plane_work_cancel(dcrtc, dplane); - - if (!dplane->state.vsync_update) { - work->fn(dcrtc, work); - if (work->old_fb) - drm_framebuffer_unreference(work->old_fb); - return 0; - } - - /* Queue it for update on the next interrupt if we are enabled */ - ret = armada_drm_plane_work_queue(dcrtc, work); - if (ret) { - work->fn(dcrtc, work); - if (work->old_fb) - drm_framebuffer_unreference(work->old_fb); - } - - dplane->next_work = !dplane->next_work; - - return 0; -} - -int armada_drm_plane_disable(struct drm_plane *plane, - struct drm_modeset_acquire_ctx *ctx) -{ - struct armada_plane *dplane = drm_to_armada_plane(plane); - struct armada_crtc *dcrtc; - struct armada_plane_work *work; - unsigned int idx = 0; - u32 sram_para1, enable_mask; - - if (!plane->crtc) - return 0; - - /* - * Arrange to power down most RAMs and FIFOs if this is the primary - * plane, otherwise just the YUV FIFOs for the overlay plane. - */ - if (plane->type == DRM_PLANE_TYPE_PRIMARY) { - sram_para1 = CFG_PDWN256x32 | CFG_PDWN256x24 | CFG_PDWN256x8 | - CFG_PDWN32x32 | CFG_PDWN64x66; - enable_mask = CFG_GRA_ENA; - } else { - sram_para1 = CFG_PDWN16x66 | CFG_PDWN32x66; - enable_mask = CFG_DMA_ENA; - } - - dplane->state.ctrl0 &= ~enable_mask; - - dcrtc = drm_to_armada_crtc(plane->crtc); - - /* - * Try to disable the plane and drop our ref on the framebuffer - * at the next frame update. If we fail for any reason, disable - * the plane immediately. - */ - work = &dplane->works[dplane->next_work]; - work->fn = armada_drm_crtc_complete_disable_work; - work->cancel = armada_drm_crtc_complete_disable_work; - work->old_fb = plane->fb; - - armada_reg_queue_mod(work->regs, idx, - 0, enable_mask, LCD_SPU_DMA_CTRL0); - armada_reg_queue_mod(work->regs, idx, - sram_para1, 0, LCD_SPU_SRAM_PARA1); - armada_reg_queue_end(work->regs, idx); - - /* Wait for any preceding work to complete, but don't wedge */ - if (WARN_ON(!armada_drm_plane_work_wait(dplane, HZ))) - armada_drm_plane_work_cancel(dcrtc, dplane); - - if (armada_drm_plane_work_queue(dcrtc, work)) { - work->fn(dcrtc, work); - if (work->old_fb) - drm_framebuffer_unreference(work->old_fb); - } - - dplane->next_work = !dplane->next_work; - - return 0; -} - -static const struct drm_plane_funcs armada_primary_plane_funcs = { - .update_plane = armada_drm_primary_update, - .disable_plane = armada_drm_plane_disable, - .destroy = drm_primary_helper_destroy, -}; - -int armada_drm_plane_init(struct armada_plane *plane) -{ - unsigned int i; - - for (i = 0; i < ARRAY_SIZE(plane->works); i++) - plane->works[i].plane = &plane->base; - - init_waitqueue_head(&plane->frame_wait); - - return 0; -} - -static const struct drm_prop_enum_list armada_drm_csc_yuv_enum_list[] = { - { CSC_AUTO, "Auto" }, - { CSC_YUV_CCIR601, "CCIR601" }, - { CSC_YUV_CCIR709, "CCIR709" }, -}; - -static const struct drm_prop_enum_list armada_drm_csc_rgb_enum_list[] = { - { CSC_AUTO, "Auto" }, - { CSC_RGB_COMPUTER, "Computer system" }, - { CSC_RGB_STUDIO, "Studio" }, -}; - -static int armada_drm_crtc_create_properties(struct drm_device *dev) -{ - struct armada_private *priv = dev->dev_private; - - if (priv->csc_yuv_prop) - return 0; - - priv->csc_yuv_prop = drm_property_create_enum(dev, 0, - "CSC_YUV", armada_drm_csc_yuv_enum_list, - ARRAY_SIZE(armada_drm_csc_yuv_enum_list)); - priv->csc_rgb_prop = drm_property_create_enum(dev, 0, - "CSC_RGB", armada_drm_csc_rgb_enum_list, - ARRAY_SIZE(armada_drm_csc_rgb_enum_list)); - - if (!priv->csc_yuv_prop || !priv->csc_rgb_prop) - return -ENOMEM; - - return 0; -} - static int armada_drm_crtc_create(struct drm_device *drm, struct device *dev, struct resource *res, int irq, const struct armada_variant *variant, struct device_node *port) { struct armada_private *priv = drm->dev_private; struct armada_crtc *dcrtc; - struct armada_plane *primary; + struct drm_plane *primary; void __iomem *base; int ret; - ret = armada_drm_crtc_create_properties(drm); - if (ret) - return ret; - base = devm_ioremap_resource(dev, res); if (IS_ERR(base)) return PTR_ERR(base); @@ -1397,8 +744,6 @@ static int armada_drm_crtc_create(struct drm_device *drm, struct device *dev, dcrtc->base = base; dcrtc->num = drm->mode_config.num_crtc; dcrtc->clk = ERR_PTR(-EINVAL); - dcrtc->csc_yuv_mode = CSC_AUTO; - dcrtc->csc_rgb_mode = CSC_AUTO; dcrtc->cfg_dumb_ctrl = DUMB24_RGB888_0; dcrtc->spu_iopad_ctrl = CFG_VSCALE_LN_EN | CFG_IOPAD_DUMB24; spin_lock_init(&dcrtc->irq_lock); @@ -1415,6 +760,7 @@ static int armada_drm_crtc_create(struct drm_device *drm, struct device *dev, CFG_PDWN64x66, dcrtc->base + LCD_SPU_SRAM_PARA1); writel_relaxed(0x2032ff81, dcrtc->base + LCD_SPU_DMA_CTRL1); writel_relaxed(dcrtc->irq_ena, dcrtc->base + LCD_SPU_IRQ_ENA); + readl_relaxed(dcrtc->base + LCD_SPU_IRQ_ISR); writel_relaxed(0, dcrtc->base + LCD_SPU_IRQ_ISR); ret = devm_request_irq(dev, irq, armada_drm_irq, 0, "armada_drm_crtc", @@ -1441,39 +787,23 @@ static int armada_drm_crtc_create(struct drm_device *drm, struct device *dev, goto err_crtc; } - ret = armada_drm_plane_init(primary); - if (ret) { - kfree(primary); - goto err_crtc; - } - - ret = drm_universal_plane_init(drm, &primary->base, 0, - &armada_primary_plane_funcs, - armada_primary_formats, - ARRAY_SIZE(armada_primary_formats), - NULL, - DRM_PLANE_TYPE_PRIMARY, NULL); + ret = armada_drm_primary_plane_init(drm, primary); if (ret) { kfree(primary); goto err_crtc; } - ret = drm_crtc_init_with_planes(drm, &dcrtc->crtc, &primary->base, NULL, + ret = drm_crtc_init_with_planes(drm, &dcrtc->crtc, primary, NULL, &armada_crtc_funcs, NULL); if (ret) goto err_crtc_init; drm_crtc_helper_add(&dcrtc->crtc, &armada_crtc_helper_funcs); - drm_object_attach_property(&dcrtc->crtc.base, priv->csc_yuv_prop, - dcrtc->csc_yuv_mode); - drm_object_attach_property(&dcrtc->crtc.base, priv->csc_rgb_prop, - dcrtc->csc_rgb_mode); - return armada_overlay_plane_create(drm, 1 << dcrtc->num); err_crtc_init: - primary->base.funcs->destroy(&primary->base); + primary->funcs->destroy(primary); err_crtc: kfree(dcrtc); diff --git a/drivers/gpu/drm/armada/armada_crtc.h b/drivers/gpu/drm/armada/armada_crtc.h index 445829b8877a..7ebd337b60af 100644 --- a/drivers/gpu/drm/armada/armada_crtc.h +++ b/drivers/gpu/drm/armada/armada_crtc.h @@ -32,49 +32,8 @@ struct armada_regs { armada_reg_queue_mod(_r, _i, 0, 0, ~0) struct armada_crtc; -struct armada_plane; struct armada_variant; -struct armada_plane_work { - void (*fn)(struct armada_crtc *, struct armada_plane_work *); - void (*cancel)(struct armada_crtc *, struct armada_plane_work *); - bool need_kfree; - struct drm_plane *plane; - struct drm_framebuffer *old_fb; - struct drm_pending_vblank_event *event; - struct armada_regs regs[14]; -}; - -struct armada_plane_state { - u16 src_x; - u16 src_y; - u32 src_hw; - u32 dst_hw; - u32 dst_yx; - u32 ctrl0; - bool changed; - bool vsync_update; -}; - -struct armada_plane { - struct drm_plane base; - wait_queue_head_t frame_wait; - bool next_work; - struct armada_plane_work works[2]; - struct armada_plane_work *work; - struct armada_plane_state state; -}; -#define drm_to_armada_plane(p) container_of(p, struct armada_plane, base) - -int armada_drm_plane_init(struct armada_plane *plane); -int armada_drm_plane_work_queue(struct armada_crtc *dcrtc, - struct armada_plane_work *work); -int armada_drm_plane_work_wait(struct armada_plane *plane, long timeout); -void armada_drm_plane_work_cancel(struct armada_crtc *dcrtc, - struct armada_plane *plane); -void armada_drm_plane_calc_addrs(u32 *addrs, struct drm_framebuffer *fb, - int x, int y); - struct armada_crtc { struct drm_crtc crtc; const struct armada_variant *variant; @@ -89,10 +48,6 @@ struct armada_crtc { } v[2]; bool interlaced; bool cursor_update; - uint8_t csc_yuv_mode; - uint8_t csc_rgb_mode; - - struct drm_plane *plane; struct armada_gem_object *cursor_obj; int cursor_x; @@ -102,21 +57,22 @@ struct armada_crtc { uint32_t cursor_w; uint32_t cursor_h; - int dpms; uint32_t cfg_dumb_ctrl; - uint32_t dumb_ctrl; uint32_t spu_iopad_ctrl; spinlock_t irq_lock; uint32_t irq_ena; + + bool update_pending; + struct drm_pending_vblank_event *event; + struct armada_regs atomic_regs[32]; + struct armada_regs *regs; + unsigned int regs_idx; }; #define drm_to_armada_crtc(c) container_of(c, struct armada_crtc, crtc) void armada_drm_crtc_update_regs(struct armada_crtc *, struct armada_regs *); -int armada_drm_plane_disable(struct drm_plane *plane, - struct drm_modeset_acquire_ctx *ctx); - extern struct platform_driver armada_lcd_platform_driver; #endif diff --git a/drivers/gpu/drm/armada/armada_drm.h b/drivers/gpu/drm/armada/armada_drm.h index cc4c557c9f66..f09083ff15d3 100644 --- a/drivers/gpu/drm/armada/armada_drm.h +++ b/drivers/gpu/drm/armada/armada_drm.h @@ -42,11 +42,12 @@ struct armada_private; struct armada_variant { bool has_spu_adv_reg; - uint32_t spu_adv_reg; int (*init)(struct armada_crtc *, struct device *); int (*compute_clock)(struct armada_crtc *, const struct drm_display_mode *, uint32_t *); + void (*disable)(struct armada_crtc *); + void (*enable)(struct armada_crtc *, const struct drm_display_mode *); }; /* Variant ops */ @@ -54,14 +55,10 @@ extern const struct armada_variant armada510_ops; struct armada_private { struct drm_device drm; - struct work_struct fb_unref_work; - DECLARE_KFIFO(fb_unref, struct drm_framebuffer *, 8); struct drm_fb_helper *fbdev; struct armada_crtc *dcrtc[2]; struct drm_mm linear; /* protected by linear_lock */ struct mutex linear_lock; - struct drm_property *csc_yuv_prop; - struct drm_property *csc_rgb_prop; struct drm_property *colorkey_prop; struct drm_property *colorkey_min_prop; struct drm_property *colorkey_max_prop; @@ -76,13 +73,6 @@ struct armada_private { #endif }; -void __armada_drm_queue_unref_work(struct drm_device *, - struct drm_framebuffer *); -void armada_drm_queue_unref_work(struct drm_device *, - struct drm_framebuffer *); - -extern const struct drm_mode_config_funcs armada_drm_mode_config_funcs; - int armada_fbdev_init(struct drm_device *); void armada_fbdev_fini(struct drm_device *); diff --git a/drivers/gpu/drm/armada/armada_drv.c b/drivers/gpu/drm/armada/armada_drv.c index 4b11b6b52f1d..fa31589b4fc0 100644 --- a/drivers/gpu/drm/armada/armada_drv.c +++ b/drivers/gpu/drm/armada/armada_drv.c @@ -9,46 +9,18 @@ #include <linux/component.h> #include <linux/module.h> #include <linux/of_graph.h> +#include <drm/drm_atomic_helper.h> #include <drm/drm_crtc_helper.h> #include <drm/drm_fb_helper.h> #include <drm/drm_of.h> #include "armada_crtc.h" #include "armada_drm.h" #include "armada_gem.h" +#include "armada_fb.h" #include "armada_hw.h" #include <drm/armada_drm.h> #include "armada_ioctlP.h" -static void armada_drm_unref_work(struct work_struct *work) -{ - struct armada_private *priv = - container_of(work, struct armada_private, fb_unref_work); - struct drm_framebuffer *fb; - - while (kfifo_get(&priv->fb_unref, &fb)) - drm_framebuffer_put(fb); -} - -/* Must be called with dev->event_lock held */ -void __armada_drm_queue_unref_work(struct drm_device *dev, - struct drm_framebuffer *fb) -{ - struct armada_private *priv = dev->dev_private; - - WARN_ON(!kfifo_put(&priv->fb_unref, fb)); - schedule_work(&priv->fb_unref_work); -} - -void armada_drm_queue_unref_work(struct drm_device *dev, - struct drm_framebuffer *fb) -{ - unsigned long flags; - - spin_lock_irqsave(&dev->event_lock, flags); - __armada_drm_queue_unref_work(dev, fb); - spin_unlock_irqrestore(&dev->event_lock, flags); -} - static struct drm_ioctl_desc armada_ioctls[] = { DRM_IOCTL_DEF_DRV(ARMADA_GEM_CREATE, armada_gem_create_ioctl,0), DRM_IOCTL_DEF_DRV(ARMADA_GEM_MMAP, armada_gem_mmap_ioctl, 0), @@ -72,11 +44,18 @@ static struct drm_driver armada_drm_driver = { .desc = "Armada SoC DRM", .date = "20120730", .driver_features = DRIVER_GEM | DRIVER_MODESET | - DRIVER_PRIME, + DRIVER_PRIME | DRIVER_ATOMIC, .ioctls = armada_ioctls, .fops = &armada_drm_fops, }; +static const struct drm_mode_config_funcs armada_drm_mode_config_funcs = { + .fb_create = armada_fb_create, + .output_poll_changed = drm_fb_helper_output_poll_changed, + .atomic_check = drm_atomic_helper_check, + .atomic_commit = drm_atomic_helper_commit, +}; + static int armada_drm_bind(struct device *dev) { struct armada_private *priv; @@ -109,7 +88,7 @@ static int armada_drm_bind(struct device *dev) /* * The drm_device structure must be at the start of - * armada_private for drm_dev_unref() to work correctly. + * armada_private for drm_dev_put() to work correctly. */ BUILD_BUG_ON(offsetof(struct armada_private, drm) != 0); @@ -125,9 +104,6 @@ static int armada_drm_bind(struct device *dev) dev_set_drvdata(dev, &priv->drm); - INIT_WORK(&priv->fb_unref_work, armada_drm_unref_work); - INIT_KFIFO(priv->fb_unref); - /* Mode setting support */ drm_mode_config_init(&priv->drm); priv->drm.mode_config.min_width = 320; @@ -155,6 +131,8 @@ static int armada_drm_bind(struct device *dev) priv->drm.irq_enabled = true; + drm_mode_config_reset(&priv->drm); + ret = armada_fbdev_init(&priv->drm); if (ret) goto err_comp; @@ -179,8 +157,7 @@ static int armada_drm_bind(struct device *dev) err_kms: drm_mode_config_cleanup(&priv->drm); drm_mm_takedown(&priv->linear); - flush_work(&priv->fb_unref_work); - drm_dev_unref(&priv->drm); + drm_dev_put(&priv->drm); return ret; } @@ -198,9 +175,8 @@ static void armada_drm_unbind(struct device *dev) drm_mode_config_cleanup(&priv->drm); drm_mm_takedown(&priv->linear); - flush_work(&priv->fb_unref_work); - drm_dev_unref(&priv->drm); + drm_dev_put(&priv->drm); } static int compare_of(struct device *dev, void *data) diff --git a/drivers/gpu/drm/armada/armada_fb.c b/drivers/gpu/drm/armada/armada_fb.c index edd15126bde9..6bd638a54579 100644 --- a/drivers/gpu/drm/armada/armada_fb.c +++ b/drivers/gpu/drm/armada/armada_fb.c @@ -84,7 +84,7 @@ struct armada_framebuffer *armada_framebuffer_create(struct drm_device *dev, return dfb; } -static struct drm_framebuffer *armada_fb_create(struct drm_device *dev, +struct drm_framebuffer *armada_fb_create(struct drm_device *dev, struct drm_file *dfile, const struct drm_mode_fb_cmd2 *mode) { struct armada_gem_object *obj; @@ -138,8 +138,3 @@ static struct drm_framebuffer *armada_fb_create(struct drm_device *dev, DRM_ERROR("failed to initialize framebuffer: %d\n", ret); return ERR_PTR(ret); } - -const struct drm_mode_config_funcs armada_drm_mode_config_funcs = { - .fb_create = armada_fb_create, - .output_poll_changed = drm_fb_helper_output_poll_changed, -}; diff --git a/drivers/gpu/drm/armada/armada_fb.h b/drivers/gpu/drm/armada/armada_fb.h index 5c130ff5da77..476daad0a36a 100644 --- a/drivers/gpu/drm/armada/armada_fb.h +++ b/drivers/gpu/drm/armada/armada_fb.h @@ -19,5 +19,6 @@ struct armada_framebuffer { struct armada_framebuffer *armada_framebuffer_create(struct drm_device *, const struct drm_mode_fb_cmd2 *, struct armada_gem_object *); - +struct drm_framebuffer *armada_fb_create(struct drm_device *dev, + struct drm_file *dfile, const struct drm_mode_fb_cmd2 *mode); #endif diff --git a/drivers/gpu/drm/armada/armada_fbdev.c b/drivers/gpu/drm/armada/armada_fbdev.c index 2a59db0994b2..8d23700848df 100644 --- a/drivers/gpu/drm/armada/armada_fbdev.c +++ b/drivers/gpu/drm/armada/armada_fbdev.c @@ -24,7 +24,7 @@ static /*const*/ struct fb_ops armada_fb_ops = { .fb_imageblit = drm_fb_helper_cfb_imageblit, }; -static int armada_fb_create(struct drm_fb_helper *fbh, +static int armada_fbdev_create(struct drm_fb_helper *fbh, struct drm_fb_helper_surface_size *sizes) { struct drm_device *dev = fbh->dev; @@ -108,7 +108,7 @@ static int armada_fb_probe(struct drm_fb_helper *fbh, int ret = 0; if (!fbh->fb) { - ret = armada_fb_create(fbh, sizes); + ret = armada_fbdev_create(fbh, sizes); if (ret == 0) ret = 1; } diff --git a/drivers/gpu/drm/armada/armada_gem.c b/drivers/gpu/drm/armada/armada_gem.c index 3fb37c75c065..892c1d9304bb 100644 --- a/drivers/gpu/drm/armada/armada_gem.c +++ b/drivers/gpu/drm/armada/armada_gem.c @@ -13,25 +13,14 @@ #include <drm/armada_drm.h> #include "armada_ioctlP.h" -static int armada_gem_vm_fault(struct vm_fault *vmf) +static vm_fault_t armada_gem_vm_fault(struct vm_fault *vmf) { struct drm_gem_object *gobj = vmf->vma->vm_private_data; struct armada_gem_object *obj = drm_to_armada_gem(gobj); unsigned long pfn = obj->phys_addr >> PAGE_SHIFT; - int ret; pfn += (vmf->address - vmf->vma->vm_start) >> PAGE_SHIFT; - ret = vm_insert_pfn(vmf->vma, vmf->address, pfn); - - switch (ret) { - case 0: - case -EBUSY: - return VM_FAULT_NOPAGE; - case -ENOMEM: - return VM_FAULT_OOM; - default: - return VM_FAULT_SIGBUS; - } + return vmf_insert_pfn(vmf->vma, vmf->address, pfn); } const struct vm_operations_struct armada_gem_vm_ops = { diff --git a/drivers/gpu/drm/armada/armada_hw.h b/drivers/gpu/drm/armada/armada_hw.h index 27319a8335e2..277580b36758 100644 --- a/drivers/gpu/drm/armada/armada_hw.h +++ b/drivers/gpu/drm/armada/armada_hw.h @@ -160,6 +160,7 @@ enum { CFG_ALPHAM_GRA = 0x1 << 16, CFG_ALPHAM_CFG = 0x2 << 16, CFG_ALPHA_MASK = 0xff << 8, +#define CFG_ALPHA(x) ((x) << 8) CFG_PIXCMD_MASK = 0xff, }; @@ -315,4 +316,19 @@ enum { PWRDN_IRQ_LEVEL = 1 << 0, }; +static inline u32 armada_rect_hw_fp(struct drm_rect *r) +{ + return (drm_rect_height(r) & 0xffff0000) | drm_rect_width(r) >> 16; +} + +static inline u32 armada_rect_hw(struct drm_rect *r) +{ + return drm_rect_height(r) << 16 | (drm_rect_width(r) & 0x0000ffff); +} + +static inline u32 armada_rect_yx(struct drm_rect *r) +{ + return (r)->y1 << 16 | ((r)->x1 & 0x0000ffff); +} + #endif diff --git a/drivers/gpu/drm/armada/armada_overlay.c b/drivers/gpu/drm/armada/armada_overlay.c index c391955009d6..eb7dfb65ef47 100644 --- a/drivers/gpu/drm/armada/armada_overlay.c +++ b/drivers/gpu/drm/armada/armada_overlay.c @@ -7,346 +7,468 @@ * published by the Free Software Foundation. */ #include <drm/drmP.h> +#include <drm/drm_atomic.h> #include <drm/drm_atomic_helper.h> +#include <drm/drm_plane_helper.h> +#include <drm/armada_drm.h> #include "armada_crtc.h" #include "armada_drm.h" #include "armada_fb.h" #include "armada_gem.h" #include "armada_hw.h" -#include <drm/armada_drm.h> #include "armada_ioctlP.h" +#include "armada_plane.h" #include "armada_trace.h" -struct armada_ovl_plane_properties { - uint32_t colorkey_yr; - uint32_t colorkey_ug; - uint32_t colorkey_vb; -#define K2R(val) (((val) >> 0) & 0xff) -#define K2G(val) (((val) >> 8) & 0xff) -#define K2B(val) (((val) >> 16) & 0xff) - int16_t brightness; - uint16_t contrast; - uint16_t saturation; - uint32_t colorkey_mode; -}; - -struct armada_ovl_plane { - struct armada_plane base; - struct armada_ovl_plane_properties prop; +#define DEFAULT_BRIGHTNESS 0 +#define DEFAULT_CONTRAST 0x4000 +#define DEFAULT_SATURATION 0x4000 +#define DEFAULT_ENCODING DRM_COLOR_YCBCR_BT601 + +struct armada_overlay_state { + struct drm_plane_state base; + u32 colorkey_yr; + u32 colorkey_ug; + u32 colorkey_vb; + u32 colorkey_mode; + u32 colorkey_enable; + s16 brightness; + u16 contrast; + u16 saturation; }; -#define drm_to_armada_ovl_plane(p) \ - container_of(p, struct armada_ovl_plane, base.base) - +#define drm_to_overlay_state(s) \ + container_of(s, struct armada_overlay_state, base) -static void -armada_ovl_update_attr(struct armada_ovl_plane_properties *prop, - struct armada_crtc *dcrtc) +static inline u32 armada_spu_contrast(struct drm_plane_state *state) { - writel_relaxed(prop->colorkey_yr, dcrtc->base + LCD_SPU_COLORKEY_Y); - writel_relaxed(prop->colorkey_ug, dcrtc->base + LCD_SPU_COLORKEY_U); - writel_relaxed(prop->colorkey_vb, dcrtc->base + LCD_SPU_COLORKEY_V); + return drm_to_overlay_state(state)->brightness << 16 | + drm_to_overlay_state(state)->contrast; +} - writel_relaxed(prop->brightness << 16 | prop->contrast, - dcrtc->base + LCD_SPU_CONTRAST); +static inline u32 armada_spu_saturation(struct drm_plane_state *state) +{ /* Docs say 15:0, but it seems to actually be 31:16 on Armada 510 */ - writel_relaxed(prop->saturation << 16, - dcrtc->base + LCD_SPU_SATURATION); - writel_relaxed(0x00002000, dcrtc->base + LCD_SPU_CBSH_HUE); - - spin_lock_irq(&dcrtc->irq_lock); - armada_updatel(prop->colorkey_mode | CFG_ALPHAM_GRA, - CFG_CKMODE_MASK | CFG_ALPHAM_MASK | CFG_ALPHA_MASK, - dcrtc->base + LCD_SPU_DMA_CTRL1); + return drm_to_overlay_state(state)->saturation << 16; +} - armada_updatel(ADV_GRACOLORKEY, 0, dcrtc->base + LCD_SPU_ADV_REG); - spin_unlock_irq(&dcrtc->irq_lock); +static inline u32 armada_csc(struct drm_plane_state *state) +{ + /* + * The CFG_CSC_RGB_* settings control the output of the colour space + * converter, setting the range of output values it produces. Since + * we will be blending with the full-range graphics, we need to + * produce full-range RGB output from the conversion. + */ + return CFG_CSC_RGB_COMPUTER | + (state->color_encoding == DRM_COLOR_YCBCR_BT709 ? + CFG_CSC_YUV_CCIR709 : CFG_CSC_YUV_CCIR601); } /* === Plane support === */ -static void armada_ovl_plane_work(struct armada_crtc *dcrtc, - struct armada_plane_work *work) +static void armada_drm_overlay_plane_atomic_update(struct drm_plane *plane, + struct drm_plane_state *old_state) { - unsigned long flags; + struct drm_plane_state *state = plane->state; + struct armada_crtc *dcrtc; + struct armada_regs *regs; + unsigned int idx; + u32 cfg, cfg_mask, val; - trace_armada_ovl_plane_work(&dcrtc->crtc, work->plane); + DRM_DEBUG_KMS("[PLANE:%d:%s]\n", plane->base.id, plane->name); - spin_lock_irqsave(&dcrtc->irq_lock, flags); - armada_drm_crtc_update_regs(dcrtc, work->regs); - spin_unlock_irqrestore(&dcrtc->irq_lock, flags); -} - -static void armada_ovl_plane_update_state(struct drm_plane_state *state, - struct armada_regs *regs) -{ - struct armada_ovl_plane *dplane = drm_to_armada_ovl_plane(state->plane); - struct armada_framebuffer *dfb = drm_fb_to_armada_fb(state->fb); - const struct drm_format_info *format; - unsigned int idx = 0; - bool fb_changed; - u32 val, ctrl0; - u16 src_x, src_y; + if (!state->fb || WARN_ON(!state->crtc)) + return; - ctrl0 = CFG_DMA_FMT(dfb->fmt) | CFG_DMA_MOD(dfb->mod) | CFG_CBSH_ENA; - if (state->visible) - ctrl0 |= CFG_DMA_ENA; - if (drm_rect_width(&state->src) >> 16 != drm_rect_width(&state->dst)) - ctrl0 |= CFG_DMA_HSMOOTH; + DRM_DEBUG_KMS("[PLANE:%d:%s] is on [CRTC:%d:%s] with [FB:%d] visible %u->%u\n", + plane->base.id, plane->name, + state->crtc->base.id, state->crtc->name, + state->fb->base.id, + old_state->visible, state->visible); - /* - * Shifting a YUV packed format image by one pixel causes the U/V - * planes to swap. Compensate for it by also toggling the UV swap. - */ - format = dfb->fb.format; - if (format->num_planes == 1 && state->src.x1 >> 16 & (format->hsub - 1)) - ctrl0 ^= CFG_DMA_MOD(CFG_SWAPUV); + dcrtc = drm_to_armada_crtc(state->crtc); + regs = dcrtc->regs + dcrtc->regs_idx; - if (~dplane->base.state.ctrl0 & ctrl0 & CFG_DMA_ENA) { - /* Power up the Y/U/V FIFOs on ENA 0->1 transitions */ + idx = 0; + if (!old_state->visible && state->visible) armada_reg_queue_mod(regs, idx, 0, CFG_PDWN16x66 | CFG_PDWN32x66, LCD_SPU_SRAM_PARA1); - } - - fb_changed = dplane->base.base.fb != &dfb->fb || - dplane->base.state.src_x != state->src.x1 >> 16 || - dplane->base.state.src_y != state->src.y1 >> 16; - - dplane->base.state.vsync_update = fb_changed; - + val = armada_rect_hw_fp(&state->src); + if (armada_rect_hw_fp(&old_state->src) != val) + armada_reg_queue_set(regs, idx, val, LCD_SPU_DMA_HPXL_VLN); + val = armada_rect_yx(&state->dst); + if (armada_rect_yx(&old_state->dst) != val) + armada_reg_queue_set(regs, idx, val, LCD_SPU_DMA_OVSA_HPXL_VLN); + val = armada_rect_hw(&state->dst); + if (armada_rect_hw(&old_state->dst) != val) + armada_reg_queue_set(regs, idx, val, LCD_SPU_DZM_HPXL_VLN); /* FIXME: overlay on an interlaced display */ - if (fb_changed) { - u32 addrs[3]; - - dplane->base.state.src_y = src_y = state->src.y1 >> 16; - dplane->base.state.src_x = src_x = state->src.x1 >> 16; + if (old_state->src.x1 != state->src.x1 || + old_state->src.y1 != state->src.y1 || + old_state->fb != state->fb) { + const struct drm_format_info *format; + u16 src_x, pitches[3]; + u32 addrs[2][3]; - armada_drm_plane_calc_addrs(addrs, &dfb->fb, src_x, src_y); + armada_drm_plane_calc(state, addrs, pitches, false); - armada_reg_queue_set(regs, idx, addrs[0], + armada_reg_queue_set(regs, idx, addrs[0][0], LCD_SPU_DMA_START_ADDR_Y0); - armada_reg_queue_set(regs, idx, addrs[1], + armada_reg_queue_set(regs, idx, addrs[0][1], LCD_SPU_DMA_START_ADDR_U0); - armada_reg_queue_set(regs, idx, addrs[2], + armada_reg_queue_set(regs, idx, addrs[0][2], LCD_SPU_DMA_START_ADDR_V0); - armada_reg_queue_set(regs, idx, addrs[0], + armada_reg_queue_set(regs, idx, addrs[1][0], LCD_SPU_DMA_START_ADDR_Y1); - armada_reg_queue_set(regs, idx, addrs[1], + armada_reg_queue_set(regs, idx, addrs[1][1], LCD_SPU_DMA_START_ADDR_U1); - armada_reg_queue_set(regs, idx, addrs[2], + armada_reg_queue_set(regs, idx, addrs[1][2], LCD_SPU_DMA_START_ADDR_V1); - val = dfb->fb.pitches[0] << 16 | dfb->fb.pitches[0]; - armada_reg_queue_set(regs, idx, val, - LCD_SPU_DMA_PITCH_YC); - val = dfb->fb.pitches[1] << 16 | dfb->fb.pitches[2]; - armada_reg_queue_set(regs, idx, val, - LCD_SPU_DMA_PITCH_UV); - } + val = pitches[0] << 16 | pitches[0]; + armada_reg_queue_set(regs, idx, val, LCD_SPU_DMA_PITCH_YC); + val = pitches[1] << 16 | pitches[2]; + armada_reg_queue_set(regs, idx, val, LCD_SPU_DMA_PITCH_UV); - val = (drm_rect_height(&state->src) & 0xffff0000) | - drm_rect_width(&state->src) >> 16; - if (dplane->base.state.src_hw != val) { - dplane->base.state.src_hw = val; - armada_reg_queue_set(regs, idx, val, - LCD_SPU_DMA_HPXL_VLN); - } + cfg = CFG_DMA_FMT(drm_fb_to_armada_fb(state->fb)->fmt) | + CFG_DMA_MOD(drm_fb_to_armada_fb(state->fb)->mod) | + CFG_CBSH_ENA; + if (state->visible) + cfg |= CFG_DMA_ENA; - val = drm_rect_height(&state->dst) << 16 | drm_rect_width(&state->dst); - if (dplane->base.state.dst_hw != val) { - dplane->base.state.dst_hw = val; - armada_reg_queue_set(regs, idx, val, - LCD_SPU_DZM_HPXL_VLN); + /* + * Shifting a YUV packed format image by one pixel causes the + * U/V planes to swap. Compensate for it by also toggling + * the UV swap. + */ + format = state->fb->format; + src_x = state->src.x1 >> 16; + if (format->num_planes == 1 && src_x & (format->hsub - 1)) + cfg ^= CFG_DMA_MOD(CFG_SWAPUV); + cfg_mask = CFG_CBSH_ENA | CFG_DMAFORMAT | + CFG_DMA_MOD(CFG_SWAPRB | CFG_SWAPUV | + CFG_SWAPYU | CFG_YUV2RGB) | + CFG_DMA_FTOGGLE | CFG_DMA_TSTMODE | + CFG_DMA_ENA; + } else if (old_state->visible != state->visible) { + cfg = state->visible ? CFG_DMA_ENA : 0; + cfg_mask = CFG_DMA_ENA; + } else { + cfg = cfg_mask = 0; } - - val = state->dst.y1 << 16 | state->dst.x1; - if (dplane->base.state.dst_yx != val) { - dplane->base.state.dst_yx = val; - armada_reg_queue_set(regs, idx, val, - LCD_SPU_DMA_OVSA_HPXL_VLN); + if (drm_rect_width(&old_state->src) != drm_rect_width(&state->src) || + drm_rect_width(&old_state->dst) != drm_rect_width(&state->dst)) { + cfg_mask |= CFG_DMA_HSMOOTH; + if (drm_rect_width(&state->src) >> 16 != + drm_rect_width(&state->dst)) + cfg |= CFG_DMA_HSMOOTH; } - if (dplane->base.state.ctrl0 != ctrl0) { - dplane->base.state.ctrl0 = ctrl0; - armada_reg_queue_mod(regs, idx, ctrl0, - CFG_CBSH_ENA | CFG_DMAFORMAT | CFG_DMA_FTOGGLE | - CFG_DMA_HSMOOTH | CFG_DMA_TSTMODE | - CFG_DMA_MOD(CFG_SWAPRB | CFG_SWAPUV | CFG_SWAPYU | - CFG_YUV2RGB) | CFG_DMA_ENA, - LCD_SPU_DMA_CTRL0); - dplane->base.state.vsync_update = true; - } + if (cfg_mask) + armada_reg_queue_mod(regs, idx, cfg, cfg_mask, + LCD_SPU_DMA_CTRL0); + + val = armada_spu_contrast(state); + if ((!old_state->visible && state->visible) || + armada_spu_contrast(old_state) != val) + armada_reg_queue_set(regs, idx, val, LCD_SPU_CONTRAST); + val = armada_spu_saturation(state); + if ((!old_state->visible && state->visible) || + armada_spu_saturation(old_state) != val) + armada_reg_queue_set(regs, idx, val, LCD_SPU_SATURATION); + if (!old_state->visible && state->visible) + armada_reg_queue_set(regs, idx, 0x00002000, LCD_SPU_CBSH_HUE); + val = armada_csc(state); + if ((!old_state->visible && state->visible) || + armada_csc(old_state) != val) + armada_reg_queue_mod(regs, idx, val, CFG_CSC_MASK, + LCD_SPU_IOPAD_CONTROL); + val = drm_to_overlay_state(state)->colorkey_yr; + if ((!old_state->visible && state->visible) || + drm_to_overlay_state(old_state)->colorkey_yr != val) + armada_reg_queue_set(regs, idx, val, LCD_SPU_COLORKEY_Y); + val = drm_to_overlay_state(state)->colorkey_ug; + if ((!old_state->visible && state->visible) || + drm_to_overlay_state(old_state)->colorkey_ug != val) + armada_reg_queue_set(regs, idx, val, LCD_SPU_COLORKEY_U); + val = drm_to_overlay_state(state)->colorkey_vb; + if ((!old_state->visible && state->visible) || + drm_to_overlay_state(old_state)->colorkey_vb != val) + armada_reg_queue_set(regs, idx, val, LCD_SPU_COLORKEY_V); + val = drm_to_overlay_state(state)->colorkey_mode; + if ((!old_state->visible && state->visible) || + drm_to_overlay_state(old_state)->colorkey_mode != val) + armada_reg_queue_mod(regs, idx, val, CFG_CKMODE_MASK | + CFG_ALPHAM_MASK | CFG_ALPHA_MASK, + LCD_SPU_DMA_CTRL1); + val = drm_to_overlay_state(state)->colorkey_enable; + if (((!old_state->visible && state->visible) || + drm_to_overlay_state(old_state)->colorkey_enable != val) && + dcrtc->variant->has_spu_adv_reg) + armada_reg_queue_mod(regs, idx, val, ADV_GRACOLORKEY | + ADV_VIDCOLORKEY, LCD_SPU_ADV_REG); + + dcrtc->regs_idx += idx; +} + +static void armada_drm_overlay_plane_atomic_disable(struct drm_plane *plane, + struct drm_plane_state *old_state) +{ + struct armada_crtc *dcrtc; + struct armada_regs *regs; + unsigned int idx = 0; + + DRM_DEBUG_KMS("[PLANE:%d:%s]\n", plane->base.id, plane->name); + + if (!old_state->crtc) + return; - dplane->base.state.changed = idx != 0; + DRM_DEBUG_KMS("[PLANE:%d:%s] was on [CRTC:%d:%s] with [FB:%d]\n", + plane->base.id, plane->name, + old_state->crtc->base.id, old_state->crtc->name, + old_state->fb->base.id); - armada_reg_queue_end(regs, idx); + dcrtc = drm_to_armada_crtc(old_state->crtc); + regs = dcrtc->regs + dcrtc->regs_idx; + + /* Disable plane and power down the YUV FIFOs */ + armada_reg_queue_mod(regs, idx, 0, CFG_DMA_ENA, LCD_SPU_DMA_CTRL0); + armada_reg_queue_mod(regs, idx, CFG_PDWN16x66 | CFG_PDWN32x66, 0, + LCD_SPU_SRAM_PARA1); + + dcrtc->regs_idx += idx; } +static const struct drm_plane_helper_funcs armada_overlay_plane_helper_funcs = { + .prepare_fb = armada_drm_plane_prepare_fb, + .cleanup_fb = armada_drm_plane_cleanup_fb, + .atomic_check = armada_drm_plane_atomic_check, + .atomic_update = armada_drm_overlay_plane_atomic_update, + .atomic_disable = armada_drm_overlay_plane_atomic_disable, +}; + static int -armada_ovl_plane_update(struct drm_plane *plane, struct drm_crtc *crtc, +armada_overlay_plane_update(struct drm_plane *plane, struct drm_crtc *crtc, struct drm_framebuffer *fb, int crtc_x, int crtc_y, unsigned crtc_w, unsigned crtc_h, uint32_t src_x, uint32_t src_y, uint32_t src_w, uint32_t src_h, struct drm_modeset_acquire_ctx *ctx) { - struct armada_ovl_plane *dplane = drm_to_armada_ovl_plane(plane); - struct armada_crtc *dcrtc = drm_to_armada_crtc(crtc); - struct armada_plane_work *work; - struct drm_plane_state state = { - .plane = plane, - .crtc = crtc, - .fb = fb, - .src_x = src_x, - .src_y = src_y, - .src_w = src_w, - .src_h = src_h, - .crtc_x = crtc_x, - .crtc_y = crtc_y, - .crtc_w = crtc_w, - .crtc_h = crtc_h, - .rotation = DRM_MODE_ROTATE_0, - }; - struct drm_crtc_state crtc_state = { - .crtc = crtc, - .enable = crtc->enabled, - .mode = crtc->mode, - }; - int ret; + struct drm_atomic_state *state; + struct drm_plane_state *plane_state; + int ret = 0; trace_armada_ovl_plane_update(plane, crtc, fb, crtc_x, crtc_y, crtc_w, crtc_h, src_x, src_y, src_w, src_h); - ret = drm_atomic_helper_check_plane_state(&state, &crtc_state, 0, - INT_MAX, true, false); - if (ret) - return ret; - - work = &dplane->base.works[dplane->base.next_work]; - - if (plane->fb != fb) { - /* - * Take a reference on the new framebuffer - we want to - * hold on to it while the hardware is displaying it. - */ - drm_framebuffer_reference(fb); + state = drm_atomic_state_alloc(plane->dev); + if (!state) + return -ENOMEM; - work->old_fb = plane->fb; - } else { - work->old_fb = NULL; + state->acquire_ctx = ctx; + plane_state = drm_atomic_get_plane_state(state, plane); + if (IS_ERR(plane_state)) { + ret = PTR_ERR(plane_state); + goto fail; } - armada_ovl_plane_update_state(&state, work->regs); - - if (!dplane->base.state.changed) - return 0; - - /* Wait for pending work to complete */ - if (armada_drm_plane_work_wait(&dplane->base, HZ / 25) == 0) - armada_drm_plane_work_cancel(dcrtc, &dplane->base); + ret = drm_atomic_set_crtc_for_plane(plane_state, crtc); + if (ret != 0) + goto fail; + + drm_atomic_set_fb_for_plane(plane_state, fb); + plane_state->crtc_x = crtc_x; + plane_state->crtc_y = crtc_y; + plane_state->crtc_h = crtc_h; + plane_state->crtc_w = crtc_w; + plane_state->src_x = src_x; + plane_state->src_y = src_y; + plane_state->src_h = src_h; + plane_state->src_w = src_w; + + ret = drm_atomic_nonblocking_commit(state); +fail: + drm_atomic_state_put(state); + return ret; +} - /* Just updating the position/size? */ - if (!dplane->base.state.vsync_update) { - armada_ovl_plane_work(dcrtc, work); - return 0; - } +static void armada_ovl_plane_destroy(struct drm_plane *plane) +{ + drm_plane_cleanup(plane); + kfree(plane); +} - if (!dcrtc->plane) { - dcrtc->plane = plane; - armada_ovl_update_attr(&dplane->prop, dcrtc); +static void armada_overlay_reset(struct drm_plane *plane) +{ + struct armada_overlay_state *state; + + if (plane->state) + __drm_atomic_helper_plane_destroy_state(plane->state); + kfree(plane->state); + + state = kzalloc(sizeof(*state), GFP_KERNEL); + if (state) { + state->base.plane = plane; + state->base.color_encoding = DEFAULT_ENCODING; + state->base.color_range = DRM_COLOR_YCBCR_LIMITED_RANGE; + state->base.rotation = DRM_MODE_ROTATE_0; + state->colorkey_yr = 0xfefefe00; + state->colorkey_ug = 0x01010100; + state->colorkey_vb = 0x01010100; + state->colorkey_mode = CFG_CKMODE(CKMODE_RGB) | + CFG_ALPHAM_GRA | CFG_ALPHA(0); + state->colorkey_enable = ADV_GRACOLORKEY; + state->brightness = DEFAULT_BRIGHTNESS; + state->contrast = DEFAULT_CONTRAST; + state->saturation = DEFAULT_SATURATION; } - - /* Queue it for update on the next interrupt if we are enabled */ - ret = armada_drm_plane_work_queue(dcrtc, work); - if (ret) - DRM_ERROR("failed to queue plane work: %d\n", ret); - - dplane->base.next_work = !dplane->base.next_work; - - return 0; + plane->state = &state->base; } -static void armada_ovl_plane_destroy(struct drm_plane *plane) +struct drm_plane_state * +armada_overlay_duplicate_state(struct drm_plane *plane) { - struct armada_ovl_plane *dplane = drm_to_armada_ovl_plane(plane); + struct armada_overlay_state *state; - drm_plane_cleanup(plane); + if (WARN_ON(!plane->state)) + return NULL; - kfree(dplane); + state = kmemdup(plane->state, sizeof(*state), GFP_KERNEL); + if (state) + __drm_atomic_helper_plane_duplicate_state(plane, &state->base); + return &state->base; } -static int armada_ovl_plane_set_property(struct drm_plane *plane, - struct drm_property *property, uint64_t val) +static int armada_overlay_set_property(struct drm_plane *plane, + struct drm_plane_state *state, struct drm_property *property, + uint64_t val) { struct armada_private *priv = plane->dev->dev_private; - struct armada_ovl_plane *dplane = drm_to_armada_ovl_plane(plane); - bool update_attr = false; +#define K2R(val) (((val) >> 0) & 0xff) +#define K2G(val) (((val) >> 8) & 0xff) +#define K2B(val) (((val) >> 16) & 0xff) if (property == priv->colorkey_prop) { #define CCC(v) ((v) << 24 | (v) << 16 | (v) << 8) - dplane->prop.colorkey_yr = CCC(K2R(val)); - dplane->prop.colorkey_ug = CCC(K2G(val)); - dplane->prop.colorkey_vb = CCC(K2B(val)); + drm_to_overlay_state(state)->colorkey_yr = CCC(K2R(val)); + drm_to_overlay_state(state)->colorkey_ug = CCC(K2G(val)); + drm_to_overlay_state(state)->colorkey_vb = CCC(K2B(val)); #undef CCC - update_attr = true; } else if (property == priv->colorkey_min_prop) { - dplane->prop.colorkey_yr &= ~0x00ff0000; - dplane->prop.colorkey_yr |= K2R(val) << 16; - dplane->prop.colorkey_ug &= ~0x00ff0000; - dplane->prop.colorkey_ug |= K2G(val) << 16; - dplane->prop.colorkey_vb &= ~0x00ff0000; - dplane->prop.colorkey_vb |= K2B(val) << 16; - update_attr = true; + drm_to_overlay_state(state)->colorkey_yr &= ~0x00ff0000; + drm_to_overlay_state(state)->colorkey_yr |= K2R(val) << 16; + drm_to_overlay_state(state)->colorkey_ug &= ~0x00ff0000; + drm_to_overlay_state(state)->colorkey_ug |= K2G(val) << 16; + drm_to_overlay_state(state)->colorkey_vb &= ~0x00ff0000; + drm_to_overlay_state(state)->colorkey_vb |= K2B(val) << 16; } else if (property == priv->colorkey_max_prop) { - dplane->prop.colorkey_yr &= ~0xff000000; - dplane->prop.colorkey_yr |= K2R(val) << 24; - dplane->prop.colorkey_ug &= ~0xff000000; - dplane->prop.colorkey_ug |= K2G(val) << 24; - dplane->prop.colorkey_vb &= ~0xff000000; - dplane->prop.colorkey_vb |= K2B(val) << 24; - update_attr = true; + drm_to_overlay_state(state)->colorkey_yr &= ~0xff000000; + drm_to_overlay_state(state)->colorkey_yr |= K2R(val) << 24; + drm_to_overlay_state(state)->colorkey_ug &= ~0xff000000; + drm_to_overlay_state(state)->colorkey_ug |= K2G(val) << 24; + drm_to_overlay_state(state)->colorkey_vb &= ~0xff000000; + drm_to_overlay_state(state)->colorkey_vb |= K2B(val) << 24; } else if (property == priv->colorkey_val_prop) { - dplane->prop.colorkey_yr &= ~0x0000ff00; - dplane->prop.colorkey_yr |= K2R(val) << 8; - dplane->prop.colorkey_ug &= ~0x0000ff00; - dplane->prop.colorkey_ug |= K2G(val) << 8; - dplane->prop.colorkey_vb &= ~0x0000ff00; - dplane->prop.colorkey_vb |= K2B(val) << 8; - update_attr = true; + drm_to_overlay_state(state)->colorkey_yr &= ~0x0000ff00; + drm_to_overlay_state(state)->colorkey_yr |= K2R(val) << 8; + drm_to_overlay_state(state)->colorkey_ug &= ~0x0000ff00; + drm_to_overlay_state(state)->colorkey_ug |= K2G(val) << 8; + drm_to_overlay_state(state)->colorkey_vb &= ~0x0000ff00; + drm_to_overlay_state(state)->colorkey_vb |= K2B(val) << 8; } else if (property == priv->colorkey_alpha_prop) { - dplane->prop.colorkey_yr &= ~0x000000ff; - dplane->prop.colorkey_yr |= K2R(val); - dplane->prop.colorkey_ug &= ~0x000000ff; - dplane->prop.colorkey_ug |= K2G(val); - dplane->prop.colorkey_vb &= ~0x000000ff; - dplane->prop.colorkey_vb |= K2B(val); - update_attr = true; + drm_to_overlay_state(state)->colorkey_yr &= ~0x000000ff; + drm_to_overlay_state(state)->colorkey_yr |= K2R(val); + drm_to_overlay_state(state)->colorkey_ug &= ~0x000000ff; + drm_to_overlay_state(state)->colorkey_ug |= K2G(val); + drm_to_overlay_state(state)->colorkey_vb &= ~0x000000ff; + drm_to_overlay_state(state)->colorkey_vb |= K2B(val); } else if (property == priv->colorkey_mode_prop) { - dplane->prop.colorkey_mode &= ~CFG_CKMODE_MASK; - dplane->prop.colorkey_mode |= CFG_CKMODE(val); - update_attr = true; + if (val == CKMODE_DISABLE) { + drm_to_overlay_state(state)->colorkey_mode = + CFG_CKMODE(CKMODE_DISABLE) | + CFG_ALPHAM_CFG | CFG_ALPHA(255); + drm_to_overlay_state(state)->colorkey_enable = 0; + } else { + drm_to_overlay_state(state)->colorkey_mode = + CFG_CKMODE(val) | + CFG_ALPHAM_GRA | CFG_ALPHA(0); + drm_to_overlay_state(state)->colorkey_enable = + ADV_GRACOLORKEY; + } } else if (property == priv->brightness_prop) { - dplane->prop.brightness = val - 256; - update_attr = true; + drm_to_overlay_state(state)->brightness = val - 256; } else if (property == priv->contrast_prop) { - dplane->prop.contrast = val; - update_attr = true; + drm_to_overlay_state(state)->contrast = val; } else if (property == priv->saturation_prop) { - dplane->prop.saturation = val; - update_attr = true; + drm_to_overlay_state(state)->saturation = val; + } else { + return -EINVAL; } + return 0; +} - if (update_attr && dplane->base.base.crtc) - armada_ovl_update_attr(&dplane->prop, - drm_to_armada_crtc(dplane->base.base.crtc)); +static int armada_overlay_get_property(struct drm_plane *plane, + const struct drm_plane_state *state, struct drm_property *property, + uint64_t *val) +{ + struct armada_private *priv = plane->dev->dev_private; +#define C2K(c,s) (((c) >> (s)) & 0xff) +#define R2BGR(r,g,b,s) (C2K(r,s) << 0 | C2K(g,s) << 8 | C2K(b,s) << 16) + if (property == priv->colorkey_prop) { + /* Do best-efforts here for this property */ + *val = R2BGR(drm_to_overlay_state(state)->colorkey_yr, + drm_to_overlay_state(state)->colorkey_ug, + drm_to_overlay_state(state)->colorkey_vb, 16); + /* If min != max, or min != val, error out */ + if (*val != R2BGR(drm_to_overlay_state(state)->colorkey_yr, + drm_to_overlay_state(state)->colorkey_ug, + drm_to_overlay_state(state)->colorkey_vb, 24) || + *val != R2BGR(drm_to_overlay_state(state)->colorkey_yr, + drm_to_overlay_state(state)->colorkey_ug, + drm_to_overlay_state(state)->colorkey_vb, 8)) + return -EINVAL; + } else if (property == priv->colorkey_min_prop) { + *val = R2BGR(drm_to_overlay_state(state)->colorkey_yr, + drm_to_overlay_state(state)->colorkey_ug, + drm_to_overlay_state(state)->colorkey_vb, 16); + } else if (property == priv->colorkey_max_prop) { + *val = R2BGR(drm_to_overlay_state(state)->colorkey_yr, + drm_to_overlay_state(state)->colorkey_ug, + drm_to_overlay_state(state)->colorkey_vb, 24); + } else if (property == priv->colorkey_val_prop) { + *val = R2BGR(drm_to_overlay_state(state)->colorkey_yr, + drm_to_overlay_state(state)->colorkey_ug, + drm_to_overlay_state(state)->colorkey_vb, 8); + } else if (property == priv->colorkey_alpha_prop) { + *val = R2BGR(drm_to_overlay_state(state)->colorkey_yr, + drm_to_overlay_state(state)->colorkey_ug, + drm_to_overlay_state(state)->colorkey_vb, 0); + } else if (property == priv->colorkey_mode_prop) { + *val = (drm_to_overlay_state(state)->colorkey_mode & + CFG_CKMODE_MASK) >> ffs(CFG_CKMODE_MASK); + } else if (property == priv->brightness_prop) { + *val = drm_to_overlay_state(state)->brightness + 256; + } else if (property == priv->contrast_prop) { + *val = drm_to_overlay_state(state)->contrast; + } else if (property == priv->saturation_prop) { + *val = drm_to_overlay_state(state)->saturation; + } else { + return -EINVAL; + } return 0; } static const struct drm_plane_funcs armada_ovl_plane_funcs = { - .update_plane = armada_ovl_plane_update, - .disable_plane = armada_drm_plane_disable, + .update_plane = armada_overlay_plane_update, + .disable_plane = drm_atomic_helper_disable_plane, .destroy = armada_ovl_plane_destroy, - .set_property = armada_ovl_plane_set_property, + .reset = armada_overlay_reset, + .atomic_duplicate_state = armada_overlay_duplicate_state, + .atomic_destroy_state = drm_atomic_helper_plane_destroy_state, + .atomic_set_property = armada_overlay_set_property, + .atomic_get_property = armada_overlay_get_property, }; static const uint32_t armada_ovl_formats[] = { @@ -419,46 +541,31 @@ int armada_overlay_plane_create(struct drm_device *dev, unsigned long crtcs) { struct armada_private *priv = dev->dev_private; struct drm_mode_object *mobj; - struct armada_ovl_plane *dplane; + struct drm_plane *overlay; int ret; ret = armada_overlay_create_properties(dev); if (ret) return ret; - dplane = kzalloc(sizeof(*dplane), GFP_KERNEL); - if (!dplane) + overlay = kzalloc(sizeof(*overlay), GFP_KERNEL); + if (!overlay) return -ENOMEM; - ret = armada_drm_plane_init(&dplane->base); - if (ret) { - kfree(dplane); - return ret; - } - - dplane->base.works[0].fn = armada_ovl_plane_work; - dplane->base.works[1].fn = armada_ovl_plane_work; + drm_plane_helper_add(overlay, &armada_overlay_plane_helper_funcs); - ret = drm_universal_plane_init(dev, &dplane->base.base, crtcs, + ret = drm_universal_plane_init(dev, overlay, crtcs, &armada_ovl_plane_funcs, armada_ovl_formats, ARRAY_SIZE(armada_ovl_formats), NULL, DRM_PLANE_TYPE_OVERLAY, NULL); if (ret) { - kfree(dplane); + kfree(overlay); return ret; } - dplane->prop.colorkey_yr = 0xfefefe00; - dplane->prop.colorkey_ug = 0x01010100; - dplane->prop.colorkey_vb = 0x01010100; - dplane->prop.colorkey_mode = CFG_CKMODE(CKMODE_RGB); - dplane->prop.brightness = 0; - dplane->prop.contrast = 0x4000; - dplane->prop.saturation = 0x4000; - - mobj = &dplane->base.base.base; + mobj = &overlay->base; drm_object_attach_property(mobj, priv->colorkey_prop, 0x0101fe); drm_object_attach_property(mobj, priv->colorkey_min_prop, @@ -471,11 +578,19 @@ int armada_overlay_plane_create(struct drm_device *dev, unsigned long crtcs) 0x000000); drm_object_attach_property(mobj, priv->colorkey_mode_prop, CKMODE_RGB); - drm_object_attach_property(mobj, priv->brightness_prop, 256); + drm_object_attach_property(mobj, priv->brightness_prop, + 256 + DEFAULT_BRIGHTNESS); drm_object_attach_property(mobj, priv->contrast_prop, - dplane->prop.contrast); + DEFAULT_CONTRAST); drm_object_attach_property(mobj, priv->saturation_prop, - dplane->prop.saturation); + DEFAULT_SATURATION); - return 0; + ret = drm_plane_create_color_properties(overlay, + BIT(DRM_COLOR_YCBCR_BT601) | + BIT(DRM_COLOR_YCBCR_BT709), + BIT(DRM_COLOR_YCBCR_LIMITED_RANGE), + DEFAULT_ENCODING, + DRM_COLOR_YCBCR_LIMITED_RANGE); + + return ret; } diff --git a/drivers/gpu/drm/armada/armada_plane.c b/drivers/gpu/drm/armada/armada_plane.c new file mode 100644 index 000000000000..9f36423dd394 --- /dev/null +++ b/drivers/gpu/drm/armada/armada_plane.c @@ -0,0 +1,289 @@ +/* + * Copyright (C) 2012 Russell King + * Rewritten from the dovefb driver, and Armada510 manuals. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ +#include <drm/drmP.h> +#include <drm/drm_atomic.h> +#include <drm/drm_atomic_helper.h> +#include <drm/drm_plane_helper.h> +#include "armada_crtc.h" +#include "armada_drm.h" +#include "armada_fb.h" +#include "armada_gem.h" +#include "armada_hw.h" +#include "armada_plane.h" +#include "armada_trace.h" + +static const uint32_t armada_primary_formats[] = { + DRM_FORMAT_UYVY, + DRM_FORMAT_YUYV, + DRM_FORMAT_VYUY, + DRM_FORMAT_YVYU, + DRM_FORMAT_ARGB8888, + DRM_FORMAT_ABGR8888, + DRM_FORMAT_XRGB8888, + DRM_FORMAT_XBGR8888, + DRM_FORMAT_RGB888, + DRM_FORMAT_BGR888, + DRM_FORMAT_ARGB1555, + DRM_FORMAT_ABGR1555, + DRM_FORMAT_RGB565, + DRM_FORMAT_BGR565, +}; + +void armada_drm_plane_calc(struct drm_plane_state *state, u32 addrs[2][3], + u16 pitches[3], bool interlaced) +{ + struct drm_framebuffer *fb = state->fb; + const struct drm_format_info *format = fb->format; + unsigned int num_planes = format->num_planes; + unsigned int x = state->src.x1 >> 16; + unsigned int y = state->src.y1 >> 16; + u32 addr = drm_fb_obj(fb)->dev_addr; + int i; + + DRM_DEBUG_KMS("pitch %u x %d y %d bpp %d\n", + fb->pitches[0], x, y, format->cpp[0] * 8); + + if (num_planes > 3) + num_planes = 3; + + addrs[0][0] = addr + fb->offsets[0] + y * fb->pitches[0] + + x * format->cpp[0]; + pitches[0] = fb->pitches[0]; + + y /= format->vsub; + x /= format->hsub; + + for (i = 1; i < num_planes; i++) { + addrs[0][i] = addr + fb->offsets[i] + y * fb->pitches[i] + + x * format->cpp[i]; + pitches[i] = fb->pitches[i]; + } + for (; i < 3; i++) { + addrs[0][i] = 0; + pitches[i] = 0; + } + if (interlaced) { + for (i = 0; i < 3; i++) { + addrs[1][i] = addrs[0][i] + pitches[i]; + pitches[i] *= 2; + } + } else { + for (i = 0; i < 3; i++) + addrs[1][i] = addrs[0][i]; + } +} + +static unsigned armada_drm_crtc_calc_fb(struct drm_plane_state *state, + struct armada_regs *regs, bool interlaced) +{ + u16 pitches[3]; + u32 addrs[2][3]; + unsigned i = 0; + + armada_drm_plane_calc(state, addrs, pitches, interlaced); + + /* write offset, base, and pitch */ + armada_reg_queue_set(regs, i, addrs[0][0], LCD_CFG_GRA_START_ADDR0); + armada_reg_queue_set(regs, i, addrs[1][0], LCD_CFG_GRA_START_ADDR1); + armada_reg_queue_mod(regs, i, pitches[0], 0xffff, LCD_CFG_GRA_PITCH); + + return i; +} + +int armada_drm_plane_prepare_fb(struct drm_plane *plane, + struct drm_plane_state *state) +{ + DRM_DEBUG_KMS("[PLANE:%d:%s] [FB:%d]\n", + plane->base.id, plane->name, + state->fb ? state->fb->base.id : 0); + + /* + * Take a reference on the new framebuffer - we want to + * hold on to it while the hardware is displaying it. + */ + if (state->fb) + drm_framebuffer_get(state->fb); + return 0; +} + +void armada_drm_plane_cleanup_fb(struct drm_plane *plane, + struct drm_plane_state *old_state) +{ + DRM_DEBUG_KMS("[PLANE:%d:%s] [FB:%d]\n", + plane->base.id, plane->name, + old_state->fb ? old_state->fb->base.id : 0); + + if (old_state->fb) + drm_framebuffer_put(old_state->fb); +} + +int armada_drm_plane_atomic_check(struct drm_plane *plane, + struct drm_plane_state *state) +{ + if (state->fb && !WARN_ON(!state->crtc)) { + struct drm_crtc *crtc = state->crtc; + struct drm_crtc_state *crtc_state; + + if (state->state) + crtc_state = drm_atomic_get_existing_crtc_state(state->state, crtc); + else + crtc_state = crtc->state; + return drm_atomic_helper_check_plane_state(state, crtc_state, + 0, INT_MAX, + true, false); + } else { + state->visible = false; + } + return 0; +} + +static void armada_drm_primary_plane_atomic_update(struct drm_plane *plane, + struct drm_plane_state *old_state) +{ + struct drm_plane_state *state = plane->state; + struct armada_crtc *dcrtc; + struct armada_regs *regs; + u32 cfg, cfg_mask, val; + unsigned int idx; + + DRM_DEBUG_KMS("[PLANE:%d:%s]\n", plane->base.id, plane->name); + + if (!state->fb || WARN_ON(!state->crtc)) + return; + + DRM_DEBUG_KMS("[PLANE:%d:%s] is on [CRTC:%d:%s] with [FB:%d] visible %u->%u\n", + plane->base.id, plane->name, + state->crtc->base.id, state->crtc->name, + state->fb->base.id, + old_state->visible, state->visible); + + dcrtc = drm_to_armada_crtc(state->crtc); + regs = dcrtc->regs + dcrtc->regs_idx; + + idx = 0; + if (!old_state->visible && state->visible) { + val = CFG_PDWN64x66; + if (drm_fb_to_armada_fb(state->fb)->fmt > CFG_420) + val |= CFG_PDWN256x24; + armada_reg_queue_mod(regs, idx, 0, val, LCD_SPU_SRAM_PARA1); + } + val = armada_rect_hw_fp(&state->src); + if (armada_rect_hw_fp(&old_state->src) != val) + armada_reg_queue_set(regs, idx, val, LCD_SPU_GRA_HPXL_VLN); + val = armada_rect_yx(&state->dst); + if (armada_rect_yx(&old_state->dst) != val) + armada_reg_queue_set(regs, idx, val, LCD_SPU_GRA_OVSA_HPXL_VLN); + val = armada_rect_hw(&state->dst); + if (armada_rect_hw(&old_state->dst) != val) + armada_reg_queue_set(regs, idx, val, LCD_SPU_GZM_HPXL_VLN); + if (old_state->src.x1 != state->src.x1 || + old_state->src.y1 != state->src.y1 || + old_state->fb != state->fb || + state->crtc->state->mode_changed) { + idx += armada_drm_crtc_calc_fb(state, regs + idx, + dcrtc->interlaced); + } + if (old_state->fb != state->fb || + state->crtc->state->mode_changed) { + cfg = CFG_GRA_FMT(drm_fb_to_armada_fb(state->fb)->fmt) | + CFG_GRA_MOD(drm_fb_to_armada_fb(state->fb)->mod); + if (drm_fb_to_armada_fb(state->fb)->fmt > CFG_420) + cfg |= CFG_PALETTE_ENA; + if (state->visible) + cfg |= CFG_GRA_ENA; + if (dcrtc->interlaced) + cfg |= CFG_GRA_FTOGGLE; + cfg_mask = CFG_GRAFORMAT | + CFG_GRA_MOD(CFG_SWAPRB | CFG_SWAPUV | + CFG_SWAPYU | CFG_YUV2RGB) | + CFG_PALETTE_ENA | CFG_GRA_FTOGGLE | + CFG_GRA_ENA; + } else if (old_state->visible != state->visible) { + cfg = state->visible ? CFG_GRA_ENA : 0; + cfg_mask = CFG_GRA_ENA; + } else { + cfg = cfg_mask = 0; + } + if (drm_rect_width(&old_state->src) != drm_rect_width(&state->src) || + drm_rect_width(&old_state->dst) != drm_rect_width(&state->dst)) { + cfg_mask |= CFG_GRA_HSMOOTH; + if (drm_rect_width(&state->src) >> 16 != + drm_rect_width(&state->dst)) + cfg |= CFG_GRA_HSMOOTH; + } + + if (cfg_mask) + armada_reg_queue_mod(regs, idx, cfg, cfg_mask, + LCD_SPU_DMA_CTRL0); + + dcrtc->regs_idx += idx; +} + +static void armada_drm_primary_plane_atomic_disable(struct drm_plane *plane, + struct drm_plane_state *old_state) +{ + struct armada_crtc *dcrtc; + struct armada_regs *regs; + unsigned int idx = 0; + + DRM_DEBUG_KMS("[PLANE:%d:%s]\n", plane->base.id, plane->name); + + if (!old_state->crtc) + return; + + DRM_DEBUG_KMS("[PLANE:%d:%s] was on [CRTC:%d:%s] with [FB:%d]\n", + plane->base.id, plane->name, + old_state->crtc->base.id, old_state->crtc->name, + old_state->fb->base.id); + + dcrtc = drm_to_armada_crtc(old_state->crtc); + regs = dcrtc->regs + dcrtc->regs_idx; + + /* Disable plane and power down most RAMs and FIFOs */ + armada_reg_queue_mod(regs, idx, 0, CFG_GRA_ENA, LCD_SPU_DMA_CTRL0); + armada_reg_queue_mod(regs, idx, CFG_PDWN256x32 | CFG_PDWN256x24 | + CFG_PDWN256x8 | CFG_PDWN32x32 | CFG_PDWN64x66, + 0, LCD_SPU_SRAM_PARA1); + + dcrtc->regs_idx += idx; +} + +static const struct drm_plane_helper_funcs armada_primary_plane_helper_funcs = { + .prepare_fb = armada_drm_plane_prepare_fb, + .cleanup_fb = armada_drm_plane_cleanup_fb, + .atomic_check = armada_drm_plane_atomic_check, + .atomic_update = armada_drm_primary_plane_atomic_update, + .atomic_disable = armada_drm_primary_plane_atomic_disable, +}; + +static const struct drm_plane_funcs armada_primary_plane_funcs = { + .update_plane = drm_atomic_helper_update_plane, + .disable_plane = drm_atomic_helper_disable_plane, + .destroy = drm_primary_helper_destroy, + .reset = drm_atomic_helper_plane_reset, + .atomic_duplicate_state = drm_atomic_helper_plane_duplicate_state, + .atomic_destroy_state = drm_atomic_helper_plane_destroy_state, +}; + +int armada_drm_primary_plane_init(struct drm_device *drm, + struct drm_plane *primary) +{ + int ret; + + drm_plane_helper_add(primary, &armada_primary_plane_helper_funcs); + + ret = drm_universal_plane_init(drm, primary, 0, + &armada_primary_plane_funcs, + armada_primary_formats, + ARRAY_SIZE(armada_primary_formats), + NULL, + DRM_PLANE_TYPE_PRIMARY, NULL); + + return ret; +} diff --git a/drivers/gpu/drm/armada/armada_plane.h b/drivers/gpu/drm/armada/armada_plane.h new file mode 100644 index 000000000000..ff4281ba7fad --- /dev/null +++ b/drivers/gpu/drm/armada/armada_plane.h @@ -0,0 +1,15 @@ +#ifndef ARMADA_PLANE_H +#define ARMADA_PLANE_H + +void armada_drm_plane_calc(struct drm_plane_state *state, u32 addrs[2][3], + u16 pitches[3], bool interlaced); +int armada_drm_plane_prepare_fb(struct drm_plane *plane, + struct drm_plane_state *state); +void armada_drm_plane_cleanup_fb(struct drm_plane *plane, + struct drm_plane_state *old_state); +int armada_drm_plane_atomic_check(struct drm_plane *plane, + struct drm_plane_state *state); +int armada_drm_primary_plane_init(struct drm_device *drm, + struct drm_plane *primary); + +#endif diff --git a/drivers/gpu/drm/bridge/sil-sii8620.c b/drivers/gpu/drm/bridge/sil-sii8620.c index 250effa0e6b8..a6e8f4591e63 100644 --- a/drivers/gpu/drm/bridge/sil-sii8620.c +++ b/drivers/gpu/drm/bridge/sil-sii8620.c @@ -14,6 +14,7 @@ #include <drm/bridge/mhl.h> #include <drm/drm_crtc.h> #include <drm/drm_edid.h> +#include <drm/drm_encoder.h> #include <linux/clk.h> #include <linux/delay.h> @@ -72,9 +73,7 @@ struct sii8620 { struct regulator_bulk_data supplies[2]; struct mutex lock; /* context lock, protects fields below */ int error; - int pixel_clock; unsigned int use_packed_pixel:1; - int video_code; enum sii8620_mode mode; enum sii8620_sink_type sink_type; u8 cbus_status; @@ -82,7 +81,6 @@ struct sii8620 { u8 xstat[MHL_XDS_SIZE]; u8 devcap[MHL_DCAP_SIZE]; u8 xdevcap[MHL_XDC_SIZE]; - u8 avif[HDMI_INFOFRAME_SIZE(AVI)]; bool feature_complete; bool devcap_read; bool sink_detected; @@ -1017,21 +1015,36 @@ static void sii8620_stop_video(struct sii8620 *ctx) static void sii8620_set_format(struct sii8620 *ctx) { + u8 out_fmt; + if (sii8620_is_mhl3(ctx)) { sii8620_setbits(ctx, REG_M3_P0CTRL, BIT_M3_P0CTRL_MHL3_P0_PIXEL_MODE_PACKED, ctx->use_packed_pixel ? ~0 : 0); } else { + if (ctx->use_packed_pixel) { + sii8620_write_seq_static(ctx, + REG_VID_MODE, BIT_VID_MODE_M1080P, + REG_MHL_TOP_CTL, BIT_MHL_TOP_CTL_MHL_PP_SEL | 1, + REG_MHLTX_CTL6, 0x60 + ); + } else { sii8620_write_seq_static(ctx, REG_VID_MODE, 0, REG_MHL_TOP_CTL, 1, REG_MHLTX_CTL6, 0xa0 ); + } } + if (ctx->use_packed_pixel) + out_fmt = VAL_TPI_FORMAT(YCBCR422, FULL); + else + out_fmt = VAL_TPI_FORMAT(RGB, FULL); + sii8620_write_seq(ctx, REG_TPI_INPUT, VAL_TPI_FORMAT(RGB, FULL), - REG_TPI_OUTPUT, VAL_TPI_FORMAT(RGB, FULL), + REG_TPI_OUTPUT, out_fmt, ); } @@ -1082,18 +1095,28 @@ static ssize_t mhl3_infoframe_pack(struct mhl3_infoframe *frame, return frm_len; } -static void sii8620_set_infoframes(struct sii8620 *ctx) +static void sii8620_set_infoframes(struct sii8620 *ctx, + struct drm_display_mode *mode) { struct mhl3_infoframe mhl_frm; union hdmi_infoframe frm; u8 buf[31]; int ret; + ret = drm_hdmi_avi_infoframe_from_display_mode(&frm.avi, + mode, + true); + if (ctx->use_packed_pixel) + frm.avi.colorspace = HDMI_COLORSPACE_YUV422; + + if (!ret) + ret = hdmi_avi_infoframe_pack(&frm.avi, buf, ARRAY_SIZE(buf)); + if (ret > 0) + sii8620_write_buf(ctx, REG_TPI_AVI_CHSUM, buf + 3, ret - 3); + if (!sii8620_is_mhl3(ctx) || !ctx->use_packed_pixel) { sii8620_write(ctx, REG_TPI_SC, BIT_TPI_SC_TPI_OUTPUT_MODE_0_HDMI); - sii8620_write_buf(ctx, REG_TPI_AVI_CHSUM, ctx->avif + 3, - ARRAY_SIZE(ctx->avif) - 3); sii8620_write(ctx, REG_PKT_FILTER_0, BIT_PKT_FILTER_0_DROP_CEA_GAMUT_PKT | BIT_PKT_FILTER_0_DROP_MPEG_PKT | @@ -1102,16 +1125,6 @@ static void sii8620_set_infoframes(struct sii8620 *ctx) return; } - ret = hdmi_avi_infoframe_init(&frm.avi); - frm.avi.colorspace = HDMI_COLORSPACE_YUV422; - frm.avi.active_aspect = HDMI_ACTIVE_ASPECT_PICTURE; - frm.avi.picture_aspect = HDMI_PICTURE_ASPECT_16_9; - frm.avi.colorimetry = HDMI_COLORIMETRY_ITU_709; - frm.avi.video_code = ctx->video_code; - if (!ret) - ret = hdmi_avi_infoframe_pack(&frm.avi, buf, ARRAY_SIZE(buf)); - if (ret > 0) - sii8620_write_buf(ctx, REG_TPI_AVI_CHSUM, buf + 3, ret - 3); sii8620_write(ctx, REG_PKT_FILTER_0, BIT_PKT_FILTER_0_DROP_CEA_GAMUT_PKT | BIT_PKT_FILTER_0_DROP_MPEG_PKT | @@ -1131,6 +1144,9 @@ static void sii8620_set_infoframes(struct sii8620 *ctx) static void sii8620_start_video(struct sii8620 *ctx) { + struct drm_display_mode *mode = + &ctx->bridge.encoder->crtc->state->adjusted_mode; + if (!sii8620_is_mhl3(ctx)) sii8620_stop_video(ctx); @@ -1149,8 +1165,14 @@ static void sii8620_start_video(struct sii8620 *ctx) sii8620_set_format(ctx); if (!sii8620_is_mhl3(ctx)) { - sii8620_mt_write_stat(ctx, MHL_DST_REG(LINK_MODE), - MHL_DST_LM_CLK_MODE_NORMAL | MHL_DST_LM_PATH_ENABLED); + u8 link_mode = MHL_DST_LM_PATH_ENABLED; + + if (ctx->use_packed_pixel) + link_mode |= MHL_DST_LM_CLK_MODE_PACKED_PIXEL; + else + link_mode |= MHL_DST_LM_CLK_MODE_NORMAL; + + sii8620_mt_write_stat(ctx, MHL_DST_REG(LINK_MODE), link_mode); sii8620_set_auto_zone(ctx); } else { static const struct { @@ -1167,7 +1189,7 @@ static void sii8620_start_video(struct sii8620 *ctx) MHL_XDS_LINK_RATE_6_0_GBPS, 0x40 }, }; u8 p0_ctrl = BIT_M3_P0CTRL_MHL3_P0_PORT_EN; - int clk = ctx->pixel_clock * (ctx->use_packed_pixel ? 2 : 3); + int clk = mode->clock * (ctx->use_packed_pixel ? 2 : 3); int i; for (i = 0; i < ARRAY_SIZE(clk_spec) - 1; ++i) @@ -1196,7 +1218,7 @@ static void sii8620_start_video(struct sii8620 *ctx) clk_spec[i].link_rate); } - sii8620_set_infoframes(ctx); + sii8620_set_infoframes(ctx, mode); } static void sii8620_disable_hpd(struct sii8620 *ctx) @@ -1661,14 +1683,18 @@ static void sii8620_status_dcap_ready(struct sii8620 *ctx) static void sii8620_status_changed_path(struct sii8620 *ctx) { - if (ctx->stat[MHL_DST_LINK_MODE] & MHL_DST_LM_PATH_ENABLED) { - sii8620_mt_write_stat(ctx, MHL_DST_REG(LINK_MODE), - MHL_DST_LM_CLK_MODE_NORMAL - | MHL_DST_LM_PATH_ENABLED); - } else { - sii8620_mt_write_stat(ctx, MHL_DST_REG(LINK_MODE), - MHL_DST_LM_CLK_MODE_NORMAL); - } + u8 link_mode; + + if (ctx->use_packed_pixel) + link_mode = MHL_DST_LM_CLK_MODE_PACKED_PIXEL; + else + link_mode = MHL_DST_LM_CLK_MODE_NORMAL; + + if (ctx->stat[MHL_DST_LINK_MODE] & MHL_DST_LM_PATH_ENABLED) + link_mode |= MHL_DST_LM_PATH_ENABLED; + + sii8620_mt_write_stat(ctx, MHL_DST_REG(LINK_MODE), + link_mode); } static void sii8620_msc_mr_write_stat(struct sii8620 *ctx) @@ -2242,8 +2268,6 @@ static bool sii8620_mode_fixup(struct drm_bridge *bridge, mutex_lock(&ctx->lock); ctx->use_packed_pixel = sii8620_is_packing_required(ctx, adjusted_mode); - ctx->video_code = drm_match_cea_mode(adjusted_mode); - ctx->pixel_clock = adjusted_mode->clock; mutex_unlock(&ctx->lock); diff --git a/drivers/gpu/drm/drm_lease.c b/drivers/gpu/drm/drm_lease.c index 50c73c0a20b9..d638c0fb3418 100644 --- a/drivers/gpu/drm/drm_lease.c +++ b/drivers/gpu/drm/drm_lease.c @@ -553,24 +553,13 @@ int drm_mode_create_lease_ioctl(struct drm_device *dev, /* Clone the lessor file to create a new file for us */ DRM_DEBUG_LEASE("Allocating lease file\n"); - path_get(&lessor_file->f_path); - lessee_file = alloc_file(&lessor_file->f_path, - lessor_file->f_mode, - fops_get(lessor_file->f_inode->i_fop)); - + lessee_file = filp_clone_open(lessor_file); if (IS_ERR(lessee_file)) { ret = PTR_ERR(lessee_file); goto out_lessee; } - /* Initialize the new file for DRM */ - DRM_DEBUG_LEASE("Initializing the file with %p\n", lessee_file->f_op->open); - ret = lessee_file->f_op->open(lessee_file->f_inode, lessee_file); - if (ret) - goto out_lessee_file; - lessee_priv = lessee_file->private_data; - /* Change the file to a master one */ drm_master_put(&lessee_priv->master); lessee_priv->master = lessee; @@ -588,9 +577,6 @@ int drm_mode_create_lease_ioctl(struct drm_device *dev, DRM_DEBUG_LEASE("drm_mode_create_lease_ioctl succeeded\n"); return 0; -out_lessee_file: - fput(lessee_file); - out_lessee: drm_master_put(&lessee); diff --git a/drivers/gpu/drm/drm_mipi_dsi.c b/drivers/gpu/drm/drm_mipi_dsi.c index bc73b7f5b9fc..80b75501f5c6 100644 --- a/drivers/gpu/drm/drm_mipi_dsi.c +++ b/drivers/gpu/drm/drm_mipi_dsi.c @@ -392,6 +392,7 @@ bool mipi_dsi_packet_format_is_short(u8 type) case MIPI_DSI_DCS_SHORT_WRITE: case MIPI_DSI_DCS_SHORT_WRITE_PARAM: case MIPI_DSI_DCS_READ: + case MIPI_DSI_DCS_COMPRESSION_MODE: case MIPI_DSI_SET_MAXIMUM_RETURN_PACKET_SIZE: return true; } @@ -410,6 +411,7 @@ EXPORT_SYMBOL(mipi_dsi_packet_format_is_short); bool mipi_dsi_packet_format_is_long(u8 type) { switch (type) { + case MIPI_DSI_PPS_LONG_WRITE: case MIPI_DSI_NULL_PACKET: case MIPI_DSI_BLANKING_PACKET: case MIPI_DSI_GENERIC_LONG_WRITE: diff --git a/drivers/gpu/drm/drm_print.c b/drivers/gpu/drm/drm_print.c index b25f98f33f6c..0e7fc3e7dfb4 100644 --- a/drivers/gpu/drm/drm_print.c +++ b/drivers/gpu/drm/drm_print.c @@ -30,6 +30,100 @@ #include <drm/drmP.h> #include <drm/drm_print.h> +void __drm_puts_coredump(struct drm_printer *p, const char *str) +{ + struct drm_print_iterator *iterator = p->arg; + ssize_t len; + + if (!iterator->remain) + return; + + if (iterator->offset < iterator->start) { + ssize_t copy; + + len = strlen(str); + + if (iterator->offset + len <= iterator->start) { + iterator->offset += len; + return; + } + + copy = len - (iterator->start - iterator->offset); + + if (copy > iterator->remain) + copy = iterator->remain; + + /* Copy out the bit of the string that we need */ + memcpy(iterator->data, + str + (iterator->start - iterator->offset), copy); + + iterator->offset = iterator->start + copy; + iterator->remain -= copy; + } else { + ssize_t pos = iterator->offset - iterator->start; + + len = min_t(ssize_t, strlen(str), iterator->remain); + + memcpy(iterator->data + pos, str, len); + + iterator->offset += len; + iterator->remain -= len; + } +} +EXPORT_SYMBOL(__drm_puts_coredump); + +void __drm_printfn_coredump(struct drm_printer *p, struct va_format *vaf) +{ + struct drm_print_iterator *iterator = p->arg; + size_t len; + char *buf; + + if (!iterator->remain) + return; + + /* Figure out how big the string will be */ + len = snprintf(NULL, 0, "%pV", vaf); + + /* This is the easiest path, we've already advanced beyond the offset */ + if (iterator->offset + len <= iterator->start) { + iterator->offset += len; + return; + } + + /* Then check if we can directly copy into the target buffer */ + if ((iterator->offset >= iterator->start) && (len < iterator->remain)) { + ssize_t pos = iterator->offset - iterator->start; + + snprintf(((char *) iterator->data) + pos, + iterator->remain, "%pV", vaf); + + iterator->offset += len; + iterator->remain -= len; + + return; + } + + /* + * Finally, hit the slow path and make a temporary string to copy over + * using _drm_puts_coredump + */ + buf = kmalloc(len + 1, GFP_KERNEL | __GFP_NOWARN | __GFP_NORETRY); + if (!buf) + return; + + snprintf(buf, len + 1, "%pV", vaf); + __drm_puts_coredump(p, (const char *) buf); + + kfree(buf); +} +EXPORT_SYMBOL(__drm_printfn_coredump); + +void __drm_puts_seq_file(struct drm_printer *p, const char *str) +{ + seq_puts(p->arg, str); +} +EXPORT_SYMBOL(__drm_puts_seq_file); + void __drm_printfn_seq_file(struct drm_printer *p, struct va_format *vaf) { seq_printf(p->arg, "%pV", vaf); @@ -49,6 +143,23 @@ void __drm_printfn_debug(struct drm_printer *p, struct va_format *vaf) EXPORT_SYMBOL(__drm_printfn_debug); /** + * drm_puts - print a const string to a &drm_printer stream + * @p: the &drm printer + * @str: const string + * + * Allow &drm_printer types that have a constant string + * option to use it. + */ +void drm_puts(struct drm_printer *p, const char *str) +{ + if (p->puts) + p->puts(p, str); + else + drm_printf(p, "%s", str); +} +EXPORT_SYMBOL(drm_puts); + +/** * drm_printf - print to a &drm_printer stream * @p: the &drm_printer * @f: format string diff --git a/drivers/gpu/drm/drm_property.c b/drivers/gpu/drm/drm_property.c index 1f8031e30f53..cdb10f885a4f 100644 --- a/drivers/gpu/drm/drm_property.c +++ b/drivers/gpu/drm/drm_property.c @@ -532,7 +532,7 @@ static void drm_property_free_blob(struct kref *kref) drm_mode_object_unregister(blob->dev, &blob->base); - kfree(blob); + kvfree(blob); } /** @@ -559,7 +559,7 @@ drm_property_create_blob(struct drm_device *dev, size_t length, if (!length || length > ULONG_MAX - sizeof(struct drm_property_blob)) return ERR_PTR(-EINVAL); - blob = kzalloc(sizeof(struct drm_property_blob)+length, GFP_KERNEL); + blob = kvzalloc(sizeof(struct drm_property_blob)+length, GFP_KERNEL); if (!blob) return ERR_PTR(-ENOMEM); @@ -576,7 +576,7 @@ drm_property_create_blob(struct drm_device *dev, size_t length, ret = __drm_mode_object_add(dev, &blob->base, DRM_MODE_OBJECT_BLOB, true, drm_property_free_blob); if (ret) { - kfree(blob); + kvfree(blob); return ERR_PTR(-EINVAL); } diff --git a/drivers/gpu/drm/etnaviv/etnaviv_drv.c b/drivers/gpu/drm/etnaviv/etnaviv_drv.c index 207532c05eb8..9b2720b41571 100644 --- a/drivers/gpu/drm/etnaviv/etnaviv_drv.c +++ b/drivers/gpu/drm/etnaviv/etnaviv_drv.c @@ -630,8 +630,11 @@ static struct platform_driver etnaviv_platform_driver = { }, }; +static struct platform_device *etnaviv_drm; + static int __init etnaviv_init(void) { + struct platform_device *pdev; int ret; struct device_node *np; @@ -643,7 +646,7 @@ static int __init etnaviv_init(void) ret = platform_driver_register(&etnaviv_platform_driver); if (ret != 0) - platform_driver_unregister(&etnaviv_gpu_driver); + goto unregister_gpu_driver; /* * If the DT contains at least one available GPU device, instantiate @@ -652,20 +655,33 @@ static int __init etnaviv_init(void) for_each_compatible_node(np, NULL, "vivante,gc") { if (!of_device_is_available(np)) continue; - - platform_device_register_simple("etnaviv", -1, NULL, 0); + pdev = platform_device_register_simple("etnaviv", -1, + NULL, 0); + if (IS_ERR(pdev)) { + ret = PTR_ERR(pdev); + of_node_put(np); + goto unregister_platform_driver; + } + etnaviv_drm = pdev; of_node_put(np); break; } + return 0; + +unregister_platform_driver: + platform_driver_unregister(&etnaviv_platform_driver); +unregister_gpu_driver: + platform_driver_unregister(&etnaviv_gpu_driver); return ret; } module_init(etnaviv_init); static void __exit etnaviv_exit(void) { - platform_driver_unregister(&etnaviv_gpu_driver); + platform_device_unregister(etnaviv_drm); platform_driver_unregister(&etnaviv_platform_driver); + platform_driver_unregister(&etnaviv_gpu_driver); } module_exit(etnaviv_exit); diff --git a/drivers/gpu/drm/etnaviv/etnaviv_drv.h b/drivers/gpu/drm/etnaviv/etnaviv_drv.h index d36c7bbe66db..8d02d1b7dcf5 100644 --- a/drivers/gpu/drm/etnaviv/etnaviv_drv.h +++ b/drivers/gpu/drm/etnaviv/etnaviv_drv.h @@ -18,6 +18,7 @@ #include <linux/time64.h> #include <linux/types.h> #include <linux/sizes.h> +#include <linux/mm_types.h> #include <drm/drmP.h> #include <drm/drm_crtc_helper.h> @@ -53,7 +54,7 @@ int etnaviv_ioctl_gem_submit(struct drm_device *dev, void *data, struct drm_file *file); int etnaviv_gem_mmap(struct file *filp, struct vm_area_struct *vma); -int etnaviv_gem_fault(struct vm_fault *vmf); +vm_fault_t etnaviv_gem_fault(struct vm_fault *vmf); int etnaviv_gem_mmap_offset(struct drm_gem_object *obj, u64 *offset); struct sg_table *etnaviv_gem_prime_get_sg_table(struct drm_gem_object *obj); void *etnaviv_gem_prime_vmap(struct drm_gem_object *obj); diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gem.c b/drivers/gpu/drm/etnaviv/etnaviv_gem.c index 209ef1274b80..1fa74226db91 100644 --- a/drivers/gpu/drm/etnaviv/etnaviv_gem.c +++ b/drivers/gpu/drm/etnaviv/etnaviv_gem.c @@ -169,31 +169,30 @@ int etnaviv_gem_mmap(struct file *filp, struct vm_area_struct *vma) return obj->ops->mmap(obj, vma); } -int etnaviv_gem_fault(struct vm_fault *vmf) +vm_fault_t etnaviv_gem_fault(struct vm_fault *vmf) { struct vm_area_struct *vma = vmf->vma; struct drm_gem_object *obj = vma->vm_private_data; struct etnaviv_gem_object *etnaviv_obj = to_etnaviv_bo(obj); struct page **pages, *page; pgoff_t pgoff; - int ret; + int err; /* * Make sure we don't parallel update on a fault, nor move or remove - * something from beneath our feet. Note that vm_insert_page() is + * something from beneath our feet. Note that vmf_insert_page() is * specifically coded to take care of this, so we don't have to. */ - ret = mutex_lock_interruptible(&etnaviv_obj->lock); - if (ret) - goto out; - + err = mutex_lock_interruptible(&etnaviv_obj->lock); + if (err) + return VM_FAULT_NOPAGE; /* make sure we have pages attached now */ pages = etnaviv_gem_get_pages(etnaviv_obj); mutex_unlock(&etnaviv_obj->lock); if (IS_ERR(pages)) { - ret = PTR_ERR(pages); - goto out; + err = PTR_ERR(pages); + return vmf_error(err); } /* We don't use vmf->pgoff since that has the fake offset: */ @@ -204,25 +203,7 @@ int etnaviv_gem_fault(struct vm_fault *vmf) VERB("Inserting %p pfn %lx, pa %lx", (void *)vmf->address, page_to_pfn(page), page_to_pfn(page) << PAGE_SHIFT); - ret = vm_insert_page(vma, vmf->address, page); - -out: - switch (ret) { - case -EAGAIN: - case 0: - case -ERESTARTSYS: - case -EINTR: - case -EBUSY: - /* - * EBUSY is ok: this just means that another thread - * already did the job. - */ - return VM_FAULT_NOPAGE; - case -ENOMEM: - return VM_FAULT_OOM; - default: - return VM_FAULT_SIGBUS; - } + return vmf_insert_page(vma, vmf->address, page); } int etnaviv_gem_mmap_offset(struct drm_gem_object *obj, u64 *offset) diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gem_submit.c b/drivers/gpu/drm/etnaviv/etnaviv_gem_submit.c index 46ecd3e66ac9..983e67f19e45 100644 --- a/drivers/gpu/drm/etnaviv/etnaviv_gem_submit.c +++ b/drivers/gpu/drm/etnaviv/etnaviv_gem_submit.c @@ -388,9 +388,9 @@ static void submit_cleanup(struct kref *kref) dma_fence_put(submit->in_fence); if (submit->out_fence) { /* first remove from IDR, so fence can not be found anymore */ - mutex_lock(&submit->gpu->fence_idr_lock); + mutex_lock(&submit->gpu->fence_lock); idr_remove(&submit->gpu->fence_idr, submit->out_fence_id); - mutex_unlock(&submit->gpu->fence_idr_lock); + mutex_unlock(&submit->gpu->fence_lock); dma_fence_put(submit->out_fence); } kfree(submit->pmrs); diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gpu.c b/drivers/gpu/drm/etnaviv/etnaviv_gpu.c index 19b09a59e30e..f225fbc6edd2 100644 --- a/drivers/gpu/drm/etnaviv/etnaviv_gpu.c +++ b/drivers/gpu/drm/etnaviv/etnaviv_gpu.c @@ -799,6 +799,7 @@ int etnaviv_gpu_init(struct etnaviv_gpu *gpu) free_buffer: etnaviv_cmdbuf_free(&gpu->buffer); + gpu->buffer.suballoc = NULL; destroy_iommu: etnaviv_iommu_destroy(gpu->mmu); gpu->mmu = NULL; @@ -1726,7 +1727,7 @@ static int etnaviv_gpu_platform_probe(struct platform_device *pdev) gpu->dev = &pdev->dev; mutex_init(&gpu->lock); - mutex_init(&gpu->fence_idr_lock); + mutex_init(&gpu->fence_lock); /* Map registers: */ res = platform_get_resource(pdev, IORESOURCE_MEM, 0); diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gpu.h b/drivers/gpu/drm/etnaviv/etnaviv_gpu.h index dd430f0f8ff5..9a75a6937268 100644 --- a/drivers/gpu/drm/etnaviv/etnaviv_gpu.h +++ b/drivers/gpu/drm/etnaviv/etnaviv_gpu.h @@ -118,7 +118,7 @@ struct etnaviv_gpu { u32 idle_mask; /* Fencing support */ - struct mutex fence_idr_lock; + struct mutex fence_lock; struct idr fence_idr; u32 next_fence; u32 active_fence; @@ -131,6 +131,9 @@ struct etnaviv_gpu { struct work_struct sync_point_work; int sync_point_event; + /* hang detection */ + u32 hangcheck_dma_addr; + void __iomem *mmio; int irq; diff --git a/drivers/gpu/drm/etnaviv/etnaviv_iommu_v2.c b/drivers/gpu/drm/etnaviv/etnaviv_iommu_v2.c index 71fbc1f96cb6..f1c88d8ad5ba 100644 --- a/drivers/gpu/drm/etnaviv/etnaviv_iommu_v2.c +++ b/drivers/gpu/drm/etnaviv/etnaviv_iommu_v2.c @@ -119,8 +119,7 @@ static size_t etnaviv_iommuv2_unmap(struct etnaviv_iommu_domain *domain, static int etnaviv_iommuv2_init(struct etnaviv_iommuv2_domain *etnaviv_domain) { - u32 *p; - int ret, i; + int ret; /* allocate scratch page */ etnaviv_domain->base.bad_page_cpu = @@ -131,9 +130,9 @@ static int etnaviv_iommuv2_init(struct etnaviv_iommuv2_domain *etnaviv_domain) ret = -ENOMEM; goto fail_mem; } - p = etnaviv_domain->base.bad_page_cpu; - for (i = 0; i < SZ_4K / 4; i++) - *p++ = 0xdead55aa; + + memset32(etnaviv_domain->base.bad_page_cpu, 0xdead55aa, + SZ_4K / sizeof(u32)); etnaviv_domain->pta_cpu = dma_alloc_wc(etnaviv_domain->base.dev, SZ_4K, &etnaviv_domain->pta_dma, diff --git a/drivers/gpu/drm/etnaviv/etnaviv_sched.c b/drivers/gpu/drm/etnaviv/etnaviv_sched.c index 590e44b0d963..69e9b431bf1f 100644 --- a/drivers/gpu/drm/etnaviv/etnaviv_sched.c +++ b/drivers/gpu/drm/etnaviv/etnaviv_sched.c @@ -10,6 +10,7 @@ #include "etnaviv_gem.h" #include "etnaviv_gpu.h" #include "etnaviv_sched.h" +#include "state.xml.h" static int etnaviv_job_hang_limit = 0; module_param_named(job_hang_limit, etnaviv_job_hang_limit, int , 0444); @@ -85,6 +86,29 @@ static void etnaviv_sched_timedout_job(struct drm_sched_job *sched_job) { struct etnaviv_gem_submit *submit = to_etnaviv_submit(sched_job); struct etnaviv_gpu *gpu = submit->gpu; + u32 dma_addr; + int change; + + /* + * If the GPU managed to complete this jobs fence, the timout is + * spurious. Bail out. + */ + if (fence_completed(gpu, submit->out_fence->seqno)) + return; + + /* + * If the GPU is still making forward progress on the front-end (which + * should never loop) we shift out the timeout to give it a chance to + * finish the job. + */ + dma_addr = gpu_read(gpu, VIVS_FE_DMA_ADDRESS); + change = dma_addr - gpu->hangcheck_dma_addr; + if (change < 0 || change > 16) { + gpu->hangcheck_dma_addr = dma_addr; + schedule_delayed_work(&sched_job->work_tdr, + sched_job->sched->timeout); + return; + } /* block scheduler */ kthread_park(gpu->sched.thread); @@ -116,28 +140,38 @@ static const struct drm_sched_backend_ops etnaviv_sched_ops = { int etnaviv_sched_push_job(struct drm_sched_entity *sched_entity, struct etnaviv_gem_submit *submit) { - int ret; + int ret = 0; + + /* + * Hold the fence lock across the whole operation to avoid jobs being + * pushed out of order with regard to their sched fence seqnos as + * allocated in drm_sched_job_init. + */ + mutex_lock(&submit->gpu->fence_lock); ret = drm_sched_job_init(&submit->sched_job, sched_entity, submit->cmdbuf.ctx); if (ret) - return ret; + goto out_unlock; submit->out_fence = dma_fence_get(&submit->sched_job.s_fence->finished); - mutex_lock(&submit->gpu->fence_idr_lock); submit->out_fence_id = idr_alloc_cyclic(&submit->gpu->fence_idr, submit->out_fence, 0, INT_MAX, GFP_KERNEL); - mutex_unlock(&submit->gpu->fence_idr_lock); - if (submit->out_fence_id < 0) - return -ENOMEM; + if (submit->out_fence_id < 0) { + ret = -ENOMEM; + goto out_unlock; + } /* the scheduler holds on to the job now */ kref_get(&submit->refcount); drm_sched_entity_push_job(&submit->sched_job, sched_entity); - return 0; +out_unlock: + mutex_unlock(&submit->gpu->fence_lock); + + return ret; } int etnaviv_sched_init(struct etnaviv_gpu *gpu) diff --git a/drivers/gpu/drm/i2c/tda998x_drv.c b/drivers/gpu/drm/i2c/tda998x_drv.c index eecdc327b9f8..a7c39f39793f 100644 --- a/drivers/gpu/drm/i2c/tda998x_drv.c +++ b/drivers/gpu/drm/i2c/tda998x_drv.c @@ -69,6 +69,7 @@ struct tda998x_priv { bool edid_delay_active; struct drm_encoder encoder; + struct drm_bridge bridge; struct drm_connector connector; struct tda998x_audio_port audio_port[2]; @@ -79,9 +80,10 @@ struct tda998x_priv { #define conn_to_tda998x_priv(x) \ container_of(x, struct tda998x_priv, connector) - #define enc_to_tda998x_priv(x) \ container_of(x, struct tda998x_priv, encoder) +#define bridge_to_tda998x_priv(x) \ + container_of(x, struct tda998x_priv, bridge) /* The TDA9988 series of devices use a paged register scheme.. to simplify * things we encode the page # in upper bits of the register #. To read/ @@ -762,7 +764,7 @@ static void tda998x_detect_work(struct work_struct *work) { struct tda998x_priv *priv = container_of(work, struct tda998x_priv, detect_work); - struct drm_device *dev = priv->encoder.dev; + struct drm_device *dev = priv->connector.dev; if (dev) drm_kms_helper_hotplug_event(dev); @@ -1104,29 +1106,6 @@ static int tda998x_audio_codec_init(struct tda998x_priv *priv, /* DRM connector functions */ -static int tda998x_connector_fill_modes(struct drm_connector *connector, - uint32_t maxX, uint32_t maxY) -{ - struct tda998x_priv *priv = conn_to_tda998x_priv(connector); - int ret; - - mutex_lock(&priv->audio_mutex); - ret = drm_helper_probe_single_connector_modes(connector, maxX, maxY); - - if (connector->edid_blob_ptr) { - struct edid *edid = (void *)connector->edid_blob_ptr->data; - - cec_notifier_set_phys_addr_from_edid(priv->cec_notify, edid); - - priv->sink_has_audio = drm_detect_monitor_audio(edid); - } else { - priv->sink_has_audio = false; - } - mutex_unlock(&priv->audio_mutex); - - return ret; -} - static enum drm_connector_status tda998x_connector_detect(struct drm_connector *connector, bool force) { @@ -1145,7 +1124,7 @@ static void tda998x_connector_destroy(struct drm_connector *connector) static const struct drm_connector_funcs tda998x_connector_funcs = { .dpms = drm_helper_connector_dpms, .reset = drm_atomic_helper_connector_reset, - .fill_modes = tda998x_connector_fill_modes, + .fill_modes = drm_helper_probe_single_connector_modes, .detect = tda998x_connector_detect, .destroy = tda998x_connector_destroy, .atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state, @@ -1244,40 +1223,29 @@ static int tda998x_connector_get_modes(struct drm_connector *connector) } drm_connector_update_edid_property(connector, edid); + cec_notifier_set_phys_addr_from_edid(priv->cec_notify, edid); + + mutex_lock(&priv->audio_mutex); n = drm_add_edid_modes(connector, edid); + priv->sink_has_audio = drm_detect_monitor_audio(edid); + mutex_unlock(&priv->audio_mutex); kfree(edid); return n; } -static enum drm_mode_status tda998x_connector_mode_valid(struct drm_connector *connector, - struct drm_display_mode *mode) -{ - /* TDA19988 dotclock can go up to 165MHz */ - struct tda998x_priv *priv = conn_to_tda998x_priv(connector); - - if (mode->clock > ((priv->rev == TDA19988) ? 165000 : 150000)) - return MODE_CLOCK_HIGH; - if (mode->htotal >= BIT(13)) - return MODE_BAD_HVALUE; - if (mode->vtotal >= BIT(11)) - return MODE_BAD_VVALUE; - return MODE_OK; -} - static struct drm_encoder * tda998x_connector_best_encoder(struct drm_connector *connector) { struct tda998x_priv *priv = conn_to_tda998x_priv(connector); - return &priv->encoder; + return priv->bridge.encoder; } static const struct drm_connector_helper_funcs tda998x_connector_helper_funcs = { .get_modes = tda998x_connector_get_modes, - .mode_valid = tda998x_connector_mode_valid, .best_encoder = tda998x_connector_best_encoder, }; @@ -1301,25 +1269,48 @@ static int tda998x_connector_init(struct tda998x_priv *priv, if (ret) return ret; - drm_connector_attach_encoder(&priv->connector, &priv->encoder); + drm_connector_attach_encoder(&priv->connector, + priv->bridge.encoder); return 0; } -/* DRM encoder functions */ +/* DRM bridge functions */ -static void tda998x_encoder_dpms(struct drm_encoder *encoder, int mode) +static int tda998x_bridge_attach(struct drm_bridge *bridge) { - struct tda998x_priv *priv = enc_to_tda998x_priv(encoder); - bool on; + struct tda998x_priv *priv = bridge_to_tda998x_priv(bridge); - /* we only care about on or off: */ - on = mode == DRM_MODE_DPMS_ON; + return tda998x_connector_init(priv, bridge->dev); +} - if (on == priv->is_on) - return; +static void tda998x_bridge_detach(struct drm_bridge *bridge) +{ + struct tda998x_priv *priv = bridge_to_tda998x_priv(bridge); - if (on) { + drm_connector_cleanup(&priv->connector); +} + +static enum drm_mode_status tda998x_bridge_mode_valid(struct drm_bridge *bridge, + const struct drm_display_mode *mode) +{ + /* TDA19988 dotclock can go up to 165MHz */ + struct tda998x_priv *priv = bridge_to_tda998x_priv(bridge); + + if (mode->clock > ((priv->rev == TDA19988) ? 165000 : 150000)) + return MODE_CLOCK_HIGH; + if (mode->htotal >= BIT(13)) + return MODE_BAD_HVALUE; + if (mode->vtotal >= BIT(11)) + return MODE_BAD_VVALUE; + return MODE_OK; +} + +static void tda998x_bridge_enable(struct drm_bridge *bridge) +{ + struct tda998x_priv *priv = bridge_to_tda998x_priv(bridge); + + if (!priv->is_on) { /* enable video ports, audio will be enabled later */ reg_write(priv, REG_ENA_VP_0, 0xff); reg_write(priv, REG_ENA_VP_1, 0xff); @@ -1330,7 +1321,14 @@ static void tda998x_encoder_dpms(struct drm_encoder *encoder, int mode) reg_write(priv, REG_VIP_CNTRL_2, priv->vip_cntrl_2); priv->is_on = true; - } else { + } +} + +static void tda998x_bridge_disable(struct drm_bridge *bridge) +{ + struct tda998x_priv *priv = bridge_to_tda998x_priv(bridge); + + if (priv->is_on) { /* disable video ports */ reg_write(priv, REG_ENA_VP_0, 0x00); reg_write(priv, REG_ENA_VP_1, 0x00); @@ -1340,12 +1338,12 @@ static void tda998x_encoder_dpms(struct drm_encoder *encoder, int mode) } } -static void -tda998x_encoder_mode_set(struct drm_encoder *encoder, - struct drm_display_mode *mode, - struct drm_display_mode *adjusted_mode) +static void tda998x_bridge_mode_set(struct drm_bridge *bridge, + struct drm_display_mode *mode, + struct drm_display_mode *adjusted_mode) { - struct tda998x_priv *priv = enc_to_tda998x_priv(encoder); + struct tda998x_priv *priv = bridge_to_tda998x_priv(bridge); + unsigned long tmds_clock; u16 ref_pix, ref_line, n_pix, n_line; u16 hs_pix_s, hs_pix_e; u16 vs1_pix_s, vs1_pix_e, vs1_line_s, vs1_line_e; @@ -1416,12 +1414,19 @@ tda998x_encoder_mode_set(struct drm_encoder *encoder, (mode->vsync_end - mode->vsync_start)/2; } - div = 148500 / mode->clock; - if (div != 0) { - div--; - if (div > 3) - div = 3; - } + tmds_clock = mode->clock; + + /* + * The divisor is power-of-2. The TDA9983B datasheet gives + * this as ranges of Msample/s, which is 10x the TMDS clock: + * 0 - 800 to 1500 Msample/s + * 1 - 400 to 800 Msample/s + * 2 - 200 to 400 Msample/s + * 3 - as 2 above + */ + for (div = 0; div < 3; div++) + if (80000 >> div <= tmds_clock) + break; mutex_lock(&priv->audio_mutex); @@ -1552,26 +1557,14 @@ tda998x_encoder_mode_set(struct drm_encoder *encoder, mutex_unlock(&priv->audio_mutex); } -static void tda998x_destroy(struct tda998x_priv *priv) -{ - /* disable all IRQs and free the IRQ handler */ - cec_write(priv, REG_CEC_RXSHPDINTENA, 0); - reg_clear(priv, REG_INT_FLAGS_2, INT_FLAGS_2_EDID_BLK_RD); - - if (priv->audio_pdev) - platform_device_unregister(priv->audio_pdev); - - if (priv->hdmi->irq) - free_irq(priv->hdmi->irq, priv); - - del_timer_sync(&priv->edid_delay_timer); - cancel_work_sync(&priv->detect_work); - - i2c_unregister_device(priv->cec); - - if (priv->cec_notify) - cec_notifier_put(priv->cec_notify); -} +static const struct drm_bridge_funcs tda998x_bridge_funcs = { + .attach = tda998x_bridge_attach, + .detach = tda998x_bridge_detach, + .mode_valid = tda998x_bridge_mode_valid, + .disable = tda998x_bridge_disable, + .mode_set = tda998x_bridge_mode_set, + .enable = tda998x_bridge_enable, +}; /* I2C driver functions */ @@ -1617,16 +1610,69 @@ static int tda998x_get_audio_ports(struct tda998x_priv *priv, return 0; } -static int tda998x_create(struct i2c_client *client, struct tda998x_priv *priv) +static void tda998x_set_config(struct tda998x_priv *priv, + const struct tda998x_encoder_params *p) { + priv->vip_cntrl_0 = VIP_CNTRL_0_SWAP_A(p->swap_a) | + (p->mirr_a ? VIP_CNTRL_0_MIRR_A : 0) | + VIP_CNTRL_0_SWAP_B(p->swap_b) | + (p->mirr_b ? VIP_CNTRL_0_MIRR_B : 0); + priv->vip_cntrl_1 = VIP_CNTRL_1_SWAP_C(p->swap_c) | + (p->mirr_c ? VIP_CNTRL_1_MIRR_C : 0) | + VIP_CNTRL_1_SWAP_D(p->swap_d) | + (p->mirr_d ? VIP_CNTRL_1_MIRR_D : 0); + priv->vip_cntrl_2 = VIP_CNTRL_2_SWAP_E(p->swap_e) | + (p->mirr_e ? VIP_CNTRL_2_MIRR_E : 0) | + VIP_CNTRL_2_SWAP_F(p->swap_f) | + (p->mirr_f ? VIP_CNTRL_2_MIRR_F : 0); + + priv->audio_params = p->audio_params; +} + +static void tda998x_destroy(struct device *dev) +{ + struct tda998x_priv *priv = dev_get_drvdata(dev); + + drm_bridge_remove(&priv->bridge); + + /* disable all IRQs and free the IRQ handler */ + cec_write(priv, REG_CEC_RXSHPDINTENA, 0); + reg_clear(priv, REG_INT_FLAGS_2, INT_FLAGS_2_EDID_BLK_RD); + + if (priv->audio_pdev) + platform_device_unregister(priv->audio_pdev); + + if (priv->hdmi->irq) + free_irq(priv->hdmi->irq, priv); + + del_timer_sync(&priv->edid_delay_timer); + cancel_work_sync(&priv->detect_work); + + i2c_unregister_device(priv->cec); + + if (priv->cec_notify) + cec_notifier_put(priv->cec_notify); +} + +static int tda998x_create(struct device *dev) +{ + struct i2c_client *client = to_i2c_client(dev); struct device_node *np = client->dev.of_node; struct i2c_board_info cec_info; + struct tda998x_priv *priv; u32 video; int rev_lo, rev_hi, ret; + priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL); + if (!priv) + return -ENOMEM; + + dev_set_drvdata(dev, priv); + mutex_init(&priv->mutex); /* protect the page access */ mutex_init(&priv->audio_mutex); /* protect access from audio thread */ mutex_init(&priv->edid_mutex); + INIT_LIST_HEAD(&priv->bridge.list); init_waitqueue_head(&priv->edid_delay_waitq); timer_setup(&priv->edid_delay_timer, tda998x_edid_delay_done, 0); INIT_WORK(&priv->detect_work, tda998x_detect_work); @@ -1649,13 +1695,13 @@ static int tda998x_create(struct i2c_client *client, struct tda998x_priv *priv) /* read version: */ rev_lo = reg_read(priv, REG_VERSION_LSB); if (rev_lo < 0) { - dev_err(&client->dev, "failed to read version: %d\n", rev_lo); + dev_err(dev, "failed to read version: %d\n", rev_lo); return rev_lo; } rev_hi = reg_read(priv, REG_VERSION_MSB); if (rev_hi < 0) { - dev_err(&client->dev, "failed to read version: %d\n", rev_hi); + dev_err(dev, "failed to read version: %d\n", rev_hi); return rev_hi; } @@ -1666,20 +1712,19 @@ static int tda998x_create(struct i2c_client *client, struct tda998x_priv *priv) switch (priv->rev) { case TDA9989N2: - dev_info(&client->dev, "found TDA9989 n2"); + dev_info(dev, "found TDA9989 n2"); break; case TDA19989: - dev_info(&client->dev, "found TDA19989"); + dev_info(dev, "found TDA19989"); break; case TDA19989N2: - dev_info(&client->dev, "found TDA19989 n2"); + dev_info(dev, "found TDA19989 n2"); break; case TDA19988: - dev_info(&client->dev, "found TDA19988"); + dev_info(dev, "found TDA19988"); break; default: - dev_err(&client->dev, "found unsupported device: %04x\n", - priv->rev); + dev_err(dev, "found unsupported device: %04x\n", priv->rev); return -ENXIO; } @@ -1722,8 +1767,7 @@ static int tda998x_create(struct i2c_client *client, struct tda998x_priv *priv) tda998x_irq_thread, irq_flags, "tda998x", priv); if (ret) { - dev_err(&client->dev, - "failed to request IRQ#%u: %d\n", + dev_err(dev, "failed to request IRQ#%u: %d\n", client->irq, ret); goto err_irq; } @@ -1732,13 +1776,13 @@ static int tda998x_create(struct i2c_client *client, struct tda998x_priv *priv) cec_write(priv, REG_CEC_RXSHPDINTENA, CEC_RXSHPDLEV_HPD); } - priv->cec_notify = cec_notifier_get(&client->dev); + priv->cec_notify = cec_notifier_get(dev); if (!priv->cec_notify) { ret = -ENOMEM; goto fail; } - priv->cec_glue.parent = &client->dev; + priv->cec_glue.parent = dev; priv->cec_glue.data = priv; priv->cec_glue.init = tda998x_cec_hook_init; priv->cec_glue.exit = tda998x_cec_hook_exit; @@ -1768,61 +1812,44 @@ static int tda998x_create(struct i2c_client *client, struct tda998x_priv *priv) /* enable EDID read irq: */ reg_set(priv, REG_INT_FLAGS_2, INT_FLAGS_2_EDID_BLK_RD); - if (!np) - return 0; /* non-DT */ + if (np) { + /* get the device tree parameters */ + ret = of_property_read_u32(np, "video-ports", &video); + if (ret == 0) { + priv->vip_cntrl_0 = video >> 16; + priv->vip_cntrl_1 = video >> 8; + priv->vip_cntrl_2 = video; + } + + ret = tda998x_get_audio_ports(priv, np); + if (ret) + goto fail; - /* get the device tree parameters */ - ret = of_property_read_u32(np, "video-ports", &video); - if (ret == 0) { - priv->vip_cntrl_0 = video >> 16; - priv->vip_cntrl_1 = video >> 8; - priv->vip_cntrl_2 = video; + if (priv->audio_port[0].format != AFMT_UNUSED) + tda998x_audio_codec_init(priv, &client->dev); + } else if (dev->platform_data) { + tda998x_set_config(priv, dev->platform_data); } - ret = tda998x_get_audio_ports(priv, np); - if (ret) - goto fail; + priv->bridge.funcs = &tda998x_bridge_funcs; +#ifdef CONFIG_OF + priv->bridge.of_node = dev->of_node; +#endif - if (priv->audio_port[0].format != AFMT_UNUSED) - tda998x_audio_codec_init(priv, &client->dev); + drm_bridge_add(&priv->bridge); return 0; fail: - /* if encoder_init fails, the encoder slave is never registered, - * so cleanup here: - */ - i2c_unregister_device(priv->cec); - if (priv->cec_notify) - cec_notifier_put(priv->cec_notify); - if (client->irq) - free_irq(client->irq, priv); + tda998x_destroy(dev); err_irq: return ret; } -static void tda998x_encoder_prepare(struct drm_encoder *encoder) -{ - tda998x_encoder_dpms(encoder, DRM_MODE_DPMS_OFF); -} - -static void tda998x_encoder_commit(struct drm_encoder *encoder) -{ - tda998x_encoder_dpms(encoder, DRM_MODE_DPMS_ON); -} - -static const struct drm_encoder_helper_funcs tda998x_encoder_helper_funcs = { - .dpms = tda998x_encoder_dpms, - .prepare = tda998x_encoder_prepare, - .commit = tda998x_encoder_commit, - .mode_set = tda998x_encoder_mode_set, -}; +/* DRM encoder functions */ static void tda998x_encoder_destroy(struct drm_encoder *encoder) { - struct tda998x_priv *priv = enc_to_tda998x_priv(encoder); - - tda998x_destroy(priv); drm_encoder_cleanup(encoder); } @@ -1830,40 +1857,12 @@ static const struct drm_encoder_funcs tda998x_encoder_funcs = { .destroy = tda998x_encoder_destroy, }; -static void tda998x_set_config(struct tda998x_priv *priv, - const struct tda998x_encoder_params *p) -{ - priv->vip_cntrl_0 = VIP_CNTRL_0_SWAP_A(p->swap_a) | - (p->mirr_a ? VIP_CNTRL_0_MIRR_A : 0) | - VIP_CNTRL_0_SWAP_B(p->swap_b) | - (p->mirr_b ? VIP_CNTRL_0_MIRR_B : 0); - priv->vip_cntrl_1 = VIP_CNTRL_1_SWAP_C(p->swap_c) | - (p->mirr_c ? VIP_CNTRL_1_MIRR_C : 0) | - VIP_CNTRL_1_SWAP_D(p->swap_d) | - (p->mirr_d ? VIP_CNTRL_1_MIRR_D : 0); - priv->vip_cntrl_2 = VIP_CNTRL_2_SWAP_E(p->swap_e) | - (p->mirr_e ? VIP_CNTRL_2_MIRR_E : 0) | - VIP_CNTRL_2_SWAP_F(p->swap_f) | - (p->mirr_f ? VIP_CNTRL_2_MIRR_F : 0); - - priv->audio_params = p->audio_params; -} - -static int tda998x_bind(struct device *dev, struct device *master, void *data) +static int tda998x_encoder_init(struct device *dev, struct drm_device *drm) { - struct tda998x_encoder_params *params = dev->platform_data; - struct i2c_client *client = to_i2c_client(dev); - struct drm_device *drm = data; - struct tda998x_priv *priv; + struct tda998x_priv *priv = dev_get_drvdata(dev); u32 crtcs = 0; int ret; - priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL); - if (!priv) - return -ENOMEM; - - dev_set_drvdata(dev, priv); - if (dev->of_node) crtcs = drm_of_find_possible_crtcs(drm, dev->of_node); @@ -1875,40 +1874,36 @@ static int tda998x_bind(struct device *dev, struct device *master, void *data) priv->encoder.possible_crtcs = crtcs; - ret = tda998x_create(client, priv); - if (ret) - return ret; - - if (!dev->of_node && params) - tda998x_set_config(priv, params); - - drm_encoder_helper_add(&priv->encoder, &tda998x_encoder_helper_funcs); ret = drm_encoder_init(drm, &priv->encoder, &tda998x_encoder_funcs, DRM_MODE_ENCODER_TMDS, NULL); if (ret) goto err_encoder; - ret = tda998x_connector_init(priv, drm); + ret = drm_bridge_attach(&priv->encoder, &priv->bridge, NULL); if (ret) - goto err_connector; + goto err_bridge; return 0; -err_connector: +err_bridge: drm_encoder_cleanup(&priv->encoder); err_encoder: - tda998x_destroy(priv); return ret; } +static int tda998x_bind(struct device *dev, struct device *master, void *data) +{ + struct drm_device *drm = data; + + return tda998x_encoder_init(dev, drm); +} + static void tda998x_unbind(struct device *dev, struct device *master, void *data) { struct tda998x_priv *priv = dev_get_drvdata(dev); - drm_connector_cleanup(&priv->connector); drm_encoder_cleanup(&priv->encoder); - tda998x_destroy(priv); } static const struct component_ops tda998x_ops = { @@ -1919,16 +1914,27 @@ static const struct component_ops tda998x_ops = { static int tda998x_probe(struct i2c_client *client, const struct i2c_device_id *id) { + int ret; + if (!i2c_check_functionality(client->adapter, I2C_FUNC_I2C)) { dev_warn(&client->dev, "adapter does not support I2C\n"); return -EIO; } - return component_add(&client->dev, &tda998x_ops); + + ret = tda998x_create(&client->dev); + if (ret) + return ret; + + ret = component_add(&client->dev, &tda998x_ops); + if (ret) + tda998x_destroy(&client->dev); + return ret; } static int tda998x_remove(struct i2c_client *client) { component_del(&client->dev, &tda998x_ops); + tda998x_destroy(&client->dev); return 0; } diff --git a/drivers/gpu/drm/i915/gvt/cmd_parser.c b/drivers/gpu/drm/i915/gvt/cmd_parser.c index 0651e63b25fb..45e89b1e0481 100644 --- a/drivers/gpu/drm/i915/gvt/cmd_parser.c +++ b/drivers/gpu/drm/i915/gvt/cmd_parser.c @@ -863,6 +863,7 @@ static int cmd_reg_handler(struct parser_exec_state *s, { struct intel_vgpu *vgpu = s->vgpu; struct intel_gvt *gvt = vgpu->gvt; + u32 ctx_sr_ctl; if (offset + 4 > gvt->device_info.mmio_size) { gvt_vgpu_err("%s access to (%x) outside of MMIO range\n", @@ -895,6 +896,28 @@ static int cmd_reg_handler(struct parser_exec_state *s, patch_value(s, cmd_ptr(s, index), VGT_PVINFO_PAGE); } + /* TODO + * Right now only scan LRI command on KBL and in inhibit context. + * It's good enough to support initializing mmio by lri command in + * vgpu inhibit context on KBL. + */ + if (IS_KABYLAKE(s->vgpu->gvt->dev_priv) && + intel_gvt_mmio_is_in_ctx(gvt, offset) && + !strncmp(cmd, "lri", 3)) { + intel_gvt_hypervisor_read_gpa(s->vgpu, + s->workload->ring_context_gpa + 12, &ctx_sr_ctl, 4); + /* check inhibit context */ + if (ctx_sr_ctl & 1) { + u32 data = cmd_val(s, index + 1); + + if (intel_gvt_mmio_has_mode_mask(s->vgpu->gvt, offset)) + intel_vgpu_mask_mmio_write(vgpu, + offset, &data, 4); + else + vgpu_vreg(vgpu, offset) = data; + } + } + /* TODO: Update the global mask if this MMIO is a masked-MMIO */ intel_gvt_mmio_set_cmd_accessed(gvt, offset); return 0; diff --git a/drivers/gpu/drm/i915/gvt/display.c b/drivers/gpu/drm/i915/gvt/display.c index 6ee50cb328f8..3019dbc39aef 100644 --- a/drivers/gpu/drm/i915/gvt/display.c +++ b/drivers/gpu/drm/i915/gvt/display.c @@ -219,7 +219,7 @@ static void emulate_monitor_status_change(struct intel_vgpu *vgpu) ~(TRANS_DDI_BPC_MASK | TRANS_DDI_MODE_SELECT_MASK | TRANS_DDI_PORT_MASK); vgpu_vreg_t(vgpu, TRANS_DDI_FUNC_CTL(TRANSCODER_A)) |= - (TRANS_DDI_BPC_8 | TRANS_DDI_MODE_SELECT_DP_SST | + (TRANS_DDI_BPC_8 | TRANS_DDI_MODE_SELECT_DVI | (PORT_B << TRANS_DDI_PORT_SHIFT) | TRANS_DDI_FUNC_ENABLE); if (IS_BROADWELL(dev_priv)) { @@ -239,7 +239,7 @@ static void emulate_monitor_status_change(struct intel_vgpu *vgpu) ~(TRANS_DDI_BPC_MASK | TRANS_DDI_MODE_SELECT_MASK | TRANS_DDI_PORT_MASK); vgpu_vreg_t(vgpu, TRANS_DDI_FUNC_CTL(TRANSCODER_A)) |= - (TRANS_DDI_BPC_8 | TRANS_DDI_MODE_SELECT_DP_SST | + (TRANS_DDI_BPC_8 | TRANS_DDI_MODE_SELECT_DVI | (PORT_C << TRANS_DDI_PORT_SHIFT) | TRANS_DDI_FUNC_ENABLE); if (IS_BROADWELL(dev_priv)) { @@ -259,7 +259,7 @@ static void emulate_monitor_status_change(struct intel_vgpu *vgpu) ~(TRANS_DDI_BPC_MASK | TRANS_DDI_MODE_SELECT_MASK | TRANS_DDI_PORT_MASK); vgpu_vreg_t(vgpu, TRANS_DDI_FUNC_CTL(TRANSCODER_A)) |= - (TRANS_DDI_BPC_8 | TRANS_DDI_MODE_SELECT_DP_SST | + (TRANS_DDI_BPC_8 | TRANS_DDI_MODE_SELECT_DVI | (PORT_D << TRANS_DDI_PORT_SHIFT) | TRANS_DDI_FUNC_ENABLE); if (IS_BROADWELL(dev_priv)) { diff --git a/drivers/gpu/drm/i915/gvt/gtt.c b/drivers/gpu/drm/i915/gvt/gtt.c index 39980dfbbebd..00aad8164dec 100644 --- a/drivers/gpu/drm/i915/gvt/gtt.c +++ b/drivers/gpu/drm/i915/gvt/gtt.c @@ -1901,6 +1901,7 @@ static struct intel_vgpu_mm *intel_vgpu_create_ggtt_mm(struct intel_vgpu *vgpu) vgpu_free_mm(mm); return ERR_PTR(-ENOMEM); } + mm->ggtt_mm.last_partial_off = -1UL; return mm; } @@ -1925,6 +1926,7 @@ void _intel_vgpu_mm_release(struct kref *mm_ref) invalidate_ppgtt_mm(mm); } else { vfree(mm->ggtt_mm.virtual_ggtt); + mm->ggtt_mm.last_partial_off = -1UL; } vgpu_free_mm(mm); @@ -2177,6 +2179,62 @@ static int emulate_ggtt_mmio_write(struct intel_vgpu *vgpu, unsigned int off, memcpy((void *)&e.val64 + (off & (info->gtt_entry_size - 1)), p_data, bytes); + /* If ggtt entry size is 8 bytes, and it's split into two 4 bytes + * write, we assume the two 4 bytes writes are consecutive. + * Otherwise, we abort and report error + */ + if (bytes < info->gtt_entry_size) { + if (ggtt_mm->ggtt_mm.last_partial_off == -1UL) { + /* the first partial part*/ + ggtt_mm->ggtt_mm.last_partial_off = off; + ggtt_mm->ggtt_mm.last_partial_data = e.val64; + return 0; + } else if ((g_gtt_index == + (ggtt_mm->ggtt_mm.last_partial_off >> + info->gtt_entry_size_shift)) && + (off != ggtt_mm->ggtt_mm.last_partial_off)) { + /* the second partial part */ + + int last_off = ggtt_mm->ggtt_mm.last_partial_off & + (info->gtt_entry_size - 1); + + memcpy((void *)&e.val64 + last_off, + (void *)&ggtt_mm->ggtt_mm.last_partial_data + + last_off, bytes); + + ggtt_mm->ggtt_mm.last_partial_off = -1UL; + } else { + int last_offset; + + gvt_vgpu_err("failed to populate guest ggtt entry: abnormal ggtt entry write sequence, last_partial_off=%lx, offset=%x, bytes=%d, ggtt entry size=%d\n", + ggtt_mm->ggtt_mm.last_partial_off, off, + bytes, info->gtt_entry_size); + + /* set host ggtt entry to scratch page and clear + * virtual ggtt entry as not present for last + * partially write offset + */ + last_offset = ggtt_mm->ggtt_mm.last_partial_off & + (~(info->gtt_entry_size - 1)); + + ggtt_get_host_entry(ggtt_mm, &m, last_offset); + ggtt_invalidate_pte(vgpu, &m); + ops->set_pfn(&m, gvt->gtt.scratch_mfn); + ops->clear_present(&m); + ggtt_set_host_entry(ggtt_mm, &m, last_offset); + ggtt_invalidate(gvt->dev_priv); + + ggtt_get_guest_entry(ggtt_mm, &e, last_offset); + ops->clear_present(&e); + ggtt_set_guest_entry(ggtt_mm, &e, last_offset); + + ggtt_mm->ggtt_mm.last_partial_off = off; + ggtt_mm->ggtt_mm.last_partial_data = e.val64; + + return 0; + } + } + if (ops->test_present(&e)) { gfn = ops->get_pfn(&e); m = e; diff --git a/drivers/gpu/drm/i915/gvt/gtt.h b/drivers/gpu/drm/i915/gvt/gtt.h index b7bf68cc8418..7a9b36176efb 100644 --- a/drivers/gpu/drm/i915/gvt/gtt.h +++ b/drivers/gpu/drm/i915/gvt/gtt.h @@ -157,6 +157,8 @@ struct intel_vgpu_mm { } ppgtt_mm; struct { void *virtual_ggtt; + unsigned long last_partial_off; + u64 last_partial_data; } ggtt_mm; }; }; diff --git a/drivers/gpu/drm/i915/gvt/gvt.h b/drivers/gpu/drm/i915/gvt/gvt.h index de2a3a2580be..9a9671522774 100644 --- a/drivers/gpu/drm/i915/gvt/gvt.h +++ b/drivers/gpu/drm/i915/gvt/gvt.h @@ -274,6 +274,8 @@ struct intel_gvt_mmio { #define F_CMD_ACCESSED (1 << 5) /* This reg could be accessed by unaligned address */ #define F_UNALIGN (1 << 6) +/* This reg is saved/restored in context */ +#define F_IN_CTX (1 << 7) struct gvt_mmio_block *mmio_block; unsigned int num_mmio_block; @@ -655,6 +657,33 @@ static inline bool intel_gvt_mmio_has_mode_mask( return gvt->mmio.mmio_attribute[offset >> 2] & F_MODE_MASK; } +/** + * intel_gvt_mmio_is_in_ctx - check if a MMIO has in-ctx mask + * @gvt: a GVT device + * @offset: register offset + * + * Returns: + * True if a MMIO has a in-context mask, false if it isn't. + * + */ +static inline bool intel_gvt_mmio_is_in_ctx( + struct intel_gvt *gvt, unsigned int offset) +{ + return gvt->mmio.mmio_attribute[offset >> 2] & F_IN_CTX; +} + +/** + * intel_gvt_mmio_set_in_ctx - mask a MMIO in logical context + * @gvt: a GVT device + * @offset: register offset + * + */ +static inline void intel_gvt_mmio_set_in_ctx( + struct intel_gvt *gvt, unsigned int offset) +{ + gvt->mmio.mmio_attribute[offset >> 2] |= F_IN_CTX; +} + int intel_gvt_debugfs_add_vgpu(struct intel_vgpu *vgpu); void intel_gvt_debugfs_remove_vgpu(struct intel_vgpu *vgpu); int intel_gvt_debugfs_init(struct intel_gvt *gvt); diff --git a/drivers/gpu/drm/i915/gvt/handlers.c b/drivers/gpu/drm/i915/gvt/handlers.c index 6b50f850dc28..7a58ca555197 100644 --- a/drivers/gpu/drm/i915/gvt/handlers.c +++ b/drivers/gpu/drm/i915/gvt/handlers.c @@ -3388,6 +3388,30 @@ int intel_vgpu_default_mmio_write(struct intel_vgpu *vgpu, unsigned int offset, } /** + * intel_vgpu_mask_mmio_write - write mask register + * @vgpu: a vGPU + * @offset: access offset + * @p_data: write data buffer + * @bytes: access data length + * + * Returns: + * Zero on success, negative error code if failed. + */ +int intel_vgpu_mask_mmio_write(struct intel_vgpu *vgpu, unsigned int offset, + void *p_data, unsigned int bytes) +{ + u32 mask, old_vreg; + + old_vreg = vgpu_vreg(vgpu, offset); + write_vreg(vgpu, offset, p_data, bytes); + mask = vgpu_vreg(vgpu, offset) >> 16; + vgpu_vreg(vgpu, offset) = (old_vreg & ~mask) | + (vgpu_vreg(vgpu, offset) & mask); + + return 0; +} + +/** * intel_gvt_in_force_nonpriv_whitelist - if a mmio is in whitelist to be * force-nopriv register * diff --git a/drivers/gpu/drm/i915/gvt/mmio.h b/drivers/gpu/drm/i915/gvt/mmio.h index e474188b46d2..1ffc69eba30e 100644 --- a/drivers/gpu/drm/i915/gvt/mmio.h +++ b/drivers/gpu/drm/i915/gvt/mmio.h @@ -99,4 +99,6 @@ bool intel_gvt_in_force_nonpriv_whitelist(struct intel_gvt *gvt, int intel_vgpu_mmio_reg_rw(struct intel_vgpu *vgpu, unsigned int offset, void *pdata, unsigned int bytes, bool is_read); +int intel_vgpu_mask_mmio_write(struct intel_vgpu *vgpu, unsigned int offset, + void *p_data, unsigned int bytes); #endif diff --git a/drivers/gpu/drm/i915/gvt/mmio_context.c b/drivers/gpu/drm/i915/gvt/mmio_context.c index 20be9a92600f..42e1e6bdcc2c 100644 --- a/drivers/gpu/drm/i915/gvt/mmio_context.c +++ b/drivers/gpu/drm/i915/gvt/mmio_context.c @@ -587,7 +587,9 @@ void intel_gvt_init_engine_mmio_context(struct intel_gvt *gvt) for (mmio = gvt->engine_mmio_list.mmio; i915_mmio_reg_valid(mmio->reg); mmio++) { - if (mmio->in_context) + if (mmio->in_context) { gvt->engine_mmio_list.ctx_mmio_count[mmio->ring_id]++; + intel_gvt_mmio_set_in_ctx(gvt, mmio->reg.reg); + } } } diff --git a/drivers/gpu/drm/i915/i915_gpu_error.c b/drivers/gpu/drm/i915/i915_gpu_error.c index 8c81cf3aa182..f7f2aa71d8d9 100644 --- a/drivers/gpu/drm/i915/i915_gpu_error.c +++ b/drivers/gpu/drm/i915/i915_gpu_error.c @@ -31,6 +31,7 @@ #include <linux/stop_machine.h> #include <linux/zlib.h> #include <drm/drm_print.h> +#include <linux/ascii85.h> #include "i915_gpu_error.h" #include "i915_drv.h" @@ -517,35 +518,12 @@ void i915_error_printf(struct drm_i915_error_state_buf *e, const char *f, ...) va_end(args); } -static int -ascii85_encode_len(int len) -{ - return DIV_ROUND_UP(len, 4); -} - -static bool -ascii85_encode(u32 in, char *out) -{ - int i; - - if (in == 0) - return false; - - out[5] = '\0'; - for (i = 5; i--; ) { - out[i] = '!' + in % 85; - in /= 85; - } - - return true; -} - static void print_error_obj(struct drm_i915_error_state_buf *m, struct intel_engine_cs *engine, const char *name, struct drm_i915_error_object *obj) { - char out[6]; + char out[ASCII85_BUFSZ]; int page; if (!obj) @@ -567,12 +545,8 @@ static void print_error_obj(struct drm_i915_error_state_buf *m, len -= obj->unused; len = ascii85_encode_len(len); - for (i = 0; i < len; i++) { - if (ascii85_encode(obj->pages[page][i], out)) - err_puts(m, out); - else - err_puts(m, "z"); - } + for (i = 0; i < len; i++) + err_puts(m, ascii85_encode(obj->pages[page][i], out)); } err_puts(m, "\n"); } diff --git a/drivers/gpu/drm/msm/Kconfig b/drivers/gpu/drm/msm/Kconfig index 38cbde971b48..843a9d40c05e 100644 --- a/drivers/gpu/drm/msm/Kconfig +++ b/drivers/gpu/drm/msm/Kconfig @@ -12,6 +12,7 @@ config DRM_MSM select SHMEM select TMPFS select QCOM_SCM + select WANT_DEV_COREDUMP select SND_SOC_HDMI_CODEC if SND_SOC select SYNC_FILE select PM_OPP diff --git a/drivers/gpu/drm/msm/Makefile b/drivers/gpu/drm/msm/Makefile index cd40c050b2d7..7c773e003663 100644 --- a/drivers/gpu/drm/msm/Makefile +++ b/drivers/gpu/drm/msm/Makefile @@ -1,5 +1,6 @@ # SPDX-License-Identifier: GPL-2.0 ccflags-y := -Idrivers/gpu/drm/msm +ccflags-y += -Idrivers/gpu/drm/msm/disp/dpu1 ccflags-$(CONFIG_DRM_MSM_DSI) += -Idrivers/gpu/drm/msm/dsi msm-y := \ @@ -45,6 +46,33 @@ msm-y := \ disp/mdp5/mdp5_mixer.o \ disp/mdp5/mdp5_plane.o \ disp/mdp5/mdp5_smp.o \ + disp/dpu1/dpu_core_irq.o \ + disp/dpu1/dpu_core_perf.o \ + disp/dpu1/dpu_crtc.o \ + disp/dpu1/dpu_encoder.o \ + disp/dpu1/dpu_encoder_phys_cmd.o \ + disp/dpu1/dpu_encoder_phys_vid.o \ + disp/dpu1/dpu_formats.o \ + disp/dpu1/dpu_hw_blk.o \ + disp/dpu1/dpu_hw_catalog.o \ + disp/dpu1/dpu_hw_cdm.o \ + disp/dpu1/dpu_hw_ctl.o \ + disp/dpu1/dpu_hw_interrupts.o \ + disp/dpu1/dpu_hw_intf.o \ + disp/dpu1/dpu_hw_lm.o \ + disp/dpu1/dpu_hw_pingpong.o \ + disp/dpu1/dpu_hw_sspp.o \ + disp/dpu1/dpu_hw_top.o \ + disp/dpu1/dpu_hw_util.o \ + disp/dpu1/dpu_hw_vbif.o \ + disp/dpu1/dpu_io_util.o \ + disp/dpu1/dpu_irq.o \ + disp/dpu1/dpu_kms.o \ + disp/dpu1/dpu_mdss.o \ + disp/dpu1/dpu_plane.o \ + disp/dpu1/dpu_power_handle.o \ + disp/dpu1/dpu_rm.o \ + disp/dpu1/dpu_vbif.o \ msm_atomic.o \ msm_debugfs.o \ msm_drv.o \ @@ -62,7 +90,8 @@ msm-y := \ msm_ringbuffer.o \ msm_submitqueue.o -msm-$(CONFIG_DEBUG_FS) += adreno/a5xx_debugfs.o +msm-$(CONFIG_DEBUG_FS) += adreno/a5xx_debugfs.o \ + disp/dpu1/dpu_dbg.o msm-$(CONFIG_DRM_FBDEV_EMULATION) += msm_fbdev.o msm-$(CONFIG_COMMON_CLK) += disp/mdp4/mdp4_lvds_pll.o diff --git a/drivers/gpu/drm/msm/adreno/a3xx_gpu.c b/drivers/gpu/drm/msm/adreno/a3xx_gpu.c index 3ebbeb3a9b68..669c2d4b070d 100644 --- a/drivers/gpu/drm/msm/adreno/a3xx_gpu.c +++ b/drivers/gpu/drm/msm/adreno/a3xx_gpu.c @@ -411,15 +411,6 @@ static const unsigned int a3xx_registers[] = { ~0 /* sentinel */ }; -#ifdef CONFIG_DEBUG_FS -static void a3xx_show(struct msm_gpu *gpu, struct seq_file *m) -{ - seq_printf(m, "status: %08x\n", - gpu_read(gpu, REG_A3XX_RBBM_STATUS)); - adreno_show(gpu, m); -} -#endif - /* would be nice to not have to duplicate the _show() stuff with printk(): */ static void a3xx_dump(struct msm_gpu *gpu) { @@ -427,6 +418,21 @@ static void a3xx_dump(struct msm_gpu *gpu) gpu_read(gpu, REG_A3XX_RBBM_STATUS)); adreno_dump(gpu); } + +static struct msm_gpu_state *a3xx_gpu_state_get(struct msm_gpu *gpu) +{ + struct msm_gpu_state *state = kzalloc(sizeof(*state), GFP_KERNEL); + + if (!state) + return ERR_PTR(-ENOMEM); + + adreno_gpu_state_get(gpu, state); + + state->rbbm_status = gpu_read(gpu, REG_A3XX_RBBM_STATUS); + + return state; +} + /* Register offset defines for A3XX */ static const unsigned int a3xx_register_offsets[REG_ADRENO_REGISTER_MAX] = { REG_ADRENO_DEFINE(REG_ADRENO_CP_RB_BASE, REG_AXXX_CP_RB_BASE), @@ -450,9 +456,11 @@ static const struct adreno_gpu_funcs funcs = { .active_ring = adreno_active_ring, .irq = a3xx_irq, .destroy = a3xx_destroy, -#ifdef CONFIG_DEBUG_FS - .show = a3xx_show, +#if defined(CONFIG_DEBUG_FS) || defined(CONFIG_DEV_COREDUMP) + .show = adreno_show, #endif + .gpu_state_get = a3xx_gpu_state_get, + .gpu_state_put = adreno_gpu_state_put, }, }; diff --git a/drivers/gpu/drm/msm/adreno/a4xx_gpu.c b/drivers/gpu/drm/msm/adreno/a4xx_gpu.c index 16d3d596638e..7c4e6dc1ed59 100644 --- a/drivers/gpu/drm/msm/adreno/a4xx_gpu.c +++ b/drivers/gpu/drm/msm/adreno/a4xx_gpu.c @@ -455,15 +455,19 @@ static const unsigned int a4xx_registers[] = { ~0 /* sentinel */ }; -#ifdef CONFIG_DEBUG_FS -static void a4xx_show(struct msm_gpu *gpu, struct seq_file *m) +static struct msm_gpu_state *a4xx_gpu_state_get(struct msm_gpu *gpu) { - seq_printf(m, "status: %08x\n", - gpu_read(gpu, REG_A4XX_RBBM_STATUS)); - adreno_show(gpu, m); + struct msm_gpu_state *state = kzalloc(sizeof(*state), GFP_KERNEL); + + if (!state) + return ERR_PTR(-ENOMEM); + + adreno_gpu_state_get(gpu, state); + state->rbbm_status = gpu_read(gpu, REG_A4XX_RBBM_STATUS); + + return state; } -#endif /* Register offset defines for A4XX, in order of enum adreno_regs */ static const unsigned int a4xx_register_offsets[REG_ADRENO_REGISTER_MAX] = { @@ -538,9 +542,11 @@ static const struct adreno_gpu_funcs funcs = { .active_ring = adreno_active_ring, .irq = a4xx_irq, .destroy = a4xx_destroy, -#ifdef CONFIG_DEBUG_FS - .show = a4xx_show, +#if defined(CONFIG_DEBUG_FS) || defined(CONFIG_DEV_COREDUMP) + .show = adreno_show, #endif + .gpu_state_get = a4xx_gpu_state_get, + .gpu_state_put = adreno_gpu_state_put, }, .get_timestamp = a4xx_get_timestamp, }; diff --git a/drivers/gpu/drm/msm/adreno/a5xx_gpu.c b/drivers/gpu/drm/msm/adreno/a5xx_gpu.c index d39400e5bc42..bd84f71d27d8 100644 --- a/drivers/gpu/drm/msm/adreno/a5xx_gpu.c +++ b/drivers/gpu/drm/msm/adreno/a5xx_gpu.c @@ -19,6 +19,7 @@ #include <linux/soc/qcom/mdt_loader.h> #include <linux/pm_opp.h> #include <linux/nvmem-consumer.h> +#include <linux/iopoll.h> #include "msm_gem.h" #include "msm_mmu.h" #include "a5xx_gpu.h" @@ -1123,8 +1124,9 @@ static const u32 a5xx_registers[] = { 0xE800, 0xE806, 0xE810, 0xE89A, 0xE8A0, 0xE8A4, 0xE8AA, 0xE8EB, 0xE900, 0xE905, 0xEB80, 0xEB8F, 0xEBB0, 0xEBB0, 0xEC00, 0xEC05, 0xEC08, 0xECE9, 0xECF0, 0xECF0, 0xEA80, 0xEA80, 0xEA82, 0xEAA3, - 0xEAA5, 0xEAC2, 0xA800, 0xA8FF, 0xAC60, 0xAC60, 0xB000, 0xB97F, - 0xB9A0, 0xB9BF, ~0 + 0xEAA5, 0xEAC2, 0xA800, 0xA800, 0xA820, 0xA828, 0xA840, 0xA87D, + 0XA880, 0xA88D, 0xA890, 0xA8A3, 0xA8D0, 0xA8D8, 0xA8E0, 0xA8F5, + 0xAC60, 0xAC60, ~0, }; static void a5xx_dump(struct msm_gpu *gpu) @@ -1195,19 +1197,231 @@ static int a5xx_get_timestamp(struct msm_gpu *gpu, uint64_t *value) return 0; } -#ifdef CONFIG_DEBUG_FS -static void a5xx_show(struct msm_gpu *gpu, struct seq_file *m) +struct a5xx_crashdumper { + void *ptr; + struct drm_gem_object *bo; + u64 iova; +}; + +struct a5xx_gpu_state { + struct msm_gpu_state base; + u32 *hlsqregs; +}; + +#define gpu_poll_timeout(gpu, addr, val, cond, interval, timeout) \ + readl_poll_timeout((gpu)->mmio + ((addr) << 2), val, cond, \ + interval, timeout) + +static int a5xx_crashdumper_init(struct msm_gpu *gpu, + struct a5xx_crashdumper *dumper) { - seq_printf(m, "status: %08x\n", - gpu_read(gpu, REG_A5XX_RBBM_STATUS)); + dumper->ptr = msm_gem_kernel_new_locked(gpu->dev, + SZ_1M, MSM_BO_UNCACHED, gpu->aspace, + &dumper->bo, &dumper->iova); - /* - * Temporarily disable hardware clock gating before going into - * adreno_show to avoid issues while reading the registers - */ + if (IS_ERR(dumper->ptr)) + return PTR_ERR(dumper->ptr); + + return 0; +} + +static void a5xx_crashdumper_free(struct msm_gpu *gpu, + struct a5xx_crashdumper *dumper) +{ + msm_gem_put_iova(dumper->bo, gpu->aspace); + msm_gem_put_vaddr(dumper->bo); + + drm_gem_object_unreference(dumper->bo); +} + +static int a5xx_crashdumper_run(struct msm_gpu *gpu, + struct a5xx_crashdumper *dumper) +{ + u32 val; + + if (IS_ERR_OR_NULL(dumper->ptr)) + return -EINVAL; + + gpu_write64(gpu, REG_A5XX_CP_CRASH_SCRIPT_BASE_LO, + REG_A5XX_CP_CRASH_SCRIPT_BASE_HI, dumper->iova); + + gpu_write(gpu, REG_A5XX_CP_CRASH_DUMP_CNTL, 1); + + return gpu_poll_timeout(gpu, REG_A5XX_CP_CRASH_DUMP_CNTL, val, + val & 0x04, 100, 10000); +} + +/* + * These are a list of the registers that need to be read through the HLSQ + * aperture through the crashdumper. These are not nominally accessible from + * the CPU on a secure platform. + */ +static const struct { + u32 type; + u32 regoffset; + u32 count; +} a5xx_hlsq_aperture_regs[] = { + { 0x35, 0xe00, 0x32 }, /* HSLQ non-context */ + { 0x31, 0x2080, 0x1 }, /* HLSQ 2D context 0 */ + { 0x33, 0x2480, 0x1 }, /* HLSQ 2D context 1 */ + { 0x32, 0xe780, 0x62 }, /* HLSQ 3D context 0 */ + { 0x34, 0xef80, 0x62 }, /* HLSQ 3D context 1 */ + { 0x3f, 0x0ec0, 0x40 }, /* SP non-context */ + { 0x3d, 0x2040, 0x1 }, /* SP 2D context 0 */ + { 0x3b, 0x2440, 0x1 }, /* SP 2D context 1 */ + { 0x3e, 0xe580, 0x170 }, /* SP 3D context 0 */ + { 0x3c, 0xed80, 0x170 }, /* SP 3D context 1 */ + { 0x3a, 0x0f00, 0x1c }, /* TP non-context */ + { 0x38, 0x2000, 0xa }, /* TP 2D context 0 */ + { 0x36, 0x2400, 0xa }, /* TP 2D context 1 */ + { 0x39, 0xe700, 0x80 }, /* TP 3D context 0 */ + { 0x37, 0xef00, 0x80 }, /* TP 3D context 1 */ +}; + +static void a5xx_gpu_state_get_hlsq_regs(struct msm_gpu *gpu, + struct a5xx_gpu_state *a5xx_state) +{ + struct a5xx_crashdumper dumper = { 0 }; + u32 offset, count = 0; + u64 *ptr; + int i; + + if (a5xx_crashdumper_init(gpu, &dumper)) + return; + + /* The script will be written at offset 0 */ + ptr = dumper.ptr; + + /* Start writing the data at offset 256k */ + offset = dumper.iova + (256 * SZ_1K); + + /* Count how many additional registers to get from the HLSQ aperture */ + for (i = 0; i < ARRAY_SIZE(a5xx_hlsq_aperture_regs); i++) + count += a5xx_hlsq_aperture_regs[i].count; + + a5xx_state->hlsqregs = kcalloc(count, sizeof(u32), GFP_KERNEL); + if (!a5xx_state->hlsqregs) + return; + + /* Build the crashdump script */ + for (i = 0; i < ARRAY_SIZE(a5xx_hlsq_aperture_regs); i++) { + u32 type = a5xx_hlsq_aperture_regs[i].type; + u32 c = a5xx_hlsq_aperture_regs[i].count; + + /* Write the register to select the desired bank */ + *ptr++ = ((u64) type << 8); + *ptr++ = (((u64) REG_A5XX_HLSQ_DBG_READ_SEL) << 44) | + (1 << 21) | 1; + + *ptr++ = offset; + *ptr++ = (((u64) REG_A5XX_HLSQ_DBG_AHB_READ_APERTURE) << 44) + | c; + + offset += c * sizeof(u32); + } + + /* Write two zeros to close off the script */ + *ptr++ = 0; + *ptr++ = 0; + + if (a5xx_crashdumper_run(gpu, &dumper)) { + kfree(a5xx_state->hlsqregs); + a5xx_crashdumper_free(gpu, &dumper); + return; + } + + /* Copy the data from the crashdumper to the state */ + memcpy(a5xx_state->hlsqregs, dumper.ptr + (256 * SZ_1K), + count * sizeof(u32)); + + a5xx_crashdumper_free(gpu, &dumper); +} + +static struct msm_gpu_state *a5xx_gpu_state_get(struct msm_gpu *gpu) +{ + struct a5xx_gpu_state *a5xx_state = kzalloc(sizeof(*a5xx_state), + GFP_KERNEL); + + if (!a5xx_state) + return ERR_PTR(-ENOMEM); + + /* Temporarily disable hardware clock gating before reading the hw */ a5xx_set_hwcg(gpu, false); - adreno_show(gpu, m); + + /* First get the generic state from the adreno core */ + adreno_gpu_state_get(gpu, &(a5xx_state->base)); + + a5xx_state->base.rbbm_status = gpu_read(gpu, REG_A5XX_RBBM_STATUS); + + /* Get the HLSQ regs with the help of the crashdumper */ + a5xx_gpu_state_get_hlsq_regs(gpu, a5xx_state); + a5xx_set_hwcg(gpu, true); + + return &a5xx_state->base; +} + +static void a5xx_gpu_state_destroy(struct kref *kref) +{ + struct msm_gpu_state *state = container_of(kref, + struct msm_gpu_state, ref); + struct a5xx_gpu_state *a5xx_state = container_of(state, + struct a5xx_gpu_state, base); + + kfree(a5xx_state->hlsqregs); + + adreno_gpu_state_destroy(state); + kfree(a5xx_state); +} + +int a5xx_gpu_state_put(struct msm_gpu_state *state) +{ + if (IS_ERR_OR_NULL(state)) + return 1; + + return kref_put(&state->ref, a5xx_gpu_state_destroy); +} + + +#if defined(CONFIG_DEBUG_FS) || defined(CONFIG_DEV_COREDUMP) +void a5xx_show(struct msm_gpu *gpu, struct msm_gpu_state *state, + struct drm_printer *p) +{ + int i, j; + u32 pos = 0; + struct a5xx_gpu_state *a5xx_state = container_of(state, + struct a5xx_gpu_state, base); + + if (IS_ERR_OR_NULL(state)) + return; + + adreno_show(gpu, state, p); + + /* Dump the additional a5xx HLSQ registers */ + if (!a5xx_state->hlsqregs) + return; + + drm_printf(p, "registers-hlsq:\n"); + + for (i = 0; i < ARRAY_SIZE(a5xx_hlsq_aperture_regs); i++) { + u32 o = a5xx_hlsq_aperture_regs[i].regoffset; + u32 c = a5xx_hlsq_aperture_regs[i].count; + + for (j = 0; j < c; j++, pos++, o++) { + /* + * To keep the crashdump simple we pull the entire range + * for each register type but not all of the registers + * in the range are valid. Fortunately invalid registers + * stick out like a sore thumb with a value of + * 0xdeadbeef + */ + if (a5xx_state->hlsqregs[pos] == 0xdeadbeef) + continue; + + drm_printf(p, " - { offset: 0x%04x, value: 0x%08x }\n", + o << 2, a5xx_state->hlsqregs[pos]); + } + } } #endif @@ -1239,11 +1453,15 @@ static const struct adreno_gpu_funcs funcs = { .active_ring = a5xx_active_ring, .irq = a5xx_irq, .destroy = a5xx_destroy, -#ifdef CONFIG_DEBUG_FS +#if defined(CONFIG_DEBUG_FS) || defined(CONFIG_DEV_COREDUMP) .show = a5xx_show, +#endif +#if defined(CONFIG_DEBUG_FS) .debugfs_init = a5xx_debugfs_init, #endif .gpu_busy = a5xx_gpu_busy, + .gpu_state_get = a5xx_gpu_state_get, + .gpu_state_put = a5xx_gpu_state_put, }, .get_timestamp = a5xx_get_timestamp, }; diff --git a/drivers/gpu/drm/msm/adreno/adreno_device.c b/drivers/gpu/drm/msm/adreno/adreno_device.c index 0ae5ace65462..44813624a286 100644 --- a/drivers/gpu/drm/msm/adreno/adreno_device.c +++ b/drivers/gpu/drm/msm/adreno/adreno_device.c @@ -35,6 +35,7 @@ static const struct adreno_info gpulist[] = { [ADRENO_FW_PFP] = "a300_pfp.fw", }, .gmem = SZ_256K, + .inactive_period = DRM_MSM_INACTIVE_PERIOD, .init = a3xx_gpu_init, }, { .rev = ADRENO_REV(3, 0, 6, 0), @@ -45,6 +46,7 @@ static const struct adreno_info gpulist[] = { [ADRENO_FW_PFP] = "a300_pfp.fw", }, .gmem = SZ_128K, + .inactive_period = DRM_MSM_INACTIVE_PERIOD, .init = a3xx_gpu_init, }, { .rev = ADRENO_REV(3, 2, ANY_ID, ANY_ID), @@ -55,6 +57,7 @@ static const struct adreno_info gpulist[] = { [ADRENO_FW_PFP] = "a300_pfp.fw", }, .gmem = SZ_512K, + .inactive_period = DRM_MSM_INACTIVE_PERIOD, .init = a3xx_gpu_init, }, { .rev = ADRENO_REV(3, 3, 0, ANY_ID), @@ -65,6 +68,7 @@ static const struct adreno_info gpulist[] = { [ADRENO_FW_PFP] = "a330_pfp.fw", }, .gmem = SZ_1M, + .inactive_period = DRM_MSM_INACTIVE_PERIOD, .init = a3xx_gpu_init, }, { .rev = ADRENO_REV(4, 2, 0, ANY_ID), @@ -75,6 +79,7 @@ static const struct adreno_info gpulist[] = { [ADRENO_FW_PFP] = "a420_pfp.fw", }, .gmem = (SZ_1M + SZ_512K), + .inactive_period = DRM_MSM_INACTIVE_PERIOD, .init = a4xx_gpu_init, }, { .rev = ADRENO_REV(4, 3, 0, ANY_ID), @@ -85,6 +90,7 @@ static const struct adreno_info gpulist[] = { [ADRENO_FW_PFP] = "a420_pfp.fw", }, .gmem = (SZ_1M + SZ_512K), + .inactive_period = DRM_MSM_INACTIVE_PERIOD, .init = a4xx_gpu_init, }, { .rev = ADRENO_REV(5, 3, 0, 2), @@ -96,6 +102,11 @@ static const struct adreno_info gpulist[] = { [ADRENO_FW_GPMU] = "a530v3_gpmu.fw2", }, .gmem = SZ_1M, + /* + * Increase inactive period to 250 to avoid bouncing + * the GDSC which appears to make it grumpy + */ + .inactive_period = 250, .quirks = ADRENO_QUIRK_TWO_PASS_USE_WFI | ADRENO_QUIRK_FAULT_DETECT_MASK, .init = a5xx_gpu_init, @@ -158,7 +169,7 @@ struct msm_gpu *adreno_load_gpu(struct drm_device *dev) mutex_lock(&dev->struct_mutex); ret = msm_gpu_hw_init(gpu); mutex_unlock(&dev->struct_mutex); - pm_runtime_put_sync(&pdev->dev); + pm_runtime_put_autosuspend(&pdev->dev); if (ret) { dev_err(dev->dev, "gpu hw init failed: %d\n", ret); return NULL; @@ -316,6 +327,7 @@ static int adreno_suspend(struct device *dev) #endif static const struct dev_pm_ops adreno_pm_ops = { + SET_SYSTEM_SLEEP_PM_OPS(pm_runtime_force_suspend, pm_runtime_force_resume) SET_RUNTIME_PM_OPS(adreno_suspend, adreno_resume, NULL) }; diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.c b/drivers/gpu/drm/msm/adreno/adreno_gpu.c index 17d0506d058c..38ac50b73829 100644 --- a/drivers/gpu/drm/msm/adreno/adreno_gpu.c +++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.c @@ -17,6 +17,7 @@ * this program. If not, see <http://www.gnu.org/licenses/>. */ +#include <linux/ascii85.h> #include <linux/pm_opp.h> #include "adreno_gpu.h" #include "msm_gem.h" @@ -368,40 +369,185 @@ bool adreno_idle(struct msm_gpu *gpu, struct msm_ringbuffer *ring) return false; } -#ifdef CONFIG_DEBUG_FS -void adreno_show(struct msm_gpu *gpu, struct seq_file *m) +int adreno_gpu_state_get(struct msm_gpu *gpu, struct msm_gpu_state *state) { struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu); + int i, count = 0; + + kref_init(&state->ref); + + ktime_get_real_ts64(&state->time); + + for (i = 0; i < gpu->nr_rings; i++) { + int size = 0, j; + + state->ring[i].fence = gpu->rb[i]->memptrs->fence; + state->ring[i].iova = gpu->rb[i]->iova; + state->ring[i].seqno = gpu->rb[i]->seqno; + state->ring[i].rptr = get_rptr(adreno_gpu, gpu->rb[i]); + state->ring[i].wptr = get_wptr(gpu->rb[i]); + + /* Copy at least 'wptr' dwords of the data */ + size = state->ring[i].wptr; + + /* After wptr find the last non zero dword to save space */ + for (j = state->ring[i].wptr; j < MSM_GPU_RINGBUFFER_SZ >> 2; j++) + if (gpu->rb[i]->start[j]) + size = j + 1; + + if (size) { + state->ring[i].data = kmalloc(size << 2, GFP_KERNEL); + if (state->ring[i].data) { + memcpy(state->ring[i].data, gpu->rb[i]->start, size << 2); + state->ring[i].data_size = size << 2; + } + } + } + + /* Count the number of registers */ + for (i = 0; adreno_gpu->registers[i] != ~0; i += 2) + count += adreno_gpu->registers[i + 1] - + adreno_gpu->registers[i] + 1; + + state->registers = kcalloc(count * 2, sizeof(u32), GFP_KERNEL); + if (state->registers) { + int pos = 0; + + for (i = 0; adreno_gpu->registers[i] != ~0; i += 2) { + u32 start = adreno_gpu->registers[i]; + u32 end = adreno_gpu->registers[i + 1]; + u32 addr; + + for (addr = start; addr <= end; addr++) { + state->registers[pos++] = addr; + state->registers[pos++] = gpu_read(gpu, addr); + } + } + + state->nr_registers = count; + } + + return 0; +} + +void adreno_gpu_state_destroy(struct msm_gpu_state *state) +{ int i; - seq_printf(m, "revision: %d (%d.%d.%d.%d)\n", + for (i = 0; i < ARRAY_SIZE(state->ring); i++) + kfree(state->ring[i].data); + + for (i = 0; state->bos && i < state->nr_bos; i++) + kvfree(state->bos[i].data); + + kfree(state->bos); + kfree(state->comm); + kfree(state->cmd); + kfree(state->registers); +} + +static void adreno_gpu_state_kref_destroy(struct kref *kref) +{ + struct msm_gpu_state *state = container_of(kref, + struct msm_gpu_state, ref); + + adreno_gpu_state_destroy(state); + kfree(state); +} + +int adreno_gpu_state_put(struct msm_gpu_state *state) +{ + if (IS_ERR_OR_NULL(state)) + return 1; + + return kref_put(&state->ref, adreno_gpu_state_kref_destroy); +} + +#if defined(CONFIG_DEBUG_FS) || defined(CONFIG_DEV_COREDUMP) + +static void adreno_show_object(struct drm_printer *p, u32 *ptr, int len) +{ + char out[ASCII85_BUFSZ]; + long l, datalen, i; + + if (!ptr || !len) + return; + + /* + * Only dump the non-zero part of the buffer - rarely will any data + * completely fill the entire allocated size of the buffer + */ + for (datalen = 0, i = 0; i < len >> 2; i++) { + if (ptr[i]) + datalen = (i << 2) + 1; + } + + /* Skip printing the object if it is empty */ + if (datalen == 0) + return; + + l = ascii85_encode_len(datalen); + + drm_puts(p, " data: !!ascii85 |\n"); + drm_puts(p, " "); + + for (i = 0; i < l; i++) + drm_puts(p, ascii85_encode(ptr[i], out)); + + drm_puts(p, "\n"); +} + +void adreno_show(struct msm_gpu *gpu, struct msm_gpu_state *state, + struct drm_printer *p) +{ + struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu); + int i; + + if (IS_ERR_OR_NULL(state)) + return; + + drm_printf(p, "revision: %d (%d.%d.%d.%d)\n", adreno_gpu->info->revn, adreno_gpu->rev.core, adreno_gpu->rev.major, adreno_gpu->rev.minor, adreno_gpu->rev.patchid); - for (i = 0; i < gpu->nr_rings; i++) { - struct msm_ringbuffer *ring = gpu->rb[i]; + drm_printf(p, "rbbm-status: 0x%08x\n", state->rbbm_status); - seq_printf(m, "rb %d: fence: %d/%d\n", i, - ring->memptrs->fence, ring->seqno); + drm_puts(p, "ringbuffer:\n"); - seq_printf(m, " rptr: %d\n", - get_rptr(adreno_gpu, ring)); - seq_printf(m, "rb wptr: %d\n", get_wptr(ring)); + for (i = 0; i < gpu->nr_rings; i++) { + drm_printf(p, " - id: %d\n", i); + drm_printf(p, " iova: 0x%016llx\n", state->ring[i].iova); + drm_printf(p, " last-fence: %d\n", state->ring[i].seqno); + drm_printf(p, " retired-fence: %d\n", state->ring[i].fence); + drm_printf(p, " rptr: %d\n", state->ring[i].rptr); + drm_printf(p, " wptr: %d\n", state->ring[i].wptr); + drm_printf(p, " size: %d\n", MSM_GPU_RINGBUFFER_SZ); + + adreno_show_object(p, state->ring[i].data, + state->ring[i].data_size); } - /* dump these out in a form that can be parsed by demsm: */ - seq_printf(m, "IO:region %s 00000000 00020000\n", gpu->name); - for (i = 0; adreno_gpu->registers[i] != ~0; i += 2) { - uint32_t start = adreno_gpu->registers[i]; - uint32_t end = adreno_gpu->registers[i+1]; - uint32_t addr; + if (state->bos) { + drm_puts(p, "bos:\n"); - for (addr = start; addr <= end; addr++) { - uint32_t val = gpu_read(gpu, addr); - seq_printf(m, "IO:R %08x %08x\n", addr<<2, val); + for (i = 0; i < state->nr_bos; i++) { + drm_printf(p, " - iova: 0x%016llx\n", + state->bos[i].iova); + drm_printf(p, " size: %zd\n", state->bos[i].size); + + adreno_show_object(p, state->bos[i].data, + state->bos[i].size); } } + + drm_puts(p, "registers:\n"); + + for (i = 0; i < state->nr_registers; i++) { + drm_printf(p, " - { offset: 0x%04x, value: 0x%08x }\n", + state->registers[i * 2] << 2, + state->registers[(i * 2) + 1]); + } } #endif @@ -565,7 +711,8 @@ int adreno_gpu_init(struct drm_device *drm, struct platform_device *pdev, adreno_get_pwrlevels(&pdev->dev, gpu); - pm_runtime_set_autosuspend_delay(&pdev->dev, DRM_MSM_INACTIVE_PERIOD); + pm_runtime_set_autosuspend_delay(&pdev->dev, + adreno_gpu->info->inactive_period); pm_runtime_use_autosuspend(&pdev->dev); pm_runtime_enable(&pdev->dev); diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.h b/drivers/gpu/drm/msm/adreno/adreno_gpu.h index d6b0e7b813f4..4406776597fd 100644 --- a/drivers/gpu/drm/msm/adreno/adreno_gpu.h +++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.h @@ -84,6 +84,7 @@ struct adreno_info { enum adreno_quirks quirks; struct msm_gpu *(*init)(struct drm_device *dev); const char *zapfw; + u32 inactive_period; }; const struct adreno_info *adreno_info(struct adreno_rev rev); @@ -214,8 +215,9 @@ void adreno_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit, struct msm_file_private *ctx); void adreno_flush(struct msm_gpu *gpu, struct msm_ringbuffer *ring); bool adreno_idle(struct msm_gpu *gpu, struct msm_ringbuffer *ring); -#ifdef CONFIG_DEBUG_FS -void adreno_show(struct msm_gpu *gpu, struct seq_file *m); +#if defined(CONFIG_DEBUG_FS) || defined(CONFIG_DEV_COREDUMP) +void adreno_show(struct msm_gpu *gpu, struct msm_gpu_state *state, + struct drm_printer *p); #endif void adreno_dump_info(struct msm_gpu *gpu); void adreno_dump(struct msm_gpu *gpu); @@ -228,6 +230,11 @@ int adreno_gpu_init(struct drm_device *drm, struct platform_device *pdev, void adreno_gpu_cleanup(struct adreno_gpu *gpu); +void adreno_gpu_state_destroy(struct msm_gpu_state *state); + +int adreno_gpu_state_get(struct msm_gpu *gpu, struct msm_gpu_state *state); +int adreno_gpu_state_put(struct msm_gpu_state *state); + /* ringbuffer helpers (the parts that are adreno specific) */ static inline void diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_core_irq.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_core_irq.c new file mode 100644 index 000000000000..879c13fe74e0 --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_core_irq.c @@ -0,0 +1,479 @@ +/* Copyright (c) 2015-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#define pr_fmt(fmt) "[drm:%s:%d] " fmt, __func__, __LINE__ + +#include <linux/debugfs.h> +#include <linux/irqdomain.h> +#include <linux/irq.h> +#include <linux/kthread.h> + +#include "dpu_core_irq.h" +#include "dpu_trace.h" + +/** + * dpu_core_irq_callback_handler - dispatch core interrupts + * @arg: private data of callback handler + * @irq_idx: interrupt index + */ +static void dpu_core_irq_callback_handler(void *arg, int irq_idx) +{ + struct dpu_kms *dpu_kms = arg; + struct dpu_irq *irq_obj = &dpu_kms->irq_obj; + struct dpu_irq_callback *cb; + unsigned long irq_flags; + + pr_debug("irq_idx=%d\n", irq_idx); + + if (list_empty(&irq_obj->irq_cb_tbl[irq_idx])) { + DRM_ERROR("no registered cb, idx:%d enable_count:%d\n", irq_idx, + atomic_read(&dpu_kms->irq_obj.enable_counts[irq_idx])); + } + + atomic_inc(&irq_obj->irq_counts[irq_idx]); + + /* + * Perform registered function callback + */ + spin_lock_irqsave(&dpu_kms->irq_obj.cb_lock, irq_flags); + list_for_each_entry(cb, &irq_obj->irq_cb_tbl[irq_idx], list) + if (cb->func) + cb->func(cb->arg, irq_idx); + spin_unlock_irqrestore(&dpu_kms->irq_obj.cb_lock, irq_flags); + + /* + * Clear pending interrupt status in HW. + * NOTE: dpu_core_irq_callback_handler is protected by top-level + * spinlock, so it is safe to clear any interrupt status here. + */ + dpu_kms->hw_intr->ops.clear_intr_status_nolock( + dpu_kms->hw_intr, + irq_idx); +} + +int dpu_core_irq_idx_lookup(struct dpu_kms *dpu_kms, + enum dpu_intr_type intr_type, u32 instance_idx) +{ + if (!dpu_kms || !dpu_kms->hw_intr || + !dpu_kms->hw_intr->ops.irq_idx_lookup) + return -EINVAL; + + return dpu_kms->hw_intr->ops.irq_idx_lookup(intr_type, + instance_idx); +} + +/** + * _dpu_core_irq_enable - enable core interrupt given by the index + * @dpu_kms: Pointer to dpu kms context + * @irq_idx: interrupt index + */ +static int _dpu_core_irq_enable(struct dpu_kms *dpu_kms, int irq_idx) +{ + unsigned long irq_flags; + int ret = 0, enable_count; + + if (!dpu_kms || !dpu_kms->hw_intr || + !dpu_kms->irq_obj.enable_counts || + !dpu_kms->irq_obj.irq_counts) { + DPU_ERROR("invalid params\n"); + return -EINVAL; + } + + if (irq_idx < 0 || irq_idx >= dpu_kms->hw_intr->irq_idx_tbl_size) { + DPU_ERROR("invalid IRQ index: [%d]\n", irq_idx); + return -EINVAL; + } + + enable_count = atomic_read(&dpu_kms->irq_obj.enable_counts[irq_idx]); + DRM_DEBUG_KMS("irq_idx=%d enable_count=%d\n", irq_idx, enable_count); + trace_dpu_core_irq_enable_idx(irq_idx, enable_count); + + if (atomic_inc_return(&dpu_kms->irq_obj.enable_counts[irq_idx]) == 1) { + ret = dpu_kms->hw_intr->ops.enable_irq( + dpu_kms->hw_intr, + irq_idx); + if (ret) + DPU_ERROR("Fail to enable IRQ for irq_idx:%d\n", + irq_idx); + + DPU_DEBUG("irq_idx=%d ret=%d\n", irq_idx, ret); + + spin_lock_irqsave(&dpu_kms->irq_obj.cb_lock, irq_flags); + /* empty callback list but interrupt is enabled */ + if (list_empty(&dpu_kms->irq_obj.irq_cb_tbl[irq_idx])) + DPU_ERROR("irq_idx=%d enabled with no callback\n", + irq_idx); + spin_unlock_irqrestore(&dpu_kms->irq_obj.cb_lock, irq_flags); + } + + return ret; +} + +int dpu_core_irq_enable(struct dpu_kms *dpu_kms, int *irq_idxs, u32 irq_count) +{ + int i, ret = 0, counts; + + if (!dpu_kms || !irq_idxs || !irq_count) { + DPU_ERROR("invalid params\n"); + return -EINVAL; + } + + counts = atomic_read(&dpu_kms->irq_obj.enable_counts[irq_idxs[0]]); + if (counts) + DRM_ERROR("irq_idx=%d enable_count=%d\n", irq_idxs[0], counts); + + for (i = 0; (i < irq_count) && !ret; i++) + ret = _dpu_core_irq_enable(dpu_kms, irq_idxs[i]); + + return ret; +} + +/** + * _dpu_core_irq_disable - disable core interrupt given by the index + * @dpu_kms: Pointer to dpu kms context + * @irq_idx: interrupt index + */ +static int _dpu_core_irq_disable(struct dpu_kms *dpu_kms, int irq_idx) +{ + int ret = 0, enable_count; + + if (!dpu_kms || !dpu_kms->hw_intr || !dpu_kms->irq_obj.enable_counts) { + DPU_ERROR("invalid params\n"); + return -EINVAL; + } + + if (irq_idx < 0 || irq_idx >= dpu_kms->hw_intr->irq_idx_tbl_size) { + DPU_ERROR("invalid IRQ index: [%d]\n", irq_idx); + return -EINVAL; + } + + enable_count = atomic_read(&dpu_kms->irq_obj.enable_counts[irq_idx]); + DRM_DEBUG_KMS("irq_idx=%d enable_count=%d\n", irq_idx, enable_count); + trace_dpu_core_irq_disable_idx(irq_idx, enable_count); + + if (atomic_dec_return(&dpu_kms->irq_obj.enable_counts[irq_idx]) == 0) { + ret = dpu_kms->hw_intr->ops.disable_irq( + dpu_kms->hw_intr, + irq_idx); + if (ret) + DPU_ERROR("Fail to disable IRQ for irq_idx:%d\n", + irq_idx); + DPU_DEBUG("irq_idx=%d ret=%d\n", irq_idx, ret); + } + + return ret; +} + +int dpu_core_irq_disable(struct dpu_kms *dpu_kms, int *irq_idxs, u32 irq_count) +{ + int i, ret = 0, counts; + + if (!dpu_kms || !irq_idxs || !irq_count) { + DPU_ERROR("invalid params\n"); + return -EINVAL; + } + + counts = atomic_read(&dpu_kms->irq_obj.enable_counts[irq_idxs[0]]); + if (counts == 2) + DRM_ERROR("irq_idx=%d enable_count=%d\n", irq_idxs[0], counts); + + for (i = 0; (i < irq_count) && !ret; i++) + ret = _dpu_core_irq_disable(dpu_kms, irq_idxs[i]); + + return ret; +} + +u32 dpu_core_irq_read(struct dpu_kms *dpu_kms, int irq_idx, bool clear) +{ + if (!dpu_kms || !dpu_kms->hw_intr || + !dpu_kms->hw_intr->ops.get_interrupt_status) + return 0; + + if (irq_idx < 0) { + DPU_ERROR("[%pS] invalid irq_idx=%d\n", + __builtin_return_address(0), irq_idx); + return 0; + } + + return dpu_kms->hw_intr->ops.get_interrupt_status(dpu_kms->hw_intr, + irq_idx, clear); +} + +int dpu_core_irq_register_callback(struct dpu_kms *dpu_kms, int irq_idx, + struct dpu_irq_callback *register_irq_cb) +{ + unsigned long irq_flags; + + if (!dpu_kms || !dpu_kms->irq_obj.irq_cb_tbl) { + DPU_ERROR("invalid params\n"); + return -EINVAL; + } + + if (!register_irq_cb || !register_irq_cb->func) { + DPU_ERROR("invalid irq_cb:%d func:%d\n", + register_irq_cb != NULL, + register_irq_cb ? + register_irq_cb->func != NULL : -1); + return -EINVAL; + } + + if (irq_idx < 0 || irq_idx >= dpu_kms->hw_intr->irq_idx_tbl_size) { + DPU_ERROR("invalid IRQ index: [%d]\n", irq_idx); + return -EINVAL; + } + + DPU_DEBUG("[%pS] irq_idx=%d\n", __builtin_return_address(0), irq_idx); + + spin_lock_irqsave(&dpu_kms->irq_obj.cb_lock, irq_flags); + trace_dpu_core_irq_register_callback(irq_idx, register_irq_cb); + list_del_init(®ister_irq_cb->list); + list_add_tail(®ister_irq_cb->list, + &dpu_kms->irq_obj.irq_cb_tbl[irq_idx]); + spin_unlock_irqrestore(&dpu_kms->irq_obj.cb_lock, irq_flags); + + return 0; +} + +int dpu_core_irq_unregister_callback(struct dpu_kms *dpu_kms, int irq_idx, + struct dpu_irq_callback *register_irq_cb) +{ + unsigned long irq_flags; + + if (!dpu_kms || !dpu_kms->irq_obj.irq_cb_tbl) { + DPU_ERROR("invalid params\n"); + return -EINVAL; + } + + if (!register_irq_cb || !register_irq_cb->func) { + DPU_ERROR("invalid irq_cb:%d func:%d\n", + register_irq_cb != NULL, + register_irq_cb ? + register_irq_cb->func != NULL : -1); + return -EINVAL; + } + + if (irq_idx < 0 || irq_idx >= dpu_kms->hw_intr->irq_idx_tbl_size) { + DPU_ERROR("invalid IRQ index: [%d]\n", irq_idx); + return -EINVAL; + } + + DPU_DEBUG("[%pS] irq_idx=%d\n", __builtin_return_address(0), irq_idx); + + spin_lock_irqsave(&dpu_kms->irq_obj.cb_lock, irq_flags); + trace_dpu_core_irq_unregister_callback(irq_idx, register_irq_cb); + list_del_init(®ister_irq_cb->list); + /* empty callback list but interrupt is still enabled */ + if (list_empty(&dpu_kms->irq_obj.irq_cb_tbl[irq_idx]) && + atomic_read(&dpu_kms->irq_obj.enable_counts[irq_idx])) + DPU_ERROR("irq_idx=%d enabled with no callback\n", irq_idx); + spin_unlock_irqrestore(&dpu_kms->irq_obj.cb_lock, irq_flags); + + return 0; +} + +static void dpu_clear_all_irqs(struct dpu_kms *dpu_kms) +{ + if (!dpu_kms || !dpu_kms->hw_intr || + !dpu_kms->hw_intr->ops.clear_all_irqs) + return; + + dpu_kms->hw_intr->ops.clear_all_irqs(dpu_kms->hw_intr); +} + +static void dpu_disable_all_irqs(struct dpu_kms *dpu_kms) +{ + if (!dpu_kms || !dpu_kms->hw_intr || + !dpu_kms->hw_intr->ops.disable_all_irqs) + return; + + dpu_kms->hw_intr->ops.disable_all_irqs(dpu_kms->hw_intr); +} + +#ifdef CONFIG_DEBUG_FS +#define DEFINE_DPU_DEBUGFS_SEQ_FOPS(__prefix) \ +static int __prefix ## _open(struct inode *inode, struct file *file) \ +{ \ + return single_open(file, __prefix ## _show, inode->i_private); \ +} \ +static const struct file_operations __prefix ## _fops = { \ + .owner = THIS_MODULE, \ + .open = __prefix ## _open, \ + .release = single_release, \ + .read = seq_read, \ + .llseek = seq_lseek, \ +} + +static int dpu_debugfs_core_irq_show(struct seq_file *s, void *v) +{ + struct dpu_irq *irq_obj = s->private; + struct dpu_irq_callback *cb; + unsigned long irq_flags; + int i, irq_count, enable_count, cb_count; + + if (!irq_obj || !irq_obj->enable_counts || !irq_obj->irq_cb_tbl) { + DPU_ERROR("invalid parameters\n"); + return 0; + } + + for (i = 0; i < irq_obj->total_irqs; i++) { + spin_lock_irqsave(&irq_obj->cb_lock, irq_flags); + cb_count = 0; + irq_count = atomic_read(&irq_obj->irq_counts[i]); + enable_count = atomic_read(&irq_obj->enable_counts[i]); + list_for_each_entry(cb, &irq_obj->irq_cb_tbl[i], list) + cb_count++; + spin_unlock_irqrestore(&irq_obj->cb_lock, irq_flags); + + if (irq_count || enable_count || cb_count) + seq_printf(s, "idx:%d irq:%d enable:%d cb:%d\n", + i, irq_count, enable_count, cb_count); + } + + return 0; +} + +DEFINE_DPU_DEBUGFS_SEQ_FOPS(dpu_debugfs_core_irq); + +int dpu_debugfs_core_irq_init(struct dpu_kms *dpu_kms, + struct dentry *parent) +{ + dpu_kms->irq_obj.debugfs_file = debugfs_create_file("core_irq", 0600, + parent, &dpu_kms->irq_obj, + &dpu_debugfs_core_irq_fops); + + return 0; +} + +void dpu_debugfs_core_irq_destroy(struct dpu_kms *dpu_kms) +{ + debugfs_remove(dpu_kms->irq_obj.debugfs_file); + dpu_kms->irq_obj.debugfs_file = NULL; +} + +#else +int dpu_debugfs_core_irq_init(struct dpu_kms *dpu_kms, + struct dentry *parent) +{ + return 0; +} + +void dpu_debugfs_core_irq_destroy(struct dpu_kms *dpu_kms) +{ +} +#endif + +void dpu_core_irq_preinstall(struct dpu_kms *dpu_kms) +{ + struct msm_drm_private *priv; + int i; + + if (!dpu_kms) { + DPU_ERROR("invalid dpu_kms\n"); + return; + } else if (!dpu_kms->dev) { + DPU_ERROR("invalid drm device\n"); + return; + } else if (!dpu_kms->dev->dev_private) { + DPU_ERROR("invalid device private\n"); + return; + } + priv = dpu_kms->dev->dev_private; + + pm_runtime_get_sync(&dpu_kms->pdev->dev); + dpu_clear_all_irqs(dpu_kms); + dpu_disable_all_irqs(dpu_kms); + pm_runtime_put_sync(&dpu_kms->pdev->dev); + + spin_lock_init(&dpu_kms->irq_obj.cb_lock); + + /* Create irq callbacks for all possible irq_idx */ + dpu_kms->irq_obj.total_irqs = dpu_kms->hw_intr->irq_idx_tbl_size; + dpu_kms->irq_obj.irq_cb_tbl = kcalloc(dpu_kms->irq_obj.total_irqs, + sizeof(struct list_head), GFP_KERNEL); + dpu_kms->irq_obj.enable_counts = kcalloc(dpu_kms->irq_obj.total_irqs, + sizeof(atomic_t), GFP_KERNEL); + dpu_kms->irq_obj.irq_counts = kcalloc(dpu_kms->irq_obj.total_irqs, + sizeof(atomic_t), GFP_KERNEL); + for (i = 0; i < dpu_kms->irq_obj.total_irqs; i++) { + INIT_LIST_HEAD(&dpu_kms->irq_obj.irq_cb_tbl[i]); + atomic_set(&dpu_kms->irq_obj.enable_counts[i], 0); + atomic_set(&dpu_kms->irq_obj.irq_counts[i], 0); + } +} + +int dpu_core_irq_postinstall(struct dpu_kms *dpu_kms) +{ + return 0; +} + +void dpu_core_irq_uninstall(struct dpu_kms *dpu_kms) +{ + struct msm_drm_private *priv; + int i; + + if (!dpu_kms) { + DPU_ERROR("invalid dpu_kms\n"); + return; + } else if (!dpu_kms->dev) { + DPU_ERROR("invalid drm device\n"); + return; + } else if (!dpu_kms->dev->dev_private) { + DPU_ERROR("invalid device private\n"); + return; + } + priv = dpu_kms->dev->dev_private; + + pm_runtime_get_sync(&dpu_kms->pdev->dev); + for (i = 0; i < dpu_kms->irq_obj.total_irqs; i++) + if (atomic_read(&dpu_kms->irq_obj.enable_counts[i]) || + !list_empty(&dpu_kms->irq_obj.irq_cb_tbl[i])) + DPU_ERROR("irq_idx=%d still enabled/registered\n", i); + + dpu_clear_all_irqs(dpu_kms); + dpu_disable_all_irqs(dpu_kms); + pm_runtime_put_sync(&dpu_kms->pdev->dev); + + kfree(dpu_kms->irq_obj.irq_cb_tbl); + kfree(dpu_kms->irq_obj.enable_counts); + kfree(dpu_kms->irq_obj.irq_counts); + dpu_kms->irq_obj.irq_cb_tbl = NULL; + dpu_kms->irq_obj.enable_counts = NULL; + dpu_kms->irq_obj.irq_counts = NULL; + dpu_kms->irq_obj.total_irqs = 0; +} + +irqreturn_t dpu_core_irq(struct dpu_kms *dpu_kms) +{ + /* + * Read interrupt status from all sources. Interrupt status are + * stored within hw_intr. + * Function will also clear the interrupt status after reading. + * Individual interrupt status bit will only get stored if it + * is enabled. + */ + dpu_kms->hw_intr->ops.get_interrupt_statuses(dpu_kms->hw_intr); + + /* + * Dispatch to HW driver to handle interrupt lookup that is being + * fired. When matching interrupt is located, HW driver will call to + * dpu_core_irq_callback_handler with the irq_idx from the lookup table. + * dpu_core_irq_callback_handler will perform the registered function + * callback, and do the interrupt status clearing once the registered + * callback is finished. + */ + dpu_kms->hw_intr->ops.dispatch_irqs( + dpu_kms->hw_intr, + dpu_core_irq_callback_handler, + dpu_kms); + + return IRQ_HANDLED; +} diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_core_irq.h b/drivers/gpu/drm/msm/disp/dpu1/dpu_core_irq.h new file mode 100644 index 000000000000..5e98bba46af5 --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_core_irq.h @@ -0,0 +1,153 @@ +/* Copyright (c) 2015-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#ifndef __DPU_CORE_IRQ_H__ +#define __DPU_CORE_IRQ_H__ + +#include "dpu_kms.h" +#include "dpu_hw_interrupts.h" + +/** + * dpu_core_irq_preinstall - perform pre-installation of core IRQ handler + * @dpu_kms: DPU handle + * @return: none + */ +void dpu_core_irq_preinstall(struct dpu_kms *dpu_kms); + +/** + * dpu_core_irq_postinstall - perform post-installation of core IRQ handler + * @dpu_kms: DPU handle + * @return: 0 if success; error code otherwise + */ +int dpu_core_irq_postinstall(struct dpu_kms *dpu_kms); + +/** + * dpu_core_irq_uninstall - uninstall core IRQ handler + * @dpu_kms: DPU handle + * @return: none + */ +void dpu_core_irq_uninstall(struct dpu_kms *dpu_kms); + +/** + * dpu_core_irq - core IRQ handler + * @dpu_kms: DPU handle + * @return: interrupt handling status + */ +irqreturn_t dpu_core_irq(struct dpu_kms *dpu_kms); + +/** + * dpu_core_irq_idx_lookup - IRQ helper function for lookup irq_idx from HW + * interrupt mapping table. + * @dpu_kms: DPU handle + * @intr_type: DPU HW interrupt type for lookup + * @instance_idx: DPU HW block instance defined in dpu_hw_mdss.h + * @return: irq_idx or -EINVAL when fail to lookup + */ +int dpu_core_irq_idx_lookup( + struct dpu_kms *dpu_kms, + enum dpu_intr_type intr_type, + uint32_t instance_idx); + +/** + * dpu_core_irq_enable - IRQ helper function for enabling one or more IRQs + * @dpu_kms: DPU handle + * @irq_idxs: Array of irq index + * @irq_count: Number of irq_idx provided in the array + * @return: 0 for success enabling IRQ, otherwise failure + * + * This function increments count on each enable and decrements on each + * disable. Interrupts is enabled if count is 0 before increment. + */ +int dpu_core_irq_enable( + struct dpu_kms *dpu_kms, + int *irq_idxs, + uint32_t irq_count); + +/** + * dpu_core_irq_disable - IRQ helper function for disabling one of more IRQs + * @dpu_kms: DPU handle + * @irq_idxs: Array of irq index + * @irq_count: Number of irq_idx provided in the array + * @return: 0 for success disabling IRQ, otherwise failure + * + * This function increments count on each enable and decrements on each + * disable. Interrupts is disabled if count is 0 after decrement. + */ +int dpu_core_irq_disable( + struct dpu_kms *dpu_kms, + int *irq_idxs, + uint32_t irq_count); + +/** + * dpu_core_irq_read - IRQ helper function for reading IRQ status + * @dpu_kms: DPU handle + * @irq_idx: irq index + * @clear: True to clear the irq after read + * @return: non-zero if irq detected; otherwise no irq detected + */ +u32 dpu_core_irq_read( + struct dpu_kms *dpu_kms, + int irq_idx, + bool clear); + +/** + * dpu_core_irq_register_callback - For registering callback function on IRQ + * interrupt + * @dpu_kms: DPU handle + * @irq_idx: irq index + * @irq_cb: IRQ callback structure, containing callback function + * and argument. Passing NULL for irq_cb will unregister + * the callback for the given irq_idx + * This must exist until un-registration. + * @return: 0 for success registering callback, otherwise failure + * + * This function supports registration of multiple callbacks for each interrupt. + */ +int dpu_core_irq_register_callback( + struct dpu_kms *dpu_kms, + int irq_idx, + struct dpu_irq_callback *irq_cb); + +/** + * dpu_core_irq_unregister_callback - For unregistering callback function on IRQ + * interrupt + * @dpu_kms: DPU handle + * @irq_idx: irq index + * @irq_cb: IRQ callback structure, containing callback function + * and argument. Passing NULL for irq_cb will unregister + * the callback for the given irq_idx + * This must match with registration. + * @return: 0 for success registering callback, otherwise failure + * + * This function supports registration of multiple callbacks for each interrupt. + */ +int dpu_core_irq_unregister_callback( + struct dpu_kms *dpu_kms, + int irq_idx, + struct dpu_irq_callback *irq_cb); + +/** + * dpu_debugfs_core_irq_init - register core irq debugfs + * @dpu_kms: pointer to kms + * @parent: debugfs directory root + * @Return: 0 on success + */ +int dpu_debugfs_core_irq_init(struct dpu_kms *dpu_kms, + struct dentry *parent); + +/** + * dpu_debugfs_core_irq_destroy - deregister core irq debugfs + * @dpu_kms: pointer to kms + */ +void dpu_debugfs_core_irq_destroy(struct dpu_kms *dpu_kms); + +#endif /* __DPU_CORE_IRQ_H__ */ diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_core_perf.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_core_perf.c new file mode 100644 index 000000000000..41c5191f9056 --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_core_perf.c @@ -0,0 +1,637 @@ +/* Copyright (c) 2016-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#define pr_fmt(fmt) "[drm:%s:%d] " fmt, __func__, __LINE__ + +#include <linux/debugfs.h> +#include <linux/errno.h> +#include <linux/mutex.h> +#include <linux/sort.h> +#include <linux/clk.h> +#include <linux/bitmap.h> + +#include "dpu_kms.h" +#include "dpu_trace.h" +#include "dpu_crtc.h" +#include "dpu_core_perf.h" + +#define DPU_PERF_MODE_STRING_SIZE 128 + +/** + * enum dpu_perf_mode - performance tuning mode + * @DPU_PERF_MODE_NORMAL: performance controlled by user mode client + * @DPU_PERF_MODE_MINIMUM: performance bounded by minimum setting + * @DPU_PERF_MODE_FIXED: performance bounded by fixed setting + */ +enum dpu_perf_mode { + DPU_PERF_MODE_NORMAL, + DPU_PERF_MODE_MINIMUM, + DPU_PERF_MODE_FIXED, + DPU_PERF_MODE_MAX +}; + +static struct dpu_kms *_dpu_crtc_get_kms(struct drm_crtc *crtc) +{ + struct msm_drm_private *priv; + + if (!crtc->dev || !crtc->dev->dev_private) { + DPU_ERROR("invalid device\n"); + return NULL; + } + + priv = crtc->dev->dev_private; + if (!priv || !priv->kms) { + DPU_ERROR("invalid kms\n"); + return NULL; + } + + return to_dpu_kms(priv->kms); +} + +static bool _dpu_core_perf_crtc_is_power_on(struct drm_crtc *crtc) +{ + return dpu_crtc_is_enabled(crtc); +} + +static bool _dpu_core_video_mode_intf_connected(struct drm_crtc *crtc) +{ + struct drm_crtc *tmp_crtc; + bool intf_connected = false; + + if (!crtc) + goto end; + + drm_for_each_crtc(tmp_crtc, crtc->dev) { + if ((dpu_crtc_get_intf_mode(tmp_crtc) == INTF_MODE_VIDEO) && + _dpu_core_perf_crtc_is_power_on(tmp_crtc)) { + DPU_DEBUG("video interface connected crtc:%d\n", + tmp_crtc->base.id); + intf_connected = true; + goto end; + } + } + +end: + return intf_connected; +} + +static void _dpu_core_perf_calc_crtc(struct dpu_kms *kms, + struct drm_crtc *crtc, + struct drm_crtc_state *state, + struct dpu_core_perf_params *perf) +{ + struct dpu_crtc_state *dpu_cstate; + int i; + + if (!kms || !kms->catalog || !crtc || !state || !perf) { + DPU_ERROR("invalid parameters\n"); + return; + } + + dpu_cstate = to_dpu_crtc_state(state); + memset(perf, 0, sizeof(struct dpu_core_perf_params)); + + if (!dpu_cstate->bw_control) { + for (i = 0; i < DPU_POWER_HANDLE_DBUS_ID_MAX; i++) { + perf->bw_ctl[i] = kms->catalog->perf.max_bw_high * + 1000ULL; + perf->max_per_pipe_ib[i] = perf->bw_ctl[i]; + } + perf->core_clk_rate = kms->perf.max_core_clk_rate; + } else if (kms->perf.perf_tune.mode == DPU_PERF_MODE_MINIMUM) { + for (i = 0; i < DPU_POWER_HANDLE_DBUS_ID_MAX; i++) { + perf->bw_ctl[i] = 0; + perf->max_per_pipe_ib[i] = 0; + } + perf->core_clk_rate = 0; + } else if (kms->perf.perf_tune.mode == DPU_PERF_MODE_FIXED) { + for (i = 0; i < DPU_POWER_HANDLE_DBUS_ID_MAX; i++) { + perf->bw_ctl[i] = kms->perf.fix_core_ab_vote; + perf->max_per_pipe_ib[i] = kms->perf.fix_core_ib_vote; + } + perf->core_clk_rate = kms->perf.fix_core_clk_rate; + } + + DPU_DEBUG( + "crtc=%d clk_rate=%llu core_ib=%llu core_ab=%llu llcc_ib=%llu llcc_ab=%llu mem_ib=%llu mem_ab=%llu\n", + crtc->base.id, perf->core_clk_rate, + perf->max_per_pipe_ib[DPU_POWER_HANDLE_DBUS_ID_MNOC], + perf->bw_ctl[DPU_POWER_HANDLE_DBUS_ID_MNOC], + perf->max_per_pipe_ib[DPU_POWER_HANDLE_DBUS_ID_LLCC], + perf->bw_ctl[DPU_POWER_HANDLE_DBUS_ID_LLCC], + perf->max_per_pipe_ib[DPU_POWER_HANDLE_DBUS_ID_EBI], + perf->bw_ctl[DPU_POWER_HANDLE_DBUS_ID_EBI]); +} + +int dpu_core_perf_crtc_check(struct drm_crtc *crtc, + struct drm_crtc_state *state) +{ + u32 bw, threshold; + u64 bw_sum_of_intfs = 0; + enum dpu_crtc_client_type curr_client_type; + bool is_video_mode; + struct dpu_crtc_state *dpu_cstate; + struct drm_crtc *tmp_crtc; + struct dpu_kms *kms; + int i; + + if (!crtc || !state) { + DPU_ERROR("invalid crtc\n"); + return -EINVAL; + } + + kms = _dpu_crtc_get_kms(crtc); + if (!kms || !kms->catalog) { + DPU_ERROR("invalid parameters\n"); + return 0; + } + + /* we only need bandwidth check on real-time clients (interfaces) */ + if (dpu_crtc_get_client_type(crtc) == NRT_CLIENT) + return 0; + + dpu_cstate = to_dpu_crtc_state(state); + + /* obtain new values */ + _dpu_core_perf_calc_crtc(kms, crtc, state, &dpu_cstate->new_perf); + + for (i = DPU_POWER_HANDLE_DBUS_ID_MNOC; + i < DPU_POWER_HANDLE_DBUS_ID_MAX; i++) { + bw_sum_of_intfs = dpu_cstate->new_perf.bw_ctl[i]; + curr_client_type = dpu_crtc_get_client_type(crtc); + + drm_for_each_crtc(tmp_crtc, crtc->dev) { + if (_dpu_core_perf_crtc_is_power_on(tmp_crtc) && + (dpu_crtc_get_client_type(tmp_crtc) == + curr_client_type) && + (tmp_crtc != crtc)) { + struct dpu_crtc_state *tmp_cstate = + to_dpu_crtc_state(tmp_crtc->state); + + DPU_DEBUG("crtc:%d bw:%llu ctrl:%d\n", + tmp_crtc->base.id, + tmp_cstate->new_perf.bw_ctl[i], + tmp_cstate->bw_control); + /* + * For bw check only use the bw if the + * atomic property has been already set + */ + if (tmp_cstate->bw_control) + bw_sum_of_intfs += + tmp_cstate->new_perf.bw_ctl[i]; + } + } + + /* convert bandwidth to kb */ + bw = DIV_ROUND_UP_ULL(bw_sum_of_intfs, 1000); + DPU_DEBUG("calculated bandwidth=%uk\n", bw); + + is_video_mode = dpu_crtc_get_intf_mode(crtc) == INTF_MODE_VIDEO; + threshold = (is_video_mode || + _dpu_core_video_mode_intf_connected(crtc)) ? + kms->catalog->perf.max_bw_low : + kms->catalog->perf.max_bw_high; + + DPU_DEBUG("final threshold bw limit = %d\n", threshold); + + if (!dpu_cstate->bw_control) { + DPU_DEBUG("bypass bandwidth check\n"); + } else if (!threshold) { + DPU_ERROR("no bandwidth limits specified\n"); + return -E2BIG; + } else if (bw > threshold) { + DPU_ERROR("exceeds bandwidth: %ukb > %ukb\n", bw, + threshold); + return -E2BIG; + } + } + + return 0; +} + +static int _dpu_core_perf_crtc_update_bus(struct dpu_kms *kms, + struct drm_crtc *crtc, u32 bus_id) +{ + struct dpu_core_perf_params perf = { { 0 } }; + enum dpu_crtc_client_type curr_client_type + = dpu_crtc_get_client_type(crtc); + struct drm_crtc *tmp_crtc; + struct dpu_crtc_state *dpu_cstate; + int ret = 0; + + drm_for_each_crtc(tmp_crtc, crtc->dev) { + if (_dpu_core_perf_crtc_is_power_on(tmp_crtc) && + curr_client_type == + dpu_crtc_get_client_type(tmp_crtc)) { + dpu_cstate = to_dpu_crtc_state(tmp_crtc->state); + + perf.max_per_pipe_ib[bus_id] = + max(perf.max_per_pipe_ib[bus_id], + dpu_cstate->new_perf.max_per_pipe_ib[bus_id]); + + DPU_DEBUG("crtc=%d bus_id=%d bw=%llu\n", + tmp_crtc->base.id, bus_id, + dpu_cstate->new_perf.bw_ctl[bus_id]); + } + } + return ret; +} + +/** + * @dpu_core_perf_crtc_release_bw() - request zero bandwidth + * @crtc - pointer to a crtc + * + * Function checks a state variable for the crtc, if all pending commit + * requests are done, meaning no more bandwidth is needed, release + * bandwidth request. + */ +void dpu_core_perf_crtc_release_bw(struct drm_crtc *crtc) +{ + struct drm_crtc *tmp_crtc; + struct dpu_crtc *dpu_crtc; + struct dpu_crtc_state *dpu_cstate; + struct dpu_kms *kms; + int i; + + if (!crtc) { + DPU_ERROR("invalid crtc\n"); + return; + } + + kms = _dpu_crtc_get_kms(crtc); + if (!kms || !kms->catalog) { + DPU_ERROR("invalid kms\n"); + return; + } + + dpu_crtc = to_dpu_crtc(crtc); + dpu_cstate = to_dpu_crtc_state(crtc->state); + + /* only do this for command mode rt client */ + if (dpu_crtc_get_intf_mode(crtc) != INTF_MODE_CMD) + return; + + /* + * If video interface present, cmd panel bandwidth cannot be + * released. + */ + if (dpu_crtc_get_intf_mode(crtc) == INTF_MODE_CMD) + drm_for_each_crtc(tmp_crtc, crtc->dev) { + if (_dpu_core_perf_crtc_is_power_on(tmp_crtc) && + dpu_crtc_get_intf_mode(tmp_crtc) == + INTF_MODE_VIDEO) + return; + } + + /* Release the bandwidth */ + if (kms->perf.enable_bw_release) { + trace_dpu_cmd_release_bw(crtc->base.id); + DPU_DEBUG("Release BW crtc=%d\n", crtc->base.id); + for (i = 0; i < DPU_POWER_HANDLE_DBUS_ID_MAX; i++) { + dpu_crtc->cur_perf.bw_ctl[i] = 0; + _dpu_core_perf_crtc_update_bus(kms, crtc, i); + } + } +} + +static int _dpu_core_perf_set_core_clk_rate(struct dpu_kms *kms, u64 rate) +{ + struct dss_clk *core_clk = kms->perf.core_clk; + + if (core_clk->max_rate && (rate > core_clk->max_rate)) + rate = core_clk->max_rate; + + core_clk->rate = rate; + return msm_dss_clk_set_rate(core_clk, 1); +} + +static u64 _dpu_core_perf_get_core_clk_rate(struct dpu_kms *kms) +{ + u64 clk_rate = kms->perf.perf_tune.min_core_clk; + struct drm_crtc *crtc; + struct dpu_crtc_state *dpu_cstate; + + drm_for_each_crtc(crtc, kms->dev) { + if (_dpu_core_perf_crtc_is_power_on(crtc)) { + dpu_cstate = to_dpu_crtc_state(crtc->state); + clk_rate = max(dpu_cstate->new_perf.core_clk_rate, + clk_rate); + clk_rate = clk_round_rate(kms->perf.core_clk->clk, + clk_rate); + } + } + + if (kms->perf.perf_tune.mode == DPU_PERF_MODE_FIXED) + clk_rate = kms->perf.fix_core_clk_rate; + + DPU_DEBUG("clk:%llu\n", clk_rate); + + return clk_rate; +} + +int dpu_core_perf_crtc_update(struct drm_crtc *crtc, + int params_changed, bool stop_req) +{ + struct dpu_core_perf_params *new, *old; + int update_bus = 0, update_clk = 0; + u64 clk_rate = 0; + struct dpu_crtc *dpu_crtc; + struct dpu_crtc_state *dpu_cstate; + int i; + struct msm_drm_private *priv; + struct dpu_kms *kms; + int ret; + + if (!crtc) { + DPU_ERROR("invalid crtc\n"); + return -EINVAL; + } + + kms = _dpu_crtc_get_kms(crtc); + if (!kms || !kms->catalog) { + DPU_ERROR("invalid kms\n"); + return -EINVAL; + } + priv = kms->dev->dev_private; + + dpu_crtc = to_dpu_crtc(crtc); + dpu_cstate = to_dpu_crtc_state(crtc->state); + + DPU_DEBUG("crtc:%d stop_req:%d core_clk:%llu\n", + crtc->base.id, stop_req, kms->perf.core_clk_rate); + + old = &dpu_crtc->cur_perf; + new = &dpu_cstate->new_perf; + + if (_dpu_core_perf_crtc_is_power_on(crtc) && !stop_req) { + for (i = 0; i < DPU_POWER_HANDLE_DBUS_ID_MAX; i++) { + /* + * cases for bus bandwidth update. + * 1. new bandwidth vote - "ab or ib vote" is higher + * than current vote for update request. + * 2. new bandwidth vote - "ab or ib vote" is lower + * than current vote at end of commit or stop. + */ + if ((params_changed && ((new->bw_ctl[i] > + old->bw_ctl[i]) || + (new->max_per_pipe_ib[i] > + old->max_per_pipe_ib[i]))) || + (!params_changed && ((new->bw_ctl[i] < + old->bw_ctl[i]) || + (new->max_per_pipe_ib[i] < + old->max_per_pipe_ib[i])))) { + DPU_DEBUG( + "crtc=%d p=%d new_bw=%llu,old_bw=%llu\n", + crtc->base.id, params_changed, + new->bw_ctl[i], old->bw_ctl[i]); + old->bw_ctl[i] = new->bw_ctl[i]; + old->max_per_pipe_ib[i] = + new->max_per_pipe_ib[i]; + update_bus |= BIT(i); + } + } + + if ((params_changed && + (new->core_clk_rate > old->core_clk_rate)) || + (!params_changed && + (new->core_clk_rate < old->core_clk_rate))) { + old->core_clk_rate = new->core_clk_rate; + update_clk = 1; + } + } else { + DPU_DEBUG("crtc=%d disable\n", crtc->base.id); + memset(old, 0, sizeof(*old)); + memset(new, 0, sizeof(*new)); + update_bus = ~0; + update_clk = 1; + } + trace_dpu_perf_crtc_update(crtc->base.id, + new->bw_ctl[DPU_POWER_HANDLE_DBUS_ID_MNOC], + new->bw_ctl[DPU_POWER_HANDLE_DBUS_ID_LLCC], + new->bw_ctl[DPU_POWER_HANDLE_DBUS_ID_EBI], + new->core_clk_rate, stop_req, + update_bus, update_clk); + + for (i = 0; i < DPU_POWER_HANDLE_DBUS_ID_MAX; i++) { + if (update_bus & BIT(i)) { + ret = _dpu_core_perf_crtc_update_bus(kms, crtc, i); + if (ret) { + DPU_ERROR("crtc-%d: failed to update bw vote for bus-%d\n", + crtc->base.id, i); + return ret; + } + } + } + + /* + * Update the clock after bandwidth vote to ensure + * bandwidth is available before clock rate is increased. + */ + if (update_clk) { + clk_rate = _dpu_core_perf_get_core_clk_rate(kms); + + trace_dpu_core_perf_update_clk(kms->dev, stop_req, clk_rate); + + ret = _dpu_core_perf_set_core_clk_rate(kms, clk_rate); + if (ret) { + DPU_ERROR("failed to set %s clock rate %llu\n", + kms->perf.core_clk->clk_name, clk_rate); + return ret; + } + + kms->perf.core_clk_rate = clk_rate; + DPU_DEBUG("update clk rate = %lld HZ\n", clk_rate); + } + return 0; +} + +#ifdef CONFIG_DEBUG_FS + +static ssize_t _dpu_core_perf_mode_write(struct file *file, + const char __user *user_buf, size_t count, loff_t *ppos) +{ + struct dpu_core_perf *perf = file->private_data; + struct dpu_perf_cfg *cfg = &perf->catalog->perf; + u32 perf_mode = 0; + char buf[10]; + + if (!perf) + return -ENODEV; + + if (count >= sizeof(buf)) + return -EFAULT; + + if (copy_from_user(buf, user_buf, count)) + return -EFAULT; + + buf[count] = 0; /* end of string */ + + if (kstrtouint(buf, 0, &perf_mode)) + return -EFAULT; + + if (perf_mode >= DPU_PERF_MODE_MAX) + return -EFAULT; + + if (perf_mode == DPU_PERF_MODE_FIXED) { + DRM_INFO("fix performance mode\n"); + } else if (perf_mode == DPU_PERF_MODE_MINIMUM) { + /* run the driver with max clk and BW vote */ + perf->perf_tune.min_core_clk = perf->max_core_clk_rate; + perf->perf_tune.min_bus_vote = + (u64) cfg->max_bw_high * 1000; + DRM_INFO("minimum performance mode\n"); + } else if (perf_mode == DPU_PERF_MODE_NORMAL) { + /* reset the perf tune params to 0 */ + perf->perf_tune.min_core_clk = 0; + perf->perf_tune.min_bus_vote = 0; + DRM_INFO("normal performance mode\n"); + } + perf->perf_tune.mode = perf_mode; + + return count; +} + +static ssize_t _dpu_core_perf_mode_read(struct file *file, + char __user *buff, size_t count, loff_t *ppos) +{ + struct dpu_core_perf *perf = file->private_data; + int len = 0; + char buf[DPU_PERF_MODE_STRING_SIZE] = {'\0'}; + + if (!perf) + return -ENODEV; + + if (*ppos) + return 0; /* the end */ + + len = snprintf(buf, sizeof(buf), + "mode %d min_mdp_clk %llu min_bus_vote %llu\n", + perf->perf_tune.mode, + perf->perf_tune.min_core_clk, + perf->perf_tune.min_bus_vote); + if (len < 0 || len >= sizeof(buf)) + return 0; + + if ((count < sizeof(buf)) || copy_to_user(buff, buf, len)) + return -EFAULT; + + *ppos += len; /* increase offset */ + + return len; +} + +static const struct file_operations dpu_core_perf_mode_fops = { + .open = simple_open, + .read = _dpu_core_perf_mode_read, + .write = _dpu_core_perf_mode_write, +}; + +static void dpu_core_perf_debugfs_destroy(struct dpu_core_perf *perf) +{ + debugfs_remove_recursive(perf->debugfs_root); + perf->debugfs_root = NULL; +} + +int dpu_core_perf_debugfs_init(struct dpu_core_perf *perf, + struct dentry *parent) +{ + struct dpu_mdss_cfg *catalog = perf->catalog; + struct msm_drm_private *priv; + struct dpu_kms *dpu_kms; + + priv = perf->dev->dev_private; + if (!priv || !priv->kms) { + DPU_ERROR("invalid KMS reference\n"); + return -EINVAL; + } + + dpu_kms = to_dpu_kms(priv->kms); + + perf->debugfs_root = debugfs_create_dir("core_perf", parent); + if (!perf->debugfs_root) { + DPU_ERROR("failed to create core perf debugfs\n"); + return -EINVAL; + } + + debugfs_create_u64("max_core_clk_rate", 0600, perf->debugfs_root, + &perf->max_core_clk_rate); + debugfs_create_u64("core_clk_rate", 0600, perf->debugfs_root, + &perf->core_clk_rate); + debugfs_create_u32("enable_bw_release", 0600, perf->debugfs_root, + (u32 *)&perf->enable_bw_release); + debugfs_create_u32("threshold_low", 0600, perf->debugfs_root, + (u32 *)&catalog->perf.max_bw_low); + debugfs_create_u32("threshold_high", 0600, perf->debugfs_root, + (u32 *)&catalog->perf.max_bw_high); + debugfs_create_u32("min_core_ib", 0600, perf->debugfs_root, + (u32 *)&catalog->perf.min_core_ib); + debugfs_create_u32("min_llcc_ib", 0600, perf->debugfs_root, + (u32 *)&catalog->perf.min_llcc_ib); + debugfs_create_u32("min_dram_ib", 0600, perf->debugfs_root, + (u32 *)&catalog->perf.min_dram_ib); + debugfs_create_file("perf_mode", 0600, perf->debugfs_root, + (u32 *)perf, &dpu_core_perf_mode_fops); + debugfs_create_u64("fix_core_clk_rate", 0600, perf->debugfs_root, + &perf->fix_core_clk_rate); + debugfs_create_u64("fix_core_ib_vote", 0600, perf->debugfs_root, + &perf->fix_core_ib_vote); + debugfs_create_u64("fix_core_ab_vote", 0600, perf->debugfs_root, + &perf->fix_core_ab_vote); + + return 0; +} +#else +static void dpu_core_perf_debugfs_destroy(struct dpu_core_perf *perf) +{ +} + +int dpu_core_perf_debugfs_init(struct dpu_core_perf *perf, + struct dentry *parent) +{ + return 0; +} +#endif + +void dpu_core_perf_destroy(struct dpu_core_perf *perf) +{ + if (!perf) { + DPU_ERROR("invalid parameters\n"); + return; + } + + dpu_core_perf_debugfs_destroy(perf); + perf->max_core_clk_rate = 0; + perf->core_clk = NULL; + perf->phandle = NULL; + perf->catalog = NULL; + perf->dev = NULL; +} + +int dpu_core_perf_init(struct dpu_core_perf *perf, + struct drm_device *dev, + struct dpu_mdss_cfg *catalog, + struct dpu_power_handle *phandle, + struct dss_clk *core_clk) +{ + perf->dev = dev; + perf->catalog = catalog; + perf->phandle = phandle; + perf->core_clk = core_clk; + + perf->max_core_clk_rate = core_clk->max_rate; + if (!perf->max_core_clk_rate) { + DPU_DEBUG("optional max core clk rate, use default\n"); + perf->max_core_clk_rate = DPU_PERF_DEFAULT_MAX_CORE_CLK_RATE; + } + + return 0; +} diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_core_perf.h b/drivers/gpu/drm/msm/disp/dpu1/dpu_core_perf.h new file mode 100644 index 000000000000..fbcbe0c7527a --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_core_perf.h @@ -0,0 +1,133 @@ +/* Copyright (c) 2016-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#ifndef _DPU_CORE_PERF_H_ +#define _DPU_CORE_PERF_H_ + +#include <linux/types.h> +#include <linux/dcache.h> +#include <linux/mutex.h> +#include <drm/drm_crtc.h> + +#include "dpu_hw_catalog.h" +#include "dpu_power_handle.h" + +#define DPU_PERF_DEFAULT_MAX_CORE_CLK_RATE 412500000 + +/** + * struct dpu_core_perf_params - definition of performance parameters + * @max_per_pipe_ib: maximum instantaneous bandwidth request + * @bw_ctl: arbitrated bandwidth request + * @core_clk_rate: core clock rate request + */ +struct dpu_core_perf_params { + u64 max_per_pipe_ib[DPU_POWER_HANDLE_DBUS_ID_MAX]; + u64 bw_ctl[DPU_POWER_HANDLE_DBUS_ID_MAX]; + u64 core_clk_rate; +}; + +/** + * struct dpu_core_perf_tune - definition of performance tuning control + * @mode: performance mode + * @min_core_clk: minimum core clock + * @min_bus_vote: minimum bus vote + */ +struct dpu_core_perf_tune { + u32 mode; + u64 min_core_clk; + u64 min_bus_vote; +}; + +/** + * struct dpu_core_perf - definition of core performance context + * @dev: Pointer to drm device + * @debugfs_root: top level debug folder + * @catalog: Pointer to catalog configuration + * @phandle: Pointer to power handler + * @core_clk: Pointer to core clock structure + * @core_clk_rate: current core clock rate + * @max_core_clk_rate: maximum allowable core clock rate + * @perf_tune: debug control for performance tuning + * @enable_bw_release: debug control for bandwidth release + * @fix_core_clk_rate: fixed core clock request in Hz used in mode 2 + * @fix_core_ib_vote: fixed core ib vote in bps used in mode 2 + * @fix_core_ab_vote: fixed core ab vote in bps used in mode 2 + */ +struct dpu_core_perf { + struct drm_device *dev; + struct dentry *debugfs_root; + struct dpu_mdss_cfg *catalog; + struct dpu_power_handle *phandle; + struct dss_clk *core_clk; + u64 core_clk_rate; + u64 max_core_clk_rate; + struct dpu_core_perf_tune perf_tune; + u32 enable_bw_release; + u64 fix_core_clk_rate; + u64 fix_core_ib_vote; + u64 fix_core_ab_vote; +}; + +/** + * dpu_core_perf_crtc_check - validate performance of the given crtc state + * @crtc: Pointer to crtc + * @state: Pointer to new crtc state + * return: zero if success, or error code otherwise + */ +int dpu_core_perf_crtc_check(struct drm_crtc *crtc, + struct drm_crtc_state *state); + +/** + * dpu_core_perf_crtc_update - update performance of the given crtc + * @crtc: Pointer to crtc + * @params_changed: true if crtc parameters are modified + * @stop_req: true if this is a stop request + * return: zero if success, or error code otherwise + */ +int dpu_core_perf_crtc_update(struct drm_crtc *crtc, + int params_changed, bool stop_req); + +/** + * dpu_core_perf_crtc_release_bw - release bandwidth of the given crtc + * @crtc: Pointer to crtc + */ +void dpu_core_perf_crtc_release_bw(struct drm_crtc *crtc); + +/** + * dpu_core_perf_destroy - destroy the given core performance context + * @perf: Pointer to core performance context + */ +void dpu_core_perf_destroy(struct dpu_core_perf *perf); + +/** + * dpu_core_perf_init - initialize the given core performance context + * @perf: Pointer to core performance context + * @dev: Pointer to drm device + * @catalog: Pointer to catalog + * @phandle: Pointer to power handle + * @core_clk: pointer to core clock + */ +int dpu_core_perf_init(struct dpu_core_perf *perf, + struct drm_device *dev, + struct dpu_mdss_cfg *catalog, + struct dpu_power_handle *phandle, + struct dss_clk *core_clk); + +/** + * dpu_core_perf_debugfs_init - initialize debugfs for core performance context + * @perf: Pointer to core performance context + * @debugfs_parent: Pointer to parent debugfs + */ +int dpu_core_perf_debugfs_init(struct dpu_core_perf *perf, + struct dentry *parent); + +#endif /* _DPU_CORE_PERF_H_ */ diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_crtc.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_crtc.c new file mode 100644 index 000000000000..80cbf75bc2ff --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_crtc.c @@ -0,0 +1,2138 @@ +/* + * Copyright (c) 2014-2018 The Linux Foundation. All rights reserved. + * Copyright (C) 2013 Red Hat + * Author: Rob Clark <robdclark@gmail.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published by + * the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program. If not, see <http://www.gnu.org/licenses/>. + */ + +#define pr_fmt(fmt) "[drm:%s:%d] " fmt, __func__, __LINE__ +#include <linux/sort.h> +#include <linux/debugfs.h> +#include <linux/ktime.h> +#include <drm/drm_mode.h> +#include <drm/drm_crtc.h> +#include <drm/drm_crtc_helper.h> +#include <drm/drm_flip_work.h> +#include <drm/drm_rect.h> + +#include "dpu_kms.h" +#include "dpu_hw_lm.h" +#include "dpu_hw_ctl.h" +#include "dpu_crtc.h" +#include "dpu_plane.h" +#include "dpu_encoder.h" +#include "dpu_vbif.h" +#include "dpu_power_handle.h" +#include "dpu_core_perf.h" +#include "dpu_trace.h" + +#define DPU_DRM_BLEND_OP_NOT_DEFINED 0 +#define DPU_DRM_BLEND_OP_OPAQUE 1 +#define DPU_DRM_BLEND_OP_PREMULTIPLIED 2 +#define DPU_DRM_BLEND_OP_COVERAGE 3 +#define DPU_DRM_BLEND_OP_MAX 4 + +/* layer mixer index on dpu_crtc */ +#define LEFT_MIXER 0 +#define RIGHT_MIXER 1 + +#define MISR_BUFF_SIZE 256 + +static inline struct dpu_kms *_dpu_crtc_get_kms(struct drm_crtc *crtc) +{ + struct msm_drm_private *priv; + + if (!crtc || !crtc->dev || !crtc->dev->dev_private) { + DPU_ERROR("invalid crtc\n"); + return NULL; + } + priv = crtc->dev->dev_private; + if (!priv || !priv->kms) { + DPU_ERROR("invalid kms\n"); + return NULL; + } + + return to_dpu_kms(priv->kms); +} + +static inline int _dpu_crtc_power_enable(struct dpu_crtc *dpu_crtc, bool enable) +{ + struct drm_crtc *crtc; + struct msm_drm_private *priv; + struct dpu_kms *dpu_kms; + + if (!dpu_crtc) { + DPU_ERROR("invalid dpu crtc\n"); + return -EINVAL; + } + + crtc = &dpu_crtc->base; + if (!crtc->dev || !crtc->dev->dev_private) { + DPU_ERROR("invalid drm device\n"); + return -EINVAL; + } + + priv = crtc->dev->dev_private; + if (!priv->kms) { + DPU_ERROR("invalid kms\n"); + return -EINVAL; + } + + dpu_kms = to_dpu_kms(priv->kms); + + if (enable) + pm_runtime_get_sync(&dpu_kms->pdev->dev); + else + pm_runtime_put_sync(&dpu_kms->pdev->dev); + + return 0; +} + +/** + * _dpu_crtc_rp_to_crtc - get crtc from resource pool object + * @rp: Pointer to resource pool + * return: Pointer to drm crtc if success; null otherwise + */ +static struct drm_crtc *_dpu_crtc_rp_to_crtc(struct dpu_crtc_respool *rp) +{ + if (!rp) + return NULL; + + return container_of(rp, struct dpu_crtc_state, rp)->base.crtc; +} + +/** + * _dpu_crtc_rp_reclaim - reclaim unused, or all if forced, resources in pool + * @rp: Pointer to resource pool + * @force: True to reclaim all resources; otherwise, reclaim only unused ones + * return: None + */ +static void _dpu_crtc_rp_reclaim(struct dpu_crtc_respool *rp, bool force) +{ + struct dpu_crtc_res *res, *next; + struct drm_crtc *crtc; + + crtc = _dpu_crtc_rp_to_crtc(rp); + if (!crtc) { + DPU_ERROR("invalid crtc\n"); + return; + } + + DPU_DEBUG("crtc%d.%u %s\n", crtc->base.id, rp->sequence_id, + force ? "destroy" : "free_unused"); + + list_for_each_entry_safe(res, next, &rp->res_list, list) { + if (!force && !(res->flags & DPU_CRTC_RES_FLAG_FREE)) + continue; + DPU_DEBUG("crtc%d.%u reclaim res:0x%x/0x%llx/%pK/%d\n", + crtc->base.id, rp->sequence_id, + res->type, res->tag, res->val, + atomic_read(&res->refcount)); + list_del(&res->list); + if (res->ops.put) + res->ops.put(res->val); + kfree(res); + } +} + +/** + * _dpu_crtc_rp_free_unused - free unused resource in pool + * @rp: Pointer to resource pool + * return: none + */ +static void _dpu_crtc_rp_free_unused(struct dpu_crtc_respool *rp) +{ + mutex_lock(rp->rp_lock); + _dpu_crtc_rp_reclaim(rp, false); + mutex_unlock(rp->rp_lock); +} + +/** + * _dpu_crtc_rp_destroy - destroy resource pool + * @rp: Pointer to resource pool + * return: None + */ +static void _dpu_crtc_rp_destroy(struct dpu_crtc_respool *rp) +{ + mutex_lock(rp->rp_lock); + list_del_init(&rp->rp_list); + _dpu_crtc_rp_reclaim(rp, true); + mutex_unlock(rp->rp_lock); +} + +/** + * _dpu_crtc_hw_blk_get - get callback for hardware block + * @val: Resource handle + * @type: Resource type + * @tag: Search tag for given resource + * return: Resource handle + */ +static void *_dpu_crtc_hw_blk_get(void *val, u32 type, u64 tag) +{ + DPU_DEBUG("res:%d/0x%llx/%pK\n", type, tag, val); + return dpu_hw_blk_get(val, type, tag); +} + +/** + * _dpu_crtc_hw_blk_put - put callback for hardware block + * @val: Resource handle + * return: None + */ +static void _dpu_crtc_hw_blk_put(void *val) +{ + DPU_DEBUG("res://%pK\n", val); + dpu_hw_blk_put(val); +} + +/** + * _dpu_crtc_rp_duplicate - duplicate resource pool and reset reference count + * @rp: Pointer to original resource pool + * @dup_rp: Pointer to duplicated resource pool + * return: None + */ +static void _dpu_crtc_rp_duplicate(struct dpu_crtc_respool *rp, + struct dpu_crtc_respool *dup_rp) +{ + struct dpu_crtc_res *res, *dup_res; + struct drm_crtc *crtc; + + if (!rp || !dup_rp || !rp->rp_head) { + DPU_ERROR("invalid resource pool\n"); + return; + } + + crtc = _dpu_crtc_rp_to_crtc(rp); + if (!crtc) { + DPU_ERROR("invalid crtc\n"); + return; + } + + DPU_DEBUG("crtc%d.%u duplicate\n", crtc->base.id, rp->sequence_id); + + mutex_lock(rp->rp_lock); + dup_rp->sequence_id = rp->sequence_id + 1; + INIT_LIST_HEAD(&dup_rp->res_list); + dup_rp->ops = rp->ops; + list_for_each_entry(res, &rp->res_list, list) { + dup_res = kzalloc(sizeof(struct dpu_crtc_res), GFP_KERNEL); + if (!dup_res) { + mutex_unlock(rp->rp_lock); + return; + } + INIT_LIST_HEAD(&dup_res->list); + atomic_set(&dup_res->refcount, 0); + dup_res->type = res->type; + dup_res->tag = res->tag; + dup_res->val = res->val; + dup_res->ops = res->ops; + dup_res->flags = DPU_CRTC_RES_FLAG_FREE; + DPU_DEBUG("crtc%d.%u dup res:0x%x/0x%llx/%pK/%d\n", + crtc->base.id, dup_rp->sequence_id, + dup_res->type, dup_res->tag, dup_res->val, + atomic_read(&dup_res->refcount)); + list_add_tail(&dup_res->list, &dup_rp->res_list); + if (dup_res->ops.get) + dup_res->ops.get(dup_res->val, 0, -1); + } + + dup_rp->rp_lock = rp->rp_lock; + dup_rp->rp_head = rp->rp_head; + INIT_LIST_HEAD(&dup_rp->rp_list); + list_add_tail(&dup_rp->rp_list, rp->rp_head); + mutex_unlock(rp->rp_lock); +} + +/** + * _dpu_crtc_rp_reset - reset resource pool after allocation + * @rp: Pointer to original resource pool + * @rp_lock: Pointer to serialization resource pool lock + * @rp_head: Pointer to crtc resource pool head + * return: None + */ +static void _dpu_crtc_rp_reset(struct dpu_crtc_respool *rp, + struct mutex *rp_lock, struct list_head *rp_head) +{ + if (!rp || !rp_lock || !rp_head) { + DPU_ERROR("invalid resource pool\n"); + return; + } + + mutex_lock(rp_lock); + rp->rp_lock = rp_lock; + rp->rp_head = rp_head; + INIT_LIST_HEAD(&rp->rp_list); + rp->sequence_id = 0; + INIT_LIST_HEAD(&rp->res_list); + rp->ops.get = _dpu_crtc_hw_blk_get; + rp->ops.put = _dpu_crtc_hw_blk_put; + list_add_tail(&rp->rp_list, rp->rp_head); + mutex_unlock(rp_lock); +} + +static void dpu_crtc_destroy(struct drm_crtc *crtc) +{ + struct dpu_crtc *dpu_crtc = to_dpu_crtc(crtc); + + DPU_DEBUG("\n"); + + if (!crtc) + return; + + dpu_crtc->phandle = NULL; + + drm_crtc_cleanup(crtc); + mutex_destroy(&dpu_crtc->crtc_lock); + kfree(dpu_crtc); +} + +static void _dpu_crtc_setup_blend_cfg(struct dpu_crtc_mixer *mixer, + struct dpu_plane_state *pstate) +{ + struct dpu_hw_mixer *lm = mixer->hw_lm; + + /* default to opaque blending */ + lm->ops.setup_blend_config(lm, pstate->stage, 0XFF, 0, + DPU_BLEND_FG_ALPHA_FG_CONST | + DPU_BLEND_BG_ALPHA_BG_CONST); +} + +static void _dpu_crtc_program_lm_output_roi(struct drm_crtc *crtc) +{ + struct dpu_crtc *dpu_crtc; + struct dpu_crtc_state *crtc_state; + int lm_idx, lm_horiz_position; + + dpu_crtc = to_dpu_crtc(crtc); + crtc_state = to_dpu_crtc_state(crtc->state); + + lm_horiz_position = 0; + for (lm_idx = 0; lm_idx < dpu_crtc->num_mixers; lm_idx++) { + const struct drm_rect *lm_roi = &crtc_state->lm_bounds[lm_idx]; + struct dpu_hw_mixer *hw_lm = dpu_crtc->mixers[lm_idx].hw_lm; + struct dpu_hw_mixer_cfg cfg; + + if (!lm_roi || !drm_rect_visible(lm_roi)) + continue; + + cfg.out_width = drm_rect_width(lm_roi); + cfg.out_height = drm_rect_height(lm_roi); + cfg.right_mixer = lm_horiz_position++; + cfg.flags = 0; + hw_lm->ops.setup_mixer_out(hw_lm, &cfg); + } +} + +static void _dpu_crtc_blend_setup_mixer(struct drm_crtc *crtc, + struct dpu_crtc *dpu_crtc, struct dpu_crtc_mixer *mixer) +{ + struct drm_plane *plane; + struct drm_framebuffer *fb; + struct drm_plane_state *state; + struct dpu_crtc_state *cstate; + struct dpu_plane_state *pstate = NULL; + struct dpu_format *format; + struct dpu_hw_ctl *ctl; + struct dpu_hw_mixer *lm; + struct dpu_hw_stage_cfg *stage_cfg; + + u32 flush_mask; + uint32_t stage_idx, lm_idx; + int zpos_cnt[DPU_STAGE_MAX + 1] = { 0 }; + bool bg_alpha_enable = false; + + if (!dpu_crtc || !mixer) { + DPU_ERROR("invalid dpu_crtc or mixer\n"); + return; + } + + ctl = mixer->hw_ctl; + lm = mixer->hw_lm; + stage_cfg = &dpu_crtc->stage_cfg; + cstate = to_dpu_crtc_state(crtc->state); + + drm_atomic_crtc_for_each_plane(plane, crtc) { + state = plane->state; + if (!state) + continue; + + pstate = to_dpu_plane_state(state); + fb = state->fb; + + dpu_plane_get_ctl_flush(plane, ctl, &flush_mask); + + DPU_DEBUG("crtc %d stage:%d - plane %d sspp %d fb %d\n", + crtc->base.id, + pstate->stage, + plane->base.id, + dpu_plane_pipe(plane) - SSPP_VIG0, + state->fb ? state->fb->base.id : -1); + + format = to_dpu_format(msm_framebuffer_format(pstate->base.fb)); + if (!format) { + DPU_ERROR("invalid format\n"); + return; + } + + if (pstate->stage == DPU_STAGE_BASE && format->alpha_enable) + bg_alpha_enable = true; + + stage_idx = zpos_cnt[pstate->stage]++; + stage_cfg->stage[pstate->stage][stage_idx] = + dpu_plane_pipe(plane); + stage_cfg->multirect_index[pstate->stage][stage_idx] = + pstate->multirect_index; + + trace_dpu_crtc_setup_mixer(DRMID(crtc), DRMID(plane), + state, pstate, stage_idx, + dpu_plane_pipe(plane) - SSPP_VIG0, + format->base.pixel_format, + fb ? fb->modifier : 0); + + /* blend config update */ + for (lm_idx = 0; lm_idx < dpu_crtc->num_mixers; lm_idx++) { + _dpu_crtc_setup_blend_cfg(mixer + lm_idx, pstate); + + mixer[lm_idx].flush_mask |= flush_mask; + + if (bg_alpha_enable && !format->alpha_enable) + mixer[lm_idx].mixer_op_mode = 0; + else + mixer[lm_idx].mixer_op_mode |= + 1 << pstate->stage; + } + } + + _dpu_crtc_program_lm_output_roi(crtc); +} + +/** + * _dpu_crtc_blend_setup - configure crtc mixers + * @crtc: Pointer to drm crtc structure + */ +static void _dpu_crtc_blend_setup(struct drm_crtc *crtc) +{ + struct dpu_crtc *dpu_crtc; + struct dpu_crtc_state *dpu_crtc_state; + struct dpu_crtc_mixer *mixer; + struct dpu_hw_ctl *ctl; + struct dpu_hw_mixer *lm; + + int i; + + if (!crtc) + return; + + dpu_crtc = to_dpu_crtc(crtc); + dpu_crtc_state = to_dpu_crtc_state(crtc->state); + mixer = dpu_crtc->mixers; + + DPU_DEBUG("%s\n", dpu_crtc->name); + + if (dpu_crtc->num_mixers > CRTC_DUAL_MIXERS) { + DPU_ERROR("invalid number mixers: %d\n", dpu_crtc->num_mixers); + return; + } + + for (i = 0; i < dpu_crtc->num_mixers; i++) { + if (!mixer[i].hw_lm || !mixer[i].hw_ctl) { + DPU_ERROR("invalid lm or ctl assigned to mixer\n"); + return; + } + mixer[i].mixer_op_mode = 0; + mixer[i].flush_mask = 0; + if (mixer[i].hw_ctl->ops.clear_all_blendstages) + mixer[i].hw_ctl->ops.clear_all_blendstages( + mixer[i].hw_ctl); + } + + /* initialize stage cfg */ + memset(&dpu_crtc->stage_cfg, 0, sizeof(struct dpu_hw_stage_cfg)); + + _dpu_crtc_blend_setup_mixer(crtc, dpu_crtc, mixer); + + for (i = 0; i < dpu_crtc->num_mixers; i++) { + ctl = mixer[i].hw_ctl; + lm = mixer[i].hw_lm; + + lm->ops.setup_alpha_out(lm, mixer[i].mixer_op_mode); + + mixer[i].flush_mask |= ctl->ops.get_bitmask_mixer(ctl, + mixer[i].hw_lm->idx); + + /* stage config flush mask */ + ctl->ops.update_pending_flush(ctl, mixer[i].flush_mask); + + DPU_DEBUG("lm %d, op_mode 0x%X, ctl %d, flush mask 0x%x\n", + mixer[i].hw_lm->idx - LM_0, + mixer[i].mixer_op_mode, + ctl->idx - CTL_0, + mixer[i].flush_mask); + + ctl->ops.setup_blendstage(ctl, mixer[i].hw_lm->idx, + &dpu_crtc->stage_cfg); + } +} + +/** + * _dpu_crtc_complete_flip - signal pending page_flip events + * Any pending vblank events are added to the vblank_event_list + * so that the next vblank interrupt shall signal them. + * However PAGE_FLIP events are not handled through the vblank_event_list. + * This API signals any pending PAGE_FLIP events requested through + * DRM_IOCTL_MODE_PAGE_FLIP and are cached in the dpu_crtc->event. + * @crtc: Pointer to drm crtc structure + */ +static void _dpu_crtc_complete_flip(struct drm_crtc *crtc) +{ + struct dpu_crtc *dpu_crtc = to_dpu_crtc(crtc); + struct drm_device *dev = crtc->dev; + unsigned long flags; + + spin_lock_irqsave(&dev->event_lock, flags); + if (dpu_crtc->event) { + DRM_DEBUG_VBL("%s: send event: %pK\n", dpu_crtc->name, + dpu_crtc->event); + trace_dpu_crtc_complete_flip(DRMID(crtc)); + drm_crtc_send_vblank_event(crtc, dpu_crtc->event); + dpu_crtc->event = NULL; + } + spin_unlock_irqrestore(&dev->event_lock, flags); +} + +enum dpu_intf_mode dpu_crtc_get_intf_mode(struct drm_crtc *crtc) +{ + struct drm_encoder *encoder; + + if (!crtc || !crtc->dev) { + DPU_ERROR("invalid crtc\n"); + return INTF_MODE_NONE; + } + + drm_for_each_encoder(encoder, crtc->dev) + if (encoder->crtc == crtc) + return dpu_encoder_get_intf_mode(encoder); + + return INTF_MODE_NONE; +} + +static void dpu_crtc_vblank_cb(void *data) +{ + struct drm_crtc *crtc = (struct drm_crtc *)data; + struct dpu_crtc *dpu_crtc = to_dpu_crtc(crtc); + + /* keep statistics on vblank callback - with auto reset via debugfs */ + if (ktime_compare(dpu_crtc->vblank_cb_time, ktime_set(0, 0)) == 0) + dpu_crtc->vblank_cb_time = ktime_get(); + else + dpu_crtc->vblank_cb_count++; + _dpu_crtc_complete_flip(crtc); + drm_crtc_handle_vblank(crtc); + trace_dpu_crtc_vblank_cb(DRMID(crtc)); +} + +static void dpu_crtc_frame_event_work(struct kthread_work *work) +{ + struct msm_drm_private *priv; + struct dpu_crtc_frame_event *fevent; + struct drm_crtc *crtc; + struct dpu_crtc *dpu_crtc; + struct dpu_kms *dpu_kms; + unsigned long flags; + bool frame_done = false; + + if (!work) { + DPU_ERROR("invalid work handle\n"); + return; + } + + fevent = container_of(work, struct dpu_crtc_frame_event, work); + if (!fevent->crtc || !fevent->crtc->state) { + DPU_ERROR("invalid crtc\n"); + return; + } + + crtc = fevent->crtc; + dpu_crtc = to_dpu_crtc(crtc); + + dpu_kms = _dpu_crtc_get_kms(crtc); + if (!dpu_kms) { + DPU_ERROR("invalid kms handle\n"); + return; + } + priv = dpu_kms->dev->dev_private; + DPU_ATRACE_BEGIN("crtc_frame_event"); + + DRM_DEBUG_KMS("crtc%d event:%u ts:%lld\n", crtc->base.id, fevent->event, + ktime_to_ns(fevent->ts)); + + if (fevent->event & (DPU_ENCODER_FRAME_EVENT_DONE + | DPU_ENCODER_FRAME_EVENT_ERROR + | DPU_ENCODER_FRAME_EVENT_PANEL_DEAD)) { + + if (atomic_read(&dpu_crtc->frame_pending) < 1) { + /* this should not happen */ + DRM_ERROR("crtc%d ev:%u ts:%lld frame_pending:%d\n", + crtc->base.id, + fevent->event, + ktime_to_ns(fevent->ts), + atomic_read(&dpu_crtc->frame_pending)); + } else if (atomic_dec_return(&dpu_crtc->frame_pending) == 0) { + /* release bandwidth and other resources */ + trace_dpu_crtc_frame_event_done(DRMID(crtc), + fevent->event); + dpu_core_perf_crtc_release_bw(crtc); + } else { + trace_dpu_crtc_frame_event_more_pending(DRMID(crtc), + fevent->event); + } + + if (fevent->event & DPU_ENCODER_FRAME_EVENT_DONE) + dpu_core_perf_crtc_update(crtc, 0, false); + + if (fevent->event & (DPU_ENCODER_FRAME_EVENT_DONE + | DPU_ENCODER_FRAME_EVENT_ERROR)) + frame_done = true; + } + + if (fevent->event & DPU_ENCODER_FRAME_EVENT_PANEL_DEAD) + DPU_ERROR("crtc%d ts:%lld received panel dead event\n", + crtc->base.id, ktime_to_ns(fevent->ts)); + + if (frame_done) + complete_all(&dpu_crtc->frame_done_comp); + + spin_lock_irqsave(&dpu_crtc->spin_lock, flags); + list_add_tail(&fevent->list, &dpu_crtc->frame_event_list); + spin_unlock_irqrestore(&dpu_crtc->spin_lock, flags); + DPU_ATRACE_END("crtc_frame_event"); +} + +/* + * dpu_crtc_frame_event_cb - crtc frame event callback API. CRTC module + * registers this API to encoder for all frame event callbacks like + * frame_error, frame_done, idle_timeout, etc. Encoder may call different events + * from different context - IRQ, user thread, commit_thread, etc. Each event + * should be carefully reviewed and should be processed in proper task context + * to avoid schedulin delay or properly manage the irq context's bottom half + * processing. + */ +static void dpu_crtc_frame_event_cb(void *data, u32 event) +{ + struct drm_crtc *crtc = (struct drm_crtc *)data; + struct dpu_crtc *dpu_crtc; + struct msm_drm_private *priv; + struct dpu_crtc_frame_event *fevent; + unsigned long flags; + u32 crtc_id; + + if (!crtc || !crtc->dev || !crtc->dev->dev_private) { + DPU_ERROR("invalid parameters\n"); + return; + } + + /* Nothing to do on idle event */ + if (event & DPU_ENCODER_FRAME_EVENT_IDLE) + return; + + dpu_crtc = to_dpu_crtc(crtc); + priv = crtc->dev->dev_private; + crtc_id = drm_crtc_index(crtc); + + trace_dpu_crtc_frame_event_cb(DRMID(crtc), event); + + spin_lock_irqsave(&dpu_crtc->spin_lock, flags); + fevent = list_first_entry_or_null(&dpu_crtc->frame_event_list, + struct dpu_crtc_frame_event, list); + if (fevent) + list_del_init(&fevent->list); + spin_unlock_irqrestore(&dpu_crtc->spin_lock, flags); + + if (!fevent) { + DRM_ERROR("crtc%d event %d overflow\n", crtc->base.id, event); + return; + } + + fevent->event = event; + fevent->crtc = crtc; + fevent->ts = ktime_get(); + kthread_queue_work(&priv->event_thread[crtc_id].worker, &fevent->work); +} + +void dpu_crtc_complete_commit(struct drm_crtc *crtc, + struct drm_crtc_state *old_state) +{ + if (!crtc || !crtc->state) { + DPU_ERROR("invalid crtc\n"); + return; + } + trace_dpu_crtc_complete_commit(DRMID(crtc)); +} + +static void _dpu_crtc_setup_mixer_for_encoder( + struct drm_crtc *crtc, + struct drm_encoder *enc) +{ + struct dpu_crtc *dpu_crtc = to_dpu_crtc(crtc); + struct dpu_kms *dpu_kms = _dpu_crtc_get_kms(crtc); + struct dpu_rm *rm = &dpu_kms->rm; + struct dpu_crtc_mixer *mixer; + struct dpu_hw_ctl *last_valid_ctl = NULL; + int i; + struct dpu_rm_hw_iter lm_iter, ctl_iter; + + dpu_rm_init_hw_iter(&lm_iter, enc->base.id, DPU_HW_BLK_LM); + dpu_rm_init_hw_iter(&ctl_iter, enc->base.id, DPU_HW_BLK_CTL); + + /* Set up all the mixers and ctls reserved by this encoder */ + for (i = dpu_crtc->num_mixers; i < ARRAY_SIZE(dpu_crtc->mixers); i++) { + mixer = &dpu_crtc->mixers[i]; + + if (!dpu_rm_get_hw(rm, &lm_iter)) + break; + mixer->hw_lm = (struct dpu_hw_mixer *)lm_iter.hw; + + /* CTL may be <= LMs, if <, multiple LMs controlled by 1 CTL */ + if (!dpu_rm_get_hw(rm, &ctl_iter)) { + DPU_DEBUG("no ctl assigned to lm %d, using previous\n", + mixer->hw_lm->idx - LM_0); + mixer->hw_ctl = last_valid_ctl; + } else { + mixer->hw_ctl = (struct dpu_hw_ctl *)ctl_iter.hw; + last_valid_ctl = mixer->hw_ctl; + } + + /* Shouldn't happen, mixers are always >= ctls */ + if (!mixer->hw_ctl) { + DPU_ERROR("no valid ctls found for lm %d\n", + mixer->hw_lm->idx - LM_0); + return; + } + + mixer->encoder = enc; + + dpu_crtc->num_mixers++; + DPU_DEBUG("setup mixer %d: lm %d\n", + i, mixer->hw_lm->idx - LM_0); + DPU_DEBUG("setup mixer %d: ctl %d\n", + i, mixer->hw_ctl->idx - CTL_0); + } +} + +static void _dpu_crtc_setup_mixers(struct drm_crtc *crtc) +{ + struct dpu_crtc *dpu_crtc = to_dpu_crtc(crtc); + struct drm_encoder *enc; + + dpu_crtc->num_mixers = 0; + dpu_crtc->mixers_swapped = false; + memset(dpu_crtc->mixers, 0, sizeof(dpu_crtc->mixers)); + + mutex_lock(&dpu_crtc->crtc_lock); + /* Check for mixers on all encoders attached to this crtc */ + list_for_each_entry(enc, &crtc->dev->mode_config.encoder_list, head) { + if (enc->crtc != crtc) + continue; + + _dpu_crtc_setup_mixer_for_encoder(crtc, enc); + } + + mutex_unlock(&dpu_crtc->crtc_lock); +} + +static void _dpu_crtc_setup_lm_bounds(struct drm_crtc *crtc, + struct drm_crtc_state *state) +{ + struct dpu_crtc *dpu_crtc; + struct dpu_crtc_state *cstate; + struct drm_display_mode *adj_mode; + u32 crtc_split_width; + int i; + + if (!crtc || !state) { + DPU_ERROR("invalid args\n"); + return; + } + + dpu_crtc = to_dpu_crtc(crtc); + cstate = to_dpu_crtc_state(state); + + adj_mode = &state->adjusted_mode; + crtc_split_width = dpu_crtc_get_mixer_width(dpu_crtc, cstate, adj_mode); + + for (i = 0; i < dpu_crtc->num_mixers; i++) { + struct drm_rect *r = &cstate->lm_bounds[i]; + r->x1 = crtc_split_width * i; + r->y1 = 0; + r->x2 = r->x1 + crtc_split_width; + r->y2 = dpu_crtc_get_mixer_height(dpu_crtc, cstate, adj_mode); + + trace_dpu_crtc_setup_lm_bounds(DRMID(crtc), i, r); + } + + drm_mode_debug_printmodeline(adj_mode); +} + +static void dpu_crtc_atomic_begin(struct drm_crtc *crtc, + struct drm_crtc_state *old_state) +{ + struct dpu_crtc *dpu_crtc; + struct drm_encoder *encoder; + struct drm_device *dev; + unsigned long flags; + struct dpu_crtc_smmu_state_data *smmu_state; + + if (!crtc) { + DPU_ERROR("invalid crtc\n"); + return; + } + + if (!crtc->state->enable) { + DPU_DEBUG("crtc%d -> enable %d, skip atomic_begin\n", + crtc->base.id, crtc->state->enable); + return; + } + + DPU_DEBUG("crtc%d\n", crtc->base.id); + + dpu_crtc = to_dpu_crtc(crtc); + dev = crtc->dev; + smmu_state = &dpu_crtc->smmu_state; + + if (!dpu_crtc->num_mixers) { + _dpu_crtc_setup_mixers(crtc); + _dpu_crtc_setup_lm_bounds(crtc, crtc->state); + } + + if (dpu_crtc->event) { + WARN_ON(dpu_crtc->event); + } else { + spin_lock_irqsave(&dev->event_lock, flags); + dpu_crtc->event = crtc->state->event; + crtc->state->event = NULL; + spin_unlock_irqrestore(&dev->event_lock, flags); + } + + list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) { + if (encoder->crtc != crtc) + continue; + + /* encoder will trigger pending mask now */ + dpu_encoder_trigger_kickoff_pending(encoder); + } + + /* + * If no mixers have been allocated in dpu_crtc_atomic_check(), + * it means we are trying to flush a CRTC whose state is disabled: + * nothing else needs to be done. + */ + if (unlikely(!dpu_crtc->num_mixers)) + return; + + _dpu_crtc_blend_setup(crtc); + + /* + * PP_DONE irq is only used by command mode for now. + * It is better to request pending before FLUSH and START trigger + * to make sure no pp_done irq missed. + * This is safe because no pp_done will happen before SW trigger + * in command mode. + */ +} + +static void dpu_crtc_atomic_flush(struct drm_crtc *crtc, + struct drm_crtc_state *old_crtc_state) +{ + struct dpu_crtc *dpu_crtc; + struct drm_device *dev; + struct drm_plane *plane; + struct msm_drm_private *priv; + struct msm_drm_thread *event_thread; + unsigned long flags; + struct dpu_crtc_state *cstate; + + if (!crtc || !crtc->dev || !crtc->dev->dev_private) { + DPU_ERROR("invalid crtc\n"); + return; + } + + if (!crtc->state->enable) { + DPU_DEBUG("crtc%d -> enable %d, skip atomic_flush\n", + crtc->base.id, crtc->state->enable); + return; + } + + DPU_DEBUG("crtc%d\n", crtc->base.id); + + dpu_crtc = to_dpu_crtc(crtc); + cstate = to_dpu_crtc_state(crtc->state); + dev = crtc->dev; + priv = dev->dev_private; + + if (crtc->index >= ARRAY_SIZE(priv->event_thread)) { + DPU_ERROR("invalid crtc index[%d]\n", crtc->index); + return; + } + + event_thread = &priv->event_thread[crtc->index]; + + if (dpu_crtc->event) { + DPU_DEBUG("already received dpu_crtc->event\n"); + } else { + spin_lock_irqsave(&dev->event_lock, flags); + dpu_crtc->event = crtc->state->event; + crtc->state->event = NULL; + spin_unlock_irqrestore(&dev->event_lock, flags); + } + + /* + * If no mixers has been allocated in dpu_crtc_atomic_check(), + * it means we are trying to flush a CRTC whose state is disabled: + * nothing else needs to be done. + */ + if (unlikely(!dpu_crtc->num_mixers)) + return; + + /* + * For planes without commit update, drm framework will not add + * those planes to current state since hardware update is not + * required. However, if those planes were power collapsed since + * last commit cycle, driver has to restore the hardware state + * of those planes explicitly here prior to plane flush. + */ + drm_atomic_crtc_for_each_plane(plane, crtc) + dpu_plane_restore(plane); + + /* update performance setting before crtc kickoff */ + dpu_core_perf_crtc_update(crtc, 1, false); + + /* + * Final plane updates: Give each plane a chance to complete all + * required writes/flushing before crtc's "flush + * everything" call below. + */ + drm_atomic_crtc_for_each_plane(plane, crtc) { + if (dpu_crtc->smmu_state.transition_error) + dpu_plane_set_error(plane, true); + dpu_plane_flush(plane); + } + + /* Kickoff will be scheduled by outer layer */ +} + +/** + * dpu_crtc_destroy_state - state destroy hook + * @crtc: drm CRTC + * @state: CRTC state object to release + */ +static void dpu_crtc_destroy_state(struct drm_crtc *crtc, + struct drm_crtc_state *state) +{ + struct dpu_crtc *dpu_crtc; + struct dpu_crtc_state *cstate; + + if (!crtc || !state) { + DPU_ERROR("invalid argument(s)\n"); + return; + } + + dpu_crtc = to_dpu_crtc(crtc); + cstate = to_dpu_crtc_state(state); + + DPU_DEBUG("crtc%d\n", crtc->base.id); + + _dpu_crtc_rp_destroy(&cstate->rp); + + __drm_atomic_helper_crtc_destroy_state(state); + + kfree(cstate); +} + +static int _dpu_crtc_wait_for_frame_done(struct drm_crtc *crtc) +{ + struct dpu_crtc *dpu_crtc; + int ret, rc = 0; + + if (!crtc) { + DPU_ERROR("invalid argument\n"); + return -EINVAL; + } + dpu_crtc = to_dpu_crtc(crtc); + + if (!atomic_read(&dpu_crtc->frame_pending)) { + DPU_DEBUG("no frames pending\n"); + return 0; + } + + DPU_ATRACE_BEGIN("frame done completion wait"); + ret = wait_for_completion_timeout(&dpu_crtc->frame_done_comp, + msecs_to_jiffies(DPU_FRAME_DONE_TIMEOUT)); + if (!ret) { + DRM_ERROR("frame done wait timed out, ret:%d\n", ret); + rc = -ETIMEDOUT; + } + DPU_ATRACE_END("frame done completion wait"); + + return rc; +} + +void dpu_crtc_commit_kickoff(struct drm_crtc *crtc) +{ + struct drm_encoder *encoder; + struct drm_device *dev; + struct dpu_crtc *dpu_crtc; + struct msm_drm_private *priv; + struct dpu_kms *dpu_kms; + struct dpu_crtc_state *cstate; + int ret; + + if (!crtc) { + DPU_ERROR("invalid argument\n"); + return; + } + dev = crtc->dev; + dpu_crtc = to_dpu_crtc(crtc); + dpu_kms = _dpu_crtc_get_kms(crtc); + + if (!dpu_kms || !dpu_kms->dev || !dpu_kms->dev->dev_private) { + DPU_ERROR("invalid argument\n"); + return; + } + + priv = dpu_kms->dev->dev_private; + cstate = to_dpu_crtc_state(crtc->state); + + /* + * If no mixers has been allocated in dpu_crtc_atomic_check(), + * it means we are trying to start a CRTC whose state is disabled: + * nothing else needs to be done. + */ + if (unlikely(!dpu_crtc->num_mixers)) + return; + + DPU_ATRACE_BEGIN("crtc_commit"); + + list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) { + struct dpu_encoder_kickoff_params params = { 0 }; + + if (encoder->crtc != crtc) + continue; + + /* + * Encoder will flush/start now, unless it has a tx pending. + * If so, it may delay and flush at an irq event (e.g. ppdone) + */ + dpu_encoder_prepare_for_kickoff(encoder, ¶ms); + } + + /* wait for frame_event_done completion */ + DPU_ATRACE_BEGIN("wait_for_frame_done_event"); + ret = _dpu_crtc_wait_for_frame_done(crtc); + DPU_ATRACE_END("wait_for_frame_done_event"); + if (ret) { + DPU_ERROR("crtc%d wait for frame done failed;frame_pending%d\n", + crtc->base.id, + atomic_read(&dpu_crtc->frame_pending)); + goto end; + } + + if (atomic_inc_return(&dpu_crtc->frame_pending) == 1) { + /* acquire bandwidth and other resources */ + DPU_DEBUG("crtc%d first commit\n", crtc->base.id); + } else + DPU_DEBUG("crtc%d commit\n", crtc->base.id); + + dpu_crtc->play_count++; + + dpu_vbif_clear_errors(dpu_kms); + + list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) { + if (encoder->crtc != crtc) + continue; + + dpu_encoder_kickoff(encoder); + } + +end: + reinit_completion(&dpu_crtc->frame_done_comp); + DPU_ATRACE_END("crtc_commit"); +} + +/** + * _dpu_crtc_vblank_enable_no_lock - update power resource and vblank request + * @dpu_crtc: Pointer to dpu crtc structure + * @enable: Whether to enable/disable vblanks + * + * @Return: error code + */ +static int _dpu_crtc_vblank_enable_no_lock( + struct dpu_crtc *dpu_crtc, bool enable) +{ + struct drm_device *dev; + struct drm_crtc *crtc; + struct drm_encoder *enc; + + if (!dpu_crtc) { + DPU_ERROR("invalid crtc\n"); + return -EINVAL; + } + + crtc = &dpu_crtc->base; + dev = crtc->dev; + + if (enable) { + int ret; + + /* drop lock since power crtc cb may try to re-acquire lock */ + mutex_unlock(&dpu_crtc->crtc_lock); + ret = _dpu_crtc_power_enable(dpu_crtc, true); + mutex_lock(&dpu_crtc->crtc_lock); + if (ret) + return ret; + + list_for_each_entry(enc, &dev->mode_config.encoder_list, head) { + if (enc->crtc != crtc) + continue; + + trace_dpu_crtc_vblank_enable(DRMID(&dpu_crtc->base), + DRMID(enc), enable, + dpu_crtc); + + dpu_encoder_register_vblank_callback(enc, + dpu_crtc_vblank_cb, (void *)crtc); + } + } else { + list_for_each_entry(enc, &dev->mode_config.encoder_list, head) { + if (enc->crtc != crtc) + continue; + + trace_dpu_crtc_vblank_enable(DRMID(&dpu_crtc->base), + DRMID(enc), enable, + dpu_crtc); + + dpu_encoder_register_vblank_callback(enc, NULL, NULL); + } + + /* drop lock since power crtc cb may try to re-acquire lock */ + mutex_unlock(&dpu_crtc->crtc_lock); + _dpu_crtc_power_enable(dpu_crtc, false); + mutex_lock(&dpu_crtc->crtc_lock); + } + + return 0; +} + +/** + * _dpu_crtc_set_suspend - notify crtc of suspend enable/disable + * @crtc: Pointer to drm crtc object + * @enable: true to enable suspend, false to indicate resume + */ +static void _dpu_crtc_set_suspend(struct drm_crtc *crtc, bool enable) +{ + struct dpu_crtc *dpu_crtc; + struct msm_drm_private *priv; + struct dpu_kms *dpu_kms; + int ret = 0; + + if (!crtc || !crtc->dev || !crtc->dev->dev_private) { + DPU_ERROR("invalid crtc\n"); + return; + } + dpu_crtc = to_dpu_crtc(crtc); + priv = crtc->dev->dev_private; + + if (!priv->kms) { + DPU_ERROR("invalid crtc kms\n"); + return; + } + dpu_kms = to_dpu_kms(priv->kms); + + DRM_DEBUG_KMS("crtc%d suspend = %d\n", crtc->base.id, enable); + + mutex_lock(&dpu_crtc->crtc_lock); + + /* + * If the vblank is enabled, release a power reference on suspend + * and take it back during resume (if it is still enabled). + */ + trace_dpu_crtc_set_suspend(DRMID(&dpu_crtc->base), enable, dpu_crtc); + if (dpu_crtc->suspend == enable) + DPU_DEBUG("crtc%d suspend already set to %d, ignoring update\n", + crtc->base.id, enable); + else if (dpu_crtc->enabled && dpu_crtc->vblank_requested) { + ret = _dpu_crtc_vblank_enable_no_lock(dpu_crtc, !enable); + if (ret) + DPU_ERROR("%s vblank enable failed: %d\n", + dpu_crtc->name, ret); + } + + dpu_crtc->suspend = enable; + mutex_unlock(&dpu_crtc->crtc_lock); +} + +/** + * dpu_crtc_duplicate_state - state duplicate hook + * @crtc: Pointer to drm crtc structure + * @Returns: Pointer to new drm_crtc_state structure + */ +static struct drm_crtc_state *dpu_crtc_duplicate_state(struct drm_crtc *crtc) +{ + struct dpu_crtc *dpu_crtc; + struct dpu_crtc_state *cstate, *old_cstate; + + if (!crtc || !crtc->state) { + DPU_ERROR("invalid argument(s)\n"); + return NULL; + } + + dpu_crtc = to_dpu_crtc(crtc); + old_cstate = to_dpu_crtc_state(crtc->state); + cstate = kmemdup(old_cstate, sizeof(*old_cstate), GFP_KERNEL); + if (!cstate) { + DPU_ERROR("failed to allocate state\n"); + return NULL; + } + + /* duplicate base helper */ + __drm_atomic_helper_crtc_duplicate_state(crtc, &cstate->base); + + _dpu_crtc_rp_duplicate(&old_cstate->rp, &cstate->rp); + + return &cstate->base; +} + +/** + * dpu_crtc_reset - reset hook for CRTCs + * Resets the atomic state for @crtc by freeing the state pointer (which might + * be NULL, e.g. at driver load time) and allocating a new empty state object. + * @crtc: Pointer to drm crtc structure + */ +static void dpu_crtc_reset(struct drm_crtc *crtc) +{ + struct dpu_crtc *dpu_crtc; + struct dpu_crtc_state *cstate; + + if (!crtc) { + DPU_ERROR("invalid crtc\n"); + return; + } + + /* revert suspend actions, if necessary */ + if (dpu_kms_is_suspend_state(crtc->dev)) + _dpu_crtc_set_suspend(crtc, false); + + /* remove previous state, if present */ + if (crtc->state) { + dpu_crtc_destroy_state(crtc, crtc->state); + crtc->state = 0; + } + + dpu_crtc = to_dpu_crtc(crtc); + cstate = kzalloc(sizeof(*cstate), GFP_KERNEL); + if (!cstate) { + DPU_ERROR("failed to allocate state\n"); + return; + } + + _dpu_crtc_rp_reset(&cstate->rp, &dpu_crtc->rp_lock, + &dpu_crtc->rp_head); + + cstate->base.crtc = crtc; + crtc->state = &cstate->base; +} + +static void dpu_crtc_handle_power_event(u32 event_type, void *arg) +{ + struct drm_crtc *crtc = arg; + struct dpu_crtc *dpu_crtc; + struct drm_encoder *encoder; + struct dpu_crtc_mixer *m; + u32 i, misr_status; + + if (!crtc) { + DPU_ERROR("invalid crtc\n"); + return; + } + dpu_crtc = to_dpu_crtc(crtc); + + mutex_lock(&dpu_crtc->crtc_lock); + + trace_dpu_crtc_handle_power_event(DRMID(crtc), event_type); + + switch (event_type) { + case DPU_POWER_EVENT_POST_ENABLE: + /* restore encoder; crtc will be programmed during commit */ + drm_for_each_encoder(encoder, crtc->dev) { + if (encoder->crtc != crtc) + continue; + + dpu_encoder_virt_restore(encoder); + } + + for (i = 0; i < dpu_crtc->num_mixers; ++i) { + m = &dpu_crtc->mixers[i]; + if (!m->hw_lm || !m->hw_lm->ops.setup_misr || + !dpu_crtc->misr_enable) + continue; + + m->hw_lm->ops.setup_misr(m->hw_lm, true, + dpu_crtc->misr_frame_count); + } + break; + case DPU_POWER_EVENT_PRE_DISABLE: + for (i = 0; i < dpu_crtc->num_mixers; ++i) { + m = &dpu_crtc->mixers[i]; + if (!m->hw_lm || !m->hw_lm->ops.collect_misr || + !dpu_crtc->misr_enable) + continue; + + misr_status = m->hw_lm->ops.collect_misr(m->hw_lm); + dpu_crtc->misr_data[i] = misr_status ? misr_status : + dpu_crtc->misr_data[i]; + } + break; + case DPU_POWER_EVENT_POST_DISABLE: + /** + * Nothing to do. All the planes on the CRTC will be + * programmed for every frame + */ + break; + default: + DPU_DEBUG("event:%d not handled\n", event_type); + break; + } + + mutex_unlock(&dpu_crtc->crtc_lock); +} + +static void dpu_crtc_disable(struct drm_crtc *crtc) +{ + struct dpu_crtc *dpu_crtc; + struct dpu_crtc_state *cstate; + struct drm_display_mode *mode; + struct drm_encoder *encoder; + struct msm_drm_private *priv; + int ret; + unsigned long flags; + + if (!crtc || !crtc->dev || !crtc->dev->dev_private || !crtc->state) { + DPU_ERROR("invalid crtc\n"); + return; + } + dpu_crtc = to_dpu_crtc(crtc); + cstate = to_dpu_crtc_state(crtc->state); + mode = &cstate->base.adjusted_mode; + priv = crtc->dev->dev_private; + + DRM_DEBUG_KMS("crtc%d\n", crtc->base.id); + + if (dpu_kms_is_suspend_state(crtc->dev)) + _dpu_crtc_set_suspend(crtc, true); + + /* Disable/save vblank irq handling */ + drm_crtc_vblank_off(crtc); + + mutex_lock(&dpu_crtc->crtc_lock); + + /* wait for frame_event_done completion */ + if (_dpu_crtc_wait_for_frame_done(crtc)) + DPU_ERROR("crtc%d wait for frame done failed;frame_pending%d\n", + crtc->base.id, + atomic_read(&dpu_crtc->frame_pending)); + + trace_dpu_crtc_disable(DRMID(crtc), false, dpu_crtc); + if (dpu_crtc->enabled && !dpu_crtc->suspend && + dpu_crtc->vblank_requested) { + ret = _dpu_crtc_vblank_enable_no_lock(dpu_crtc, false); + if (ret) + DPU_ERROR("%s vblank enable failed: %d\n", + dpu_crtc->name, ret); + } + dpu_crtc->enabled = false; + + if (atomic_read(&dpu_crtc->frame_pending)) { + trace_dpu_crtc_disable_frame_pending(DRMID(crtc), + atomic_read(&dpu_crtc->frame_pending)); + dpu_core_perf_crtc_release_bw(crtc); + atomic_set(&dpu_crtc->frame_pending, 0); + } + + dpu_core_perf_crtc_update(crtc, 0, true); + + drm_for_each_encoder(encoder, crtc->dev) { + if (encoder->crtc != crtc) + continue; + dpu_encoder_register_frame_event_callback(encoder, NULL, NULL); + } + + if (dpu_crtc->power_event) + dpu_power_handle_unregister_event(dpu_crtc->phandle, + dpu_crtc->power_event); + + memset(dpu_crtc->mixers, 0, sizeof(dpu_crtc->mixers)); + dpu_crtc->num_mixers = 0; + dpu_crtc->mixers_swapped = false; + + /* disable clk & bw control until clk & bw properties are set */ + cstate->bw_control = false; + cstate->bw_split_vote = false; + + mutex_unlock(&dpu_crtc->crtc_lock); + + if (crtc->state->event && !crtc->state->active) { + spin_lock_irqsave(&crtc->dev->event_lock, flags); + drm_crtc_send_vblank_event(crtc, crtc->state->event); + crtc->state->event = NULL; + spin_unlock_irqrestore(&crtc->dev->event_lock, flags); + } +} + +static void dpu_crtc_enable(struct drm_crtc *crtc, + struct drm_crtc_state *old_crtc_state) +{ + struct dpu_crtc *dpu_crtc; + struct drm_encoder *encoder; + struct msm_drm_private *priv; + int ret; + + if (!crtc || !crtc->dev || !crtc->dev->dev_private) { + DPU_ERROR("invalid crtc\n"); + return; + } + priv = crtc->dev->dev_private; + + DRM_DEBUG_KMS("crtc%d\n", crtc->base.id); + dpu_crtc = to_dpu_crtc(crtc); + + drm_for_each_encoder(encoder, crtc->dev) { + if (encoder->crtc != crtc) + continue; + dpu_encoder_register_frame_event_callback(encoder, + dpu_crtc_frame_event_cb, (void *)crtc); + } + + mutex_lock(&dpu_crtc->crtc_lock); + trace_dpu_crtc_enable(DRMID(crtc), true, dpu_crtc); + if (!dpu_crtc->enabled && !dpu_crtc->suspend && + dpu_crtc->vblank_requested) { + ret = _dpu_crtc_vblank_enable_no_lock(dpu_crtc, true); + if (ret) + DPU_ERROR("%s vblank enable failed: %d\n", + dpu_crtc->name, ret); + } + dpu_crtc->enabled = true; + + mutex_unlock(&dpu_crtc->crtc_lock); + + /* Enable/restore vblank irq handling */ + drm_crtc_vblank_on(crtc); + + dpu_crtc->power_event = dpu_power_handle_register_event( + dpu_crtc->phandle, + DPU_POWER_EVENT_POST_ENABLE | DPU_POWER_EVENT_POST_DISABLE | + DPU_POWER_EVENT_PRE_DISABLE, + dpu_crtc_handle_power_event, crtc, dpu_crtc->name); + +} + +struct plane_state { + struct dpu_plane_state *dpu_pstate; + const struct drm_plane_state *drm_pstate; + int stage; + u32 pipe_id; +}; + +static int dpu_crtc_atomic_check(struct drm_crtc *crtc, + struct drm_crtc_state *state) +{ + struct dpu_crtc *dpu_crtc; + struct plane_state *pstates; + struct dpu_crtc_state *cstate; + + const struct drm_plane_state *pstate; + struct drm_plane *plane; + struct drm_display_mode *mode; + + int cnt = 0, rc = 0, mixer_width, i, z_pos; + + struct dpu_multirect_plane_states multirect_plane[DPU_STAGE_MAX * 2]; + int multirect_count = 0; + const struct drm_plane_state *pipe_staged[SSPP_MAX]; + int left_zpos_cnt = 0, right_zpos_cnt = 0; + struct drm_rect crtc_rect = { 0 }; + + if (!crtc) { + DPU_ERROR("invalid crtc\n"); + return -EINVAL; + } + + pstates = kzalloc(sizeof(*pstates) * DPU_STAGE_MAX * 4, GFP_KERNEL); + + dpu_crtc = to_dpu_crtc(crtc); + cstate = to_dpu_crtc_state(state); + + if (!state->enable || !state->active) { + DPU_DEBUG("crtc%d -> enable %d, active %d, skip atomic_check\n", + crtc->base.id, state->enable, state->active); + goto end; + } + + mode = &state->adjusted_mode; + DPU_DEBUG("%s: check", dpu_crtc->name); + + /* force a full mode set if active state changed */ + if (state->active_changed) + state->mode_changed = true; + + memset(pipe_staged, 0, sizeof(pipe_staged)); + + mixer_width = dpu_crtc_get_mixer_width(dpu_crtc, cstate, mode); + + _dpu_crtc_setup_lm_bounds(crtc, state); + + crtc_rect.x2 = mode->hdisplay; + crtc_rect.y2 = mode->vdisplay; + + /* get plane state for all drm planes associated with crtc state */ + drm_atomic_crtc_state_for_each_plane_state(plane, pstate, state) { + struct drm_rect dst, clip = crtc_rect; + + if (IS_ERR_OR_NULL(pstate)) { + rc = PTR_ERR(pstate); + DPU_ERROR("%s: failed to get plane%d state, %d\n", + dpu_crtc->name, plane->base.id, rc); + goto end; + } + if (cnt >= DPU_STAGE_MAX * 4) + continue; + + pstates[cnt].dpu_pstate = to_dpu_plane_state(pstate); + pstates[cnt].drm_pstate = pstate; + pstates[cnt].stage = pstate->normalized_zpos; + pstates[cnt].pipe_id = dpu_plane_pipe(plane); + + if (pipe_staged[pstates[cnt].pipe_id]) { + multirect_plane[multirect_count].r0 = + pipe_staged[pstates[cnt].pipe_id]; + multirect_plane[multirect_count].r1 = pstate; + multirect_count++; + + pipe_staged[pstates[cnt].pipe_id] = NULL; + } else { + pipe_staged[pstates[cnt].pipe_id] = pstate; + } + + cnt++; + + dst = drm_plane_state_dest(pstate); + if (!drm_rect_intersect(&clip, &dst) || + !drm_rect_equals(&clip, &dst)) { + DPU_ERROR("invalid vertical/horizontal destination\n"); + DPU_ERROR("display: " DRM_RECT_FMT " plane: " + DRM_RECT_FMT "\n", DRM_RECT_ARG(&crtc_rect), + DRM_RECT_ARG(&dst)); + rc = -E2BIG; + goto end; + } + } + + for (i = 1; i < SSPP_MAX; i++) { + if (pipe_staged[i]) { + dpu_plane_clear_multirect(pipe_staged[i]); + + if (is_dpu_plane_virtual(pipe_staged[i]->plane)) { + DPU_ERROR( + "r1 only virt plane:%d not supported\n", + pipe_staged[i]->plane->base.id); + rc = -EINVAL; + goto end; + } + } + } + + z_pos = -1; + for (i = 0; i < cnt; i++) { + /* reset counts at every new blend stage */ + if (pstates[i].stage != z_pos) { + left_zpos_cnt = 0; + right_zpos_cnt = 0; + z_pos = pstates[i].stage; + } + + /* verify z_pos setting before using it */ + if (z_pos >= DPU_STAGE_MAX - DPU_STAGE_0) { + DPU_ERROR("> %d plane stages assigned\n", + DPU_STAGE_MAX - DPU_STAGE_0); + rc = -EINVAL; + goto end; + } else if (pstates[i].drm_pstate->crtc_x < mixer_width) { + if (left_zpos_cnt == 2) { + DPU_ERROR("> 2 planes @ stage %d on left\n", + z_pos); + rc = -EINVAL; + goto end; + } + left_zpos_cnt++; + + } else { + if (right_zpos_cnt == 2) { + DPU_ERROR("> 2 planes @ stage %d on right\n", + z_pos); + rc = -EINVAL; + goto end; + } + right_zpos_cnt++; + } + + pstates[i].dpu_pstate->stage = z_pos + DPU_STAGE_0; + DPU_DEBUG("%s: zpos %d", dpu_crtc->name, z_pos); + } + + for (i = 0; i < multirect_count; i++) { + if (dpu_plane_validate_multirect_v2(&multirect_plane[i])) { + DPU_ERROR( + "multirect validation failed for planes (%d - %d)\n", + multirect_plane[i].r0->plane->base.id, + multirect_plane[i].r1->plane->base.id); + rc = -EINVAL; + goto end; + } + } + + rc = dpu_core_perf_crtc_check(crtc, state); + if (rc) { + DPU_ERROR("crtc%d failed performance check %d\n", + crtc->base.id, rc); + goto end; + } + + /* validate source split: + * use pstates sorted by stage to check planes on same stage + * we assume that all pipes are in source split so its valid to compare + * without taking into account left/right mixer placement + */ + for (i = 1; i < cnt; i++) { + struct plane_state *prv_pstate, *cur_pstate; + struct drm_rect left_rect, right_rect; + int32_t left_pid, right_pid; + int32_t stage; + + prv_pstate = &pstates[i - 1]; + cur_pstate = &pstates[i]; + if (prv_pstate->stage != cur_pstate->stage) + continue; + + stage = cur_pstate->stage; + + left_pid = prv_pstate->dpu_pstate->base.plane->base.id; + left_rect = drm_plane_state_dest(prv_pstate->drm_pstate); + + right_pid = cur_pstate->dpu_pstate->base.plane->base.id; + right_rect = drm_plane_state_dest(cur_pstate->drm_pstate); + + if (right_rect.x1 < left_rect.x1) { + swap(left_pid, right_pid); + swap(left_rect, right_rect); + } + + /** + * - planes are enumerated in pipe-priority order such that + * planes with lower drm_id must be left-most in a shared + * blend-stage when using source split. + * - planes in source split must be contiguous in width + * - planes in source split must have same dest yoff and height + */ + if (right_pid < left_pid) { + DPU_ERROR( + "invalid src split cfg. priority mismatch. stage: %d left: %d right: %d\n", + stage, left_pid, right_pid); + rc = -EINVAL; + goto end; + } else if (right_rect.x1 != drm_rect_width(&left_rect)) { + DPU_ERROR("non-contiguous coordinates for src split. " + "stage: %d left: " DRM_RECT_FMT " right: " + DRM_RECT_FMT "\n", stage, + DRM_RECT_ARG(&left_rect), + DRM_RECT_ARG(&right_rect)); + rc = -EINVAL; + goto end; + } else if (left_rect.y1 != right_rect.y1 || + drm_rect_height(&left_rect) != drm_rect_height(&right_rect)) { + DPU_ERROR("source split at stage: %d. invalid " + "yoff/height: left: " DRM_RECT_FMT " right: " + DRM_RECT_FMT "\n", stage, + DRM_RECT_ARG(&left_rect), + DRM_RECT_ARG(&right_rect)); + rc = -EINVAL; + goto end; + } + } + +end: + _dpu_crtc_rp_free_unused(&cstate->rp); + kfree(pstates); + return rc; +} + +int dpu_crtc_vblank(struct drm_crtc *crtc, bool en) +{ + struct dpu_crtc *dpu_crtc; + int ret; + + if (!crtc) { + DPU_ERROR("invalid crtc\n"); + return -EINVAL; + } + dpu_crtc = to_dpu_crtc(crtc); + + mutex_lock(&dpu_crtc->crtc_lock); + trace_dpu_crtc_vblank(DRMID(&dpu_crtc->base), en, dpu_crtc); + if (dpu_crtc->enabled && !dpu_crtc->suspend) { + ret = _dpu_crtc_vblank_enable_no_lock(dpu_crtc, en); + if (ret) + DPU_ERROR("%s vblank enable failed: %d\n", + dpu_crtc->name, ret); + } + dpu_crtc->vblank_requested = en; + mutex_unlock(&dpu_crtc->crtc_lock); + + return 0; +} + +#ifdef CONFIG_DEBUG_FS +static int _dpu_debugfs_status_show(struct seq_file *s, void *data) +{ + struct dpu_crtc *dpu_crtc; + struct dpu_plane_state *pstate = NULL; + struct dpu_crtc_mixer *m; + + struct drm_crtc *crtc; + struct drm_plane *plane; + struct drm_display_mode *mode; + struct drm_framebuffer *fb; + struct drm_plane_state *state; + struct dpu_crtc_state *cstate; + + int i, out_width; + + if (!s || !s->private) + return -EINVAL; + + dpu_crtc = s->private; + crtc = &dpu_crtc->base; + cstate = to_dpu_crtc_state(crtc->state); + + mutex_lock(&dpu_crtc->crtc_lock); + mode = &crtc->state->adjusted_mode; + out_width = dpu_crtc_get_mixer_width(dpu_crtc, cstate, mode); + + seq_printf(s, "crtc:%d width:%d height:%d\n", crtc->base.id, + mode->hdisplay, mode->vdisplay); + + seq_puts(s, "\n"); + + for (i = 0; i < dpu_crtc->num_mixers; ++i) { + m = &dpu_crtc->mixers[i]; + if (!m->hw_lm) + seq_printf(s, "\tmixer[%d] has no lm\n", i); + else if (!m->hw_ctl) + seq_printf(s, "\tmixer[%d] has no ctl\n", i); + else + seq_printf(s, "\tmixer:%d ctl:%d width:%d height:%d\n", + m->hw_lm->idx - LM_0, m->hw_ctl->idx - CTL_0, + out_width, mode->vdisplay); + } + + seq_puts(s, "\n"); + + drm_atomic_crtc_for_each_plane(plane, crtc) { + pstate = to_dpu_plane_state(plane->state); + state = plane->state; + + if (!pstate || !state) + continue; + + seq_printf(s, "\tplane:%u stage:%d\n", plane->base.id, + pstate->stage); + + if (plane->state->fb) { + fb = plane->state->fb; + + seq_printf(s, "\tfb:%d image format:%4.4s wxh:%ux%u ", + fb->base.id, (char *) &fb->format->format, + fb->width, fb->height); + for (i = 0; i < ARRAY_SIZE(fb->format->cpp); ++i) + seq_printf(s, "cpp[%d]:%u ", + i, fb->format->cpp[i]); + seq_puts(s, "\n\t"); + + seq_printf(s, "modifier:%8llu ", fb->modifier); + seq_puts(s, "\n"); + + seq_puts(s, "\t"); + for (i = 0; i < ARRAY_SIZE(fb->pitches); i++) + seq_printf(s, "pitches[%d]:%8u ", i, + fb->pitches[i]); + seq_puts(s, "\n"); + + seq_puts(s, "\t"); + for (i = 0; i < ARRAY_SIZE(fb->offsets); i++) + seq_printf(s, "offsets[%d]:%8u ", i, + fb->offsets[i]); + seq_puts(s, "\n"); + } + + seq_printf(s, "\tsrc_x:%4d src_y:%4d src_w:%4d src_h:%4d\n", + state->src_x, state->src_y, state->src_w, state->src_h); + + seq_printf(s, "\tdst x:%4d dst_y:%4d dst_w:%4d dst_h:%4d\n", + state->crtc_x, state->crtc_y, state->crtc_w, + state->crtc_h); + seq_printf(s, "\tmultirect: mode: %d index: %d\n", + pstate->multirect_mode, pstate->multirect_index); + + seq_puts(s, "\n"); + } + if (dpu_crtc->vblank_cb_count) { + ktime_t diff = ktime_sub(ktime_get(), dpu_crtc->vblank_cb_time); + s64 diff_ms = ktime_to_ms(diff); + s64 fps = diff_ms ? div_s64( + dpu_crtc->vblank_cb_count * 1000, diff_ms) : 0; + + seq_printf(s, + "vblank fps:%lld count:%u total:%llums total_framecount:%llu\n", + fps, dpu_crtc->vblank_cb_count, + ktime_to_ms(diff), dpu_crtc->play_count); + + /* reset time & count for next measurement */ + dpu_crtc->vblank_cb_count = 0; + dpu_crtc->vblank_cb_time = ktime_set(0, 0); + } + + seq_printf(s, "vblank_enable:%d\n", dpu_crtc->vblank_requested); + + mutex_unlock(&dpu_crtc->crtc_lock); + + return 0; +} + +static int _dpu_debugfs_status_open(struct inode *inode, struct file *file) +{ + return single_open(file, _dpu_debugfs_status_show, inode->i_private); +} + +static ssize_t _dpu_crtc_misr_setup(struct file *file, + const char __user *user_buf, size_t count, loff_t *ppos) +{ + struct dpu_crtc *dpu_crtc; + struct dpu_crtc_mixer *m; + int i = 0, rc; + char buf[MISR_BUFF_SIZE + 1]; + u32 frame_count, enable; + size_t buff_copy; + + if (!file || !file->private_data) + return -EINVAL; + + dpu_crtc = file->private_data; + buff_copy = min_t(size_t, count, MISR_BUFF_SIZE); + if (copy_from_user(buf, user_buf, buff_copy)) { + DPU_ERROR("buffer copy failed\n"); + return -EINVAL; + } + + buf[buff_copy] = 0; /* end of string */ + + if (sscanf(buf, "%u %u", &enable, &frame_count) != 2) + return -EINVAL; + + rc = _dpu_crtc_power_enable(dpu_crtc, true); + if (rc) + return rc; + + mutex_lock(&dpu_crtc->crtc_lock); + dpu_crtc->misr_enable = enable; + dpu_crtc->misr_frame_count = frame_count; + for (i = 0; i < dpu_crtc->num_mixers; ++i) { + dpu_crtc->misr_data[i] = 0; + m = &dpu_crtc->mixers[i]; + if (!m->hw_lm || !m->hw_lm->ops.setup_misr) + continue; + + m->hw_lm->ops.setup_misr(m->hw_lm, enable, frame_count); + } + mutex_unlock(&dpu_crtc->crtc_lock); + _dpu_crtc_power_enable(dpu_crtc, false); + + return count; +} + +static ssize_t _dpu_crtc_misr_read(struct file *file, + char __user *user_buff, size_t count, loff_t *ppos) +{ + struct dpu_crtc *dpu_crtc; + struct dpu_crtc_mixer *m; + int i = 0, rc; + u32 misr_status; + ssize_t len = 0; + char buf[MISR_BUFF_SIZE + 1] = {'\0'}; + + if (*ppos) + return 0; + + if (!file || !file->private_data) + return -EINVAL; + + dpu_crtc = file->private_data; + rc = _dpu_crtc_power_enable(dpu_crtc, true); + if (rc) + return rc; + + mutex_lock(&dpu_crtc->crtc_lock); + if (!dpu_crtc->misr_enable) { + len += snprintf(buf + len, MISR_BUFF_SIZE - len, + "disabled\n"); + goto buff_check; + } + + for (i = 0; i < dpu_crtc->num_mixers; ++i) { + m = &dpu_crtc->mixers[i]; + if (!m->hw_lm || !m->hw_lm->ops.collect_misr) + continue; + + misr_status = m->hw_lm->ops.collect_misr(m->hw_lm); + dpu_crtc->misr_data[i] = misr_status ? misr_status : + dpu_crtc->misr_data[i]; + len += snprintf(buf + len, MISR_BUFF_SIZE - len, "lm idx:%d\n", + m->hw_lm->idx - LM_0); + len += snprintf(buf + len, MISR_BUFF_SIZE - len, "0x%x\n", + dpu_crtc->misr_data[i]); + } + +buff_check: + if (count <= len) { + len = 0; + goto end; + } + + if (copy_to_user(user_buff, buf, len)) { + len = -EFAULT; + goto end; + } + + *ppos += len; /* increase offset */ + +end: + mutex_unlock(&dpu_crtc->crtc_lock); + _dpu_crtc_power_enable(dpu_crtc, false); + return len; +} + +#define DEFINE_DPU_DEBUGFS_SEQ_FOPS(__prefix) \ +static int __prefix ## _open(struct inode *inode, struct file *file) \ +{ \ + return single_open(file, __prefix ## _show, inode->i_private); \ +} \ +static const struct file_operations __prefix ## _fops = { \ + .owner = THIS_MODULE, \ + .open = __prefix ## _open, \ + .release = single_release, \ + .read = seq_read, \ + .llseek = seq_lseek, \ +} + +static int dpu_crtc_debugfs_state_show(struct seq_file *s, void *v) +{ + struct drm_crtc *crtc = (struct drm_crtc *) s->private; + struct dpu_crtc *dpu_crtc = to_dpu_crtc(crtc); + struct dpu_crtc_res *res; + struct dpu_crtc_respool *rp; + int i; + + seq_printf(s, "client type: %d\n", dpu_crtc_get_client_type(crtc)); + seq_printf(s, "intf_mode: %d\n", dpu_crtc_get_intf_mode(crtc)); + seq_printf(s, "core_clk_rate: %llu\n", + dpu_crtc->cur_perf.core_clk_rate); + for (i = DPU_POWER_HANDLE_DBUS_ID_MNOC; + i < DPU_POWER_HANDLE_DBUS_ID_MAX; i++) { + seq_printf(s, "bw_ctl[%s]: %llu\n", + dpu_power_handle_get_dbus_name(i), + dpu_crtc->cur_perf.bw_ctl[i]); + seq_printf(s, "max_per_pipe_ib[%s]: %llu\n", + dpu_power_handle_get_dbus_name(i), + dpu_crtc->cur_perf.max_per_pipe_ib[i]); + } + + mutex_lock(&dpu_crtc->rp_lock); + list_for_each_entry(rp, &dpu_crtc->rp_head, rp_list) { + seq_printf(s, "rp.%d: ", rp->sequence_id); + list_for_each_entry(res, &rp->res_list, list) + seq_printf(s, "0x%x/0x%llx/%pK/%d ", + res->type, res->tag, res->val, + atomic_read(&res->refcount)); + seq_puts(s, "\n"); + } + mutex_unlock(&dpu_crtc->rp_lock); + + return 0; +} +DEFINE_DPU_DEBUGFS_SEQ_FOPS(dpu_crtc_debugfs_state); + +static int _dpu_crtc_init_debugfs(struct drm_crtc *crtc) +{ + struct dpu_crtc *dpu_crtc; + struct dpu_kms *dpu_kms; + + static const struct file_operations debugfs_status_fops = { + .open = _dpu_debugfs_status_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, + }; + static const struct file_operations debugfs_misr_fops = { + .open = simple_open, + .read = _dpu_crtc_misr_read, + .write = _dpu_crtc_misr_setup, + }; + + if (!crtc) + return -EINVAL; + dpu_crtc = to_dpu_crtc(crtc); + + dpu_kms = _dpu_crtc_get_kms(crtc); + if (!dpu_kms) + return -EINVAL; + + dpu_crtc->debugfs_root = debugfs_create_dir(dpu_crtc->name, + crtc->dev->primary->debugfs_root); + if (!dpu_crtc->debugfs_root) + return -ENOMEM; + + /* don't error check these */ + debugfs_create_file("status", 0400, + dpu_crtc->debugfs_root, + dpu_crtc, &debugfs_status_fops); + debugfs_create_file("state", 0600, + dpu_crtc->debugfs_root, + &dpu_crtc->base, + &dpu_crtc_debugfs_state_fops); + debugfs_create_file("misr_data", 0600, dpu_crtc->debugfs_root, + dpu_crtc, &debugfs_misr_fops); + + return 0; +} + +static void _dpu_crtc_destroy_debugfs(struct drm_crtc *crtc) +{ + struct dpu_crtc *dpu_crtc; + + if (!crtc) + return; + dpu_crtc = to_dpu_crtc(crtc); + debugfs_remove_recursive(dpu_crtc->debugfs_root); +} +#else +static int _dpu_crtc_init_debugfs(struct drm_crtc *crtc) +{ + return 0; +} + +static void _dpu_crtc_destroy_debugfs(struct drm_crtc *crtc) +{ +} +#endif /* CONFIG_DEBUG_FS */ + +static int dpu_crtc_late_register(struct drm_crtc *crtc) +{ + return _dpu_crtc_init_debugfs(crtc); +} + +static void dpu_crtc_early_unregister(struct drm_crtc *crtc) +{ + _dpu_crtc_destroy_debugfs(crtc); +} + +static const struct drm_crtc_funcs dpu_crtc_funcs = { + .set_config = drm_atomic_helper_set_config, + .destroy = dpu_crtc_destroy, + .page_flip = drm_atomic_helper_page_flip, + .reset = dpu_crtc_reset, + .atomic_duplicate_state = dpu_crtc_duplicate_state, + .atomic_destroy_state = dpu_crtc_destroy_state, + .late_register = dpu_crtc_late_register, + .early_unregister = dpu_crtc_early_unregister, +}; + +static const struct drm_crtc_helper_funcs dpu_crtc_helper_funcs = { + .disable = dpu_crtc_disable, + .atomic_enable = dpu_crtc_enable, + .atomic_check = dpu_crtc_atomic_check, + .atomic_begin = dpu_crtc_atomic_begin, + .atomic_flush = dpu_crtc_atomic_flush, +}; + +/* initialize crtc */ +struct drm_crtc *dpu_crtc_init(struct drm_device *dev, struct drm_plane *plane) +{ + struct drm_crtc *crtc = NULL; + struct dpu_crtc *dpu_crtc = NULL; + struct msm_drm_private *priv = NULL; + struct dpu_kms *kms = NULL; + int i; + + priv = dev->dev_private; + kms = to_dpu_kms(priv->kms); + + dpu_crtc = kzalloc(sizeof(*dpu_crtc), GFP_KERNEL); + if (!dpu_crtc) + return ERR_PTR(-ENOMEM); + + crtc = &dpu_crtc->base; + crtc->dev = dev; + + mutex_init(&dpu_crtc->crtc_lock); + spin_lock_init(&dpu_crtc->spin_lock); + atomic_set(&dpu_crtc->frame_pending, 0); + + mutex_init(&dpu_crtc->rp_lock); + INIT_LIST_HEAD(&dpu_crtc->rp_head); + + init_completion(&dpu_crtc->frame_done_comp); + + INIT_LIST_HEAD(&dpu_crtc->frame_event_list); + + for (i = 0; i < ARRAY_SIZE(dpu_crtc->frame_events); i++) { + INIT_LIST_HEAD(&dpu_crtc->frame_events[i].list); + list_add(&dpu_crtc->frame_events[i].list, + &dpu_crtc->frame_event_list); + kthread_init_work(&dpu_crtc->frame_events[i].work, + dpu_crtc_frame_event_work); + } + + drm_crtc_init_with_planes(dev, crtc, plane, NULL, &dpu_crtc_funcs, + NULL); + + drm_crtc_helper_add(crtc, &dpu_crtc_helper_funcs); + plane->crtc = crtc; + + /* save user friendly CRTC name for later */ + snprintf(dpu_crtc->name, DPU_CRTC_NAME_SIZE, "crtc%u", crtc->base.id); + + /* initialize event handling */ + spin_lock_init(&dpu_crtc->event_lock); + + dpu_crtc->phandle = &kms->phandle; + + DPU_DEBUG("%s: successfully initialized crtc\n", dpu_crtc->name); + return crtc; +} diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_crtc.h b/drivers/gpu/drm/msm/disp/dpu1/dpu_crtc.h new file mode 100644 index 000000000000..e87109e608e9 --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_crtc.h @@ -0,0 +1,423 @@ +/* + * Copyright (c) 2015-2018 The Linux Foundation. All rights reserved. + * Copyright (C) 2013 Red Hat + * Author: Rob Clark <robdclark@gmail.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published by + * the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program. If not, see <http://www.gnu.org/licenses/>. + */ + +#ifndef _DPU_CRTC_H_ +#define _DPU_CRTC_H_ + +#include <linux/kthread.h> +#include <drm/drm_crtc.h> +#include "dpu_kms.h" +#include "dpu_core_perf.h" +#include "dpu_hw_blk.h" + +#define DPU_CRTC_NAME_SIZE 12 + +/* define the maximum number of in-flight frame events */ +#define DPU_CRTC_FRAME_EVENT_SIZE 4 + +/** + * enum dpu_crtc_client_type: crtc client type + * @RT_CLIENT: RealTime client like video/cmd mode display + * voting through apps rsc + * @NRT_CLIENT: Non-RealTime client like WB display + * voting through apps rsc + */ +enum dpu_crtc_client_type { + RT_CLIENT, + NRT_CLIENT, +}; + +/** + * enum dpu_crtc_smmu_state: smmu state + * @ATTACHED: all the context banks are attached. + * @DETACHED: all the context banks are detached. + * @ATTACH_ALL_REQ: transient state of attaching context banks. + * @DETACH_ALL_REQ: transient state of detaching context banks. + */ +enum dpu_crtc_smmu_state { + ATTACHED = 0, + DETACHED, + ATTACH_ALL_REQ, + DETACH_ALL_REQ, +}; + +/** + * enum dpu_crtc_smmu_state_transition_type: state transition type + * @NONE: no pending state transitions + * @PRE_COMMIT: state transitions should be done before processing the commit + * @POST_COMMIT: state transitions to be done after processing the commit. + */ +enum dpu_crtc_smmu_state_transition_type { + NONE, + PRE_COMMIT, + POST_COMMIT +}; + +/** + * struct dpu_crtc_smmu_state_data: stores the smmu state and transition type + * @state: current state of smmu context banks + * @transition_type: transition request type + * @transition_error: whether there is error while transitioning the state + */ +struct dpu_crtc_smmu_state_data { + uint32_t state; + uint32_t transition_type; + uint32_t transition_error; +}; + +/** + * struct dpu_crtc_mixer: stores the map for each virtual pipeline in the CRTC + * @hw_lm: LM HW Driver context + * @hw_ctl: CTL Path HW driver context + * @encoder: Encoder attached to this lm & ctl + * @mixer_op_mode: mixer blending operation mode + * @flush_mask: mixer flush mask for ctl, mixer and pipe + */ +struct dpu_crtc_mixer { + struct dpu_hw_mixer *hw_lm; + struct dpu_hw_ctl *hw_ctl; + struct drm_encoder *encoder; + u32 mixer_op_mode; + u32 flush_mask; +}; + +/** + * struct dpu_crtc_frame_event: stores crtc frame event for crtc processing + * @work: base work structure + * @crtc: Pointer to crtc handling this event + * @list: event list + * @ts: timestamp at queue entry + * @event: event identifier + */ +struct dpu_crtc_frame_event { + struct kthread_work work; + struct drm_crtc *crtc; + struct list_head list; + ktime_t ts; + u32 event; +}; + +/* + * Maximum number of free event structures to cache + */ +#define DPU_CRTC_MAX_EVENT_COUNT 16 + +/** + * struct dpu_crtc - virtualized CRTC data structure + * @base : Base drm crtc structure + * @name : ASCII description of this crtc + * @num_ctls : Number of ctl paths in use + * @num_mixers : Number of mixers in use + * @mixers_swapped: Whether the mixers have been swapped for left/right update + * especially in the case of DSC Merge. + * @mixers : List of active mixers + * @event : Pointer to last received drm vblank event. If there is a + * pending vblank event, this will be non-null. + * @vsync_count : Running count of received vsync events + * @drm_requested_vblank : Whether vblanks have been enabled in the encoder + * @property_info : Opaque structure for generic property support + * @property_defaults : Array of default values for generic property support + * @stage_cfg : H/w mixer stage configuration + * @debugfs_root : Parent of debugfs node + * @vblank_cb_count : count of vblank callback since last reset + * @play_count : frame count between crtc enable and disable + * @vblank_cb_time : ktime at vblank count reset + * @vblank_requested : whether the user has requested vblank events + * @suspend : whether or not a suspend operation is in progress + * @enabled : whether the DPU CRTC is currently enabled. updated in the + * commit-thread, not state-swap time which is earlier, so + * safe to make decisions on during VBLANK on/off work + * @feature_list : list of color processing features supported on a crtc + * @active_list : list of color processing features are active + * @dirty_list : list of color processing features are dirty + * @ad_dirty: list containing ad properties that are dirty + * @ad_active: list containing ad properties that are active + * @crtc_lock : crtc lock around create, destroy and access. + * @frame_pending : Whether or not an update is pending + * @frame_events : static allocation of in-flight frame events + * @frame_event_list : available frame event list + * @spin_lock : spin lock for frame event, transaction status, etc... + * @frame_done_comp : for frame_event_done synchronization + * @event_thread : Pointer to event handler thread + * @event_worker : Event worker queue + * @event_lock : Spinlock around event handling code + * @misr_enable : boolean entry indicates misr enable/disable status. + * @misr_frame_count : misr frame count provided by client + * @misr_data : store misr data before turning off the clocks. + * @phandle: Pointer to power handler + * @power_event : registered power event handle + * @cur_perf : current performance committed to clock/bandwidth driver + * @rp_lock : serialization lock for resource pool + * @rp_head : list of active resource pool + * @scl3_cfg_lut : qseed3 lut config + */ +struct dpu_crtc { + struct drm_crtc base; + char name[DPU_CRTC_NAME_SIZE]; + + /* HW Resources reserved for the crtc */ + u32 num_ctls; + u32 num_mixers; + bool mixers_swapped; + struct dpu_crtc_mixer mixers[CRTC_DUAL_MIXERS]; + struct dpu_hw_scaler3_lut_cfg *scl3_lut_cfg; + + struct drm_pending_vblank_event *event; + u32 vsync_count; + + struct dpu_hw_stage_cfg stage_cfg; + struct dentry *debugfs_root; + + u32 vblank_cb_count; + u64 play_count; + ktime_t vblank_cb_time; + bool vblank_requested; + bool suspend; + bool enabled; + + struct list_head feature_list; + struct list_head active_list; + struct list_head dirty_list; + struct list_head ad_dirty; + struct list_head ad_active; + + struct mutex crtc_lock; + + atomic_t frame_pending; + struct dpu_crtc_frame_event frame_events[DPU_CRTC_FRAME_EVENT_SIZE]; + struct list_head frame_event_list; + spinlock_t spin_lock; + struct completion frame_done_comp; + + /* for handling internal event thread */ + spinlock_t event_lock; + bool misr_enable; + u32 misr_frame_count; + u32 misr_data[CRTC_DUAL_MIXERS]; + + struct dpu_power_handle *phandle; + struct dpu_power_event *power_event; + + struct dpu_core_perf_params cur_perf; + + struct mutex rp_lock; + struct list_head rp_head; + + struct dpu_crtc_smmu_state_data smmu_state; +}; + +#define to_dpu_crtc(x) container_of(x, struct dpu_crtc, base) + +/** + * struct dpu_crtc_res_ops - common operations for crtc resources + * @get: get given resource + * @put: put given resource + */ +struct dpu_crtc_res_ops { + void *(*get)(void *val, u32 type, u64 tag); + void (*put)(void *val); +}; + +#define DPU_CRTC_RES_FLAG_FREE BIT(0) + +/** + * struct dpu_crtc_res - definition of crtc resources + * @list: list of crtc resource + * @type: crtc resource type + * @tag: unique identifier per type + * @refcount: reference/usage count + * @ops: callback operations + * @val: resource handle associated with type/tag + * @flags: customization flags + */ +struct dpu_crtc_res { + struct list_head list; + u32 type; + u64 tag; + atomic_t refcount; + struct dpu_crtc_res_ops ops; + void *val; + u32 flags; +}; + +/** + * dpu_crtc_respool - crtc resource pool + * @rp_lock: pointer to serialization lock + * @rp_head: pointer to head of active resource pools of this crtc + * @rp_list: list of crtc resource pool + * @sequence_id: sequence identifier, incremented per state duplication + * @res_list: list of resource managed by this resource pool + * @ops: resource operations for parent resource pool + */ +struct dpu_crtc_respool { + struct mutex *rp_lock; + struct list_head *rp_head; + struct list_head rp_list; + u32 sequence_id; + struct list_head res_list; + struct dpu_crtc_res_ops ops; +}; + +/** + * struct dpu_crtc_state - dpu container for atomic crtc state + * @base: Base drm crtc state structure + * @is_ppsplit : Whether current topology requires PPSplit special handling + * @bw_control : true if bw/clk controlled by core bw/clk properties + * @bw_split_vote : true if bw controlled by llcc/dram bw properties + * @lm_bounds : LM boundaries based on current mode full resolution, no ROI. + * Origin top left of CRTC. + * @property_state: Local storage for msm_prop properties + * @property_values: Current crtc property values + * @input_fence_timeout_ns : Cached input fence timeout, in ns + * @new_perf: new performance state being requested + */ +struct dpu_crtc_state { + struct drm_crtc_state base; + + bool bw_control; + bool bw_split_vote; + + bool is_ppsplit; + struct drm_rect lm_bounds[CRTC_DUAL_MIXERS]; + + uint64_t input_fence_timeout_ns; + + struct dpu_core_perf_params new_perf; + struct dpu_crtc_respool rp; +}; + +#define to_dpu_crtc_state(x) \ + container_of(x, struct dpu_crtc_state, base) + +/** + * dpu_crtc_get_mixer_width - get the mixer width + * Mixer width will be same as panel width(/2 for split) + */ +static inline int dpu_crtc_get_mixer_width(struct dpu_crtc *dpu_crtc, + struct dpu_crtc_state *cstate, struct drm_display_mode *mode) +{ + u32 mixer_width; + + if (!dpu_crtc || !cstate || !mode) + return 0; + + mixer_width = (dpu_crtc->num_mixers == CRTC_DUAL_MIXERS ? + mode->hdisplay / CRTC_DUAL_MIXERS : mode->hdisplay); + + return mixer_width; +} + +/** + * dpu_crtc_get_mixer_height - get the mixer height + * Mixer height will be same as panel height + */ +static inline int dpu_crtc_get_mixer_height(struct dpu_crtc *dpu_crtc, + struct dpu_crtc_state *cstate, struct drm_display_mode *mode) +{ + if (!dpu_crtc || !cstate || !mode) + return 0; + + return mode->vdisplay; +} + +/** + * dpu_crtc_frame_pending - retun the number of pending frames + * @crtc: Pointer to drm crtc object + */ +static inline int dpu_crtc_frame_pending(struct drm_crtc *crtc) +{ + struct dpu_crtc *dpu_crtc; + + if (!crtc) + return -EINVAL; + + dpu_crtc = to_dpu_crtc(crtc); + return atomic_read(&dpu_crtc->frame_pending); +} + +/** + * dpu_crtc_vblank - enable or disable vblanks for this crtc + * @crtc: Pointer to drm crtc object + * @en: true to enable vblanks, false to disable + */ +int dpu_crtc_vblank(struct drm_crtc *crtc, bool en); + +/** + * dpu_crtc_commit_kickoff - trigger kickoff of the commit for this crtc + * @crtc: Pointer to drm crtc object + */ +void dpu_crtc_commit_kickoff(struct drm_crtc *crtc); + +/** + * dpu_crtc_complete_commit - callback signalling completion of current commit + * @crtc: Pointer to drm crtc object + * @old_state: Pointer to drm crtc old state object + */ +void dpu_crtc_complete_commit(struct drm_crtc *crtc, + struct drm_crtc_state *old_state); + +/** + * dpu_crtc_init - create a new crtc object + * @dev: dpu device + * @plane: base plane + * @Return: new crtc object or error + */ +struct drm_crtc *dpu_crtc_init(struct drm_device *dev, struct drm_plane *plane); + +/** + * dpu_crtc_register_custom_event - api for enabling/disabling crtc event + * @kms: Pointer to dpu_kms + * @crtc_drm: Pointer to crtc object + * @event: Event that client is interested + * @en: Flag to enable/disable the event + */ +int dpu_crtc_register_custom_event(struct dpu_kms *kms, + struct drm_crtc *crtc_drm, u32 event, bool en); + +/** + * dpu_crtc_get_intf_mode - get interface mode of the given crtc + * @crtc: Pointert to crtc + */ +enum dpu_intf_mode dpu_crtc_get_intf_mode(struct drm_crtc *crtc); + +/** + * dpu_crtc_get_client_type - check the crtc type- rt, nrt etc. + * @crtc: Pointer to crtc + */ +static inline enum dpu_crtc_client_type dpu_crtc_get_client_type( + struct drm_crtc *crtc) +{ + struct dpu_crtc_state *cstate = + crtc ? to_dpu_crtc_state(crtc->state) : NULL; + + if (!cstate) + return NRT_CLIENT; + + return RT_CLIENT; +} + +/** + * dpu_crtc_is_enabled - check if dpu crtc is enabled or not + * @crtc: Pointer to crtc + */ +static inline bool dpu_crtc_is_enabled(struct drm_crtc *crtc) +{ + return crtc ? crtc->enabled : false; +} + +#endif /* _DPU_CRTC_H_ */ diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_dbg.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_dbg.c new file mode 100644 index 000000000000..ae2aee7ed9e1 --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_dbg.c @@ -0,0 +1,2393 @@ +/* Copyright (c) 2009-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#define pr_fmt(fmt) "[drm:%s:%d] " fmt, __func__, __LINE__ + +#include <linux/delay.h> +#include <linux/spinlock.h> +#include <linux/ktime.h> +#include <linux/debugfs.h> +#include <linux/uaccess.h> +#include <linux/dma-buf.h> +#include <linux/slab.h> +#include <linux/list_sort.h> +#include <linux/pm_runtime.h> + +#include "dpu_dbg.h" +#include "disp/dpu1/dpu_hw_catalog.h" + + +#define DEFAULT_DBGBUS_DPU DPU_DBG_DUMP_IN_MEM +#define DEFAULT_DBGBUS_VBIFRT DPU_DBG_DUMP_IN_MEM +#define REG_BASE_NAME_LEN 80 + +#define DBGBUS_FLAGS_DSPP BIT(0) +#define DBGBUS_DSPP_STATUS 0x34C + +#define DBGBUS_NAME_DPU "dpu" +#define DBGBUS_NAME_VBIF_RT "vbif_rt" + +/* offsets from dpu top address for the debug buses */ +#define DBGBUS_SSPP0 0x188 +#define DBGBUS_AXI_INTF 0x194 +#define DBGBUS_SSPP1 0x298 +#define DBGBUS_DSPP 0x348 +#define DBGBUS_PERIPH 0x418 + +#define TEST_MASK(id, tp) ((id << 4) | (tp << 1) | BIT(0)) + +/* following offsets are with respect to MDP VBIF base for DBG BUS access */ +#define MMSS_VBIF_CLKON 0x4 +#define MMSS_VBIF_TEST_BUS_OUT_CTRL 0x210 +#define MMSS_VBIF_TEST_BUS_OUT 0x230 + +/* Vbif error info */ +#define MMSS_VBIF_PND_ERR 0x190 +#define MMSS_VBIF_SRC_ERR 0x194 +#define MMSS_VBIF_XIN_HALT_CTRL1 0x204 +#define MMSS_VBIF_ERR_INFO 0X1a0 +#define MMSS_VBIF_ERR_INFO_1 0x1a4 +#define MMSS_VBIF_CLIENT_NUM 14 + +/** + * struct dpu_dbg_reg_base - register region base. + * may sub-ranges: sub-ranges are used for dumping + * or may not have sub-ranges: dumping is base -> max_offset + * @reg_base_head: head of this node + * @name: register base name + * @base: base pointer + * @off: cached offset of region for manual register dumping + * @cnt: cached range of region for manual register dumping + * @max_offset: length of region + * @buf: buffer used for manual register dumping + * @buf_len: buffer length used for manual register dumping + * @cb: callback for external dump function, null if not defined + * @cb_ptr: private pointer to callback function + */ +struct dpu_dbg_reg_base { + struct list_head reg_base_head; + char name[REG_BASE_NAME_LEN]; + void __iomem *base; + size_t off; + size_t cnt; + size_t max_offset; + char *buf; + size_t buf_len; + void (*cb)(void *ptr); + void *cb_ptr; +}; + +struct dpu_debug_bus_entry { + u32 wr_addr; + u32 block_id; + u32 test_id; + void (*analyzer)(void __iomem *mem_base, + struct dpu_debug_bus_entry *entry, u32 val); +}; + +struct vbif_debug_bus_entry { + u32 disable_bus_addr; + u32 block_bus_addr; + u32 bit_offset; + u32 block_cnt; + u32 test_pnt_start; + u32 test_pnt_cnt; +}; + +struct dpu_dbg_debug_bus_common { + char *name; + u32 enable_mask; + bool include_in_deferred_work; + u32 flags; + u32 entries_size; + u32 *dumped_content; +}; + +struct dpu_dbg_dpu_debug_bus { + struct dpu_dbg_debug_bus_common cmn; + struct dpu_debug_bus_entry *entries; + u32 top_blk_off; +}; + +struct dpu_dbg_vbif_debug_bus { + struct dpu_dbg_debug_bus_common cmn; + struct vbif_debug_bus_entry *entries; +}; + +/** + * struct dpu_dbg_base - global dpu debug base structure + * @reg_base_list: list of register dumping regions + * @dev: device pointer + * @dump_work: work struct for deferring register dump work to separate thread + * @dbgbus_dpu: debug bus structure for the dpu + * @dbgbus_vbif_rt: debug bus structure for the realtime vbif + */ +static struct dpu_dbg_base { + struct list_head reg_base_list; + struct device *dev; + + struct work_struct dump_work; + + struct dpu_dbg_dpu_debug_bus dbgbus_dpu; + struct dpu_dbg_vbif_debug_bus dbgbus_vbif_rt; +} dpu_dbg_base; + +static void _dpu_debug_bus_xbar_dump(void __iomem *mem_base, + struct dpu_debug_bus_entry *entry, u32 val) +{ + dev_err(dpu_dbg_base.dev, "xbar 0x%x %d %d 0x%x\n", + entry->wr_addr, entry->block_id, entry->test_id, val); +} + +static void _dpu_debug_bus_lm_dump(void __iomem *mem_base, + struct dpu_debug_bus_entry *entry, u32 val) +{ + if (!(val & 0xFFF000)) + return; + + dev_err(dpu_dbg_base.dev, "lm 0x%x %d %d 0x%x\n", + entry->wr_addr, entry->block_id, entry->test_id, val); +} + +static void _dpu_debug_bus_ppb0_dump(void __iomem *mem_base, + struct dpu_debug_bus_entry *entry, u32 val) +{ + if (!(val & BIT(15))) + return; + + dev_err(dpu_dbg_base.dev, "ppb0 0x%x %d %d 0x%x\n", + entry->wr_addr, entry->block_id, entry->test_id, val); +} + +static void _dpu_debug_bus_ppb1_dump(void __iomem *mem_base, + struct dpu_debug_bus_entry *entry, u32 val) +{ + if (!(val & BIT(15))) + return; + + dev_err(dpu_dbg_base.dev, "ppb1 0x%x %d %d 0x%x\n", + entry->wr_addr, entry->block_id, entry->test_id, val); +} + +static struct dpu_debug_bus_entry dbg_bus_dpu_8998[] = { + + /* Unpack 0 sspp 0*/ + { DBGBUS_SSPP0, 50, 2 }, + { DBGBUS_SSPP0, 60, 2 }, + { DBGBUS_SSPP0, 70, 2 }, + { DBGBUS_SSPP0, 85, 2 }, + + /* Upack 0 sspp 1*/ + { DBGBUS_SSPP1, 50, 2 }, + { DBGBUS_SSPP1, 60, 2 }, + { DBGBUS_SSPP1, 70, 2 }, + { DBGBUS_SSPP1, 85, 2 }, + + /* scheduler */ + { DBGBUS_DSPP, 130, 0 }, + { DBGBUS_DSPP, 130, 1 }, + { DBGBUS_DSPP, 130, 2 }, + { DBGBUS_DSPP, 130, 3 }, + { DBGBUS_DSPP, 130, 4 }, + { DBGBUS_DSPP, 130, 5 }, + + /* qseed */ + { DBGBUS_SSPP0, 6, 0}, + { DBGBUS_SSPP0, 6, 1}, + { DBGBUS_SSPP0, 26, 0}, + { DBGBUS_SSPP0, 26, 1}, + { DBGBUS_SSPP1, 6, 0}, + { DBGBUS_SSPP1, 6, 1}, + { DBGBUS_SSPP1, 26, 0}, + { DBGBUS_SSPP1, 26, 1}, + + /* scale */ + { DBGBUS_SSPP0, 16, 0}, + { DBGBUS_SSPP0, 16, 1}, + { DBGBUS_SSPP0, 36, 0}, + { DBGBUS_SSPP0, 36, 1}, + { DBGBUS_SSPP1, 16, 0}, + { DBGBUS_SSPP1, 16, 1}, + { DBGBUS_SSPP1, 36, 0}, + { DBGBUS_SSPP1, 36, 1}, + + /* fetch sspp0 */ + + /* vig 0 */ + { DBGBUS_SSPP0, 0, 0 }, + { DBGBUS_SSPP0, 0, 1 }, + { DBGBUS_SSPP0, 0, 2 }, + { DBGBUS_SSPP0, 0, 3 }, + { DBGBUS_SSPP0, 0, 4 }, + { DBGBUS_SSPP0, 0, 5 }, + { DBGBUS_SSPP0, 0, 6 }, + { DBGBUS_SSPP0, 0, 7 }, + + { DBGBUS_SSPP0, 1, 0 }, + { DBGBUS_SSPP0, 1, 1 }, + { DBGBUS_SSPP0, 1, 2 }, + { DBGBUS_SSPP0, 1, 3 }, + { DBGBUS_SSPP0, 1, 4 }, + { DBGBUS_SSPP0, 1, 5 }, + { DBGBUS_SSPP0, 1, 6 }, + { DBGBUS_SSPP0, 1, 7 }, + + { DBGBUS_SSPP0, 2, 0 }, + { DBGBUS_SSPP0, 2, 1 }, + { DBGBUS_SSPP0, 2, 2 }, + { DBGBUS_SSPP0, 2, 3 }, + { DBGBUS_SSPP0, 2, 4 }, + { DBGBUS_SSPP0, 2, 5 }, + { DBGBUS_SSPP0, 2, 6 }, + { DBGBUS_SSPP0, 2, 7 }, + + { DBGBUS_SSPP0, 4, 0 }, + { DBGBUS_SSPP0, 4, 1 }, + { DBGBUS_SSPP0, 4, 2 }, + { DBGBUS_SSPP0, 4, 3 }, + { DBGBUS_SSPP0, 4, 4 }, + { DBGBUS_SSPP0, 4, 5 }, + { DBGBUS_SSPP0, 4, 6 }, + { DBGBUS_SSPP0, 4, 7 }, + + { DBGBUS_SSPP0, 5, 0 }, + { DBGBUS_SSPP0, 5, 1 }, + { DBGBUS_SSPP0, 5, 2 }, + { DBGBUS_SSPP0, 5, 3 }, + { DBGBUS_SSPP0, 5, 4 }, + { DBGBUS_SSPP0, 5, 5 }, + { DBGBUS_SSPP0, 5, 6 }, + { DBGBUS_SSPP0, 5, 7 }, + + /* vig 2 */ + { DBGBUS_SSPP0, 20, 0 }, + { DBGBUS_SSPP0, 20, 1 }, + { DBGBUS_SSPP0, 20, 2 }, + { DBGBUS_SSPP0, 20, 3 }, + { DBGBUS_SSPP0, 20, 4 }, + { DBGBUS_SSPP0, 20, 5 }, + { DBGBUS_SSPP0, 20, 6 }, + { DBGBUS_SSPP0, 20, 7 }, + + { DBGBUS_SSPP0, 21, 0 }, + { DBGBUS_SSPP0, 21, 1 }, + { DBGBUS_SSPP0, 21, 2 }, + { DBGBUS_SSPP0, 21, 3 }, + { DBGBUS_SSPP0, 21, 4 }, + { DBGBUS_SSPP0, 21, 5 }, + { DBGBUS_SSPP0, 21, 6 }, + { DBGBUS_SSPP0, 21, 7 }, + + { DBGBUS_SSPP0, 22, 0 }, + { DBGBUS_SSPP0, 22, 1 }, + { DBGBUS_SSPP0, 22, 2 }, + { DBGBUS_SSPP0, 22, 3 }, + { DBGBUS_SSPP0, 22, 4 }, + { DBGBUS_SSPP0, 22, 5 }, + { DBGBUS_SSPP0, 22, 6 }, + { DBGBUS_SSPP0, 22, 7 }, + + { DBGBUS_SSPP0, 24, 0 }, + { DBGBUS_SSPP0, 24, 1 }, + { DBGBUS_SSPP0, 24, 2 }, + { DBGBUS_SSPP0, 24, 3 }, + { DBGBUS_SSPP0, 24, 4 }, + { DBGBUS_SSPP0, 24, 5 }, + { DBGBUS_SSPP0, 24, 6 }, + { DBGBUS_SSPP0, 24, 7 }, + + { DBGBUS_SSPP0, 25, 0 }, + { DBGBUS_SSPP0, 25, 1 }, + { DBGBUS_SSPP0, 25, 2 }, + { DBGBUS_SSPP0, 25, 3 }, + { DBGBUS_SSPP0, 25, 4 }, + { DBGBUS_SSPP0, 25, 5 }, + { DBGBUS_SSPP0, 25, 6 }, + { DBGBUS_SSPP0, 25, 7 }, + + /* dma 2 */ + { DBGBUS_SSPP0, 30, 0 }, + { DBGBUS_SSPP0, 30, 1 }, + { DBGBUS_SSPP0, 30, 2 }, + { DBGBUS_SSPP0, 30, 3 }, + { DBGBUS_SSPP0, 30, 4 }, + { DBGBUS_SSPP0, 30, 5 }, + { DBGBUS_SSPP0, 30, 6 }, + { DBGBUS_SSPP0, 30, 7 }, + + { DBGBUS_SSPP0, 31, 0 }, + { DBGBUS_SSPP0, 31, 1 }, + { DBGBUS_SSPP0, 31, 2 }, + { DBGBUS_SSPP0, 31, 3 }, + { DBGBUS_SSPP0, 31, 4 }, + { DBGBUS_SSPP0, 31, 5 }, + { DBGBUS_SSPP0, 31, 6 }, + { DBGBUS_SSPP0, 31, 7 }, + + { DBGBUS_SSPP0, 32, 0 }, + { DBGBUS_SSPP0, 32, 1 }, + { DBGBUS_SSPP0, 32, 2 }, + { DBGBUS_SSPP0, 32, 3 }, + { DBGBUS_SSPP0, 32, 4 }, + { DBGBUS_SSPP0, 32, 5 }, + { DBGBUS_SSPP0, 32, 6 }, + { DBGBUS_SSPP0, 32, 7 }, + + { DBGBUS_SSPP0, 33, 0 }, + { DBGBUS_SSPP0, 33, 1 }, + { DBGBUS_SSPP0, 33, 2 }, + { DBGBUS_SSPP0, 33, 3 }, + { DBGBUS_SSPP0, 33, 4 }, + { DBGBUS_SSPP0, 33, 5 }, + { DBGBUS_SSPP0, 33, 6 }, + { DBGBUS_SSPP0, 33, 7 }, + + { DBGBUS_SSPP0, 34, 0 }, + { DBGBUS_SSPP0, 34, 1 }, + { DBGBUS_SSPP0, 34, 2 }, + { DBGBUS_SSPP0, 34, 3 }, + { DBGBUS_SSPP0, 34, 4 }, + { DBGBUS_SSPP0, 34, 5 }, + { DBGBUS_SSPP0, 34, 6 }, + { DBGBUS_SSPP0, 34, 7 }, + + { DBGBUS_SSPP0, 35, 0 }, + { DBGBUS_SSPP0, 35, 1 }, + { DBGBUS_SSPP0, 35, 2 }, + { DBGBUS_SSPP0, 35, 3 }, + + /* dma 0 */ + { DBGBUS_SSPP0, 40, 0 }, + { DBGBUS_SSPP0, 40, 1 }, + { DBGBUS_SSPP0, 40, 2 }, + { DBGBUS_SSPP0, 40, 3 }, + { DBGBUS_SSPP0, 40, 4 }, + { DBGBUS_SSPP0, 40, 5 }, + { DBGBUS_SSPP0, 40, 6 }, + { DBGBUS_SSPP0, 40, 7 }, + + { DBGBUS_SSPP0, 41, 0 }, + { DBGBUS_SSPP0, 41, 1 }, + { DBGBUS_SSPP0, 41, 2 }, + { DBGBUS_SSPP0, 41, 3 }, + { DBGBUS_SSPP0, 41, 4 }, + { DBGBUS_SSPP0, 41, 5 }, + { DBGBUS_SSPP0, 41, 6 }, + { DBGBUS_SSPP0, 41, 7 }, + + { DBGBUS_SSPP0, 42, 0 }, + { DBGBUS_SSPP0, 42, 1 }, + { DBGBUS_SSPP0, 42, 2 }, + { DBGBUS_SSPP0, 42, 3 }, + { DBGBUS_SSPP0, 42, 4 }, + { DBGBUS_SSPP0, 42, 5 }, + { DBGBUS_SSPP0, 42, 6 }, + { DBGBUS_SSPP0, 42, 7 }, + + { DBGBUS_SSPP0, 44, 0 }, + { DBGBUS_SSPP0, 44, 1 }, + { DBGBUS_SSPP0, 44, 2 }, + { DBGBUS_SSPP0, 44, 3 }, + { DBGBUS_SSPP0, 44, 4 }, + { DBGBUS_SSPP0, 44, 5 }, + { DBGBUS_SSPP0, 44, 6 }, + { DBGBUS_SSPP0, 44, 7 }, + + { DBGBUS_SSPP0, 45, 0 }, + { DBGBUS_SSPP0, 45, 1 }, + { DBGBUS_SSPP0, 45, 2 }, + { DBGBUS_SSPP0, 45, 3 }, + { DBGBUS_SSPP0, 45, 4 }, + { DBGBUS_SSPP0, 45, 5 }, + { DBGBUS_SSPP0, 45, 6 }, + { DBGBUS_SSPP0, 45, 7 }, + + /* fetch sspp1 */ + /* vig 1 */ + { DBGBUS_SSPP1, 0, 0 }, + { DBGBUS_SSPP1, 0, 1 }, + { DBGBUS_SSPP1, 0, 2 }, + { DBGBUS_SSPP1, 0, 3 }, + { DBGBUS_SSPP1, 0, 4 }, + { DBGBUS_SSPP1, 0, 5 }, + { DBGBUS_SSPP1, 0, 6 }, + { DBGBUS_SSPP1, 0, 7 }, + + { DBGBUS_SSPP1, 1, 0 }, + { DBGBUS_SSPP1, 1, 1 }, + { DBGBUS_SSPP1, 1, 2 }, + { DBGBUS_SSPP1, 1, 3 }, + { DBGBUS_SSPP1, 1, 4 }, + { DBGBUS_SSPP1, 1, 5 }, + { DBGBUS_SSPP1, 1, 6 }, + { DBGBUS_SSPP1, 1, 7 }, + + { DBGBUS_SSPP1, 2, 0 }, + { DBGBUS_SSPP1, 2, 1 }, + { DBGBUS_SSPP1, 2, 2 }, + { DBGBUS_SSPP1, 2, 3 }, + { DBGBUS_SSPP1, 2, 4 }, + { DBGBUS_SSPP1, 2, 5 }, + { DBGBUS_SSPP1, 2, 6 }, + { DBGBUS_SSPP1, 2, 7 }, + + { DBGBUS_SSPP1, 4, 0 }, + { DBGBUS_SSPP1, 4, 1 }, + { DBGBUS_SSPP1, 4, 2 }, + { DBGBUS_SSPP1, 4, 3 }, + { DBGBUS_SSPP1, 4, 4 }, + { DBGBUS_SSPP1, 4, 5 }, + { DBGBUS_SSPP1, 4, 6 }, + { DBGBUS_SSPP1, 4, 7 }, + + { DBGBUS_SSPP1, 5, 0 }, + { DBGBUS_SSPP1, 5, 1 }, + { DBGBUS_SSPP1, 5, 2 }, + { DBGBUS_SSPP1, 5, 3 }, + { DBGBUS_SSPP1, 5, 4 }, + { DBGBUS_SSPP1, 5, 5 }, + { DBGBUS_SSPP1, 5, 6 }, + { DBGBUS_SSPP1, 5, 7 }, + + /* vig 3 */ + { DBGBUS_SSPP1, 20, 0 }, + { DBGBUS_SSPP1, 20, 1 }, + { DBGBUS_SSPP1, 20, 2 }, + { DBGBUS_SSPP1, 20, 3 }, + { DBGBUS_SSPP1, 20, 4 }, + { DBGBUS_SSPP1, 20, 5 }, + { DBGBUS_SSPP1, 20, 6 }, + { DBGBUS_SSPP1, 20, 7 }, + + { DBGBUS_SSPP1, 21, 0 }, + { DBGBUS_SSPP1, 21, 1 }, + { DBGBUS_SSPP1, 21, 2 }, + { DBGBUS_SSPP1, 21, 3 }, + { DBGBUS_SSPP1, 21, 4 }, + { DBGBUS_SSPP1, 21, 5 }, + { DBGBUS_SSPP1, 21, 6 }, + { DBGBUS_SSPP1, 21, 7 }, + + { DBGBUS_SSPP1, 22, 0 }, + { DBGBUS_SSPP1, 22, 1 }, + { DBGBUS_SSPP1, 22, 2 }, + { DBGBUS_SSPP1, 22, 3 }, + { DBGBUS_SSPP1, 22, 4 }, + { DBGBUS_SSPP1, 22, 5 }, + { DBGBUS_SSPP1, 22, 6 }, + { DBGBUS_SSPP1, 22, 7 }, + + { DBGBUS_SSPP1, 24, 0 }, + { DBGBUS_SSPP1, 24, 1 }, + { DBGBUS_SSPP1, 24, 2 }, + { DBGBUS_SSPP1, 24, 3 }, + { DBGBUS_SSPP1, 24, 4 }, + { DBGBUS_SSPP1, 24, 5 }, + { DBGBUS_SSPP1, 24, 6 }, + { DBGBUS_SSPP1, 24, 7 }, + + { DBGBUS_SSPP1, 25, 0 }, + { DBGBUS_SSPP1, 25, 1 }, + { DBGBUS_SSPP1, 25, 2 }, + { DBGBUS_SSPP1, 25, 3 }, + { DBGBUS_SSPP1, 25, 4 }, + { DBGBUS_SSPP1, 25, 5 }, + { DBGBUS_SSPP1, 25, 6 }, + { DBGBUS_SSPP1, 25, 7 }, + + /* dma 3 */ + { DBGBUS_SSPP1, 30, 0 }, + { DBGBUS_SSPP1, 30, 1 }, + { DBGBUS_SSPP1, 30, 2 }, + { DBGBUS_SSPP1, 30, 3 }, + { DBGBUS_SSPP1, 30, 4 }, + { DBGBUS_SSPP1, 30, 5 }, + { DBGBUS_SSPP1, 30, 6 }, + { DBGBUS_SSPP1, 30, 7 }, + + { DBGBUS_SSPP1, 31, 0 }, + { DBGBUS_SSPP1, 31, 1 }, + { DBGBUS_SSPP1, 31, 2 }, + { DBGBUS_SSPP1, 31, 3 }, + { DBGBUS_SSPP1, 31, 4 }, + { DBGBUS_SSPP1, 31, 5 }, + { DBGBUS_SSPP1, 31, 6 }, + { DBGBUS_SSPP1, 31, 7 }, + + { DBGBUS_SSPP1, 32, 0 }, + { DBGBUS_SSPP1, 32, 1 }, + { DBGBUS_SSPP1, 32, 2 }, + { DBGBUS_SSPP1, 32, 3 }, + { DBGBUS_SSPP1, 32, 4 }, + { DBGBUS_SSPP1, 32, 5 }, + { DBGBUS_SSPP1, 32, 6 }, + { DBGBUS_SSPP1, 32, 7 }, + + { DBGBUS_SSPP1, 33, 0 }, + { DBGBUS_SSPP1, 33, 1 }, + { DBGBUS_SSPP1, 33, 2 }, + { DBGBUS_SSPP1, 33, 3 }, + { DBGBUS_SSPP1, 33, 4 }, + { DBGBUS_SSPP1, 33, 5 }, + { DBGBUS_SSPP1, 33, 6 }, + { DBGBUS_SSPP1, 33, 7 }, + + { DBGBUS_SSPP1, 34, 0 }, + { DBGBUS_SSPP1, 34, 1 }, + { DBGBUS_SSPP1, 34, 2 }, + { DBGBUS_SSPP1, 34, 3 }, + { DBGBUS_SSPP1, 34, 4 }, + { DBGBUS_SSPP1, 34, 5 }, + { DBGBUS_SSPP1, 34, 6 }, + { DBGBUS_SSPP1, 34, 7 }, + + { DBGBUS_SSPP1, 35, 0 }, + { DBGBUS_SSPP1, 35, 1 }, + { DBGBUS_SSPP1, 35, 2 }, + + /* dma 1 */ + { DBGBUS_SSPP1, 40, 0 }, + { DBGBUS_SSPP1, 40, 1 }, + { DBGBUS_SSPP1, 40, 2 }, + { DBGBUS_SSPP1, 40, 3 }, + { DBGBUS_SSPP1, 40, 4 }, + { DBGBUS_SSPP1, 40, 5 }, + { DBGBUS_SSPP1, 40, 6 }, + { DBGBUS_SSPP1, 40, 7 }, + + { DBGBUS_SSPP1, 41, 0 }, + { DBGBUS_SSPP1, 41, 1 }, + { DBGBUS_SSPP1, 41, 2 }, + { DBGBUS_SSPP1, 41, 3 }, + { DBGBUS_SSPP1, 41, 4 }, + { DBGBUS_SSPP1, 41, 5 }, + { DBGBUS_SSPP1, 41, 6 }, + { DBGBUS_SSPP1, 41, 7 }, + + { DBGBUS_SSPP1, 42, 0 }, + { DBGBUS_SSPP1, 42, 1 }, + { DBGBUS_SSPP1, 42, 2 }, + { DBGBUS_SSPP1, 42, 3 }, + { DBGBUS_SSPP1, 42, 4 }, + { DBGBUS_SSPP1, 42, 5 }, + { DBGBUS_SSPP1, 42, 6 }, + { DBGBUS_SSPP1, 42, 7 }, + + { DBGBUS_SSPP1, 44, 0 }, + { DBGBUS_SSPP1, 44, 1 }, + { DBGBUS_SSPP1, 44, 2 }, + { DBGBUS_SSPP1, 44, 3 }, + { DBGBUS_SSPP1, 44, 4 }, + { DBGBUS_SSPP1, 44, 5 }, + { DBGBUS_SSPP1, 44, 6 }, + { DBGBUS_SSPP1, 44, 7 }, + + { DBGBUS_SSPP1, 45, 0 }, + { DBGBUS_SSPP1, 45, 1 }, + { DBGBUS_SSPP1, 45, 2 }, + { DBGBUS_SSPP1, 45, 3 }, + { DBGBUS_SSPP1, 45, 4 }, + { DBGBUS_SSPP1, 45, 5 }, + { DBGBUS_SSPP1, 45, 6 }, + { DBGBUS_SSPP1, 45, 7 }, + + /* cursor 1 */ + { DBGBUS_SSPP1, 80, 0 }, + { DBGBUS_SSPP1, 80, 1 }, + { DBGBUS_SSPP1, 80, 2 }, + { DBGBUS_SSPP1, 80, 3 }, + { DBGBUS_SSPP1, 80, 4 }, + { DBGBUS_SSPP1, 80, 5 }, + { DBGBUS_SSPP1, 80, 6 }, + { DBGBUS_SSPP1, 80, 7 }, + + { DBGBUS_SSPP1, 81, 0 }, + { DBGBUS_SSPP1, 81, 1 }, + { DBGBUS_SSPP1, 81, 2 }, + { DBGBUS_SSPP1, 81, 3 }, + { DBGBUS_SSPP1, 81, 4 }, + { DBGBUS_SSPP1, 81, 5 }, + { DBGBUS_SSPP1, 81, 6 }, + { DBGBUS_SSPP1, 81, 7 }, + + { DBGBUS_SSPP1, 82, 0 }, + { DBGBUS_SSPP1, 82, 1 }, + { DBGBUS_SSPP1, 82, 2 }, + { DBGBUS_SSPP1, 82, 3 }, + { DBGBUS_SSPP1, 82, 4 }, + { DBGBUS_SSPP1, 82, 5 }, + { DBGBUS_SSPP1, 82, 6 }, + { DBGBUS_SSPP1, 82, 7 }, + + { DBGBUS_SSPP1, 83, 0 }, + { DBGBUS_SSPP1, 83, 1 }, + { DBGBUS_SSPP1, 83, 2 }, + { DBGBUS_SSPP1, 83, 3 }, + { DBGBUS_SSPP1, 83, 4 }, + { DBGBUS_SSPP1, 83, 5 }, + { DBGBUS_SSPP1, 83, 6 }, + { DBGBUS_SSPP1, 83, 7 }, + + { DBGBUS_SSPP1, 84, 0 }, + { DBGBUS_SSPP1, 84, 1 }, + { DBGBUS_SSPP1, 84, 2 }, + { DBGBUS_SSPP1, 84, 3 }, + { DBGBUS_SSPP1, 84, 4 }, + { DBGBUS_SSPP1, 84, 5 }, + { DBGBUS_SSPP1, 84, 6 }, + { DBGBUS_SSPP1, 84, 7 }, + + /* dspp */ + { DBGBUS_DSPP, 13, 0 }, + { DBGBUS_DSPP, 19, 0 }, + { DBGBUS_DSPP, 14, 0 }, + { DBGBUS_DSPP, 14, 1 }, + { DBGBUS_DSPP, 14, 3 }, + { DBGBUS_DSPP, 20, 0 }, + { DBGBUS_DSPP, 20, 1 }, + { DBGBUS_DSPP, 20, 3 }, + + /* ppb_0 */ + { DBGBUS_DSPP, 31, 0, _dpu_debug_bus_ppb0_dump }, + { DBGBUS_DSPP, 33, 0, _dpu_debug_bus_ppb0_dump }, + { DBGBUS_DSPP, 35, 0, _dpu_debug_bus_ppb0_dump }, + { DBGBUS_DSPP, 42, 0, _dpu_debug_bus_ppb0_dump }, + + /* ppb_1 */ + { DBGBUS_DSPP, 32, 0, _dpu_debug_bus_ppb1_dump }, + { DBGBUS_DSPP, 34, 0, _dpu_debug_bus_ppb1_dump }, + { DBGBUS_DSPP, 36, 0, _dpu_debug_bus_ppb1_dump }, + { DBGBUS_DSPP, 43, 0, _dpu_debug_bus_ppb1_dump }, + + /* lm_lut */ + { DBGBUS_DSPP, 109, 0 }, + { DBGBUS_DSPP, 105, 0 }, + { DBGBUS_DSPP, 103, 0 }, + + /* tear-check */ + { DBGBUS_PERIPH, 63, 0 }, + { DBGBUS_PERIPH, 64, 0 }, + { DBGBUS_PERIPH, 65, 0 }, + { DBGBUS_PERIPH, 73, 0 }, + { DBGBUS_PERIPH, 74, 0 }, + + /* crossbar */ + { DBGBUS_DSPP, 0, 0, _dpu_debug_bus_xbar_dump }, + + /* rotator */ + { DBGBUS_DSPP, 9, 0}, + + /* blend */ + /* LM0 */ + { DBGBUS_DSPP, 63, 0}, + { DBGBUS_DSPP, 63, 1}, + { DBGBUS_DSPP, 63, 2}, + { DBGBUS_DSPP, 63, 3}, + { DBGBUS_DSPP, 63, 4}, + { DBGBUS_DSPP, 63, 5}, + { DBGBUS_DSPP, 63, 6}, + { DBGBUS_DSPP, 63, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 64, 0}, + { DBGBUS_DSPP, 64, 1}, + { DBGBUS_DSPP, 64, 2}, + { DBGBUS_DSPP, 64, 3}, + { DBGBUS_DSPP, 64, 4}, + { DBGBUS_DSPP, 64, 5}, + { DBGBUS_DSPP, 64, 6}, + { DBGBUS_DSPP, 64, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 65, 0}, + { DBGBUS_DSPP, 65, 1}, + { DBGBUS_DSPP, 65, 2}, + { DBGBUS_DSPP, 65, 3}, + { DBGBUS_DSPP, 65, 4}, + { DBGBUS_DSPP, 65, 5}, + { DBGBUS_DSPP, 65, 6}, + { DBGBUS_DSPP, 65, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 66, 0}, + { DBGBUS_DSPP, 66, 1}, + { DBGBUS_DSPP, 66, 2}, + { DBGBUS_DSPP, 66, 3}, + { DBGBUS_DSPP, 66, 4}, + { DBGBUS_DSPP, 66, 5}, + { DBGBUS_DSPP, 66, 6}, + { DBGBUS_DSPP, 66, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 67, 0}, + { DBGBUS_DSPP, 67, 1}, + { DBGBUS_DSPP, 67, 2}, + { DBGBUS_DSPP, 67, 3}, + { DBGBUS_DSPP, 67, 4}, + { DBGBUS_DSPP, 67, 5}, + { DBGBUS_DSPP, 67, 6}, + { DBGBUS_DSPP, 67, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 68, 0}, + { DBGBUS_DSPP, 68, 1}, + { DBGBUS_DSPP, 68, 2}, + { DBGBUS_DSPP, 68, 3}, + { DBGBUS_DSPP, 68, 4}, + { DBGBUS_DSPP, 68, 5}, + { DBGBUS_DSPP, 68, 6}, + { DBGBUS_DSPP, 68, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 69, 0}, + { DBGBUS_DSPP, 69, 1}, + { DBGBUS_DSPP, 69, 2}, + { DBGBUS_DSPP, 69, 3}, + { DBGBUS_DSPP, 69, 4}, + { DBGBUS_DSPP, 69, 5}, + { DBGBUS_DSPP, 69, 6}, + { DBGBUS_DSPP, 69, 7, _dpu_debug_bus_lm_dump }, + + /* LM1 */ + { DBGBUS_DSPP, 70, 0}, + { DBGBUS_DSPP, 70, 1}, + { DBGBUS_DSPP, 70, 2}, + { DBGBUS_DSPP, 70, 3}, + { DBGBUS_DSPP, 70, 4}, + { DBGBUS_DSPP, 70, 5}, + { DBGBUS_DSPP, 70, 6}, + { DBGBUS_DSPP, 70, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 71, 0}, + { DBGBUS_DSPP, 71, 1}, + { DBGBUS_DSPP, 71, 2}, + { DBGBUS_DSPP, 71, 3}, + { DBGBUS_DSPP, 71, 4}, + { DBGBUS_DSPP, 71, 5}, + { DBGBUS_DSPP, 71, 6}, + { DBGBUS_DSPP, 71, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 72, 0}, + { DBGBUS_DSPP, 72, 1}, + { DBGBUS_DSPP, 72, 2}, + { DBGBUS_DSPP, 72, 3}, + { DBGBUS_DSPP, 72, 4}, + { DBGBUS_DSPP, 72, 5}, + { DBGBUS_DSPP, 72, 6}, + { DBGBUS_DSPP, 72, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 73, 0}, + { DBGBUS_DSPP, 73, 1}, + { DBGBUS_DSPP, 73, 2}, + { DBGBUS_DSPP, 73, 3}, + { DBGBUS_DSPP, 73, 4}, + { DBGBUS_DSPP, 73, 5}, + { DBGBUS_DSPP, 73, 6}, + { DBGBUS_DSPP, 73, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 74, 0}, + { DBGBUS_DSPP, 74, 1}, + { DBGBUS_DSPP, 74, 2}, + { DBGBUS_DSPP, 74, 3}, + { DBGBUS_DSPP, 74, 4}, + { DBGBUS_DSPP, 74, 5}, + { DBGBUS_DSPP, 74, 6}, + { DBGBUS_DSPP, 74, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 75, 0}, + { DBGBUS_DSPP, 75, 1}, + { DBGBUS_DSPP, 75, 2}, + { DBGBUS_DSPP, 75, 3}, + { DBGBUS_DSPP, 75, 4}, + { DBGBUS_DSPP, 75, 5}, + { DBGBUS_DSPP, 75, 6}, + { DBGBUS_DSPP, 75, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 76, 0}, + { DBGBUS_DSPP, 76, 1}, + { DBGBUS_DSPP, 76, 2}, + { DBGBUS_DSPP, 76, 3}, + { DBGBUS_DSPP, 76, 4}, + { DBGBUS_DSPP, 76, 5}, + { DBGBUS_DSPP, 76, 6}, + { DBGBUS_DSPP, 76, 7, _dpu_debug_bus_lm_dump }, + + /* LM2 */ + { DBGBUS_DSPP, 77, 0}, + { DBGBUS_DSPP, 77, 1}, + { DBGBUS_DSPP, 77, 2}, + { DBGBUS_DSPP, 77, 3}, + { DBGBUS_DSPP, 77, 4}, + { DBGBUS_DSPP, 77, 5}, + { DBGBUS_DSPP, 77, 6}, + { DBGBUS_DSPP, 77, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 78, 0}, + { DBGBUS_DSPP, 78, 1}, + { DBGBUS_DSPP, 78, 2}, + { DBGBUS_DSPP, 78, 3}, + { DBGBUS_DSPP, 78, 4}, + { DBGBUS_DSPP, 78, 5}, + { DBGBUS_DSPP, 78, 6}, + { DBGBUS_DSPP, 78, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 79, 0}, + { DBGBUS_DSPP, 79, 1}, + { DBGBUS_DSPP, 79, 2}, + { DBGBUS_DSPP, 79, 3}, + { DBGBUS_DSPP, 79, 4}, + { DBGBUS_DSPP, 79, 5}, + { DBGBUS_DSPP, 79, 6}, + { DBGBUS_DSPP, 79, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 80, 0}, + { DBGBUS_DSPP, 80, 1}, + { DBGBUS_DSPP, 80, 2}, + { DBGBUS_DSPP, 80, 3}, + { DBGBUS_DSPP, 80, 4}, + { DBGBUS_DSPP, 80, 5}, + { DBGBUS_DSPP, 80, 6}, + { DBGBUS_DSPP, 80, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 81, 0}, + { DBGBUS_DSPP, 81, 1}, + { DBGBUS_DSPP, 81, 2}, + { DBGBUS_DSPP, 81, 3}, + { DBGBUS_DSPP, 81, 4}, + { DBGBUS_DSPP, 81, 5}, + { DBGBUS_DSPP, 81, 6}, + { DBGBUS_DSPP, 81, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 82, 0}, + { DBGBUS_DSPP, 82, 1}, + { DBGBUS_DSPP, 82, 2}, + { DBGBUS_DSPP, 82, 3}, + { DBGBUS_DSPP, 82, 4}, + { DBGBUS_DSPP, 82, 5}, + { DBGBUS_DSPP, 82, 6}, + { DBGBUS_DSPP, 82, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 83, 0}, + { DBGBUS_DSPP, 83, 1}, + { DBGBUS_DSPP, 83, 2}, + { DBGBUS_DSPP, 83, 3}, + { DBGBUS_DSPP, 83, 4}, + { DBGBUS_DSPP, 83, 5}, + { DBGBUS_DSPP, 83, 6}, + { DBGBUS_DSPP, 83, 7, _dpu_debug_bus_lm_dump }, + + /* csc */ + { DBGBUS_SSPP0, 7, 0}, + { DBGBUS_SSPP0, 7, 1}, + { DBGBUS_SSPP0, 27, 0}, + { DBGBUS_SSPP0, 27, 1}, + { DBGBUS_SSPP1, 7, 0}, + { DBGBUS_SSPP1, 7, 1}, + { DBGBUS_SSPP1, 27, 0}, + { DBGBUS_SSPP1, 27, 1}, + + /* pcc */ + { DBGBUS_SSPP0, 3, 3}, + { DBGBUS_SSPP0, 23, 3}, + { DBGBUS_SSPP0, 33, 3}, + { DBGBUS_SSPP0, 43, 3}, + { DBGBUS_SSPP1, 3, 3}, + { DBGBUS_SSPP1, 23, 3}, + { DBGBUS_SSPP1, 33, 3}, + { DBGBUS_SSPP1, 43, 3}, + + /* spa */ + { DBGBUS_SSPP0, 8, 0}, + { DBGBUS_SSPP0, 28, 0}, + { DBGBUS_SSPP1, 8, 0}, + { DBGBUS_SSPP1, 28, 0}, + { DBGBUS_DSPP, 13, 0}, + { DBGBUS_DSPP, 19, 0}, + + /* igc */ + { DBGBUS_SSPP0, 9, 0}, + { DBGBUS_SSPP0, 9, 1}, + { DBGBUS_SSPP0, 9, 3}, + { DBGBUS_SSPP0, 29, 0}, + { DBGBUS_SSPP0, 29, 1}, + { DBGBUS_SSPP0, 29, 3}, + { DBGBUS_SSPP0, 17, 0}, + { DBGBUS_SSPP0, 17, 1}, + { DBGBUS_SSPP0, 17, 3}, + { DBGBUS_SSPP0, 37, 0}, + { DBGBUS_SSPP0, 37, 1}, + { DBGBUS_SSPP0, 37, 3}, + { DBGBUS_SSPP0, 46, 0}, + { DBGBUS_SSPP0, 46, 1}, + { DBGBUS_SSPP0, 46, 3}, + + { DBGBUS_SSPP1, 9, 0}, + { DBGBUS_SSPP1, 9, 1}, + { DBGBUS_SSPP1, 9, 3}, + { DBGBUS_SSPP1, 29, 0}, + { DBGBUS_SSPP1, 29, 1}, + { DBGBUS_SSPP1, 29, 3}, + { DBGBUS_SSPP1, 17, 0}, + { DBGBUS_SSPP1, 17, 1}, + { DBGBUS_SSPP1, 17, 3}, + { DBGBUS_SSPP1, 37, 0}, + { DBGBUS_SSPP1, 37, 1}, + { DBGBUS_SSPP1, 37, 3}, + { DBGBUS_SSPP1, 46, 0}, + { DBGBUS_SSPP1, 46, 1}, + { DBGBUS_SSPP1, 46, 3}, + + { DBGBUS_DSPP, 14, 0}, + { DBGBUS_DSPP, 14, 1}, + { DBGBUS_DSPP, 14, 3}, + { DBGBUS_DSPP, 20, 0}, + { DBGBUS_DSPP, 20, 1}, + { DBGBUS_DSPP, 20, 3}, + + { DBGBUS_PERIPH, 60, 0}, +}; + +static struct dpu_debug_bus_entry dbg_bus_dpu_sdm845[] = { + + /* Unpack 0 sspp 0*/ + { DBGBUS_SSPP0, 50, 2 }, + { DBGBUS_SSPP0, 60, 2 }, + { DBGBUS_SSPP0, 70, 2 }, + + /* Upack 0 sspp 1*/ + { DBGBUS_SSPP1, 50, 2 }, + { DBGBUS_SSPP1, 60, 2 }, + { DBGBUS_SSPP1, 70, 2 }, + + /* scheduler */ + { DBGBUS_DSPP, 130, 0 }, + { DBGBUS_DSPP, 130, 1 }, + { DBGBUS_DSPP, 130, 2 }, + { DBGBUS_DSPP, 130, 3 }, + { DBGBUS_DSPP, 130, 4 }, + { DBGBUS_DSPP, 130, 5 }, + + /* qseed */ + { DBGBUS_SSPP0, 6, 0}, + { DBGBUS_SSPP0, 6, 1}, + { DBGBUS_SSPP0, 26, 0}, + { DBGBUS_SSPP0, 26, 1}, + { DBGBUS_SSPP1, 6, 0}, + { DBGBUS_SSPP1, 6, 1}, + { DBGBUS_SSPP1, 26, 0}, + { DBGBUS_SSPP1, 26, 1}, + + /* scale */ + { DBGBUS_SSPP0, 16, 0}, + { DBGBUS_SSPP0, 16, 1}, + { DBGBUS_SSPP0, 36, 0}, + { DBGBUS_SSPP0, 36, 1}, + { DBGBUS_SSPP1, 16, 0}, + { DBGBUS_SSPP1, 16, 1}, + { DBGBUS_SSPP1, 36, 0}, + { DBGBUS_SSPP1, 36, 1}, + + /* fetch sspp0 */ + + /* vig 0 */ + { DBGBUS_SSPP0, 0, 0 }, + { DBGBUS_SSPP0, 0, 1 }, + { DBGBUS_SSPP0, 0, 2 }, + { DBGBUS_SSPP0, 0, 3 }, + { DBGBUS_SSPP0, 0, 4 }, + { DBGBUS_SSPP0, 0, 5 }, + { DBGBUS_SSPP0, 0, 6 }, + { DBGBUS_SSPP0, 0, 7 }, + + { DBGBUS_SSPP0, 1, 0 }, + { DBGBUS_SSPP0, 1, 1 }, + { DBGBUS_SSPP0, 1, 2 }, + { DBGBUS_SSPP0, 1, 3 }, + { DBGBUS_SSPP0, 1, 4 }, + { DBGBUS_SSPP0, 1, 5 }, + { DBGBUS_SSPP0, 1, 6 }, + { DBGBUS_SSPP0, 1, 7 }, + + { DBGBUS_SSPP0, 2, 0 }, + { DBGBUS_SSPP0, 2, 1 }, + { DBGBUS_SSPP0, 2, 2 }, + { DBGBUS_SSPP0, 2, 3 }, + { DBGBUS_SSPP0, 2, 4 }, + { DBGBUS_SSPP0, 2, 5 }, + { DBGBUS_SSPP0, 2, 6 }, + { DBGBUS_SSPP0, 2, 7 }, + + { DBGBUS_SSPP0, 4, 0 }, + { DBGBUS_SSPP0, 4, 1 }, + { DBGBUS_SSPP0, 4, 2 }, + { DBGBUS_SSPP0, 4, 3 }, + { DBGBUS_SSPP0, 4, 4 }, + { DBGBUS_SSPP0, 4, 5 }, + { DBGBUS_SSPP0, 4, 6 }, + { DBGBUS_SSPP0, 4, 7 }, + + { DBGBUS_SSPP0, 5, 0 }, + { DBGBUS_SSPP0, 5, 1 }, + { DBGBUS_SSPP0, 5, 2 }, + { DBGBUS_SSPP0, 5, 3 }, + { DBGBUS_SSPP0, 5, 4 }, + { DBGBUS_SSPP0, 5, 5 }, + { DBGBUS_SSPP0, 5, 6 }, + { DBGBUS_SSPP0, 5, 7 }, + + /* vig 2 */ + { DBGBUS_SSPP0, 20, 0 }, + { DBGBUS_SSPP0, 20, 1 }, + { DBGBUS_SSPP0, 20, 2 }, + { DBGBUS_SSPP0, 20, 3 }, + { DBGBUS_SSPP0, 20, 4 }, + { DBGBUS_SSPP0, 20, 5 }, + { DBGBUS_SSPP0, 20, 6 }, + { DBGBUS_SSPP0, 20, 7 }, + + { DBGBUS_SSPP0, 21, 0 }, + { DBGBUS_SSPP0, 21, 1 }, + { DBGBUS_SSPP0, 21, 2 }, + { DBGBUS_SSPP0, 21, 3 }, + { DBGBUS_SSPP0, 21, 4 }, + { DBGBUS_SSPP0, 21, 5 }, + { DBGBUS_SSPP0, 21, 6 }, + { DBGBUS_SSPP0, 21, 7 }, + + { DBGBUS_SSPP0, 22, 0 }, + { DBGBUS_SSPP0, 22, 1 }, + { DBGBUS_SSPP0, 22, 2 }, + { DBGBUS_SSPP0, 22, 3 }, + { DBGBUS_SSPP0, 22, 4 }, + { DBGBUS_SSPP0, 22, 5 }, + { DBGBUS_SSPP0, 22, 6 }, + { DBGBUS_SSPP0, 22, 7 }, + + { DBGBUS_SSPP0, 24, 0 }, + { DBGBUS_SSPP0, 24, 1 }, + { DBGBUS_SSPP0, 24, 2 }, + { DBGBUS_SSPP0, 24, 3 }, + { DBGBUS_SSPP0, 24, 4 }, + { DBGBUS_SSPP0, 24, 5 }, + { DBGBUS_SSPP0, 24, 6 }, + { DBGBUS_SSPP0, 24, 7 }, + + { DBGBUS_SSPP0, 25, 0 }, + { DBGBUS_SSPP0, 25, 1 }, + { DBGBUS_SSPP0, 25, 2 }, + { DBGBUS_SSPP0, 25, 3 }, + { DBGBUS_SSPP0, 25, 4 }, + { DBGBUS_SSPP0, 25, 5 }, + { DBGBUS_SSPP0, 25, 6 }, + { DBGBUS_SSPP0, 25, 7 }, + + /* dma 2 */ + { DBGBUS_SSPP0, 30, 0 }, + { DBGBUS_SSPP0, 30, 1 }, + { DBGBUS_SSPP0, 30, 2 }, + { DBGBUS_SSPP0, 30, 3 }, + { DBGBUS_SSPP0, 30, 4 }, + { DBGBUS_SSPP0, 30, 5 }, + { DBGBUS_SSPP0, 30, 6 }, + { DBGBUS_SSPP0, 30, 7 }, + + { DBGBUS_SSPP0, 31, 0 }, + { DBGBUS_SSPP0, 31, 1 }, + { DBGBUS_SSPP0, 31, 2 }, + { DBGBUS_SSPP0, 31, 3 }, + { DBGBUS_SSPP0, 31, 4 }, + { DBGBUS_SSPP0, 31, 5 }, + { DBGBUS_SSPP0, 31, 6 }, + { DBGBUS_SSPP0, 31, 7 }, + + { DBGBUS_SSPP0, 32, 0 }, + { DBGBUS_SSPP0, 32, 1 }, + { DBGBUS_SSPP0, 32, 2 }, + { DBGBUS_SSPP0, 32, 3 }, + { DBGBUS_SSPP0, 32, 4 }, + { DBGBUS_SSPP0, 32, 5 }, + { DBGBUS_SSPP0, 32, 6 }, + { DBGBUS_SSPP0, 32, 7 }, + + { DBGBUS_SSPP0, 33, 0 }, + { DBGBUS_SSPP0, 33, 1 }, + { DBGBUS_SSPP0, 33, 2 }, + { DBGBUS_SSPP0, 33, 3 }, + { DBGBUS_SSPP0, 33, 4 }, + { DBGBUS_SSPP0, 33, 5 }, + { DBGBUS_SSPP0, 33, 6 }, + { DBGBUS_SSPP0, 33, 7 }, + + { DBGBUS_SSPP0, 34, 0 }, + { DBGBUS_SSPP0, 34, 1 }, + { DBGBUS_SSPP0, 34, 2 }, + { DBGBUS_SSPP0, 34, 3 }, + { DBGBUS_SSPP0, 34, 4 }, + { DBGBUS_SSPP0, 34, 5 }, + { DBGBUS_SSPP0, 34, 6 }, + { DBGBUS_SSPP0, 34, 7 }, + + { DBGBUS_SSPP0, 35, 0 }, + { DBGBUS_SSPP0, 35, 1 }, + { DBGBUS_SSPP0, 35, 2 }, + { DBGBUS_SSPP0, 35, 3 }, + + /* dma 0 */ + { DBGBUS_SSPP0, 40, 0 }, + { DBGBUS_SSPP0, 40, 1 }, + { DBGBUS_SSPP0, 40, 2 }, + { DBGBUS_SSPP0, 40, 3 }, + { DBGBUS_SSPP0, 40, 4 }, + { DBGBUS_SSPP0, 40, 5 }, + { DBGBUS_SSPP0, 40, 6 }, + { DBGBUS_SSPP0, 40, 7 }, + + { DBGBUS_SSPP0, 41, 0 }, + { DBGBUS_SSPP0, 41, 1 }, + { DBGBUS_SSPP0, 41, 2 }, + { DBGBUS_SSPP0, 41, 3 }, + { DBGBUS_SSPP0, 41, 4 }, + { DBGBUS_SSPP0, 41, 5 }, + { DBGBUS_SSPP0, 41, 6 }, + { DBGBUS_SSPP0, 41, 7 }, + + { DBGBUS_SSPP0, 42, 0 }, + { DBGBUS_SSPP0, 42, 1 }, + { DBGBUS_SSPP0, 42, 2 }, + { DBGBUS_SSPP0, 42, 3 }, + { DBGBUS_SSPP0, 42, 4 }, + { DBGBUS_SSPP0, 42, 5 }, + { DBGBUS_SSPP0, 42, 6 }, + { DBGBUS_SSPP0, 42, 7 }, + + { DBGBUS_SSPP0, 44, 0 }, + { DBGBUS_SSPP0, 44, 1 }, + { DBGBUS_SSPP0, 44, 2 }, + { DBGBUS_SSPP0, 44, 3 }, + { DBGBUS_SSPP0, 44, 4 }, + { DBGBUS_SSPP0, 44, 5 }, + { DBGBUS_SSPP0, 44, 6 }, + { DBGBUS_SSPP0, 44, 7 }, + + { DBGBUS_SSPP0, 45, 0 }, + { DBGBUS_SSPP0, 45, 1 }, + { DBGBUS_SSPP0, 45, 2 }, + { DBGBUS_SSPP0, 45, 3 }, + { DBGBUS_SSPP0, 45, 4 }, + { DBGBUS_SSPP0, 45, 5 }, + { DBGBUS_SSPP0, 45, 6 }, + { DBGBUS_SSPP0, 45, 7 }, + + /* fetch sspp1 */ + /* vig 1 */ + { DBGBUS_SSPP1, 0, 0 }, + { DBGBUS_SSPP1, 0, 1 }, + { DBGBUS_SSPP1, 0, 2 }, + { DBGBUS_SSPP1, 0, 3 }, + { DBGBUS_SSPP1, 0, 4 }, + { DBGBUS_SSPP1, 0, 5 }, + { DBGBUS_SSPP1, 0, 6 }, + { DBGBUS_SSPP1, 0, 7 }, + + { DBGBUS_SSPP1, 1, 0 }, + { DBGBUS_SSPP1, 1, 1 }, + { DBGBUS_SSPP1, 1, 2 }, + { DBGBUS_SSPP1, 1, 3 }, + { DBGBUS_SSPP1, 1, 4 }, + { DBGBUS_SSPP1, 1, 5 }, + { DBGBUS_SSPP1, 1, 6 }, + { DBGBUS_SSPP1, 1, 7 }, + + { DBGBUS_SSPP1, 2, 0 }, + { DBGBUS_SSPP1, 2, 1 }, + { DBGBUS_SSPP1, 2, 2 }, + { DBGBUS_SSPP1, 2, 3 }, + { DBGBUS_SSPP1, 2, 4 }, + { DBGBUS_SSPP1, 2, 5 }, + { DBGBUS_SSPP1, 2, 6 }, + { DBGBUS_SSPP1, 2, 7 }, + + { DBGBUS_SSPP1, 4, 0 }, + { DBGBUS_SSPP1, 4, 1 }, + { DBGBUS_SSPP1, 4, 2 }, + { DBGBUS_SSPP1, 4, 3 }, + { DBGBUS_SSPP1, 4, 4 }, + { DBGBUS_SSPP1, 4, 5 }, + { DBGBUS_SSPP1, 4, 6 }, + { DBGBUS_SSPP1, 4, 7 }, + + { DBGBUS_SSPP1, 5, 0 }, + { DBGBUS_SSPP1, 5, 1 }, + { DBGBUS_SSPP1, 5, 2 }, + { DBGBUS_SSPP1, 5, 3 }, + { DBGBUS_SSPP1, 5, 4 }, + { DBGBUS_SSPP1, 5, 5 }, + { DBGBUS_SSPP1, 5, 6 }, + { DBGBUS_SSPP1, 5, 7 }, + + /* vig 3 */ + { DBGBUS_SSPP1, 20, 0 }, + { DBGBUS_SSPP1, 20, 1 }, + { DBGBUS_SSPP1, 20, 2 }, + { DBGBUS_SSPP1, 20, 3 }, + { DBGBUS_SSPP1, 20, 4 }, + { DBGBUS_SSPP1, 20, 5 }, + { DBGBUS_SSPP1, 20, 6 }, + { DBGBUS_SSPP1, 20, 7 }, + + { DBGBUS_SSPP1, 21, 0 }, + { DBGBUS_SSPP1, 21, 1 }, + { DBGBUS_SSPP1, 21, 2 }, + { DBGBUS_SSPP1, 21, 3 }, + { DBGBUS_SSPP1, 21, 4 }, + { DBGBUS_SSPP1, 21, 5 }, + { DBGBUS_SSPP1, 21, 6 }, + { DBGBUS_SSPP1, 21, 7 }, + + { DBGBUS_SSPP1, 22, 0 }, + { DBGBUS_SSPP1, 22, 1 }, + { DBGBUS_SSPP1, 22, 2 }, + { DBGBUS_SSPP1, 22, 3 }, + { DBGBUS_SSPP1, 22, 4 }, + { DBGBUS_SSPP1, 22, 5 }, + { DBGBUS_SSPP1, 22, 6 }, + { DBGBUS_SSPP1, 22, 7 }, + + { DBGBUS_SSPP1, 24, 0 }, + { DBGBUS_SSPP1, 24, 1 }, + { DBGBUS_SSPP1, 24, 2 }, + { DBGBUS_SSPP1, 24, 3 }, + { DBGBUS_SSPP1, 24, 4 }, + { DBGBUS_SSPP1, 24, 5 }, + { DBGBUS_SSPP1, 24, 6 }, + { DBGBUS_SSPP1, 24, 7 }, + + { DBGBUS_SSPP1, 25, 0 }, + { DBGBUS_SSPP1, 25, 1 }, + { DBGBUS_SSPP1, 25, 2 }, + { DBGBUS_SSPP1, 25, 3 }, + { DBGBUS_SSPP1, 25, 4 }, + { DBGBUS_SSPP1, 25, 5 }, + { DBGBUS_SSPP1, 25, 6 }, + { DBGBUS_SSPP1, 25, 7 }, + + /* dma 3 */ + { DBGBUS_SSPP1, 30, 0 }, + { DBGBUS_SSPP1, 30, 1 }, + { DBGBUS_SSPP1, 30, 2 }, + { DBGBUS_SSPP1, 30, 3 }, + { DBGBUS_SSPP1, 30, 4 }, + { DBGBUS_SSPP1, 30, 5 }, + { DBGBUS_SSPP1, 30, 6 }, + { DBGBUS_SSPP1, 30, 7 }, + + { DBGBUS_SSPP1, 31, 0 }, + { DBGBUS_SSPP1, 31, 1 }, + { DBGBUS_SSPP1, 31, 2 }, + { DBGBUS_SSPP1, 31, 3 }, + { DBGBUS_SSPP1, 31, 4 }, + { DBGBUS_SSPP1, 31, 5 }, + { DBGBUS_SSPP1, 31, 6 }, + { DBGBUS_SSPP1, 31, 7 }, + + { DBGBUS_SSPP1, 32, 0 }, + { DBGBUS_SSPP1, 32, 1 }, + { DBGBUS_SSPP1, 32, 2 }, + { DBGBUS_SSPP1, 32, 3 }, + { DBGBUS_SSPP1, 32, 4 }, + { DBGBUS_SSPP1, 32, 5 }, + { DBGBUS_SSPP1, 32, 6 }, + { DBGBUS_SSPP1, 32, 7 }, + + { DBGBUS_SSPP1, 33, 0 }, + { DBGBUS_SSPP1, 33, 1 }, + { DBGBUS_SSPP1, 33, 2 }, + { DBGBUS_SSPP1, 33, 3 }, + { DBGBUS_SSPP1, 33, 4 }, + { DBGBUS_SSPP1, 33, 5 }, + { DBGBUS_SSPP1, 33, 6 }, + { DBGBUS_SSPP1, 33, 7 }, + + { DBGBUS_SSPP1, 34, 0 }, + { DBGBUS_SSPP1, 34, 1 }, + { DBGBUS_SSPP1, 34, 2 }, + { DBGBUS_SSPP1, 34, 3 }, + { DBGBUS_SSPP1, 34, 4 }, + { DBGBUS_SSPP1, 34, 5 }, + { DBGBUS_SSPP1, 34, 6 }, + { DBGBUS_SSPP1, 34, 7 }, + + { DBGBUS_SSPP1, 35, 0 }, + { DBGBUS_SSPP1, 35, 1 }, + { DBGBUS_SSPP1, 35, 2 }, + + /* dma 1 */ + { DBGBUS_SSPP1, 40, 0 }, + { DBGBUS_SSPP1, 40, 1 }, + { DBGBUS_SSPP1, 40, 2 }, + { DBGBUS_SSPP1, 40, 3 }, + { DBGBUS_SSPP1, 40, 4 }, + { DBGBUS_SSPP1, 40, 5 }, + { DBGBUS_SSPP1, 40, 6 }, + { DBGBUS_SSPP1, 40, 7 }, + + { DBGBUS_SSPP1, 41, 0 }, + { DBGBUS_SSPP1, 41, 1 }, + { DBGBUS_SSPP1, 41, 2 }, + { DBGBUS_SSPP1, 41, 3 }, + { DBGBUS_SSPP1, 41, 4 }, + { DBGBUS_SSPP1, 41, 5 }, + { DBGBUS_SSPP1, 41, 6 }, + { DBGBUS_SSPP1, 41, 7 }, + + { DBGBUS_SSPP1, 42, 0 }, + { DBGBUS_SSPP1, 42, 1 }, + { DBGBUS_SSPP1, 42, 2 }, + { DBGBUS_SSPP1, 42, 3 }, + { DBGBUS_SSPP1, 42, 4 }, + { DBGBUS_SSPP1, 42, 5 }, + { DBGBUS_SSPP1, 42, 6 }, + { DBGBUS_SSPP1, 42, 7 }, + + { DBGBUS_SSPP1, 44, 0 }, + { DBGBUS_SSPP1, 44, 1 }, + { DBGBUS_SSPP1, 44, 2 }, + { DBGBUS_SSPP1, 44, 3 }, + { DBGBUS_SSPP1, 44, 4 }, + { DBGBUS_SSPP1, 44, 5 }, + { DBGBUS_SSPP1, 44, 6 }, + { DBGBUS_SSPP1, 44, 7 }, + + { DBGBUS_SSPP1, 45, 0 }, + { DBGBUS_SSPP1, 45, 1 }, + { DBGBUS_SSPP1, 45, 2 }, + { DBGBUS_SSPP1, 45, 3 }, + { DBGBUS_SSPP1, 45, 4 }, + { DBGBUS_SSPP1, 45, 5 }, + { DBGBUS_SSPP1, 45, 6 }, + { DBGBUS_SSPP1, 45, 7 }, + + /* dspp */ + { DBGBUS_DSPP, 13, 0 }, + { DBGBUS_DSPP, 19, 0 }, + { DBGBUS_DSPP, 14, 0 }, + { DBGBUS_DSPP, 14, 1 }, + { DBGBUS_DSPP, 14, 3 }, + { DBGBUS_DSPP, 20, 0 }, + { DBGBUS_DSPP, 20, 1 }, + { DBGBUS_DSPP, 20, 3 }, + + /* ppb_0 */ + { DBGBUS_DSPP, 31, 0, _dpu_debug_bus_ppb0_dump }, + { DBGBUS_DSPP, 33, 0, _dpu_debug_bus_ppb0_dump }, + { DBGBUS_DSPP, 35, 0, _dpu_debug_bus_ppb0_dump }, + { DBGBUS_DSPP, 42, 0, _dpu_debug_bus_ppb0_dump }, + + /* ppb_1 */ + { DBGBUS_DSPP, 32, 0, _dpu_debug_bus_ppb1_dump }, + { DBGBUS_DSPP, 34, 0, _dpu_debug_bus_ppb1_dump }, + { DBGBUS_DSPP, 36, 0, _dpu_debug_bus_ppb1_dump }, + { DBGBUS_DSPP, 43, 0, _dpu_debug_bus_ppb1_dump }, + + /* lm_lut */ + { DBGBUS_DSPP, 109, 0 }, + { DBGBUS_DSPP, 105, 0 }, + { DBGBUS_DSPP, 103, 0 }, + + /* crossbar */ + { DBGBUS_DSPP, 0, 0, _dpu_debug_bus_xbar_dump }, + + /* rotator */ + { DBGBUS_DSPP, 9, 0}, + + /* blend */ + /* LM0 */ + { DBGBUS_DSPP, 63, 1}, + { DBGBUS_DSPP, 63, 2}, + { DBGBUS_DSPP, 63, 3}, + { DBGBUS_DSPP, 63, 4}, + { DBGBUS_DSPP, 63, 5}, + { DBGBUS_DSPP, 63, 6}, + { DBGBUS_DSPP, 63, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 64, 1}, + { DBGBUS_DSPP, 64, 2}, + { DBGBUS_DSPP, 64, 3}, + { DBGBUS_DSPP, 64, 4}, + { DBGBUS_DSPP, 64, 5}, + { DBGBUS_DSPP, 64, 6}, + { DBGBUS_DSPP, 64, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 65, 1}, + { DBGBUS_DSPP, 65, 2}, + { DBGBUS_DSPP, 65, 3}, + { DBGBUS_DSPP, 65, 4}, + { DBGBUS_DSPP, 65, 5}, + { DBGBUS_DSPP, 65, 6}, + { DBGBUS_DSPP, 65, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 66, 1}, + { DBGBUS_DSPP, 66, 2}, + { DBGBUS_DSPP, 66, 3}, + { DBGBUS_DSPP, 66, 4}, + { DBGBUS_DSPP, 66, 5}, + { DBGBUS_DSPP, 66, 6}, + { DBGBUS_DSPP, 66, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 67, 1}, + { DBGBUS_DSPP, 67, 2}, + { DBGBUS_DSPP, 67, 3}, + { DBGBUS_DSPP, 67, 4}, + { DBGBUS_DSPP, 67, 5}, + { DBGBUS_DSPP, 67, 6}, + { DBGBUS_DSPP, 67, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 68, 1}, + { DBGBUS_DSPP, 68, 2}, + { DBGBUS_DSPP, 68, 3}, + { DBGBUS_DSPP, 68, 4}, + { DBGBUS_DSPP, 68, 5}, + { DBGBUS_DSPP, 68, 6}, + { DBGBUS_DSPP, 68, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 69, 1}, + { DBGBUS_DSPP, 69, 2}, + { DBGBUS_DSPP, 69, 3}, + { DBGBUS_DSPP, 69, 4}, + { DBGBUS_DSPP, 69, 5}, + { DBGBUS_DSPP, 69, 6}, + { DBGBUS_DSPP, 69, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 84, 1}, + { DBGBUS_DSPP, 84, 2}, + { DBGBUS_DSPP, 84, 3}, + { DBGBUS_DSPP, 84, 4}, + { DBGBUS_DSPP, 84, 5}, + { DBGBUS_DSPP, 84, 6}, + { DBGBUS_DSPP, 84, 7, _dpu_debug_bus_lm_dump }, + + + { DBGBUS_DSPP, 85, 1}, + { DBGBUS_DSPP, 85, 2}, + { DBGBUS_DSPP, 85, 3}, + { DBGBUS_DSPP, 85, 4}, + { DBGBUS_DSPP, 85, 5}, + { DBGBUS_DSPP, 85, 6}, + { DBGBUS_DSPP, 85, 7, _dpu_debug_bus_lm_dump }, + + + { DBGBUS_DSPP, 86, 1}, + { DBGBUS_DSPP, 86, 2}, + { DBGBUS_DSPP, 86, 3}, + { DBGBUS_DSPP, 86, 4}, + { DBGBUS_DSPP, 86, 5}, + { DBGBUS_DSPP, 86, 6}, + { DBGBUS_DSPP, 86, 7, _dpu_debug_bus_lm_dump }, + + + { DBGBUS_DSPP, 87, 1}, + { DBGBUS_DSPP, 87, 2}, + { DBGBUS_DSPP, 87, 3}, + { DBGBUS_DSPP, 87, 4}, + { DBGBUS_DSPP, 87, 5}, + { DBGBUS_DSPP, 87, 6}, + { DBGBUS_DSPP, 87, 7, _dpu_debug_bus_lm_dump }, + + /* LM1 */ + { DBGBUS_DSPP, 70, 1}, + { DBGBUS_DSPP, 70, 2}, + { DBGBUS_DSPP, 70, 3}, + { DBGBUS_DSPP, 70, 4}, + { DBGBUS_DSPP, 70, 5}, + { DBGBUS_DSPP, 70, 6}, + { DBGBUS_DSPP, 70, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 71, 1}, + { DBGBUS_DSPP, 71, 2}, + { DBGBUS_DSPP, 71, 3}, + { DBGBUS_DSPP, 71, 4}, + { DBGBUS_DSPP, 71, 5}, + { DBGBUS_DSPP, 71, 6}, + { DBGBUS_DSPP, 71, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 72, 1}, + { DBGBUS_DSPP, 72, 2}, + { DBGBUS_DSPP, 72, 3}, + { DBGBUS_DSPP, 72, 4}, + { DBGBUS_DSPP, 72, 5}, + { DBGBUS_DSPP, 72, 6}, + { DBGBUS_DSPP, 72, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 73, 1}, + { DBGBUS_DSPP, 73, 2}, + { DBGBUS_DSPP, 73, 3}, + { DBGBUS_DSPP, 73, 4}, + { DBGBUS_DSPP, 73, 5}, + { DBGBUS_DSPP, 73, 6}, + { DBGBUS_DSPP, 73, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 74, 1}, + { DBGBUS_DSPP, 74, 2}, + { DBGBUS_DSPP, 74, 3}, + { DBGBUS_DSPP, 74, 4}, + { DBGBUS_DSPP, 74, 5}, + { DBGBUS_DSPP, 74, 6}, + { DBGBUS_DSPP, 74, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 75, 1}, + { DBGBUS_DSPP, 75, 2}, + { DBGBUS_DSPP, 75, 3}, + { DBGBUS_DSPP, 75, 4}, + { DBGBUS_DSPP, 75, 5}, + { DBGBUS_DSPP, 75, 6}, + { DBGBUS_DSPP, 75, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 76, 1}, + { DBGBUS_DSPP, 76, 2}, + { DBGBUS_DSPP, 76, 3}, + { DBGBUS_DSPP, 76, 4}, + { DBGBUS_DSPP, 76, 5}, + { DBGBUS_DSPP, 76, 6}, + { DBGBUS_DSPP, 76, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 88, 1}, + { DBGBUS_DSPP, 88, 2}, + { DBGBUS_DSPP, 88, 3}, + { DBGBUS_DSPP, 88, 4}, + { DBGBUS_DSPP, 88, 5}, + { DBGBUS_DSPP, 88, 6}, + { DBGBUS_DSPP, 88, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 89, 1}, + { DBGBUS_DSPP, 89, 2}, + { DBGBUS_DSPP, 89, 3}, + { DBGBUS_DSPP, 89, 4}, + { DBGBUS_DSPP, 89, 5}, + { DBGBUS_DSPP, 89, 6}, + { DBGBUS_DSPP, 89, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 90, 1}, + { DBGBUS_DSPP, 90, 2}, + { DBGBUS_DSPP, 90, 3}, + { DBGBUS_DSPP, 90, 4}, + { DBGBUS_DSPP, 90, 5}, + { DBGBUS_DSPP, 90, 6}, + { DBGBUS_DSPP, 90, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 91, 1}, + { DBGBUS_DSPP, 91, 2}, + { DBGBUS_DSPP, 91, 3}, + { DBGBUS_DSPP, 91, 4}, + { DBGBUS_DSPP, 91, 5}, + { DBGBUS_DSPP, 91, 6}, + { DBGBUS_DSPP, 91, 7, _dpu_debug_bus_lm_dump }, + + /* LM2 */ + { DBGBUS_DSPP, 77, 0}, + { DBGBUS_DSPP, 77, 1}, + { DBGBUS_DSPP, 77, 2}, + { DBGBUS_DSPP, 77, 3}, + { DBGBUS_DSPP, 77, 4}, + { DBGBUS_DSPP, 77, 5}, + { DBGBUS_DSPP, 77, 6}, + { DBGBUS_DSPP, 77, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 78, 0}, + { DBGBUS_DSPP, 78, 1}, + { DBGBUS_DSPP, 78, 2}, + { DBGBUS_DSPP, 78, 3}, + { DBGBUS_DSPP, 78, 4}, + { DBGBUS_DSPP, 78, 5}, + { DBGBUS_DSPP, 78, 6}, + { DBGBUS_DSPP, 78, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 79, 0}, + { DBGBUS_DSPP, 79, 1}, + { DBGBUS_DSPP, 79, 2}, + { DBGBUS_DSPP, 79, 3}, + { DBGBUS_DSPP, 79, 4}, + { DBGBUS_DSPP, 79, 5}, + { DBGBUS_DSPP, 79, 6}, + { DBGBUS_DSPP, 79, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 80, 0}, + { DBGBUS_DSPP, 80, 1}, + { DBGBUS_DSPP, 80, 2}, + { DBGBUS_DSPP, 80, 3}, + { DBGBUS_DSPP, 80, 4}, + { DBGBUS_DSPP, 80, 5}, + { DBGBUS_DSPP, 80, 6}, + { DBGBUS_DSPP, 80, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 81, 0}, + { DBGBUS_DSPP, 81, 1}, + { DBGBUS_DSPP, 81, 2}, + { DBGBUS_DSPP, 81, 3}, + { DBGBUS_DSPP, 81, 4}, + { DBGBUS_DSPP, 81, 5}, + { DBGBUS_DSPP, 81, 6}, + { DBGBUS_DSPP, 81, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 82, 0}, + { DBGBUS_DSPP, 82, 1}, + { DBGBUS_DSPP, 82, 2}, + { DBGBUS_DSPP, 82, 3}, + { DBGBUS_DSPP, 82, 4}, + { DBGBUS_DSPP, 82, 5}, + { DBGBUS_DSPP, 82, 6}, + { DBGBUS_DSPP, 82, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 83, 0}, + { DBGBUS_DSPP, 83, 1}, + { DBGBUS_DSPP, 83, 2}, + { DBGBUS_DSPP, 83, 3}, + { DBGBUS_DSPP, 83, 4}, + { DBGBUS_DSPP, 83, 5}, + { DBGBUS_DSPP, 83, 6}, + { DBGBUS_DSPP, 83, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 92, 1}, + { DBGBUS_DSPP, 92, 2}, + { DBGBUS_DSPP, 92, 3}, + { DBGBUS_DSPP, 92, 4}, + { DBGBUS_DSPP, 92, 5}, + { DBGBUS_DSPP, 92, 6}, + { DBGBUS_DSPP, 92, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 93, 1}, + { DBGBUS_DSPP, 93, 2}, + { DBGBUS_DSPP, 93, 3}, + { DBGBUS_DSPP, 93, 4}, + { DBGBUS_DSPP, 93, 5}, + { DBGBUS_DSPP, 93, 6}, + { DBGBUS_DSPP, 93, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 94, 1}, + { DBGBUS_DSPP, 94, 2}, + { DBGBUS_DSPP, 94, 3}, + { DBGBUS_DSPP, 94, 4}, + { DBGBUS_DSPP, 94, 5}, + { DBGBUS_DSPP, 94, 6}, + { DBGBUS_DSPP, 94, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 95, 1}, + { DBGBUS_DSPP, 95, 2}, + { DBGBUS_DSPP, 95, 3}, + { DBGBUS_DSPP, 95, 4}, + { DBGBUS_DSPP, 95, 5}, + { DBGBUS_DSPP, 95, 6}, + { DBGBUS_DSPP, 95, 7, _dpu_debug_bus_lm_dump }, + + /* LM5 */ + { DBGBUS_DSPP, 110, 1}, + { DBGBUS_DSPP, 110, 2}, + { DBGBUS_DSPP, 110, 3}, + { DBGBUS_DSPP, 110, 4}, + { DBGBUS_DSPP, 110, 5}, + { DBGBUS_DSPP, 110, 6}, + { DBGBUS_DSPP, 110, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 111, 1}, + { DBGBUS_DSPP, 111, 2}, + { DBGBUS_DSPP, 111, 3}, + { DBGBUS_DSPP, 111, 4}, + { DBGBUS_DSPP, 111, 5}, + { DBGBUS_DSPP, 111, 6}, + { DBGBUS_DSPP, 111, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 112, 1}, + { DBGBUS_DSPP, 112, 2}, + { DBGBUS_DSPP, 112, 3}, + { DBGBUS_DSPP, 112, 4}, + { DBGBUS_DSPP, 112, 5}, + { DBGBUS_DSPP, 112, 6}, + { DBGBUS_DSPP, 112, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 113, 1}, + { DBGBUS_DSPP, 113, 2}, + { DBGBUS_DSPP, 113, 3}, + { DBGBUS_DSPP, 113, 4}, + { DBGBUS_DSPP, 113, 5}, + { DBGBUS_DSPP, 113, 6}, + { DBGBUS_DSPP, 113, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 114, 1}, + { DBGBUS_DSPP, 114, 2}, + { DBGBUS_DSPP, 114, 3}, + { DBGBUS_DSPP, 114, 4}, + { DBGBUS_DSPP, 114, 5}, + { DBGBUS_DSPP, 114, 6}, + { DBGBUS_DSPP, 114, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 115, 1}, + { DBGBUS_DSPP, 115, 2}, + { DBGBUS_DSPP, 115, 3}, + { DBGBUS_DSPP, 115, 4}, + { DBGBUS_DSPP, 115, 5}, + { DBGBUS_DSPP, 115, 6}, + { DBGBUS_DSPP, 115, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 116, 1}, + { DBGBUS_DSPP, 116, 2}, + { DBGBUS_DSPP, 116, 3}, + { DBGBUS_DSPP, 116, 4}, + { DBGBUS_DSPP, 116, 5}, + { DBGBUS_DSPP, 116, 6}, + { DBGBUS_DSPP, 116, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 117, 1}, + { DBGBUS_DSPP, 117, 2}, + { DBGBUS_DSPP, 117, 3}, + { DBGBUS_DSPP, 117, 4}, + { DBGBUS_DSPP, 117, 5}, + { DBGBUS_DSPP, 117, 6}, + { DBGBUS_DSPP, 117, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 118, 1}, + { DBGBUS_DSPP, 118, 2}, + { DBGBUS_DSPP, 118, 3}, + { DBGBUS_DSPP, 118, 4}, + { DBGBUS_DSPP, 118, 5}, + { DBGBUS_DSPP, 118, 6}, + { DBGBUS_DSPP, 118, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 119, 1}, + { DBGBUS_DSPP, 119, 2}, + { DBGBUS_DSPP, 119, 3}, + { DBGBUS_DSPP, 119, 4}, + { DBGBUS_DSPP, 119, 5}, + { DBGBUS_DSPP, 119, 6}, + { DBGBUS_DSPP, 119, 7, _dpu_debug_bus_lm_dump }, + + { DBGBUS_DSPP, 120, 1}, + { DBGBUS_DSPP, 120, 2}, + { DBGBUS_DSPP, 120, 3}, + { DBGBUS_DSPP, 120, 4}, + { DBGBUS_DSPP, 120, 5}, + { DBGBUS_DSPP, 120, 6}, + { DBGBUS_DSPP, 120, 7, _dpu_debug_bus_lm_dump }, + + /* csc */ + { DBGBUS_SSPP0, 7, 0}, + { DBGBUS_SSPP0, 7, 1}, + { DBGBUS_SSPP0, 27, 0}, + { DBGBUS_SSPP0, 27, 1}, + { DBGBUS_SSPP1, 7, 0}, + { DBGBUS_SSPP1, 7, 1}, + { DBGBUS_SSPP1, 27, 0}, + { DBGBUS_SSPP1, 27, 1}, + + /* pcc */ + { DBGBUS_SSPP0, 3, 3}, + { DBGBUS_SSPP0, 23, 3}, + { DBGBUS_SSPP0, 33, 3}, + { DBGBUS_SSPP0, 43, 3}, + { DBGBUS_SSPP1, 3, 3}, + { DBGBUS_SSPP1, 23, 3}, + { DBGBUS_SSPP1, 33, 3}, + { DBGBUS_SSPP1, 43, 3}, + + /* spa */ + { DBGBUS_SSPP0, 8, 0}, + { DBGBUS_SSPP0, 28, 0}, + { DBGBUS_SSPP1, 8, 0}, + { DBGBUS_SSPP1, 28, 0}, + { DBGBUS_DSPP, 13, 0}, + { DBGBUS_DSPP, 19, 0}, + + /* igc */ + { DBGBUS_SSPP0, 17, 0}, + { DBGBUS_SSPP0, 17, 1}, + { DBGBUS_SSPP0, 17, 3}, + { DBGBUS_SSPP0, 37, 0}, + { DBGBUS_SSPP0, 37, 1}, + { DBGBUS_SSPP0, 37, 3}, + { DBGBUS_SSPP0, 46, 0}, + { DBGBUS_SSPP0, 46, 1}, + { DBGBUS_SSPP0, 46, 3}, + + { DBGBUS_SSPP1, 17, 0}, + { DBGBUS_SSPP1, 17, 1}, + { DBGBUS_SSPP1, 17, 3}, + { DBGBUS_SSPP1, 37, 0}, + { DBGBUS_SSPP1, 37, 1}, + { DBGBUS_SSPP1, 37, 3}, + { DBGBUS_SSPP1, 46, 0}, + { DBGBUS_SSPP1, 46, 1}, + { DBGBUS_SSPP1, 46, 3}, + + { DBGBUS_DSPP, 14, 0}, + { DBGBUS_DSPP, 14, 1}, + { DBGBUS_DSPP, 14, 3}, + { DBGBUS_DSPP, 20, 0}, + { DBGBUS_DSPP, 20, 1}, + { DBGBUS_DSPP, 20, 3}, + + /* intf0-3 */ + { DBGBUS_PERIPH, 0, 0}, + { DBGBUS_PERIPH, 1, 0}, + { DBGBUS_PERIPH, 2, 0}, + { DBGBUS_PERIPH, 3, 0}, + + /* te counter wrapper */ + { DBGBUS_PERIPH, 60, 0}, + + /* dsc0 */ + { DBGBUS_PERIPH, 47, 0}, + { DBGBUS_PERIPH, 47, 1}, + { DBGBUS_PERIPH, 47, 2}, + { DBGBUS_PERIPH, 47, 3}, + { DBGBUS_PERIPH, 47, 4}, + { DBGBUS_PERIPH, 47, 5}, + { DBGBUS_PERIPH, 47, 6}, + { DBGBUS_PERIPH, 47, 7}, + + /* dsc1 */ + { DBGBUS_PERIPH, 48, 0}, + { DBGBUS_PERIPH, 48, 1}, + { DBGBUS_PERIPH, 48, 2}, + { DBGBUS_PERIPH, 48, 3}, + { DBGBUS_PERIPH, 48, 4}, + { DBGBUS_PERIPH, 48, 5}, + { DBGBUS_PERIPH, 48, 6}, + { DBGBUS_PERIPH, 48, 7}, + + /* dsc2 */ + { DBGBUS_PERIPH, 51, 0}, + { DBGBUS_PERIPH, 51, 1}, + { DBGBUS_PERIPH, 51, 2}, + { DBGBUS_PERIPH, 51, 3}, + { DBGBUS_PERIPH, 51, 4}, + { DBGBUS_PERIPH, 51, 5}, + { DBGBUS_PERIPH, 51, 6}, + { DBGBUS_PERIPH, 51, 7}, + + /* dsc3 */ + { DBGBUS_PERIPH, 52, 0}, + { DBGBUS_PERIPH, 52, 1}, + { DBGBUS_PERIPH, 52, 2}, + { DBGBUS_PERIPH, 52, 3}, + { DBGBUS_PERIPH, 52, 4}, + { DBGBUS_PERIPH, 52, 5}, + { DBGBUS_PERIPH, 52, 6}, + { DBGBUS_PERIPH, 52, 7}, + + /* tear-check */ + { DBGBUS_PERIPH, 63, 0 }, + { DBGBUS_PERIPH, 64, 0 }, + { DBGBUS_PERIPH, 65, 0 }, + { DBGBUS_PERIPH, 73, 0 }, + { DBGBUS_PERIPH, 74, 0 }, + + /* cdwn */ + { DBGBUS_PERIPH, 80, 0}, + { DBGBUS_PERIPH, 80, 1}, + { DBGBUS_PERIPH, 80, 2}, + + { DBGBUS_PERIPH, 81, 0}, + { DBGBUS_PERIPH, 81, 1}, + { DBGBUS_PERIPH, 81, 2}, + + { DBGBUS_PERIPH, 82, 0}, + { DBGBUS_PERIPH, 82, 1}, + { DBGBUS_PERIPH, 82, 2}, + { DBGBUS_PERIPH, 82, 3}, + { DBGBUS_PERIPH, 82, 4}, + { DBGBUS_PERIPH, 82, 5}, + { DBGBUS_PERIPH, 82, 6}, + { DBGBUS_PERIPH, 82, 7}, + + /* hdmi */ + { DBGBUS_PERIPH, 68, 0}, + { DBGBUS_PERIPH, 68, 1}, + { DBGBUS_PERIPH, 68, 2}, + { DBGBUS_PERIPH, 68, 3}, + { DBGBUS_PERIPH, 68, 4}, + { DBGBUS_PERIPH, 68, 5}, + + /* edp */ + { DBGBUS_PERIPH, 69, 0}, + { DBGBUS_PERIPH, 69, 1}, + { DBGBUS_PERIPH, 69, 2}, + { DBGBUS_PERIPH, 69, 3}, + { DBGBUS_PERIPH, 69, 4}, + { DBGBUS_PERIPH, 69, 5}, + + /* dsi0 */ + { DBGBUS_PERIPH, 70, 0}, + { DBGBUS_PERIPH, 70, 1}, + { DBGBUS_PERIPH, 70, 2}, + { DBGBUS_PERIPH, 70, 3}, + { DBGBUS_PERIPH, 70, 4}, + { DBGBUS_PERIPH, 70, 5}, + + /* dsi1 */ + { DBGBUS_PERIPH, 71, 0}, + { DBGBUS_PERIPH, 71, 1}, + { DBGBUS_PERIPH, 71, 2}, + { DBGBUS_PERIPH, 71, 3}, + { DBGBUS_PERIPH, 71, 4}, + { DBGBUS_PERIPH, 71, 5}, +}; + +static struct vbif_debug_bus_entry vbif_dbg_bus_msm8998[] = { + {0x214, 0x21c, 16, 2, 0x0, 0xd}, /* arb clients */ + {0x214, 0x21c, 16, 2, 0x80, 0xc0}, /* arb clients */ + {0x214, 0x21c, 16, 2, 0x100, 0x140}, /* arb clients */ + {0x214, 0x21c, 0, 16, 0x0, 0xf}, /* xin blocks - axi side */ + {0x214, 0x21c, 0, 16, 0x80, 0xa4}, /* xin blocks - axi side */ + {0x214, 0x21c, 0, 15, 0x100, 0x124}, /* xin blocks - axi side */ + {0x21c, 0x214, 0, 14, 0, 0xc}, /* xin blocks - clock side */ +}; + +/** + * _dpu_dbg_enable_power - use callback to turn power on for hw register access + * @enable: whether to turn power on or off + */ +static inline void _dpu_dbg_enable_power(int enable) +{ + if (enable) + pm_runtime_get_sync(dpu_dbg_base.dev); + else + pm_runtime_put_sync(dpu_dbg_base.dev); +} + +static void _dpu_dbg_dump_dpu_dbg_bus(struct dpu_dbg_dpu_debug_bus *bus) +{ + bool in_log, in_mem; + u32 **dump_mem = NULL; + u32 *dump_addr = NULL; + u32 status = 0; + struct dpu_debug_bus_entry *head; + phys_addr_t phys = 0; + int list_size; + int i; + u32 offset; + void __iomem *mem_base = NULL; + struct dpu_dbg_reg_base *reg_base; + + if (!bus || !bus->cmn.entries_size) + return; + + list_for_each_entry(reg_base, &dpu_dbg_base.reg_base_list, + reg_base_head) + if (strlen(reg_base->name) && + !strcmp(reg_base->name, bus->cmn.name)) + mem_base = reg_base->base + bus->top_blk_off; + + if (!mem_base) { + pr_err("unable to find mem_base for %s\n", bus->cmn.name); + return; + } + + dump_mem = &bus->cmn.dumped_content; + + /* will keep in memory 4 entries of 4 bytes each */ + list_size = (bus->cmn.entries_size * 4 * 4); + + in_log = (bus->cmn.enable_mask & DPU_DBG_DUMP_IN_LOG); + in_mem = (bus->cmn.enable_mask & DPU_DBG_DUMP_IN_MEM); + + if (!in_log && !in_mem) + return; + + dev_info(dpu_dbg_base.dev, "======== start %s dump =========\n", + bus->cmn.name); + + if (in_mem) { + if (!(*dump_mem)) + *dump_mem = dma_alloc_coherent(dpu_dbg_base.dev, + list_size, &phys, GFP_KERNEL); + + if (*dump_mem) { + dump_addr = *dump_mem; + dev_info(dpu_dbg_base.dev, + "%s: start_addr:0x%pK len:0x%x\n", + __func__, dump_addr, list_size); + } else { + in_mem = false; + pr_err("dump_mem: allocation fails\n"); + } + } + + _dpu_dbg_enable_power(true); + for (i = 0; i < bus->cmn.entries_size; i++) { + head = bus->entries + i; + writel_relaxed(TEST_MASK(head->block_id, head->test_id), + mem_base + head->wr_addr); + wmb(); /* make sure test bits were written */ + + if (bus->cmn.flags & DBGBUS_FLAGS_DSPP) { + offset = DBGBUS_DSPP_STATUS; + /* keep DSPP test point enabled */ + if (head->wr_addr != DBGBUS_DSPP) + writel_relaxed(0xF, mem_base + DBGBUS_DSPP); + } else { + offset = head->wr_addr + 0x4; + } + + status = readl_relaxed(mem_base + offset); + + if (in_log) + dev_info(dpu_dbg_base.dev, + "waddr=0x%x blk=%d tst=%d val=0x%x\n", + head->wr_addr, head->block_id, + head->test_id, status); + + if (dump_addr && in_mem) { + dump_addr[i*4] = head->wr_addr; + dump_addr[i*4 + 1] = head->block_id; + dump_addr[i*4 + 2] = head->test_id; + dump_addr[i*4 + 3] = status; + } + + if (head->analyzer) + head->analyzer(mem_base, head, status); + + /* Disable debug bus once we are done */ + writel_relaxed(0, mem_base + head->wr_addr); + if (bus->cmn.flags & DBGBUS_FLAGS_DSPP && + head->wr_addr != DBGBUS_DSPP) + writel_relaxed(0x0, mem_base + DBGBUS_DSPP); + } + _dpu_dbg_enable_power(false); + + dev_info(dpu_dbg_base.dev, "======== end %s dump =========\n", + bus->cmn.name); +} + +static void _dpu_dbg_dump_vbif_debug_bus_entry( + struct vbif_debug_bus_entry *head, void __iomem *mem_base, + u32 *dump_addr, bool in_log) +{ + int i, j; + u32 val; + + if (!dump_addr && !in_log) + return; + + for (i = 0; i < head->block_cnt; i++) { + writel_relaxed(1 << (i + head->bit_offset), + mem_base + head->block_bus_addr); + /* make sure that current bus blcok enable */ + wmb(); + for (j = head->test_pnt_start; j < head->test_pnt_cnt; j++) { + writel_relaxed(j, mem_base + head->block_bus_addr + 4); + /* make sure that test point is enabled */ + wmb(); + val = readl_relaxed(mem_base + MMSS_VBIF_TEST_BUS_OUT); + if (dump_addr) { + *dump_addr++ = head->block_bus_addr; + *dump_addr++ = i; + *dump_addr++ = j; + *dump_addr++ = val; + } + if (in_log) + dev_info(dpu_dbg_base.dev, + "testpoint:%x arb/xin id=%d index=%d val=0x%x\n", + head->block_bus_addr, i, j, val); + } + } +} + +static void _dpu_dbg_dump_vbif_dbg_bus(struct dpu_dbg_vbif_debug_bus *bus) +{ + bool in_log, in_mem; + u32 **dump_mem = NULL; + u32 *dump_addr = NULL; + u32 value, d0, d1; + unsigned long reg, reg1, reg2; + struct vbif_debug_bus_entry *head; + phys_addr_t phys = 0; + int i, list_size = 0; + void __iomem *mem_base = NULL; + struct vbif_debug_bus_entry *dbg_bus; + u32 bus_size; + struct dpu_dbg_reg_base *reg_base; + + if (!bus || !bus->cmn.entries_size) + return; + + list_for_each_entry(reg_base, &dpu_dbg_base.reg_base_list, + reg_base_head) + if (strlen(reg_base->name) && + !strcmp(reg_base->name, bus->cmn.name)) + mem_base = reg_base->base; + + if (!mem_base) { + pr_err("unable to find mem_base for %s\n", bus->cmn.name); + return; + } + + dbg_bus = bus->entries; + bus_size = bus->cmn.entries_size; + list_size = bus->cmn.entries_size; + dump_mem = &bus->cmn.dumped_content; + + dev_info(dpu_dbg_base.dev, "======== start %s dump =========\n", + bus->cmn.name); + + if (!dump_mem || !dbg_bus || !bus_size || !list_size) + return; + + /* allocate memory for each test point */ + for (i = 0; i < bus_size; i++) { + head = dbg_bus + i; + list_size += (head->block_cnt * head->test_pnt_cnt); + } + + /* 4 bytes * 4 entries for each test point*/ + list_size *= 16; + + in_log = (bus->cmn.enable_mask & DPU_DBG_DUMP_IN_LOG); + in_mem = (bus->cmn.enable_mask & DPU_DBG_DUMP_IN_MEM); + + if (!in_log && !in_mem) + return; + + if (in_mem) { + if (!(*dump_mem)) + *dump_mem = dma_alloc_coherent(dpu_dbg_base.dev, + list_size, &phys, GFP_KERNEL); + + if (*dump_mem) { + dump_addr = *dump_mem; + dev_info(dpu_dbg_base.dev, + "%s: start_addr:0x%pK len:0x%x\n", + __func__, dump_addr, list_size); + } else { + in_mem = false; + pr_err("dump_mem: allocation fails\n"); + } + } + + _dpu_dbg_enable_power(true); + + value = readl_relaxed(mem_base + MMSS_VBIF_CLKON); + writel_relaxed(value | BIT(1), mem_base + MMSS_VBIF_CLKON); + + /* make sure that vbif core is on */ + wmb(); + + /** + * Extract VBIF error info based on XIN halt and error status. + * If the XIN client is not in HALT state, or an error is detected, + * then retrieve the VBIF error info for it. + */ + reg = readl_relaxed(mem_base + MMSS_VBIF_XIN_HALT_CTRL1); + reg1 = readl_relaxed(mem_base + MMSS_VBIF_PND_ERR); + reg2 = readl_relaxed(mem_base + MMSS_VBIF_SRC_ERR); + dev_err(dpu_dbg_base.dev, + "XIN HALT:0x%lX, PND ERR:0x%lX, SRC ERR:0x%lX\n", + reg, reg1, reg2); + reg >>= 16; + reg &= ~(reg1 | reg2); + for (i = 0; i < MMSS_VBIF_CLIENT_NUM; i++) { + if (!test_bit(0, ®)) { + writel_relaxed(i, mem_base + MMSS_VBIF_ERR_INFO); + /* make sure reg write goes through */ + wmb(); + + d0 = readl_relaxed(mem_base + MMSS_VBIF_ERR_INFO); + d1 = readl_relaxed(mem_base + MMSS_VBIF_ERR_INFO_1); + + dev_err(dpu_dbg_base.dev, + "Client:%d, errinfo=0x%X, errinfo1=0x%X\n", + i, d0, d1); + } + reg >>= 1; + } + + for (i = 0; i < bus_size; i++) { + head = dbg_bus + i; + + writel_relaxed(0, mem_base + head->disable_bus_addr); + writel_relaxed(BIT(0), mem_base + MMSS_VBIF_TEST_BUS_OUT_CTRL); + /* make sure that other bus is off */ + wmb(); + + _dpu_dbg_dump_vbif_debug_bus_entry(head, mem_base, dump_addr, + in_log); + if (dump_addr) + dump_addr += (head->block_cnt * head->test_pnt_cnt * 4); + } + + _dpu_dbg_enable_power(false); + + dev_info(dpu_dbg_base.dev, "======== end %s dump =========\n", + bus->cmn.name); +} + +/** + * _dpu_dump_array - dump array of register bases + * @name: string indicating origin of dump + * @dump_dbgbus_dpu: whether to dump the dpu debug bus + * @dump_dbgbus_vbif_rt: whether to dump the vbif rt debug bus + */ +static void _dpu_dump_array(const char *name, bool dump_dbgbus_dpu, + bool dump_dbgbus_vbif_rt) +{ + if (dump_dbgbus_dpu) + _dpu_dbg_dump_dpu_dbg_bus(&dpu_dbg_base.dbgbus_dpu); + + if (dump_dbgbus_vbif_rt) + _dpu_dbg_dump_vbif_dbg_bus(&dpu_dbg_base.dbgbus_vbif_rt); +} + +/** + * _dpu_dump_work - deferred dump work function + * @work: work structure + */ +static void _dpu_dump_work(struct work_struct *work) +{ + _dpu_dump_array("dpudump_workitem", + dpu_dbg_base.dbgbus_dpu.cmn.include_in_deferred_work, + dpu_dbg_base.dbgbus_vbif_rt.cmn.include_in_deferred_work); +} + +void dpu_dbg_dump(bool queue_work, const char *name, bool dump_dbgbus_dpu, + bool dump_dbgbus_vbif_rt) +{ + if (queue_work && work_pending(&dpu_dbg_base.dump_work)) + return; + + if (!queue_work) { + _dpu_dump_array(name, dump_dbgbus_dpu, dump_dbgbus_vbif_rt); + return; + } + + /* schedule work to dump later */ + dpu_dbg_base.dbgbus_dpu.cmn.include_in_deferred_work = dump_dbgbus_dpu; + dpu_dbg_base.dbgbus_vbif_rt.cmn.include_in_deferred_work = + dump_dbgbus_vbif_rt; + schedule_work(&dpu_dbg_base.dump_work); +} + +/* + * dpu_dbg_debugfs_open - debugfs open handler for debug dump + * @inode: debugfs inode + * @file: file handle + */ +static int dpu_dbg_debugfs_open(struct inode *inode, struct file *file) +{ + /* non-seekable */ + file->f_mode &= ~(FMODE_LSEEK | FMODE_PREAD | FMODE_PWRITE); + file->private_data = inode->i_private; + return 0; +} + +/** + * dpu_dbg_dump_write - debugfs write handler for debug dump + * @file: file handler + * @user_buf: user buffer content from debugfs + * @count: size of user buffer + * @ppos: position offset of user buffer + */ +static ssize_t dpu_dbg_dump_write(struct file *file, + const char __user *user_buf, size_t count, loff_t *ppos) +{ + _dpu_dump_array("dump_debugfs", true, true); + return count; +} + +static const struct file_operations dpu_dbg_dump_fops = { + .open = dpu_dbg_debugfs_open, + .write = dpu_dbg_dump_write, +}; + +int dpu_dbg_debugfs_register(struct dentry *debugfs_root) +{ + static struct dpu_dbg_base *dbg = &dpu_dbg_base; + char debug_name[80] = ""; + + if (!debugfs_root) + return -EINVAL; + + debugfs_create_file("dump", 0600, debugfs_root, NULL, + &dpu_dbg_dump_fops); + + if (dbg->dbgbus_dpu.entries) { + dbg->dbgbus_dpu.cmn.name = DBGBUS_NAME_DPU; + snprintf(debug_name, sizeof(debug_name), "%s_dbgbus", + dbg->dbgbus_dpu.cmn.name); + dbg->dbgbus_dpu.cmn.enable_mask = DEFAULT_DBGBUS_DPU; + debugfs_create_u32(debug_name, 0600, debugfs_root, + &dbg->dbgbus_dpu.cmn.enable_mask); + } + + if (dbg->dbgbus_vbif_rt.entries) { + dbg->dbgbus_vbif_rt.cmn.name = DBGBUS_NAME_VBIF_RT; + snprintf(debug_name, sizeof(debug_name), "%s_dbgbus", + dbg->dbgbus_vbif_rt.cmn.name); + dbg->dbgbus_vbif_rt.cmn.enable_mask = DEFAULT_DBGBUS_VBIFRT; + debugfs_create_u32(debug_name, 0600, debugfs_root, + &dbg->dbgbus_vbif_rt.cmn.enable_mask); + } + + return 0; +} + +static void _dpu_dbg_debugfs_destroy(void) +{ +} + +void dpu_dbg_init_dbg_buses(u32 hwversion) +{ + static struct dpu_dbg_base *dbg = &dpu_dbg_base; + + memset(&dbg->dbgbus_dpu, 0, sizeof(dbg->dbgbus_dpu)); + memset(&dbg->dbgbus_vbif_rt, 0, sizeof(dbg->dbgbus_vbif_rt)); + + if (IS_MSM8998_TARGET(hwversion)) { + dbg->dbgbus_dpu.entries = dbg_bus_dpu_8998; + dbg->dbgbus_dpu.cmn.entries_size = ARRAY_SIZE(dbg_bus_dpu_8998); + dbg->dbgbus_dpu.cmn.flags = DBGBUS_FLAGS_DSPP; + + dbg->dbgbus_vbif_rt.entries = vbif_dbg_bus_msm8998; + dbg->dbgbus_vbif_rt.cmn.entries_size = + ARRAY_SIZE(vbif_dbg_bus_msm8998); + } else if (IS_SDM845_TARGET(hwversion) || IS_SDM670_TARGET(hwversion)) { + dbg->dbgbus_dpu.entries = dbg_bus_dpu_sdm845; + dbg->dbgbus_dpu.cmn.entries_size = + ARRAY_SIZE(dbg_bus_dpu_sdm845); + dbg->dbgbus_dpu.cmn.flags = DBGBUS_FLAGS_DSPP; + + /* vbif is unchanged vs 8998 */ + dbg->dbgbus_vbif_rt.entries = vbif_dbg_bus_msm8998; + dbg->dbgbus_vbif_rt.cmn.entries_size = + ARRAY_SIZE(vbif_dbg_bus_msm8998); + } else { + pr_err("unsupported chipset id %X\n", hwversion); + } +} + +int dpu_dbg_init(struct device *dev) +{ + if (!dev) { + pr_err("invalid params\n"); + return -EINVAL; + } + + INIT_LIST_HEAD(&dpu_dbg_base.reg_base_list); + dpu_dbg_base.dev = dev; + + INIT_WORK(&dpu_dbg_base.dump_work, _dpu_dump_work); + + return 0; +} + +/** + * dpu_dbg_destroy - destroy dpu debug facilities + */ +void dpu_dbg_destroy(void) +{ + _dpu_dbg_debugfs_destroy(); +} + +void dpu_dbg_set_dpu_top_offset(u32 blk_off) +{ + dpu_dbg_base.dbgbus_dpu.top_blk_off = blk_off; +} diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_dbg.h b/drivers/gpu/drm/msm/disp/dpu1/dpu_dbg.h new file mode 100644 index 000000000000..1e6fa945f98b --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_dbg.h @@ -0,0 +1,103 @@ +/* Copyright (c) 2016-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#ifndef DPU_DBG_H_ +#define DPU_DBG_H_ + +#include <stdarg.h> +#include <linux/debugfs.h> +#include <linux/list.h> + +enum dpu_dbg_dump_flag { + DPU_DBG_DUMP_IN_LOG = BIT(0), + DPU_DBG_DUMP_IN_MEM = BIT(1), +}; + +#if defined(CONFIG_DEBUG_FS) + +/** + * dpu_dbg_init_dbg_buses - initialize debug bus dumping support for the chipset + * @hwversion: Chipset revision + */ +void dpu_dbg_init_dbg_buses(u32 hwversion); + +/** + * dpu_dbg_init - initialize global dpu debug facilities: regdump + * @dev: device handle + * Returns: 0 or -ERROR + */ +int dpu_dbg_init(struct device *dev); + +/** + * dpu_dbg_debugfs_register - register entries at the given debugfs dir + * @debugfs_root: debugfs root in which to create dpu debug entries + * Returns: 0 or -ERROR + */ +int dpu_dbg_debugfs_register(struct dentry *debugfs_root); + +/** + * dpu_dbg_destroy - destroy the global dpu debug facilities + * Returns: none + */ +void dpu_dbg_destroy(void); + +/** + * dpu_dbg_dump - trigger dumping of all dpu_dbg facilities + * @queue_work: whether to queue the dumping work to the work_struct + * @name: string indicating origin of dump + * @dump_dbgbus: dump the dpu debug bus + * @dump_vbif_rt: dump the vbif rt bus + * Returns: none + */ +void dpu_dbg_dump(bool queue_work, const char *name, bool dump_dbgbus_dpu, + bool dump_dbgbus_vbif_rt); + +/** + * dpu_dbg_set_dpu_top_offset - set the target specific offset from mdss base + * address of the top registers. Used for accessing debug bus controls. + * @blk_off: offset from mdss base of the top block + */ +void dpu_dbg_set_dpu_top_offset(u32 blk_off); + +#else + +static inline void dpu_dbg_init_dbg_buses(u32 hwversion) +{ +} + +static inline int dpu_dbg_init(struct device *dev) +{ + return 0; +} + +static inline int dpu_dbg_debugfs_register(struct dentry *debugfs_root) +{ + return 0; +} + +static inline void dpu_dbg_destroy(void) +{ +} + +static inline void dpu_dbg_dump(bool queue_work, const char *name, + bool dump_dbgbus_dpu, bool dump_dbgbus_vbif_rt) +{ +} + +static inline void dpu_dbg_set_dpu_top_offset(u32 blk_off) +{ +} + +#endif /* defined(CONFIG_DEBUG_FS) */ + + +#endif /* DPU_DBG_H_ */ diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.c new file mode 100644 index 000000000000..0bd3eda93e22 --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.c @@ -0,0 +1,2498 @@ +/* + * Copyright (c) 2014-2018, The Linux Foundation. All rights reserved. + * Copyright (C) 2013 Red Hat + * Author: Rob Clark <robdclark@gmail.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published by + * the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program. If not, see <http://www.gnu.org/licenses/>. + */ + +#define pr_fmt(fmt) "[drm:%s:%d] " fmt, __func__, __LINE__ +#include <linux/kthread.h> +#include <linux/debugfs.h> +#include <linux/seq_file.h> + +#include "msm_drv.h" +#include "dpu_kms.h" +#include <drm/drm_crtc.h> +#include <drm/drm_crtc_helper.h> +#include "dpu_hwio.h" +#include "dpu_hw_catalog.h" +#include "dpu_hw_intf.h" +#include "dpu_hw_ctl.h" +#include "dpu_formats.h" +#include "dpu_encoder_phys.h" +#include "dpu_crtc.h" +#include "dpu_trace.h" +#include "dpu_core_irq.h" + +#define DPU_DEBUG_ENC(e, fmt, ...) DPU_DEBUG("enc%d " fmt,\ + (e) ? (e)->base.base.id : -1, ##__VA_ARGS__) + +#define DPU_ERROR_ENC(e, fmt, ...) DPU_ERROR("enc%d " fmt,\ + (e) ? (e)->base.base.id : -1, ##__VA_ARGS__) + +#define DPU_DEBUG_PHYS(p, fmt, ...) DPU_DEBUG("enc%d intf%d pp%d " fmt,\ + (p) ? (p)->parent->base.id : -1, \ + (p) ? (p)->intf_idx - INTF_0 : -1, \ + (p) ? ((p)->hw_pp ? (p)->hw_pp->idx - PINGPONG_0 : -1) : -1, \ + ##__VA_ARGS__) + +#define DPU_ERROR_PHYS(p, fmt, ...) DPU_ERROR("enc%d intf%d pp%d " fmt,\ + (p) ? (p)->parent->base.id : -1, \ + (p) ? (p)->intf_idx - INTF_0 : -1, \ + (p) ? ((p)->hw_pp ? (p)->hw_pp->idx - PINGPONG_0 : -1) : -1, \ + ##__VA_ARGS__) + +/* + * Two to anticipate panels that can do cmd/vid dynamic switching + * plan is to create all possible physical encoder types, and switch between + * them at runtime + */ +#define NUM_PHYS_ENCODER_TYPES 2 + +#define MAX_PHYS_ENCODERS_PER_VIRTUAL \ + (MAX_H_TILES_PER_DISPLAY * NUM_PHYS_ENCODER_TYPES) + +#define MAX_CHANNELS_PER_ENC 2 + +#define MISR_BUFF_SIZE 256 + +#define IDLE_SHORT_TIMEOUT 1 + +#define MAX_VDISPLAY_SPLIT 1080 + +/** + * enum dpu_enc_rc_events - events for resource control state machine + * @DPU_ENC_RC_EVENT_KICKOFF: + * This event happens at NORMAL priority. + * Event that signals the start of the transfer. When this event is + * received, enable MDP/DSI core clocks. Regardless of the previous + * state, the resource should be in ON state at the end of this event. + * @DPU_ENC_RC_EVENT_FRAME_DONE: + * This event happens at INTERRUPT level. + * Event signals the end of the data transfer after the PP FRAME_DONE + * event. At the end of this event, a delayed work is scheduled to go to + * IDLE_PC state after IDLE_TIMEOUT time. + * @DPU_ENC_RC_EVENT_PRE_STOP: + * This event happens at NORMAL priority. + * This event, when received during the ON state, leave the RC STATE + * in the PRE_OFF state. It should be followed by the STOP event as + * part of encoder disable. + * If received during IDLE or OFF states, it will do nothing. + * @DPU_ENC_RC_EVENT_STOP: + * This event happens at NORMAL priority. + * When this event is received, disable all the MDP/DSI core clocks, and + * disable IRQs. It should be called from the PRE_OFF or IDLE states. + * IDLE is expected when IDLE_PC has run, and PRE_OFF did nothing. + * PRE_OFF is expected when PRE_STOP was executed during the ON state. + * Resource state should be in OFF at the end of the event. + * @DPU_ENC_RC_EVENT_ENTER_IDLE: + * This event happens at NORMAL priority from a work item. + * Event signals that there were no frame updates for IDLE_TIMEOUT time. + * This would disable MDP/DSI core clocks and change the resource state + * to IDLE. + */ +enum dpu_enc_rc_events { + DPU_ENC_RC_EVENT_KICKOFF = 1, + DPU_ENC_RC_EVENT_FRAME_DONE, + DPU_ENC_RC_EVENT_PRE_STOP, + DPU_ENC_RC_EVENT_STOP, + DPU_ENC_RC_EVENT_ENTER_IDLE +}; + +/* + * enum dpu_enc_rc_states - states that the resource control maintains + * @DPU_ENC_RC_STATE_OFF: Resource is in OFF state + * @DPU_ENC_RC_STATE_PRE_OFF: Resource is transitioning to OFF state + * @DPU_ENC_RC_STATE_ON: Resource is in ON state + * @DPU_ENC_RC_STATE_MODESET: Resource is in modeset state + * @DPU_ENC_RC_STATE_IDLE: Resource is in IDLE state + */ +enum dpu_enc_rc_states { + DPU_ENC_RC_STATE_OFF, + DPU_ENC_RC_STATE_PRE_OFF, + DPU_ENC_RC_STATE_ON, + DPU_ENC_RC_STATE_IDLE +}; + +/** + * struct dpu_encoder_virt - virtual encoder. Container of one or more physical + * encoders. Virtual encoder manages one "logical" display. Physical + * encoders manage one intf block, tied to a specific panel/sub-panel. + * Virtual encoder defers as much as possible to the physical encoders. + * Virtual encoder registers itself with the DRM Framework as the encoder. + * @base: drm_encoder base class for registration with DRM + * @enc_spin_lock: Virtual-Encoder-Wide Spin Lock for IRQ purposes + * @bus_scaling_client: Client handle to the bus scaling interface + * @num_phys_encs: Actual number of physical encoders contained. + * @phys_encs: Container of physical encoders managed. + * @cur_master: Pointer to the current master in this mode. Optimization + * Only valid after enable. Cleared as disable. + * @hw_pp Handle to the pingpong blocks used for the display. No. + * pingpong blocks can be different than num_phys_encs. + * @intfs_swapped Whether or not the phys_enc interfaces have been swapped + * for partial update right-only cases, such as pingpong + * split where virtual pingpong does not generate IRQs + * @crtc_vblank_cb: Callback into the upper layer / CRTC for + * notification of the VBLANK + * @crtc_vblank_cb_data: Data from upper layer for VBLANK notification + * @crtc_kickoff_cb: Callback into CRTC that will flush & start + * all CTL paths + * @crtc_kickoff_cb_data: Opaque user data given to crtc_kickoff_cb + * @debugfs_root: Debug file system root file node + * @enc_lock: Lock around physical encoder create/destroy and + access. + * @frame_busy_mask: Bitmask tracking which phys_enc we are still + * busy processing current command. + * Bit0 = phys_encs[0] etc. + * @crtc_frame_event_cb: callback handler for frame event + * @crtc_frame_event_cb_data: callback handler private data + * @frame_done_timeout: frame done timeout in Hz + * @frame_done_timer: watchdog timer for frame done event + * @vsync_event_timer: vsync timer + * @disp_info: local copy of msm_display_info struct + * @misr_enable: misr enable/disable status + * @misr_frame_count: misr frame count before start capturing the data + * @idle_pc_supported: indicate if idle power collaps is supported + * @rc_lock: resource control mutex lock to protect + * virt encoder over various state changes + * @rc_state: resource controller state + * @delayed_off_work: delayed worker to schedule disabling of + * clks and resources after IDLE_TIMEOUT time. + * @vsync_event_work: worker to handle vsync event for autorefresh + * @topology: topology of the display + * @mode_set_complete: flag to indicate modeset completion + * @idle_timeout: idle timeout duration in milliseconds + */ +struct dpu_encoder_virt { + struct drm_encoder base; + spinlock_t enc_spinlock; + uint32_t bus_scaling_client; + + uint32_t display_num_of_h_tiles; + + unsigned int num_phys_encs; + struct dpu_encoder_phys *phys_encs[MAX_PHYS_ENCODERS_PER_VIRTUAL]; + struct dpu_encoder_phys *cur_master; + struct dpu_hw_pingpong *hw_pp[MAX_CHANNELS_PER_ENC]; + + bool intfs_swapped; + + void (*crtc_vblank_cb)(void *); + void *crtc_vblank_cb_data; + + struct dentry *debugfs_root; + struct mutex enc_lock; + DECLARE_BITMAP(frame_busy_mask, MAX_PHYS_ENCODERS_PER_VIRTUAL); + void (*crtc_frame_event_cb)(void *, u32 event); + void *crtc_frame_event_cb_data; + + atomic_t frame_done_timeout; + struct timer_list frame_done_timer; + struct timer_list vsync_event_timer; + + struct msm_display_info disp_info; + bool misr_enable; + u32 misr_frame_count; + + bool idle_pc_supported; + struct mutex rc_lock; + enum dpu_enc_rc_states rc_state; + struct kthread_delayed_work delayed_off_work; + struct kthread_work vsync_event_work; + struct msm_display_topology topology; + bool mode_set_complete; + + u32 idle_timeout; +}; + +#define to_dpu_encoder_virt(x) container_of(x, struct dpu_encoder_virt, base) +static inline int _dpu_encoder_power_enable(struct dpu_encoder_virt *dpu_enc, + bool enable) +{ + struct drm_encoder *drm_enc; + struct msm_drm_private *priv; + struct dpu_kms *dpu_kms; + + if (!dpu_enc) { + DPU_ERROR("invalid dpu enc\n"); + return -EINVAL; + } + + drm_enc = &dpu_enc->base; + if (!drm_enc->dev || !drm_enc->dev->dev_private) { + DPU_ERROR("drm device invalid\n"); + return -EINVAL; + } + + priv = drm_enc->dev->dev_private; + if (!priv->kms) { + DPU_ERROR("invalid kms\n"); + return -EINVAL; + } + + dpu_kms = to_dpu_kms(priv->kms); + + if (enable) + pm_runtime_get_sync(&dpu_kms->pdev->dev); + else + pm_runtime_put_sync(&dpu_kms->pdev->dev); + + return 0; +} + +void dpu_encoder_helper_report_irq_timeout(struct dpu_encoder_phys *phys_enc, + enum dpu_intr_idx intr_idx) +{ + DRM_ERROR("irq timeout id=%u, intf=%d, pp=%d, intr=%d\n", + DRMID(phys_enc->parent), phys_enc->intf_idx - INTF_0, + phys_enc->hw_pp->idx - PINGPONG_0, intr_idx); + + if (phys_enc->parent_ops->handle_frame_done) + phys_enc->parent_ops->handle_frame_done( + phys_enc->parent, phys_enc, + DPU_ENCODER_FRAME_EVENT_ERROR); +} + +static int dpu_encoder_helper_wait_event_timeout(int32_t drm_id, + int32_t hw_id, struct dpu_encoder_wait_info *info); + +int dpu_encoder_helper_wait_for_irq(struct dpu_encoder_phys *phys_enc, + enum dpu_intr_idx intr_idx, + struct dpu_encoder_wait_info *wait_info) +{ + struct dpu_encoder_irq *irq; + u32 irq_status; + int ret; + + if (!phys_enc || !wait_info || intr_idx >= INTR_IDX_MAX) { + DPU_ERROR("invalid params\n"); + return -EINVAL; + } + irq = &phys_enc->irq[intr_idx]; + + /* note: do master / slave checking outside */ + + /* return EWOULDBLOCK since we know the wait isn't necessary */ + if (phys_enc->enable_state == DPU_ENC_DISABLED) { + DRM_ERROR("encoder is disabled id=%u, intr=%d, hw=%d, irq=%d", + DRMID(phys_enc->parent), intr_idx, irq->hw_idx, + irq->irq_idx); + return -EWOULDBLOCK; + } + + if (irq->irq_idx < 0) { + DRM_DEBUG_KMS("skip irq wait id=%u, intr=%d, hw=%d, irq=%s", + DRMID(phys_enc->parent), intr_idx, irq->hw_idx, + irq->name); + return 0; + } + + DRM_DEBUG_KMS("id=%u, intr=%d, hw=%d, irq=%d, pp=%d, pending_cnt=%d", + DRMID(phys_enc->parent), intr_idx, irq->hw_idx, + irq->irq_idx, phys_enc->hw_pp->idx - PINGPONG_0, + atomic_read(wait_info->atomic_cnt)); + + ret = dpu_encoder_helper_wait_event_timeout( + DRMID(phys_enc->parent), + irq->hw_idx, + wait_info); + + if (ret <= 0) { + irq_status = dpu_core_irq_read(phys_enc->dpu_kms, + irq->irq_idx, true); + if (irq_status) { + unsigned long flags; + + DRM_DEBUG_KMS("irq not triggered id=%u, intr=%d, " + "hw=%d, irq=%d, pp=%d, atomic_cnt=%d", + DRMID(phys_enc->parent), intr_idx, + irq->hw_idx, irq->irq_idx, + phys_enc->hw_pp->idx - PINGPONG_0, + atomic_read(wait_info->atomic_cnt)); + local_irq_save(flags); + irq->cb.func(phys_enc, irq->irq_idx); + local_irq_restore(flags); + ret = 0; + } else { + ret = -ETIMEDOUT; + DRM_DEBUG_KMS("irq timeout id=%u, intr=%d, " + "hw=%d, irq=%d, pp=%d, atomic_cnt=%d", + DRMID(phys_enc->parent), intr_idx, + irq->hw_idx, irq->irq_idx, + phys_enc->hw_pp->idx - PINGPONG_0, + atomic_read(wait_info->atomic_cnt)); + } + } else { + ret = 0; + trace_dpu_enc_irq_wait_success(DRMID(phys_enc->parent), + intr_idx, irq->hw_idx, irq->irq_idx, + phys_enc->hw_pp->idx - PINGPONG_0, + atomic_read(wait_info->atomic_cnt)); + } + + return ret; +} + +int dpu_encoder_helper_register_irq(struct dpu_encoder_phys *phys_enc, + enum dpu_intr_idx intr_idx) +{ + struct dpu_encoder_irq *irq; + int ret = 0; + + if (!phys_enc || intr_idx >= INTR_IDX_MAX) { + DPU_ERROR("invalid params\n"); + return -EINVAL; + } + irq = &phys_enc->irq[intr_idx]; + + if (irq->irq_idx >= 0) { + DPU_DEBUG_PHYS(phys_enc, + "skipping already registered irq %s type %d\n", + irq->name, irq->intr_type); + return 0; + } + + irq->irq_idx = dpu_core_irq_idx_lookup(phys_enc->dpu_kms, + irq->intr_type, irq->hw_idx); + if (irq->irq_idx < 0) { + DPU_ERROR_PHYS(phys_enc, + "failed to lookup IRQ index for %s type:%d\n", + irq->name, irq->intr_type); + return -EINVAL; + } + + ret = dpu_core_irq_register_callback(phys_enc->dpu_kms, irq->irq_idx, + &irq->cb); + if (ret) { + DPU_ERROR_PHYS(phys_enc, + "failed to register IRQ callback for %s\n", + irq->name); + irq->irq_idx = -EINVAL; + return ret; + } + + ret = dpu_core_irq_enable(phys_enc->dpu_kms, &irq->irq_idx, 1); + if (ret) { + DRM_ERROR("enable failed id=%u, intr=%d, hw=%d, irq=%d", + DRMID(phys_enc->parent), intr_idx, irq->hw_idx, + irq->irq_idx); + dpu_core_irq_unregister_callback(phys_enc->dpu_kms, + irq->irq_idx, &irq->cb); + irq->irq_idx = -EINVAL; + return ret; + } + + trace_dpu_enc_irq_register_success(DRMID(phys_enc->parent), intr_idx, + irq->hw_idx, irq->irq_idx); + + return ret; +} + +int dpu_encoder_helper_unregister_irq(struct dpu_encoder_phys *phys_enc, + enum dpu_intr_idx intr_idx) +{ + struct dpu_encoder_irq *irq; + int ret; + + if (!phys_enc) { + DPU_ERROR("invalid encoder\n"); + return -EINVAL; + } + irq = &phys_enc->irq[intr_idx]; + + /* silently skip irqs that weren't registered */ + if (irq->irq_idx < 0) { + DRM_ERROR("duplicate unregister id=%u, intr=%d, hw=%d, irq=%d", + DRMID(phys_enc->parent), intr_idx, irq->hw_idx, + irq->irq_idx); + return 0; + } + + ret = dpu_core_irq_disable(phys_enc->dpu_kms, &irq->irq_idx, 1); + if (ret) { + DRM_ERROR("diable failed id=%u, intr=%d, hw=%d, irq=%d ret=%d", + DRMID(phys_enc->parent), intr_idx, irq->hw_idx, + irq->irq_idx, ret); + } + + ret = dpu_core_irq_unregister_callback(phys_enc->dpu_kms, irq->irq_idx, + &irq->cb); + if (ret) { + DRM_ERROR("unreg cb fail id=%u, intr=%d, hw=%d, irq=%d ret=%d", + DRMID(phys_enc->parent), intr_idx, irq->hw_idx, + irq->irq_idx, ret); + } + + trace_dpu_enc_irq_unregister_success(DRMID(phys_enc->parent), intr_idx, + irq->hw_idx, irq->irq_idx); + + irq->irq_idx = -EINVAL; + + return 0; +} + +void dpu_encoder_get_hw_resources(struct drm_encoder *drm_enc, + struct dpu_encoder_hw_resources *hw_res, + struct drm_connector_state *conn_state) +{ + struct dpu_encoder_virt *dpu_enc = NULL; + int i = 0; + + if (!hw_res || !drm_enc || !conn_state) { + DPU_ERROR("invalid argument(s), drm_enc %d, res %d, state %d\n", + drm_enc != 0, hw_res != 0, conn_state != 0); + return; + } + + dpu_enc = to_dpu_encoder_virt(drm_enc); + DPU_DEBUG_ENC(dpu_enc, "\n"); + + /* Query resources used by phys encs, expected to be without overlap */ + memset(hw_res, 0, sizeof(*hw_res)); + hw_res->display_num_of_h_tiles = dpu_enc->display_num_of_h_tiles; + + for (i = 0; i < dpu_enc->num_phys_encs; i++) { + struct dpu_encoder_phys *phys = dpu_enc->phys_encs[i]; + + if (phys && phys->ops.get_hw_resources) + phys->ops.get_hw_resources(phys, hw_res, conn_state); + } +} + +static void dpu_encoder_destroy(struct drm_encoder *drm_enc) +{ + struct dpu_encoder_virt *dpu_enc = NULL; + int i = 0; + + if (!drm_enc) { + DPU_ERROR("invalid encoder\n"); + return; + } + + dpu_enc = to_dpu_encoder_virt(drm_enc); + DPU_DEBUG_ENC(dpu_enc, "\n"); + + mutex_lock(&dpu_enc->enc_lock); + + for (i = 0; i < dpu_enc->num_phys_encs; i++) { + struct dpu_encoder_phys *phys = dpu_enc->phys_encs[i]; + + if (phys && phys->ops.destroy) { + phys->ops.destroy(phys); + --dpu_enc->num_phys_encs; + dpu_enc->phys_encs[i] = NULL; + } + } + + if (dpu_enc->num_phys_encs) + DPU_ERROR_ENC(dpu_enc, "expected 0 num_phys_encs not %d\n", + dpu_enc->num_phys_encs); + dpu_enc->num_phys_encs = 0; + mutex_unlock(&dpu_enc->enc_lock); + + drm_encoder_cleanup(drm_enc); + mutex_destroy(&dpu_enc->enc_lock); + + kfree(dpu_enc); +} + +void dpu_encoder_helper_split_config( + struct dpu_encoder_phys *phys_enc, + enum dpu_intf interface) +{ + struct dpu_encoder_virt *dpu_enc; + struct split_pipe_cfg cfg = { 0 }; + struct dpu_hw_mdp *hw_mdptop; + struct msm_display_info *disp_info; + + if (!phys_enc || !phys_enc->hw_mdptop || !phys_enc->parent) { + DPU_ERROR("invalid arg(s), encoder %d\n", phys_enc != 0); + return; + } + + dpu_enc = to_dpu_encoder_virt(phys_enc->parent); + hw_mdptop = phys_enc->hw_mdptop; + disp_info = &dpu_enc->disp_info; + + if (disp_info->intf_type != DRM_MODE_CONNECTOR_DSI) + return; + + /** + * disable split modes since encoder will be operating in as the only + * encoder, either for the entire use case in the case of, for example, + * single DSI, or for this frame in the case of left/right only partial + * update. + */ + if (phys_enc->split_role == ENC_ROLE_SOLO) { + if (hw_mdptop->ops.setup_split_pipe) + hw_mdptop->ops.setup_split_pipe(hw_mdptop, &cfg); + return; + } + + cfg.en = true; + cfg.mode = phys_enc->intf_mode; + cfg.intf = interface; + + if (cfg.en && phys_enc->ops.needs_single_flush && + phys_enc->ops.needs_single_flush(phys_enc)) + cfg.split_flush_en = true; + + if (phys_enc->split_role == ENC_ROLE_MASTER) { + DPU_DEBUG_ENC(dpu_enc, "enable %d\n", cfg.en); + + if (hw_mdptop->ops.setup_split_pipe) + hw_mdptop->ops.setup_split_pipe(hw_mdptop, &cfg); + } +} + +static void _dpu_encoder_adjust_mode(struct drm_connector *connector, + struct drm_display_mode *adj_mode) +{ + struct drm_display_mode *cur_mode; + + if (!connector || !adj_mode) + return; + + list_for_each_entry(cur_mode, &connector->modes, head) { + if (cur_mode->vdisplay == adj_mode->vdisplay && + cur_mode->hdisplay == adj_mode->hdisplay && + cur_mode->vrefresh == adj_mode->vrefresh) { + adj_mode->private = cur_mode->private; + adj_mode->private_flags |= cur_mode->private_flags; + } + } +} + +static struct msm_display_topology dpu_encoder_get_topology( + struct dpu_encoder_virt *dpu_enc, + struct dpu_kms *dpu_kms, + struct drm_display_mode *mode) +{ + struct msm_display_topology topology; + int i, intf_count = 0; + + for (i = 0; i < MAX_PHYS_ENCODERS_PER_VIRTUAL; i++) + if (dpu_enc->phys_encs[i]) + intf_count++; + + /* User split topology for width > 1080 */ + topology.num_lm = (mode->vdisplay > MAX_VDISPLAY_SPLIT) ? 2 : 1; + topology.num_enc = 0; + topology.num_intf = intf_count; + + return topology; +} +static int dpu_encoder_virt_atomic_check( + struct drm_encoder *drm_enc, + struct drm_crtc_state *crtc_state, + struct drm_connector_state *conn_state) +{ + struct dpu_encoder_virt *dpu_enc; + struct msm_drm_private *priv; + struct dpu_kms *dpu_kms; + const struct drm_display_mode *mode; + struct drm_display_mode *adj_mode; + struct msm_display_topology topology; + int i = 0; + int ret = 0; + + if (!drm_enc || !crtc_state || !conn_state) { + DPU_ERROR("invalid arg(s), drm_enc %d, crtc/conn state %d/%d\n", + drm_enc != 0, crtc_state != 0, conn_state != 0); + return -EINVAL; + } + + dpu_enc = to_dpu_encoder_virt(drm_enc); + DPU_DEBUG_ENC(dpu_enc, "\n"); + + priv = drm_enc->dev->dev_private; + dpu_kms = to_dpu_kms(priv->kms); + mode = &crtc_state->mode; + adj_mode = &crtc_state->adjusted_mode; + trace_dpu_enc_atomic_check(DRMID(drm_enc)); + + /* + * display drivers may populate private fields of the drm display mode + * structure while registering possible modes of a connector with DRM. + * These private fields are not populated back while DRM invokes + * the mode_set callbacks. This module retrieves and populates the + * private fields of the given mode. + */ + _dpu_encoder_adjust_mode(conn_state->connector, adj_mode); + + /* perform atomic check on the first physical encoder (master) */ + for (i = 0; i < dpu_enc->num_phys_encs; i++) { + struct dpu_encoder_phys *phys = dpu_enc->phys_encs[i]; + + if (phys && phys->ops.atomic_check) + ret = phys->ops.atomic_check(phys, crtc_state, + conn_state); + else if (phys && phys->ops.mode_fixup) + if (!phys->ops.mode_fixup(phys, mode, adj_mode)) + ret = -EINVAL; + + if (ret) { + DPU_ERROR_ENC(dpu_enc, + "mode unsupported, phys idx %d\n", i); + break; + } + } + + topology = dpu_encoder_get_topology(dpu_enc, dpu_kms, adj_mode); + + /* Reserve dynamic resources now. Indicating AtomicTest phase */ + if (!ret) { + /* + * Avoid reserving resources when mode set is pending. Topology + * info may not be available to complete reservation. + */ + if (drm_atomic_crtc_needs_modeset(crtc_state) + && dpu_enc->mode_set_complete) { + ret = dpu_rm_reserve(&dpu_kms->rm, drm_enc, crtc_state, + conn_state, topology, true); + dpu_enc->mode_set_complete = false; + } + } + + if (!ret) + drm_mode_set_crtcinfo(adj_mode, 0); + + trace_dpu_enc_atomic_check_flags(DRMID(drm_enc), adj_mode->flags, + adj_mode->private_flags); + + return ret; +} + +static void _dpu_encoder_update_vsync_source(struct dpu_encoder_virt *dpu_enc, + struct msm_display_info *disp_info) +{ + struct dpu_vsync_source_cfg vsync_cfg = { 0 }; + struct msm_drm_private *priv; + struct dpu_kms *dpu_kms; + struct dpu_hw_mdp *hw_mdptop; + struct drm_encoder *drm_enc; + int i; + + if (!dpu_enc || !disp_info) { + DPU_ERROR("invalid param dpu_enc:%d or disp_info:%d\n", + dpu_enc != NULL, disp_info != NULL); + return; + } else if (dpu_enc->num_phys_encs > ARRAY_SIZE(dpu_enc->hw_pp)) { + DPU_ERROR("invalid num phys enc %d/%d\n", + dpu_enc->num_phys_encs, + (int) ARRAY_SIZE(dpu_enc->hw_pp)); + return; + } + + drm_enc = &dpu_enc->base; + /* this pointers are checked in virt_enable_helper */ + priv = drm_enc->dev->dev_private; + + dpu_kms = to_dpu_kms(priv->kms); + if (!dpu_kms) { + DPU_ERROR("invalid dpu_kms\n"); + return; + } + + hw_mdptop = dpu_kms->hw_mdp; + if (!hw_mdptop) { + DPU_ERROR("invalid mdptop\n"); + return; + } + + if (hw_mdptop->ops.setup_vsync_source && + disp_info->capabilities & MSM_DISPLAY_CAP_CMD_MODE) { + for (i = 0; i < dpu_enc->num_phys_encs; i++) + vsync_cfg.ppnumber[i] = dpu_enc->hw_pp[i]->idx; + + vsync_cfg.pp_count = dpu_enc->num_phys_encs; + if (disp_info->is_te_using_watchdog_timer) + vsync_cfg.vsync_source = DPU_VSYNC_SOURCE_WD_TIMER_0; + else + vsync_cfg.vsync_source = DPU_VSYNC0_SOURCE_GPIO; + + hw_mdptop->ops.setup_vsync_source(hw_mdptop, &vsync_cfg); + } +} + +static void _dpu_encoder_irq_control(struct drm_encoder *drm_enc, bool enable) +{ + struct dpu_encoder_virt *dpu_enc; + int i; + + if (!drm_enc) { + DPU_ERROR("invalid encoder\n"); + return; + } + + dpu_enc = to_dpu_encoder_virt(drm_enc); + + DPU_DEBUG_ENC(dpu_enc, "enable:%d\n", enable); + for (i = 0; i < dpu_enc->num_phys_encs; i++) { + struct dpu_encoder_phys *phys = dpu_enc->phys_encs[i]; + + if (phys && phys->ops.irq_control) + phys->ops.irq_control(phys, enable); + } + +} + +static void _dpu_encoder_resource_control_helper(struct drm_encoder *drm_enc, + bool enable) +{ + struct msm_drm_private *priv; + struct dpu_kms *dpu_kms; + struct dpu_encoder_virt *dpu_enc; + + dpu_enc = to_dpu_encoder_virt(drm_enc); + priv = drm_enc->dev->dev_private; + dpu_kms = to_dpu_kms(priv->kms); + + trace_dpu_enc_rc_helper(DRMID(drm_enc), enable); + + if (!dpu_enc->cur_master) { + DPU_ERROR("encoder master not set\n"); + return; + } + + if (enable) { + /* enable DPU core clks */ + pm_runtime_get_sync(&dpu_kms->pdev->dev); + + /* enable all the irq */ + _dpu_encoder_irq_control(drm_enc, true); + + } else { + /* disable all the irq */ + _dpu_encoder_irq_control(drm_enc, false); + + /* disable DPU core clks */ + pm_runtime_put_sync(&dpu_kms->pdev->dev); + } + +} + +static int dpu_encoder_resource_control(struct drm_encoder *drm_enc, + u32 sw_event) +{ + struct dpu_encoder_virt *dpu_enc; + struct msm_drm_private *priv; + struct msm_drm_thread *disp_thread; + bool is_vid_mode = false; + + if (!drm_enc || !drm_enc->dev || !drm_enc->dev->dev_private || + !drm_enc->crtc) { + DPU_ERROR("invalid parameters\n"); + return -EINVAL; + } + dpu_enc = to_dpu_encoder_virt(drm_enc); + priv = drm_enc->dev->dev_private; + is_vid_mode = dpu_enc->disp_info.capabilities & + MSM_DISPLAY_CAP_VID_MODE; + + if (drm_enc->crtc->index >= ARRAY_SIZE(priv->disp_thread)) { + DPU_ERROR("invalid crtc index\n"); + return -EINVAL; + } + disp_thread = &priv->disp_thread[drm_enc->crtc->index]; + + /* + * when idle_pc is not supported, process only KICKOFF, STOP and MODESET + * events and return early for other events (ie wb display). + */ + if (!dpu_enc->idle_pc_supported && + (sw_event != DPU_ENC_RC_EVENT_KICKOFF && + sw_event != DPU_ENC_RC_EVENT_STOP && + sw_event != DPU_ENC_RC_EVENT_PRE_STOP)) + return 0; + + trace_dpu_enc_rc(DRMID(drm_enc), sw_event, dpu_enc->idle_pc_supported, + dpu_enc->rc_state, "begin"); + + switch (sw_event) { + case DPU_ENC_RC_EVENT_KICKOFF: + /* cancel delayed off work, if any */ + if (kthread_cancel_delayed_work_sync( + &dpu_enc->delayed_off_work)) + DPU_DEBUG_ENC(dpu_enc, "sw_event:%d, work cancelled\n", + sw_event); + + mutex_lock(&dpu_enc->rc_lock); + + /* return if the resource control is already in ON state */ + if (dpu_enc->rc_state == DPU_ENC_RC_STATE_ON) { + DRM_DEBUG_KMS("id;%u, sw_event:%d, rc in ON state\n", + DRMID(drm_enc), sw_event); + mutex_unlock(&dpu_enc->rc_lock); + return 0; + } else if (dpu_enc->rc_state != DPU_ENC_RC_STATE_OFF && + dpu_enc->rc_state != DPU_ENC_RC_STATE_IDLE) { + DRM_DEBUG_KMS("id;%u, sw_event:%d, rc in state %d\n", + DRMID(drm_enc), sw_event, + dpu_enc->rc_state); + mutex_unlock(&dpu_enc->rc_lock); + return -EINVAL; + } + + if (is_vid_mode && dpu_enc->rc_state == DPU_ENC_RC_STATE_IDLE) + _dpu_encoder_irq_control(drm_enc, true); + else + _dpu_encoder_resource_control_helper(drm_enc, true); + + dpu_enc->rc_state = DPU_ENC_RC_STATE_ON; + + trace_dpu_enc_rc(DRMID(drm_enc), sw_event, + dpu_enc->idle_pc_supported, dpu_enc->rc_state, + "kickoff"); + + mutex_unlock(&dpu_enc->rc_lock); + break; + + case DPU_ENC_RC_EVENT_FRAME_DONE: + /* + * mutex lock is not used as this event happens at interrupt + * context. And locking is not required as, the other events + * like KICKOFF and STOP does a wait-for-idle before executing + * the resource_control + */ + if (dpu_enc->rc_state != DPU_ENC_RC_STATE_ON) { + DRM_DEBUG_KMS("id:%d, sw_event:%d,rc:%d-unexpected\n", + DRMID(drm_enc), sw_event, + dpu_enc->rc_state); + return -EINVAL; + } + + /* + * schedule off work item only when there are no + * frames pending + */ + if (dpu_crtc_frame_pending(drm_enc->crtc) > 1) { + DRM_DEBUG_KMS("id:%d skip schedule work\n", + DRMID(drm_enc)); + return 0; + } + + kthread_queue_delayed_work( + &disp_thread->worker, + &dpu_enc->delayed_off_work, + msecs_to_jiffies(dpu_enc->idle_timeout)); + + trace_dpu_enc_rc(DRMID(drm_enc), sw_event, + dpu_enc->idle_pc_supported, dpu_enc->rc_state, + "frame done"); + break; + + case DPU_ENC_RC_EVENT_PRE_STOP: + /* cancel delayed off work, if any */ + if (kthread_cancel_delayed_work_sync( + &dpu_enc->delayed_off_work)) + DPU_DEBUG_ENC(dpu_enc, "sw_event:%d, work cancelled\n", + sw_event); + + mutex_lock(&dpu_enc->rc_lock); + + if (is_vid_mode && + dpu_enc->rc_state == DPU_ENC_RC_STATE_IDLE) { + _dpu_encoder_irq_control(drm_enc, true); + } + /* skip if is already OFF or IDLE, resources are off already */ + else if (dpu_enc->rc_state == DPU_ENC_RC_STATE_OFF || + dpu_enc->rc_state == DPU_ENC_RC_STATE_IDLE) { + DRM_DEBUG_KMS("id:%u, sw_event:%d, rc in %d state\n", + DRMID(drm_enc), sw_event, + dpu_enc->rc_state); + mutex_unlock(&dpu_enc->rc_lock); + return 0; + } + + dpu_enc->rc_state = DPU_ENC_RC_STATE_PRE_OFF; + + trace_dpu_enc_rc(DRMID(drm_enc), sw_event, + dpu_enc->idle_pc_supported, dpu_enc->rc_state, + "pre stop"); + + mutex_unlock(&dpu_enc->rc_lock); + break; + + case DPU_ENC_RC_EVENT_STOP: + mutex_lock(&dpu_enc->rc_lock); + + /* return if the resource control is already in OFF state */ + if (dpu_enc->rc_state == DPU_ENC_RC_STATE_OFF) { + DRM_DEBUG_KMS("id: %u, sw_event:%d, rc in OFF state\n", + DRMID(drm_enc), sw_event); + mutex_unlock(&dpu_enc->rc_lock); + return 0; + } else if (dpu_enc->rc_state == DPU_ENC_RC_STATE_ON) { + DRM_ERROR("id: %u, sw_event:%d, rc in state %d\n", + DRMID(drm_enc), sw_event, dpu_enc->rc_state); + mutex_unlock(&dpu_enc->rc_lock); + return -EINVAL; + } + + /** + * expect to arrive here only if in either idle state or pre-off + * and in IDLE state the resources are already disabled + */ + if (dpu_enc->rc_state == DPU_ENC_RC_STATE_PRE_OFF) + _dpu_encoder_resource_control_helper(drm_enc, false); + + dpu_enc->rc_state = DPU_ENC_RC_STATE_OFF; + + trace_dpu_enc_rc(DRMID(drm_enc), sw_event, + dpu_enc->idle_pc_supported, dpu_enc->rc_state, + "stop"); + + mutex_unlock(&dpu_enc->rc_lock); + break; + + case DPU_ENC_RC_EVENT_ENTER_IDLE: + mutex_lock(&dpu_enc->rc_lock); + + if (dpu_enc->rc_state != DPU_ENC_RC_STATE_ON) { + DRM_ERROR("id: %u, sw_event:%d, rc:%d !ON state\n", + DRMID(drm_enc), sw_event, dpu_enc->rc_state); + mutex_unlock(&dpu_enc->rc_lock); + return 0; + } + + /* + * if we are in ON but a frame was just kicked off, + * ignore the IDLE event, it's probably a stale timer event + */ + if (dpu_enc->frame_busy_mask[0]) { + DRM_ERROR("id:%u, sw_event:%d, rc:%d frame pending\n", + DRMID(drm_enc), sw_event, dpu_enc->rc_state); + mutex_unlock(&dpu_enc->rc_lock); + return 0; + } + + if (is_vid_mode) + _dpu_encoder_irq_control(drm_enc, false); + else + _dpu_encoder_resource_control_helper(drm_enc, false); + + dpu_enc->rc_state = DPU_ENC_RC_STATE_IDLE; + + trace_dpu_enc_rc(DRMID(drm_enc), sw_event, + dpu_enc->idle_pc_supported, dpu_enc->rc_state, + "idle"); + + mutex_unlock(&dpu_enc->rc_lock); + break; + + default: + DRM_ERROR("id:%u, unexpected sw_event: %d\n", DRMID(drm_enc), + sw_event); + trace_dpu_enc_rc(DRMID(drm_enc), sw_event, + dpu_enc->idle_pc_supported, dpu_enc->rc_state, + "error"); + break; + } + + trace_dpu_enc_rc(DRMID(drm_enc), sw_event, + dpu_enc->idle_pc_supported, dpu_enc->rc_state, + "end"); + return 0; +} + +static void dpu_encoder_virt_mode_set(struct drm_encoder *drm_enc, + struct drm_display_mode *mode, + struct drm_display_mode *adj_mode) +{ + struct dpu_encoder_virt *dpu_enc; + struct msm_drm_private *priv; + struct dpu_kms *dpu_kms; + struct list_head *connector_list; + struct drm_connector *conn = NULL, *conn_iter; + struct dpu_rm_hw_iter pp_iter; + struct msm_display_topology topology; + enum dpu_rm_topology_name topology_name; + int i = 0, ret; + + if (!drm_enc) { + DPU_ERROR("invalid encoder\n"); + return; + } + + dpu_enc = to_dpu_encoder_virt(drm_enc); + DPU_DEBUG_ENC(dpu_enc, "\n"); + + priv = drm_enc->dev->dev_private; + dpu_kms = to_dpu_kms(priv->kms); + connector_list = &dpu_kms->dev->mode_config.connector_list; + + trace_dpu_enc_mode_set(DRMID(drm_enc)); + + list_for_each_entry(conn_iter, connector_list, head) + if (conn_iter->encoder == drm_enc) + conn = conn_iter; + + if (!conn) { + DPU_ERROR_ENC(dpu_enc, "failed to find attached connector\n"); + return; + } else if (!conn->state) { + DPU_ERROR_ENC(dpu_enc, "invalid connector state\n"); + return; + } + + topology = dpu_encoder_get_topology(dpu_enc, dpu_kms, adj_mode); + + /* Reserve dynamic resources now. Indicating non-AtomicTest phase */ + ret = dpu_rm_reserve(&dpu_kms->rm, drm_enc, drm_enc->crtc->state, + conn->state, topology, false); + if (ret) { + DPU_ERROR_ENC(dpu_enc, + "failed to reserve hw resources, %d\n", ret); + return; + } + + dpu_rm_init_hw_iter(&pp_iter, drm_enc->base.id, DPU_HW_BLK_PINGPONG); + for (i = 0; i < MAX_CHANNELS_PER_ENC; i++) { + dpu_enc->hw_pp[i] = NULL; + if (!dpu_rm_get_hw(&dpu_kms->rm, &pp_iter)) + break; + dpu_enc->hw_pp[i] = (struct dpu_hw_pingpong *) pp_iter.hw; + } + + topology_name = dpu_rm_get_topology_name(topology); + for (i = 0; i < dpu_enc->num_phys_encs; i++) { + struct dpu_encoder_phys *phys = dpu_enc->phys_encs[i]; + + if (phys) { + if (!dpu_enc->hw_pp[i]) { + DPU_ERROR_ENC(dpu_enc, + "invalid pingpong block for the encoder\n"); + return; + } + phys->hw_pp = dpu_enc->hw_pp[i]; + phys->connector = conn->state->connector; + phys->topology_name = topology_name; + if (phys->ops.mode_set) + phys->ops.mode_set(phys, mode, adj_mode); + } + } + + dpu_enc->mode_set_complete = true; +} + +static void _dpu_encoder_virt_enable_helper(struct drm_encoder *drm_enc) +{ + struct dpu_encoder_virt *dpu_enc = NULL; + struct msm_drm_private *priv; + struct dpu_kms *dpu_kms; + + if (!drm_enc || !drm_enc->dev || !drm_enc->dev->dev_private) { + DPU_ERROR("invalid parameters\n"); + return; + } + + priv = drm_enc->dev->dev_private; + dpu_kms = to_dpu_kms(priv->kms); + if (!dpu_kms) { + DPU_ERROR("invalid dpu_kms\n"); + return; + } + + dpu_enc = to_dpu_encoder_virt(drm_enc); + if (!dpu_enc || !dpu_enc->cur_master) { + DPU_ERROR("invalid dpu encoder/master\n"); + return; + } + + if (dpu_enc->disp_info.intf_type == DRM_MODE_CONNECTOR_DisplayPort && + dpu_enc->cur_master->hw_mdptop && + dpu_enc->cur_master->hw_mdptop->ops.intf_audio_select) + dpu_enc->cur_master->hw_mdptop->ops.intf_audio_select( + dpu_enc->cur_master->hw_mdptop); + + if (dpu_enc->cur_master->hw_mdptop && + dpu_enc->cur_master->hw_mdptop->ops.reset_ubwc) + dpu_enc->cur_master->hw_mdptop->ops.reset_ubwc( + dpu_enc->cur_master->hw_mdptop, + dpu_kms->catalog); + + _dpu_encoder_update_vsync_source(dpu_enc, &dpu_enc->disp_info); +} + +void dpu_encoder_virt_restore(struct drm_encoder *drm_enc) +{ + struct dpu_encoder_virt *dpu_enc = NULL; + int i; + + if (!drm_enc) { + DPU_ERROR("invalid encoder\n"); + return; + } + dpu_enc = to_dpu_encoder_virt(drm_enc); + + for (i = 0; i < dpu_enc->num_phys_encs; i++) { + struct dpu_encoder_phys *phys = dpu_enc->phys_encs[i]; + + if (phys && (phys != dpu_enc->cur_master) && phys->ops.restore) + phys->ops.restore(phys); + } + + if (dpu_enc->cur_master && dpu_enc->cur_master->ops.restore) + dpu_enc->cur_master->ops.restore(dpu_enc->cur_master); + + _dpu_encoder_virt_enable_helper(drm_enc); +} + +static void dpu_encoder_virt_enable(struct drm_encoder *drm_enc) +{ + struct dpu_encoder_virt *dpu_enc = NULL; + int i, ret = 0; + struct drm_display_mode *cur_mode = NULL; + + if (!drm_enc) { + DPU_ERROR("invalid encoder\n"); + return; + } + dpu_enc = to_dpu_encoder_virt(drm_enc); + cur_mode = &dpu_enc->base.crtc->state->adjusted_mode; + + trace_dpu_enc_enable(DRMID(drm_enc), cur_mode->hdisplay, + cur_mode->vdisplay); + + dpu_enc->cur_master = NULL; + for (i = 0; i < dpu_enc->num_phys_encs; i++) { + struct dpu_encoder_phys *phys = dpu_enc->phys_encs[i]; + + if (phys && phys->ops.is_master && phys->ops.is_master(phys)) { + DPU_DEBUG_ENC(dpu_enc, "master is now idx %d\n", i); + dpu_enc->cur_master = phys; + break; + } + } + + if (!dpu_enc->cur_master) { + DPU_ERROR("virt encoder has no master! num_phys %d\n", i); + return; + } + + ret = dpu_encoder_resource_control(drm_enc, DPU_ENC_RC_EVENT_KICKOFF); + if (ret) { + DPU_ERROR_ENC(dpu_enc, "dpu resource control failed: %d\n", + ret); + return; + } + + for (i = 0; i < dpu_enc->num_phys_encs; i++) { + struct dpu_encoder_phys *phys = dpu_enc->phys_encs[i]; + + if (!phys) + continue; + + if (phys != dpu_enc->cur_master) { + if (phys->ops.enable) + phys->ops.enable(phys); + } + + if (dpu_enc->misr_enable && (dpu_enc->disp_info.capabilities & + MSM_DISPLAY_CAP_VID_MODE) && phys->ops.setup_misr) + phys->ops.setup_misr(phys, true, + dpu_enc->misr_frame_count); + } + + if (dpu_enc->cur_master->ops.enable) + dpu_enc->cur_master->ops.enable(dpu_enc->cur_master); + + _dpu_encoder_virt_enable_helper(drm_enc); +} + +static void dpu_encoder_virt_disable(struct drm_encoder *drm_enc) +{ + struct dpu_encoder_virt *dpu_enc = NULL; + struct msm_drm_private *priv; + struct dpu_kms *dpu_kms; + struct drm_display_mode *mode; + int i = 0; + + if (!drm_enc) { + DPU_ERROR("invalid encoder\n"); + return; + } else if (!drm_enc->dev) { + DPU_ERROR("invalid dev\n"); + return; + } else if (!drm_enc->dev->dev_private) { + DPU_ERROR("invalid dev_private\n"); + return; + } + + mode = &drm_enc->crtc->state->adjusted_mode; + + dpu_enc = to_dpu_encoder_virt(drm_enc); + DPU_DEBUG_ENC(dpu_enc, "\n"); + + priv = drm_enc->dev->dev_private; + dpu_kms = to_dpu_kms(priv->kms); + + trace_dpu_enc_disable(DRMID(drm_enc)); + + /* wait for idle */ + dpu_encoder_wait_for_event(drm_enc, MSM_ENC_TX_COMPLETE); + + dpu_encoder_resource_control(drm_enc, DPU_ENC_RC_EVENT_PRE_STOP); + + for (i = 0; i < dpu_enc->num_phys_encs; i++) { + struct dpu_encoder_phys *phys = dpu_enc->phys_encs[i]; + + if (phys && phys->ops.disable) + phys->ops.disable(phys); + } + + /* after phys waits for frame-done, should be no more frames pending */ + if (atomic_xchg(&dpu_enc->frame_done_timeout, 0)) { + DPU_ERROR("enc%d timeout pending\n", drm_enc->base.id); + del_timer_sync(&dpu_enc->frame_done_timer); + } + + dpu_encoder_resource_control(drm_enc, DPU_ENC_RC_EVENT_STOP); + + for (i = 0; i < dpu_enc->num_phys_encs; i++) { + if (dpu_enc->phys_encs[i]) + dpu_enc->phys_encs[i]->connector = NULL; + } + + dpu_enc->cur_master = NULL; + + DPU_DEBUG_ENC(dpu_enc, "encoder disabled\n"); + + dpu_rm_release(&dpu_kms->rm, drm_enc); +} + +static enum dpu_intf dpu_encoder_get_intf(struct dpu_mdss_cfg *catalog, + enum dpu_intf_type type, u32 controller_id) +{ + int i = 0; + + for (i = 0; i < catalog->intf_count; i++) { + if (catalog->intf[i].type == type + && catalog->intf[i].controller_id == controller_id) { + return catalog->intf[i].id; + } + } + + return INTF_MAX; +} + +static void dpu_encoder_vblank_callback(struct drm_encoder *drm_enc, + struct dpu_encoder_phys *phy_enc) +{ + struct dpu_encoder_virt *dpu_enc = NULL; + unsigned long lock_flags; + + if (!drm_enc || !phy_enc) + return; + + DPU_ATRACE_BEGIN("encoder_vblank_callback"); + dpu_enc = to_dpu_encoder_virt(drm_enc); + + spin_lock_irqsave(&dpu_enc->enc_spinlock, lock_flags); + if (dpu_enc->crtc_vblank_cb) + dpu_enc->crtc_vblank_cb(dpu_enc->crtc_vblank_cb_data); + spin_unlock_irqrestore(&dpu_enc->enc_spinlock, lock_flags); + + atomic_inc(&phy_enc->vsync_cnt); + DPU_ATRACE_END("encoder_vblank_callback"); +} + +static void dpu_encoder_underrun_callback(struct drm_encoder *drm_enc, + struct dpu_encoder_phys *phy_enc) +{ + if (!phy_enc) + return; + + DPU_ATRACE_BEGIN("encoder_underrun_callback"); + atomic_inc(&phy_enc->underrun_cnt); + trace_dpu_enc_underrun_cb(DRMID(drm_enc), + atomic_read(&phy_enc->underrun_cnt)); + DPU_ATRACE_END("encoder_underrun_callback"); +} + +void dpu_encoder_register_vblank_callback(struct drm_encoder *drm_enc, + void (*vbl_cb)(void *), void *vbl_data) +{ + struct dpu_encoder_virt *dpu_enc = to_dpu_encoder_virt(drm_enc); + unsigned long lock_flags; + bool enable; + int i; + + enable = vbl_cb ? true : false; + + if (!drm_enc) { + DPU_ERROR("invalid encoder\n"); + return; + } + trace_dpu_enc_vblank_cb(DRMID(drm_enc), enable); + + spin_lock_irqsave(&dpu_enc->enc_spinlock, lock_flags); + dpu_enc->crtc_vblank_cb = vbl_cb; + dpu_enc->crtc_vblank_cb_data = vbl_data; + spin_unlock_irqrestore(&dpu_enc->enc_spinlock, lock_flags); + + for (i = 0; i < dpu_enc->num_phys_encs; i++) { + struct dpu_encoder_phys *phys = dpu_enc->phys_encs[i]; + + if (phys && phys->ops.control_vblank_irq) + phys->ops.control_vblank_irq(phys, enable); + } +} + +void dpu_encoder_register_frame_event_callback(struct drm_encoder *drm_enc, + void (*frame_event_cb)(void *, u32 event), + void *frame_event_cb_data) +{ + struct dpu_encoder_virt *dpu_enc = to_dpu_encoder_virt(drm_enc); + unsigned long lock_flags; + bool enable; + + enable = frame_event_cb ? true : false; + + if (!drm_enc) { + DPU_ERROR("invalid encoder\n"); + return; + } + trace_dpu_enc_frame_event_cb(DRMID(drm_enc), enable); + + spin_lock_irqsave(&dpu_enc->enc_spinlock, lock_flags); + dpu_enc->crtc_frame_event_cb = frame_event_cb; + dpu_enc->crtc_frame_event_cb_data = frame_event_cb_data; + spin_unlock_irqrestore(&dpu_enc->enc_spinlock, lock_flags); +} + +static void dpu_encoder_frame_done_callback( + struct drm_encoder *drm_enc, + struct dpu_encoder_phys *ready_phys, u32 event) +{ + struct dpu_encoder_virt *dpu_enc = to_dpu_encoder_virt(drm_enc); + unsigned int i; + + if (event & (DPU_ENCODER_FRAME_EVENT_DONE + | DPU_ENCODER_FRAME_EVENT_ERROR + | DPU_ENCODER_FRAME_EVENT_PANEL_DEAD)) { + + if (!dpu_enc->frame_busy_mask[0]) { + /** + * suppress frame_done without waiter, + * likely autorefresh + */ + trace_dpu_enc_frame_done_cb_not_busy(DRMID(drm_enc), + event, ready_phys->intf_idx); + return; + } + + /* One of the physical encoders has become idle */ + for (i = 0; i < dpu_enc->num_phys_encs; i++) { + if (dpu_enc->phys_encs[i] == ready_phys) { + clear_bit(i, dpu_enc->frame_busy_mask); + trace_dpu_enc_frame_done_cb(DRMID(drm_enc), i, + dpu_enc->frame_busy_mask[0]); + } + } + + if (!dpu_enc->frame_busy_mask[0]) { + atomic_set(&dpu_enc->frame_done_timeout, 0); + del_timer(&dpu_enc->frame_done_timer); + + dpu_encoder_resource_control(drm_enc, + DPU_ENC_RC_EVENT_FRAME_DONE); + + if (dpu_enc->crtc_frame_event_cb) + dpu_enc->crtc_frame_event_cb( + dpu_enc->crtc_frame_event_cb_data, + event); + } + } else { + if (dpu_enc->crtc_frame_event_cb) + dpu_enc->crtc_frame_event_cb( + dpu_enc->crtc_frame_event_cb_data, event); + } +} + +static void dpu_encoder_off_work(struct kthread_work *work) +{ + struct dpu_encoder_virt *dpu_enc = container_of(work, + struct dpu_encoder_virt, delayed_off_work.work); + + if (!dpu_enc) { + DPU_ERROR("invalid dpu encoder\n"); + return; + } + + dpu_encoder_resource_control(&dpu_enc->base, + DPU_ENC_RC_EVENT_ENTER_IDLE); + + dpu_encoder_frame_done_callback(&dpu_enc->base, NULL, + DPU_ENCODER_FRAME_EVENT_IDLE); +} + +/** + * _dpu_encoder_trigger_flush - trigger flush for a physical encoder + * drm_enc: Pointer to drm encoder structure + * phys: Pointer to physical encoder structure + * extra_flush_bits: Additional bit mask to include in flush trigger + */ +static inline void _dpu_encoder_trigger_flush(struct drm_encoder *drm_enc, + struct dpu_encoder_phys *phys, uint32_t extra_flush_bits) +{ + struct dpu_hw_ctl *ctl; + int pending_kickoff_cnt; + u32 ret = UINT_MAX; + + if (!drm_enc || !phys) { + DPU_ERROR("invalid argument(s), drm_enc %d, phys_enc %d\n", + drm_enc != 0, phys != 0); + return; + } + + if (!phys->hw_pp) { + DPU_ERROR("invalid pingpong hw\n"); + return; + } + + ctl = phys->hw_ctl; + if (!ctl || !ctl->ops.trigger_flush) { + DPU_ERROR("missing trigger cb\n"); + return; + } + + pending_kickoff_cnt = dpu_encoder_phys_inc_pending(phys); + + if (extra_flush_bits && ctl->ops.update_pending_flush) + ctl->ops.update_pending_flush(ctl, extra_flush_bits); + + ctl->ops.trigger_flush(ctl); + + if (ctl->ops.get_pending_flush) + ret = ctl->ops.get_pending_flush(ctl); + + trace_dpu_enc_trigger_flush(DRMID(drm_enc), phys->intf_idx, + pending_kickoff_cnt, ctl->idx, ret); +} + +/** + * _dpu_encoder_trigger_start - trigger start for a physical encoder + * phys: Pointer to physical encoder structure + */ +static inline void _dpu_encoder_trigger_start(struct dpu_encoder_phys *phys) +{ + if (!phys) { + DPU_ERROR("invalid argument(s)\n"); + return; + } + + if (!phys->hw_pp) { + DPU_ERROR("invalid pingpong hw\n"); + return; + } + + if (phys->ops.trigger_start && phys->enable_state != DPU_ENC_DISABLED) + phys->ops.trigger_start(phys); +} + +void dpu_encoder_helper_trigger_start(struct dpu_encoder_phys *phys_enc) +{ + struct dpu_hw_ctl *ctl; + + if (!phys_enc) { + DPU_ERROR("invalid encoder\n"); + return; + } + + ctl = phys_enc->hw_ctl; + if (ctl && ctl->ops.trigger_start) { + ctl->ops.trigger_start(ctl); + trace_dpu_enc_trigger_start(DRMID(phys_enc->parent), ctl->idx); + } +} + +static int dpu_encoder_helper_wait_event_timeout( + int32_t drm_id, + int32_t hw_id, + struct dpu_encoder_wait_info *info) +{ + int rc = 0; + s64 expected_time = ktime_to_ms(ktime_get()) + info->timeout_ms; + s64 jiffies = msecs_to_jiffies(info->timeout_ms); + s64 time; + + do { + rc = wait_event_timeout(*(info->wq), + atomic_read(info->atomic_cnt) == 0, jiffies); + time = ktime_to_ms(ktime_get()); + + trace_dpu_enc_wait_event_timeout(drm_id, hw_id, rc, time, + expected_time, + atomic_read(info->atomic_cnt)); + /* If we timed out, counter is valid and time is less, wait again */ + } while (atomic_read(info->atomic_cnt) && (rc == 0) && + (time < expected_time)); + + return rc; +} + +void dpu_encoder_helper_hw_reset(struct dpu_encoder_phys *phys_enc) +{ + struct dpu_encoder_virt *dpu_enc; + struct dpu_hw_ctl *ctl; + int rc; + + if (!phys_enc) { + DPU_ERROR("invalid encoder\n"); + return; + } + dpu_enc = to_dpu_encoder_virt(phys_enc->parent); + ctl = phys_enc->hw_ctl; + + if (!ctl || !ctl->ops.reset) + return; + + DRM_DEBUG_KMS("id:%u ctl %d reset\n", DRMID(phys_enc->parent), + ctl->idx); + + rc = ctl->ops.reset(ctl); + if (rc) { + DPU_ERROR_ENC(dpu_enc, "ctl %d reset failure\n", ctl->idx); + dpu_dbg_dump(false, __func__, true, true); + } + + phys_enc->enable_state = DPU_ENC_ENABLED; +} + +/** + * _dpu_encoder_kickoff_phys - handle physical encoder kickoff + * Iterate through the physical encoders and perform consolidated flush + * and/or control start triggering as needed. This is done in the virtual + * encoder rather than the individual physical ones in order to handle + * use cases that require visibility into multiple physical encoders at + * a time. + * dpu_enc: Pointer to virtual encoder structure + */ +static void _dpu_encoder_kickoff_phys(struct dpu_encoder_virt *dpu_enc) +{ + struct dpu_hw_ctl *ctl; + uint32_t i, pending_flush; + unsigned long lock_flags; + + if (!dpu_enc) { + DPU_ERROR("invalid encoder\n"); + return; + } + + pending_flush = 0x0; + + /* update pending counts and trigger kickoff ctl flush atomically */ + spin_lock_irqsave(&dpu_enc->enc_spinlock, lock_flags); + + /* don't perform flush/start operations for slave encoders */ + for (i = 0; i < dpu_enc->num_phys_encs; i++) { + struct dpu_encoder_phys *phys = dpu_enc->phys_encs[i]; + + if (!phys || phys->enable_state == DPU_ENC_DISABLED) + continue; + + ctl = phys->hw_ctl; + if (!ctl) + continue; + + if (phys->split_role != ENC_ROLE_SLAVE) + set_bit(i, dpu_enc->frame_busy_mask); + if (!phys->ops.needs_single_flush || + !phys->ops.needs_single_flush(phys)) + _dpu_encoder_trigger_flush(&dpu_enc->base, phys, 0x0); + else if (ctl->ops.get_pending_flush) + pending_flush |= ctl->ops.get_pending_flush(ctl); + } + + /* for split flush, combine pending flush masks and send to master */ + if (pending_flush && dpu_enc->cur_master) { + _dpu_encoder_trigger_flush( + &dpu_enc->base, + dpu_enc->cur_master, + pending_flush); + } + + _dpu_encoder_trigger_start(dpu_enc->cur_master); + + spin_unlock_irqrestore(&dpu_enc->enc_spinlock, lock_flags); +} + +void dpu_encoder_trigger_kickoff_pending(struct drm_encoder *drm_enc) +{ + struct dpu_encoder_virt *dpu_enc; + struct dpu_encoder_phys *phys; + unsigned int i; + struct dpu_hw_ctl *ctl; + struct msm_display_info *disp_info; + + if (!drm_enc) { + DPU_ERROR("invalid encoder\n"); + return; + } + dpu_enc = to_dpu_encoder_virt(drm_enc); + disp_info = &dpu_enc->disp_info; + + for (i = 0; i < dpu_enc->num_phys_encs; i++) { + phys = dpu_enc->phys_encs[i]; + + if (phys && phys->hw_ctl) { + ctl = phys->hw_ctl; + if (ctl->ops.clear_pending_flush) + ctl->ops.clear_pending_flush(ctl); + + /* update only for command mode primary ctl */ + if ((phys == dpu_enc->cur_master) && + (disp_info->capabilities & MSM_DISPLAY_CAP_CMD_MODE) + && ctl->ops.trigger_pending) + ctl->ops.trigger_pending(ctl); + } + } +} + +static u32 _dpu_encoder_calculate_linetime(struct dpu_encoder_virt *dpu_enc, + struct drm_display_mode *mode) +{ + u64 pclk_rate; + u32 pclk_period; + u32 line_time; + + /* + * For linetime calculation, only operate on master encoder. + */ + if (!dpu_enc->cur_master) + return 0; + + if (!dpu_enc->cur_master->ops.get_line_count) { + DPU_ERROR("get_line_count function not defined\n"); + return 0; + } + + pclk_rate = mode->clock; /* pixel clock in kHz */ + if (pclk_rate == 0) { + DPU_ERROR("pclk is 0, cannot calculate line time\n"); + return 0; + } + + pclk_period = DIV_ROUND_UP_ULL(1000000000ull, pclk_rate); + if (pclk_period == 0) { + DPU_ERROR("pclk period is 0\n"); + return 0; + } + + /* + * Line time calculation based on Pixel clock and HTOTAL. + * Final unit is in ns. + */ + line_time = (pclk_period * mode->htotal) / 1000; + if (line_time == 0) { + DPU_ERROR("line time calculation is 0\n"); + return 0; + } + + DPU_DEBUG_ENC(dpu_enc, + "clk_rate=%lldkHz, clk_period=%d, linetime=%dns\n", + pclk_rate, pclk_period, line_time); + + return line_time; +} + +static int _dpu_encoder_wakeup_time(struct drm_encoder *drm_enc, + ktime_t *wakeup_time) +{ + struct drm_display_mode *mode; + struct dpu_encoder_virt *dpu_enc; + u32 cur_line; + u32 line_time; + u32 vtotal, time_to_vsync; + ktime_t cur_time; + + dpu_enc = to_dpu_encoder_virt(drm_enc); + + if (!drm_enc->crtc || !drm_enc->crtc->state) { + DPU_ERROR("crtc/crtc state object is NULL\n"); + return -EINVAL; + } + mode = &drm_enc->crtc->state->adjusted_mode; + + line_time = _dpu_encoder_calculate_linetime(dpu_enc, mode); + if (!line_time) + return -EINVAL; + + cur_line = dpu_enc->cur_master->ops.get_line_count(dpu_enc->cur_master); + + vtotal = mode->vtotal; + if (cur_line >= vtotal) + time_to_vsync = line_time * vtotal; + else + time_to_vsync = line_time * (vtotal - cur_line); + + if (time_to_vsync == 0) { + DPU_ERROR("time to vsync should not be zero, vtotal=%d\n", + vtotal); + return -EINVAL; + } + + cur_time = ktime_get(); + *wakeup_time = ktime_add_ns(cur_time, time_to_vsync); + + DPU_DEBUG_ENC(dpu_enc, + "cur_line=%u vtotal=%u time_to_vsync=%u, cur_time=%lld, wakeup_time=%lld\n", + cur_line, vtotal, time_to_vsync, + ktime_to_ms(cur_time), + ktime_to_ms(*wakeup_time)); + return 0; +} + +static void dpu_encoder_vsync_event_handler(struct timer_list *t) +{ + struct dpu_encoder_virt *dpu_enc = from_timer(dpu_enc, t, + vsync_event_timer); + struct drm_encoder *drm_enc = &dpu_enc->base; + struct msm_drm_private *priv; + struct msm_drm_thread *event_thread; + + if (!drm_enc->dev || !drm_enc->dev->dev_private || + !drm_enc->crtc) { + DPU_ERROR("invalid parameters\n"); + return; + } + + priv = drm_enc->dev->dev_private; + + if (drm_enc->crtc->index >= ARRAY_SIZE(priv->event_thread)) { + DPU_ERROR("invalid crtc index\n"); + return; + } + event_thread = &priv->event_thread[drm_enc->crtc->index]; + if (!event_thread) { + DPU_ERROR("event_thread not found for crtc:%d\n", + drm_enc->crtc->index); + return; + } + + del_timer(&dpu_enc->vsync_event_timer); +} + +static void dpu_encoder_vsync_event_work_handler(struct kthread_work *work) +{ + struct dpu_encoder_virt *dpu_enc = container_of(work, + struct dpu_encoder_virt, vsync_event_work); + ktime_t wakeup_time; + + if (!dpu_enc) { + DPU_ERROR("invalid dpu encoder\n"); + return; + } + + if (_dpu_encoder_wakeup_time(&dpu_enc->base, &wakeup_time)) + return; + + trace_dpu_enc_vsync_event_work(DRMID(&dpu_enc->base), wakeup_time); + mod_timer(&dpu_enc->vsync_event_timer, + nsecs_to_jiffies(ktime_to_ns(wakeup_time))); +} + +void dpu_encoder_prepare_for_kickoff(struct drm_encoder *drm_enc, + struct dpu_encoder_kickoff_params *params) +{ + struct dpu_encoder_virt *dpu_enc; + struct dpu_encoder_phys *phys; + bool needs_hw_reset = false; + unsigned int i; + + if (!drm_enc || !params) { + DPU_ERROR("invalid args\n"); + return; + } + dpu_enc = to_dpu_encoder_virt(drm_enc); + + trace_dpu_enc_prepare_kickoff(DRMID(drm_enc)); + + /* prepare for next kickoff, may include waiting on previous kickoff */ + DPU_ATRACE_BEGIN("enc_prepare_for_kickoff"); + for (i = 0; i < dpu_enc->num_phys_encs; i++) { + phys = dpu_enc->phys_encs[i]; + if (phys) { + if (phys->ops.prepare_for_kickoff) + phys->ops.prepare_for_kickoff(phys, params); + if (phys->enable_state == DPU_ENC_ERR_NEEDS_HW_RESET) + needs_hw_reset = true; + } + } + DPU_ATRACE_END("enc_prepare_for_kickoff"); + + dpu_encoder_resource_control(drm_enc, DPU_ENC_RC_EVENT_KICKOFF); + + /* if any phys needs reset, reset all phys, in-order */ + if (needs_hw_reset) { + trace_dpu_enc_prepare_kickoff_reset(DRMID(drm_enc)); + for (i = 0; i < dpu_enc->num_phys_encs; i++) { + phys = dpu_enc->phys_encs[i]; + if (phys && phys->ops.hw_reset) + phys->ops.hw_reset(phys); + } + } +} + +void dpu_encoder_kickoff(struct drm_encoder *drm_enc) +{ + struct dpu_encoder_virt *dpu_enc; + struct dpu_encoder_phys *phys; + ktime_t wakeup_time; + unsigned int i; + + if (!drm_enc) { + DPU_ERROR("invalid encoder\n"); + return; + } + DPU_ATRACE_BEGIN("encoder_kickoff"); + dpu_enc = to_dpu_encoder_virt(drm_enc); + + trace_dpu_enc_kickoff(DRMID(drm_enc)); + + atomic_set(&dpu_enc->frame_done_timeout, + DPU_FRAME_DONE_TIMEOUT * 1000 / + drm_enc->crtc->state->adjusted_mode.vrefresh); + mod_timer(&dpu_enc->frame_done_timer, jiffies + + ((atomic_read(&dpu_enc->frame_done_timeout) * HZ) / 1000)); + + /* All phys encs are ready to go, trigger the kickoff */ + _dpu_encoder_kickoff_phys(dpu_enc); + + /* allow phys encs to handle any post-kickoff business */ + for (i = 0; i < dpu_enc->num_phys_encs; i++) { + phys = dpu_enc->phys_encs[i]; + if (phys && phys->ops.handle_post_kickoff) + phys->ops.handle_post_kickoff(phys); + } + + if (dpu_enc->disp_info.intf_type == DRM_MODE_CONNECTOR_DSI && + !_dpu_encoder_wakeup_time(drm_enc, &wakeup_time)) { + trace_dpu_enc_early_kickoff(DRMID(drm_enc), + ktime_to_ms(wakeup_time)); + mod_timer(&dpu_enc->vsync_event_timer, + nsecs_to_jiffies(ktime_to_ns(wakeup_time))); + } + + DPU_ATRACE_END("encoder_kickoff"); +} + +void dpu_encoder_prepare_commit(struct drm_encoder *drm_enc) +{ + struct dpu_encoder_virt *dpu_enc; + struct dpu_encoder_phys *phys; + int i; + + if (!drm_enc) { + DPU_ERROR("invalid encoder\n"); + return; + } + dpu_enc = to_dpu_encoder_virt(drm_enc); + + for (i = 0; i < dpu_enc->num_phys_encs; i++) { + phys = dpu_enc->phys_encs[i]; + if (phys && phys->ops.prepare_commit) + phys->ops.prepare_commit(phys); + } +} + +#ifdef CONFIG_DEBUG_FS +static int _dpu_encoder_status_show(struct seq_file *s, void *data) +{ + struct dpu_encoder_virt *dpu_enc; + int i; + + if (!s || !s->private) + return -EINVAL; + + dpu_enc = s->private; + + mutex_lock(&dpu_enc->enc_lock); + for (i = 0; i < dpu_enc->num_phys_encs; i++) { + struct dpu_encoder_phys *phys = dpu_enc->phys_encs[i]; + + if (!phys) + continue; + + seq_printf(s, "intf:%d vsync:%8d underrun:%8d ", + phys->intf_idx - INTF_0, + atomic_read(&phys->vsync_cnt), + atomic_read(&phys->underrun_cnt)); + + switch (phys->intf_mode) { + case INTF_MODE_VIDEO: + seq_puts(s, "mode: video\n"); + break; + case INTF_MODE_CMD: + seq_puts(s, "mode: command\n"); + break; + default: + seq_puts(s, "mode: ???\n"); + break; + } + } + mutex_unlock(&dpu_enc->enc_lock); + + return 0; +} + +static int _dpu_encoder_debugfs_status_open(struct inode *inode, + struct file *file) +{ + return single_open(file, _dpu_encoder_status_show, inode->i_private); +} + +static ssize_t _dpu_encoder_misr_setup(struct file *file, + const char __user *user_buf, size_t count, loff_t *ppos) +{ + struct dpu_encoder_virt *dpu_enc; + int i = 0, rc; + char buf[MISR_BUFF_SIZE + 1]; + size_t buff_copy; + u32 frame_count, enable; + + if (!file || !file->private_data) + return -EINVAL; + + dpu_enc = file->private_data; + + buff_copy = min_t(size_t, count, MISR_BUFF_SIZE); + if (copy_from_user(buf, user_buf, buff_copy)) + return -EINVAL; + + buf[buff_copy] = 0; /* end of string */ + + if (sscanf(buf, "%u %u", &enable, &frame_count) != 2) + return -EINVAL; + + rc = _dpu_encoder_power_enable(dpu_enc, true); + if (rc) + return rc; + + mutex_lock(&dpu_enc->enc_lock); + dpu_enc->misr_enable = enable; + dpu_enc->misr_frame_count = frame_count; + for (i = 0; i < dpu_enc->num_phys_encs; i++) { + struct dpu_encoder_phys *phys = dpu_enc->phys_encs[i]; + + if (!phys || !phys->ops.setup_misr) + continue; + + phys->ops.setup_misr(phys, enable, frame_count); + } + mutex_unlock(&dpu_enc->enc_lock); + _dpu_encoder_power_enable(dpu_enc, false); + + return count; +} + +static ssize_t _dpu_encoder_misr_read(struct file *file, + char __user *user_buff, size_t count, loff_t *ppos) +{ + struct dpu_encoder_virt *dpu_enc; + int i = 0, len = 0; + char buf[MISR_BUFF_SIZE + 1] = {'\0'}; + int rc; + + if (*ppos) + return 0; + + if (!file || !file->private_data) + return -EINVAL; + + dpu_enc = file->private_data; + + rc = _dpu_encoder_power_enable(dpu_enc, true); + if (rc) + return rc; + + mutex_lock(&dpu_enc->enc_lock); + if (!dpu_enc->misr_enable) { + len += snprintf(buf + len, MISR_BUFF_SIZE - len, + "disabled\n"); + goto buff_check; + } else if (dpu_enc->disp_info.capabilities & + ~MSM_DISPLAY_CAP_VID_MODE) { + len += snprintf(buf + len, MISR_BUFF_SIZE - len, + "unsupported\n"); + goto buff_check; + } + + for (i = 0; i < dpu_enc->num_phys_encs; i++) { + struct dpu_encoder_phys *phys = dpu_enc->phys_encs[i]; + + if (!phys || !phys->ops.collect_misr) + continue; + + len += snprintf(buf + len, MISR_BUFF_SIZE - len, + "Intf idx:%d\n", phys->intf_idx - INTF_0); + len += snprintf(buf + len, MISR_BUFF_SIZE - len, "0x%x\n", + phys->ops.collect_misr(phys)); + } + +buff_check: + if (count <= len) { + len = 0; + goto end; + } + + if (copy_to_user(user_buff, buf, len)) { + len = -EFAULT; + goto end; + } + + *ppos += len; /* increase offset */ + +end: + mutex_unlock(&dpu_enc->enc_lock); + _dpu_encoder_power_enable(dpu_enc, false); + return len; +} + +static int _dpu_encoder_init_debugfs(struct drm_encoder *drm_enc) +{ + struct dpu_encoder_virt *dpu_enc; + struct msm_drm_private *priv; + struct dpu_kms *dpu_kms; + int i; + + static const struct file_operations debugfs_status_fops = { + .open = _dpu_encoder_debugfs_status_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, + }; + + static const struct file_operations debugfs_misr_fops = { + .open = simple_open, + .read = _dpu_encoder_misr_read, + .write = _dpu_encoder_misr_setup, + }; + + char name[DPU_NAME_SIZE]; + + if (!drm_enc || !drm_enc->dev || !drm_enc->dev->dev_private) { + DPU_ERROR("invalid encoder or kms\n"); + return -EINVAL; + } + + dpu_enc = to_dpu_encoder_virt(drm_enc); + priv = drm_enc->dev->dev_private; + dpu_kms = to_dpu_kms(priv->kms); + + snprintf(name, DPU_NAME_SIZE, "encoder%u", drm_enc->base.id); + + /* create overall sub-directory for the encoder */ + dpu_enc->debugfs_root = debugfs_create_dir(name, + drm_enc->dev->primary->debugfs_root); + if (!dpu_enc->debugfs_root) + return -ENOMEM; + + /* don't error check these */ + debugfs_create_file("status", 0600, + dpu_enc->debugfs_root, dpu_enc, &debugfs_status_fops); + + debugfs_create_file("misr_data", 0600, + dpu_enc->debugfs_root, dpu_enc, &debugfs_misr_fops); + + for (i = 0; i < dpu_enc->num_phys_encs; i++) + if (dpu_enc->phys_encs[i] && + dpu_enc->phys_encs[i]->ops.late_register) + dpu_enc->phys_encs[i]->ops.late_register( + dpu_enc->phys_encs[i], + dpu_enc->debugfs_root); + + return 0; +} + +static void _dpu_encoder_destroy_debugfs(struct drm_encoder *drm_enc) +{ + struct dpu_encoder_virt *dpu_enc; + + if (!drm_enc) + return; + + dpu_enc = to_dpu_encoder_virt(drm_enc); + debugfs_remove_recursive(dpu_enc->debugfs_root); +} +#else +static int _dpu_encoder_init_debugfs(struct drm_encoder *drm_enc) +{ + return 0; +} + +static void _dpu_encoder_destroy_debugfs(struct drm_encoder *drm_enc) +{ +} +#endif + +static int dpu_encoder_late_register(struct drm_encoder *encoder) +{ + return _dpu_encoder_init_debugfs(encoder); +} + +static void dpu_encoder_early_unregister(struct drm_encoder *encoder) +{ + _dpu_encoder_destroy_debugfs(encoder); +} + +static int dpu_encoder_virt_add_phys_encs( + u32 display_caps, + struct dpu_encoder_virt *dpu_enc, + struct dpu_enc_phys_init_params *params) +{ + struct dpu_encoder_phys *enc = NULL; + + DPU_DEBUG_ENC(dpu_enc, "\n"); + + /* + * We may create up to NUM_PHYS_ENCODER_TYPES physical encoder types + * in this function, check up-front. + */ + if (dpu_enc->num_phys_encs + NUM_PHYS_ENCODER_TYPES >= + ARRAY_SIZE(dpu_enc->phys_encs)) { + DPU_ERROR_ENC(dpu_enc, "too many physical encoders %d\n", + dpu_enc->num_phys_encs); + return -EINVAL; + } + + if (display_caps & MSM_DISPLAY_CAP_VID_MODE) { + enc = dpu_encoder_phys_vid_init(params); + + if (IS_ERR_OR_NULL(enc)) { + DPU_ERROR_ENC(dpu_enc, "failed to init vid enc: %ld\n", + PTR_ERR(enc)); + return enc == 0 ? -EINVAL : PTR_ERR(enc); + } + + dpu_enc->phys_encs[dpu_enc->num_phys_encs] = enc; + ++dpu_enc->num_phys_encs; + } + + if (display_caps & MSM_DISPLAY_CAP_CMD_MODE) { + enc = dpu_encoder_phys_cmd_init(params); + + if (IS_ERR_OR_NULL(enc)) { + DPU_ERROR_ENC(dpu_enc, "failed to init cmd enc: %ld\n", + PTR_ERR(enc)); + return enc == 0 ? -EINVAL : PTR_ERR(enc); + } + + dpu_enc->phys_encs[dpu_enc->num_phys_encs] = enc; + ++dpu_enc->num_phys_encs; + } + + return 0; +} + +static const struct dpu_encoder_virt_ops dpu_encoder_parent_ops = { + .handle_vblank_virt = dpu_encoder_vblank_callback, + .handle_underrun_virt = dpu_encoder_underrun_callback, + .handle_frame_done = dpu_encoder_frame_done_callback, +}; + +static int dpu_encoder_setup_display(struct dpu_encoder_virt *dpu_enc, + struct dpu_kms *dpu_kms, + struct msm_display_info *disp_info, + int *drm_enc_mode) +{ + int ret = 0; + int i = 0; + enum dpu_intf_type intf_type; + struct dpu_enc_phys_init_params phys_params; + + if (!dpu_enc || !dpu_kms) { + DPU_ERROR("invalid arg(s), enc %d kms %d\n", + dpu_enc != 0, dpu_kms != 0); + return -EINVAL; + } + + memset(&phys_params, 0, sizeof(phys_params)); + phys_params.dpu_kms = dpu_kms; + phys_params.parent = &dpu_enc->base; + phys_params.parent_ops = &dpu_encoder_parent_ops; + phys_params.enc_spinlock = &dpu_enc->enc_spinlock; + + DPU_DEBUG("\n"); + + if (disp_info->intf_type == DRM_MODE_CONNECTOR_DSI) { + *drm_enc_mode = DRM_MODE_ENCODER_DSI; + intf_type = INTF_DSI; + } else if (disp_info->intf_type == DRM_MODE_CONNECTOR_HDMIA) { + *drm_enc_mode = DRM_MODE_ENCODER_TMDS; + intf_type = INTF_HDMI; + } else if (disp_info->intf_type == DRM_MODE_CONNECTOR_DisplayPort) { + *drm_enc_mode = DRM_MODE_ENCODER_TMDS; + intf_type = INTF_DP; + } else { + DPU_ERROR_ENC(dpu_enc, "unsupported display interface type\n"); + return -EINVAL; + } + + WARN_ON(disp_info->num_of_h_tiles < 1); + + dpu_enc->display_num_of_h_tiles = disp_info->num_of_h_tiles; + + DPU_DEBUG("dsi_info->num_of_h_tiles %d\n", disp_info->num_of_h_tiles); + + if ((disp_info->capabilities & MSM_DISPLAY_CAP_CMD_MODE) || + (disp_info->capabilities & MSM_DISPLAY_CAP_VID_MODE)) + dpu_enc->idle_pc_supported = + dpu_kms->catalog->caps->has_idle_pc; + + mutex_lock(&dpu_enc->enc_lock); + for (i = 0; i < disp_info->num_of_h_tiles && !ret; i++) { + /* + * Left-most tile is at index 0, content is controller id + * h_tile_instance_ids[2] = {0, 1}; DSI0 = left, DSI1 = right + * h_tile_instance_ids[2] = {1, 0}; DSI1 = left, DSI0 = right + */ + u32 controller_id = disp_info->h_tile_instance[i]; + + if (disp_info->num_of_h_tiles > 1) { + if (i == 0) + phys_params.split_role = ENC_ROLE_MASTER; + else + phys_params.split_role = ENC_ROLE_SLAVE; + } else { + phys_params.split_role = ENC_ROLE_SOLO; + } + + DPU_DEBUG("h_tile_instance %d = %d, split_role %d\n", + i, controller_id, phys_params.split_role); + + phys_params.intf_idx = dpu_encoder_get_intf(dpu_kms->catalog, + intf_type, + controller_id); + if (phys_params.intf_idx == INTF_MAX) { + DPU_ERROR_ENC(dpu_enc, "could not get intf: type %d, id %d\n", + intf_type, controller_id); + ret = -EINVAL; + } + + if (!ret) { + ret = dpu_encoder_virt_add_phys_encs(disp_info->capabilities, + dpu_enc, + &phys_params); + if (ret) + DPU_ERROR_ENC(dpu_enc, "failed to add phys encs\n"); + } + } + + for (i = 0; i < dpu_enc->num_phys_encs; i++) { + struct dpu_encoder_phys *phys = dpu_enc->phys_encs[i]; + + if (phys) { + atomic_set(&phys->vsync_cnt, 0); + atomic_set(&phys->underrun_cnt, 0); + } + } + mutex_unlock(&dpu_enc->enc_lock); + + return ret; +} + +static void dpu_encoder_frame_done_timeout(struct timer_list *t) +{ + struct dpu_encoder_virt *dpu_enc = from_timer(dpu_enc, t, + frame_done_timer); + struct drm_encoder *drm_enc = &dpu_enc->base; + struct msm_drm_private *priv; + u32 event; + + if (!drm_enc->dev || !drm_enc->dev->dev_private) { + DPU_ERROR("invalid parameters\n"); + return; + } + priv = drm_enc->dev->dev_private; + + if (!dpu_enc->frame_busy_mask[0] || !dpu_enc->crtc_frame_event_cb) { + DRM_DEBUG_KMS("id:%u invalid timeout frame_busy_mask=%lu\n", + DRMID(drm_enc), dpu_enc->frame_busy_mask[0]); + return; + } else if (!atomic_xchg(&dpu_enc->frame_done_timeout, 0)) { + DRM_DEBUG_KMS("id:%u invalid timeout\n", DRMID(drm_enc)); + return; + } + + DPU_ERROR_ENC(dpu_enc, "frame done timeout\n"); + + event = DPU_ENCODER_FRAME_EVENT_ERROR; + trace_dpu_enc_frame_done_timeout(DRMID(drm_enc), event); + dpu_enc->crtc_frame_event_cb(dpu_enc->crtc_frame_event_cb_data, event); +} + +static const struct drm_encoder_helper_funcs dpu_encoder_helper_funcs = { + .mode_set = dpu_encoder_virt_mode_set, + .disable = dpu_encoder_virt_disable, + .enable = dpu_kms_encoder_enable, + .atomic_check = dpu_encoder_virt_atomic_check, + + /* This is called by dpu_kms_encoder_enable */ + .commit = dpu_encoder_virt_enable, +}; + +static const struct drm_encoder_funcs dpu_encoder_funcs = { + .destroy = dpu_encoder_destroy, + .late_register = dpu_encoder_late_register, + .early_unregister = dpu_encoder_early_unregister, +}; + +int dpu_encoder_setup(struct drm_device *dev, struct drm_encoder *enc, + struct msm_display_info *disp_info) +{ + struct msm_drm_private *priv = dev->dev_private; + struct dpu_kms *dpu_kms = to_dpu_kms(priv->kms); + struct drm_encoder *drm_enc = NULL; + struct dpu_encoder_virt *dpu_enc = NULL; + int drm_enc_mode = DRM_MODE_ENCODER_NONE; + int ret = 0; + + dpu_enc = to_dpu_encoder_virt(enc); + + mutex_init(&dpu_enc->enc_lock); + ret = dpu_encoder_setup_display(dpu_enc, dpu_kms, disp_info, + &drm_enc_mode); + if (ret) + goto fail; + + dpu_enc->cur_master = NULL; + spin_lock_init(&dpu_enc->enc_spinlock); + + atomic_set(&dpu_enc->frame_done_timeout, 0); + timer_setup(&dpu_enc->frame_done_timer, + dpu_encoder_frame_done_timeout, 0); + + if (disp_info->intf_type == DRM_MODE_CONNECTOR_DSI) + timer_setup(&dpu_enc->vsync_event_timer, + dpu_encoder_vsync_event_handler, + 0); + + + mutex_init(&dpu_enc->rc_lock); + kthread_init_delayed_work(&dpu_enc->delayed_off_work, + dpu_encoder_off_work); + dpu_enc->idle_timeout = IDLE_TIMEOUT; + + kthread_init_work(&dpu_enc->vsync_event_work, + dpu_encoder_vsync_event_work_handler); + + memcpy(&dpu_enc->disp_info, disp_info, sizeof(*disp_info)); + + DPU_DEBUG_ENC(dpu_enc, "created\n"); + + return ret; + +fail: + DPU_ERROR("failed to create encoder\n"); + if (drm_enc) + dpu_encoder_destroy(drm_enc); + + return ret; + + +} + +struct drm_encoder *dpu_encoder_init(struct drm_device *dev, + int drm_enc_mode) +{ + struct dpu_encoder_virt *dpu_enc = NULL; + int rc = 0; + + dpu_enc = devm_kzalloc(dev->dev, sizeof(*dpu_enc), GFP_KERNEL); + if (!dpu_enc) + return ERR_PTR(ENOMEM); + + rc = drm_encoder_init(dev, &dpu_enc->base, &dpu_encoder_funcs, + drm_enc_mode, NULL); + if (rc) { + devm_kfree(dev->dev, dpu_enc); + return ERR_PTR(rc); + } + + drm_encoder_helper_add(&dpu_enc->base, &dpu_encoder_helper_funcs); + + return &dpu_enc->base; +} + +int dpu_encoder_wait_for_event(struct drm_encoder *drm_enc, + enum msm_event_wait event) +{ + int (*fn_wait)(struct dpu_encoder_phys *phys_enc) = NULL; + struct dpu_encoder_virt *dpu_enc = NULL; + int i, ret = 0; + + if (!drm_enc) { + DPU_ERROR("invalid encoder\n"); + return -EINVAL; + } + dpu_enc = to_dpu_encoder_virt(drm_enc); + DPU_DEBUG_ENC(dpu_enc, "\n"); + + for (i = 0; i < dpu_enc->num_phys_encs; i++) { + struct dpu_encoder_phys *phys = dpu_enc->phys_encs[i]; + + switch (event) { + case MSM_ENC_COMMIT_DONE: + fn_wait = phys->ops.wait_for_commit_done; + break; + case MSM_ENC_TX_COMPLETE: + fn_wait = phys->ops.wait_for_tx_complete; + break; + case MSM_ENC_VBLANK: + fn_wait = phys->ops.wait_for_vblank; + break; + default: + DPU_ERROR_ENC(dpu_enc, "unknown wait event %d\n", + event); + return -EINVAL; + }; + + if (phys && fn_wait) { + DPU_ATRACE_BEGIN("wait_for_completion_event"); + ret = fn_wait(phys); + DPU_ATRACE_END("wait_for_completion_event"); + if (ret) + return ret; + } + } + + return ret; +} + +enum dpu_intf_mode dpu_encoder_get_intf_mode(struct drm_encoder *encoder) +{ + struct dpu_encoder_virt *dpu_enc = NULL; + int i; + + if (!encoder) { + DPU_ERROR("invalid encoder\n"); + return INTF_MODE_NONE; + } + dpu_enc = to_dpu_encoder_virt(encoder); + + if (dpu_enc->cur_master) + return dpu_enc->cur_master->intf_mode; + + for (i = 0; i < dpu_enc->num_phys_encs; i++) { + struct dpu_encoder_phys *phys = dpu_enc->phys_encs[i]; + + if (phys) + return phys->intf_mode; + } + + return INTF_MODE_NONE; +} diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.h b/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.h new file mode 100644 index 000000000000..60f809fc7c13 --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.h @@ -0,0 +1,177 @@ +/* + * Copyright (c) 2015-2018, The Linux Foundation. All rights reserved. + * Copyright (C) 2013 Red Hat + * Author: Rob Clark <robdclark@gmail.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published by + * the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program. If not, see <http://www.gnu.org/licenses/>. + */ + +#ifndef __DPU_ENCODER_H__ +#define __DPU_ENCODER_H__ + +#include <drm/drm_crtc.h> +#include "dpu_hw_mdss.h" + +#define DPU_ENCODER_FRAME_EVENT_DONE BIT(0) +#define DPU_ENCODER_FRAME_EVENT_ERROR BIT(1) +#define DPU_ENCODER_FRAME_EVENT_PANEL_DEAD BIT(2) +#define DPU_ENCODER_FRAME_EVENT_IDLE BIT(3) + +#define IDLE_TIMEOUT (66 - 16/2) + +/** + * Encoder functions and data types + * @intfs: Interfaces this encoder is using, INTF_MODE_NONE if unused + * @needs_cdm: Encoder requests a CDM based on pixel format conversion needs + * @display_num_of_h_tiles: Number of horizontal tiles in case of split + * interface + * @topology: Topology of the display + */ +struct dpu_encoder_hw_resources { + enum dpu_intf_mode intfs[INTF_MAX]; + bool needs_cdm; + u32 display_num_of_h_tiles; +}; + +/** + * dpu_encoder_kickoff_params - info encoder requires at kickoff + * @affected_displays: bitmask, bit set means the ROI of the commit lies within + * the bounds of the physical display at the bit index + */ +struct dpu_encoder_kickoff_params { + unsigned long affected_displays; +}; + +/** + * dpu_encoder_get_hw_resources - Populate table of required hardware resources + * @encoder: encoder pointer + * @hw_res: resource table to populate with encoder required resources + * @conn_state: report hw reqs based on this proposed connector state + */ +void dpu_encoder_get_hw_resources(struct drm_encoder *encoder, + struct dpu_encoder_hw_resources *hw_res, + struct drm_connector_state *conn_state); + +/** + * dpu_encoder_register_vblank_callback - provide callback to encoder that + * will be called on the next vblank. + * @encoder: encoder pointer + * @cb: callback pointer, provide NULL to deregister and disable IRQs + * @data: user data provided to callback + */ +void dpu_encoder_register_vblank_callback(struct drm_encoder *encoder, + void (*cb)(void *), void *data); + +/** + * dpu_encoder_register_frame_event_callback - provide callback to encoder that + * will be called after the request is complete, or other events. + * @encoder: encoder pointer + * @cb: callback pointer, provide NULL to deregister + * @data: user data provided to callback + */ +void dpu_encoder_register_frame_event_callback(struct drm_encoder *encoder, + void (*cb)(void *, u32), void *data); + +/** + * dpu_encoder_prepare_for_kickoff - schedule double buffer flip of the ctl + * path (i.e. ctl flush and start) at next appropriate time. + * Immediately: if no previous commit is outstanding. + * Delayed: Block until next trigger can be issued. + * @encoder: encoder pointer + * @params: kickoff time parameters + */ +void dpu_encoder_prepare_for_kickoff(struct drm_encoder *encoder, + struct dpu_encoder_kickoff_params *params); + +/** + * dpu_encoder_trigger_kickoff_pending - Clear the flush bits from previous + * kickoff and trigger the ctl prepare progress for command mode display. + * @encoder: encoder pointer + */ +void dpu_encoder_trigger_kickoff_pending(struct drm_encoder *encoder); + +/** + * dpu_encoder_kickoff - trigger a double buffer flip of the ctl path + * (i.e. ctl flush and start) immediately. + * @encoder: encoder pointer + */ +void dpu_encoder_kickoff(struct drm_encoder *encoder); + +/** + * dpu_encoder_wait_for_event - Waits for encoder events + * @encoder: encoder pointer + * @event: event to wait for + * MSM_ENC_COMMIT_DONE - Wait for hardware to have flushed the current pending + * frames to hardware at a vblank or ctl_start + * Encoders will map this differently depending on the + * panel type. + * vid mode -> vsync_irq + * cmd mode -> ctl_start + * MSM_ENC_TX_COMPLETE - Wait for the hardware to transfer all the pixels to + * the panel. Encoders will map this differently + * depending on the panel type. + * vid mode -> vsync_irq + * cmd mode -> pp_done + * Returns: 0 on success, -EWOULDBLOCK if already signaled, error otherwise + */ +int dpu_encoder_wait_for_event(struct drm_encoder *drm_encoder, + enum msm_event_wait event); + +/* + * dpu_encoder_get_intf_mode - get interface mode of the given encoder + * @encoder: Pointer to drm encoder object + */ +enum dpu_intf_mode dpu_encoder_get_intf_mode(struct drm_encoder *encoder); + +/** + * dpu_encoder_virt_restore - restore the encoder configs + * @encoder: encoder pointer + */ +void dpu_encoder_virt_restore(struct drm_encoder *encoder); + +/** + * dpu_encoder_init - initialize virtual encoder object + * @dev: Pointer to drm device structure + * @disp_info: Pointer to display information structure + * Returns: Pointer to newly created drm encoder + */ +struct drm_encoder *dpu_encoder_init( + struct drm_device *dev, + int drm_enc_mode); + +/** + * dpu_encoder_setup - setup dpu_encoder for the display probed + * @dev: Pointer to drm device structure + * @enc: Pointer to the drm_encoder + * @disp_info: Pointer to the display info + */ +int dpu_encoder_setup(struct drm_device *dev, struct drm_encoder *enc, + struct msm_display_info *disp_info); + +/** + * dpu_encoder_prepare_commit - prepare encoder at the very beginning of an + * atomic commit, before any registers are written + * @drm_enc: Pointer to previously created drm encoder structure + */ +void dpu_encoder_prepare_commit(struct drm_encoder *drm_enc); + +/** + * dpu_encoder_set_idle_timeout - set the idle timeout for video + * and command mode encoders. + * @drm_enc: Pointer to previously created drm encoder structure + * @idle_timeout: idle timeout duration in milliseconds + */ +void dpu_encoder_set_idle_timeout(struct drm_encoder *drm_enc, + u32 idle_timeout); + +#endif /* __DPU_ENCODER_H__ */ diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder_phys.h b/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder_phys.h new file mode 100644 index 000000000000..c7df8aad6613 --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder_phys.h @@ -0,0 +1,430 @@ +/* + * Copyright (c) 2015-2018 The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + */ + +#ifndef __DPU_ENCODER_PHYS_H__ +#define __DPU_ENCODER_PHYS_H__ + +#include <linux/jiffies.h> + +#include "dpu_kms.h" +#include "dpu_hw_intf.h" +#include "dpu_hw_pingpong.h" +#include "dpu_hw_ctl.h" +#include "dpu_hw_top.h" +#include "dpu_hw_cdm.h" +#include "dpu_encoder.h" + +#define DPU_ENCODER_NAME_MAX 16 + +/* wait for at most 2 vsync for lowest refresh rate (24hz) */ +#define KICKOFF_TIMEOUT_MS 84 +#define KICKOFF_TIMEOUT_JIFFIES msecs_to_jiffies(KICKOFF_TIMEOUT_MS) + +/** + * enum dpu_enc_split_role - Role this physical encoder will play in a + * split-panel configuration, where one panel is master, and others slaves. + * Masters have extra responsibilities, like managing the VBLANK IRQ. + * @ENC_ROLE_SOLO: This is the one and only panel. This encoder is master. + * @ENC_ROLE_MASTER: This encoder is the master of a split panel config. + * @ENC_ROLE_SLAVE: This encoder is not the master of a split panel config. + */ +enum dpu_enc_split_role { + ENC_ROLE_SOLO, + ENC_ROLE_MASTER, + ENC_ROLE_SLAVE, +}; + +/** + * enum dpu_enc_enable_state - current enabled state of the physical encoder + * @DPU_ENC_DISABLING: Encoder transitioning to disable state + * Events bounding transition are encoder type specific + * @DPU_ENC_DISABLED: Encoder is disabled + * @DPU_ENC_ENABLING: Encoder transitioning to enabled + * Events bounding transition are encoder type specific + * @DPU_ENC_ENABLED: Encoder is enabled + * @DPU_ENC_ERR_NEEDS_HW_RESET: Encoder is enabled, but requires a hw_reset + * to recover from a previous error + */ +enum dpu_enc_enable_state { + DPU_ENC_DISABLING, + DPU_ENC_DISABLED, + DPU_ENC_ENABLING, + DPU_ENC_ENABLED, + DPU_ENC_ERR_NEEDS_HW_RESET +}; + +struct dpu_encoder_phys; + +/** + * struct dpu_encoder_virt_ops - Interface the containing virtual encoder + * provides for the physical encoders to use to callback. + * @handle_vblank_virt: Notify virtual encoder of vblank IRQ reception + * Note: This is called from IRQ handler context. + * @handle_underrun_virt: Notify virtual encoder of underrun IRQ reception + * Note: This is called from IRQ handler context. + * @handle_frame_done: Notify virtual encoder that this phys encoder + * completes last request frame. + */ +struct dpu_encoder_virt_ops { + void (*handle_vblank_virt)(struct drm_encoder *, + struct dpu_encoder_phys *phys); + void (*handle_underrun_virt)(struct drm_encoder *, + struct dpu_encoder_phys *phys); + void (*handle_frame_done)(struct drm_encoder *, + struct dpu_encoder_phys *phys, u32 event); +}; + +/** + * struct dpu_encoder_phys_ops - Interface the physical encoders provide to + * the containing virtual encoder. + * @late_register: DRM Call. Add Userspace interfaces, debugfs. + * @prepare_commit: MSM Atomic Call, start of atomic commit sequence + * @is_master: Whether this phys_enc is the current master + * encoder. Can be switched at enable time. Based + * on split_role and current mode (CMD/VID). + * @mode_fixup: DRM Call. Fixup a DRM mode. + * @mode_set: DRM Call. Set a DRM mode. + * This likely caches the mode, for use at enable. + * @enable: DRM Call. Enable a DRM mode. + * @disable: DRM Call. Disable mode. + * @atomic_check: DRM Call. Atomic check new DRM state. + * @destroy: DRM Call. Destroy and release resources. + * @get_hw_resources: Populate the structure with the hardware + * resources that this phys_enc is using. + * Expect no overlap between phys_encs. + * @control_vblank_irq Register/Deregister for VBLANK IRQ + * @wait_for_commit_done: Wait for hardware to have flushed the + * current pending frames to hardware + * @wait_for_tx_complete: Wait for hardware to transfer the pixels + * to the panel + * @wait_for_vblank: Wait for VBLANK, for sub-driver internal use + * @prepare_for_kickoff: Do any work necessary prior to a kickoff + * For CMD encoder, may wait for previous tx done + * @handle_post_kickoff: Do any work necessary post-kickoff work + * @trigger_start: Process start event on physical encoder + * @needs_single_flush: Whether encoder slaves need to be flushed + * @setup_misr: Sets up MISR, enable and disables based on sysfs + * @collect_misr: Collects MISR data on frame update + * @hw_reset: Issue HW recovery such as CTL reset and clear + * DPU_ENC_ERR_NEEDS_HW_RESET state + * @irq_control: Handler to enable/disable all the encoder IRQs + * @prepare_idle_pc: phys encoder can update the vsync_enable status + * on idle power collapse prepare + * @restore: Restore all the encoder configs. + * @get_line_count: Obtain current vertical line count + */ + +struct dpu_encoder_phys_ops { + int (*late_register)(struct dpu_encoder_phys *encoder, + struct dentry *debugfs_root); + void (*prepare_commit)(struct dpu_encoder_phys *encoder); + bool (*is_master)(struct dpu_encoder_phys *encoder); + bool (*mode_fixup)(struct dpu_encoder_phys *encoder, + const struct drm_display_mode *mode, + struct drm_display_mode *adjusted_mode); + void (*mode_set)(struct dpu_encoder_phys *encoder, + struct drm_display_mode *mode, + struct drm_display_mode *adjusted_mode); + void (*enable)(struct dpu_encoder_phys *encoder); + void (*disable)(struct dpu_encoder_phys *encoder); + int (*atomic_check)(struct dpu_encoder_phys *encoder, + struct drm_crtc_state *crtc_state, + struct drm_connector_state *conn_state); + void (*destroy)(struct dpu_encoder_phys *encoder); + void (*get_hw_resources)(struct dpu_encoder_phys *encoder, + struct dpu_encoder_hw_resources *hw_res, + struct drm_connector_state *conn_state); + int (*control_vblank_irq)(struct dpu_encoder_phys *enc, bool enable); + int (*wait_for_commit_done)(struct dpu_encoder_phys *phys_enc); + int (*wait_for_tx_complete)(struct dpu_encoder_phys *phys_enc); + int (*wait_for_vblank)(struct dpu_encoder_phys *phys_enc); + void (*prepare_for_kickoff)(struct dpu_encoder_phys *phys_enc, + struct dpu_encoder_kickoff_params *params); + void (*handle_post_kickoff)(struct dpu_encoder_phys *phys_enc); + void (*trigger_start)(struct dpu_encoder_phys *phys_enc); + bool (*needs_single_flush)(struct dpu_encoder_phys *phys_enc); + + void (*setup_misr)(struct dpu_encoder_phys *phys_encs, + bool enable, u32 frame_count); + u32 (*collect_misr)(struct dpu_encoder_phys *phys_enc); + void (*hw_reset)(struct dpu_encoder_phys *phys_enc); + void (*irq_control)(struct dpu_encoder_phys *phys, bool enable); + void (*prepare_idle_pc)(struct dpu_encoder_phys *phys_enc); + void (*restore)(struct dpu_encoder_phys *phys); + int (*get_line_count)(struct dpu_encoder_phys *phys); +}; + +/** + * enum dpu_intr_idx - dpu encoder interrupt index + * @INTR_IDX_VSYNC: Vsync interrupt for video mode panel + * @INTR_IDX_PINGPONG: Pingpong done unterrupt for cmd mode panel + * @INTR_IDX_UNDERRUN: Underrun unterrupt for video and cmd mode panel + * @INTR_IDX_RDPTR: Readpointer done unterrupt for cmd mode panel + */ +enum dpu_intr_idx { + INTR_IDX_VSYNC, + INTR_IDX_PINGPONG, + INTR_IDX_UNDERRUN, + INTR_IDX_CTL_START, + INTR_IDX_RDPTR, + INTR_IDX_MAX, +}; + +/** + * dpu_encoder_irq - tracking structure for interrupts + * @name: string name of interrupt + * @intr_type: Encoder interrupt type + * @intr_idx: Encoder interrupt enumeration + * @hw_idx: HW Block ID + * @irq_idx: IRQ interface lookup index from DPU IRQ framework + * will be -EINVAL if IRQ is not registered + * @irq_cb: interrupt callback + */ +struct dpu_encoder_irq { + const char *name; + enum dpu_intr_type intr_type; + enum dpu_intr_idx intr_idx; + int hw_idx; + int irq_idx; + struct dpu_irq_callback cb; +}; + +/** + * struct dpu_encoder_phys - physical encoder that drives a single INTF block + * tied to a specific panel / sub-panel. Abstract type, sub-classed by + * phys_vid or phys_cmd for video mode or command mode encs respectively. + * @parent: Pointer to the containing virtual encoder + * @connector: If a mode is set, cached pointer to the active connector + * @ops: Operations exposed to the virtual encoder + * @parent_ops: Callbacks exposed by the parent to the phys_enc + * @hw_mdptop: Hardware interface to the top registers + * @hw_ctl: Hardware interface to the ctl registers + * @hw_cdm: Hardware interface to the cdm registers + * @cdm_cfg: Chroma-down hardware configuration + * @hw_pp: Hardware interface to the ping pong registers + * @dpu_kms: Pointer to the dpu_kms top level + * @cached_mode: DRM mode cached at mode_set time, acted on in enable + * @enabled: Whether the encoder has enabled and running a mode + * @split_role: Role to play in a split-panel configuration + * @intf_mode: Interface mode + * @intf_idx: Interface index on dpu hardware + * @topology_name: topology selected for the display + * @enc_spinlock: Virtual-Encoder-Wide Spin Lock for IRQ purposes + * @enable_state: Enable state tracking + * @vblank_refcount: Reference count of vblank request + * @vsync_cnt: Vsync count for the physical encoder + * @underrun_cnt: Underrun count for the physical encoder + * @pending_kickoff_cnt: Atomic counter tracking the number of kickoffs + * vs. the number of done/vblank irqs. Should hover + * between 0-2 Incremented when a new kickoff is + * scheduled. Decremented in irq handler + * @pending_ctlstart_cnt: Atomic counter tracking the number of ctl start + * pending. + * @pending_kickoff_wq: Wait queue for blocking until kickoff completes + * @irq: IRQ tracking structures + */ +struct dpu_encoder_phys { + struct drm_encoder *parent; + struct drm_connector *connector; + struct dpu_encoder_phys_ops ops; + const struct dpu_encoder_virt_ops *parent_ops; + struct dpu_hw_mdp *hw_mdptop; + struct dpu_hw_ctl *hw_ctl; + struct dpu_hw_cdm *hw_cdm; + struct dpu_hw_cdm_cfg cdm_cfg; + struct dpu_hw_pingpong *hw_pp; + struct dpu_kms *dpu_kms; + struct drm_display_mode cached_mode; + enum dpu_enc_split_role split_role; + enum dpu_intf_mode intf_mode; + enum dpu_intf intf_idx; + enum dpu_rm_topology_name topology_name; + spinlock_t *enc_spinlock; + enum dpu_enc_enable_state enable_state; + atomic_t vblank_refcount; + atomic_t vsync_cnt; + atomic_t underrun_cnt; + atomic_t pending_ctlstart_cnt; + atomic_t pending_kickoff_cnt; + wait_queue_head_t pending_kickoff_wq; + struct dpu_encoder_irq irq[INTR_IDX_MAX]; +}; + +static inline int dpu_encoder_phys_inc_pending(struct dpu_encoder_phys *phys) +{ + atomic_inc_return(&phys->pending_ctlstart_cnt); + return atomic_inc_return(&phys->pending_kickoff_cnt); +} + +/** + * struct dpu_encoder_phys_vid - sub-class of dpu_encoder_phys to handle video + * mode specific operations + * @base: Baseclass physical encoder structure + * @hw_intf: Hardware interface to the intf registers + * @timing_params: Current timing parameter + */ +struct dpu_encoder_phys_vid { + struct dpu_encoder_phys base; + struct dpu_hw_intf *hw_intf; + struct intf_timing_params timing_params; +}; + +/** + * struct dpu_encoder_phys_cmd - sub-class of dpu_encoder_phys to handle command + * mode specific operations + * @base: Baseclass physical encoder structure + * @intf_idx: Intf Block index used by this phys encoder + * @stream_sel: Stream selection for multi-stream interfaces + * @serialize_wait4pp: serialize wait4pp feature waits for pp_done interrupt + * after ctl_start instead of before next frame kickoff + * @pp_timeout_report_cnt: number of pingpong done irq timeout errors + * @pending_vblank_cnt: Atomic counter tracking pending wait for VBLANK + * @pending_vblank_wq: Wait queue for blocking until VBLANK received + */ +struct dpu_encoder_phys_cmd { + struct dpu_encoder_phys base; + int stream_sel; + bool serialize_wait4pp; + int pp_timeout_report_cnt; + atomic_t pending_vblank_cnt; + wait_queue_head_t pending_vblank_wq; +}; + +/** + * struct dpu_enc_phys_init_params - initialization parameters for phys encs + * @dpu_kms: Pointer to the dpu_kms top level + * @parent: Pointer to the containing virtual encoder + * @parent_ops: Callbacks exposed by the parent to the phys_enc + * @split_role: Role to play in a split-panel configuration + * @intf_idx: Interface index this phys_enc will control + * @enc_spinlock: Virtual-Encoder-Wide Spin Lock for IRQ purposes + */ +struct dpu_enc_phys_init_params { + struct dpu_kms *dpu_kms; + struct drm_encoder *parent; + const struct dpu_encoder_virt_ops *parent_ops; + enum dpu_enc_split_role split_role; + enum dpu_intf intf_idx; + spinlock_t *enc_spinlock; +}; + +/** + * dpu_encoder_wait_info - container for passing arguments to irq wait functions + * @wq: wait queue structure + * @atomic_cnt: wait until atomic_cnt equals zero + * @timeout_ms: timeout value in milliseconds + */ +struct dpu_encoder_wait_info { + wait_queue_head_t *wq; + atomic_t *atomic_cnt; + s64 timeout_ms; +}; + +/** + * dpu_encoder_phys_vid_init - Construct a new video mode physical encoder + * @p: Pointer to init params structure + * Return: Error code or newly allocated encoder + */ +struct dpu_encoder_phys *dpu_encoder_phys_vid_init( + struct dpu_enc_phys_init_params *p); + +/** + * dpu_encoder_phys_cmd_init - Construct a new command mode physical encoder + * @p: Pointer to init params structure + * Return: Error code or newly allocated encoder + */ +struct dpu_encoder_phys *dpu_encoder_phys_cmd_init( + struct dpu_enc_phys_init_params *p); + +/** + * dpu_encoder_helper_trigger_start - control start helper function + * This helper function may be optionally specified by physical + * encoders if they require ctl_start triggering. + * @phys_enc: Pointer to physical encoder structure + */ +void dpu_encoder_helper_trigger_start(struct dpu_encoder_phys *phys_enc); + +/** + * dpu_encoder_helper_hw_reset - issue ctl hw reset + * This helper function may be optionally specified by physical + * encoders if they require ctl hw reset. If state is currently + * DPU_ENC_ERR_NEEDS_HW_RESET, it is set back to DPU_ENC_ENABLED. + * @phys_enc: Pointer to physical encoder structure + */ +void dpu_encoder_helper_hw_reset(struct dpu_encoder_phys *phys_enc); + +static inline enum dpu_3d_blend_mode dpu_encoder_helper_get_3d_blend_mode( + struct dpu_encoder_phys *phys_enc) +{ + if (!phys_enc || phys_enc->enable_state == DPU_ENC_DISABLING) + return BLEND_3D_NONE; + + if (phys_enc->split_role == ENC_ROLE_SOLO && + phys_enc->topology_name == DPU_RM_TOPOLOGY_DUALPIPE_3DMERGE) + return BLEND_3D_H_ROW_INT; + + return BLEND_3D_NONE; +} + +/** + * dpu_encoder_helper_split_config - split display configuration helper function + * This helper function may be used by physical encoders to configure + * the split display related registers. + * @phys_enc: Pointer to physical encoder structure + * @interface: enum dpu_intf setting + */ +void dpu_encoder_helper_split_config( + struct dpu_encoder_phys *phys_enc, + enum dpu_intf interface); + +/** + * dpu_encoder_helper_report_irq_timeout - utility to report error that irq has + * timed out, including reporting frame error event to crtc and debug dump + * @phys_enc: Pointer to physical encoder structure + * @intr_idx: Failing interrupt index + */ +void dpu_encoder_helper_report_irq_timeout(struct dpu_encoder_phys *phys_enc, + enum dpu_intr_idx intr_idx); + +/** + * dpu_encoder_helper_wait_for_irq - utility to wait on an irq. + * note: will call dpu_encoder_helper_wait_for_irq on timeout + * @phys_enc: Pointer to physical encoder structure + * @intr_idx: encoder interrupt index + * @wait_info: wait info struct + * @Return: 0 or -ERROR + */ +int dpu_encoder_helper_wait_for_irq(struct dpu_encoder_phys *phys_enc, + enum dpu_intr_idx intr_idx, + struct dpu_encoder_wait_info *wait_info); + +/** + * dpu_encoder_helper_register_irq - register and enable an irq + * @phys_enc: Pointer to physical encoder structure + * @intr_idx: encoder interrupt index + * @Return: 0 or -ERROR + */ +int dpu_encoder_helper_register_irq(struct dpu_encoder_phys *phys_enc, + enum dpu_intr_idx intr_idx); + +/** + * dpu_encoder_helper_unregister_irq - unregister and disable an irq + * @phys_enc: Pointer to physical encoder structure + * @intr_idx: encoder interrupt index + * @Return: 0 or -ERROR + */ +int dpu_encoder_helper_unregister_irq(struct dpu_encoder_phys *phys_enc, + enum dpu_intr_idx intr_idx); + +#endif /* __dpu_encoder_phys_H__ */ diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder_phys_cmd.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder_phys_cmd.c new file mode 100644 index 000000000000..3084675ed425 --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder_phys_cmd.c @@ -0,0 +1,905 @@ +/* + * Copyright (c) 2015-2018 The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + */ + +#define pr_fmt(fmt) "[drm:%s:%d] " fmt, __func__, __LINE__ +#include "dpu_encoder_phys.h" +#include "dpu_hw_interrupts.h" +#include "dpu_core_irq.h" +#include "dpu_formats.h" +#include "dpu_trace.h" + +#define DPU_DEBUG_CMDENC(e, fmt, ...) DPU_DEBUG("enc%d intf%d " fmt, \ + (e) && (e)->base.parent ? \ + (e)->base.parent->base.id : -1, \ + (e) ? (e)->base.intf_idx - INTF_0 : -1, ##__VA_ARGS__) + +#define DPU_ERROR_CMDENC(e, fmt, ...) DPU_ERROR("enc%d intf%d " fmt, \ + (e) && (e)->base.parent ? \ + (e)->base.parent->base.id : -1, \ + (e) ? (e)->base.intf_idx - INTF_0 : -1, ##__VA_ARGS__) + +#define to_dpu_encoder_phys_cmd(x) \ + container_of(x, struct dpu_encoder_phys_cmd, base) + +#define PP_TIMEOUT_MAX_TRIALS 10 + +/* + * Tearcheck sync start and continue thresholds are empirically found + * based on common panels In the future, may want to allow panels to override + * these default values + */ +#define DEFAULT_TEARCHECK_SYNC_THRESH_START 4 +#define DEFAULT_TEARCHECK_SYNC_THRESH_CONTINUE 4 + +#define DPU_ENC_WR_PTR_START_TIMEOUT_US 20000 + +static inline int _dpu_encoder_phys_cmd_get_idle_timeout( + struct dpu_encoder_phys_cmd *cmd_enc) +{ + return KICKOFF_TIMEOUT_MS; +} + +static inline bool dpu_encoder_phys_cmd_is_master( + struct dpu_encoder_phys *phys_enc) +{ + return (phys_enc->split_role != ENC_ROLE_SLAVE) ? true : false; +} + +static bool dpu_encoder_phys_cmd_mode_fixup( + struct dpu_encoder_phys *phys_enc, + const struct drm_display_mode *mode, + struct drm_display_mode *adj_mode) +{ + if (phys_enc) + DPU_DEBUG_CMDENC(to_dpu_encoder_phys_cmd(phys_enc), "\n"); + return true; +} + +static void _dpu_encoder_phys_cmd_update_intf_cfg( + struct dpu_encoder_phys *phys_enc) +{ + struct dpu_encoder_phys_cmd *cmd_enc = + to_dpu_encoder_phys_cmd(phys_enc); + struct dpu_hw_ctl *ctl; + struct dpu_hw_intf_cfg intf_cfg = { 0 }; + + if (!phys_enc) + return; + + ctl = phys_enc->hw_ctl; + if (!ctl || !ctl->ops.setup_intf_cfg) + return; + + intf_cfg.intf = phys_enc->intf_idx; + intf_cfg.intf_mode_sel = DPU_CTL_MODE_SEL_CMD; + intf_cfg.stream_sel = cmd_enc->stream_sel; + intf_cfg.mode_3d = dpu_encoder_helper_get_3d_blend_mode(phys_enc); + ctl->ops.setup_intf_cfg(ctl, &intf_cfg); +} + +static void dpu_encoder_phys_cmd_pp_tx_done_irq(void *arg, int irq_idx) +{ + struct dpu_encoder_phys *phys_enc = arg; + unsigned long lock_flags; + int new_cnt; + u32 event = DPU_ENCODER_FRAME_EVENT_DONE; + + if (!phys_enc || !phys_enc->hw_pp) + return; + + DPU_ATRACE_BEGIN("pp_done_irq"); + /* notify all synchronous clients first, then asynchronous clients */ + if (phys_enc->parent_ops->handle_frame_done) + phys_enc->parent_ops->handle_frame_done(phys_enc->parent, + phys_enc, event); + + spin_lock_irqsave(phys_enc->enc_spinlock, lock_flags); + new_cnt = atomic_add_unless(&phys_enc->pending_kickoff_cnt, -1, 0); + spin_unlock_irqrestore(phys_enc->enc_spinlock, lock_flags); + + trace_dpu_enc_phys_cmd_pp_tx_done(DRMID(phys_enc->parent), + phys_enc->hw_pp->idx - PINGPONG_0, + new_cnt, event); + + /* Signal any waiting atomic commit thread */ + wake_up_all(&phys_enc->pending_kickoff_wq); + DPU_ATRACE_END("pp_done_irq"); +} + +static void dpu_encoder_phys_cmd_pp_rd_ptr_irq(void *arg, int irq_idx) +{ + struct dpu_encoder_phys *phys_enc = arg; + struct dpu_encoder_phys_cmd *cmd_enc; + + if (!phys_enc || !phys_enc->hw_pp) + return; + + DPU_ATRACE_BEGIN("rd_ptr_irq"); + cmd_enc = to_dpu_encoder_phys_cmd(phys_enc); + + if (phys_enc->parent_ops->handle_vblank_virt) + phys_enc->parent_ops->handle_vblank_virt(phys_enc->parent, + phys_enc); + + atomic_add_unless(&cmd_enc->pending_vblank_cnt, -1, 0); + wake_up_all(&cmd_enc->pending_vblank_wq); + DPU_ATRACE_END("rd_ptr_irq"); +} + +static void dpu_encoder_phys_cmd_ctl_start_irq(void *arg, int irq_idx) +{ + struct dpu_encoder_phys *phys_enc = arg; + struct dpu_encoder_phys_cmd *cmd_enc; + + if (!phys_enc || !phys_enc->hw_ctl) + return; + + DPU_ATRACE_BEGIN("ctl_start_irq"); + cmd_enc = to_dpu_encoder_phys_cmd(phys_enc); + + atomic_add_unless(&phys_enc->pending_ctlstart_cnt, -1, 0); + + /* Signal any waiting ctl start interrupt */ + wake_up_all(&phys_enc->pending_kickoff_wq); + DPU_ATRACE_END("ctl_start_irq"); +} + +static void dpu_encoder_phys_cmd_underrun_irq(void *arg, int irq_idx) +{ + struct dpu_encoder_phys *phys_enc = arg; + + if (!phys_enc) + return; + + if (phys_enc->parent_ops->handle_underrun_virt) + phys_enc->parent_ops->handle_underrun_virt(phys_enc->parent, + phys_enc); +} + +static void _dpu_encoder_phys_cmd_setup_irq_hw_idx( + struct dpu_encoder_phys *phys_enc) +{ + struct dpu_encoder_irq *irq; + + irq = &phys_enc->irq[INTR_IDX_CTL_START]; + irq->hw_idx = phys_enc->hw_ctl->idx; + irq->irq_idx = -EINVAL; + + irq = &phys_enc->irq[INTR_IDX_PINGPONG]; + irq->hw_idx = phys_enc->hw_pp->idx; + irq->irq_idx = -EINVAL; + + irq = &phys_enc->irq[INTR_IDX_RDPTR]; + irq->hw_idx = phys_enc->hw_pp->idx; + irq->irq_idx = -EINVAL; + + irq = &phys_enc->irq[INTR_IDX_UNDERRUN]; + irq->hw_idx = phys_enc->intf_idx; + irq->irq_idx = -EINVAL; +} + +static void dpu_encoder_phys_cmd_mode_set( + struct dpu_encoder_phys *phys_enc, + struct drm_display_mode *mode, + struct drm_display_mode *adj_mode) +{ + struct dpu_encoder_phys_cmd *cmd_enc = + to_dpu_encoder_phys_cmd(phys_enc); + struct dpu_rm *rm = &phys_enc->dpu_kms->rm; + struct dpu_rm_hw_iter iter; + int i, instance; + + if (!phys_enc || !mode || !adj_mode) { + DPU_ERROR("invalid args\n"); + return; + } + phys_enc->cached_mode = *adj_mode; + DPU_DEBUG_CMDENC(cmd_enc, "caching mode:\n"); + drm_mode_debug_printmodeline(adj_mode); + + instance = phys_enc->split_role == ENC_ROLE_SLAVE ? 1 : 0; + + /* Retrieve previously allocated HW Resources. Shouldn't fail */ + dpu_rm_init_hw_iter(&iter, phys_enc->parent->base.id, DPU_HW_BLK_CTL); + for (i = 0; i <= instance; i++) { + if (dpu_rm_get_hw(rm, &iter)) + phys_enc->hw_ctl = (struct dpu_hw_ctl *)iter.hw; + } + + if (IS_ERR_OR_NULL(phys_enc->hw_ctl)) { + DPU_ERROR_CMDENC(cmd_enc, "failed to init ctl: %ld\n", + PTR_ERR(phys_enc->hw_ctl)); + phys_enc->hw_ctl = NULL; + return; + } + + _dpu_encoder_phys_cmd_setup_irq_hw_idx(phys_enc); +} + +static int _dpu_encoder_phys_cmd_handle_ppdone_timeout( + struct dpu_encoder_phys *phys_enc) +{ + struct dpu_encoder_phys_cmd *cmd_enc = + to_dpu_encoder_phys_cmd(phys_enc); + u32 frame_event = DPU_ENCODER_FRAME_EVENT_ERROR; + bool do_log = false; + + if (!phys_enc || !phys_enc->hw_pp || !phys_enc->hw_ctl) + return -EINVAL; + + cmd_enc->pp_timeout_report_cnt++; + if (cmd_enc->pp_timeout_report_cnt == PP_TIMEOUT_MAX_TRIALS) { + frame_event |= DPU_ENCODER_FRAME_EVENT_PANEL_DEAD; + do_log = true; + } else if (cmd_enc->pp_timeout_report_cnt == 1) { + do_log = true; + } + + trace_dpu_enc_phys_cmd_pdone_timeout(DRMID(phys_enc->parent), + phys_enc->hw_pp->idx - PINGPONG_0, + cmd_enc->pp_timeout_report_cnt, + atomic_read(&phys_enc->pending_kickoff_cnt), + frame_event); + + /* to avoid flooding, only log first time, and "dead" time */ + if (do_log) { + DRM_ERROR("id:%d pp:%d kickoff timeout %d cnt %d koff_cnt %d\n", + DRMID(phys_enc->parent), + phys_enc->hw_pp->idx - PINGPONG_0, + phys_enc->hw_ctl->idx - CTL_0, + cmd_enc->pp_timeout_report_cnt, + atomic_read(&phys_enc->pending_kickoff_cnt)); + + dpu_encoder_helper_unregister_irq(phys_enc, INTR_IDX_RDPTR); + dpu_dbg_dump(false, __func__, true, true); + } + + atomic_add_unless(&phys_enc->pending_kickoff_cnt, -1, 0); + + /* request a ctl reset before the next kickoff */ + phys_enc->enable_state = DPU_ENC_ERR_NEEDS_HW_RESET; + + if (phys_enc->parent_ops->handle_frame_done) + phys_enc->parent_ops->handle_frame_done( + phys_enc->parent, phys_enc, frame_event); + + return -ETIMEDOUT; +} + +static int _dpu_encoder_phys_cmd_wait_for_idle( + struct dpu_encoder_phys *phys_enc) +{ + struct dpu_encoder_phys_cmd *cmd_enc = + to_dpu_encoder_phys_cmd(phys_enc); + struct dpu_encoder_wait_info wait_info; + int ret; + + if (!phys_enc) { + DPU_ERROR("invalid encoder\n"); + return -EINVAL; + } + + wait_info.wq = &phys_enc->pending_kickoff_wq; + wait_info.atomic_cnt = &phys_enc->pending_kickoff_cnt; + wait_info.timeout_ms = KICKOFF_TIMEOUT_MS; + + ret = dpu_encoder_helper_wait_for_irq(phys_enc, INTR_IDX_PINGPONG, + &wait_info); + if (ret == -ETIMEDOUT) + _dpu_encoder_phys_cmd_handle_ppdone_timeout(phys_enc); + else if (!ret) + cmd_enc->pp_timeout_report_cnt = 0; + + return ret; +} + +static int dpu_encoder_phys_cmd_control_vblank_irq( + struct dpu_encoder_phys *phys_enc, + bool enable) +{ + int ret = 0; + int refcount; + + if (!phys_enc || !phys_enc->hw_pp) { + DPU_ERROR("invalid encoder\n"); + return -EINVAL; + } + + refcount = atomic_read(&phys_enc->vblank_refcount); + + /* Slave encoders don't report vblank */ + if (!dpu_encoder_phys_cmd_is_master(phys_enc)) + goto end; + + /* protect against negative */ + if (!enable && refcount == 0) { + ret = -EINVAL; + goto end; + } + + DRM_DEBUG_KMS("id:%u pp:%d enable=%s/%d\n", DRMID(phys_enc->parent), + phys_enc->hw_pp->idx - PINGPONG_0, + enable ? "true" : "false", refcount); + + if (enable && atomic_inc_return(&phys_enc->vblank_refcount) == 1) + ret = dpu_encoder_helper_register_irq(phys_enc, INTR_IDX_RDPTR); + else if (!enable && atomic_dec_return(&phys_enc->vblank_refcount) == 0) + ret = dpu_encoder_helper_unregister_irq(phys_enc, + INTR_IDX_RDPTR); + +end: + if (ret) { + DRM_ERROR("vblank irq err id:%u pp:%d ret:%d, enable %s/%d\n", + DRMID(phys_enc->parent), + phys_enc->hw_pp->idx - PINGPONG_0, ret, + enable ? "true" : "false", refcount); + } + + return ret; +} + +static void dpu_encoder_phys_cmd_irq_control(struct dpu_encoder_phys *phys_enc, + bool enable) +{ + struct dpu_encoder_phys_cmd *cmd_enc; + + if (!phys_enc) + return; + + cmd_enc = to_dpu_encoder_phys_cmd(phys_enc); + + trace_dpu_enc_phys_cmd_irq_ctrl(DRMID(phys_enc->parent), + phys_enc->hw_pp->idx - PINGPONG_0, + enable, atomic_read(&phys_enc->vblank_refcount)); + + if (enable) { + dpu_encoder_helper_register_irq(phys_enc, INTR_IDX_PINGPONG); + dpu_encoder_helper_register_irq(phys_enc, INTR_IDX_UNDERRUN); + dpu_encoder_phys_cmd_control_vblank_irq(phys_enc, true); + + if (dpu_encoder_phys_cmd_is_master(phys_enc)) + dpu_encoder_helper_register_irq(phys_enc, + INTR_IDX_CTL_START); + } else { + if (dpu_encoder_phys_cmd_is_master(phys_enc)) + dpu_encoder_helper_unregister_irq(phys_enc, + INTR_IDX_CTL_START); + + dpu_encoder_helper_unregister_irq(phys_enc, INTR_IDX_UNDERRUN); + dpu_encoder_phys_cmd_control_vblank_irq(phys_enc, false); + dpu_encoder_helper_unregister_irq(phys_enc, INTR_IDX_PINGPONG); + } +} + +static void dpu_encoder_phys_cmd_tearcheck_config( + struct dpu_encoder_phys *phys_enc) +{ + struct dpu_encoder_phys_cmd *cmd_enc = + to_dpu_encoder_phys_cmd(phys_enc); + struct dpu_hw_tear_check tc_cfg = { 0 }; + struct drm_display_mode *mode; + bool tc_enable = true; + u32 vsync_hz; + struct msm_drm_private *priv; + struct dpu_kms *dpu_kms; + + if (!phys_enc || !phys_enc->hw_pp) { + DPU_ERROR("invalid encoder\n"); + return; + } + mode = &phys_enc->cached_mode; + + DPU_DEBUG_CMDENC(cmd_enc, "pp %d\n", phys_enc->hw_pp->idx - PINGPONG_0); + + if (!phys_enc->hw_pp->ops.setup_tearcheck || + !phys_enc->hw_pp->ops.enable_tearcheck) { + DPU_DEBUG_CMDENC(cmd_enc, "tearcheck not supported\n"); + return; + } + + dpu_kms = phys_enc->dpu_kms; + if (!dpu_kms || !dpu_kms->dev || !dpu_kms->dev->dev_private) { + DPU_ERROR("invalid device\n"); + return; + } + priv = dpu_kms->dev->dev_private; + + /* + * TE default: dsi byte clock calculated base on 70 fps; + * around 14 ms to complete a kickoff cycle if te disabled; + * vclk_line base on 60 fps; write is faster than read; + * init == start == rdptr; + * + * vsync_count is ratio of MDP VSYNC clock frequency to LCD panel + * frequency divided by the no. of rows (lines) in the LCDpanel. + */ + vsync_hz = dpu_kms_get_clk_rate(dpu_kms, "vsync"); + if (vsync_hz <= 0) { + DPU_DEBUG_CMDENC(cmd_enc, "invalid - vsync_hz %u\n", + vsync_hz); + return; + } + + tc_cfg.vsync_count = vsync_hz / (mode->vtotal * mode->vrefresh); + + /* enable external TE after kickoff to avoid premature autorefresh */ + tc_cfg.hw_vsync_mode = 0; + + /* + * By setting sync_cfg_height to near max register value, we essentially + * disable dpu hw generated TE signal, since hw TE will arrive first. + * Only caveat is if due to error, we hit wrap-around. + */ + tc_cfg.sync_cfg_height = 0xFFF0; + tc_cfg.vsync_init_val = mode->vdisplay; + tc_cfg.sync_threshold_start = DEFAULT_TEARCHECK_SYNC_THRESH_START; + tc_cfg.sync_threshold_continue = DEFAULT_TEARCHECK_SYNC_THRESH_CONTINUE; + tc_cfg.start_pos = mode->vdisplay; + tc_cfg.rd_ptr_irq = mode->vdisplay + 1; + + DPU_DEBUG_CMDENC(cmd_enc, + "tc %d vsync_clk_speed_hz %u vtotal %u vrefresh %u\n", + phys_enc->hw_pp->idx - PINGPONG_0, vsync_hz, + mode->vtotal, mode->vrefresh); + DPU_DEBUG_CMDENC(cmd_enc, + "tc %d enable %u start_pos %u rd_ptr_irq %u\n", + phys_enc->hw_pp->idx - PINGPONG_0, tc_enable, tc_cfg.start_pos, + tc_cfg.rd_ptr_irq); + DPU_DEBUG_CMDENC(cmd_enc, + "tc %d hw_vsync_mode %u vsync_count %u vsync_init_val %u\n", + phys_enc->hw_pp->idx - PINGPONG_0, tc_cfg.hw_vsync_mode, + tc_cfg.vsync_count, tc_cfg.vsync_init_val); + DPU_DEBUG_CMDENC(cmd_enc, + "tc %d cfgheight %u thresh_start %u thresh_cont %u\n", + phys_enc->hw_pp->idx - PINGPONG_0, tc_cfg.sync_cfg_height, + tc_cfg.sync_threshold_start, tc_cfg.sync_threshold_continue); + + phys_enc->hw_pp->ops.setup_tearcheck(phys_enc->hw_pp, &tc_cfg); + phys_enc->hw_pp->ops.enable_tearcheck(phys_enc->hw_pp, tc_enable); +} + +static void _dpu_encoder_phys_cmd_pingpong_config( + struct dpu_encoder_phys *phys_enc) +{ + struct dpu_encoder_phys_cmd *cmd_enc = + to_dpu_encoder_phys_cmd(phys_enc); + + if (!phys_enc || !phys_enc->hw_ctl || !phys_enc->hw_pp + || !phys_enc->hw_ctl->ops.setup_intf_cfg) { + DPU_ERROR("invalid arg(s), enc %d\n", phys_enc != 0); + return; + } + + DPU_DEBUG_CMDENC(cmd_enc, "pp %d, enabling mode:\n", + phys_enc->hw_pp->idx - PINGPONG_0); + drm_mode_debug_printmodeline(&phys_enc->cached_mode); + + _dpu_encoder_phys_cmd_update_intf_cfg(phys_enc); + dpu_encoder_phys_cmd_tearcheck_config(phys_enc); +} + +static bool dpu_encoder_phys_cmd_needs_single_flush( + struct dpu_encoder_phys *phys_enc) +{ + /** + * we do separate flush for each CTL and let + * CTL_START synchronize them + */ + return false; +} + +static void dpu_encoder_phys_cmd_enable_helper( + struct dpu_encoder_phys *phys_enc) +{ + struct dpu_hw_ctl *ctl; + u32 flush_mask = 0; + + if (!phys_enc || !phys_enc->hw_ctl || !phys_enc->hw_pp) { + DPU_ERROR("invalid arg(s), encoder %d\n", phys_enc != 0); + return; + } + + dpu_encoder_helper_split_config(phys_enc, phys_enc->intf_idx); + + _dpu_encoder_phys_cmd_pingpong_config(phys_enc); + + if (!dpu_encoder_phys_cmd_is_master(phys_enc)) + goto skip_flush; + + ctl = phys_enc->hw_ctl; + ctl->ops.get_bitmask_intf(ctl, &flush_mask, phys_enc->intf_idx); + ctl->ops.update_pending_flush(ctl, flush_mask); + +skip_flush: + return; +} + +static void dpu_encoder_phys_cmd_enable(struct dpu_encoder_phys *phys_enc) +{ + struct dpu_encoder_phys_cmd *cmd_enc = + to_dpu_encoder_phys_cmd(phys_enc); + + if (!phys_enc || !phys_enc->hw_pp) { + DPU_ERROR("invalid phys encoder\n"); + return; + } + + DPU_DEBUG_CMDENC(cmd_enc, "pp %d\n", phys_enc->hw_pp->idx - PINGPONG_0); + + if (phys_enc->enable_state == DPU_ENC_ENABLED) { + DPU_ERROR("already enabled\n"); + return; + } + + dpu_encoder_phys_cmd_enable_helper(phys_enc); + phys_enc->enable_state = DPU_ENC_ENABLED; +} + +static void _dpu_encoder_phys_cmd_connect_te( + struct dpu_encoder_phys *phys_enc, bool enable) +{ + if (!phys_enc || !phys_enc->hw_pp || + !phys_enc->hw_pp->ops.connect_external_te) + return; + + trace_dpu_enc_phys_cmd_connect_te(DRMID(phys_enc->parent), enable); + phys_enc->hw_pp->ops.connect_external_te(phys_enc->hw_pp, enable); +} + +static void dpu_encoder_phys_cmd_prepare_idle_pc( + struct dpu_encoder_phys *phys_enc) +{ + _dpu_encoder_phys_cmd_connect_te(phys_enc, false); +} + +static int dpu_encoder_phys_cmd_get_line_count( + struct dpu_encoder_phys *phys_enc) +{ + struct dpu_hw_pingpong *hw_pp; + + if (!phys_enc || !phys_enc->hw_pp) + return -EINVAL; + + if (!dpu_encoder_phys_cmd_is_master(phys_enc)) + return -EINVAL; + + hw_pp = phys_enc->hw_pp; + if (!hw_pp->ops.get_line_count) + return -EINVAL; + + return hw_pp->ops.get_line_count(hw_pp); +} + +static void dpu_encoder_phys_cmd_disable(struct dpu_encoder_phys *phys_enc) +{ + struct dpu_encoder_phys_cmd *cmd_enc = + to_dpu_encoder_phys_cmd(phys_enc); + + if (!phys_enc || !phys_enc->hw_pp) { + DPU_ERROR("invalid encoder\n"); + return; + } + DRM_DEBUG_KMS("id:%u pp:%d state:%d\n", DRMID(phys_enc->parent), + phys_enc->hw_pp->idx - PINGPONG_0, + phys_enc->enable_state); + + if (phys_enc->enable_state == DPU_ENC_DISABLED) { + DPU_ERROR_CMDENC(cmd_enc, "already disabled\n"); + return; + } + + if (phys_enc->hw_pp->ops.enable_tearcheck) + phys_enc->hw_pp->ops.enable_tearcheck(phys_enc->hw_pp, false); + phys_enc->enable_state = DPU_ENC_DISABLED; +} + +static void dpu_encoder_phys_cmd_destroy(struct dpu_encoder_phys *phys_enc) +{ + struct dpu_encoder_phys_cmd *cmd_enc = + to_dpu_encoder_phys_cmd(phys_enc); + + if (!phys_enc) { + DPU_ERROR("invalid encoder\n"); + return; + } + kfree(cmd_enc); +} + +static void dpu_encoder_phys_cmd_get_hw_resources( + struct dpu_encoder_phys *phys_enc, + struct dpu_encoder_hw_resources *hw_res, + struct drm_connector_state *conn_state) +{ + struct dpu_encoder_phys_cmd *cmd_enc = + to_dpu_encoder_phys_cmd(phys_enc); + + if (!phys_enc) { + DPU_ERROR("invalid encoder\n"); + return; + } + + if ((phys_enc->intf_idx - INTF_0) >= INTF_MAX) { + DPU_ERROR("invalid intf idx:%d\n", phys_enc->intf_idx); + return; + } + + DPU_DEBUG_CMDENC(cmd_enc, "\n"); + hw_res->intfs[phys_enc->intf_idx - INTF_0] = INTF_MODE_CMD; +} + +static void dpu_encoder_phys_cmd_prepare_for_kickoff( + struct dpu_encoder_phys *phys_enc, + struct dpu_encoder_kickoff_params *params) +{ + struct dpu_encoder_phys_cmd *cmd_enc = + to_dpu_encoder_phys_cmd(phys_enc); + int ret; + + if (!phys_enc || !phys_enc->hw_pp) { + DPU_ERROR("invalid encoder\n"); + return; + } + DRM_DEBUG_KMS("id:%u pp:%d pending_cnt:%d\n", DRMID(phys_enc->parent), + phys_enc->hw_pp->idx - PINGPONG_0, + atomic_read(&phys_enc->pending_kickoff_cnt)); + + /* + * Mark kickoff request as outstanding. If there are more than one, + * outstanding, then we have to wait for the previous one to complete + */ + ret = _dpu_encoder_phys_cmd_wait_for_idle(phys_enc); + if (ret) { + /* force pending_kickoff_cnt 0 to discard failed kickoff */ + atomic_set(&phys_enc->pending_kickoff_cnt, 0); + DRM_ERROR("failed wait_for_idle: id:%u ret:%d pp:%d\n", + DRMID(phys_enc->parent), ret, + phys_enc->hw_pp->idx - PINGPONG_0); + } + + DPU_DEBUG_CMDENC(cmd_enc, "pp:%d pending_cnt %d\n", + phys_enc->hw_pp->idx - PINGPONG_0, + atomic_read(&phys_enc->pending_kickoff_cnt)); +} + +static int _dpu_encoder_phys_cmd_wait_for_ctl_start( + struct dpu_encoder_phys *phys_enc) +{ + struct dpu_encoder_phys_cmd *cmd_enc = + to_dpu_encoder_phys_cmd(phys_enc); + struct dpu_encoder_wait_info wait_info; + int ret; + + if (!phys_enc || !phys_enc->hw_ctl) { + DPU_ERROR("invalid argument(s)\n"); + return -EINVAL; + } + + wait_info.wq = &phys_enc->pending_kickoff_wq; + wait_info.atomic_cnt = &phys_enc->pending_ctlstart_cnt; + wait_info.timeout_ms = KICKOFF_TIMEOUT_MS; + + ret = dpu_encoder_helper_wait_for_irq(phys_enc, INTR_IDX_CTL_START, + &wait_info); + if (ret == -ETIMEDOUT) { + DPU_ERROR_CMDENC(cmd_enc, "ctl start interrupt wait failed\n"); + ret = -EINVAL; + } else if (!ret) + ret = 0; + + return ret; +} + +static int dpu_encoder_phys_cmd_wait_for_tx_complete( + struct dpu_encoder_phys *phys_enc) +{ + int rc; + struct dpu_encoder_phys_cmd *cmd_enc; + + if (!phys_enc) + return -EINVAL; + + cmd_enc = to_dpu_encoder_phys_cmd(phys_enc); + + rc = _dpu_encoder_phys_cmd_wait_for_idle(phys_enc); + if (rc) { + DRM_ERROR("failed wait_for_idle: id:%u ret:%d intf:%d\n", + DRMID(phys_enc->parent), rc, + phys_enc->intf_idx - INTF_0); + } + + return rc; +} + +static int dpu_encoder_phys_cmd_wait_for_commit_done( + struct dpu_encoder_phys *phys_enc) +{ + int rc = 0; + struct dpu_encoder_phys_cmd *cmd_enc; + + if (!phys_enc) + return -EINVAL; + + cmd_enc = to_dpu_encoder_phys_cmd(phys_enc); + + /* only required for master controller */ + if (dpu_encoder_phys_cmd_is_master(phys_enc)) + rc = _dpu_encoder_phys_cmd_wait_for_ctl_start(phys_enc); + + /* required for both controllers */ + if (!rc && cmd_enc->serialize_wait4pp) + dpu_encoder_phys_cmd_prepare_for_kickoff(phys_enc, NULL); + + return rc; +} + +static int dpu_encoder_phys_cmd_wait_for_vblank( + struct dpu_encoder_phys *phys_enc) +{ + int rc = 0; + struct dpu_encoder_phys_cmd *cmd_enc; + struct dpu_encoder_wait_info wait_info; + + if (!phys_enc) + return -EINVAL; + + cmd_enc = to_dpu_encoder_phys_cmd(phys_enc); + + /* only required for master controller */ + if (!dpu_encoder_phys_cmd_is_master(phys_enc)) + return rc; + + wait_info.wq = &cmd_enc->pending_vblank_wq; + wait_info.atomic_cnt = &cmd_enc->pending_vblank_cnt; + wait_info.timeout_ms = _dpu_encoder_phys_cmd_get_idle_timeout(cmd_enc); + + atomic_inc(&cmd_enc->pending_vblank_cnt); + + rc = dpu_encoder_helper_wait_for_irq(phys_enc, INTR_IDX_RDPTR, + &wait_info); + + return rc; +} + +static void dpu_encoder_phys_cmd_handle_post_kickoff( + struct dpu_encoder_phys *phys_enc) +{ + if (!phys_enc) + return; + + /** + * re-enable external TE, either for the first time after enabling + * or if disabled for Autorefresh + */ + _dpu_encoder_phys_cmd_connect_te(phys_enc, true); +} + +static void dpu_encoder_phys_cmd_trigger_start( + struct dpu_encoder_phys *phys_enc) +{ + if (!phys_enc) + return; + + dpu_encoder_helper_trigger_start(phys_enc); +} + +static void dpu_encoder_phys_cmd_init_ops( + struct dpu_encoder_phys_ops *ops) +{ + ops->is_master = dpu_encoder_phys_cmd_is_master; + ops->mode_set = dpu_encoder_phys_cmd_mode_set; + ops->mode_fixup = dpu_encoder_phys_cmd_mode_fixup; + ops->enable = dpu_encoder_phys_cmd_enable; + ops->disable = dpu_encoder_phys_cmd_disable; + ops->destroy = dpu_encoder_phys_cmd_destroy; + ops->get_hw_resources = dpu_encoder_phys_cmd_get_hw_resources; + ops->control_vblank_irq = dpu_encoder_phys_cmd_control_vblank_irq; + ops->wait_for_commit_done = dpu_encoder_phys_cmd_wait_for_commit_done; + ops->prepare_for_kickoff = dpu_encoder_phys_cmd_prepare_for_kickoff; + ops->wait_for_tx_complete = dpu_encoder_phys_cmd_wait_for_tx_complete; + ops->wait_for_vblank = dpu_encoder_phys_cmd_wait_for_vblank; + ops->trigger_start = dpu_encoder_phys_cmd_trigger_start; + ops->needs_single_flush = dpu_encoder_phys_cmd_needs_single_flush; + ops->hw_reset = dpu_encoder_helper_hw_reset; + ops->irq_control = dpu_encoder_phys_cmd_irq_control; + ops->restore = dpu_encoder_phys_cmd_enable_helper; + ops->prepare_idle_pc = dpu_encoder_phys_cmd_prepare_idle_pc; + ops->handle_post_kickoff = dpu_encoder_phys_cmd_handle_post_kickoff; + ops->get_line_count = dpu_encoder_phys_cmd_get_line_count; +} + +struct dpu_encoder_phys *dpu_encoder_phys_cmd_init( + struct dpu_enc_phys_init_params *p) +{ + struct dpu_encoder_phys *phys_enc = NULL; + struct dpu_encoder_phys_cmd *cmd_enc = NULL; + struct dpu_hw_mdp *hw_mdp; + struct dpu_encoder_irq *irq; + int i, ret = 0; + + DPU_DEBUG("intf %d\n", p->intf_idx - INTF_0); + + cmd_enc = kzalloc(sizeof(*cmd_enc), GFP_KERNEL); + if (!cmd_enc) { + ret = -ENOMEM; + DPU_ERROR("failed to allocate\n"); + goto fail; + } + phys_enc = &cmd_enc->base; + + hw_mdp = dpu_rm_get_mdp(&p->dpu_kms->rm); + if (IS_ERR_OR_NULL(hw_mdp)) { + ret = PTR_ERR(hw_mdp); + DPU_ERROR("failed to get mdptop\n"); + goto fail_mdp_init; + } + phys_enc->hw_mdptop = hw_mdp; + phys_enc->intf_idx = p->intf_idx; + + dpu_encoder_phys_cmd_init_ops(&phys_enc->ops); + phys_enc->parent = p->parent; + phys_enc->parent_ops = p->parent_ops; + phys_enc->dpu_kms = p->dpu_kms; + phys_enc->split_role = p->split_role; + phys_enc->intf_mode = INTF_MODE_CMD; + phys_enc->enc_spinlock = p->enc_spinlock; + cmd_enc->stream_sel = 0; + phys_enc->enable_state = DPU_ENC_DISABLED; + for (i = 0; i < INTR_IDX_MAX; i++) { + irq = &phys_enc->irq[i]; + INIT_LIST_HEAD(&irq->cb.list); + irq->irq_idx = -EINVAL; + irq->hw_idx = -EINVAL; + irq->cb.arg = phys_enc; + } + + irq = &phys_enc->irq[INTR_IDX_CTL_START]; + irq->name = "ctl_start"; + irq->intr_type = DPU_IRQ_TYPE_CTL_START; + irq->intr_idx = INTR_IDX_CTL_START; + irq->cb.func = dpu_encoder_phys_cmd_ctl_start_irq; + + irq = &phys_enc->irq[INTR_IDX_PINGPONG]; + irq->name = "pp_done"; + irq->intr_type = DPU_IRQ_TYPE_PING_PONG_COMP; + irq->intr_idx = INTR_IDX_PINGPONG; + irq->cb.func = dpu_encoder_phys_cmd_pp_tx_done_irq; + + irq = &phys_enc->irq[INTR_IDX_RDPTR]; + irq->name = "pp_rd_ptr"; + irq->intr_type = DPU_IRQ_TYPE_PING_PONG_RD_PTR; + irq->intr_idx = INTR_IDX_RDPTR; + irq->cb.func = dpu_encoder_phys_cmd_pp_rd_ptr_irq; + + irq = &phys_enc->irq[INTR_IDX_UNDERRUN]; + irq->name = "underrun"; + irq->intr_type = DPU_IRQ_TYPE_INTF_UNDER_RUN; + irq->intr_idx = INTR_IDX_UNDERRUN; + irq->cb.func = dpu_encoder_phys_cmd_underrun_irq; + + atomic_set(&phys_enc->vblank_refcount, 0); + atomic_set(&phys_enc->pending_kickoff_cnt, 0); + atomic_set(&phys_enc->pending_ctlstart_cnt, 0); + atomic_set(&cmd_enc->pending_vblank_cnt, 0); + init_waitqueue_head(&phys_enc->pending_kickoff_wq); + init_waitqueue_head(&cmd_enc->pending_vblank_wq); + + DPU_DEBUG_CMDENC(cmd_enc, "created\n"); + + return phys_enc; + +fail_mdp_init: + kfree(cmd_enc); +fail: + return ERR_PTR(ret); +} diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder_phys_vid.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder_phys_vid.c new file mode 100644 index 000000000000..14fc7c2a6bb7 --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder_phys_vid.c @@ -0,0 +1,922 @@ +/* Copyright (c) 2015-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#define pr_fmt(fmt) "[drm:%s:%d] " fmt, __func__, __LINE__ +#include "dpu_encoder_phys.h" +#include "dpu_hw_interrupts.h" +#include "dpu_core_irq.h" +#include "dpu_formats.h" +#include "dpu_trace.h" + +#define DPU_DEBUG_VIDENC(e, fmt, ...) DPU_DEBUG("enc%d intf%d " fmt, \ + (e) && (e)->base.parent ? \ + (e)->base.parent->base.id : -1, \ + (e) && (e)->hw_intf ? \ + (e)->hw_intf->idx - INTF_0 : -1, ##__VA_ARGS__) + +#define DPU_ERROR_VIDENC(e, fmt, ...) DPU_ERROR("enc%d intf%d " fmt, \ + (e) && (e)->base.parent ? \ + (e)->base.parent->base.id : -1, \ + (e) && (e)->hw_intf ? \ + (e)->hw_intf->idx - INTF_0 : -1, ##__VA_ARGS__) + +#define to_dpu_encoder_phys_vid(x) \ + container_of(x, struct dpu_encoder_phys_vid, base) + +static bool dpu_encoder_phys_vid_is_master( + struct dpu_encoder_phys *phys_enc) +{ + bool ret = false; + + if (phys_enc->split_role != ENC_ROLE_SLAVE) + ret = true; + + return ret; +} + +static void drm_mode_to_intf_timing_params( + const struct dpu_encoder_phys_vid *vid_enc, + const struct drm_display_mode *mode, + struct intf_timing_params *timing) +{ + memset(timing, 0, sizeof(*timing)); + + if ((mode->htotal < mode->hsync_end) + || (mode->hsync_start < mode->hdisplay) + || (mode->vtotal < mode->vsync_end) + || (mode->vsync_start < mode->vdisplay) + || (mode->hsync_end < mode->hsync_start) + || (mode->vsync_end < mode->vsync_start)) { + DPU_ERROR( + "invalid params - hstart:%d,hend:%d,htot:%d,hdisplay:%d\n", + mode->hsync_start, mode->hsync_end, + mode->htotal, mode->hdisplay); + DPU_ERROR("vstart:%d,vend:%d,vtot:%d,vdisplay:%d\n", + mode->vsync_start, mode->vsync_end, + mode->vtotal, mode->vdisplay); + return; + } + + /* + * https://www.kernel.org/doc/htmldocs/drm/ch02s05.html + * Active Region Front Porch Sync Back Porch + * <-----------------><------------><-----><-----------> + * <- [hv]display ---> + * <--------- [hv]sync_start ------> + * <----------------- [hv]sync_end -------> + * <---------------------------- [hv]total -------------> + */ + timing->width = mode->hdisplay; /* active width */ + timing->height = mode->vdisplay; /* active height */ + timing->xres = timing->width; + timing->yres = timing->height; + timing->h_back_porch = mode->htotal - mode->hsync_end; + timing->h_front_porch = mode->hsync_start - mode->hdisplay; + timing->v_back_porch = mode->vtotal - mode->vsync_end; + timing->v_front_porch = mode->vsync_start - mode->vdisplay; + timing->hsync_pulse_width = mode->hsync_end - mode->hsync_start; + timing->vsync_pulse_width = mode->vsync_end - mode->vsync_start; + timing->hsync_polarity = (mode->flags & DRM_MODE_FLAG_NHSYNC) ? 1 : 0; + timing->vsync_polarity = (mode->flags & DRM_MODE_FLAG_NVSYNC) ? 1 : 0; + timing->border_clr = 0; + timing->underflow_clr = 0xff; + timing->hsync_skew = mode->hskew; + + /* DSI controller cannot handle active-low sync signals. */ + if (vid_enc->hw_intf->cap->type == INTF_DSI) { + timing->hsync_polarity = 0; + timing->vsync_polarity = 0; + } + + /* + * For edp only: + * DISPLAY_V_START = (VBP * HCYCLE) + HBP + * DISPLAY_V_END = (VBP + VACTIVE) * HCYCLE - 1 - HFP + */ + /* + * if (vid_enc->hw->cap->type == INTF_EDP) { + * display_v_start += mode->htotal - mode->hsync_start; + * display_v_end -= mode->hsync_start - mode->hdisplay; + * } + */ +} + +static inline u32 get_horizontal_total(const struct intf_timing_params *timing) +{ + u32 active = timing->xres; + u32 inactive = + timing->h_back_porch + timing->h_front_porch + + timing->hsync_pulse_width; + return active + inactive; +} + +static inline u32 get_vertical_total(const struct intf_timing_params *timing) +{ + u32 active = timing->yres; + u32 inactive = + timing->v_back_porch + timing->v_front_porch + + timing->vsync_pulse_width; + return active + inactive; +} + +/* + * programmable_fetch_get_num_lines: + * Number of fetch lines in vertical front porch + * @timing: Pointer to the intf timing information for the requested mode + * + * Returns the number of fetch lines in vertical front porch at which mdp + * can start fetching the next frame. + * + * Number of needed prefetch lines is anything that cannot be absorbed in the + * start of frame time (back porch + vsync pulse width). + * + * Some panels have very large VFP, however we only need a total number of + * lines based on the chip worst case latencies. + */ +static u32 programmable_fetch_get_num_lines( + struct dpu_encoder_phys_vid *vid_enc, + const struct intf_timing_params *timing) +{ + u32 worst_case_needed_lines = + vid_enc->hw_intf->cap->prog_fetch_lines_worst_case; + u32 start_of_frame_lines = + timing->v_back_porch + timing->vsync_pulse_width; + u32 needed_vfp_lines = worst_case_needed_lines - start_of_frame_lines; + u32 actual_vfp_lines = 0; + + /* Fetch must be outside active lines, otherwise undefined. */ + if (start_of_frame_lines >= worst_case_needed_lines) { + DPU_DEBUG_VIDENC(vid_enc, + "prog fetch is not needed, large vbp+vsw\n"); + actual_vfp_lines = 0; + } else if (timing->v_front_porch < needed_vfp_lines) { + /* Warn fetch needed, but not enough porch in panel config */ + pr_warn_once + ("low vbp+vfp may lead to perf issues in some cases\n"); + DPU_DEBUG_VIDENC(vid_enc, + "less vfp than fetch req, using entire vfp\n"); + actual_vfp_lines = timing->v_front_porch; + } else { + DPU_DEBUG_VIDENC(vid_enc, "room in vfp for needed prefetch\n"); + actual_vfp_lines = needed_vfp_lines; + } + + DPU_DEBUG_VIDENC(vid_enc, + "v_front_porch %u v_back_porch %u vsync_pulse_width %u\n", + timing->v_front_porch, timing->v_back_porch, + timing->vsync_pulse_width); + DPU_DEBUG_VIDENC(vid_enc, + "wc_lines %u needed_vfp_lines %u actual_vfp_lines %u\n", + worst_case_needed_lines, needed_vfp_lines, actual_vfp_lines); + + return actual_vfp_lines; +} + +/* + * programmable_fetch_config: Programs HW to prefetch lines by offsetting + * the start of fetch into the vertical front porch for cases where the + * vsync pulse width and vertical back porch time is insufficient + * + * Gets # of lines to pre-fetch, then calculate VSYNC counter value. + * HW layer requires VSYNC counter of first pixel of tgt VFP line. + * + * @timing: Pointer to the intf timing information for the requested mode + */ +static void programmable_fetch_config(struct dpu_encoder_phys *phys_enc, + const struct intf_timing_params *timing) +{ + struct dpu_encoder_phys_vid *vid_enc = + to_dpu_encoder_phys_vid(phys_enc); + struct intf_prog_fetch f = { 0 }; + u32 vfp_fetch_lines = 0; + u32 horiz_total = 0; + u32 vert_total = 0; + u32 vfp_fetch_start_vsync_counter = 0; + unsigned long lock_flags; + + if (WARN_ON_ONCE(!vid_enc->hw_intf->ops.setup_prg_fetch)) + return; + + vfp_fetch_lines = programmable_fetch_get_num_lines(vid_enc, timing); + if (vfp_fetch_lines) { + vert_total = get_vertical_total(timing); + horiz_total = get_horizontal_total(timing); + vfp_fetch_start_vsync_counter = + (vert_total - vfp_fetch_lines) * horiz_total + 1; + f.enable = 1; + f.fetch_start = vfp_fetch_start_vsync_counter; + } + + DPU_DEBUG_VIDENC(vid_enc, + "vfp_fetch_lines %u vfp_fetch_start_vsync_counter %u\n", + vfp_fetch_lines, vfp_fetch_start_vsync_counter); + + spin_lock_irqsave(phys_enc->enc_spinlock, lock_flags); + vid_enc->hw_intf->ops.setup_prg_fetch(vid_enc->hw_intf, &f); + spin_unlock_irqrestore(phys_enc->enc_spinlock, lock_flags); +} + +static bool dpu_encoder_phys_vid_mode_fixup( + struct dpu_encoder_phys *phys_enc, + const struct drm_display_mode *mode, + struct drm_display_mode *adj_mode) +{ + if (phys_enc) + DPU_DEBUG_VIDENC(to_dpu_encoder_phys_vid(phys_enc), "\n"); + + /* + * Modifying mode has consequences when the mode comes back to us + */ + return true; +} + +static void dpu_encoder_phys_vid_setup_timing_engine( + struct dpu_encoder_phys *phys_enc) +{ + struct dpu_encoder_phys_vid *vid_enc; + struct drm_display_mode mode; + struct intf_timing_params timing_params = { 0 }; + const struct dpu_format *fmt = NULL; + u32 fmt_fourcc = DRM_FORMAT_RGB888; + unsigned long lock_flags; + struct dpu_hw_intf_cfg intf_cfg = { 0 }; + + if (!phys_enc || !phys_enc->hw_ctl->ops.setup_intf_cfg) { + DPU_ERROR("invalid encoder %d\n", phys_enc != 0); + return; + } + + mode = phys_enc->cached_mode; + vid_enc = to_dpu_encoder_phys_vid(phys_enc); + if (!vid_enc->hw_intf->ops.setup_timing_gen) { + DPU_ERROR("timing engine setup is not supported\n"); + return; + } + + DPU_DEBUG_VIDENC(vid_enc, "enabling mode:\n"); + drm_mode_debug_printmodeline(&mode); + + if (phys_enc->split_role != ENC_ROLE_SOLO) { + mode.hdisplay >>= 1; + mode.htotal >>= 1; + mode.hsync_start >>= 1; + mode.hsync_end >>= 1; + + DPU_DEBUG_VIDENC(vid_enc, + "split_role %d, halve horizontal %d %d %d %d\n", + phys_enc->split_role, + mode.hdisplay, mode.htotal, + mode.hsync_start, mode.hsync_end); + } + + drm_mode_to_intf_timing_params(vid_enc, &mode, &timing_params); + + fmt = dpu_get_dpu_format(fmt_fourcc); + DPU_DEBUG_VIDENC(vid_enc, "fmt_fourcc 0x%X\n", fmt_fourcc); + + intf_cfg.intf = vid_enc->hw_intf->idx; + intf_cfg.intf_mode_sel = DPU_CTL_MODE_SEL_VID; + intf_cfg.stream_sel = 0; /* Don't care value for video mode */ + intf_cfg.mode_3d = dpu_encoder_helper_get_3d_blend_mode(phys_enc); + + spin_lock_irqsave(phys_enc->enc_spinlock, lock_flags); + vid_enc->hw_intf->ops.setup_timing_gen(vid_enc->hw_intf, + &timing_params, fmt); + phys_enc->hw_ctl->ops.setup_intf_cfg(phys_enc->hw_ctl, &intf_cfg); + spin_unlock_irqrestore(phys_enc->enc_spinlock, lock_flags); + + programmable_fetch_config(phys_enc, &timing_params); + + vid_enc->timing_params = timing_params; +} + +static void dpu_encoder_phys_vid_vblank_irq(void *arg, int irq_idx) +{ + struct dpu_encoder_phys *phys_enc = arg; + struct dpu_hw_ctl *hw_ctl; + unsigned long lock_flags; + u32 flush_register = 0; + int new_cnt = -1, old_cnt = -1; + + if (!phys_enc) + return; + + hw_ctl = phys_enc->hw_ctl; + if (!hw_ctl) + return; + + DPU_ATRACE_BEGIN("vblank_irq"); + + if (phys_enc->parent_ops->handle_vblank_virt) + phys_enc->parent_ops->handle_vblank_virt(phys_enc->parent, + phys_enc); + + old_cnt = atomic_read(&phys_enc->pending_kickoff_cnt); + + /* + * only decrement the pending flush count if we've actually flushed + * hardware. due to sw irq latency, vblank may have already happened + * so we need to double-check with hw that it accepted the flush bits + */ + spin_lock_irqsave(phys_enc->enc_spinlock, lock_flags); + if (hw_ctl && hw_ctl->ops.get_flush_register) + flush_register = hw_ctl->ops.get_flush_register(hw_ctl); + + if (flush_register == 0) + new_cnt = atomic_add_unless(&phys_enc->pending_kickoff_cnt, + -1, 0); + spin_unlock_irqrestore(phys_enc->enc_spinlock, lock_flags); + + /* Signal any waiting atomic commit thread */ + wake_up_all(&phys_enc->pending_kickoff_wq); + DPU_ATRACE_END("vblank_irq"); +} + +static void dpu_encoder_phys_vid_underrun_irq(void *arg, int irq_idx) +{ + struct dpu_encoder_phys *phys_enc = arg; + + if (!phys_enc) + return; + + if (phys_enc->parent_ops->handle_underrun_virt) + phys_enc->parent_ops->handle_underrun_virt(phys_enc->parent, + phys_enc); +} + +static bool _dpu_encoder_phys_is_dual_ctl(struct dpu_encoder_phys *phys_enc) +{ + if (!phys_enc) + return false; + + if (phys_enc->topology_name == DPU_RM_TOPOLOGY_DUALPIPE) + return true; + + return false; +} + +static bool dpu_encoder_phys_vid_needs_single_flush( + struct dpu_encoder_phys *phys_enc) +{ + return (phys_enc && _dpu_encoder_phys_is_dual_ctl(phys_enc)); +} + +static void _dpu_encoder_phys_vid_setup_irq_hw_idx( + struct dpu_encoder_phys *phys_enc) +{ + struct dpu_encoder_irq *irq; + + /* + * Initialize irq->hw_idx only when irq is not registered. + * Prevent invalidating irq->irq_idx as modeset may be + * called many times during dfps. + */ + + irq = &phys_enc->irq[INTR_IDX_VSYNC]; + if (irq->irq_idx < 0) + irq->hw_idx = phys_enc->intf_idx; + + irq = &phys_enc->irq[INTR_IDX_UNDERRUN]; + if (irq->irq_idx < 0) + irq->hw_idx = phys_enc->intf_idx; +} + +static void dpu_encoder_phys_vid_mode_set( + struct dpu_encoder_phys *phys_enc, + struct drm_display_mode *mode, + struct drm_display_mode *adj_mode) +{ + struct dpu_rm *rm; + struct dpu_rm_hw_iter iter; + int i, instance; + struct dpu_encoder_phys_vid *vid_enc; + + if (!phys_enc || !phys_enc->dpu_kms) { + DPU_ERROR("invalid encoder/kms\n"); + return; + } + + rm = &phys_enc->dpu_kms->rm; + vid_enc = to_dpu_encoder_phys_vid(phys_enc); + + if (adj_mode) { + phys_enc->cached_mode = *adj_mode; + drm_mode_debug_printmodeline(adj_mode); + DPU_DEBUG_VIDENC(vid_enc, "caching mode:\n"); + } + + instance = phys_enc->split_role == ENC_ROLE_SLAVE ? 1 : 0; + + /* Retrieve previously allocated HW Resources. Shouldn't fail */ + dpu_rm_init_hw_iter(&iter, phys_enc->parent->base.id, DPU_HW_BLK_CTL); + for (i = 0; i <= instance; i++) { + if (dpu_rm_get_hw(rm, &iter)) + phys_enc->hw_ctl = (struct dpu_hw_ctl *)iter.hw; + } + if (IS_ERR_OR_NULL(phys_enc->hw_ctl)) { + DPU_ERROR_VIDENC(vid_enc, "failed to init ctl, %ld\n", + PTR_ERR(phys_enc->hw_ctl)); + phys_enc->hw_ctl = NULL; + return; + } + + _dpu_encoder_phys_vid_setup_irq_hw_idx(phys_enc); +} + +static int dpu_encoder_phys_vid_control_vblank_irq( + struct dpu_encoder_phys *phys_enc, + bool enable) +{ + int ret = 0; + struct dpu_encoder_phys_vid *vid_enc; + int refcount; + + if (!phys_enc) { + DPU_ERROR("invalid encoder\n"); + return -EINVAL; + } + + refcount = atomic_read(&phys_enc->vblank_refcount); + vid_enc = to_dpu_encoder_phys_vid(phys_enc); + + /* Slave encoders don't report vblank */ + if (!dpu_encoder_phys_vid_is_master(phys_enc)) + goto end; + + /* protect against negative */ + if (!enable && refcount == 0) { + ret = -EINVAL; + goto end; + } + + DRM_DEBUG_KMS("id:%u enable=%d/%d\n", DRMID(phys_enc->parent), enable, + atomic_read(&phys_enc->vblank_refcount)); + + if (enable && atomic_inc_return(&phys_enc->vblank_refcount) == 1) + ret = dpu_encoder_helper_register_irq(phys_enc, INTR_IDX_VSYNC); + else if (!enable && atomic_dec_return(&phys_enc->vblank_refcount) == 0) + ret = dpu_encoder_helper_unregister_irq(phys_enc, + INTR_IDX_VSYNC); + +end: + if (ret) { + DRM_ERROR("failed: id:%u intf:%d ret:%d enable:%d refcnt:%d\n", + DRMID(phys_enc->parent), + vid_enc->hw_intf->idx - INTF_0, ret, enable, + refcount); + } + return ret; +} + +static void dpu_encoder_phys_vid_enable(struct dpu_encoder_phys *phys_enc) +{ + struct msm_drm_private *priv; + struct dpu_encoder_phys_vid *vid_enc; + struct dpu_hw_intf *intf; + struct dpu_hw_ctl *ctl; + u32 flush_mask = 0; + + if (!phys_enc || !phys_enc->parent || !phys_enc->parent->dev || + !phys_enc->parent->dev->dev_private) { + DPU_ERROR("invalid encoder/device\n"); + return; + } + priv = phys_enc->parent->dev->dev_private; + + vid_enc = to_dpu_encoder_phys_vid(phys_enc); + intf = vid_enc->hw_intf; + ctl = phys_enc->hw_ctl; + if (!vid_enc->hw_intf || !phys_enc->hw_ctl) { + DPU_ERROR("invalid hw_intf %d hw_ctl %d\n", + vid_enc->hw_intf != 0, phys_enc->hw_ctl != 0); + return; + } + + DPU_DEBUG_VIDENC(vid_enc, "\n"); + + if (WARN_ON(!vid_enc->hw_intf->ops.enable_timing)) + return; + + dpu_encoder_helper_split_config(phys_enc, vid_enc->hw_intf->idx); + + dpu_encoder_phys_vid_setup_timing_engine(phys_enc); + + /* + * For single flush cases (dual-ctl or pp-split), skip setting the + * flush bit for the slave intf, since both intfs use same ctl + * and HW will only flush the master. + */ + if (dpu_encoder_phys_vid_needs_single_flush(phys_enc) && + !dpu_encoder_phys_vid_is_master(phys_enc)) + goto skip_flush; + + ctl->ops.get_bitmask_intf(ctl, &flush_mask, intf->idx); + ctl->ops.update_pending_flush(ctl, flush_mask); + +skip_flush: + DPU_DEBUG_VIDENC(vid_enc, "update pending flush ctl %d flush_mask %x\n", + ctl->idx - CTL_0, flush_mask); + + /* ctl_flush & timing engine enable will be triggered by framework */ + if (phys_enc->enable_state == DPU_ENC_DISABLED) + phys_enc->enable_state = DPU_ENC_ENABLING; +} + +static void dpu_encoder_phys_vid_destroy(struct dpu_encoder_phys *phys_enc) +{ + struct dpu_encoder_phys_vid *vid_enc; + + if (!phys_enc) { + DPU_ERROR("invalid encoder\n"); + return; + } + + vid_enc = to_dpu_encoder_phys_vid(phys_enc); + DPU_DEBUG_VIDENC(vid_enc, "\n"); + kfree(vid_enc); +} + +static void dpu_encoder_phys_vid_get_hw_resources( + struct dpu_encoder_phys *phys_enc, + struct dpu_encoder_hw_resources *hw_res, + struct drm_connector_state *conn_state) +{ + struct dpu_encoder_phys_vid *vid_enc; + + if (!phys_enc || !hw_res) { + DPU_ERROR("invalid arg(s), enc %d hw_res %d conn_state %d\n", + phys_enc != 0, hw_res != 0, conn_state != 0); + return; + } + + vid_enc = to_dpu_encoder_phys_vid(phys_enc); + if (!vid_enc->hw_intf) { + DPU_ERROR("invalid arg(s), hw_intf\n"); + return; + } + + DPU_DEBUG_VIDENC(vid_enc, "\n"); + hw_res->intfs[vid_enc->hw_intf->idx - INTF_0] = INTF_MODE_VIDEO; +} + +static int _dpu_encoder_phys_vid_wait_for_vblank( + struct dpu_encoder_phys *phys_enc, bool notify) +{ + struct dpu_encoder_wait_info wait_info; + int ret; + + if (!phys_enc) { + pr_err("invalid encoder\n"); + return -EINVAL; + } + + wait_info.wq = &phys_enc->pending_kickoff_wq; + wait_info.atomic_cnt = &phys_enc->pending_kickoff_cnt; + wait_info.timeout_ms = KICKOFF_TIMEOUT_MS; + + if (!dpu_encoder_phys_vid_is_master(phys_enc)) { + if (notify && phys_enc->parent_ops->handle_frame_done) + phys_enc->parent_ops->handle_frame_done( + phys_enc->parent, phys_enc, + DPU_ENCODER_FRAME_EVENT_DONE); + return 0; + } + + /* Wait for kickoff to complete */ + ret = dpu_encoder_helper_wait_for_irq(phys_enc, INTR_IDX_VSYNC, + &wait_info); + + if (ret == -ETIMEDOUT) { + dpu_encoder_helper_report_irq_timeout(phys_enc, INTR_IDX_VSYNC); + } else if (!ret && notify && phys_enc->parent_ops->handle_frame_done) + phys_enc->parent_ops->handle_frame_done( + phys_enc->parent, phys_enc, + DPU_ENCODER_FRAME_EVENT_DONE); + + return ret; +} + +static int dpu_encoder_phys_vid_wait_for_vblank( + struct dpu_encoder_phys *phys_enc) +{ + return _dpu_encoder_phys_vid_wait_for_vblank(phys_enc, true); +} + +static void dpu_encoder_phys_vid_prepare_for_kickoff( + struct dpu_encoder_phys *phys_enc, + struct dpu_encoder_kickoff_params *params) +{ + struct dpu_encoder_phys_vid *vid_enc; + struct dpu_hw_ctl *ctl; + int rc; + + if (!phys_enc || !params) { + DPU_ERROR("invalid encoder/parameters\n"); + return; + } + vid_enc = to_dpu_encoder_phys_vid(phys_enc); + + ctl = phys_enc->hw_ctl; + if (!ctl || !ctl->ops.wait_reset_status) + return; + + /* + * hw supports hardware initiated ctl reset, so before we kickoff a new + * frame, need to check and wait for hw initiated ctl reset completion + */ + rc = ctl->ops.wait_reset_status(ctl); + if (rc) { + DPU_ERROR_VIDENC(vid_enc, "ctl %d reset failure: %d\n", + ctl->idx, rc); + dpu_encoder_helper_unregister_irq(phys_enc, INTR_IDX_VSYNC); + dpu_dbg_dump(false, __func__, true, true); + } +} + +static void dpu_encoder_phys_vid_disable(struct dpu_encoder_phys *phys_enc) +{ + struct msm_drm_private *priv; + struct dpu_encoder_phys_vid *vid_enc; + unsigned long lock_flags; + int ret; + + if (!phys_enc || !phys_enc->parent || !phys_enc->parent->dev || + !phys_enc->parent->dev->dev_private) { + DPU_ERROR("invalid encoder/device\n"); + return; + } + priv = phys_enc->parent->dev->dev_private; + + vid_enc = to_dpu_encoder_phys_vid(phys_enc); + if (!vid_enc->hw_intf || !phys_enc->hw_ctl) { + DPU_ERROR("invalid hw_intf %d hw_ctl %d\n", + vid_enc->hw_intf != 0, phys_enc->hw_ctl != 0); + return; + } + + DPU_DEBUG_VIDENC(vid_enc, "\n"); + + if (WARN_ON(!vid_enc->hw_intf->ops.enable_timing)) + return; + + if (phys_enc->enable_state == DPU_ENC_DISABLED) { + DPU_ERROR("already disabled\n"); + return; + } + + spin_lock_irqsave(phys_enc->enc_spinlock, lock_flags); + vid_enc->hw_intf->ops.enable_timing(vid_enc->hw_intf, 0); + if (dpu_encoder_phys_vid_is_master(phys_enc)) + dpu_encoder_phys_inc_pending(phys_enc); + spin_unlock_irqrestore(phys_enc->enc_spinlock, lock_flags); + + /* + * Wait for a vsync so we know the ENABLE=0 latched before + * the (connector) source of the vsync's gets disabled, + * otherwise we end up in a funny state if we re-enable + * before the disable latches, which results that some of + * the settings changes for the new modeset (like new + * scanout buffer) don't latch properly.. + */ + if (dpu_encoder_phys_vid_is_master(phys_enc)) { + ret = _dpu_encoder_phys_vid_wait_for_vblank(phys_enc, false); + if (ret) { + atomic_set(&phys_enc->pending_kickoff_cnt, 0); + DRM_ERROR("wait disable failed: id:%u intf:%d ret:%d\n", + DRMID(phys_enc->parent), + vid_enc->hw_intf->idx - INTF_0, ret); + } + } + + phys_enc->enable_state = DPU_ENC_DISABLED; +} + +static void dpu_encoder_phys_vid_handle_post_kickoff( + struct dpu_encoder_phys *phys_enc) +{ + unsigned long lock_flags; + struct dpu_encoder_phys_vid *vid_enc; + + if (!phys_enc) { + DPU_ERROR("invalid encoder\n"); + return; + } + + vid_enc = to_dpu_encoder_phys_vid(phys_enc); + DPU_DEBUG_VIDENC(vid_enc, "enable_state %d\n", phys_enc->enable_state); + + /* + * Video mode must flush CTL before enabling timing engine + * Video encoders need to turn on their interfaces now + */ + if (phys_enc->enable_state == DPU_ENC_ENABLING) { + trace_dpu_enc_phys_vid_post_kickoff(DRMID(phys_enc->parent), + vid_enc->hw_intf->idx - INTF_0); + spin_lock_irqsave(phys_enc->enc_spinlock, lock_flags); + vid_enc->hw_intf->ops.enable_timing(vid_enc->hw_intf, 1); + spin_unlock_irqrestore(phys_enc->enc_spinlock, lock_flags); + phys_enc->enable_state = DPU_ENC_ENABLED; + } +} + +static void dpu_encoder_phys_vid_irq_control(struct dpu_encoder_phys *phys_enc, + bool enable) +{ + struct dpu_encoder_phys_vid *vid_enc; + int ret; + + if (!phys_enc) + return; + + vid_enc = to_dpu_encoder_phys_vid(phys_enc); + + trace_dpu_enc_phys_vid_irq_ctrl(DRMID(phys_enc->parent), + vid_enc->hw_intf->idx - INTF_0, + enable, + atomic_read(&phys_enc->vblank_refcount)); + + if (enable) { + ret = dpu_encoder_phys_vid_control_vblank_irq(phys_enc, true); + if (ret) + return; + + dpu_encoder_helper_register_irq(phys_enc, INTR_IDX_UNDERRUN); + } else { + dpu_encoder_phys_vid_control_vblank_irq(phys_enc, false); + dpu_encoder_helper_unregister_irq(phys_enc, INTR_IDX_UNDERRUN); + } +} + +static void dpu_encoder_phys_vid_setup_misr(struct dpu_encoder_phys *phys_enc, + bool enable, u32 frame_count) +{ + struct dpu_encoder_phys_vid *vid_enc; + + if (!phys_enc) + return; + vid_enc = to_dpu_encoder_phys_vid(phys_enc); + + if (vid_enc->hw_intf && vid_enc->hw_intf->ops.setup_misr) + vid_enc->hw_intf->ops.setup_misr(vid_enc->hw_intf, + enable, frame_count); +} + +static u32 dpu_encoder_phys_vid_collect_misr(struct dpu_encoder_phys *phys_enc) +{ + struct dpu_encoder_phys_vid *vid_enc; + + if (!phys_enc) + return 0; + vid_enc = to_dpu_encoder_phys_vid(phys_enc); + + return vid_enc->hw_intf && vid_enc->hw_intf->ops.collect_misr ? + vid_enc->hw_intf->ops.collect_misr(vid_enc->hw_intf) : 0; +} + +static int dpu_encoder_phys_vid_get_line_count( + struct dpu_encoder_phys *phys_enc) +{ + struct dpu_encoder_phys_vid *vid_enc; + + if (!phys_enc) + return -EINVAL; + + if (!dpu_encoder_phys_vid_is_master(phys_enc)) + return -EINVAL; + + vid_enc = to_dpu_encoder_phys_vid(phys_enc); + if (!vid_enc->hw_intf || !vid_enc->hw_intf->ops.get_line_count) + return -EINVAL; + + return vid_enc->hw_intf->ops.get_line_count(vid_enc->hw_intf); +} + +static void dpu_encoder_phys_vid_init_ops(struct dpu_encoder_phys_ops *ops) +{ + ops->is_master = dpu_encoder_phys_vid_is_master; + ops->mode_set = dpu_encoder_phys_vid_mode_set; + ops->mode_fixup = dpu_encoder_phys_vid_mode_fixup; + ops->enable = dpu_encoder_phys_vid_enable; + ops->disable = dpu_encoder_phys_vid_disable; + ops->destroy = dpu_encoder_phys_vid_destroy; + ops->get_hw_resources = dpu_encoder_phys_vid_get_hw_resources; + ops->control_vblank_irq = dpu_encoder_phys_vid_control_vblank_irq; + ops->wait_for_commit_done = dpu_encoder_phys_vid_wait_for_vblank; + ops->wait_for_vblank = dpu_encoder_phys_vid_wait_for_vblank; + ops->wait_for_tx_complete = dpu_encoder_phys_vid_wait_for_vblank; + ops->irq_control = dpu_encoder_phys_vid_irq_control; + ops->prepare_for_kickoff = dpu_encoder_phys_vid_prepare_for_kickoff; + ops->handle_post_kickoff = dpu_encoder_phys_vid_handle_post_kickoff; + ops->needs_single_flush = dpu_encoder_phys_vid_needs_single_flush; + ops->setup_misr = dpu_encoder_phys_vid_setup_misr; + ops->collect_misr = dpu_encoder_phys_vid_collect_misr; + ops->hw_reset = dpu_encoder_helper_hw_reset; + ops->get_line_count = dpu_encoder_phys_vid_get_line_count; +} + +struct dpu_encoder_phys *dpu_encoder_phys_vid_init( + struct dpu_enc_phys_init_params *p) +{ + struct dpu_encoder_phys *phys_enc = NULL; + struct dpu_encoder_phys_vid *vid_enc = NULL; + struct dpu_rm_hw_iter iter; + struct dpu_hw_mdp *hw_mdp; + struct dpu_encoder_irq *irq; + int i, ret = 0; + + if (!p) { + ret = -EINVAL; + goto fail; + } + + vid_enc = kzalloc(sizeof(*vid_enc), GFP_KERNEL); + if (!vid_enc) { + ret = -ENOMEM; + goto fail; + } + + phys_enc = &vid_enc->base; + + hw_mdp = dpu_rm_get_mdp(&p->dpu_kms->rm); + if (IS_ERR_OR_NULL(hw_mdp)) { + ret = PTR_ERR(hw_mdp); + DPU_ERROR("failed to get mdptop\n"); + goto fail; + } + phys_enc->hw_mdptop = hw_mdp; + phys_enc->intf_idx = p->intf_idx; + + /** + * hw_intf resource permanently assigned to this encoder + * Other resources allocated at atomic commit time by use case + */ + dpu_rm_init_hw_iter(&iter, 0, DPU_HW_BLK_INTF); + while (dpu_rm_get_hw(&p->dpu_kms->rm, &iter)) { + struct dpu_hw_intf *hw_intf = (struct dpu_hw_intf *)iter.hw; + + if (hw_intf->idx == p->intf_idx) { + vid_enc->hw_intf = hw_intf; + break; + } + } + + if (!vid_enc->hw_intf) { + ret = -EINVAL; + DPU_ERROR("failed to get hw_intf\n"); + goto fail; + } + + DPU_DEBUG_VIDENC(vid_enc, "\n"); + + dpu_encoder_phys_vid_init_ops(&phys_enc->ops); + phys_enc->parent = p->parent; + phys_enc->parent_ops = p->parent_ops; + phys_enc->dpu_kms = p->dpu_kms; + phys_enc->split_role = p->split_role; + phys_enc->intf_mode = INTF_MODE_VIDEO; + phys_enc->enc_spinlock = p->enc_spinlock; + for (i = 0; i < INTR_IDX_MAX; i++) { + irq = &phys_enc->irq[i]; + INIT_LIST_HEAD(&irq->cb.list); + irq->irq_idx = -EINVAL; + irq->hw_idx = -EINVAL; + irq->cb.arg = phys_enc; + } + + irq = &phys_enc->irq[INTR_IDX_VSYNC]; + irq->name = "vsync_irq"; + irq->intr_type = DPU_IRQ_TYPE_INTF_VSYNC; + irq->intr_idx = INTR_IDX_VSYNC; + irq->cb.func = dpu_encoder_phys_vid_vblank_irq; + + irq = &phys_enc->irq[INTR_IDX_UNDERRUN]; + irq->name = "underrun"; + irq->intr_type = DPU_IRQ_TYPE_INTF_UNDER_RUN; + irq->intr_idx = INTR_IDX_UNDERRUN; + irq->cb.func = dpu_encoder_phys_vid_underrun_irq; + + atomic_set(&phys_enc->vblank_refcount, 0); + atomic_set(&phys_enc->pending_kickoff_cnt, 0); + init_waitqueue_head(&phys_enc->pending_kickoff_wq); + phys_enc->enable_state = DPU_ENC_DISABLED; + + DPU_DEBUG_VIDENC(vid_enc, "created intf idx:%d\n", p->intf_idx); + + return phys_enc; + +fail: + DPU_ERROR("failed to create encoder\n"); + if (vid_enc) + dpu_encoder_phys_vid_destroy(phys_enc); + + return ERR_PTR(ret); +} diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_formats.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_formats.c new file mode 100644 index 000000000000..bfcd165e96df --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_formats.c @@ -0,0 +1,1173 @@ +/* Copyright (c) 2015-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#define pr_fmt(fmt) "[drm:%s:%d] " fmt, __func__, __LINE__ + +#include <uapi/drm/drm_fourcc.h> + +#include "msm_media_info.h" +#include "dpu_kms.h" +#include "dpu_formats.h" + +#define DPU_UBWC_META_MACRO_W_H 16 +#define DPU_UBWC_META_BLOCK_SIZE 256 +#define DPU_UBWC_PLANE_SIZE_ALIGNMENT 4096 + +#define DPU_TILE_HEIGHT_DEFAULT 1 +#define DPU_TILE_HEIGHT_TILED 4 +#define DPU_TILE_HEIGHT_UBWC 4 +#define DPU_TILE_HEIGHT_NV12 8 + +#define DPU_MAX_IMG_WIDTH 0x3FFF +#define DPU_MAX_IMG_HEIGHT 0x3FFF + +/** + * DPU supported format packing, bpp, and other format + * information. + * DPU currently only supports interleaved RGB formats + * UBWC support for a pixel format is indicated by the flag, + * there is additional meta data plane for such formats + */ + +#define INTERLEAVED_RGB_FMT(fmt, a, r, g, b, e0, e1, e2, e3, uc, alpha, \ +bp, flg, fm, np) \ +{ \ + .base.pixel_format = DRM_FORMAT_ ## fmt, \ + .fetch_planes = DPU_PLANE_INTERLEAVED, \ + .alpha_enable = alpha, \ + .element = { (e0), (e1), (e2), (e3) }, \ + .bits = { g, b, r, a }, \ + .chroma_sample = DPU_CHROMA_RGB, \ + .unpack_align_msb = 0, \ + .unpack_tight = 1, \ + .unpack_count = uc, \ + .bpp = bp, \ + .fetch_mode = fm, \ + .flag = {(flg)}, \ + .num_planes = np, \ + .tile_height = DPU_TILE_HEIGHT_DEFAULT \ +} + +#define INTERLEAVED_RGB_FMT_TILED(fmt, a, r, g, b, e0, e1, e2, e3, uc, \ +alpha, bp, flg, fm, np, th) \ +{ \ + .base.pixel_format = DRM_FORMAT_ ## fmt, \ + .fetch_planes = DPU_PLANE_INTERLEAVED, \ + .alpha_enable = alpha, \ + .element = { (e0), (e1), (e2), (e3) }, \ + .bits = { g, b, r, a }, \ + .chroma_sample = DPU_CHROMA_RGB, \ + .unpack_align_msb = 0, \ + .unpack_tight = 1, \ + .unpack_count = uc, \ + .bpp = bp, \ + .fetch_mode = fm, \ + .flag = {(flg)}, \ + .num_planes = np, \ + .tile_height = th \ +} + + +#define INTERLEAVED_YUV_FMT(fmt, a, r, g, b, e0, e1, e2, e3, \ +alpha, chroma, count, bp, flg, fm, np) \ +{ \ + .base.pixel_format = DRM_FORMAT_ ## fmt, \ + .fetch_planes = DPU_PLANE_INTERLEAVED, \ + .alpha_enable = alpha, \ + .element = { (e0), (e1), (e2), (e3)}, \ + .bits = { g, b, r, a }, \ + .chroma_sample = chroma, \ + .unpack_align_msb = 0, \ + .unpack_tight = 1, \ + .unpack_count = count, \ + .bpp = bp, \ + .fetch_mode = fm, \ + .flag = {(flg)}, \ + .num_planes = np, \ + .tile_height = DPU_TILE_HEIGHT_DEFAULT \ +} + +#define PSEUDO_YUV_FMT(fmt, a, r, g, b, e0, e1, chroma, flg, fm, np) \ +{ \ + .base.pixel_format = DRM_FORMAT_ ## fmt, \ + .fetch_planes = DPU_PLANE_PSEUDO_PLANAR, \ + .alpha_enable = false, \ + .element = { (e0), (e1), 0, 0 }, \ + .bits = { g, b, r, a }, \ + .chroma_sample = chroma, \ + .unpack_align_msb = 0, \ + .unpack_tight = 1, \ + .unpack_count = 2, \ + .bpp = 2, \ + .fetch_mode = fm, \ + .flag = {(flg)}, \ + .num_planes = np, \ + .tile_height = DPU_TILE_HEIGHT_DEFAULT \ +} + +#define PSEUDO_YUV_FMT_TILED(fmt, a, r, g, b, e0, e1, chroma, \ +flg, fm, np, th) \ +{ \ + .base.pixel_format = DRM_FORMAT_ ## fmt, \ + .fetch_planes = DPU_PLANE_PSEUDO_PLANAR, \ + .alpha_enable = false, \ + .element = { (e0), (e1), 0, 0 }, \ + .bits = { g, b, r, a }, \ + .chroma_sample = chroma, \ + .unpack_align_msb = 0, \ + .unpack_tight = 1, \ + .unpack_count = 2, \ + .bpp = 2, \ + .fetch_mode = fm, \ + .flag = {(flg)}, \ + .num_planes = np, \ + .tile_height = th \ +} + +#define PSEUDO_YUV_FMT_LOOSE(fmt, a, r, g, b, e0, e1, chroma, flg, fm, np)\ +{ \ + .base.pixel_format = DRM_FORMAT_ ## fmt, \ + .fetch_planes = DPU_PLANE_PSEUDO_PLANAR, \ + .alpha_enable = false, \ + .element = { (e0), (e1), 0, 0 }, \ + .bits = { g, b, r, a }, \ + .chroma_sample = chroma, \ + .unpack_align_msb = 1, \ + .unpack_tight = 0, \ + .unpack_count = 2, \ + .bpp = 2, \ + .fetch_mode = fm, \ + .flag = {(flg)}, \ + .num_planes = np, \ + .tile_height = DPU_TILE_HEIGHT_DEFAULT \ +} + +#define PSEUDO_YUV_FMT_LOOSE_TILED(fmt, a, r, g, b, e0, e1, chroma, \ +flg, fm, np, th) \ +{ \ + .base.pixel_format = DRM_FORMAT_ ## fmt, \ + .fetch_planes = DPU_PLANE_PSEUDO_PLANAR, \ + .alpha_enable = false, \ + .element = { (e0), (e1), 0, 0 }, \ + .bits = { g, b, r, a }, \ + .chroma_sample = chroma, \ + .unpack_align_msb = 1, \ + .unpack_tight = 0, \ + .unpack_count = 2, \ + .bpp = 2, \ + .fetch_mode = fm, \ + .flag = {(flg)}, \ + .num_planes = np, \ + .tile_height = th \ +} + + +#define PLANAR_YUV_FMT(fmt, a, r, g, b, e0, e1, e2, alpha, chroma, bp, \ +flg, fm, np) \ +{ \ + .base.pixel_format = DRM_FORMAT_ ## fmt, \ + .fetch_planes = DPU_PLANE_PLANAR, \ + .alpha_enable = alpha, \ + .element = { (e0), (e1), (e2), 0 }, \ + .bits = { g, b, r, a }, \ + .chroma_sample = chroma, \ + .unpack_align_msb = 0, \ + .unpack_tight = 1, \ + .unpack_count = 1, \ + .bpp = bp, \ + .fetch_mode = fm, \ + .flag = {(flg)}, \ + .num_planes = np, \ + .tile_height = DPU_TILE_HEIGHT_DEFAULT \ +} + +/* + * struct dpu_media_color_map - maps drm format to media format + * @format: DRM base pixel format + * @color: Media API color related to DRM format + */ +struct dpu_media_color_map { + uint32_t format; + uint32_t color; +}; + +static const struct dpu_format dpu_format_map[] = { + INTERLEAVED_RGB_FMT(ARGB8888, + COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C1_B_Cb, C0_G_Y, C2_R_Cr, C3_ALPHA, 4, + true, 4, 0, + DPU_FETCH_LINEAR, 1), + + INTERLEAVED_RGB_FMT(ABGR8888, + COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C2_R_Cr, C0_G_Y, C1_B_Cb, C3_ALPHA, 4, + true, 4, 0, + DPU_FETCH_LINEAR, 1), + + INTERLEAVED_RGB_FMT(XBGR8888, + COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C2_R_Cr, C0_G_Y, C1_B_Cb, C3_ALPHA, 4, + true, 4, 0, + DPU_FETCH_LINEAR, 1), + + INTERLEAVED_RGB_FMT(RGBA8888, + COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C3_ALPHA, C1_B_Cb, C0_G_Y, C2_R_Cr, 4, + true, 4, 0, + DPU_FETCH_LINEAR, 1), + + INTERLEAVED_RGB_FMT(BGRA8888, + COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C3_ALPHA, C2_R_Cr, C0_G_Y, C1_B_Cb, 4, + true, 4, 0, + DPU_FETCH_LINEAR, 1), + + INTERLEAVED_RGB_FMT(BGRX8888, + COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C3_ALPHA, C2_R_Cr, C0_G_Y, C1_B_Cb, 4, + false, 4, 0, + DPU_FETCH_LINEAR, 1), + + INTERLEAVED_RGB_FMT(XRGB8888, + COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C1_B_Cb, C0_G_Y, C2_R_Cr, C3_ALPHA, 4, + false, 4, 0, + DPU_FETCH_LINEAR, 1), + + INTERLEAVED_RGB_FMT(RGBX8888, + COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C3_ALPHA, C1_B_Cb, C0_G_Y, C2_R_Cr, 4, + false, 4, 0, + DPU_FETCH_LINEAR, 1), + + INTERLEAVED_RGB_FMT(RGB888, + 0, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C1_B_Cb, C0_G_Y, C2_R_Cr, 0, 3, + false, 3, 0, + DPU_FETCH_LINEAR, 1), + + INTERLEAVED_RGB_FMT(BGR888, + 0, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C2_R_Cr, C0_G_Y, C1_B_Cb, 0, 3, + false, 3, 0, + DPU_FETCH_LINEAR, 1), + + INTERLEAVED_RGB_FMT(RGB565, + 0, COLOR_5BIT, COLOR_6BIT, COLOR_5BIT, + C1_B_Cb, C0_G_Y, C2_R_Cr, 0, 3, + false, 2, 0, + DPU_FETCH_LINEAR, 1), + + INTERLEAVED_RGB_FMT(BGR565, + 0, COLOR_5BIT, COLOR_6BIT, COLOR_5BIT, + C2_R_Cr, C0_G_Y, C1_B_Cb, 0, 3, + false, 2, 0, + DPU_FETCH_LINEAR, 1), + + INTERLEAVED_RGB_FMT(ARGB1555, + COLOR_ALPHA_1BIT, COLOR_5BIT, COLOR_5BIT, COLOR_5BIT, + C1_B_Cb, C0_G_Y, C2_R_Cr, C3_ALPHA, 4, + true, 2, 0, + DPU_FETCH_LINEAR, 1), + + INTERLEAVED_RGB_FMT(ABGR1555, + COLOR_ALPHA_1BIT, COLOR_5BIT, COLOR_5BIT, COLOR_5BIT, + C2_R_Cr, C0_G_Y, C1_B_Cb, C3_ALPHA, 4, + true, 2, 0, + DPU_FETCH_LINEAR, 1), + + INTERLEAVED_RGB_FMT(RGBA5551, + COLOR_ALPHA_1BIT, COLOR_5BIT, COLOR_5BIT, COLOR_5BIT, + C3_ALPHA, C1_B_Cb, C0_G_Y, C2_R_Cr, 4, + true, 2, 0, + DPU_FETCH_LINEAR, 1), + + INTERLEAVED_RGB_FMT(BGRA5551, + COLOR_ALPHA_1BIT, COLOR_5BIT, COLOR_5BIT, COLOR_5BIT, + C3_ALPHA, C2_R_Cr, C0_G_Y, C1_B_Cb, 4, + true, 2, 0, + DPU_FETCH_LINEAR, 1), + + INTERLEAVED_RGB_FMT(XRGB1555, + COLOR_ALPHA_1BIT, COLOR_5BIT, COLOR_5BIT, COLOR_5BIT, + C1_B_Cb, C0_G_Y, C2_R_Cr, C3_ALPHA, 4, + false, 2, 0, + DPU_FETCH_LINEAR, 1), + + INTERLEAVED_RGB_FMT(XBGR1555, + COLOR_ALPHA_1BIT, COLOR_5BIT, COLOR_5BIT, COLOR_5BIT, + C2_R_Cr, C0_G_Y, C1_B_Cb, C3_ALPHA, 4, + false, 2, 0, + DPU_FETCH_LINEAR, 1), + + INTERLEAVED_RGB_FMT(RGBX5551, + COLOR_ALPHA_1BIT, COLOR_5BIT, COLOR_5BIT, COLOR_5BIT, + C3_ALPHA, C1_B_Cb, C0_G_Y, C2_R_Cr, 4, + false, 2, 0, + DPU_FETCH_LINEAR, 1), + + INTERLEAVED_RGB_FMT(BGRX5551, + COLOR_ALPHA_1BIT, COLOR_5BIT, COLOR_5BIT, COLOR_5BIT, + C3_ALPHA, C2_R_Cr, C0_G_Y, C1_B_Cb, 4, + false, 2, 0, + DPU_FETCH_LINEAR, 1), + + INTERLEAVED_RGB_FMT(ARGB4444, + COLOR_ALPHA_4BIT, COLOR_4BIT, COLOR_4BIT, COLOR_4BIT, + C1_B_Cb, C0_G_Y, C2_R_Cr, C3_ALPHA, 4, + true, 2, 0, + DPU_FETCH_LINEAR, 1), + + INTERLEAVED_RGB_FMT(ABGR4444, + COLOR_ALPHA_4BIT, COLOR_4BIT, COLOR_4BIT, COLOR_4BIT, + C2_R_Cr, C0_G_Y, C1_B_Cb, C3_ALPHA, 4, + true, 2, 0, + DPU_FETCH_LINEAR, 1), + + INTERLEAVED_RGB_FMT(RGBA4444, + COLOR_ALPHA_4BIT, COLOR_4BIT, COLOR_4BIT, COLOR_4BIT, + C3_ALPHA, C1_B_Cb, C0_G_Y, C2_R_Cr, 4, + true, 2, 0, + DPU_FETCH_LINEAR, 1), + + INTERLEAVED_RGB_FMT(BGRA4444, + COLOR_ALPHA_4BIT, COLOR_4BIT, COLOR_4BIT, COLOR_4BIT, + C3_ALPHA, C2_R_Cr, C0_G_Y, C1_B_Cb, 4, + true, 2, 0, + DPU_FETCH_LINEAR, 1), + + INTERLEAVED_RGB_FMT(XRGB4444, + COLOR_ALPHA_4BIT, COLOR_4BIT, COLOR_4BIT, COLOR_4BIT, + C1_B_Cb, C0_G_Y, C2_R_Cr, C3_ALPHA, 4, + false, 2, 0, + DPU_FETCH_LINEAR, 1), + + INTERLEAVED_RGB_FMT(XBGR4444, + COLOR_ALPHA_4BIT, COLOR_4BIT, COLOR_4BIT, COLOR_4BIT, + C2_R_Cr, C0_G_Y, C1_B_Cb, C3_ALPHA, 4, + false, 2, 0, + DPU_FETCH_LINEAR, 1), + + INTERLEAVED_RGB_FMT(RGBX4444, + COLOR_ALPHA_4BIT, COLOR_4BIT, COLOR_4BIT, COLOR_4BIT, + C3_ALPHA, C1_B_Cb, C0_G_Y, C2_R_Cr, 4, + false, 2, 0, + DPU_FETCH_LINEAR, 1), + + INTERLEAVED_RGB_FMT(BGRX4444, + COLOR_ALPHA_4BIT, COLOR_4BIT, COLOR_4BIT, COLOR_4BIT, + C3_ALPHA, C2_R_Cr, C0_G_Y, C1_B_Cb, 4, + false, 2, 0, + DPU_FETCH_LINEAR, 1), + + INTERLEAVED_RGB_FMT(BGRA1010102, + COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C3_ALPHA, C2_R_Cr, C0_G_Y, C1_B_Cb, 4, + true, 4, DPU_FORMAT_FLAG_DX, + DPU_FETCH_LINEAR, 1), + + INTERLEAVED_RGB_FMT(RGBA1010102, + COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C3_ALPHA, C1_B_Cb, C0_G_Y, C2_R_Cr, 4, + true, 4, DPU_FORMAT_FLAG_DX, + DPU_FETCH_LINEAR, 1), + + INTERLEAVED_RGB_FMT(ABGR2101010, + COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C2_R_Cr, C0_G_Y, C1_B_Cb, C3_ALPHA, 4, + true, 4, DPU_FORMAT_FLAG_DX, + DPU_FETCH_LINEAR, 1), + + INTERLEAVED_RGB_FMT(ARGB2101010, + COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C1_B_Cb, C0_G_Y, C2_R_Cr, C3_ALPHA, 4, + true, 4, DPU_FORMAT_FLAG_DX, + DPU_FETCH_LINEAR, 1), + + INTERLEAVED_RGB_FMT(XRGB2101010, + COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C1_B_Cb, C0_G_Y, C2_R_Cr, C3_ALPHA, 4, + false, 4, DPU_FORMAT_FLAG_DX, + DPU_FETCH_LINEAR, 1), + + INTERLEAVED_RGB_FMT(BGRX1010102, + COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C3_ALPHA, C2_R_Cr, C0_G_Y, C1_B_Cb, 4, + false, 4, DPU_FORMAT_FLAG_DX, + DPU_FETCH_LINEAR, 1), + + INTERLEAVED_RGB_FMT(XBGR2101010, + COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C2_R_Cr, C0_G_Y, C1_B_Cb, C3_ALPHA, 4, + false, 4, DPU_FORMAT_FLAG_DX, + DPU_FETCH_LINEAR, 1), + + INTERLEAVED_RGB_FMT(RGBX1010102, + COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C3_ALPHA, C1_B_Cb, C0_G_Y, C2_R_Cr, 4, + false, 4, DPU_FORMAT_FLAG_DX, + DPU_FETCH_LINEAR, 1), + + PSEUDO_YUV_FMT(NV12, + 0, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C1_B_Cb, C2_R_Cr, + DPU_CHROMA_420, DPU_FORMAT_FLAG_YUV, + DPU_FETCH_LINEAR, 2), + + PSEUDO_YUV_FMT(NV21, + 0, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C2_R_Cr, C1_B_Cb, + DPU_CHROMA_420, DPU_FORMAT_FLAG_YUV, + DPU_FETCH_LINEAR, 2), + + PSEUDO_YUV_FMT(NV16, + 0, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C1_B_Cb, C2_R_Cr, + DPU_CHROMA_H2V1, DPU_FORMAT_FLAG_YUV, + DPU_FETCH_LINEAR, 2), + + PSEUDO_YUV_FMT(NV61, + 0, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C2_R_Cr, C1_B_Cb, + DPU_CHROMA_H2V1, DPU_FORMAT_FLAG_YUV, + DPU_FETCH_LINEAR, 2), + + INTERLEAVED_YUV_FMT(VYUY, + 0, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C2_R_Cr, C0_G_Y, C1_B_Cb, C0_G_Y, + false, DPU_CHROMA_H2V1, 4, 2, DPU_FORMAT_FLAG_YUV, + DPU_FETCH_LINEAR, 2), + + INTERLEAVED_YUV_FMT(UYVY, + 0, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C1_B_Cb, C0_G_Y, C2_R_Cr, C0_G_Y, + false, DPU_CHROMA_H2V1, 4, 2, DPU_FORMAT_FLAG_YUV, + DPU_FETCH_LINEAR, 2), + + INTERLEAVED_YUV_FMT(YUYV, + 0, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C0_G_Y, C1_B_Cb, C0_G_Y, C2_R_Cr, + false, DPU_CHROMA_H2V1, 4, 2, DPU_FORMAT_FLAG_YUV, + DPU_FETCH_LINEAR, 2), + + INTERLEAVED_YUV_FMT(YVYU, + 0, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C0_G_Y, C2_R_Cr, C0_G_Y, C1_B_Cb, + false, DPU_CHROMA_H2V1, 4, 2, DPU_FORMAT_FLAG_YUV, + DPU_FETCH_LINEAR, 2), + + PLANAR_YUV_FMT(YUV420, + 0, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C2_R_Cr, C1_B_Cb, C0_G_Y, + false, DPU_CHROMA_420, 1, DPU_FORMAT_FLAG_YUV, + DPU_FETCH_LINEAR, 3), + + PLANAR_YUV_FMT(YVU420, + 0, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C1_B_Cb, C2_R_Cr, C0_G_Y, + false, DPU_CHROMA_420, 1, DPU_FORMAT_FLAG_YUV, + DPU_FETCH_LINEAR, 3), +}; + +/* + * A5x tile formats tables: + * These tables hold the A5x tile formats supported. + */ +static const struct dpu_format dpu_format_map_tile[] = { + INTERLEAVED_RGB_FMT_TILED(BGR565, + 0, COLOR_5BIT, COLOR_6BIT, COLOR_5BIT, + C2_R_Cr, C0_G_Y, C1_B_Cb, 0, 3, + false, 2, 0, + DPU_FETCH_UBWC, 1, DPU_TILE_HEIGHT_TILED), + + INTERLEAVED_RGB_FMT_TILED(ARGB8888, + COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C3_ALPHA, C2_R_Cr, C0_G_Y, C1_B_Cb, 4, + true, 4, 0, + DPU_FETCH_UBWC, 1, DPU_TILE_HEIGHT_TILED), + + INTERLEAVED_RGB_FMT_TILED(ABGR8888, + COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C3_ALPHA, C1_B_Cb, C0_G_Y, C2_R_Cr, 4, + true, 4, 0, + DPU_FETCH_UBWC, 1, DPU_TILE_HEIGHT_TILED), + + INTERLEAVED_RGB_FMT_TILED(XBGR8888, + COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C2_R_Cr, C0_G_Y, C1_B_Cb, C3_ALPHA, 4, + false, 4, 0, + DPU_FETCH_UBWC, 1, DPU_TILE_HEIGHT_TILED), + + INTERLEAVED_RGB_FMT_TILED(RGBA8888, + COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C2_R_Cr, C0_G_Y, C1_B_Cb, C3_ALPHA, 4, + true, 4, 0, + DPU_FETCH_UBWC, 1, DPU_TILE_HEIGHT_TILED), + + INTERLEAVED_RGB_FMT_TILED(BGRA8888, + COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C1_B_Cb, C0_G_Y, C2_R_Cr, C3_ALPHA, 4, + true, 4, 0, + DPU_FETCH_UBWC, 1, DPU_TILE_HEIGHT_TILED), + + INTERLEAVED_RGB_FMT_TILED(BGRX8888, + COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C1_B_Cb, C0_G_Y, C2_R_Cr, C3_ALPHA, 4, + false, 4, 0, + DPU_FETCH_UBWC, 1, DPU_TILE_HEIGHT_TILED), + + INTERLEAVED_RGB_FMT_TILED(XRGB8888, + COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C3_ALPHA, C2_R_Cr, C0_G_Y, C1_B_Cb, 4, + false, 4, 0, + DPU_FETCH_UBWC, 1, DPU_TILE_HEIGHT_TILED), + + INTERLEAVED_RGB_FMT_TILED(RGBX8888, + COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C2_R_Cr, C0_G_Y, C1_B_Cb, C3_ALPHA, 4, + false, 4, 0, + DPU_FETCH_UBWC, 1, DPU_TILE_HEIGHT_TILED), + + INTERLEAVED_RGB_FMT_TILED(ABGR2101010, + COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C2_R_Cr, C0_G_Y, C1_B_Cb, C3_ALPHA, 4, + true, 4, DPU_FORMAT_FLAG_DX, + DPU_FETCH_UBWC, 1, DPU_TILE_HEIGHT_TILED), + + INTERLEAVED_RGB_FMT_TILED(XBGR2101010, + COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C2_R_Cr, C0_G_Y, C1_B_Cb, C3_ALPHA, 4, + true, 4, DPU_FORMAT_FLAG_DX, + DPU_FETCH_UBWC, 1, DPU_TILE_HEIGHT_TILED), + + PSEUDO_YUV_FMT_TILED(NV12, + 0, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C1_B_Cb, C2_R_Cr, + DPU_CHROMA_420, DPU_FORMAT_FLAG_YUV, + DPU_FETCH_UBWC, 2, DPU_TILE_HEIGHT_NV12), + + PSEUDO_YUV_FMT_TILED(NV21, + 0, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C2_R_Cr, C1_B_Cb, + DPU_CHROMA_420, DPU_FORMAT_FLAG_YUV, + DPU_FETCH_UBWC, 2, DPU_TILE_HEIGHT_NV12), +}; + +/* + * UBWC formats table: + * This table holds the UBWC formats supported. + * If a compression ratio needs to be used for this or any other format, + * the data will be passed by user-space. + */ +static const struct dpu_format dpu_format_map_ubwc[] = { + INTERLEAVED_RGB_FMT_TILED(BGR565, + 0, COLOR_5BIT, COLOR_6BIT, COLOR_5BIT, + C2_R_Cr, C0_G_Y, C1_B_Cb, 0, 3, + false, 2, DPU_FORMAT_FLAG_COMPRESSED, + DPU_FETCH_UBWC, 2, DPU_TILE_HEIGHT_UBWC), + + INTERLEAVED_RGB_FMT_TILED(ABGR8888, + COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C2_R_Cr, C0_G_Y, C1_B_Cb, C3_ALPHA, 4, + true, 4, DPU_FORMAT_FLAG_COMPRESSED, + DPU_FETCH_UBWC, 2, DPU_TILE_HEIGHT_UBWC), + + INTERLEAVED_RGB_FMT_TILED(XBGR8888, + COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C2_R_Cr, C0_G_Y, C1_B_Cb, C3_ALPHA, 4, + false, 4, DPU_FORMAT_FLAG_COMPRESSED, + DPU_FETCH_UBWC, 2, DPU_TILE_HEIGHT_UBWC), + + INTERLEAVED_RGB_FMT_TILED(ABGR2101010, + COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C2_R_Cr, C0_G_Y, C1_B_Cb, C3_ALPHA, 4, + true, 4, DPU_FORMAT_FLAG_DX | DPU_FORMAT_FLAG_COMPRESSED, + DPU_FETCH_UBWC, 2, DPU_TILE_HEIGHT_UBWC), + + INTERLEAVED_RGB_FMT_TILED(XBGR2101010, + COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C2_R_Cr, C0_G_Y, C1_B_Cb, C3_ALPHA, 4, + true, 4, DPU_FORMAT_FLAG_DX | DPU_FORMAT_FLAG_COMPRESSED, + DPU_FETCH_UBWC, 2, DPU_TILE_HEIGHT_UBWC), + + PSEUDO_YUV_FMT_TILED(NV12, + 0, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C1_B_Cb, C2_R_Cr, + DPU_CHROMA_420, DPU_FORMAT_FLAG_YUV | + DPU_FORMAT_FLAG_COMPRESSED, + DPU_FETCH_UBWC, 4, DPU_TILE_HEIGHT_NV12), +}; + +static const struct dpu_format dpu_format_map_p010[] = { + PSEUDO_YUV_FMT_LOOSE(NV12, + 0, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C1_B_Cb, C2_R_Cr, + DPU_CHROMA_420, (DPU_FORMAT_FLAG_YUV | DPU_FORMAT_FLAG_DX), + DPU_FETCH_LINEAR, 2), +}; + +static const struct dpu_format dpu_format_map_p010_ubwc[] = { + PSEUDO_YUV_FMT_LOOSE_TILED(NV12, + 0, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C1_B_Cb, C2_R_Cr, + DPU_CHROMA_420, (DPU_FORMAT_FLAG_YUV | DPU_FORMAT_FLAG_DX | + DPU_FORMAT_FLAG_COMPRESSED), + DPU_FETCH_UBWC, 4, DPU_TILE_HEIGHT_NV12), +}; + +static const struct dpu_format dpu_format_map_tp10_ubwc[] = { + PSEUDO_YUV_FMT_TILED(NV12, + 0, COLOR_8BIT, COLOR_8BIT, COLOR_8BIT, + C1_B_Cb, C2_R_Cr, + DPU_CHROMA_420, (DPU_FORMAT_FLAG_YUV | DPU_FORMAT_FLAG_DX | + DPU_FORMAT_FLAG_COMPRESSED), + DPU_FETCH_UBWC, 4, DPU_TILE_HEIGHT_NV12), +}; + +/* _dpu_get_v_h_subsample_rate - Get subsample rates for all formats we support + * Note: Not using the drm_format_*_subsampling since we have formats + */ +static void _dpu_get_v_h_subsample_rate( + enum dpu_chroma_samp_type chroma_sample, + uint32_t *v_sample, + uint32_t *h_sample) +{ + if (!v_sample || !h_sample) + return; + + switch (chroma_sample) { + case DPU_CHROMA_H2V1: + *v_sample = 1; + *h_sample = 2; + break; + case DPU_CHROMA_H1V2: + *v_sample = 2; + *h_sample = 1; + break; + case DPU_CHROMA_420: + *v_sample = 2; + *h_sample = 2; + break; + default: + *v_sample = 1; + *h_sample = 1; + break; + } +} + +static int _dpu_format_get_media_color_ubwc(const struct dpu_format *fmt) +{ + static const struct dpu_media_color_map dpu_media_ubwc_map[] = { + {DRM_FORMAT_ABGR8888, COLOR_FMT_RGBA8888_UBWC}, + {DRM_FORMAT_XBGR8888, COLOR_FMT_RGBA8888_UBWC}, + {DRM_FORMAT_ABGR2101010, COLOR_FMT_RGBA1010102_UBWC}, + {DRM_FORMAT_XBGR2101010, COLOR_FMT_RGBA1010102_UBWC}, + {DRM_FORMAT_BGR565, COLOR_FMT_RGB565_UBWC}, + }; + int color_fmt = -1; + int i; + + if (fmt->base.pixel_format == DRM_FORMAT_NV12) { + if (DPU_FORMAT_IS_DX(fmt)) { + if (fmt->unpack_tight) + color_fmt = COLOR_FMT_NV12_BPP10_UBWC; + else + color_fmt = COLOR_FMT_P010_UBWC; + } else + color_fmt = COLOR_FMT_NV12_UBWC; + return color_fmt; + } + + for (i = 0; i < ARRAY_SIZE(dpu_media_ubwc_map); ++i) + if (fmt->base.pixel_format == dpu_media_ubwc_map[i].format) { + color_fmt = dpu_media_ubwc_map[i].color; + break; + } + return color_fmt; +} + +static int _dpu_format_get_plane_sizes_ubwc( + const struct dpu_format *fmt, + const uint32_t width, + const uint32_t height, + struct dpu_hw_fmt_layout *layout) +{ + int i; + int color; + bool meta = DPU_FORMAT_IS_UBWC(fmt); + + memset(layout, 0, sizeof(struct dpu_hw_fmt_layout)); + layout->format = fmt; + layout->width = width; + layout->height = height; + layout->num_planes = fmt->num_planes; + + color = _dpu_format_get_media_color_ubwc(fmt); + if (color < 0) { + DRM_ERROR("UBWC format not supported for fmt: %4.4s\n", + (char *)&fmt->base.pixel_format); + return -EINVAL; + } + + if (DPU_FORMAT_IS_YUV(layout->format)) { + uint32_t y_sclines, uv_sclines; + uint32_t y_meta_scanlines = 0; + uint32_t uv_meta_scanlines = 0; + + layout->num_planes = 2; + layout->plane_pitch[0] = VENUS_Y_STRIDE(color, width); + y_sclines = VENUS_Y_SCANLINES(color, height); + layout->plane_size[0] = MSM_MEDIA_ALIGN(layout->plane_pitch[0] * + y_sclines, DPU_UBWC_PLANE_SIZE_ALIGNMENT); + + layout->plane_pitch[1] = VENUS_UV_STRIDE(color, width); + uv_sclines = VENUS_UV_SCANLINES(color, height); + layout->plane_size[1] = MSM_MEDIA_ALIGN(layout->plane_pitch[1] * + uv_sclines, DPU_UBWC_PLANE_SIZE_ALIGNMENT); + + if (!meta) + goto done; + + layout->num_planes += 2; + layout->plane_pitch[2] = VENUS_Y_META_STRIDE(color, width); + y_meta_scanlines = VENUS_Y_META_SCANLINES(color, height); + layout->plane_size[2] = MSM_MEDIA_ALIGN(layout->plane_pitch[2] * + y_meta_scanlines, DPU_UBWC_PLANE_SIZE_ALIGNMENT); + + layout->plane_pitch[3] = VENUS_UV_META_STRIDE(color, width); + uv_meta_scanlines = VENUS_UV_META_SCANLINES(color, height); + layout->plane_size[3] = MSM_MEDIA_ALIGN(layout->plane_pitch[3] * + uv_meta_scanlines, DPU_UBWC_PLANE_SIZE_ALIGNMENT); + + } else { + uint32_t rgb_scanlines, rgb_meta_scanlines; + + layout->num_planes = 1; + + layout->plane_pitch[0] = VENUS_RGB_STRIDE(color, width); + rgb_scanlines = VENUS_RGB_SCANLINES(color, height); + layout->plane_size[0] = MSM_MEDIA_ALIGN(layout->plane_pitch[0] * + rgb_scanlines, DPU_UBWC_PLANE_SIZE_ALIGNMENT); + + if (!meta) + goto done; + layout->num_planes += 2; + layout->plane_pitch[2] = VENUS_RGB_META_STRIDE(color, width); + rgb_meta_scanlines = VENUS_RGB_META_SCANLINES(color, height); + layout->plane_size[2] = MSM_MEDIA_ALIGN(layout->plane_pitch[2] * + rgb_meta_scanlines, DPU_UBWC_PLANE_SIZE_ALIGNMENT); + } + +done: + for (i = 0; i < DPU_MAX_PLANES; i++) + layout->total_size += layout->plane_size[i]; + + return 0; +} + +static int _dpu_format_get_plane_sizes_linear( + const struct dpu_format *fmt, + const uint32_t width, + const uint32_t height, + struct dpu_hw_fmt_layout *layout, + const uint32_t *pitches) +{ + int i; + + memset(layout, 0, sizeof(struct dpu_hw_fmt_layout)); + layout->format = fmt; + layout->width = width; + layout->height = height; + layout->num_planes = fmt->num_planes; + + /* Due to memset above, only need to set planes of interest */ + if (fmt->fetch_planes == DPU_PLANE_INTERLEAVED) { + layout->num_planes = 1; + layout->plane_size[0] = width * height * layout->format->bpp; + layout->plane_pitch[0] = width * layout->format->bpp; + } else { + uint32_t v_subsample, h_subsample; + uint32_t chroma_samp; + uint32_t bpp = 1; + + chroma_samp = fmt->chroma_sample; + _dpu_get_v_h_subsample_rate(chroma_samp, &v_subsample, + &h_subsample); + + if (width % h_subsample || height % v_subsample) { + DRM_ERROR("mismatch in subsample vs dimensions\n"); + return -EINVAL; + } + + if ((fmt->base.pixel_format == DRM_FORMAT_NV12) && + (DPU_FORMAT_IS_DX(fmt))) + bpp = 2; + layout->plane_pitch[0] = width * bpp; + layout->plane_pitch[1] = layout->plane_pitch[0] / h_subsample; + layout->plane_size[0] = layout->plane_pitch[0] * height; + layout->plane_size[1] = layout->plane_pitch[1] * + (height / v_subsample); + + if (fmt->fetch_planes == DPU_PLANE_PSEUDO_PLANAR) { + layout->num_planes = 2; + layout->plane_size[1] *= 2; + layout->plane_pitch[1] *= 2; + } else { + /* planar */ + layout->num_planes = 3; + layout->plane_size[2] = layout->plane_size[1]; + layout->plane_pitch[2] = layout->plane_pitch[1]; + } + } + + /* + * linear format: allow user allocated pitches if they are greater than + * the requirement. + * ubwc format: pitch values are computed uniformly across + * all the components based on ubwc specifications. + */ + for (i = 0; i < layout->num_planes && i < DPU_MAX_PLANES; ++i) { + if (pitches && layout->plane_pitch[i] < pitches[i]) + layout->plane_pitch[i] = pitches[i]; + } + + for (i = 0; i < DPU_MAX_PLANES; i++) + layout->total_size += layout->plane_size[i]; + + return 0; +} + +static int dpu_format_get_plane_sizes( + const struct dpu_format *fmt, + const uint32_t w, + const uint32_t h, + struct dpu_hw_fmt_layout *layout, + const uint32_t *pitches) +{ + if (!layout || !fmt) { + DRM_ERROR("invalid pointer\n"); + return -EINVAL; + } + + if ((w > DPU_MAX_IMG_WIDTH) || (h > DPU_MAX_IMG_HEIGHT)) { + DRM_ERROR("image dimensions outside max range\n"); + return -ERANGE; + } + + if (DPU_FORMAT_IS_UBWC(fmt) || DPU_FORMAT_IS_TILE(fmt)) + return _dpu_format_get_plane_sizes_ubwc(fmt, w, h, layout); + + return _dpu_format_get_plane_sizes_linear(fmt, w, h, layout, pitches); +} + +static int _dpu_format_populate_addrs_ubwc( + struct msm_gem_address_space *aspace, + struct drm_framebuffer *fb, + struct dpu_hw_fmt_layout *layout) +{ + uint32_t base_addr = 0; + bool meta; + + if (!fb || !layout) { + DRM_ERROR("invalid pointers\n"); + return -EINVAL; + } + + if (aspace) + base_addr = msm_framebuffer_iova(fb, aspace, 0); + if (!base_addr) { + DRM_ERROR("failed to retrieve base addr\n"); + return -EFAULT; + } + + meta = DPU_FORMAT_IS_UBWC(layout->format); + + /* Per-format logic for verifying active planes */ + if (DPU_FORMAT_IS_YUV(layout->format)) { + /************************************************/ + /* UBWC ** */ + /* buffer ** DPU PLANE */ + /* format ** */ + /************************************************/ + /* ------------------- ** -------------------- */ + /* | Y meta | ** | Y bitstream | */ + /* | data | ** | plane | */ + /* ------------------- ** -------------------- */ + /* | Y bitstream | ** | CbCr bitstream | */ + /* | data | ** | plane | */ + /* ------------------- ** -------------------- */ + /* | Cbcr metadata | ** | Y meta | */ + /* | data | ** | plane | */ + /* ------------------- ** -------------------- */ + /* | CbCr bitstream | ** | CbCr meta | */ + /* | data | ** | plane | */ + /* ------------------- ** -------------------- */ + /************************************************/ + + /* configure Y bitstream plane */ + layout->plane_addr[0] = base_addr + layout->plane_size[2]; + + /* configure CbCr bitstream plane */ + layout->plane_addr[1] = base_addr + layout->plane_size[0] + + layout->plane_size[2] + layout->plane_size[3]; + + if (!meta) + goto done; + + /* configure Y metadata plane */ + layout->plane_addr[2] = base_addr; + + /* configure CbCr metadata plane */ + layout->plane_addr[3] = base_addr + layout->plane_size[0] + + layout->plane_size[2]; + + } else { + /************************************************/ + /* UBWC ** */ + /* buffer ** DPU PLANE */ + /* format ** */ + /************************************************/ + /* ------------------- ** -------------------- */ + /* | RGB meta | ** | RGB bitstream | */ + /* | data | ** | plane | */ + /* ------------------- ** -------------------- */ + /* | RGB bitstream | ** | NONE | */ + /* | data | ** | | */ + /* ------------------- ** -------------------- */ + /* ** | RGB meta | */ + /* ** | plane | */ + /* ** -------------------- */ + /************************************************/ + + layout->plane_addr[0] = base_addr + layout->plane_size[2]; + layout->plane_addr[1] = 0; + + if (!meta) + goto done; + + layout->plane_addr[2] = base_addr; + layout->plane_addr[3] = 0; + } +done: + return 0; +} + +static int _dpu_format_populate_addrs_linear( + struct msm_gem_address_space *aspace, + struct drm_framebuffer *fb, + struct dpu_hw_fmt_layout *layout) +{ + unsigned int i; + + /* Can now check the pitches given vs pitches expected */ + for (i = 0; i < layout->num_planes; ++i) { + if (layout->plane_pitch[i] > fb->pitches[i]) { + DRM_ERROR("plane %u expected pitch %u, fb %u\n", + i, layout->plane_pitch[i], fb->pitches[i]); + return -EINVAL; + } + } + + /* Populate addresses for simple formats here */ + for (i = 0; i < layout->num_planes; ++i) { + if (aspace) + layout->plane_addr[i] = + msm_framebuffer_iova(fb, aspace, i); + if (!layout->plane_addr[i]) { + DRM_ERROR("failed to retrieve base addr\n"); + return -EFAULT; + } + } + + return 0; +} + +int dpu_format_populate_layout( + struct msm_gem_address_space *aspace, + struct drm_framebuffer *fb, + struct dpu_hw_fmt_layout *layout) +{ + uint32_t plane_addr[DPU_MAX_PLANES]; + int i, ret; + + if (!fb || !layout) { + DRM_ERROR("invalid arguments\n"); + return -EINVAL; + } + + if ((fb->width > DPU_MAX_IMG_WIDTH) || + (fb->height > DPU_MAX_IMG_HEIGHT)) { + DRM_ERROR("image dimensions outside max range\n"); + return -ERANGE; + } + + layout->format = to_dpu_format(msm_framebuffer_format(fb)); + + /* Populate the plane sizes etc via get_format */ + ret = dpu_format_get_plane_sizes(layout->format, fb->width, fb->height, + layout, fb->pitches); + if (ret) + return ret; + + for (i = 0; i < DPU_MAX_PLANES; ++i) + plane_addr[i] = layout->plane_addr[i]; + + /* Populate the addresses given the fb */ + if (DPU_FORMAT_IS_UBWC(layout->format) || + DPU_FORMAT_IS_TILE(layout->format)) + ret = _dpu_format_populate_addrs_ubwc(aspace, fb, layout); + else + ret = _dpu_format_populate_addrs_linear(aspace, fb, layout); + + /* check if anything changed */ + if (!ret && !memcmp(plane_addr, layout->plane_addr, sizeof(plane_addr))) + ret = -EAGAIN; + + return ret; +} + +int dpu_format_check_modified_format( + const struct msm_kms *kms, + const struct msm_format *msm_fmt, + const struct drm_mode_fb_cmd2 *cmd, + struct drm_gem_object **bos) +{ + int ret, i, num_base_fmt_planes; + const struct dpu_format *fmt; + struct dpu_hw_fmt_layout layout; + uint32_t bos_total_size = 0; + + if (!msm_fmt || !cmd || !bos) { + DRM_ERROR("invalid arguments\n"); + return -EINVAL; + } + + fmt = to_dpu_format(msm_fmt); + num_base_fmt_planes = drm_format_num_planes(fmt->base.pixel_format); + + ret = dpu_format_get_plane_sizes(fmt, cmd->width, cmd->height, + &layout, cmd->pitches); + if (ret) + return ret; + + for (i = 0; i < num_base_fmt_planes; i++) { + if (!bos[i]) { + DRM_ERROR("invalid handle for plane %d\n", i); + return -EINVAL; + } + if ((i == 0) || (bos[i] != bos[0])) + bos_total_size += bos[i]->size; + } + + if (bos_total_size < layout.total_size) { + DRM_ERROR("buffers total size too small %u expected %u\n", + bos_total_size, layout.total_size); + return -EINVAL; + } + + return 0; +} + +const struct dpu_format *dpu_get_dpu_format_ext( + const uint32_t format, + const uint64_t modifier) +{ + uint32_t i = 0; + const struct dpu_format *fmt = NULL; + const struct dpu_format *map = NULL; + ssize_t map_size = 0; + + /* + * Currently only support exactly zero or one modifier. + * All planes use the same modifier. + */ + DPU_DEBUG("plane format modifier 0x%llX\n", modifier); + + switch (modifier) { + case 0: + map = dpu_format_map; + map_size = ARRAY_SIZE(dpu_format_map); + break; + case DRM_FORMAT_MOD_QCOM_COMPRESSED: + map = dpu_format_map_ubwc; + map_size = ARRAY_SIZE(dpu_format_map_ubwc); + DPU_DEBUG("found fmt: %4.4s DRM_FORMAT_MOD_QCOM_COMPRESSED\n", + (char *)&format); + break; + default: + DPU_ERROR("unsupported format modifier %llX\n", modifier); + return NULL; + } + + for (i = 0; i < map_size; i++) { + if (format == map[i].base.pixel_format) { + fmt = &map[i]; + break; + } + } + + if (fmt == NULL) + DPU_ERROR("unsupported fmt: %4.4s modifier 0x%llX\n", + (char *)&format, modifier); + else + DPU_DEBUG("fmt %4.4s mod 0x%llX ubwc %d yuv %d\n", + (char *)&format, modifier, + DPU_FORMAT_IS_UBWC(fmt), + DPU_FORMAT_IS_YUV(fmt)); + + return fmt; +} + +const struct msm_format *dpu_get_msm_format( + struct msm_kms *kms, + const uint32_t format, + const uint64_t modifiers) +{ + const struct dpu_format *fmt = dpu_get_dpu_format_ext(format, + modifiers); + if (fmt) + return &fmt->base; + return NULL; +} + +uint32_t dpu_populate_formats( + const struct dpu_format_extended *format_list, + uint32_t *pixel_formats, + uint64_t *pixel_modifiers, + uint32_t pixel_formats_max) +{ + uint32_t i, fourcc_format; + + if (!format_list || !pixel_formats) + return 0; + + for (i = 0, fourcc_format = 0; + format_list->fourcc_format && i < pixel_formats_max; + ++format_list) { + /* verify if listed format is in dpu_format_map? */ + + /* optionally return modified formats */ + if (pixel_modifiers) { + /* assume same modifier for all fb planes */ + pixel_formats[i] = format_list->fourcc_format; + pixel_modifiers[i++] = format_list->modifier; + } else { + /* assume base formats grouped together */ + if (fourcc_format != format_list->fourcc_format) { + fourcc_format = format_list->fourcc_format; + pixel_formats[i++] = fourcc_format; + } + } + } + + return i; +} diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_formats.h b/drivers/gpu/drm/msm/disp/dpu1/dpu_formats.h new file mode 100644 index 000000000000..a54451d8d011 --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_formats.h @@ -0,0 +1,88 @@ +/* Copyright (c) 2015-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#ifndef _DPU_FORMATS_H +#define _DPU_FORMATS_H + +#include <drm/drm_fourcc.h> +#include "msm_gem.h" +#include "dpu_hw_mdss.h" + +/** + * dpu_get_dpu_format_ext() - Returns dpu format structure pointer. + * @format: DRM FourCC Code + * @modifiers: format modifier array from client, one per plane + */ +const struct dpu_format *dpu_get_dpu_format_ext( + const uint32_t format, + const uint64_t modifier); + +#define dpu_get_dpu_format(f) dpu_get_dpu_format_ext(f, 0) + +/** + * dpu_get_msm_format - get an dpu_format by its msm_format base + * callback function registers with the msm_kms layer + * @kms: kms driver + * @format: DRM FourCC Code + * @modifiers: data layout modifier + */ +const struct msm_format *dpu_get_msm_format( + struct msm_kms *kms, + const uint32_t format, + const uint64_t modifiers); + +/** + * dpu_populate_formats - populate the given array with fourcc codes supported + * @format_list: pointer to list of possible formats + * @pixel_formats: array to populate with fourcc codes + * @pixel_modifiers: array to populate with drm modifiers, can be NULL + * @pixel_formats_max: length of pixel formats array + * Return: number of elements populated + */ +uint32_t dpu_populate_formats( + const struct dpu_format_extended *format_list, + uint32_t *pixel_formats, + uint64_t *pixel_modifiers, + uint32_t pixel_formats_max); + +/** + * dpu_format_check_modified_format - validate format and buffers for + * dpu non-standard, i.e. modified format + * @kms: kms driver + * @msm_fmt: pointer to the msm_fmt base pointer of an dpu_format + * @cmd: fb_cmd2 structure user request + * @bos: gem buffer object list + * + * Return: error code on failure, 0 on success + */ +int dpu_format_check_modified_format( + const struct msm_kms *kms, + const struct msm_format *msm_fmt, + const struct drm_mode_fb_cmd2 *cmd, + struct drm_gem_object **bos); + +/** + * dpu_format_populate_layout - populate the given format layout based on + * mmu, fb, and format found in the fb + * @aspace: address space pointer + * @fb: framebuffer pointer + * @fmtl: format layout structure to populate + * + * Return: error code on failure, -EAGAIN if success but the addresses + * are the same as before or 0 if new addresses were populated + */ +int dpu_format_populate_layout( + struct msm_gem_address_space *aspace, + struct drm_framebuffer *fb, + struct dpu_hw_fmt_layout *fmtl); + +#endif /*_DPU_FORMATS_H */ diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_blk.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_blk.c new file mode 100644 index 000000000000..58d29e43faef --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_blk.c @@ -0,0 +1,155 @@ +/* Copyright (c) 2017-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#define pr_fmt(fmt) "[drm:%s:%d] " fmt, __func__, __LINE__ + +#include <linux/mutex.h> +#include <linux/errno.h> +#include <linux/slab.h> + +#include "dpu_hw_mdss.h" +#include "dpu_hw_blk.h" + +/* Serialization lock for dpu_hw_blk_list */ +static DEFINE_MUTEX(dpu_hw_blk_lock); + +/* List of all hw block objects */ +static LIST_HEAD(dpu_hw_blk_list); + +/** + * dpu_hw_blk_init - initialize hw block object + * @type: hw block type - enum dpu_hw_blk_type + * @id: instance id of the hw block + * @ops: Pointer to block operations + * return: 0 if success; error code otherwise + */ +int dpu_hw_blk_init(struct dpu_hw_blk *hw_blk, u32 type, int id, + struct dpu_hw_blk_ops *ops) +{ + if (!hw_blk) { + pr_err("invalid parameters\n"); + return -EINVAL; + } + + INIT_LIST_HEAD(&hw_blk->list); + hw_blk->type = type; + hw_blk->id = id; + atomic_set(&hw_blk->refcount, 0); + + if (ops) + hw_blk->ops = *ops; + + mutex_lock(&dpu_hw_blk_lock); + list_add(&hw_blk->list, &dpu_hw_blk_list); + mutex_unlock(&dpu_hw_blk_lock); + + return 0; +} + +/** + * dpu_hw_blk_destroy - destroy hw block object. + * @hw_blk: pointer to hw block object + * return: none + */ +void dpu_hw_blk_destroy(struct dpu_hw_blk *hw_blk) +{ + if (!hw_blk) { + pr_err("invalid parameters\n"); + return; + } + + if (atomic_read(&hw_blk->refcount)) + pr_err("hw_blk:%d.%d invalid refcount\n", hw_blk->type, + hw_blk->id); + + mutex_lock(&dpu_hw_blk_lock); + list_del(&hw_blk->list); + mutex_unlock(&dpu_hw_blk_lock); +} + +/** + * dpu_hw_blk_get - get hw_blk from free pool + * @hw_blk: if specified, increment reference count only + * @type: if hw_blk is not specified, allocate the next available of this type + * @id: if specified (>= 0), allocate the given instance of the above type + * return: pointer to hw block object + */ +struct dpu_hw_blk *dpu_hw_blk_get(struct dpu_hw_blk *hw_blk, u32 type, int id) +{ + struct dpu_hw_blk *curr; + int rc, refcount; + + if (!hw_blk) { + mutex_lock(&dpu_hw_blk_lock); + list_for_each_entry(curr, &dpu_hw_blk_list, list) { + if ((curr->type != type) || + (id >= 0 && curr->id != id) || + (id < 0 && + atomic_read(&curr->refcount))) + continue; + + hw_blk = curr; + break; + } + mutex_unlock(&dpu_hw_blk_lock); + } + + if (!hw_blk) { + pr_debug("no hw_blk:%d\n", type); + return NULL; + } + + refcount = atomic_inc_return(&hw_blk->refcount); + + if (refcount == 1 && hw_blk->ops.start) { + rc = hw_blk->ops.start(hw_blk); + if (rc) { + pr_err("failed to start hw_blk:%d rc:%d\n", type, rc); + goto error_start; + } + } + + pr_debug("hw_blk:%d.%d refcount:%d\n", hw_blk->type, + hw_blk->id, refcount); + return hw_blk; + +error_start: + dpu_hw_blk_put(hw_blk); + return ERR_PTR(rc); +} + +/** + * dpu_hw_blk_put - put hw_blk to free pool if decremented refcount is zero + * @hw_blk: hw block to be freed + * @free_blk: function to be called when reference count goes to zero + */ +void dpu_hw_blk_put(struct dpu_hw_blk *hw_blk) +{ + if (!hw_blk) { + pr_err("invalid parameters\n"); + return; + } + + pr_debug("hw_blk:%d.%d refcount:%d\n", hw_blk->type, hw_blk->id, + atomic_read(&hw_blk->refcount)); + + if (!atomic_read(&hw_blk->refcount)) { + pr_err("hw_blk:%d.%d invalid put\n", hw_blk->type, hw_blk->id); + return; + } + + if (atomic_dec_return(&hw_blk->refcount)) + return; + + if (hw_blk->ops.stop) + hw_blk->ops.stop(hw_blk); +} diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_blk.h b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_blk.h new file mode 100644 index 000000000000..0f4ca8af1ec5 --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_blk.h @@ -0,0 +1,53 @@ +/* Copyright (c) 2017-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#ifndef _DPU_HW_BLK_H +#define _DPU_HW_BLK_H + +#include <linux/types.h> +#include <linux/list.h> +#include <linux/atomic.h> + +struct dpu_hw_blk; + +/** + * struct dpu_hw_blk_ops - common hardware block operations + * @start: start operation on first get + * @stop: stop operation on last put + */ +struct dpu_hw_blk_ops { + int (*start)(struct dpu_hw_blk *); + void (*stop)(struct dpu_hw_blk *); +}; + +/** + * struct dpu_hw_blk - definition of hardware block object + * @list: list of hardware blocks + * @type: hardware block type + * @id: instance id + * @refcount: reference/usage count + */ +struct dpu_hw_blk { + struct list_head list; + u32 type; + int id; + atomic_t refcount; + struct dpu_hw_blk_ops ops; +}; + +int dpu_hw_blk_init(struct dpu_hw_blk *hw_blk, u32 type, int id, + struct dpu_hw_blk_ops *ops); +void dpu_hw_blk_destroy(struct dpu_hw_blk *hw_blk); + +struct dpu_hw_blk *dpu_hw_blk_get(struct dpu_hw_blk *hw_blk, u32 type, int id); +void dpu_hw_blk_put(struct dpu_hw_blk *hw_blk); +#endif /*_DPU_HW_BLK_H */ diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_catalog.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_catalog.c new file mode 100644 index 000000000000..44ee06398b1d --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_catalog.c @@ -0,0 +1,511 @@ +/* Copyright (c) 2015-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#define pr_fmt(fmt) "[drm:%s:%d] " fmt, __func__, __LINE__ +#include <linux/slab.h> +#include <linux/of_address.h> +#include <linux/platform_device.h> +#include "dpu_hw_mdss.h" +#include "dpu_hw_catalog.h" +#include "dpu_hw_catalog_format.h" +#include "dpu_kms.h" + +#define VIG_SDM845_MASK \ + (BIT(DPU_SSPP_SRC) | BIT(DPU_SSPP_SCALER_QSEED3) | BIT(DPU_SSPP_QOS) |\ + BIT(DPU_SSPP_CSC_10BIT) | BIT(DPU_SSPP_CDP) | BIT(DPU_SSPP_QOS_8LVL) |\ + BIT(DPU_SSPP_TS_PREFILL) | BIT(DPU_SSPP_EXCL_RECT)) + +#define DMA_SDM845_MASK \ + (BIT(DPU_SSPP_SRC) | BIT(DPU_SSPP_QOS) | BIT(DPU_SSPP_QOS_8LVL) |\ + BIT(DPU_SSPP_TS_PREFILL) | BIT(DPU_SSPP_TS_PREFILL_REC1) |\ + BIT(DPU_SSPP_CDP) | BIT(DPU_SSPP_EXCL_RECT)) + +#define MIXER_SDM845_MASK \ + (BIT(DPU_MIXER_SOURCESPLIT) | BIT(DPU_DIM_LAYER)) + +#define PINGPONG_SDM845_MASK BIT(DPU_PINGPONG_DITHER) + +#define PINGPONG_SDM845_SPLIT_MASK \ + (PINGPONG_SDM845_MASK | BIT(DPU_PINGPONG_TE2)) + +#define DEFAULT_PIXEL_RAM_SIZE (50 * 1024) +#define DEFAULT_DPU_LINE_WIDTH 2048 +#define DEFAULT_DPU_OUTPUT_LINE_WIDTH 2560 + +#define MAX_HORZ_DECIMATION 4 +#define MAX_VERT_DECIMATION 4 + +#define MAX_UPSCALE_RATIO 20 +#define MAX_DOWNSCALE_RATIO 4 +#define SSPP_UNITY_SCALE 1 + +#define STRCAT(X, Y) (X Y) + +/************************************************************* + * DPU sub blocks config + *************************************************************/ +/* DPU top level caps */ +static const struct dpu_caps sdm845_dpu_caps = { + .max_mixer_width = DEFAULT_DPU_OUTPUT_LINE_WIDTH, + .max_mixer_blendstages = 0xb, + .qseed_type = DPU_SSPP_SCALER_QSEED3, + .smart_dma_rev = DPU_SSPP_SMART_DMA_V2, + .ubwc_version = DPU_HW_UBWC_VER_20, + .has_src_split = true, + .has_dim_layer = true, + .has_idle_pc = true, +}; + +static struct dpu_mdp_cfg sdm845_mdp[] = { + { + .name = "top_0", .id = MDP_TOP, + .base = 0x0, .len = 0x45C, + .features = 0, + .highest_bank_bit = 0x2, + .has_dest_scaler = true, + .clk_ctrls[DPU_CLK_CTRL_VIG0] = { + .reg_off = 0x2AC, .bit_off = 0}, + .clk_ctrls[DPU_CLK_CTRL_VIG1] = { + .reg_off = 0x2B4, .bit_off = 0}, + .clk_ctrls[DPU_CLK_CTRL_VIG2] = { + .reg_off = 0x2BC, .bit_off = 0}, + .clk_ctrls[DPU_CLK_CTRL_VIG3] = { + .reg_off = 0x2C4, .bit_off = 0}, + .clk_ctrls[DPU_CLK_CTRL_DMA0] = { + .reg_off = 0x2AC, .bit_off = 8}, + .clk_ctrls[DPU_CLK_CTRL_DMA1] = { + .reg_off = 0x2B4, .bit_off = 8}, + .clk_ctrls[DPU_CLK_CTRL_CURSOR0] = { + .reg_off = 0x2BC, .bit_off = 8}, + .clk_ctrls[DPU_CLK_CTRL_CURSOR1] = { + .reg_off = 0x2C4, .bit_off = 8}, + }, +}; + +/************************************************************* + * CTL sub blocks config + *************************************************************/ +static struct dpu_ctl_cfg sdm845_ctl[] = { + { + .name = "ctl_0", .id = CTL_0, + .base = 0x1000, .len = 0xE4, + .features = BIT(DPU_CTL_SPLIT_DISPLAY) + }, + { + .name = "ctl_1", .id = CTL_1, + .base = 0x1200, .len = 0xE4, + .features = BIT(DPU_CTL_SPLIT_DISPLAY) + }, + { + .name = "ctl_2", .id = CTL_2, + .base = 0x1400, .len = 0xE4, + .features = 0 + }, + { + .name = "ctl_3", .id = CTL_3, + .base = 0x1600, .len = 0xE4, + .features = 0 + }, + { + .name = "ctl_4", .id = CTL_4, + .base = 0x1800, .len = 0xE4, + .features = 0 + }, +}; + +/************************************************************* + * SSPP sub blocks config + *************************************************************/ + +/* SSPP common configuration */ +static const struct dpu_sspp_blks_common sdm845_sspp_common = { + .maxlinewidth = DEFAULT_DPU_OUTPUT_LINE_WIDTH, + .pixel_ram_size = DEFAULT_PIXEL_RAM_SIZE, + .maxhdeciexp = MAX_HORZ_DECIMATION, + .maxvdeciexp = MAX_VERT_DECIMATION, +}; + +#define _VIG_SBLK(num, sdma_pri) \ + { \ + .common = &sdm845_sspp_common, \ + .maxdwnscale = MAX_DOWNSCALE_RATIO, \ + .maxupscale = MAX_UPSCALE_RATIO, \ + .smart_dma_priority = sdma_pri, \ + .src_blk = {.name = STRCAT("sspp_src_", num), \ + .id = DPU_SSPP_SRC, .base = 0x00, .len = 0x150,}, \ + .scaler_blk = {.name = STRCAT("sspp_scaler", num), \ + .id = DPU_SSPP_SCALER_QSEED3, \ + .base = 0xa00, .len = 0xa0,}, \ + .csc_blk = {.name = STRCAT("sspp_csc", num), \ + .id = DPU_SSPP_CSC_10BIT, \ + .base = 0x1a00, .len = 0x100,}, \ + .format_list = plane_formats_yuv, \ + .virt_format_list = plane_formats, \ + } + +#define _DMA_SBLK(num, sdma_pri) \ + { \ + .common = &sdm845_sspp_common, \ + .maxdwnscale = SSPP_UNITY_SCALE, \ + .maxupscale = SSPP_UNITY_SCALE, \ + .smart_dma_priority = sdma_pri, \ + .src_blk = {.name = STRCAT("sspp_src_", num), \ + .id = DPU_SSPP_SRC, .base = 0x00, .len = 0x150,}, \ + .format_list = plane_formats, \ + .virt_format_list = plane_formats, \ + } + +static const struct dpu_sspp_sub_blks sdm845_vig_sblk_0 = _VIG_SBLK("0", 5); +static const struct dpu_sspp_sub_blks sdm845_vig_sblk_1 = _VIG_SBLK("1", 6); +static const struct dpu_sspp_sub_blks sdm845_vig_sblk_2 = _VIG_SBLK("2", 7); +static const struct dpu_sspp_sub_blks sdm845_vig_sblk_3 = _VIG_SBLK("3", 8); + +static const struct dpu_sspp_sub_blks sdm845_dma_sblk_0 = _DMA_SBLK("8", 1); +static const struct dpu_sspp_sub_blks sdm845_dma_sblk_1 = _DMA_SBLK("9", 2); +static const struct dpu_sspp_sub_blks sdm845_dma_sblk_2 = _DMA_SBLK("10", 3); +static const struct dpu_sspp_sub_blks sdm845_dma_sblk_3 = _DMA_SBLK("11", 4); + +#define SSPP_VIG_BLK(_name, _id, _base, _sblk, _xinid, _clkctrl) \ + { \ + .name = _name, .id = _id, \ + .base = _base, .len = 0x1c8, \ + .features = VIG_SDM845_MASK, \ + .sblk = &_sblk, \ + .xin_id = _xinid, \ + .type = SSPP_TYPE_VIG, \ + .clk_ctrl = _clkctrl \ + } + +#define SSPP_DMA_BLK(_name, _id, _base, _sblk, _xinid, _clkctrl) \ + { \ + .name = _name, .id = _id, \ + .base = _base, .len = 0x1c8, \ + .features = DMA_SDM845_MASK, \ + .sblk = &_sblk, \ + .xin_id = _xinid, \ + .type = SSPP_TYPE_DMA, \ + .clk_ctrl = _clkctrl \ + } + +static struct dpu_sspp_cfg sdm845_sspp[] = { + SSPP_VIG_BLK("sspp_0", SSPP_VIG0, 0x4000, + sdm845_vig_sblk_0, 0, DPU_CLK_CTRL_VIG0), + SSPP_VIG_BLK("sspp_1", SSPP_VIG1, 0x6000, + sdm845_vig_sblk_1, 4, DPU_CLK_CTRL_VIG1), + SSPP_VIG_BLK("sspp_2", SSPP_VIG2, 0x8000, + sdm845_vig_sblk_2, 8, DPU_CLK_CTRL_VIG2), + SSPP_VIG_BLK("sspp_3", SSPP_VIG3, 0xa000, + sdm845_vig_sblk_3, 12, DPU_CLK_CTRL_VIG3), + SSPP_DMA_BLK("sspp_8", SSPP_DMA0, 0x24000, + sdm845_dma_sblk_0, 1, DPU_CLK_CTRL_DMA0), + SSPP_DMA_BLK("sspp_9", SSPP_DMA1, 0x26000, + sdm845_dma_sblk_1, 5, DPU_CLK_CTRL_DMA1), + SSPP_DMA_BLK("sspp_10", SSPP_DMA2, 0x28000, + sdm845_dma_sblk_2, 9, DPU_CLK_CTRL_CURSOR0), + SSPP_DMA_BLK("sspp_11", SSPP_DMA3, 0x2a000, + sdm845_dma_sblk_3, 13, DPU_CLK_CTRL_CURSOR1), +}; + +/************************************************************* + * MIXER sub blocks config + *************************************************************/ +static const struct dpu_lm_sub_blks sdm845_lm_sblk = { + .maxwidth = DEFAULT_DPU_OUTPUT_LINE_WIDTH, + .maxblendstages = 11, /* excluding base layer */ + .blendstage_base = { /* offsets relative to mixer base */ + 0x20, 0x38, 0x50, 0x68, 0x80, 0x98, + 0xb0, 0xc8, 0xe0, 0xf8, 0x110 + }, +}; + +#define LM_BLK(_name, _id, _base, _ds, _pp, _lmpair) \ + { \ + .name = _name, .id = _id, \ + .base = _base, .len = 0x320, \ + .features = MIXER_SDM845_MASK, \ + .sblk = &sdm845_lm_sblk, \ + .ds = _ds, \ + .pingpong = _pp, \ + .lm_pair_mask = (1 << _lmpair) \ + } + +static struct dpu_lm_cfg sdm845_lm[] = { + LM_BLK("lm_0", LM_0, 0x44000, DS_0, PINGPONG_0, LM_1), + LM_BLK("lm_1", LM_1, 0x45000, DS_1, PINGPONG_1, LM_0), + LM_BLK("lm_2", LM_2, 0x46000, DS_MAX, PINGPONG_2, LM_5), + LM_BLK("lm_3", LM_3, 0x0, DS_MAX, PINGPONG_MAX, 0), + LM_BLK("lm_4", LM_4, 0x0, DS_MAX, PINGPONG_MAX, 0), + LM_BLK("lm_5", LM_5, 0x49000, DS_MAX, PINGPONG_3, LM_2), +}; + +/************************************************************* + * DS sub blocks config + *************************************************************/ +static const struct dpu_ds_top_cfg sdm845_ds_top = { + .name = "ds_top_0", .id = DS_TOP, + .base = 0x60000, .len = 0xc, + .maxinputwidth = DEFAULT_DPU_LINE_WIDTH, + .maxoutputwidth = DEFAULT_DPU_OUTPUT_LINE_WIDTH, + .maxupscale = MAX_UPSCALE_RATIO, +}; + +#define DS_BLK(_name, _id, _base) \ + {\ + .name = _name, .id = _id, \ + .base = _base, .len = 0x800, \ + .features = DPU_SSPP_SCALER_QSEED3, \ + .top = &sdm845_ds_top \ + } + +static struct dpu_ds_cfg sdm845_ds[] = { + DS_BLK("ds_0", DS_0, 0x800), + DS_BLK("ds_1", DS_1, 0x1000), +}; + +/************************************************************* + * PINGPONG sub blocks config + *************************************************************/ +static const struct dpu_pingpong_sub_blks sdm845_pp_sblk_te = { + .te2 = {.id = DPU_PINGPONG_TE2, .base = 0x2000, .len = 0x0, + .version = 0x1}, + .dither = {.id = DPU_PINGPONG_DITHER, .base = 0x30e0, + .len = 0x20, .version = 0x10000}, +}; + +static const struct dpu_pingpong_sub_blks sdm845_pp_sblk = { + .dither = {.id = DPU_PINGPONG_DITHER, .base = 0x30e0, + .len = 0x20, .version = 0x10000}, +}; + +#define PP_BLK_TE(_name, _id, _base) \ + {\ + .name = _name, .id = _id, \ + .base = _base, .len = 0xd4, \ + .features = PINGPONG_SDM845_SPLIT_MASK, \ + .sblk = &sdm845_pp_sblk_te \ + } +#define PP_BLK(_name, _id, _base) \ + {\ + .name = _name, .id = _id, \ + .base = _base, .len = 0xd4, \ + .features = PINGPONG_SDM845_MASK, \ + .sblk = &sdm845_pp_sblk \ + } + +static struct dpu_pingpong_cfg sdm845_pp[] = { + PP_BLK_TE("pingpong_0", PINGPONG_0, 0x70000), + PP_BLK_TE("pingpong_1", PINGPONG_1, 0x70800), + PP_BLK("pingpong_2", PINGPONG_2, 0x71000), + PP_BLK("pingpong_3", PINGPONG_3, 0x71800), +}; + +/************************************************************* + * INTF sub blocks config + *************************************************************/ +#define INTF_BLK(_name, _id, _base, _type, _ctrl_id) \ + {\ + .name = _name, .id = _id, \ + .base = _base, .len = 0x280, \ + .type = _type, \ + .controller_id = _ctrl_id, \ + .prog_fetch_lines_worst_case = 24 \ + } + +static struct dpu_intf_cfg sdm845_intf[] = { + INTF_BLK("intf_0", INTF_0, 0x6A000, INTF_DP, 0), + INTF_BLK("intf_1", INTF_1, 0x6A800, INTF_DSI, 0), + INTF_BLK("intf_2", INTF_2, 0x6B000, INTF_DSI, 1), + INTF_BLK("intf_3", INTF_3, 0x6B800, INTF_DP, 1), +}; + +/************************************************************* + * CDM sub blocks config + *************************************************************/ +static struct dpu_cdm_cfg sdm845_cdm[] = { + { + .name = "cdm_0", .id = CDM_0, + .base = 0x79200, .len = 0x224, + .features = 0, + .intf_connect = BIT(INTF_3), + }, +}; + +/************************************************************* + * VBIF sub blocks config + *************************************************************/ +/* VBIF QOS remap */ +static u32 sdm845_rt_pri_lvl[] = {3, 3, 4, 4, 5, 5, 6, 6}; +static u32 sdm845_nrt_pri_lvl[] = {3, 3, 3, 3, 3, 3, 3, 3}; + +static struct dpu_vbif_cfg sdm845_vbif[] = { + { + .name = "vbif_0", .id = VBIF_0, + .base = 0, .len = 0x1040, + .features = BIT(DPU_VBIF_QOS_REMAP), + .xin_halt_timeout = 0x4000, + .qos_rt_tbl = { + .npriority_lvl = ARRAY_SIZE(sdm845_rt_pri_lvl), + .priority_lvl = sdm845_rt_pri_lvl, + }, + .qos_nrt_tbl = { + .npriority_lvl = ARRAY_SIZE(sdm845_nrt_pri_lvl), + .priority_lvl = sdm845_nrt_pri_lvl, + }, + .memtype_count = 14, + .memtype = {3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3}, + }, +}; + +static struct dpu_reg_dma_cfg sdm845_regdma = { + .base = 0x0, .version = 0x1, .trigger_sel_off = 0x119c +}; + +/************************************************************* + * PERF data config + *************************************************************/ + +/* SSPP QOS LUTs */ +static struct dpu_qos_lut_entry sdm845_qos_linear[] = { + {.fl = 4, .lut = 0x357}, + {.fl = 5, .lut = 0x3357}, + {.fl = 6, .lut = 0x23357}, + {.fl = 7, .lut = 0x223357}, + {.fl = 8, .lut = 0x2223357}, + {.fl = 9, .lut = 0x22223357}, + {.fl = 10, .lut = 0x222223357}, + {.fl = 11, .lut = 0x2222223357}, + {.fl = 12, .lut = 0x22222223357}, + {.fl = 13, .lut = 0x222222223357}, + {.fl = 14, .lut = 0x1222222223357}, + {.fl = 0, .lut = 0x11222222223357} +}; + +static struct dpu_qos_lut_entry sdm845_qos_macrotile[] = { + {.fl = 10, .lut = 0x344556677}, + {.fl = 11, .lut = 0x3344556677}, + {.fl = 12, .lut = 0x23344556677}, + {.fl = 13, .lut = 0x223344556677}, + {.fl = 14, .lut = 0x1223344556677}, + {.fl = 0, .lut = 0x112233344556677}, +}; + +static struct dpu_qos_lut_entry sdm845_qos_nrt[] = { + {.fl = 0, .lut = 0x0}, +}; + +static struct dpu_perf_cfg sdm845_perf_data = { + .max_bw_low = 6800000, + .max_bw_high = 6800000, + .min_core_ib = 2400000, + .min_llcc_ib = 800000, + .min_dram_ib = 800000, + .core_ib_ff = "6.0", + .core_clk_ff = "1.0", + .comp_ratio_rt = + "NV12/5/1/1.23 AB24/5/1/1.23 XB24/5/1/1.23", + .comp_ratio_nrt = + "NV12/5/1/1.25 AB24/5/1/1.25 XB24/5/1/1.25", + .undersized_prefill_lines = 2, + .xtra_prefill_lines = 2, + .dest_scale_prefill_lines = 3, + .macrotile_prefill_lines = 4, + .yuv_nv12_prefill_lines = 8, + .linear_prefill_lines = 1, + .downscaling_prefill_lines = 1, + .amortizable_threshold = 25, + .min_prefill_lines = 24, + .danger_lut_tbl = {0xf, 0xffff, 0x0}, + .qos_lut_tbl = { + {.nentry = ARRAY_SIZE(sdm845_qos_linear), + .entries = sdm845_qos_linear + }, + {.nentry = ARRAY_SIZE(sdm845_qos_macrotile), + .entries = sdm845_qos_macrotile + }, + {.nentry = ARRAY_SIZE(sdm845_qos_nrt), + .entries = sdm845_qos_nrt + }, + }, + .cdp_cfg = { + {.rd_enable = 1, .wr_enable = 1}, + {.rd_enable = 1, .wr_enable = 0} + }, +}; + +/************************************************************* + * Hardware catalog init + *************************************************************/ + +/* + * sdm845_cfg_init(): populate sdm845 dpu sub-blocks reg offsets + * and instance counts. + */ +static void sdm845_cfg_init(struct dpu_mdss_cfg *dpu_cfg) +{ + *dpu_cfg = (struct dpu_mdss_cfg){ + .caps = &sdm845_dpu_caps, + .mdp_count = ARRAY_SIZE(sdm845_mdp), + .mdp = sdm845_mdp, + .ctl_count = ARRAY_SIZE(sdm845_ctl), + .ctl = sdm845_ctl, + .sspp_count = ARRAY_SIZE(sdm845_sspp), + .sspp = sdm845_sspp, + .mixer_count = ARRAY_SIZE(sdm845_lm), + .mixer = sdm845_lm, + .ds_count = ARRAY_SIZE(sdm845_ds), + .ds = sdm845_ds, + .pingpong_count = ARRAY_SIZE(sdm845_pp), + .pingpong = sdm845_pp, + .cdm_count = ARRAY_SIZE(sdm845_cdm), + .cdm = sdm845_cdm, + .intf_count = ARRAY_SIZE(sdm845_intf), + .intf = sdm845_intf, + .vbif_count = ARRAY_SIZE(sdm845_vbif), + .vbif = sdm845_vbif, + .reg_dma_count = 1, + .dma_cfg = sdm845_regdma, + .perf = sdm845_perf_data, + }; +} + +static struct dpu_mdss_hw_cfg_handler cfg_handler[] = { + { .hw_rev = DPU_HW_VER_400, .cfg_init = sdm845_cfg_init}, + { .hw_rev = DPU_HW_VER_401, .cfg_init = sdm845_cfg_init}, +}; + +void dpu_hw_catalog_deinit(struct dpu_mdss_cfg *dpu_cfg) +{ + kfree(dpu_cfg); +} + +struct dpu_mdss_cfg *dpu_hw_catalog_init(u32 hw_rev) +{ + int i; + struct dpu_mdss_cfg *dpu_cfg; + + dpu_cfg = kzalloc(sizeof(*dpu_cfg), GFP_KERNEL); + if (!dpu_cfg) + return ERR_PTR(-ENOMEM); + + for (i = 0; i < ARRAY_SIZE(cfg_handler); i++) { + if (cfg_handler[i].hw_rev == hw_rev) { + cfg_handler[i].cfg_init(dpu_cfg); + dpu_cfg->hwversion = hw_rev; + return dpu_cfg; + } + } + + DPU_ERROR("unsupported chipset id:%X\n", hw_rev); + dpu_hw_catalog_deinit(dpu_cfg); + return ERR_PTR(-ENODEV); +} + diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_catalog.h b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_catalog.h new file mode 100644 index 000000000000..f0cb0d4fc80e --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_catalog.h @@ -0,0 +1,804 @@ +/* Copyright (c) 2015-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#ifndef _DPU_HW_CATALOG_H +#define _DPU_HW_CATALOG_H + +#include <linux/kernel.h> +#include <linux/bug.h> +#include <linux/bitmap.h> +#include <linux/err.h> +#include <drm/drmP.h> + +/** + * Max hardware block count: For ex: max 12 SSPP pipes or + * 5 ctl paths. In all cases, it can have max 12 hardware blocks + * based on current design + */ +#define MAX_BLOCKS 12 + +#define DPU_HW_VER(MAJOR, MINOR, STEP) (((MAJOR & 0xF) << 28) |\ + ((MINOR & 0xFFF) << 16) |\ + (STEP & 0xFFFF)) + +#define DPU_HW_MAJOR(rev) ((rev) >> 28) +#define DPU_HW_MINOR(rev) (((rev) >> 16) & 0xFFF) +#define DPU_HW_STEP(rev) ((rev) & 0xFFFF) +#define DPU_HW_MAJOR_MINOR(rev) ((rev) >> 16) + +#define IS_DPU_MAJOR_MINOR_SAME(rev1, rev2) \ + (DPU_HW_MAJOR_MINOR((rev1)) == DPU_HW_MAJOR_MINOR((rev2))) + +#define DPU_HW_VER_170 DPU_HW_VER(1, 7, 0) /* 8996 v1.0 */ +#define DPU_HW_VER_171 DPU_HW_VER(1, 7, 1) /* 8996 v2.0 */ +#define DPU_HW_VER_172 DPU_HW_VER(1, 7, 2) /* 8996 v3.0 */ +#define DPU_HW_VER_300 DPU_HW_VER(3, 0, 0) /* 8998 v1.0 */ +#define DPU_HW_VER_301 DPU_HW_VER(3, 0, 1) /* 8998 v1.1 */ +#define DPU_HW_VER_400 DPU_HW_VER(4, 0, 0) /* sdm845 v1.0 */ +#define DPU_HW_VER_401 DPU_HW_VER(4, 0, 1) /* sdm845 v2.0 */ +#define DPU_HW_VER_410 DPU_HW_VER(4, 1, 0) /* sdm670 v1.0 */ +#define DPU_HW_VER_500 DPU_HW_VER(5, 0, 0) /* sdm855 v1.0 */ + + +#define IS_MSM8996_TARGET(rev) IS_DPU_MAJOR_MINOR_SAME((rev), DPU_HW_VER_170) +#define IS_MSM8998_TARGET(rev) IS_DPU_MAJOR_MINOR_SAME((rev), DPU_HW_VER_300) +#define IS_SDM845_TARGET(rev) IS_DPU_MAJOR_MINOR_SAME((rev), DPU_HW_VER_400) +#define IS_SDM670_TARGET(rev) IS_DPU_MAJOR_MINOR_SAME((rev), DPU_HW_VER_410) +#define IS_SDM855_TARGET(rev) IS_DPU_MAJOR_MINOR_SAME((rev), DPU_HW_VER_500) + + +#define DPU_HW_BLK_NAME_LEN 16 + +#define MAX_IMG_WIDTH 0x3fff +#define MAX_IMG_HEIGHT 0x3fff + +#define CRTC_DUAL_MIXERS 2 + +#define MAX_XIN_COUNT 16 + +/** + * Supported UBWC feature versions + */ +enum { + DPU_HW_UBWC_VER_10 = 0x100, + DPU_HW_UBWC_VER_20 = 0x200, + DPU_HW_UBWC_VER_30 = 0x300, +}; + +#define IS_UBWC_20_SUPPORTED(rev) ((rev) >= DPU_HW_UBWC_VER_20) + +/** + * MDP TOP BLOCK features + * @DPU_MDP_PANIC_PER_PIPE Panic configuration needs to be be done per pipe + * @DPU_MDP_10BIT_SUPPORT, Chipset supports 10 bit pixel formats + * @DPU_MDP_BWC, MDSS HW supports Bandwidth compression. + * @DPU_MDP_UBWC_1_0, This chipsets supports Universal Bandwidth + * compression initial revision + * @DPU_MDP_UBWC_1_5, Universal Bandwidth compression version 1.5 + * @DPU_MDP_MAX Maximum value + + */ +enum { + DPU_MDP_PANIC_PER_PIPE = 0x1, + DPU_MDP_10BIT_SUPPORT, + DPU_MDP_BWC, + DPU_MDP_UBWC_1_0, + DPU_MDP_UBWC_1_5, + DPU_MDP_MAX +}; + +/** + * SSPP sub-blocks/features + * @DPU_SSPP_SRC Src and fetch part of the pipes, + * @DPU_SSPP_SCALER_QSEED2, QSEED2 algorithm support + * @DPU_SSPP_SCALER_QSEED3, QSEED3 alogorithm support + * @DPU_SSPP_SCALER_RGB, RGB Scaler, supported by RGB pipes + * @DPU_SSPP_CSC, Support of Color space converion + * @DPU_SSPP_CSC_10BIT, Support of 10-bit Color space conversion + * @DPU_SSPP_CURSOR, SSPP can be used as a cursor layer + * @DPU_SSPP_QOS, SSPP support QoS control, danger/safe/creq + * @DPU_SSPP_QOS_8LVL, SSPP support 8-level QoS control + * @DPU_SSPP_EXCL_RECT, SSPP supports exclusion rect + * @DPU_SSPP_SMART_DMA_V1, SmartDMA 1.0 support + * @DPU_SSPP_SMART_DMA_V2, SmartDMA 2.0 support + * @DPU_SSPP_TS_PREFILL Supports prefill with traffic shaper + * @DPU_SSPP_TS_PREFILL_REC1 Supports prefill with traffic shaper multirec + * @DPU_SSPP_CDP Supports client driven prefetch + * @DPU_SSPP_MAX maximum value + */ +enum { + DPU_SSPP_SRC = 0x1, + DPU_SSPP_SCALER_QSEED2, + DPU_SSPP_SCALER_QSEED3, + DPU_SSPP_SCALER_RGB, + DPU_SSPP_CSC, + DPU_SSPP_CSC_10BIT, + DPU_SSPP_CURSOR, + DPU_SSPP_QOS, + DPU_SSPP_QOS_8LVL, + DPU_SSPP_EXCL_RECT, + DPU_SSPP_SMART_DMA_V1, + DPU_SSPP_SMART_DMA_V2, + DPU_SSPP_TS_PREFILL, + DPU_SSPP_TS_PREFILL_REC1, + DPU_SSPP_CDP, + DPU_SSPP_MAX +}; + +/* + * MIXER sub-blocks/features + * @DPU_MIXER_LAYER Layer mixer layer blend configuration, + * @DPU_MIXER_SOURCESPLIT Layer mixer supports source-split configuration + * @DPU_MIXER_GC Gamma correction block + * @DPU_DIM_LAYER Layer mixer supports dim layer + * @DPU_MIXER_MAX maximum value + */ +enum { + DPU_MIXER_LAYER = 0x1, + DPU_MIXER_SOURCESPLIT, + DPU_MIXER_GC, + DPU_DIM_LAYER, + DPU_MIXER_MAX +}; + +/** + * PINGPONG sub-blocks + * @DPU_PINGPONG_TE Tear check block + * @DPU_PINGPONG_TE2 Additional tear check block for split pipes + * @DPU_PINGPONG_SPLIT PP block supports split fifo + * @DPU_PINGPONG_SLAVE PP block is a suitable slave for split fifo + * @DPU_PINGPONG_DITHER, Dither blocks + * @DPU_PINGPONG_MAX + */ +enum { + DPU_PINGPONG_TE = 0x1, + DPU_PINGPONG_TE2, + DPU_PINGPONG_SPLIT, + DPU_PINGPONG_SLAVE, + DPU_PINGPONG_DITHER, + DPU_PINGPONG_MAX +}; + +/** + * CTL sub-blocks + * @DPU_CTL_SPLIT_DISPLAY CTL supports video mode split display + * @DPU_CTL_MAX + */ +enum { + DPU_CTL_SPLIT_DISPLAY = 0x1, + DPU_CTL_MAX +}; + +/** + * VBIF sub-blocks and features + * @DPU_VBIF_QOS_OTLIM VBIF supports OT Limit + * @DPU_VBIF_QOS_REMAP VBIF supports QoS priority remap + * @DPU_VBIF_MAX maximum value + */ +enum { + DPU_VBIF_QOS_OTLIM = 0x1, + DPU_VBIF_QOS_REMAP, + DPU_VBIF_MAX +}; + +/** + * MACRO DPU_HW_BLK_INFO - information of HW blocks inside DPU + * @name: string name for debug purposes + * @id: enum identifying this block + * @base: register base offset to mdss + * @len: length of hardware block + * @features bit mask identifying sub-blocks/features + */ +#define DPU_HW_BLK_INFO \ + char name[DPU_HW_BLK_NAME_LEN]; \ + u32 id; \ + u32 base; \ + u32 len; \ + unsigned long features + +/** + * MACRO DPU_HW_SUBBLK_INFO - information of HW sub-block inside DPU + * @name: string name for debug purposes + * @id: enum identifying this sub-block + * @base: offset of this sub-block relative to the block + * offset + * @len register block length of this sub-block + */ +#define DPU_HW_SUBBLK_INFO \ + char name[DPU_HW_BLK_NAME_LEN]; \ + u32 id; \ + u32 base; \ + u32 len + +/** + * struct dpu_src_blk: SSPP part of the source pipes + * @info: HW register and features supported by this sub-blk + */ +struct dpu_src_blk { + DPU_HW_SUBBLK_INFO; +}; + +/** + * struct dpu_scaler_blk: Scaler information + * @info: HW register and features supported by this sub-blk + * @version: qseed block revision + */ +struct dpu_scaler_blk { + DPU_HW_SUBBLK_INFO; + u32 version; +}; + +struct dpu_csc_blk { + DPU_HW_SUBBLK_INFO; +}; + +/** + * struct dpu_pp_blk : Pixel processing sub-blk information + * @info: HW register and features supported by this sub-blk + * @version: HW Algorithm version + */ +struct dpu_pp_blk { + DPU_HW_SUBBLK_INFO; + u32 version; +}; + +/** + * struct dpu_format_extended - define dpu specific pixel format+modifier + * @fourcc_format: Base FOURCC pixel format code + * @modifier: 64-bit drm format modifier, same modifier must be applied to all + * framebuffer planes + */ +struct dpu_format_extended { + uint32_t fourcc_format; + uint64_t modifier; +}; + +/** + * enum dpu_qos_lut_usage - define QoS LUT use cases + */ +enum dpu_qos_lut_usage { + DPU_QOS_LUT_USAGE_LINEAR, + DPU_QOS_LUT_USAGE_MACROTILE, + DPU_QOS_LUT_USAGE_NRT, + DPU_QOS_LUT_USAGE_MAX, +}; + +/** + * struct dpu_qos_lut_entry - define QoS LUT table entry + * @fl: fill level, or zero on last entry to indicate default lut + * @lut: lut to use if equal to or less than fill level + */ +struct dpu_qos_lut_entry { + u32 fl; + u64 lut; +}; + +/** + * struct dpu_qos_lut_tbl - define QoS LUT table + * @nentry: number of entry in this table + * @entries: Pointer to table entries + */ +struct dpu_qos_lut_tbl { + u32 nentry; + struct dpu_qos_lut_entry *entries; +}; + +/** + * struct dpu_caps - define DPU capabilities + * @max_mixer_width max layer mixer line width support. + * @max_mixer_blendstages max layer mixer blend stages or + * supported z order + * @qseed_type qseed2 or qseed3 support. + * @smart_dma_rev Supported version of SmartDMA feature. + * @ubwc_version UBWC feature version (0x0 for not supported) + * @has_src_split source split feature status + * @has_dim_layer dim layer feature status + * @has_idle_pc indicate if idle power collapse feature is supported + */ +struct dpu_caps { + u32 max_mixer_width; + u32 max_mixer_blendstages; + u32 qseed_type; + u32 smart_dma_rev; + u32 ubwc_version; + bool has_src_split; + bool has_dim_layer; + bool has_idle_pc; +}; + +/** + * struct dpu_sspp_blks_common : SSPP sub-blocks common configuration + * @maxwidth: max pixelwidth supported by this pipe + * @pixel_ram_size: size of latency hiding and de-tiling buffer in bytes + * @maxhdeciexp: max horizontal decimation supported by this pipe + * (max is 2^value) + * @maxvdeciexp: max vertical decimation supported by this pipe + * (max is 2^value) + */ +struct dpu_sspp_blks_common { + u32 maxlinewidth; + u32 pixel_ram_size; + u32 maxhdeciexp; + u32 maxvdeciexp; +}; + +/** + * struct dpu_sspp_sub_blks : SSPP sub-blocks + * common: Pointer to common configurations shared by sub blocks + * @creq_vblank: creq priority during vertical blanking + * @danger_vblank: danger priority during vertical blanking + * @maxdwnscale: max downscale ratio supported(without DECIMATION) + * @maxupscale: maxupscale ratio supported + * @smart_dma_priority: hw priority of rect1 of multirect pipe + * @max_per_pipe_bw: maximum allowable bandwidth of this pipe in kBps + * @src_blk: + * @scaler_blk: + * @csc_blk: + * @hsic: + * @memcolor: + * @pcc_blk: + * @igc_blk: + * @format_list: Pointer to list of supported formats + * @virt_format_list: Pointer to list of supported formats for virtual planes + */ +struct dpu_sspp_sub_blks { + const struct dpu_sspp_blks_common *common; + u32 creq_vblank; + u32 danger_vblank; + u32 maxdwnscale; + u32 maxupscale; + u32 smart_dma_priority; + u32 max_per_pipe_bw; + struct dpu_src_blk src_blk; + struct dpu_scaler_blk scaler_blk; + struct dpu_pp_blk csc_blk; + struct dpu_pp_blk hsic_blk; + struct dpu_pp_blk memcolor_blk; + struct dpu_pp_blk pcc_blk; + struct dpu_pp_blk igc_blk; + + const struct dpu_format_extended *format_list; + const struct dpu_format_extended *virt_format_list; +}; + +/** + * struct dpu_lm_sub_blks: information of mixer block + * @maxwidth: Max pixel width supported by this mixer + * @maxblendstages: Max number of blend-stages supported + * @blendstage_base: Blend-stage register base offset + * @gc: gamma correction block + */ +struct dpu_lm_sub_blks { + u32 maxwidth; + u32 maxblendstages; + u32 blendstage_base[MAX_BLOCKS]; + struct dpu_pp_blk gc; +}; + +struct dpu_pingpong_sub_blks { + struct dpu_pp_blk te; + struct dpu_pp_blk te2; + struct dpu_pp_blk dither; +}; + +/** + * dpu_clk_ctrl_type - Defines top level clock control signals + */ +enum dpu_clk_ctrl_type { + DPU_CLK_CTRL_NONE, + DPU_CLK_CTRL_VIG0, + DPU_CLK_CTRL_VIG1, + DPU_CLK_CTRL_VIG2, + DPU_CLK_CTRL_VIG3, + DPU_CLK_CTRL_VIG4, + DPU_CLK_CTRL_RGB0, + DPU_CLK_CTRL_RGB1, + DPU_CLK_CTRL_RGB2, + DPU_CLK_CTRL_RGB3, + DPU_CLK_CTRL_DMA0, + DPU_CLK_CTRL_DMA1, + DPU_CLK_CTRL_CURSOR0, + DPU_CLK_CTRL_CURSOR1, + DPU_CLK_CTRL_INLINE_ROT0_SSPP, + DPU_CLK_CTRL_MAX, +}; + +/* struct dpu_clk_ctrl_reg : Clock control register + * @reg_off: register offset + * @bit_off: bit offset + */ +struct dpu_clk_ctrl_reg { + u32 reg_off; + u32 bit_off; +}; + +/* struct dpu_mdp_cfg : MDP TOP-BLK instance info + * @id: index identifying this block + * @base: register base offset to mdss + * @features bit mask identifying sub-blocks/features + * @highest_bank_bit: UBWC parameter + * @ubwc_static: ubwc static configuration + * @ubwc_swizzle: ubwc default swizzle setting + * @has_dest_scaler: indicates support of destination scaler + * @clk_ctrls clock control register definition + */ +struct dpu_mdp_cfg { + DPU_HW_BLK_INFO; + u32 highest_bank_bit; + u32 ubwc_static; + u32 ubwc_swizzle; + bool has_dest_scaler; + struct dpu_clk_ctrl_reg clk_ctrls[DPU_CLK_CTRL_MAX]; +}; + +/* struct dpu_mdp_cfg : MDP TOP-BLK instance info + * @id: index identifying this block + * @base: register base offset to mdss + * @features bit mask identifying sub-blocks/features + */ +struct dpu_ctl_cfg { + DPU_HW_BLK_INFO; +}; + +/** + * struct dpu_sspp_cfg - information of source pipes + * @id: index identifying this block + * @base register offset of this block + * @features bit mask identifying sub-blocks/features + * @sblk: SSPP sub-blocks information + * @xin_id: bus client identifier + * @clk_ctrl clock control identifier + * @type sspp type identifier + */ +struct dpu_sspp_cfg { + DPU_HW_BLK_INFO; + const struct dpu_sspp_sub_blks *sblk; + u32 xin_id; + enum dpu_clk_ctrl_type clk_ctrl; + u32 type; +}; + +/** + * struct dpu_lm_cfg - information of layer mixer blocks + * @id: index identifying this block + * @base register offset of this block + * @features bit mask identifying sub-blocks/features + * @sblk: LM Sub-blocks information + * @pingpong: ID of connected PingPong, PINGPONG_MAX if unsupported + * @ds: ID of connected DS, DS_MAX if unsupported + * @lm_pair_mask: Bitmask of LMs that can be controlled by same CTL + */ +struct dpu_lm_cfg { + DPU_HW_BLK_INFO; + const struct dpu_lm_sub_blks *sblk; + u32 pingpong; + u32 ds; + unsigned long lm_pair_mask; +}; + +/** + * struct dpu_ds_top_cfg - information of dest scaler top + * @id enum identifying this block + * @base register offset of this block + * @features bit mask identifying features + * @version hw version of dest scaler + * @maxinputwidth maximum input line width + * @maxoutputwidth maximum output line width + * @maxupscale maximum upscale ratio + */ +struct dpu_ds_top_cfg { + DPU_HW_BLK_INFO; + u32 version; + u32 maxinputwidth; + u32 maxoutputwidth; + u32 maxupscale; +}; + +/** + * struct dpu_ds_cfg - information of dest scaler blocks + * @id enum identifying this block + * @base register offset wrt DS top offset + * @features bit mask identifying features + * @version hw version of the qseed block + * @top DS top information + */ +struct dpu_ds_cfg { + DPU_HW_BLK_INFO; + u32 version; + const struct dpu_ds_top_cfg *top; +}; + +/** + * struct dpu_pingpong_cfg - information of PING-PONG blocks + * @id enum identifying this block + * @base register offset of this block + * @features bit mask identifying sub-blocks/features + * @sblk sub-blocks information + */ +struct dpu_pingpong_cfg { + DPU_HW_BLK_INFO; + const struct dpu_pingpong_sub_blks *sblk; +}; + +/** + * struct dpu_cdm_cfg - information of chroma down blocks + * @id enum identifying this block + * @base register offset of this block + * @features bit mask identifying sub-blocks/features + * @intf_connect Bitmask of INTF IDs this CDM can connect to + */ +struct dpu_cdm_cfg { + DPU_HW_BLK_INFO; + unsigned long intf_connect; +}; + +/** + * struct dpu_intf_cfg - information of timing engine blocks + * @id enum identifying this block + * @base register offset of this block + * @features bit mask identifying sub-blocks/features + * @type: Interface type(DSI, DP, HDMI) + * @controller_id: Controller Instance ID in case of multiple of intf type + * @prog_fetch_lines_worst_case Worst case latency num lines needed to prefetch + */ +struct dpu_intf_cfg { + DPU_HW_BLK_INFO; + u32 type; /* interface type*/ + u32 controller_id; + u32 prog_fetch_lines_worst_case; +}; + +/** + * struct dpu_vbif_dynamic_ot_cfg - dynamic OT setting + * @pps pixel per seconds + * @ot_limit OT limit to use up to specified pixel per second + */ +struct dpu_vbif_dynamic_ot_cfg { + u64 pps; + u32 ot_limit; +}; + +/** + * struct dpu_vbif_dynamic_ot_tbl - dynamic OT setting table + * @count length of cfg + * @cfg pointer to array of configuration settings with + * ascending requirements + */ +struct dpu_vbif_dynamic_ot_tbl { + u32 count; + struct dpu_vbif_dynamic_ot_cfg *cfg; +}; + +/** + * struct dpu_vbif_qos_tbl - QoS priority table + * @npriority_lvl num of priority level + * @priority_lvl pointer to array of priority level in ascending order + */ +struct dpu_vbif_qos_tbl { + u32 npriority_lvl; + u32 *priority_lvl; +}; + +/** + * struct dpu_vbif_cfg - information of VBIF blocks + * @id enum identifying this block + * @base register offset of this block + * @features bit mask identifying sub-blocks/features + * @ot_rd_limit default OT read limit + * @ot_wr_limit default OT write limit + * @xin_halt_timeout maximum time (in usec) for xin to halt + * @dynamic_ot_rd_tbl dynamic OT read configuration table + * @dynamic_ot_wr_tbl dynamic OT write configuration table + * @qos_rt_tbl real-time QoS priority table + * @qos_nrt_tbl non-real-time QoS priority table + * @memtype_count number of defined memtypes + * @memtype array of xin memtype definitions + */ +struct dpu_vbif_cfg { + DPU_HW_BLK_INFO; + u32 default_ot_rd_limit; + u32 default_ot_wr_limit; + u32 xin_halt_timeout; + struct dpu_vbif_dynamic_ot_tbl dynamic_ot_rd_tbl; + struct dpu_vbif_dynamic_ot_tbl dynamic_ot_wr_tbl; + struct dpu_vbif_qos_tbl qos_rt_tbl; + struct dpu_vbif_qos_tbl qos_nrt_tbl; + u32 memtype_count; + u32 memtype[MAX_XIN_COUNT]; +}; +/** + * struct dpu_reg_dma_cfg - information of lut dma blocks + * @id enum identifying this block + * @base register offset of this block + * @features bit mask identifying sub-blocks/features + * @version version of lutdma hw block + * @trigger_sel_off offset to trigger select registers of lutdma + */ +struct dpu_reg_dma_cfg { + DPU_HW_BLK_INFO; + u32 version; + u32 trigger_sel_off; +}; + +/** + * Define CDP use cases + * @DPU_PERF_CDP_UDAGE_RT: real-time use cases + * @DPU_PERF_CDP_USAGE_NRT: non real-time use cases such as WFD + */ +enum { + DPU_PERF_CDP_USAGE_RT, + DPU_PERF_CDP_USAGE_NRT, + DPU_PERF_CDP_USAGE_MAX +}; + +/** + * struct dpu_perf_cdp_cfg - define CDP use case configuration + * @rd_enable: true if read pipe CDP is enabled + * @wr_enable: true if write pipe CDP is enabled + */ +struct dpu_perf_cdp_cfg { + bool rd_enable; + bool wr_enable; +}; + +/** + * struct dpu_perf_cfg - performance control settings + * @max_bw_low low threshold of maximum bandwidth (kbps) + * @max_bw_high high threshold of maximum bandwidth (kbps) + * @min_core_ib minimum bandwidth for core (kbps) + * @min_core_ib minimum mnoc ib vote in kbps + * @min_llcc_ib minimum llcc ib vote in kbps + * @min_dram_ib minimum dram ib vote in kbps + * @core_ib_ff core instantaneous bandwidth fudge factor + * @core_clk_ff core clock fudge factor + * @comp_ratio_rt string of 0 or more of <fourcc>/<ven>/<mod>/<comp ratio> + * @comp_ratio_nrt string of 0 or more of <fourcc>/<ven>/<mod>/<comp ratio> + * @undersized_prefill_lines undersized prefill in lines + * @xtra_prefill_lines extra prefill latency in lines + * @dest_scale_prefill_lines destination scaler latency in lines + * @macrotile_perfill_lines macrotile latency in lines + * @yuv_nv12_prefill_lines yuv_nv12 latency in lines + * @linear_prefill_lines linear latency in lines + * @downscaling_prefill_lines downscaling latency in lines + * @amortizable_theshold minimum y position for traffic shaping prefill + * @min_prefill_lines minimum pipeline latency in lines + * @safe_lut_tbl: LUT tables for safe signals + * @danger_lut_tbl: LUT tables for danger signals + * @qos_lut_tbl: LUT tables for QoS signals + * @cdp_cfg cdp use case configurations + */ +struct dpu_perf_cfg { + u32 max_bw_low; + u32 max_bw_high; + u32 min_core_ib; + u32 min_llcc_ib; + u32 min_dram_ib; + const char *core_ib_ff; + const char *core_clk_ff; + const char *comp_ratio_rt; + const char *comp_ratio_nrt; + u32 undersized_prefill_lines; + u32 xtra_prefill_lines; + u32 dest_scale_prefill_lines; + u32 macrotile_prefill_lines; + u32 yuv_nv12_prefill_lines; + u32 linear_prefill_lines; + u32 downscaling_prefill_lines; + u32 amortizable_threshold; + u32 min_prefill_lines; + u32 safe_lut_tbl[DPU_QOS_LUT_USAGE_MAX]; + u32 danger_lut_tbl[DPU_QOS_LUT_USAGE_MAX]; + struct dpu_qos_lut_tbl qos_lut_tbl[DPU_QOS_LUT_USAGE_MAX]; + struct dpu_perf_cdp_cfg cdp_cfg[DPU_PERF_CDP_USAGE_MAX]; +}; + +/** + * struct dpu_mdss_cfg - information of MDSS HW + * This is the main catalog data structure representing + * this HW version. Contains number of instances, + * register offsets, capabilities of the all MDSS HW sub-blocks. + * + * @dma_formats Supported formats for dma pipe + * @cursor_formats Supported formats for cursor pipe + * @vig_formats Supported formats for vig pipe + */ +struct dpu_mdss_cfg { + u32 hwversion; + + const struct dpu_caps *caps; + + u32 mdp_count; + struct dpu_mdp_cfg *mdp; + + u32 ctl_count; + struct dpu_ctl_cfg *ctl; + + u32 sspp_count; + struct dpu_sspp_cfg *sspp; + + u32 mixer_count; + struct dpu_lm_cfg *mixer; + + u32 ds_count; + struct dpu_ds_cfg *ds; + + u32 pingpong_count; + struct dpu_pingpong_cfg *pingpong; + + u32 cdm_count; + struct dpu_cdm_cfg *cdm; + + u32 intf_count; + struct dpu_intf_cfg *intf; + + u32 vbif_count; + struct dpu_vbif_cfg *vbif; + + u32 reg_dma_count; + struct dpu_reg_dma_cfg dma_cfg; + + u32 ad_count; + + /* Add additional block data structures here */ + + struct dpu_perf_cfg perf; + struct dpu_format_extended *dma_formats; + struct dpu_format_extended *cursor_formats; + struct dpu_format_extended *vig_formats; +}; + +struct dpu_mdss_hw_cfg_handler { + u32 hw_rev; + void (*cfg_init)(struct dpu_mdss_cfg *dpu_cfg); +}; + +/* + * Access Macros + */ +#define BLK_MDP(s) ((s)->mdp) +#define BLK_CTL(s) ((s)->ctl) +#define BLK_VIG(s) ((s)->vig) +#define BLK_RGB(s) ((s)->rgb) +#define BLK_DMA(s) ((s)->dma) +#define BLK_CURSOR(s) ((s)->cursor) +#define BLK_MIXER(s) ((s)->mixer) +#define BLK_DS(s) ((s)->ds) +#define BLK_PINGPONG(s) ((s)->pingpong) +#define BLK_CDM(s) ((s)->cdm) +#define BLK_INTF(s) ((s)->intf) +#define BLK_AD(s) ((s)->ad) + +/** + * dpu_hw_catalog_init - dpu hardware catalog init API retrieves + * hardcoded target specific catalog information in config structure + * @hw_rev: caller needs provide the hardware revision. + * + * Return: dpu config structure + */ +struct dpu_mdss_cfg *dpu_hw_catalog_init(u32 hw_rev); + +/** + * dpu_hw_catalog_deinit - dpu hardware catalog cleanup + * @dpu_cfg: pointer returned from init function + */ +void dpu_hw_catalog_deinit(struct dpu_mdss_cfg *dpu_cfg); + +/** + * dpu_hw_sspp_multirect_enabled - check multirect enabled for the sspp + * @cfg: pointer to sspp cfg + */ +static inline bool dpu_hw_sspp_multirect_enabled(const struct dpu_sspp_cfg *cfg) +{ + return test_bit(DPU_SSPP_SMART_DMA_V1, &cfg->features) || + test_bit(DPU_SSPP_SMART_DMA_V2, &cfg->features); +} +#endif /* _DPU_HW_CATALOG_H */ diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_catalog_format.h b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_catalog_format.h new file mode 100644 index 000000000000..3c9f028628ef --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_catalog_format.h @@ -0,0 +1,168 @@ +/* Copyright (c) 2015-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#include "dpu_hw_mdss.h" + +static const struct dpu_format_extended plane_formats[] = { + {DRM_FORMAT_ARGB8888, 0}, + {DRM_FORMAT_ABGR8888, 0}, + {DRM_FORMAT_RGBA8888, 0}, + {DRM_FORMAT_ABGR8888, DRM_FORMAT_MOD_QCOM_COMPRESSED}, + {DRM_FORMAT_BGRA8888, 0}, + {DRM_FORMAT_XRGB8888, 0}, + {DRM_FORMAT_RGBX8888, 0}, + {DRM_FORMAT_BGRX8888, 0}, + {DRM_FORMAT_XBGR8888, 0}, + {DRM_FORMAT_XBGR8888, DRM_FORMAT_MOD_QCOM_COMPRESSED}, + {DRM_FORMAT_RGB888, 0}, + {DRM_FORMAT_BGR888, 0}, + {DRM_FORMAT_RGB565, 0}, + {DRM_FORMAT_BGR565, DRM_FORMAT_MOD_QCOM_COMPRESSED}, + {DRM_FORMAT_BGR565, 0}, + {DRM_FORMAT_ARGB1555, 0}, + {DRM_FORMAT_ABGR1555, 0}, + {DRM_FORMAT_RGBA5551, 0}, + {DRM_FORMAT_BGRA5551, 0}, + {DRM_FORMAT_XRGB1555, 0}, + {DRM_FORMAT_XBGR1555, 0}, + {DRM_FORMAT_RGBX5551, 0}, + {DRM_FORMAT_BGRX5551, 0}, + {DRM_FORMAT_ARGB4444, 0}, + {DRM_FORMAT_ABGR4444, 0}, + {DRM_FORMAT_RGBA4444, 0}, + {DRM_FORMAT_BGRA4444, 0}, + {DRM_FORMAT_XRGB4444, 0}, + {DRM_FORMAT_XBGR4444, 0}, + {DRM_FORMAT_RGBX4444, 0}, + {DRM_FORMAT_BGRX4444, 0}, + {0, 0}, +}; + +static const struct dpu_format_extended plane_formats_yuv[] = { + {DRM_FORMAT_ARGB8888, 0}, + {DRM_FORMAT_ABGR8888, 0}, + {DRM_FORMAT_RGBA8888, 0}, + {DRM_FORMAT_BGRX8888, 0}, + {DRM_FORMAT_ABGR8888, DRM_FORMAT_MOD_QCOM_COMPRESSED}, + {DRM_FORMAT_BGRA8888, 0}, + {DRM_FORMAT_XRGB8888, 0}, + {DRM_FORMAT_XBGR8888, 0}, + {DRM_FORMAT_RGBX8888, 0}, + {DRM_FORMAT_XBGR8888, DRM_FORMAT_MOD_QCOM_COMPRESSED}, + {DRM_FORMAT_RGB888, 0}, + {DRM_FORMAT_BGR888, 0}, + {DRM_FORMAT_RGB565, 0}, + {DRM_FORMAT_BGR565, DRM_FORMAT_MOD_QCOM_COMPRESSED}, + {DRM_FORMAT_BGR565, 0}, + {DRM_FORMAT_ARGB1555, 0}, + {DRM_FORMAT_ABGR1555, 0}, + {DRM_FORMAT_RGBA5551, 0}, + {DRM_FORMAT_BGRA5551, 0}, + {DRM_FORMAT_XRGB1555, 0}, + {DRM_FORMAT_XBGR1555, 0}, + {DRM_FORMAT_RGBX5551, 0}, + {DRM_FORMAT_BGRX5551, 0}, + {DRM_FORMAT_ARGB4444, 0}, + {DRM_FORMAT_ABGR4444, 0}, + {DRM_FORMAT_RGBA4444, 0}, + {DRM_FORMAT_BGRA4444, 0}, + {DRM_FORMAT_XRGB4444, 0}, + {DRM_FORMAT_XBGR4444, 0}, + {DRM_FORMAT_RGBX4444, 0}, + {DRM_FORMAT_BGRX4444, 0}, + + {DRM_FORMAT_NV12, 0}, + {DRM_FORMAT_NV12, DRM_FORMAT_MOD_QCOM_COMPRESSED}, + {DRM_FORMAT_NV21, 0}, + {DRM_FORMAT_NV16, 0}, + {DRM_FORMAT_NV61, 0}, + {DRM_FORMAT_VYUY, 0}, + {DRM_FORMAT_UYVY, 0}, + {DRM_FORMAT_YUYV, 0}, + {DRM_FORMAT_YVYU, 0}, + {DRM_FORMAT_YUV420, 0}, + {DRM_FORMAT_YVU420, 0}, + {0, 0}, +}; + +static const struct dpu_format_extended cursor_formats[] = { + {DRM_FORMAT_ARGB8888, 0}, + {DRM_FORMAT_ABGR8888, 0}, + {DRM_FORMAT_RGBA8888, 0}, + {DRM_FORMAT_BGRA8888, 0}, + {DRM_FORMAT_XRGB8888, 0}, + {DRM_FORMAT_ARGB1555, 0}, + {DRM_FORMAT_ABGR1555, 0}, + {DRM_FORMAT_RGBA5551, 0}, + {DRM_FORMAT_BGRA5551, 0}, + {DRM_FORMAT_ARGB4444, 0}, + {DRM_FORMAT_ABGR4444, 0}, + {DRM_FORMAT_RGBA4444, 0}, + {DRM_FORMAT_BGRA4444, 0}, + {0, 0}, +}; + +static const struct dpu_format_extended wb2_formats[] = { + {DRM_FORMAT_RGB565, 0}, + {DRM_FORMAT_BGR565, DRM_FORMAT_MOD_QCOM_COMPRESSED}, + {DRM_FORMAT_RGB888, 0}, + {DRM_FORMAT_ARGB8888, 0}, + {DRM_FORMAT_RGBA8888, 0}, + {DRM_FORMAT_ABGR8888, DRM_FORMAT_MOD_QCOM_COMPRESSED}, + {DRM_FORMAT_XRGB8888, 0}, + {DRM_FORMAT_RGBX8888, 0}, + {DRM_FORMAT_XBGR8888, DRM_FORMAT_MOD_QCOM_COMPRESSED}, + {DRM_FORMAT_ARGB1555, 0}, + {DRM_FORMAT_RGBA5551, 0}, + {DRM_FORMAT_XRGB1555, 0}, + {DRM_FORMAT_RGBX5551, 0}, + {DRM_FORMAT_ARGB4444, 0}, + {DRM_FORMAT_RGBA4444, 0}, + {DRM_FORMAT_RGBX4444, 0}, + {DRM_FORMAT_XRGB4444, 0}, + + {DRM_FORMAT_BGR565, 0}, + {DRM_FORMAT_BGR888, 0}, + {DRM_FORMAT_ABGR8888, 0}, + {DRM_FORMAT_BGRA8888, 0}, + {DRM_FORMAT_BGRX8888, 0}, + {DRM_FORMAT_XBGR8888, 0}, + {DRM_FORMAT_ABGR1555, 0}, + {DRM_FORMAT_BGRA5551, 0}, + {DRM_FORMAT_XBGR1555, 0}, + {DRM_FORMAT_BGRX5551, 0}, + {DRM_FORMAT_ABGR4444, 0}, + {DRM_FORMAT_BGRA4444, 0}, + {DRM_FORMAT_BGRX4444, 0}, + {DRM_FORMAT_XBGR4444, 0}, + + {DRM_FORMAT_YUV420, 0}, + {DRM_FORMAT_NV12, 0}, + {DRM_FORMAT_NV12, DRM_FORMAT_MOD_QCOM_COMPRESSED}, + {DRM_FORMAT_NV16, 0}, + {DRM_FORMAT_YUYV, 0}, + + {0, 0}, +}; + +static const struct dpu_format_extended rgb_10bit_formats[] = { + {DRM_FORMAT_BGRA1010102, 0}, + {DRM_FORMAT_BGRX1010102, 0}, + {DRM_FORMAT_RGBA1010102, 0}, + {DRM_FORMAT_RGBX1010102, 0}, + {DRM_FORMAT_ABGR2101010, 0}, + {DRM_FORMAT_ABGR2101010, DRM_FORMAT_MOD_QCOM_COMPRESSED}, + {DRM_FORMAT_XBGR2101010, 0}, + {DRM_FORMAT_XBGR2101010, DRM_FORMAT_MOD_QCOM_COMPRESSED}, + {DRM_FORMAT_ARGB2101010, 0}, + {DRM_FORMAT_XRGB2101010, 0}, +}; diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_cdm.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_cdm.c new file mode 100644 index 000000000000..554874ba0c3b --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_cdm.c @@ -0,0 +1,323 @@ +/* Copyright (c) 2015-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#include "dpu_hw_mdss.h" +#include "dpu_hwio.h" +#include "dpu_hw_catalog.h" +#include "dpu_hw_cdm.h" +#include "dpu_dbg.h" +#include "dpu_kms.h" + +#define CDM_CSC_10_OPMODE 0x000 +#define CDM_CSC_10_BASE 0x004 + +#define CDM_CDWN2_OP_MODE 0x100 +#define CDM_CDWN2_CLAMP_OUT 0x104 +#define CDM_CDWN2_PARAMS_3D_0 0x108 +#define CDM_CDWN2_PARAMS_3D_1 0x10C +#define CDM_CDWN2_COEFF_COSITE_H_0 0x110 +#define CDM_CDWN2_COEFF_COSITE_H_1 0x114 +#define CDM_CDWN2_COEFF_COSITE_H_2 0x118 +#define CDM_CDWN2_COEFF_OFFSITE_H_0 0x11C +#define CDM_CDWN2_COEFF_OFFSITE_H_1 0x120 +#define CDM_CDWN2_COEFF_OFFSITE_H_2 0x124 +#define CDM_CDWN2_COEFF_COSITE_V 0x128 +#define CDM_CDWN2_COEFF_OFFSITE_V 0x12C +#define CDM_CDWN2_OUT_SIZE 0x130 + +#define CDM_HDMI_PACK_OP_MODE 0x200 +#define CDM_CSC_10_MATRIX_COEFF_0 0x004 + +/** + * Horizontal coefficients for cosite chroma downscale + * s13 representation of coefficients + */ +static u32 cosite_h_coeff[] = {0x00000016, 0x000001cc, 0x0100009e}; + +/** + * Horizontal coefficients for offsite chroma downscale + */ +static u32 offsite_h_coeff[] = {0x000b0005, 0x01db01eb, 0x00e40046}; + +/** + * Vertical coefficients for cosite chroma downscale + */ +static u32 cosite_v_coeff[] = {0x00080004}; +/** + * Vertical coefficients for offsite chroma downscale + */ +static u32 offsite_v_coeff[] = {0x00060002}; + +/* Limited Range rgb2yuv coeff with clamp and bias values for CSC 10 module */ +static struct dpu_csc_cfg rgb2yuv_cfg = { + { + 0x0083, 0x0102, 0x0032, + 0x1fb5, 0x1f6c, 0x00e1, + 0x00e1, 0x1f45, 0x1fdc + }, + { 0x00, 0x00, 0x00 }, + { 0x0040, 0x0200, 0x0200 }, + { 0x000, 0x3ff, 0x000, 0x3ff, 0x000, 0x3ff }, + { 0x040, 0x3ac, 0x040, 0x3c0, 0x040, 0x3c0 }, +}; + +static struct dpu_cdm_cfg *_cdm_offset(enum dpu_cdm cdm, + struct dpu_mdss_cfg *m, + void __iomem *addr, + struct dpu_hw_blk_reg_map *b) +{ + int i; + + for (i = 0; i < m->cdm_count; i++) { + if (cdm == m->cdm[i].id) { + b->base_off = addr; + b->blk_off = m->cdm[i].base; + b->length = m->cdm[i].len; + b->hwversion = m->hwversion; + b->log_mask = DPU_DBG_MASK_CDM; + return &m->cdm[i]; + } + } + + return ERR_PTR(-EINVAL); +} + +static int dpu_hw_cdm_setup_csc_10bit(struct dpu_hw_cdm *ctx, + struct dpu_csc_cfg *data) +{ + dpu_hw_csc_setup(&ctx->hw, CDM_CSC_10_MATRIX_COEFF_0, data, true); + + return 0; +} + +static int dpu_hw_cdm_setup_cdwn(struct dpu_hw_cdm *ctx, + struct dpu_hw_cdm_cfg *cfg) +{ + struct dpu_hw_blk_reg_map *c = &ctx->hw; + u32 opmode = 0; + u32 out_size = 0; + + if (cfg->output_bit_depth == CDM_CDWN_OUTPUT_10BIT) + opmode &= ~BIT(7); + else + opmode |= BIT(7); + + /* ENABLE DWNS_H bit */ + opmode |= BIT(1); + + switch (cfg->h_cdwn_type) { + case CDM_CDWN_DISABLE: + /* CLEAR METHOD_H field */ + opmode &= ~(0x18); + /* CLEAR DWNS_H bit */ + opmode &= ~BIT(1); + break; + case CDM_CDWN_PIXEL_DROP: + /* Clear METHOD_H field (pixel drop is 0) */ + opmode &= ~(0x18); + break; + case CDM_CDWN_AVG: + /* Clear METHOD_H field (Average is 0x1) */ + opmode &= ~(0x18); + opmode |= (0x1 << 0x3); + break; + case CDM_CDWN_COSITE: + /* Clear METHOD_H field (Average is 0x2) */ + opmode &= ~(0x18); + opmode |= (0x2 << 0x3); + /* Co-site horizontal coefficients */ + DPU_REG_WRITE(c, CDM_CDWN2_COEFF_COSITE_H_0, + cosite_h_coeff[0]); + DPU_REG_WRITE(c, CDM_CDWN2_COEFF_COSITE_H_1, + cosite_h_coeff[1]); + DPU_REG_WRITE(c, CDM_CDWN2_COEFF_COSITE_H_2, + cosite_h_coeff[2]); + break; + case CDM_CDWN_OFFSITE: + /* Clear METHOD_H field (Average is 0x3) */ + opmode &= ~(0x18); + opmode |= (0x3 << 0x3); + + /* Off-site horizontal coefficients */ + DPU_REG_WRITE(c, CDM_CDWN2_COEFF_OFFSITE_H_0, + offsite_h_coeff[0]); + DPU_REG_WRITE(c, CDM_CDWN2_COEFF_OFFSITE_H_1, + offsite_h_coeff[1]); + DPU_REG_WRITE(c, CDM_CDWN2_COEFF_OFFSITE_H_2, + offsite_h_coeff[2]); + break; + default: + pr_err("%s invalid horz down sampling type\n", __func__); + return -EINVAL; + } + + /* ENABLE DWNS_V bit */ + opmode |= BIT(2); + + switch (cfg->v_cdwn_type) { + case CDM_CDWN_DISABLE: + /* CLEAR METHOD_V field */ + opmode &= ~(0x60); + /* CLEAR DWNS_V bit */ + opmode &= ~BIT(2); + break; + case CDM_CDWN_PIXEL_DROP: + /* Clear METHOD_V field (pixel drop is 0) */ + opmode &= ~(0x60); + break; + case CDM_CDWN_AVG: + /* Clear METHOD_V field (Average is 0x1) */ + opmode &= ~(0x60); + opmode |= (0x1 << 0x5); + break; + case CDM_CDWN_COSITE: + /* Clear METHOD_V field (Average is 0x2) */ + opmode &= ~(0x60); + opmode |= (0x2 << 0x5); + /* Co-site vertical coefficients */ + DPU_REG_WRITE(c, + CDM_CDWN2_COEFF_COSITE_V, + cosite_v_coeff[0]); + break; + case CDM_CDWN_OFFSITE: + /* Clear METHOD_V field (Average is 0x3) */ + opmode &= ~(0x60); + opmode |= (0x3 << 0x5); + + /* Off-site vertical coefficients */ + DPU_REG_WRITE(c, + CDM_CDWN2_COEFF_OFFSITE_V, + offsite_v_coeff[0]); + break; + default: + return -EINVAL; + } + + if (cfg->v_cdwn_type || cfg->h_cdwn_type) + opmode |= BIT(0); /* EN CDWN module */ + else + opmode &= ~BIT(0); + + out_size = (cfg->output_width & 0xFFFF) | + ((cfg->output_height & 0xFFFF) << 16); + DPU_REG_WRITE(c, CDM_CDWN2_OUT_SIZE, out_size); + DPU_REG_WRITE(c, CDM_CDWN2_OP_MODE, opmode); + DPU_REG_WRITE(c, CDM_CDWN2_CLAMP_OUT, + ((0x3FF << 16) | 0x0)); + + return 0; +} + +static int dpu_hw_cdm_enable(struct dpu_hw_cdm *ctx, + struct dpu_hw_cdm_cfg *cdm) +{ + struct dpu_hw_blk_reg_map *c = &ctx->hw; + const struct dpu_format *fmt = cdm->output_fmt; + struct cdm_output_cfg cdm_cfg = { 0 }; + u32 opmode = 0; + u32 csc = 0; + + if (!DPU_FORMAT_IS_YUV(fmt)) + return -EINVAL; + + if (cdm->output_type == CDM_CDWN_OUTPUT_HDMI) { + if (fmt->chroma_sample != DPU_CHROMA_H1V2) + return -EINVAL; /*unsupported format */ + opmode = BIT(0); + opmode |= (fmt->chroma_sample << 1); + cdm_cfg.intf_en = true; + } + + csc |= BIT(2); + csc &= ~BIT(1); + csc |= BIT(0); + + if (ctx->hw_mdp && ctx->hw_mdp->ops.setup_cdm_output) + ctx->hw_mdp->ops.setup_cdm_output(ctx->hw_mdp, &cdm_cfg); + + DPU_REG_WRITE(c, CDM_CSC_10_OPMODE, csc); + DPU_REG_WRITE(c, CDM_HDMI_PACK_OP_MODE, opmode); + return 0; +} + +static void dpu_hw_cdm_disable(struct dpu_hw_cdm *ctx) +{ + struct cdm_output_cfg cdm_cfg = { 0 }; + + if (ctx->hw_mdp && ctx->hw_mdp->ops.setup_cdm_output) + ctx->hw_mdp->ops.setup_cdm_output(ctx->hw_mdp, &cdm_cfg); +} + +static void _setup_cdm_ops(struct dpu_hw_cdm_ops *ops, + unsigned long features) +{ + ops->setup_csc_data = dpu_hw_cdm_setup_csc_10bit; + ops->setup_cdwn = dpu_hw_cdm_setup_cdwn; + ops->enable = dpu_hw_cdm_enable; + ops->disable = dpu_hw_cdm_disable; +} + +static struct dpu_hw_blk_ops dpu_hw_ops = { + .start = NULL, + .stop = NULL, +}; + +struct dpu_hw_cdm *dpu_hw_cdm_init(enum dpu_cdm idx, + void __iomem *addr, + struct dpu_mdss_cfg *m, + struct dpu_hw_mdp *hw_mdp) +{ + struct dpu_hw_cdm *c; + struct dpu_cdm_cfg *cfg; + int rc; + + c = kzalloc(sizeof(*c), GFP_KERNEL); + if (!c) + return ERR_PTR(-ENOMEM); + + cfg = _cdm_offset(idx, m, addr, &c->hw); + if (IS_ERR_OR_NULL(cfg)) { + kfree(c); + return ERR_PTR(-EINVAL); + } + + c->idx = idx; + c->caps = cfg; + _setup_cdm_ops(&c->ops, c->caps->features); + c->hw_mdp = hw_mdp; + + rc = dpu_hw_blk_init(&c->base, DPU_HW_BLK_CDM, idx, &dpu_hw_ops); + if (rc) { + DPU_ERROR("failed to init hw blk %d\n", rc); + goto blk_init_error; + } + + /* + * Perform any default initialization for the chroma down module + * @setup default csc coefficients + */ + dpu_hw_cdm_setup_csc_10bit(c, &rgb2yuv_cfg); + + return c; + +blk_init_error: + kzfree(c); + + return ERR_PTR(rc); +} + +void dpu_hw_cdm_destroy(struct dpu_hw_cdm *cdm) +{ + if (cdm) + dpu_hw_blk_destroy(&cdm->base); + kfree(cdm); +} diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_cdm.h b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_cdm.h new file mode 100644 index 000000000000..5cceb1ecb8e0 --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_cdm.h @@ -0,0 +1,139 @@ +/* Copyright (c) 2015-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#ifndef _DPU_HW_CDM_H +#define _DPU_HW_CDM_H + +#include "dpu_hw_mdss.h" +#include "dpu_hw_top.h" +#include "dpu_hw_blk.h" + +struct dpu_hw_cdm; + +struct dpu_hw_cdm_cfg { + u32 output_width; + u32 output_height; + u32 output_bit_depth; + u32 h_cdwn_type; + u32 v_cdwn_type; + const struct dpu_format *output_fmt; + u32 output_type; + int flags; +}; + +enum dpu_hw_cdwn_type { + CDM_CDWN_DISABLE, + CDM_CDWN_PIXEL_DROP, + CDM_CDWN_AVG, + CDM_CDWN_COSITE, + CDM_CDWN_OFFSITE, +}; + +enum dpu_hw_cdwn_output_type { + CDM_CDWN_OUTPUT_HDMI, + CDM_CDWN_OUTPUT_WB, +}; + +enum dpu_hw_cdwn_output_bit_depth { + CDM_CDWN_OUTPUT_8BIT, + CDM_CDWN_OUTPUT_10BIT, +}; + +/** + * struct dpu_hw_cdm_ops : Interface to the chroma down Hw driver functions + * Assumption is these functions will be called after + * clocks are enabled + * @setup_csc: Programs the csc matrix + * @setup_cdwn: Sets up the chroma down sub module + * @enable: Enables the output to interface and programs the + * output packer + * @disable: Puts the cdm in bypass mode + */ +struct dpu_hw_cdm_ops { + /** + * Programs the CSC matrix for conversion from RGB space to YUV space, + * it is optional to call this function as this matrix is automatically + * set during initialization, user should call this if it wants + * to program a different matrix than default matrix. + * @cdm: Pointer to the chroma down context structure + * @data Pointer to CSC configuration data + * return: 0 if success; error code otherwise + */ + int (*setup_csc_data)(struct dpu_hw_cdm *cdm, + struct dpu_csc_cfg *data); + + /** + * Programs the Chroma downsample part. + * @cdm Pointer to chroma down context + */ + int (*setup_cdwn)(struct dpu_hw_cdm *cdm, + struct dpu_hw_cdm_cfg *cfg); + + /** + * Enable the CDM module + * @cdm Pointer to chroma down context + */ + int (*enable)(struct dpu_hw_cdm *cdm, + struct dpu_hw_cdm_cfg *cfg); + + /** + * Disable the CDM module + * @cdm Pointer to chroma down context + */ + void (*disable)(struct dpu_hw_cdm *cdm); +}; + +struct dpu_hw_cdm { + struct dpu_hw_blk base; + struct dpu_hw_blk_reg_map hw; + + /* chroma down */ + const struct dpu_cdm_cfg *caps; + enum dpu_cdm idx; + + /* mdp top hw driver */ + struct dpu_hw_mdp *hw_mdp; + + /* ops */ + struct dpu_hw_cdm_ops ops; +}; + +/** + * dpu_hw_cdm - convert base object dpu_hw_base to container + * @hw: Pointer to base hardware block + * return: Pointer to hardware block container + */ +static inline struct dpu_hw_cdm *to_dpu_hw_cdm(struct dpu_hw_blk *hw) +{ + return container_of(hw, struct dpu_hw_cdm, base); +} + +/** + * dpu_hw_cdm_init - initializes the cdm hw driver object. + * should be called once before accessing every cdm. + * @idx: cdm index for which driver object is required + * @addr: mapped register io address of MDP + * @m : pointer to mdss catalog data + * @hw_mdp: pointer to mdp top hw driver object + */ +struct dpu_hw_cdm *dpu_hw_cdm_init(enum dpu_cdm idx, + void __iomem *addr, + struct dpu_mdss_cfg *m, + struct dpu_hw_mdp *hw_mdp); + +/** + * dpu_hw_cdm_destroy - destroys CDM driver context + * @cdm: pointer to CDM driver context + */ +void dpu_hw_cdm_destroy(struct dpu_hw_cdm *cdm); + +#endif /*_DPU_HW_CDM_H */ diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_ctl.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_ctl.c new file mode 100644 index 000000000000..06be7cf7ce50 --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_ctl.c @@ -0,0 +1,540 @@ +/* Copyright (c) 2015-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#include <linux/delay.h> +#include "dpu_hwio.h" +#include "dpu_hw_ctl.h" +#include "dpu_dbg.h" +#include "dpu_kms.h" + +#define CTL_LAYER(lm) \ + (((lm) == LM_5) ? (0x024) : (((lm) - LM_0) * 0x004)) +#define CTL_LAYER_EXT(lm) \ + (0x40 + (((lm) - LM_0) * 0x004)) +#define CTL_LAYER_EXT2(lm) \ + (0x70 + (((lm) - LM_0) * 0x004)) +#define CTL_LAYER_EXT3(lm) \ + (0xA0 + (((lm) - LM_0) * 0x004)) +#define CTL_TOP 0x014 +#define CTL_FLUSH 0x018 +#define CTL_START 0x01C +#define CTL_PREPARE 0x0d0 +#define CTL_SW_RESET 0x030 +#define CTL_LAYER_EXTN_OFFSET 0x40 + +#define CTL_MIXER_BORDER_OUT BIT(24) +#define CTL_FLUSH_MASK_CTL BIT(17) + +#define DPU_REG_RESET_TIMEOUT_US 2000 + +static struct dpu_ctl_cfg *_ctl_offset(enum dpu_ctl ctl, + struct dpu_mdss_cfg *m, + void __iomem *addr, + struct dpu_hw_blk_reg_map *b) +{ + int i; + + for (i = 0; i < m->ctl_count; i++) { + if (ctl == m->ctl[i].id) { + b->base_off = addr; + b->blk_off = m->ctl[i].base; + b->length = m->ctl[i].len; + b->hwversion = m->hwversion; + b->log_mask = DPU_DBG_MASK_CTL; + return &m->ctl[i]; + } + } + return ERR_PTR(-ENOMEM); +} + +static int _mixer_stages(const struct dpu_lm_cfg *mixer, int count, + enum dpu_lm lm) +{ + int i; + int stages = -EINVAL; + + for (i = 0; i < count; i++) { + if (lm == mixer[i].id) { + stages = mixer[i].sblk->maxblendstages; + break; + } + } + + return stages; +} + +static inline void dpu_hw_ctl_trigger_start(struct dpu_hw_ctl *ctx) +{ + DPU_REG_WRITE(&ctx->hw, CTL_START, 0x1); +} + +static inline void dpu_hw_ctl_trigger_pending(struct dpu_hw_ctl *ctx) +{ + DPU_REG_WRITE(&ctx->hw, CTL_PREPARE, 0x1); +} + +static inline void dpu_hw_ctl_clear_pending_flush(struct dpu_hw_ctl *ctx) +{ + ctx->pending_flush_mask = 0x0; +} + +static inline void dpu_hw_ctl_update_pending_flush(struct dpu_hw_ctl *ctx, + u32 flushbits) +{ + ctx->pending_flush_mask |= flushbits; +} + +static u32 dpu_hw_ctl_get_pending_flush(struct dpu_hw_ctl *ctx) +{ + if (!ctx) + return 0x0; + + return ctx->pending_flush_mask; +} + +static inline void dpu_hw_ctl_trigger_flush(struct dpu_hw_ctl *ctx) +{ + + DPU_REG_WRITE(&ctx->hw, CTL_FLUSH, ctx->pending_flush_mask); +} + +static inline u32 dpu_hw_ctl_get_flush_register(struct dpu_hw_ctl *ctx) +{ + struct dpu_hw_blk_reg_map *c = &ctx->hw; + + return DPU_REG_READ(c, CTL_FLUSH); +} + +static inline uint32_t dpu_hw_ctl_get_bitmask_sspp(struct dpu_hw_ctl *ctx, + enum dpu_sspp sspp) +{ + uint32_t flushbits = 0; + + switch (sspp) { + case SSPP_VIG0: + flushbits = BIT(0); + break; + case SSPP_VIG1: + flushbits = BIT(1); + break; + case SSPP_VIG2: + flushbits = BIT(2); + break; + case SSPP_VIG3: + flushbits = BIT(18); + break; + case SSPP_RGB0: + flushbits = BIT(3); + break; + case SSPP_RGB1: + flushbits = BIT(4); + break; + case SSPP_RGB2: + flushbits = BIT(5); + break; + case SSPP_RGB3: + flushbits = BIT(19); + break; + case SSPP_DMA0: + flushbits = BIT(11); + break; + case SSPP_DMA1: + flushbits = BIT(12); + break; + case SSPP_DMA2: + flushbits = BIT(24); + break; + case SSPP_DMA3: + flushbits = BIT(25); + break; + case SSPP_CURSOR0: + flushbits = BIT(22); + break; + case SSPP_CURSOR1: + flushbits = BIT(23); + break; + default: + break; + } + + return flushbits; +} + +static inline uint32_t dpu_hw_ctl_get_bitmask_mixer(struct dpu_hw_ctl *ctx, + enum dpu_lm lm) +{ + uint32_t flushbits = 0; + + switch (lm) { + case LM_0: + flushbits = BIT(6); + break; + case LM_1: + flushbits = BIT(7); + break; + case LM_2: + flushbits = BIT(8); + break; + case LM_3: + flushbits = BIT(9); + break; + case LM_4: + flushbits = BIT(10); + break; + case LM_5: + flushbits = BIT(20); + break; + default: + return -EINVAL; + } + + flushbits |= CTL_FLUSH_MASK_CTL; + + return flushbits; +} + +static inline int dpu_hw_ctl_get_bitmask_intf(struct dpu_hw_ctl *ctx, + u32 *flushbits, enum dpu_intf intf) +{ + switch (intf) { + case INTF_0: + *flushbits |= BIT(31); + break; + case INTF_1: + *flushbits |= BIT(30); + break; + case INTF_2: + *flushbits |= BIT(29); + break; + case INTF_3: + *flushbits |= BIT(28); + break; + default: + return -EINVAL; + } + return 0; +} + +static inline int dpu_hw_ctl_get_bitmask_cdm(struct dpu_hw_ctl *ctx, + u32 *flushbits, enum dpu_cdm cdm) +{ + switch (cdm) { + case CDM_0: + *flushbits |= BIT(26); + break; + default: + return -EINVAL; + } + return 0; +} + +static u32 dpu_hw_ctl_poll_reset_status(struct dpu_hw_ctl *ctx, u32 timeout_us) +{ + struct dpu_hw_blk_reg_map *c = &ctx->hw; + ktime_t timeout; + u32 status; + + timeout = ktime_add_us(ktime_get(), timeout_us); + + /* + * it takes around 30us to have mdp finish resetting its ctl path + * poll every 50us so that reset should be completed at 1st poll + */ + do { + status = DPU_REG_READ(c, CTL_SW_RESET); + status &= 0x1; + if (status) + usleep_range(20, 50); + } while (status && ktime_compare_safe(ktime_get(), timeout) < 0); + + return status; +} + +static int dpu_hw_ctl_reset_control(struct dpu_hw_ctl *ctx) +{ + struct dpu_hw_blk_reg_map *c = &ctx->hw; + + pr_debug("issuing hw ctl reset for ctl:%d\n", ctx->idx); + DPU_REG_WRITE(c, CTL_SW_RESET, 0x1); + if (dpu_hw_ctl_poll_reset_status(ctx, DPU_REG_RESET_TIMEOUT_US)) + return -EINVAL; + + return 0; +} + +static int dpu_hw_ctl_wait_reset_status(struct dpu_hw_ctl *ctx) +{ + struct dpu_hw_blk_reg_map *c = &ctx->hw; + u32 status; + + status = DPU_REG_READ(c, CTL_SW_RESET); + status &= 0x01; + if (!status) + return 0; + + pr_debug("hw ctl reset is set for ctl:%d\n", ctx->idx); + if (dpu_hw_ctl_poll_reset_status(ctx, DPU_REG_RESET_TIMEOUT_US)) { + pr_err("hw recovery is not complete for ctl:%d\n", ctx->idx); + return -EINVAL; + } + + return 0; +} + +static void dpu_hw_ctl_clear_all_blendstages(struct dpu_hw_ctl *ctx) +{ + struct dpu_hw_blk_reg_map *c = &ctx->hw; + int i; + + for (i = 0; i < ctx->mixer_count; i++) { + DPU_REG_WRITE(c, CTL_LAYER(LM_0 + i), 0); + DPU_REG_WRITE(c, CTL_LAYER_EXT(LM_0 + i), 0); + DPU_REG_WRITE(c, CTL_LAYER_EXT2(LM_0 + i), 0); + DPU_REG_WRITE(c, CTL_LAYER_EXT3(LM_0 + i), 0); + } +} + +static void dpu_hw_ctl_setup_blendstage(struct dpu_hw_ctl *ctx, + enum dpu_lm lm, struct dpu_hw_stage_cfg *stage_cfg) +{ + struct dpu_hw_blk_reg_map *c = &ctx->hw; + u32 mixercfg = 0, mixercfg_ext = 0, mix, ext; + u32 mixercfg_ext2 = 0, mixercfg_ext3 = 0; + int i, j; + u8 stages; + int pipes_per_stage; + + stages = _mixer_stages(ctx->mixer_hw_caps, ctx->mixer_count, lm); + if (stages < 0) + return; + + if (test_bit(DPU_MIXER_SOURCESPLIT, + &ctx->mixer_hw_caps->features)) + pipes_per_stage = PIPES_PER_STAGE; + else + pipes_per_stage = 1; + + mixercfg = CTL_MIXER_BORDER_OUT; /* always set BORDER_OUT */ + + if (!stage_cfg) + goto exit; + + for (i = 0; i <= stages; i++) { + /* overflow to ext register if 'i + 1 > 7' */ + mix = (i + 1) & 0x7; + ext = i >= 7; + + for (j = 0 ; j < pipes_per_stage; j++) { + enum dpu_sspp_multirect_index rect_index = + stage_cfg->multirect_index[i][j]; + + switch (stage_cfg->stage[i][j]) { + case SSPP_VIG0: + if (rect_index == DPU_SSPP_RECT_1) { + mixercfg_ext3 |= ((i + 1) & 0xF) << 0; + } else { + mixercfg |= mix << 0; + mixercfg_ext |= ext << 0; + } + break; + case SSPP_VIG1: + if (rect_index == DPU_SSPP_RECT_1) { + mixercfg_ext3 |= ((i + 1) & 0xF) << 4; + } else { + mixercfg |= mix << 3; + mixercfg_ext |= ext << 2; + } + break; + case SSPP_VIG2: + if (rect_index == DPU_SSPP_RECT_1) { + mixercfg_ext3 |= ((i + 1) & 0xF) << 8; + } else { + mixercfg |= mix << 6; + mixercfg_ext |= ext << 4; + } + break; + case SSPP_VIG3: + if (rect_index == DPU_SSPP_RECT_1) { + mixercfg_ext3 |= ((i + 1) & 0xF) << 12; + } else { + mixercfg |= mix << 26; + mixercfg_ext |= ext << 6; + } + break; + case SSPP_RGB0: + mixercfg |= mix << 9; + mixercfg_ext |= ext << 8; + break; + case SSPP_RGB1: + mixercfg |= mix << 12; + mixercfg_ext |= ext << 10; + break; + case SSPP_RGB2: + mixercfg |= mix << 15; + mixercfg_ext |= ext << 12; + break; + case SSPP_RGB3: + mixercfg |= mix << 29; + mixercfg_ext |= ext << 14; + break; + case SSPP_DMA0: + if (rect_index == DPU_SSPP_RECT_1) { + mixercfg_ext2 |= ((i + 1) & 0xF) << 8; + } else { + mixercfg |= mix << 18; + mixercfg_ext |= ext << 16; + } + break; + case SSPP_DMA1: + if (rect_index == DPU_SSPP_RECT_1) { + mixercfg_ext2 |= ((i + 1) & 0xF) << 12; + } else { + mixercfg |= mix << 21; + mixercfg_ext |= ext << 18; + } + break; + case SSPP_DMA2: + if (rect_index == DPU_SSPP_RECT_1) { + mixercfg_ext2 |= ((i + 1) & 0xF) << 16; + } else { + mix |= (i + 1) & 0xF; + mixercfg_ext2 |= mix << 0; + } + break; + case SSPP_DMA3: + if (rect_index == DPU_SSPP_RECT_1) { + mixercfg_ext2 |= ((i + 1) & 0xF) << 20; + } else { + mix |= (i + 1) & 0xF; + mixercfg_ext2 |= mix << 4; + } + break; + case SSPP_CURSOR0: + mixercfg_ext |= ((i + 1) & 0xF) << 20; + break; + case SSPP_CURSOR1: + mixercfg_ext |= ((i + 1) & 0xF) << 26; + break; + default: + break; + } + } + } + +exit: + DPU_REG_WRITE(c, CTL_LAYER(lm), mixercfg); + DPU_REG_WRITE(c, CTL_LAYER_EXT(lm), mixercfg_ext); + DPU_REG_WRITE(c, CTL_LAYER_EXT2(lm), mixercfg_ext2); + DPU_REG_WRITE(c, CTL_LAYER_EXT3(lm), mixercfg_ext3); +} + +static void dpu_hw_ctl_intf_cfg(struct dpu_hw_ctl *ctx, + struct dpu_hw_intf_cfg *cfg) +{ + struct dpu_hw_blk_reg_map *c = &ctx->hw; + u32 intf_cfg = 0; + + intf_cfg |= (cfg->intf & 0xF) << 4; + + if (cfg->mode_3d) { + intf_cfg |= BIT(19); + intf_cfg |= (cfg->mode_3d - 0x1) << 20; + } + + switch (cfg->intf_mode_sel) { + case DPU_CTL_MODE_SEL_VID: + intf_cfg &= ~BIT(17); + intf_cfg &= ~(0x3 << 15); + break; + case DPU_CTL_MODE_SEL_CMD: + intf_cfg |= BIT(17); + intf_cfg |= ((cfg->stream_sel & 0x3) << 15); + break; + default: + pr_err("unknown interface type %d\n", cfg->intf_mode_sel); + return; + } + + DPU_REG_WRITE(c, CTL_TOP, intf_cfg); +} + +static void _setup_ctl_ops(struct dpu_hw_ctl_ops *ops, + unsigned long cap) +{ + ops->clear_pending_flush = dpu_hw_ctl_clear_pending_flush; + ops->update_pending_flush = dpu_hw_ctl_update_pending_flush; + ops->get_pending_flush = dpu_hw_ctl_get_pending_flush; + ops->trigger_flush = dpu_hw_ctl_trigger_flush; + ops->get_flush_register = dpu_hw_ctl_get_flush_register; + ops->trigger_start = dpu_hw_ctl_trigger_start; + ops->trigger_pending = dpu_hw_ctl_trigger_pending; + ops->setup_intf_cfg = dpu_hw_ctl_intf_cfg; + ops->reset = dpu_hw_ctl_reset_control; + ops->wait_reset_status = dpu_hw_ctl_wait_reset_status; + ops->clear_all_blendstages = dpu_hw_ctl_clear_all_blendstages; + ops->setup_blendstage = dpu_hw_ctl_setup_blendstage; + ops->get_bitmask_sspp = dpu_hw_ctl_get_bitmask_sspp; + ops->get_bitmask_mixer = dpu_hw_ctl_get_bitmask_mixer; + ops->get_bitmask_intf = dpu_hw_ctl_get_bitmask_intf; + ops->get_bitmask_cdm = dpu_hw_ctl_get_bitmask_cdm; +}; + +static struct dpu_hw_blk_ops dpu_hw_ops = { + .start = NULL, + .stop = NULL, +}; + +struct dpu_hw_ctl *dpu_hw_ctl_init(enum dpu_ctl idx, + void __iomem *addr, + struct dpu_mdss_cfg *m) +{ + struct dpu_hw_ctl *c; + struct dpu_ctl_cfg *cfg; + int rc; + + c = kzalloc(sizeof(*c), GFP_KERNEL); + if (!c) + return ERR_PTR(-ENOMEM); + + cfg = _ctl_offset(idx, m, addr, &c->hw); + if (IS_ERR_OR_NULL(cfg)) { + kfree(c); + pr_err("failed to create dpu_hw_ctl %d\n", idx); + return ERR_PTR(-EINVAL); + } + + c->caps = cfg; + _setup_ctl_ops(&c->ops, c->caps->features); + c->idx = idx; + c->mixer_count = m->mixer_count; + c->mixer_hw_caps = m->mixer; + + rc = dpu_hw_blk_init(&c->base, DPU_HW_BLK_CTL, idx, &dpu_hw_ops); + if (rc) { + DPU_ERROR("failed to init hw blk %d\n", rc); + goto blk_init_error; + } + + return c; + +blk_init_error: + kzfree(c); + + return ERR_PTR(rc); +} + +void dpu_hw_ctl_destroy(struct dpu_hw_ctl *ctx) +{ + if (ctx) + dpu_hw_blk_destroy(&ctx->base); + kfree(ctx); +} diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_ctl.h b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_ctl.h new file mode 100644 index 000000000000..c66a71f8b839 --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_ctl.h @@ -0,0 +1,218 @@ +/* Copyright (c) 2015-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#ifndef _DPU_HW_CTL_H +#define _DPU_HW_CTL_H + +#include "dpu_hw_mdss.h" +#include "dpu_hw_util.h" +#include "dpu_hw_catalog.h" +#include "dpu_hw_sspp.h" +#include "dpu_hw_blk.h" + +/** + * dpu_ctl_mode_sel: Interface mode selection + * DPU_CTL_MODE_SEL_VID: Video mode interface + * DPU_CTL_MODE_SEL_CMD: Command mode interface + */ +enum dpu_ctl_mode_sel { + DPU_CTL_MODE_SEL_VID = 0, + DPU_CTL_MODE_SEL_CMD +}; + +struct dpu_hw_ctl; +/** + * struct dpu_hw_stage_cfg - blending stage cfg + * @stage : SSPP_ID at each stage + * @multirect_index: index of the rectangle of SSPP. + */ +struct dpu_hw_stage_cfg { + enum dpu_sspp stage[DPU_STAGE_MAX][PIPES_PER_STAGE]; + enum dpu_sspp_multirect_index multirect_index + [DPU_STAGE_MAX][PIPES_PER_STAGE]; +}; + +/** + * struct dpu_hw_intf_cfg :Describes how the DPU writes data to output interface + * @intf : Interface id + * @mode_3d: 3d mux configuration + * @intf_mode_sel: Interface mode, cmd / vid + * @stream_sel: Stream selection for multi-stream interfaces + */ +struct dpu_hw_intf_cfg { + enum dpu_intf intf; + enum dpu_3d_blend_mode mode_3d; + enum dpu_ctl_mode_sel intf_mode_sel; + int stream_sel; +}; + +/** + * struct dpu_hw_ctl_ops - Interface to the wb Hw driver functions + * Assumption is these functions will be called after clocks are enabled + */ +struct dpu_hw_ctl_ops { + /** + * kickoff hw operation for Sw controlled interfaces + * DSI cmd mode and WB interface are SW controlled + * @ctx : ctl path ctx pointer + */ + void (*trigger_start)(struct dpu_hw_ctl *ctx); + + /** + * kickoff prepare is in progress hw operation for sw + * controlled interfaces: DSI cmd mode and WB interface + * are SW controlled + * @ctx : ctl path ctx pointer + */ + void (*trigger_pending)(struct dpu_hw_ctl *ctx); + + /** + * Clear the value of the cached pending_flush_mask + * No effect on hardware + * @ctx : ctl path ctx pointer + */ + void (*clear_pending_flush)(struct dpu_hw_ctl *ctx); + + /** + * Query the value of the cached pending_flush_mask + * No effect on hardware + * @ctx : ctl path ctx pointer + */ + u32 (*get_pending_flush)(struct dpu_hw_ctl *ctx); + + /** + * OR in the given flushbits to the cached pending_flush_mask + * No effect on hardware + * @ctx : ctl path ctx pointer + * @flushbits : module flushmask + */ + void (*update_pending_flush)(struct dpu_hw_ctl *ctx, + u32 flushbits); + + /** + * Write the value of the pending_flush_mask to hardware + * @ctx : ctl path ctx pointer + */ + void (*trigger_flush)(struct dpu_hw_ctl *ctx); + + /** + * Read the value of the flush register + * @ctx : ctl path ctx pointer + * @Return: value of the ctl flush register. + */ + u32 (*get_flush_register)(struct dpu_hw_ctl *ctx); + + /** + * Setup ctl_path interface config + * @ctx + * @cfg : interface config structure pointer + */ + void (*setup_intf_cfg)(struct dpu_hw_ctl *ctx, + struct dpu_hw_intf_cfg *cfg); + + int (*reset)(struct dpu_hw_ctl *c); + + /* + * wait_reset_status - checks ctl reset status + * @ctx : ctl path ctx pointer + * + * This function checks the ctl reset status bit. + * If the reset bit is set, it keeps polling the status till the hw + * reset is complete. + * Returns: 0 on success or -error if reset incomplete within interval + */ + int (*wait_reset_status)(struct dpu_hw_ctl *ctx); + + uint32_t (*get_bitmask_sspp)(struct dpu_hw_ctl *ctx, + enum dpu_sspp blk); + + uint32_t (*get_bitmask_mixer)(struct dpu_hw_ctl *ctx, + enum dpu_lm blk); + + int (*get_bitmask_intf)(struct dpu_hw_ctl *ctx, + u32 *flushbits, + enum dpu_intf blk); + + int (*get_bitmask_cdm)(struct dpu_hw_ctl *ctx, + u32 *flushbits, + enum dpu_cdm blk); + + /** + * Set all blend stages to disabled + * @ctx : ctl path ctx pointer + */ + void (*clear_all_blendstages)(struct dpu_hw_ctl *ctx); + + /** + * Configure layer mixer to pipe configuration + * @ctx : ctl path ctx pointer + * @lm : layer mixer enumeration + * @cfg : blend stage configuration + */ + void (*setup_blendstage)(struct dpu_hw_ctl *ctx, + enum dpu_lm lm, struct dpu_hw_stage_cfg *cfg); +}; + +/** + * struct dpu_hw_ctl : CTL PATH driver object + * @base: hardware block base structure + * @hw: block register map object + * @idx: control path index + * @caps: control path capabilities + * @mixer_count: number of mixers + * @mixer_hw_caps: mixer hardware capabilities + * @pending_flush_mask: storage for pending ctl_flush managed via ops + * @ops: operation list + */ +struct dpu_hw_ctl { + struct dpu_hw_blk base; + struct dpu_hw_blk_reg_map hw; + + /* ctl path */ + int idx; + const struct dpu_ctl_cfg *caps; + int mixer_count; + const struct dpu_lm_cfg *mixer_hw_caps; + u32 pending_flush_mask; + + /* ops */ + struct dpu_hw_ctl_ops ops; +}; + +/** + * dpu_hw_ctl - convert base object dpu_hw_base to container + * @hw: Pointer to base hardware block + * return: Pointer to hardware block container + */ +static inline struct dpu_hw_ctl *to_dpu_hw_ctl(struct dpu_hw_blk *hw) +{ + return container_of(hw, struct dpu_hw_ctl, base); +} + +/** + * dpu_hw_ctl_init(): Initializes the ctl_path hw driver object. + * should be called before accessing every ctl path registers. + * @idx: ctl_path index for which driver object is required + * @addr: mapped register io address of MDP + * @m : pointer to mdss catalog data + */ +struct dpu_hw_ctl *dpu_hw_ctl_init(enum dpu_ctl idx, + void __iomem *addr, + struct dpu_mdss_cfg *m); + +/** + * dpu_hw_ctl_destroy(): Destroys ctl driver context + * should be called to free the context + */ +void dpu_hw_ctl_destroy(struct dpu_hw_ctl *ctx); + +#endif /*_DPU_HW_CTL_H */ diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_interrupts.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_interrupts.c new file mode 100644 index 000000000000..c0b7f0049365 --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_interrupts.c @@ -0,0 +1,1183 @@ +/* Copyright (c) 2016-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#include <linux/bitops.h> +#include <linux/slab.h> + +#include "dpu_kms.h" +#include "dpu_hw_interrupts.h" +#include "dpu_hw_util.h" +#include "dpu_hw_mdss.h" + +/** + * Register offsets in MDSS register file for the interrupt registers + * w.r.t. to the MDP base + */ +#define MDP_SSPP_TOP0_OFF 0x0 +#define MDP_INTF_0_OFF 0x6A000 +#define MDP_INTF_1_OFF 0x6A800 +#define MDP_INTF_2_OFF 0x6B000 +#define MDP_INTF_3_OFF 0x6B800 +#define MDP_INTF_4_OFF 0x6C000 +#define MDP_AD4_0_OFF 0x7C000 +#define MDP_AD4_1_OFF 0x7D000 +#define MDP_AD4_INTR_EN_OFF 0x41c +#define MDP_AD4_INTR_CLEAR_OFF 0x424 +#define MDP_AD4_INTR_STATUS_OFF 0x420 + +/** + * WB interrupt status bit definitions + */ +#define DPU_INTR_WB_0_DONE BIT(0) +#define DPU_INTR_WB_1_DONE BIT(1) +#define DPU_INTR_WB_2_DONE BIT(4) + +/** + * WDOG timer interrupt status bit definitions + */ +#define DPU_INTR_WD_TIMER_0_DONE BIT(2) +#define DPU_INTR_WD_TIMER_1_DONE BIT(3) +#define DPU_INTR_WD_TIMER_2_DONE BIT(5) +#define DPU_INTR_WD_TIMER_3_DONE BIT(6) +#define DPU_INTR_WD_TIMER_4_DONE BIT(7) + +/** + * Pingpong interrupt status bit definitions + */ +#define DPU_INTR_PING_PONG_0_DONE BIT(8) +#define DPU_INTR_PING_PONG_1_DONE BIT(9) +#define DPU_INTR_PING_PONG_2_DONE BIT(10) +#define DPU_INTR_PING_PONG_3_DONE BIT(11) +#define DPU_INTR_PING_PONG_0_RD_PTR BIT(12) +#define DPU_INTR_PING_PONG_1_RD_PTR BIT(13) +#define DPU_INTR_PING_PONG_2_RD_PTR BIT(14) +#define DPU_INTR_PING_PONG_3_RD_PTR BIT(15) +#define DPU_INTR_PING_PONG_0_WR_PTR BIT(16) +#define DPU_INTR_PING_PONG_1_WR_PTR BIT(17) +#define DPU_INTR_PING_PONG_2_WR_PTR BIT(18) +#define DPU_INTR_PING_PONG_3_WR_PTR BIT(19) +#define DPU_INTR_PING_PONG_0_AUTOREFRESH_DONE BIT(20) +#define DPU_INTR_PING_PONG_1_AUTOREFRESH_DONE BIT(21) +#define DPU_INTR_PING_PONG_2_AUTOREFRESH_DONE BIT(22) +#define DPU_INTR_PING_PONG_3_AUTOREFRESH_DONE BIT(23) + +/** + * Interface interrupt status bit definitions + */ +#define DPU_INTR_INTF_0_UNDERRUN BIT(24) +#define DPU_INTR_INTF_1_UNDERRUN BIT(26) +#define DPU_INTR_INTF_2_UNDERRUN BIT(28) +#define DPU_INTR_INTF_3_UNDERRUN BIT(30) +#define DPU_INTR_INTF_0_VSYNC BIT(25) +#define DPU_INTR_INTF_1_VSYNC BIT(27) +#define DPU_INTR_INTF_2_VSYNC BIT(29) +#define DPU_INTR_INTF_3_VSYNC BIT(31) + +/** + * Pingpong Secondary interrupt status bit definitions + */ +#define DPU_INTR_PING_PONG_S0_AUTOREFRESH_DONE BIT(0) +#define DPU_INTR_PING_PONG_S0_WR_PTR BIT(4) +#define DPU_INTR_PING_PONG_S0_RD_PTR BIT(8) +#define DPU_INTR_PING_PONG_S0_TEAR_DETECTED BIT(22) +#define DPU_INTR_PING_PONG_S0_TE_DETECTED BIT(28) + +/** + * Pingpong TEAR detection interrupt status bit definitions + */ +#define DPU_INTR_PING_PONG_0_TEAR_DETECTED BIT(16) +#define DPU_INTR_PING_PONG_1_TEAR_DETECTED BIT(17) +#define DPU_INTR_PING_PONG_2_TEAR_DETECTED BIT(18) +#define DPU_INTR_PING_PONG_3_TEAR_DETECTED BIT(19) + +/** + * Pingpong TE detection interrupt status bit definitions + */ +#define DPU_INTR_PING_PONG_0_TE_DETECTED BIT(24) +#define DPU_INTR_PING_PONG_1_TE_DETECTED BIT(25) +#define DPU_INTR_PING_PONG_2_TE_DETECTED BIT(26) +#define DPU_INTR_PING_PONG_3_TE_DETECTED BIT(27) + +/** + * Ctl start interrupt status bit definitions + */ +#define DPU_INTR_CTL_0_START BIT(9) +#define DPU_INTR_CTL_1_START BIT(10) +#define DPU_INTR_CTL_2_START BIT(11) +#define DPU_INTR_CTL_3_START BIT(12) +#define DPU_INTR_CTL_4_START BIT(13) + +/** + * Concurrent WB overflow interrupt status bit definitions + */ +#define DPU_INTR_CWB_2_OVERFLOW BIT(14) +#define DPU_INTR_CWB_3_OVERFLOW BIT(15) + +/** + * Histogram VIG done interrupt status bit definitions + */ +#define DPU_INTR_HIST_VIG_0_DONE BIT(0) +#define DPU_INTR_HIST_VIG_1_DONE BIT(4) +#define DPU_INTR_HIST_VIG_2_DONE BIT(8) +#define DPU_INTR_HIST_VIG_3_DONE BIT(10) + +/** + * Histogram VIG reset Sequence done interrupt status bit definitions + */ +#define DPU_INTR_HIST_VIG_0_RSTSEQ_DONE BIT(1) +#define DPU_INTR_HIST_VIG_1_RSTSEQ_DONE BIT(5) +#define DPU_INTR_HIST_VIG_2_RSTSEQ_DONE BIT(9) +#define DPU_INTR_HIST_VIG_3_RSTSEQ_DONE BIT(11) + +/** + * Histogram DSPP done interrupt status bit definitions + */ +#define DPU_INTR_HIST_DSPP_0_DONE BIT(12) +#define DPU_INTR_HIST_DSPP_1_DONE BIT(16) +#define DPU_INTR_HIST_DSPP_2_DONE BIT(20) +#define DPU_INTR_HIST_DSPP_3_DONE BIT(22) + +/** + * Histogram DSPP reset Sequence done interrupt status bit definitions + */ +#define DPU_INTR_HIST_DSPP_0_RSTSEQ_DONE BIT(13) +#define DPU_INTR_HIST_DSPP_1_RSTSEQ_DONE BIT(17) +#define DPU_INTR_HIST_DSPP_2_RSTSEQ_DONE BIT(21) +#define DPU_INTR_HIST_DSPP_3_RSTSEQ_DONE BIT(23) + +/** + * INTF interrupt status bit definitions + */ +#define DPU_INTR_VIDEO_INTO_STATIC BIT(0) +#define DPU_INTR_VIDEO_OUTOF_STATIC BIT(1) +#define DPU_INTR_DSICMD_0_INTO_STATIC BIT(2) +#define DPU_INTR_DSICMD_0_OUTOF_STATIC BIT(3) +#define DPU_INTR_DSICMD_1_INTO_STATIC BIT(4) +#define DPU_INTR_DSICMD_1_OUTOF_STATIC BIT(5) +#define DPU_INTR_DSICMD_2_INTO_STATIC BIT(6) +#define DPU_INTR_DSICMD_2_OUTOF_STATIC BIT(7) +#define DPU_INTR_PROG_LINE BIT(8) + +/** + * AD4 interrupt status bit definitions + */ +#define DPU_INTR_BRIGHTPR_UPDATED BIT(4) +#define DPU_INTR_DARKENH_UPDATED BIT(3) +#define DPU_INTR_STREN_OUTROI_UPDATED BIT(2) +#define DPU_INTR_STREN_INROI_UPDATED BIT(1) +#define DPU_INTR_BACKLIGHT_UPDATED BIT(0) +/** + * struct dpu_intr_reg - array of DPU register sets + * @clr_off: offset to CLEAR reg + * @en_off: offset to ENABLE reg + * @status_off: offset to STATUS reg + */ +struct dpu_intr_reg { + u32 clr_off; + u32 en_off; + u32 status_off; +}; + +/** + * struct dpu_irq_type - maps each irq with i/f + * @intr_type: type of interrupt listed in dpu_intr_type + * @instance_idx: instance index of the associated HW block in DPU + * @irq_mask: corresponding bit in the interrupt status reg + * @reg_idx: which reg set to use + */ +struct dpu_irq_type { + u32 intr_type; + u32 instance_idx; + u32 irq_mask; + u32 reg_idx; +}; + +/** + * List of DPU interrupt registers + */ +static const struct dpu_intr_reg dpu_intr_set[] = { + { + MDP_SSPP_TOP0_OFF+INTR_CLEAR, + MDP_SSPP_TOP0_OFF+INTR_EN, + MDP_SSPP_TOP0_OFF+INTR_STATUS + }, + { + MDP_SSPP_TOP0_OFF+INTR2_CLEAR, + MDP_SSPP_TOP0_OFF+INTR2_EN, + MDP_SSPP_TOP0_OFF+INTR2_STATUS + }, + { + MDP_SSPP_TOP0_OFF+HIST_INTR_CLEAR, + MDP_SSPP_TOP0_OFF+HIST_INTR_EN, + MDP_SSPP_TOP0_OFF+HIST_INTR_STATUS + }, + { + MDP_INTF_0_OFF+INTF_INTR_CLEAR, + MDP_INTF_0_OFF+INTF_INTR_EN, + MDP_INTF_0_OFF+INTF_INTR_STATUS + }, + { + MDP_INTF_1_OFF+INTF_INTR_CLEAR, + MDP_INTF_1_OFF+INTF_INTR_EN, + MDP_INTF_1_OFF+INTF_INTR_STATUS + }, + { + MDP_INTF_2_OFF+INTF_INTR_CLEAR, + MDP_INTF_2_OFF+INTF_INTR_EN, + MDP_INTF_2_OFF+INTF_INTR_STATUS + }, + { + MDP_INTF_3_OFF+INTF_INTR_CLEAR, + MDP_INTF_3_OFF+INTF_INTR_EN, + MDP_INTF_3_OFF+INTF_INTR_STATUS + }, + { + MDP_INTF_4_OFF+INTF_INTR_CLEAR, + MDP_INTF_4_OFF+INTF_INTR_EN, + MDP_INTF_4_OFF+INTF_INTR_STATUS + }, + { + MDP_AD4_0_OFF + MDP_AD4_INTR_CLEAR_OFF, + MDP_AD4_0_OFF + MDP_AD4_INTR_EN_OFF, + MDP_AD4_0_OFF + MDP_AD4_INTR_STATUS_OFF, + }, + { + MDP_AD4_1_OFF + MDP_AD4_INTR_CLEAR_OFF, + MDP_AD4_1_OFF + MDP_AD4_INTR_EN_OFF, + MDP_AD4_1_OFF + MDP_AD4_INTR_STATUS_OFF, + } +}; + +/** + * IRQ mapping table - use for lookup an irq_idx in this table that have + * a matching interface type and instance index. + */ +static const struct dpu_irq_type dpu_irq_map[] = { + /* BEGIN MAP_RANGE: 0-31, INTR */ + /* irq_idx: 0-3 */ + { DPU_IRQ_TYPE_WB_ROT_COMP, WB_0, DPU_INTR_WB_0_DONE, 0}, + { DPU_IRQ_TYPE_WB_ROT_COMP, WB_1, DPU_INTR_WB_1_DONE, 0}, + { DPU_IRQ_TYPE_WD_TIMER, WD_TIMER_0, DPU_INTR_WD_TIMER_0_DONE, 0}, + { DPU_IRQ_TYPE_WD_TIMER, WD_TIMER_1, DPU_INTR_WD_TIMER_1_DONE, 0}, + /* irq_idx: 4-7 */ + { DPU_IRQ_TYPE_WB_WFD_COMP, WB_2, DPU_INTR_WB_2_DONE, 0}, + { DPU_IRQ_TYPE_WD_TIMER, WD_TIMER_2, DPU_INTR_WD_TIMER_2_DONE, 0}, + { DPU_IRQ_TYPE_WD_TIMER, WD_TIMER_3, DPU_INTR_WD_TIMER_3_DONE, 0}, + { DPU_IRQ_TYPE_WD_TIMER, WD_TIMER_4, DPU_INTR_WD_TIMER_4_DONE, 0}, + /* irq_idx: 8-11 */ + { DPU_IRQ_TYPE_PING_PONG_COMP, PINGPONG_0, + DPU_INTR_PING_PONG_0_DONE, 0}, + { DPU_IRQ_TYPE_PING_PONG_COMP, PINGPONG_1, + DPU_INTR_PING_PONG_1_DONE, 0}, + { DPU_IRQ_TYPE_PING_PONG_COMP, PINGPONG_2, + DPU_INTR_PING_PONG_2_DONE, 0}, + { DPU_IRQ_TYPE_PING_PONG_COMP, PINGPONG_3, + DPU_INTR_PING_PONG_3_DONE, 0}, + /* irq_idx: 12-15 */ + { DPU_IRQ_TYPE_PING_PONG_RD_PTR, PINGPONG_0, + DPU_INTR_PING_PONG_0_RD_PTR, 0}, + { DPU_IRQ_TYPE_PING_PONG_RD_PTR, PINGPONG_1, + DPU_INTR_PING_PONG_1_RD_PTR, 0}, + { DPU_IRQ_TYPE_PING_PONG_RD_PTR, PINGPONG_2, + DPU_INTR_PING_PONG_2_RD_PTR, 0}, + { DPU_IRQ_TYPE_PING_PONG_RD_PTR, PINGPONG_3, + DPU_INTR_PING_PONG_3_RD_PTR, 0}, + /* irq_idx: 16-19 */ + { DPU_IRQ_TYPE_PING_PONG_WR_PTR, PINGPONG_0, + DPU_INTR_PING_PONG_0_WR_PTR, 0}, + { DPU_IRQ_TYPE_PING_PONG_WR_PTR, PINGPONG_1, + DPU_INTR_PING_PONG_1_WR_PTR, 0}, + { DPU_IRQ_TYPE_PING_PONG_WR_PTR, PINGPONG_2, + DPU_INTR_PING_PONG_2_WR_PTR, 0}, + { DPU_IRQ_TYPE_PING_PONG_WR_PTR, PINGPONG_3, + DPU_INTR_PING_PONG_3_WR_PTR, 0}, + /* irq_idx: 20-23 */ + { DPU_IRQ_TYPE_PING_PONG_AUTO_REF, PINGPONG_0, + DPU_INTR_PING_PONG_0_AUTOREFRESH_DONE, 0}, + { DPU_IRQ_TYPE_PING_PONG_AUTO_REF, PINGPONG_1, + DPU_INTR_PING_PONG_1_AUTOREFRESH_DONE, 0}, + { DPU_IRQ_TYPE_PING_PONG_AUTO_REF, PINGPONG_2, + DPU_INTR_PING_PONG_2_AUTOREFRESH_DONE, 0}, + { DPU_IRQ_TYPE_PING_PONG_AUTO_REF, PINGPONG_3, + DPU_INTR_PING_PONG_3_AUTOREFRESH_DONE, 0}, + /* irq_idx: 24-27 */ + { DPU_IRQ_TYPE_INTF_UNDER_RUN, INTF_0, DPU_INTR_INTF_0_UNDERRUN, 0}, + { DPU_IRQ_TYPE_INTF_VSYNC, INTF_0, DPU_INTR_INTF_0_VSYNC, 0}, + { DPU_IRQ_TYPE_INTF_UNDER_RUN, INTF_1, DPU_INTR_INTF_1_UNDERRUN, 0}, + { DPU_IRQ_TYPE_INTF_VSYNC, INTF_1, DPU_INTR_INTF_1_VSYNC, 0}, + /* irq_idx: 28-31 */ + { DPU_IRQ_TYPE_INTF_UNDER_RUN, INTF_2, DPU_INTR_INTF_2_UNDERRUN, 0}, + { DPU_IRQ_TYPE_INTF_VSYNC, INTF_2, DPU_INTR_INTF_2_VSYNC, 0}, + { DPU_IRQ_TYPE_INTF_UNDER_RUN, INTF_3, DPU_INTR_INTF_3_UNDERRUN, 0}, + { DPU_IRQ_TYPE_INTF_VSYNC, INTF_3, DPU_INTR_INTF_3_VSYNC, 0}, + + /* BEGIN MAP_RANGE: 32-64, INTR2 */ + /* irq_idx: 32-35 */ + { DPU_IRQ_TYPE_PING_PONG_AUTO_REF, PINGPONG_S0, + DPU_INTR_PING_PONG_S0_AUTOREFRESH_DONE, 1}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 1}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 1}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 1}, + /* irq_idx: 36-39 */ + { DPU_IRQ_TYPE_PING_PONG_WR_PTR, PINGPONG_S0, + DPU_INTR_PING_PONG_S0_WR_PTR, 1}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 1}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 1}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 1}, + /* irq_idx: 40 */ + { DPU_IRQ_TYPE_PING_PONG_RD_PTR, PINGPONG_S0, + DPU_INTR_PING_PONG_S0_RD_PTR, 1}, + /* irq_idx: 41-45 */ + { DPU_IRQ_TYPE_CTL_START, CTL_0, + DPU_INTR_CTL_0_START, 1}, + { DPU_IRQ_TYPE_CTL_START, CTL_1, + DPU_INTR_CTL_1_START, 1}, + { DPU_IRQ_TYPE_CTL_START, CTL_2, + DPU_INTR_CTL_2_START, 1}, + { DPU_IRQ_TYPE_CTL_START, CTL_3, + DPU_INTR_CTL_3_START, 1}, + { DPU_IRQ_TYPE_CTL_START, CTL_4, + DPU_INTR_CTL_4_START, 1}, + /* irq_idx: 46-47 */ + { DPU_IRQ_TYPE_CWB_OVERFLOW, CWB_2, DPU_INTR_CWB_2_OVERFLOW, 1}, + { DPU_IRQ_TYPE_CWB_OVERFLOW, CWB_3, DPU_INTR_CWB_3_OVERFLOW, 1}, + /* irq_idx: 48-51 */ + { DPU_IRQ_TYPE_PING_PONG_TEAR_CHECK, PINGPONG_0, + DPU_INTR_PING_PONG_0_TEAR_DETECTED, 1}, + { DPU_IRQ_TYPE_PING_PONG_TEAR_CHECK, PINGPONG_1, + DPU_INTR_PING_PONG_1_TEAR_DETECTED, 1}, + { DPU_IRQ_TYPE_PING_PONG_TEAR_CHECK, PINGPONG_2, + DPU_INTR_PING_PONG_2_TEAR_DETECTED, 1}, + { DPU_IRQ_TYPE_PING_PONG_TEAR_CHECK, PINGPONG_3, + DPU_INTR_PING_PONG_3_TEAR_DETECTED, 1}, + /* irq_idx: 52-55 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 1}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 1}, + { DPU_IRQ_TYPE_PING_PONG_TEAR_CHECK, PINGPONG_S0, + DPU_INTR_PING_PONG_S0_TEAR_DETECTED, 1}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 1}, + /* irq_idx: 56-59 */ + { DPU_IRQ_TYPE_PING_PONG_TE_CHECK, PINGPONG_0, + DPU_INTR_PING_PONG_0_TE_DETECTED, 1}, + { DPU_IRQ_TYPE_PING_PONG_TE_CHECK, PINGPONG_1, + DPU_INTR_PING_PONG_1_TE_DETECTED, 1}, + { DPU_IRQ_TYPE_PING_PONG_TE_CHECK, PINGPONG_2, + DPU_INTR_PING_PONG_2_TE_DETECTED, 1}, + { DPU_IRQ_TYPE_PING_PONG_TE_CHECK, PINGPONG_3, + DPU_INTR_PING_PONG_3_TE_DETECTED, 1}, + /* irq_idx: 60-63 */ + { DPU_IRQ_TYPE_PING_PONG_TE_CHECK, PINGPONG_S0, + DPU_INTR_PING_PONG_S0_TE_DETECTED, 1}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 1}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 1}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 1}, + + /* BEGIN MAP_RANGE: 64-95 HIST */ + /* irq_idx: 64-67 */ + { DPU_IRQ_TYPE_HIST_VIG_DONE, SSPP_VIG0, DPU_INTR_HIST_VIG_0_DONE, 2}, + { DPU_IRQ_TYPE_HIST_VIG_RSTSEQ, SSPP_VIG0, + DPU_INTR_HIST_VIG_0_RSTSEQ_DONE, 2}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 2}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 2}, + /* irq_idx: 68-71 */ + { DPU_IRQ_TYPE_HIST_VIG_DONE, SSPP_VIG1, DPU_INTR_HIST_VIG_1_DONE, 2}, + { DPU_IRQ_TYPE_HIST_VIG_RSTSEQ, SSPP_VIG1, + DPU_INTR_HIST_VIG_1_RSTSEQ_DONE, 2}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 2}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 2}, + /* irq_idx: 72-75 */ + { DPU_IRQ_TYPE_HIST_VIG_DONE, SSPP_VIG2, DPU_INTR_HIST_VIG_2_DONE, 2}, + { DPU_IRQ_TYPE_HIST_VIG_RSTSEQ, SSPP_VIG2, + DPU_INTR_HIST_VIG_2_RSTSEQ_DONE, 2}, + { DPU_IRQ_TYPE_HIST_VIG_DONE, SSPP_VIG3, DPU_INTR_HIST_VIG_3_DONE, 2}, + { DPU_IRQ_TYPE_HIST_VIG_RSTSEQ, SSPP_VIG3, + DPU_INTR_HIST_VIG_3_RSTSEQ_DONE, 2}, + /* irq_idx: 76-79 */ + { DPU_IRQ_TYPE_HIST_DSPP_DONE, DSPP_0, DPU_INTR_HIST_DSPP_0_DONE, 2}, + { DPU_IRQ_TYPE_HIST_DSPP_RSTSEQ, DSPP_0, + DPU_INTR_HIST_DSPP_0_RSTSEQ_DONE, 2}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 2}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 2}, + /* irq_idx: 80-83 */ + { DPU_IRQ_TYPE_HIST_DSPP_DONE, DSPP_1, DPU_INTR_HIST_DSPP_1_DONE, 2}, + { DPU_IRQ_TYPE_HIST_DSPP_RSTSEQ, DSPP_1, + DPU_INTR_HIST_DSPP_1_RSTSEQ_DONE, 2}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 2}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 2}, + /* irq_idx: 84-87 */ + { DPU_IRQ_TYPE_HIST_DSPP_DONE, DSPP_2, DPU_INTR_HIST_DSPP_2_DONE, 2}, + { DPU_IRQ_TYPE_HIST_DSPP_RSTSEQ, DSPP_2, + DPU_INTR_HIST_DSPP_2_RSTSEQ_DONE, 2}, + { DPU_IRQ_TYPE_HIST_DSPP_DONE, DSPP_3, DPU_INTR_HIST_DSPP_3_DONE, 2}, + { DPU_IRQ_TYPE_HIST_DSPP_RSTSEQ, DSPP_3, + DPU_INTR_HIST_DSPP_3_RSTSEQ_DONE, 2}, + /* irq_idx: 88-91 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 2}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 2}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 2}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 2}, + /* irq_idx: 92-95 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 2}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 2}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 2}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 2}, + + /* BEGIN MAP_RANGE: 96-127 INTF_0_INTR */ + /* irq_idx: 96-99 */ + { DPU_IRQ_TYPE_SFI_VIDEO_IN, INTF_0, + DPU_INTR_VIDEO_INTO_STATIC, 3}, + { DPU_IRQ_TYPE_SFI_VIDEO_OUT, INTF_0, + DPU_INTR_VIDEO_OUTOF_STATIC, 3}, + { DPU_IRQ_TYPE_SFI_CMD_0_IN, INTF_0, + DPU_INTR_DSICMD_0_INTO_STATIC, 3}, + { DPU_IRQ_TYPE_SFI_CMD_0_OUT, INTF_0, + DPU_INTR_DSICMD_0_OUTOF_STATIC, 3}, + /* irq_idx: 100-103 */ + { DPU_IRQ_TYPE_SFI_CMD_1_IN, INTF_0, + DPU_INTR_DSICMD_1_INTO_STATIC, 3}, + { DPU_IRQ_TYPE_SFI_CMD_1_OUT, INTF_0, + DPU_INTR_DSICMD_1_OUTOF_STATIC, 3}, + { DPU_IRQ_TYPE_SFI_CMD_2_IN, INTF_0, + DPU_INTR_DSICMD_2_INTO_STATIC, 3}, + { DPU_IRQ_TYPE_SFI_CMD_2_OUT, INTF_0, + DPU_INTR_DSICMD_2_OUTOF_STATIC, 3}, + /* irq_idx: 104-107 */ + { DPU_IRQ_TYPE_PROG_LINE, INTF_0, DPU_INTR_PROG_LINE, 3}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 3}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 3}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 3}, + /* irq_idx: 108-111 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 3}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 3}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 3}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 3}, + /* irq_idx: 112-115 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 3}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 3}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 3}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 3}, + /* irq_idx: 116-119 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 3}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 3}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 3}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 3}, + /* irq_idx: 120-123 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 3}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 3}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 3}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 3}, + /* irq_idx: 124-127 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 3}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 3}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 3}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 3}, + + /* BEGIN MAP_RANGE: 128-159 INTF_1_INTR */ + /* irq_idx: 128-131 */ + { DPU_IRQ_TYPE_SFI_VIDEO_IN, INTF_1, + DPU_INTR_VIDEO_INTO_STATIC, 4}, + { DPU_IRQ_TYPE_SFI_VIDEO_OUT, INTF_1, + DPU_INTR_VIDEO_OUTOF_STATIC, 4}, + { DPU_IRQ_TYPE_SFI_CMD_0_IN, INTF_1, + DPU_INTR_DSICMD_0_INTO_STATIC, 4}, + { DPU_IRQ_TYPE_SFI_CMD_0_OUT, INTF_1, + DPU_INTR_DSICMD_0_OUTOF_STATIC, 4}, + /* irq_idx: 132-135 */ + { DPU_IRQ_TYPE_SFI_CMD_1_IN, INTF_1, + DPU_INTR_DSICMD_1_INTO_STATIC, 4}, + { DPU_IRQ_TYPE_SFI_CMD_1_OUT, INTF_1, + DPU_INTR_DSICMD_1_OUTOF_STATIC, 4}, + { DPU_IRQ_TYPE_SFI_CMD_2_IN, INTF_1, + DPU_INTR_DSICMD_2_INTO_STATIC, 4}, + { DPU_IRQ_TYPE_SFI_CMD_2_OUT, INTF_1, + DPU_INTR_DSICMD_2_OUTOF_STATIC, 4}, + /* irq_idx: 136-139 */ + { DPU_IRQ_TYPE_PROG_LINE, INTF_1, DPU_INTR_PROG_LINE, 4}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 4}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 4}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 4}, + /* irq_idx: 140-143 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 4}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 4}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 4}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 4}, + /* irq_idx: 144-147 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 4}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 4}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 4}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 4}, + /* irq_idx: 148-151 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 4}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 4}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 4}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 4}, + /* irq_idx: 152-155 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 4}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 4}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 4}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 4}, + /* irq_idx: 156-159 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 4}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 4}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 4}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 4}, + + /* BEGIN MAP_RANGE: 160-191 INTF_2_INTR */ + /* irq_idx: 160-163 */ + { DPU_IRQ_TYPE_SFI_VIDEO_IN, INTF_2, + DPU_INTR_VIDEO_INTO_STATIC, 5}, + { DPU_IRQ_TYPE_SFI_VIDEO_OUT, INTF_2, + DPU_INTR_VIDEO_OUTOF_STATIC, 5}, + { DPU_IRQ_TYPE_SFI_CMD_0_IN, INTF_2, + DPU_INTR_DSICMD_0_INTO_STATIC, 5}, + { DPU_IRQ_TYPE_SFI_CMD_0_OUT, INTF_2, + DPU_INTR_DSICMD_0_OUTOF_STATIC, 5}, + /* irq_idx: 164-167 */ + { DPU_IRQ_TYPE_SFI_CMD_1_IN, INTF_2, + DPU_INTR_DSICMD_1_INTO_STATIC, 5}, + { DPU_IRQ_TYPE_SFI_CMD_1_OUT, INTF_2, + DPU_INTR_DSICMD_1_OUTOF_STATIC, 5}, + { DPU_IRQ_TYPE_SFI_CMD_2_IN, INTF_2, + DPU_INTR_DSICMD_2_INTO_STATIC, 5}, + { DPU_IRQ_TYPE_SFI_CMD_2_OUT, INTF_2, + DPU_INTR_DSICMD_2_OUTOF_STATIC, 5}, + /* irq_idx: 168-171 */ + { DPU_IRQ_TYPE_PROG_LINE, INTF_2, DPU_INTR_PROG_LINE, 5}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 5}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 5}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 5}, + /* irq_idx: 172-175 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 5}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 5}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 5}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 5}, + /* irq_idx: 176-179 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 5}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 5}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 5}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 5}, + /* irq_idx: 180-183 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 5}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 5}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 5}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 5}, + /* irq_idx: 184-187 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 5}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 5}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 5}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 5}, + /* irq_idx: 188-191 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 5}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 5}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 5}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 5}, + + /* BEGIN MAP_RANGE: 192-223 INTF_3_INTR */ + /* irq_idx: 192-195 */ + { DPU_IRQ_TYPE_SFI_VIDEO_IN, INTF_3, + DPU_INTR_VIDEO_INTO_STATIC, 6}, + { DPU_IRQ_TYPE_SFI_VIDEO_OUT, INTF_3, + DPU_INTR_VIDEO_OUTOF_STATIC, 6}, + { DPU_IRQ_TYPE_SFI_CMD_0_IN, INTF_3, + DPU_INTR_DSICMD_0_INTO_STATIC, 6}, + { DPU_IRQ_TYPE_SFI_CMD_0_OUT, INTF_3, + DPU_INTR_DSICMD_0_OUTOF_STATIC, 6}, + /* irq_idx: 196-199 */ + { DPU_IRQ_TYPE_SFI_CMD_1_IN, INTF_3, + DPU_INTR_DSICMD_1_INTO_STATIC, 6}, + { DPU_IRQ_TYPE_SFI_CMD_1_OUT, INTF_3, + DPU_INTR_DSICMD_1_OUTOF_STATIC, 6}, + { DPU_IRQ_TYPE_SFI_CMD_2_IN, INTF_3, + DPU_INTR_DSICMD_2_INTO_STATIC, 6}, + { DPU_IRQ_TYPE_SFI_CMD_2_OUT, INTF_3, + DPU_INTR_DSICMD_2_OUTOF_STATIC, 6}, + /* irq_idx: 200-203 */ + { DPU_IRQ_TYPE_PROG_LINE, INTF_3, DPU_INTR_PROG_LINE, 6}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 6}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 6}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 6}, + /* irq_idx: 204-207 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 6}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 6}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 6}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 6}, + /* irq_idx: 208-211 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 6}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 6}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 6}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 6}, + /* irq_idx: 212-215 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 6}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 6}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 6}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 6}, + /* irq_idx: 216-219 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 6}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 6}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 6}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 6}, + /* irq_idx: 220-223 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 6}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 6}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 6}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 6}, + + /* BEGIN MAP_RANGE: 224-255 INTF_4_INTR */ + /* irq_idx: 224-227 */ + { DPU_IRQ_TYPE_SFI_VIDEO_IN, INTF_4, + DPU_INTR_VIDEO_INTO_STATIC, 7}, + { DPU_IRQ_TYPE_SFI_VIDEO_OUT, INTF_4, + DPU_INTR_VIDEO_OUTOF_STATIC, 7}, + { DPU_IRQ_TYPE_SFI_CMD_0_IN, INTF_4, + DPU_INTR_DSICMD_0_INTO_STATIC, 7}, + { DPU_IRQ_TYPE_SFI_CMD_0_OUT, INTF_4, + DPU_INTR_DSICMD_0_OUTOF_STATIC, 7}, + /* irq_idx: 228-231 */ + { DPU_IRQ_TYPE_SFI_CMD_1_IN, INTF_4, + DPU_INTR_DSICMD_1_INTO_STATIC, 7}, + { DPU_IRQ_TYPE_SFI_CMD_1_OUT, INTF_4, + DPU_INTR_DSICMD_1_OUTOF_STATIC, 7}, + { DPU_IRQ_TYPE_SFI_CMD_2_IN, INTF_4, + DPU_INTR_DSICMD_2_INTO_STATIC, 7}, + { DPU_IRQ_TYPE_SFI_CMD_2_OUT, INTF_4, + DPU_INTR_DSICMD_2_OUTOF_STATIC, 7}, + /* irq_idx: 232-235 */ + { DPU_IRQ_TYPE_PROG_LINE, INTF_4, DPU_INTR_PROG_LINE, 7}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 7}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 7}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 7}, + /* irq_idx: 236-239 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 7}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 7}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 7}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 7}, + /* irq_idx: 240-243 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 7}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 7}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 7}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 7}, + /* irq_idx: 244-247 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 7}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 7}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 7}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 7}, + /* irq_idx: 248-251 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 7}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 7}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 7}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 7}, + /* irq_idx: 252-255 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 7}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 7}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 7}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 7}, + + /* BEGIN MAP_RANGE: 256-287 AD4_0_INTR */ + /* irq_idx: 256-259 */ + { DPU_IRQ_TYPE_AD4_BL_DONE, DSPP_0, DPU_INTR_BACKLIGHT_UPDATED, 8}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 8}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 8}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 8}, + /* irq_idx: 260-263 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 8}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 8}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 8}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 8}, + /* irq_idx: 264-267 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 8}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 8}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 8}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 8}, + /* irq_idx: 268-271 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 8}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 8}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 8}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 8}, + /* irq_idx: 272-275 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 8}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 8}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 8}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 8}, + /* irq_idx: 276-279 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 8}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 8}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 8}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 8}, + /* irq_idx: 280-283 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 8}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 8}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 8}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 8}, + /* irq_idx: 284-287 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 8}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 8}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 8}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 8}, + + /* BEGIN MAP_RANGE: 288-319 AD4_1_INTR */ + /* irq_idx: 288-291 */ + { DPU_IRQ_TYPE_AD4_BL_DONE, DSPP_1, DPU_INTR_BACKLIGHT_UPDATED, 9}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 9}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 9}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 9}, + /* irq_idx: 292-295 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 9}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 9}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 9}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 9}, + /* irq_idx: 296-299 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 9}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 9}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 9}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 9}, + /* irq_idx: 300-303 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 9}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 9}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 9}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 9}, + /* irq_idx: 304-307 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 9}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 9}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 9}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 9}, + /* irq_idx: 308-311 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 9}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 9}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 9}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 9}, + /* irq_idx: 312-315 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 9}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 9}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 9}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 9}, + /* irq_idx: 315-319 */ + { DPU_IRQ_TYPE_RESERVED, 0, 0, 9}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 9}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 9}, + { DPU_IRQ_TYPE_RESERVED, 0, 0, 9}, +}; + +static int dpu_hw_intr_irqidx_lookup(enum dpu_intr_type intr_type, + u32 instance_idx) +{ + int i; + + for (i = 0; i < ARRAY_SIZE(dpu_irq_map); i++) { + if (intr_type == dpu_irq_map[i].intr_type && + instance_idx == dpu_irq_map[i].instance_idx) + return i; + } + + pr_debug("IRQ lookup fail!! intr_type=%d, instance_idx=%d\n", + intr_type, instance_idx); + return -EINVAL; +} + +static void dpu_hw_intr_set_mask(struct dpu_hw_intr *intr, uint32_t reg_off, + uint32_t mask) +{ + if (!intr) + return; + + DPU_REG_WRITE(&intr->hw, reg_off, mask); + + /* ensure register writes go through */ + wmb(); +} + +static void dpu_hw_intr_dispatch_irq(struct dpu_hw_intr *intr, + void (*cbfunc)(void *, int), + void *arg) +{ + int reg_idx; + int irq_idx; + int start_idx; + int end_idx; + u32 irq_status; + unsigned long irq_flags; + + if (!intr) + return; + + /* + * The dispatcher will save the IRQ status before calling here. + * Now need to go through each IRQ status and find matching + * irq lookup index. + */ + spin_lock_irqsave(&intr->irq_lock, irq_flags); + for (reg_idx = 0; reg_idx < ARRAY_SIZE(dpu_intr_set); reg_idx++) { + irq_status = intr->save_irq_status[reg_idx]; + + /* + * Each Interrupt register has a range of 32 indexes, and + * that is static for dpu_irq_map. + */ + start_idx = reg_idx * 32; + end_idx = start_idx + 32; + + if (start_idx >= ARRAY_SIZE(dpu_irq_map) || + end_idx > ARRAY_SIZE(dpu_irq_map)) + continue; + + /* + * Search through matching intr status from irq map. + * start_idx and end_idx defined the search range in + * the dpu_irq_map. + */ + for (irq_idx = start_idx; + (irq_idx < end_idx) && irq_status; + irq_idx++) + if ((irq_status & dpu_irq_map[irq_idx].irq_mask) && + (dpu_irq_map[irq_idx].reg_idx == reg_idx)) { + /* + * Once a match on irq mask, perform a callback + * to the given cbfunc. cbfunc will take care + * the interrupt status clearing. If cbfunc is + * not provided, then the interrupt clearing + * is here. + */ + if (cbfunc) + cbfunc(arg, irq_idx); + else + intr->ops.clear_intr_status_nolock( + intr, irq_idx); + + /* + * When callback finish, clear the irq_status + * with the matching mask. Once irq_status + * is all cleared, the search can be stopped. + */ + irq_status &= ~dpu_irq_map[irq_idx].irq_mask; + } + } + spin_unlock_irqrestore(&intr->irq_lock, irq_flags); +} + +static int dpu_hw_intr_enable_irq(struct dpu_hw_intr *intr, int irq_idx) +{ + int reg_idx; + unsigned long irq_flags; + const struct dpu_intr_reg *reg; + const struct dpu_irq_type *irq; + const char *dbgstr = NULL; + uint32_t cache_irq_mask; + + if (!intr) + return -EINVAL; + + if (irq_idx < 0 || irq_idx >= ARRAY_SIZE(dpu_irq_map)) { + pr_err("invalid IRQ index: [%d]\n", irq_idx); + return -EINVAL; + } + + irq = &dpu_irq_map[irq_idx]; + reg_idx = irq->reg_idx; + reg = &dpu_intr_set[reg_idx]; + + spin_lock_irqsave(&intr->irq_lock, irq_flags); + cache_irq_mask = intr->cache_irq_mask[reg_idx]; + if (cache_irq_mask & irq->irq_mask) { + dbgstr = "DPU IRQ already set:"; + } else { + dbgstr = "DPU IRQ enabled:"; + + cache_irq_mask |= irq->irq_mask; + /* Cleaning any pending interrupt */ + DPU_REG_WRITE(&intr->hw, reg->clr_off, irq->irq_mask); + /* Enabling interrupts with the new mask */ + DPU_REG_WRITE(&intr->hw, reg->en_off, cache_irq_mask); + + /* ensure register write goes through */ + wmb(); + + intr->cache_irq_mask[reg_idx] = cache_irq_mask; + } + spin_unlock_irqrestore(&intr->irq_lock, irq_flags); + + pr_debug("%s MASK:0x%.8x, CACHE-MASK:0x%.8x\n", dbgstr, + irq->irq_mask, cache_irq_mask); + + return 0; +} + +static int dpu_hw_intr_disable_irq_nolock(struct dpu_hw_intr *intr, int irq_idx) +{ + int reg_idx; + const struct dpu_intr_reg *reg; + const struct dpu_irq_type *irq; + const char *dbgstr = NULL; + uint32_t cache_irq_mask; + + if (!intr) + return -EINVAL; + + if (irq_idx < 0 || irq_idx >= ARRAY_SIZE(dpu_irq_map)) { + pr_err("invalid IRQ index: [%d]\n", irq_idx); + return -EINVAL; + } + + irq = &dpu_irq_map[irq_idx]; + reg_idx = irq->reg_idx; + reg = &dpu_intr_set[reg_idx]; + + cache_irq_mask = intr->cache_irq_mask[reg_idx]; + if ((cache_irq_mask & irq->irq_mask) == 0) { + dbgstr = "DPU IRQ is already cleared:"; + } else { + dbgstr = "DPU IRQ mask disable:"; + + cache_irq_mask &= ~irq->irq_mask; + /* Disable interrupts based on the new mask */ + DPU_REG_WRITE(&intr->hw, reg->en_off, cache_irq_mask); + /* Cleaning any pending interrupt */ + DPU_REG_WRITE(&intr->hw, reg->clr_off, irq->irq_mask); + + /* ensure register write goes through */ + wmb(); + + intr->cache_irq_mask[reg_idx] = cache_irq_mask; + } + + pr_debug("%s MASK:0x%.8x, CACHE-MASK:0x%.8x\n", dbgstr, + irq->irq_mask, cache_irq_mask); + + return 0; +} + +static int dpu_hw_intr_disable_irq(struct dpu_hw_intr *intr, int irq_idx) +{ + unsigned long irq_flags; + + if (!intr) + return -EINVAL; + + if (irq_idx < 0 || irq_idx >= ARRAY_SIZE(dpu_irq_map)) { + pr_err("invalid IRQ index: [%d]\n", irq_idx); + return -EINVAL; + } + + spin_lock_irqsave(&intr->irq_lock, irq_flags); + dpu_hw_intr_disable_irq_nolock(intr, irq_idx); + spin_unlock_irqrestore(&intr->irq_lock, irq_flags); + + return 0; +} + +static int dpu_hw_intr_clear_irqs(struct dpu_hw_intr *intr) +{ + int i; + + if (!intr) + return -EINVAL; + + for (i = 0; i < ARRAY_SIZE(dpu_intr_set); i++) + DPU_REG_WRITE(&intr->hw, dpu_intr_set[i].clr_off, 0xffffffff); + + /* ensure register writes go through */ + wmb(); + + return 0; +} + +static int dpu_hw_intr_disable_irqs(struct dpu_hw_intr *intr) +{ + int i; + + if (!intr) + return -EINVAL; + + for (i = 0; i < ARRAY_SIZE(dpu_intr_set); i++) + DPU_REG_WRITE(&intr->hw, dpu_intr_set[i].en_off, 0x00000000); + + /* ensure register writes go through */ + wmb(); + + return 0; +} + +static int dpu_hw_intr_get_valid_interrupts(struct dpu_hw_intr *intr, + uint32_t *mask) +{ + if (!intr || !mask) + return -EINVAL; + + *mask = IRQ_SOURCE_MDP | IRQ_SOURCE_DSI0 | IRQ_SOURCE_DSI1 + | IRQ_SOURCE_HDMI | IRQ_SOURCE_EDP; + + return 0; +} + +static void dpu_hw_intr_get_interrupt_statuses(struct dpu_hw_intr *intr) +{ + int i; + u32 enable_mask; + unsigned long irq_flags; + + if (!intr) + return; + + spin_lock_irqsave(&intr->irq_lock, irq_flags); + for (i = 0; i < ARRAY_SIZE(dpu_intr_set); i++) { + /* Read interrupt status */ + intr->save_irq_status[i] = DPU_REG_READ(&intr->hw, + dpu_intr_set[i].status_off); + + /* Read enable mask */ + enable_mask = DPU_REG_READ(&intr->hw, dpu_intr_set[i].en_off); + + /* and clear the interrupt */ + if (intr->save_irq_status[i]) + DPU_REG_WRITE(&intr->hw, dpu_intr_set[i].clr_off, + intr->save_irq_status[i]); + + /* Finally update IRQ status based on enable mask */ + intr->save_irq_status[i] &= enable_mask; + } + + /* ensure register writes go through */ + wmb(); + + spin_unlock_irqrestore(&intr->irq_lock, irq_flags); +} + +static void dpu_hw_intr_clear_intr_status_nolock(struct dpu_hw_intr *intr, + int irq_idx) +{ + int reg_idx; + + if (!intr) + return; + + reg_idx = dpu_irq_map[irq_idx].reg_idx; + DPU_REG_WRITE(&intr->hw, dpu_intr_set[reg_idx].clr_off, + dpu_irq_map[irq_idx].irq_mask); + + /* ensure register writes go through */ + wmb(); +} + +static void dpu_hw_intr_clear_interrupt_status(struct dpu_hw_intr *intr, + int irq_idx) +{ + unsigned long irq_flags; + + if (!intr) + return; + + spin_lock_irqsave(&intr->irq_lock, irq_flags); + dpu_hw_intr_clear_intr_status_nolock(intr, irq_idx); + spin_unlock_irqrestore(&intr->irq_lock, irq_flags); +} + +static u32 dpu_hw_intr_get_interrupt_status(struct dpu_hw_intr *intr, + int irq_idx, bool clear) +{ + int reg_idx; + unsigned long irq_flags; + u32 intr_status; + + if (!intr) + return 0; + + if (irq_idx >= ARRAY_SIZE(dpu_irq_map) || irq_idx < 0) { + pr_err("invalid IRQ index: [%d]\n", irq_idx); + return 0; + } + + spin_lock_irqsave(&intr->irq_lock, irq_flags); + + reg_idx = dpu_irq_map[irq_idx].reg_idx; + intr_status = DPU_REG_READ(&intr->hw, + dpu_intr_set[reg_idx].status_off) & + dpu_irq_map[irq_idx].irq_mask; + if (intr_status && clear) + DPU_REG_WRITE(&intr->hw, dpu_intr_set[reg_idx].clr_off, + intr_status); + + /* ensure register writes go through */ + wmb(); + + spin_unlock_irqrestore(&intr->irq_lock, irq_flags); + + return intr_status; +} + +static void __setup_intr_ops(struct dpu_hw_intr_ops *ops) +{ + ops->set_mask = dpu_hw_intr_set_mask; + ops->irq_idx_lookup = dpu_hw_intr_irqidx_lookup; + ops->enable_irq = dpu_hw_intr_enable_irq; + ops->disable_irq = dpu_hw_intr_disable_irq; + ops->dispatch_irqs = dpu_hw_intr_dispatch_irq; + ops->clear_all_irqs = dpu_hw_intr_clear_irqs; + ops->disable_all_irqs = dpu_hw_intr_disable_irqs; + ops->get_valid_interrupts = dpu_hw_intr_get_valid_interrupts; + ops->get_interrupt_statuses = dpu_hw_intr_get_interrupt_statuses; + ops->clear_interrupt_status = dpu_hw_intr_clear_interrupt_status; + ops->clear_intr_status_nolock = dpu_hw_intr_clear_intr_status_nolock; + ops->get_interrupt_status = dpu_hw_intr_get_interrupt_status; +} + +static void __intr_offset(struct dpu_mdss_cfg *m, + void __iomem *addr, struct dpu_hw_blk_reg_map *hw) +{ + hw->base_off = addr; + hw->blk_off = m->mdp[0].base; + hw->hwversion = m->hwversion; +} + +struct dpu_hw_intr *dpu_hw_intr_init(void __iomem *addr, + struct dpu_mdss_cfg *m) +{ + struct dpu_hw_intr *intr; + + if (!addr || !m) + return ERR_PTR(-EINVAL); + + intr = kzalloc(sizeof(*intr), GFP_KERNEL); + if (!intr) + return ERR_PTR(-ENOMEM); + + __intr_offset(m, addr, &intr->hw); + __setup_intr_ops(&intr->ops); + + intr->irq_idx_tbl_size = ARRAY_SIZE(dpu_irq_map); + + intr->cache_irq_mask = kcalloc(ARRAY_SIZE(dpu_intr_set), sizeof(u32), + GFP_KERNEL); + if (intr->cache_irq_mask == NULL) { + kfree(intr); + return ERR_PTR(-ENOMEM); + } + + intr->save_irq_status = kcalloc(ARRAY_SIZE(dpu_intr_set), sizeof(u32), + GFP_KERNEL); + if (intr->save_irq_status == NULL) { + kfree(intr->cache_irq_mask); + kfree(intr); + return ERR_PTR(-ENOMEM); + } + + spin_lock_init(&intr->irq_lock); + + return intr; +} + +void dpu_hw_intr_destroy(struct dpu_hw_intr *intr) +{ + if (intr) { + kfree(intr->cache_irq_mask); + kfree(intr->save_irq_status); + kfree(intr); + } +} + diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_interrupts.h b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_interrupts.h new file mode 100644 index 000000000000..61e4cba36562 --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_interrupts.h @@ -0,0 +1,257 @@ +/* Copyright (c) 2016-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#ifndef _DPU_HW_INTERRUPTS_H +#define _DPU_HW_INTERRUPTS_H + +#include <linux/types.h> + +#include "dpu_hwio.h" +#include "dpu_hw_catalog.h" +#include "dpu_hw_util.h" +#include "dpu_hw_mdss.h" + +#define IRQ_SOURCE_MDP BIT(0) +#define IRQ_SOURCE_DSI0 BIT(4) +#define IRQ_SOURCE_DSI1 BIT(5) +#define IRQ_SOURCE_HDMI BIT(8) +#define IRQ_SOURCE_EDP BIT(12) +#define IRQ_SOURCE_MHL BIT(16) + +/** + * dpu_intr_type - HW Interrupt Type + * @DPU_IRQ_TYPE_WB_ROT_COMP: WB rotator done + * @DPU_IRQ_TYPE_WB_WFD_COMP: WB WFD done + * @DPU_IRQ_TYPE_PING_PONG_COMP: PingPong done + * @DPU_IRQ_TYPE_PING_PONG_RD_PTR: PingPong read pointer + * @DPU_IRQ_TYPE_PING_PONG_WR_PTR: PingPong write pointer + * @DPU_IRQ_TYPE_PING_PONG_AUTO_REF: PingPong auto refresh + * @DPU_IRQ_TYPE_PING_PONG_TEAR_CHECK: PingPong Tear check + * @DPU_IRQ_TYPE_PING_PONG_TE_CHECK: PingPong TE detection + * @DPU_IRQ_TYPE_INTF_UNDER_RUN: INTF underrun + * @DPU_IRQ_TYPE_INTF_VSYNC: INTF VSYNC + * @DPU_IRQ_TYPE_CWB_OVERFLOW: Concurrent WB overflow + * @DPU_IRQ_TYPE_HIST_VIG_DONE: VIG Histogram done + * @DPU_IRQ_TYPE_HIST_VIG_RSTSEQ: VIG Histogram reset + * @DPU_IRQ_TYPE_HIST_DSPP_DONE: DSPP Histogram done + * @DPU_IRQ_TYPE_HIST_DSPP_RSTSEQ: DSPP Histogram reset + * @DPU_IRQ_TYPE_WD_TIMER: Watchdog timer + * @DPU_IRQ_TYPE_SFI_VIDEO_IN: Video static frame INTR into static + * @DPU_IRQ_TYPE_SFI_VIDEO_OUT: Video static frame INTR out-of static + * @DPU_IRQ_TYPE_SFI_CMD_0_IN: DSI CMD0 static frame INTR into static + * @DPU_IRQ_TYPE_SFI_CMD_0_OUT: DSI CMD0 static frame INTR out-of static + * @DPU_IRQ_TYPE_SFI_CMD_1_IN: DSI CMD1 static frame INTR into static + * @DPU_IRQ_TYPE_SFI_CMD_1_OUT: DSI CMD1 static frame INTR out-of static + * @DPU_IRQ_TYPE_SFI_CMD_2_IN: DSI CMD2 static frame INTR into static + * @DPU_IRQ_TYPE_SFI_CMD_2_OUT: DSI CMD2 static frame INTR out-of static + * @DPU_IRQ_TYPE_PROG_LINE: Programmable Line interrupt + * @DPU_IRQ_TYPE_AD4_BL_DONE: AD4 backlight + * @DPU_IRQ_TYPE_CTL_START: Control start + * @DPU_IRQ_TYPE_RESERVED: Reserved for expansion + */ +enum dpu_intr_type { + DPU_IRQ_TYPE_WB_ROT_COMP, + DPU_IRQ_TYPE_WB_WFD_COMP, + DPU_IRQ_TYPE_PING_PONG_COMP, + DPU_IRQ_TYPE_PING_PONG_RD_PTR, + DPU_IRQ_TYPE_PING_PONG_WR_PTR, + DPU_IRQ_TYPE_PING_PONG_AUTO_REF, + DPU_IRQ_TYPE_PING_PONG_TEAR_CHECK, + DPU_IRQ_TYPE_PING_PONG_TE_CHECK, + DPU_IRQ_TYPE_INTF_UNDER_RUN, + DPU_IRQ_TYPE_INTF_VSYNC, + DPU_IRQ_TYPE_CWB_OVERFLOW, + DPU_IRQ_TYPE_HIST_VIG_DONE, + DPU_IRQ_TYPE_HIST_VIG_RSTSEQ, + DPU_IRQ_TYPE_HIST_DSPP_DONE, + DPU_IRQ_TYPE_HIST_DSPP_RSTSEQ, + DPU_IRQ_TYPE_WD_TIMER, + DPU_IRQ_TYPE_SFI_VIDEO_IN, + DPU_IRQ_TYPE_SFI_VIDEO_OUT, + DPU_IRQ_TYPE_SFI_CMD_0_IN, + DPU_IRQ_TYPE_SFI_CMD_0_OUT, + DPU_IRQ_TYPE_SFI_CMD_1_IN, + DPU_IRQ_TYPE_SFI_CMD_1_OUT, + DPU_IRQ_TYPE_SFI_CMD_2_IN, + DPU_IRQ_TYPE_SFI_CMD_2_OUT, + DPU_IRQ_TYPE_PROG_LINE, + DPU_IRQ_TYPE_AD4_BL_DONE, + DPU_IRQ_TYPE_CTL_START, + DPU_IRQ_TYPE_RESERVED, +}; + +struct dpu_hw_intr; + +/** + * Interrupt operations. + */ +struct dpu_hw_intr_ops { + /** + * set_mask - Programs the given interrupt register with the + * given interrupt mask. Register value will get overwritten. + * @intr: HW interrupt handle + * @reg_off: MDSS HW register offset + * @irqmask: IRQ mask value + */ + void (*set_mask)( + struct dpu_hw_intr *intr, + uint32_t reg, + uint32_t irqmask); + + /** + * irq_idx_lookup - Lookup IRQ index on the HW interrupt type + * Used for all irq related ops + * @intr_type: Interrupt type defined in dpu_intr_type + * @instance_idx: HW interrupt block instance + * @return: irq_idx or -EINVAL for lookup fail + */ + int (*irq_idx_lookup)( + enum dpu_intr_type intr_type, + u32 instance_idx); + + /** + * enable_irq - Enable IRQ based on lookup IRQ index + * @intr: HW interrupt handle + * @irq_idx: Lookup irq index return from irq_idx_lookup + * @return: 0 for success, otherwise failure + */ + int (*enable_irq)( + struct dpu_hw_intr *intr, + int irq_idx); + + /** + * disable_irq - Disable IRQ based on lookup IRQ index + * @intr: HW interrupt handle + * @irq_idx: Lookup irq index return from irq_idx_lookup + * @return: 0 for success, otherwise failure + */ + int (*disable_irq)( + struct dpu_hw_intr *intr, + int irq_idx); + + /** + * clear_all_irqs - Clears all the interrupts (i.e. acknowledges + * any asserted IRQs). Useful during reset. + * @intr: HW interrupt handle + * @return: 0 for success, otherwise failure + */ + int (*clear_all_irqs)( + struct dpu_hw_intr *intr); + + /** + * disable_all_irqs - Disables all the interrupts. Useful during reset. + * @intr: HW interrupt handle + * @return: 0 for success, otherwise failure + */ + int (*disable_all_irqs)( + struct dpu_hw_intr *intr); + + /** + * dispatch_irqs - IRQ dispatcher will call the given callback + * function when a matching interrupt status bit is + * found in the irq mapping table. + * @intr: HW interrupt handle + * @cbfunc: Callback function pointer + * @arg: Argument to pass back during callback + */ + void (*dispatch_irqs)( + struct dpu_hw_intr *intr, + void (*cbfunc)(void *arg, int irq_idx), + void *arg); + + /** + * get_interrupt_statuses - Gets and store value from all interrupt + * status registers that are currently fired. + * @intr: HW interrupt handle + */ + void (*get_interrupt_statuses)( + struct dpu_hw_intr *intr); + + /** + * clear_interrupt_status - Clears HW interrupt status based on given + * lookup IRQ index. + * @intr: HW interrupt handle + * @irq_idx: Lookup irq index return from irq_idx_lookup + */ + void (*clear_interrupt_status)( + struct dpu_hw_intr *intr, + int irq_idx); + + /** + * clear_intr_status_nolock() - clears the HW interrupts without lock + * @intr: HW interrupt handle + * @irq_idx: Lookup irq index return from irq_idx_lookup + */ + void (*clear_intr_status_nolock)( + struct dpu_hw_intr *intr, + int irq_idx); + + /** + * get_interrupt_status - Gets HW interrupt status, and clear if set, + * based on given lookup IRQ index. + * @intr: HW interrupt handle + * @irq_idx: Lookup irq index return from irq_idx_lookup + * @clear: True to clear irq after read + */ + u32 (*get_interrupt_status)( + struct dpu_hw_intr *intr, + int irq_idx, + bool clear); + + /** + * get_valid_interrupts - Gets a mask of all valid interrupt sources + * within DPU. These are actually status bits + * within interrupt registers that specify the + * source of the interrupt in IRQs. For example, + * valid interrupt sources can be MDP, DSI, + * HDMI etc. + * @intr: HW interrupt handle + * @mask: Returning the interrupt source MASK + * @return: 0 for success, otherwise failure + */ + int (*get_valid_interrupts)( + struct dpu_hw_intr *intr, + uint32_t *mask); +}; + +/** + * struct dpu_hw_intr: hw interrupts handling data structure + * @hw: virtual address mapping + * @ops: function pointer mapping for IRQ handling + * @cache_irq_mask: array of IRQ enable masks reg storage created during init + * @save_irq_status: array of IRQ status reg storage created during init + * @irq_idx_tbl_size: total number of irq_idx mapped in the hw_interrupts + * @irq_lock: spinlock for accessing IRQ resources + */ +struct dpu_hw_intr { + struct dpu_hw_blk_reg_map hw; + struct dpu_hw_intr_ops ops; + u32 *cache_irq_mask; + u32 *save_irq_status; + u32 irq_idx_tbl_size; + spinlock_t irq_lock; +}; + +/** + * dpu_hw_intr_init(): Initializes the interrupts hw object + * @addr: mapped register io address of MDP + * @m : pointer to mdss catalog data + */ +struct dpu_hw_intr *dpu_hw_intr_init(void __iomem *addr, + struct dpu_mdss_cfg *m); + +/** + * dpu_hw_intr_destroy(): Cleanup interrutps hw object + * @intr: pointer to interrupts hw object + */ +void dpu_hw_intr_destroy(struct dpu_hw_intr *intr); +#endif diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_intf.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_intf.c new file mode 100644 index 000000000000..d280df5613c9 --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_intf.c @@ -0,0 +1,349 @@ +/* Copyright (c) 2015-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#include "dpu_hwio.h" +#include "dpu_hw_catalog.h" +#include "dpu_hw_intf.h" +#include "dpu_dbg.h" +#include "dpu_kms.h" + +#define INTF_TIMING_ENGINE_EN 0x000 +#define INTF_CONFIG 0x004 +#define INTF_HSYNC_CTL 0x008 +#define INTF_VSYNC_PERIOD_F0 0x00C +#define INTF_VSYNC_PERIOD_F1 0x010 +#define INTF_VSYNC_PULSE_WIDTH_F0 0x014 +#define INTF_VSYNC_PULSE_WIDTH_F1 0x018 +#define INTF_DISPLAY_V_START_F0 0x01C +#define INTF_DISPLAY_V_START_F1 0x020 +#define INTF_DISPLAY_V_END_F0 0x024 +#define INTF_DISPLAY_V_END_F1 0x028 +#define INTF_ACTIVE_V_START_F0 0x02C +#define INTF_ACTIVE_V_START_F1 0x030 +#define INTF_ACTIVE_V_END_F0 0x034 +#define INTF_ACTIVE_V_END_F1 0x038 +#define INTF_DISPLAY_HCTL 0x03C +#define INTF_ACTIVE_HCTL 0x040 +#define INTF_BORDER_COLOR 0x044 +#define INTF_UNDERFLOW_COLOR 0x048 +#define INTF_HSYNC_SKEW 0x04C +#define INTF_POLARITY_CTL 0x050 +#define INTF_TEST_CTL 0x054 +#define INTF_TP_COLOR0 0x058 +#define INTF_TP_COLOR1 0x05C +#define INTF_FRAME_LINE_COUNT_EN 0x0A8 +#define INTF_FRAME_COUNT 0x0AC +#define INTF_LINE_COUNT 0x0B0 + +#define INTF_DEFLICKER_CONFIG 0x0F0 +#define INTF_DEFLICKER_STRNG_COEFF 0x0F4 +#define INTF_DEFLICKER_WEAK_COEFF 0x0F8 + +#define INTF_DSI_CMD_MODE_TRIGGER_EN 0x084 +#define INTF_PANEL_FORMAT 0x090 +#define INTF_TPG_ENABLE 0x100 +#define INTF_TPG_MAIN_CONTROL 0x104 +#define INTF_TPG_VIDEO_CONFIG 0x108 +#define INTF_TPG_COMPONENT_LIMITS 0x10C +#define INTF_TPG_RECTANGLE 0x110 +#define INTF_TPG_INITIAL_VALUE 0x114 +#define INTF_TPG_BLK_WHITE_PATTERN_FRAMES 0x118 +#define INTF_TPG_RGB_MAPPING 0x11C +#define INTF_PROG_FETCH_START 0x170 +#define INTF_PROG_ROT_START 0x174 + +#define INTF_FRAME_LINE_COUNT_EN 0x0A8 +#define INTF_FRAME_COUNT 0x0AC +#define INTF_LINE_COUNT 0x0B0 + +#define INTF_MISR_CTRL 0x180 +#define INTF_MISR_SIGNATURE 0x184 + +static struct dpu_intf_cfg *_intf_offset(enum dpu_intf intf, + struct dpu_mdss_cfg *m, + void __iomem *addr, + struct dpu_hw_blk_reg_map *b) +{ + int i; + + for (i = 0; i < m->intf_count; i++) { + if ((intf == m->intf[i].id) && + (m->intf[i].type != INTF_NONE)) { + b->base_off = addr; + b->blk_off = m->intf[i].base; + b->length = m->intf[i].len; + b->hwversion = m->hwversion; + b->log_mask = DPU_DBG_MASK_INTF; + return &m->intf[i]; + } + } + + return ERR_PTR(-EINVAL); +} + +static void dpu_hw_intf_setup_timing_engine(struct dpu_hw_intf *ctx, + const struct intf_timing_params *p, + const struct dpu_format *fmt) +{ + struct dpu_hw_blk_reg_map *c = &ctx->hw; + u32 hsync_period, vsync_period; + u32 display_v_start, display_v_end; + u32 hsync_start_x, hsync_end_x; + u32 active_h_start, active_h_end; + u32 active_v_start, active_v_end; + u32 active_hctl, display_hctl, hsync_ctl; + u32 polarity_ctl, den_polarity, hsync_polarity, vsync_polarity; + u32 panel_format; + u32 intf_cfg; + + /* read interface_cfg */ + intf_cfg = DPU_REG_READ(c, INTF_CONFIG); + hsync_period = p->hsync_pulse_width + p->h_back_porch + p->width + + p->h_front_porch; + vsync_period = p->vsync_pulse_width + p->v_back_porch + p->height + + p->v_front_porch; + + display_v_start = ((p->vsync_pulse_width + p->v_back_porch) * + hsync_period) + p->hsync_skew; + display_v_end = ((vsync_period - p->v_front_porch) * hsync_period) + + p->hsync_skew - 1; + + if (ctx->cap->type == INTF_EDP || ctx->cap->type == INTF_DP) { + display_v_start += p->hsync_pulse_width + p->h_back_porch; + display_v_end -= p->h_front_porch; + } + + hsync_start_x = p->h_back_porch + p->hsync_pulse_width; + hsync_end_x = hsync_period - p->h_front_porch - 1; + + if (p->width != p->xres) { + active_h_start = hsync_start_x; + active_h_end = active_h_start + p->xres - 1; + } else { + active_h_start = 0; + active_h_end = 0; + } + + if (p->height != p->yres) { + active_v_start = display_v_start; + active_v_end = active_v_start + (p->yres * hsync_period) - 1; + } else { + active_v_start = 0; + active_v_end = 0; + } + + if (active_h_end) { + active_hctl = (active_h_end << 16) | active_h_start; + intf_cfg |= BIT(29); /* ACTIVE_H_ENABLE */ + } else { + active_hctl = 0; + } + + if (active_v_end) + intf_cfg |= BIT(30); /* ACTIVE_V_ENABLE */ + + hsync_ctl = (hsync_period << 16) | p->hsync_pulse_width; + display_hctl = (hsync_end_x << 16) | hsync_start_x; + + den_polarity = 0; + if (ctx->cap->type == INTF_HDMI) { + hsync_polarity = p->yres >= 720 ? 0 : 1; + vsync_polarity = p->yres >= 720 ? 0 : 1; + } else { + hsync_polarity = 0; + vsync_polarity = 0; + } + polarity_ctl = (den_polarity << 2) | /* DEN Polarity */ + (vsync_polarity << 1) | /* VSYNC Polarity */ + (hsync_polarity << 0); /* HSYNC Polarity */ + + if (!DPU_FORMAT_IS_YUV(fmt)) + panel_format = (fmt->bits[C0_G_Y] | + (fmt->bits[C1_B_Cb] << 2) | + (fmt->bits[C2_R_Cr] << 4) | + (0x21 << 8)); + else + /* Interface treats all the pixel data in RGB888 format */ + panel_format = (COLOR_8BIT | + (COLOR_8BIT << 2) | + (COLOR_8BIT << 4) | + (0x21 << 8)); + + DPU_REG_WRITE(c, INTF_HSYNC_CTL, hsync_ctl); + DPU_REG_WRITE(c, INTF_VSYNC_PERIOD_F0, vsync_period * hsync_period); + DPU_REG_WRITE(c, INTF_VSYNC_PULSE_WIDTH_F0, + p->vsync_pulse_width * hsync_period); + DPU_REG_WRITE(c, INTF_DISPLAY_HCTL, display_hctl); + DPU_REG_WRITE(c, INTF_DISPLAY_V_START_F0, display_v_start); + DPU_REG_WRITE(c, INTF_DISPLAY_V_END_F0, display_v_end); + DPU_REG_WRITE(c, INTF_ACTIVE_HCTL, active_hctl); + DPU_REG_WRITE(c, INTF_ACTIVE_V_START_F0, active_v_start); + DPU_REG_WRITE(c, INTF_ACTIVE_V_END_F0, active_v_end); + DPU_REG_WRITE(c, INTF_BORDER_COLOR, p->border_clr); + DPU_REG_WRITE(c, INTF_UNDERFLOW_COLOR, p->underflow_clr); + DPU_REG_WRITE(c, INTF_HSYNC_SKEW, p->hsync_skew); + DPU_REG_WRITE(c, INTF_POLARITY_CTL, polarity_ctl); + DPU_REG_WRITE(c, INTF_FRAME_LINE_COUNT_EN, 0x3); + DPU_REG_WRITE(c, INTF_CONFIG, intf_cfg); + DPU_REG_WRITE(c, INTF_PANEL_FORMAT, panel_format); +} + +static void dpu_hw_intf_enable_timing_engine( + struct dpu_hw_intf *intf, + u8 enable) +{ + struct dpu_hw_blk_reg_map *c = &intf->hw; + /* Note: Display interface select is handled in top block hw layer */ + DPU_REG_WRITE(c, INTF_TIMING_ENGINE_EN, enable != 0); +} + +static void dpu_hw_intf_setup_prg_fetch( + struct dpu_hw_intf *intf, + const struct intf_prog_fetch *fetch) +{ + struct dpu_hw_blk_reg_map *c = &intf->hw; + int fetch_enable; + + /* + * Fetch should always be outside the active lines. If the fetching + * is programmed within active region, hardware behavior is unknown. + */ + + fetch_enable = DPU_REG_READ(c, INTF_CONFIG); + if (fetch->enable) { + fetch_enable |= BIT(31); + DPU_REG_WRITE(c, INTF_PROG_FETCH_START, + fetch->fetch_start); + } else { + fetch_enable &= ~BIT(31); + } + + DPU_REG_WRITE(c, INTF_CONFIG, fetch_enable); +} + +static void dpu_hw_intf_get_status( + struct dpu_hw_intf *intf, + struct intf_status *s) +{ + struct dpu_hw_blk_reg_map *c = &intf->hw; + + s->is_en = DPU_REG_READ(c, INTF_TIMING_ENGINE_EN); + if (s->is_en) { + s->frame_count = DPU_REG_READ(c, INTF_FRAME_COUNT); + s->line_count = DPU_REG_READ(c, INTF_LINE_COUNT); + } else { + s->line_count = 0; + s->frame_count = 0; + } +} + +static void dpu_hw_intf_setup_misr(struct dpu_hw_intf *intf, + bool enable, u32 frame_count) +{ + struct dpu_hw_blk_reg_map *c = &intf->hw; + u32 config = 0; + + DPU_REG_WRITE(c, INTF_MISR_CTRL, MISR_CTRL_STATUS_CLEAR); + /* clear misr data */ + wmb(); + + if (enable) + config = (frame_count & MISR_FRAME_COUNT_MASK) | + MISR_CTRL_ENABLE | INTF_MISR_CTRL_FREE_RUN_MASK; + + DPU_REG_WRITE(c, INTF_MISR_CTRL, config); +} + +static u32 dpu_hw_intf_collect_misr(struct dpu_hw_intf *intf) +{ + struct dpu_hw_blk_reg_map *c = &intf->hw; + + return DPU_REG_READ(c, INTF_MISR_SIGNATURE); +} + +static u32 dpu_hw_intf_get_line_count(struct dpu_hw_intf *intf) +{ + struct dpu_hw_blk_reg_map *c; + + if (!intf) + return 0; + + c = &intf->hw; + + return DPU_REG_READ(c, INTF_LINE_COUNT); +} + +static void _setup_intf_ops(struct dpu_hw_intf_ops *ops, + unsigned long cap) +{ + ops->setup_timing_gen = dpu_hw_intf_setup_timing_engine; + ops->setup_prg_fetch = dpu_hw_intf_setup_prg_fetch; + ops->get_status = dpu_hw_intf_get_status; + ops->enable_timing = dpu_hw_intf_enable_timing_engine; + ops->setup_misr = dpu_hw_intf_setup_misr; + ops->collect_misr = dpu_hw_intf_collect_misr; + ops->get_line_count = dpu_hw_intf_get_line_count; +} + +static struct dpu_hw_blk_ops dpu_hw_ops = { + .start = NULL, + .stop = NULL, +}; + +struct dpu_hw_intf *dpu_hw_intf_init(enum dpu_intf idx, + void __iomem *addr, + struct dpu_mdss_cfg *m) +{ + struct dpu_hw_intf *c; + struct dpu_intf_cfg *cfg; + int rc; + + c = kzalloc(sizeof(*c), GFP_KERNEL); + if (!c) + return ERR_PTR(-ENOMEM); + + cfg = _intf_offset(idx, m, addr, &c->hw); + if (IS_ERR_OR_NULL(cfg)) { + kfree(c); + pr_err("failed to create dpu_hw_intf %d\n", idx); + return ERR_PTR(-EINVAL); + } + + /* + * Assign ops + */ + c->idx = idx; + c->cap = cfg; + c->mdss = m; + _setup_intf_ops(&c->ops, c->cap->features); + + rc = dpu_hw_blk_init(&c->base, DPU_HW_BLK_INTF, idx, &dpu_hw_ops); + if (rc) { + DPU_ERROR("failed to init hw blk %d\n", rc); + goto blk_init_error; + } + + return c; + +blk_init_error: + kzfree(c); + + return ERR_PTR(rc); +} + +void dpu_hw_intf_destroy(struct dpu_hw_intf *intf) +{ + if (intf) + dpu_hw_blk_destroy(&intf->base); + kfree(intf); +} + diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_intf.h b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_intf.h new file mode 100644 index 000000000000..a79d735da68d --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_intf.h @@ -0,0 +1,128 @@ +/* Copyright (c) 2015-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#ifndef _DPU_HW_INTF_H +#define _DPU_HW_INTF_H + +#include "dpu_hw_catalog.h" +#include "dpu_hw_mdss.h" +#include "dpu_hw_util.h" +#include "dpu_hw_blk.h" + +struct dpu_hw_intf; + +/* intf timing settings */ +struct intf_timing_params { + u32 width; /* active width */ + u32 height; /* active height */ + u32 xres; /* Display panel width */ + u32 yres; /* Display panel height */ + + u32 h_back_porch; + u32 h_front_porch; + u32 v_back_porch; + u32 v_front_porch; + u32 hsync_pulse_width; + u32 vsync_pulse_width; + u32 hsync_polarity; + u32 vsync_polarity; + u32 border_clr; + u32 underflow_clr; + u32 hsync_skew; +}; + +struct intf_prog_fetch { + u8 enable; + /* vsync counter for the front porch pixel line */ + u32 fetch_start; +}; + +struct intf_status { + u8 is_en; /* interface timing engine is enabled or not */ + u32 frame_count; /* frame count since timing engine enabled */ + u32 line_count; /* current line count including blanking */ +}; + +/** + * struct dpu_hw_intf_ops : Interface to the interface Hw driver functions + * Assumption is these functions will be called after clocks are enabled + * @ setup_timing_gen : programs the timing engine + * @ setup_prog_fetch : enables/disables the programmable fetch logic + * @ enable_timing: enable/disable timing engine + * @ get_status: returns if timing engine is enabled or not + * @ setup_misr: enables/disables MISR in HW register + * @ collect_misr: reads and stores MISR data from HW register + * @ get_line_count: reads current vertical line counter + */ +struct dpu_hw_intf_ops { + void (*setup_timing_gen)(struct dpu_hw_intf *intf, + const struct intf_timing_params *p, + const struct dpu_format *fmt); + + void (*setup_prg_fetch)(struct dpu_hw_intf *intf, + const struct intf_prog_fetch *fetch); + + void (*enable_timing)(struct dpu_hw_intf *intf, + u8 enable); + + void (*get_status)(struct dpu_hw_intf *intf, + struct intf_status *status); + + void (*setup_misr)(struct dpu_hw_intf *intf, + bool enable, u32 frame_count); + + u32 (*collect_misr)(struct dpu_hw_intf *intf); + + u32 (*get_line_count)(struct dpu_hw_intf *intf); +}; + +struct dpu_hw_intf { + struct dpu_hw_blk base; + struct dpu_hw_blk_reg_map hw; + + /* intf */ + enum dpu_intf idx; + const struct dpu_intf_cfg *cap; + const struct dpu_mdss_cfg *mdss; + + /* ops */ + struct dpu_hw_intf_ops ops; +}; + +/** + * to_dpu_hw_intf - convert base object dpu_hw_base to container + * @hw: Pointer to base hardware block + * return: Pointer to hardware block container + */ +static inline struct dpu_hw_intf *to_dpu_hw_intf(struct dpu_hw_blk *hw) +{ + return container_of(hw, struct dpu_hw_intf, base); +} + +/** + * dpu_hw_intf_init(): Initializes the intf driver for the passed + * interface idx. + * @idx: interface index for which driver object is required + * @addr: mapped register io address of MDP + * @m : pointer to mdss catalog data + */ +struct dpu_hw_intf *dpu_hw_intf_init(enum dpu_intf idx, + void __iomem *addr, + struct dpu_mdss_cfg *m); + +/** + * dpu_hw_intf_destroy(): Destroys INTF driver context + * @intf: Pointer to INTF driver context + */ +void dpu_hw_intf_destroy(struct dpu_hw_intf *intf); + +#endif /*_DPU_HW_INTF_H */ diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_lm.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_lm.c new file mode 100644 index 000000000000..4ab72b0f07a5 --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_lm.c @@ -0,0 +1,261 @@ +/* Copyright (c) 2015-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#include "dpu_kms.h" +#include "dpu_hw_catalog.h" +#include "dpu_hwio.h" +#include "dpu_hw_lm.h" +#include "dpu_hw_mdss.h" +#include "dpu_dbg.h" +#include "dpu_kms.h" + +#define LM_OP_MODE 0x00 +#define LM_OUT_SIZE 0x04 +#define LM_BORDER_COLOR_0 0x08 +#define LM_BORDER_COLOR_1 0x010 + +/* These register are offset to mixer base + stage base */ +#define LM_BLEND0_OP 0x00 +#define LM_BLEND0_CONST_ALPHA 0x04 +#define LM_FG_COLOR_FILL_COLOR_0 0x08 +#define LM_FG_COLOR_FILL_COLOR_1 0x0C +#define LM_FG_COLOR_FILL_SIZE 0x10 +#define LM_FG_COLOR_FILL_XY 0x14 + +#define LM_BLEND0_FG_ALPHA 0x04 +#define LM_BLEND0_BG_ALPHA 0x08 + +#define LM_MISR_CTRL 0x310 +#define LM_MISR_SIGNATURE 0x314 + +static struct dpu_lm_cfg *_lm_offset(enum dpu_lm mixer, + struct dpu_mdss_cfg *m, + void __iomem *addr, + struct dpu_hw_blk_reg_map *b) +{ + int i; + + for (i = 0; i < m->mixer_count; i++) { + if (mixer == m->mixer[i].id) { + b->base_off = addr; + b->blk_off = m->mixer[i].base; + b->length = m->mixer[i].len; + b->hwversion = m->hwversion; + b->log_mask = DPU_DBG_MASK_LM; + return &m->mixer[i]; + } + } + + return ERR_PTR(-ENOMEM); +} + +/** + * _stage_offset(): returns the relative offset of the blend registers + * for the stage to be setup + * @c: mixer ctx contains the mixer to be programmed + * @stage: stage index to setup + */ +static inline int _stage_offset(struct dpu_hw_mixer *ctx, enum dpu_stage stage) +{ + const struct dpu_lm_sub_blks *sblk = ctx->cap->sblk; + int rc; + + if (stage == DPU_STAGE_BASE) + rc = -EINVAL; + else if (stage <= sblk->maxblendstages) + rc = sblk->blendstage_base[stage - DPU_STAGE_0]; + else + rc = -EINVAL; + + return rc; +} + +static void dpu_hw_lm_setup_out(struct dpu_hw_mixer *ctx, + struct dpu_hw_mixer_cfg *mixer) +{ + struct dpu_hw_blk_reg_map *c = &ctx->hw; + u32 outsize; + u32 op_mode; + + op_mode = DPU_REG_READ(c, LM_OP_MODE); + + outsize = mixer->out_height << 16 | mixer->out_width; + DPU_REG_WRITE(c, LM_OUT_SIZE, outsize); + + /* SPLIT_LEFT_RIGHT */ + if (mixer->right_mixer) + op_mode |= BIT(31); + else + op_mode &= ~BIT(31); + DPU_REG_WRITE(c, LM_OP_MODE, op_mode); +} + +static void dpu_hw_lm_setup_border_color(struct dpu_hw_mixer *ctx, + struct dpu_mdss_color *color, + u8 border_en) +{ + struct dpu_hw_blk_reg_map *c = &ctx->hw; + + if (border_en) { + DPU_REG_WRITE(c, LM_BORDER_COLOR_0, + (color->color_0 & 0xFFF) | + ((color->color_1 & 0xFFF) << 0x10)); + DPU_REG_WRITE(c, LM_BORDER_COLOR_1, + (color->color_2 & 0xFFF) | + ((color->color_3 & 0xFFF) << 0x10)); + } +} + +static void dpu_hw_lm_setup_blend_config_sdm845(struct dpu_hw_mixer *ctx, + u32 stage, u32 fg_alpha, u32 bg_alpha, u32 blend_op) +{ + struct dpu_hw_blk_reg_map *c = &ctx->hw; + int stage_off; + u32 const_alpha; + + if (stage == DPU_STAGE_BASE) + return; + + stage_off = _stage_offset(ctx, stage); + if (WARN_ON(stage_off < 0)) + return; + + const_alpha = (bg_alpha & 0xFF) | ((fg_alpha & 0xFF) << 16); + DPU_REG_WRITE(c, LM_BLEND0_CONST_ALPHA + stage_off, const_alpha); + DPU_REG_WRITE(c, LM_BLEND0_OP + stage_off, blend_op); +} + +static void dpu_hw_lm_setup_blend_config(struct dpu_hw_mixer *ctx, + u32 stage, u32 fg_alpha, u32 bg_alpha, u32 blend_op) +{ + struct dpu_hw_blk_reg_map *c = &ctx->hw; + int stage_off; + + if (stage == DPU_STAGE_BASE) + return; + + stage_off = _stage_offset(ctx, stage); + if (WARN_ON(stage_off < 0)) + return; + + DPU_REG_WRITE(c, LM_BLEND0_FG_ALPHA + stage_off, fg_alpha); + DPU_REG_WRITE(c, LM_BLEND0_BG_ALPHA + stage_off, bg_alpha); + DPU_REG_WRITE(c, LM_BLEND0_OP + stage_off, blend_op); +} + +static void dpu_hw_lm_setup_color3(struct dpu_hw_mixer *ctx, + uint32_t mixer_op_mode) +{ + struct dpu_hw_blk_reg_map *c = &ctx->hw; + int op_mode; + + /* read the existing op_mode configuration */ + op_mode = DPU_REG_READ(c, LM_OP_MODE); + + op_mode = (op_mode & (BIT(31) | BIT(30))) | mixer_op_mode; + + DPU_REG_WRITE(c, LM_OP_MODE, op_mode); +} + +static void dpu_hw_lm_gc(struct dpu_hw_mixer *mixer, + void *cfg) +{ +} + +static void dpu_hw_lm_setup_misr(struct dpu_hw_mixer *ctx, + bool enable, u32 frame_count) +{ + struct dpu_hw_blk_reg_map *c = &ctx->hw; + u32 config = 0; + + DPU_REG_WRITE(c, LM_MISR_CTRL, MISR_CTRL_STATUS_CLEAR); + /* clear misr data */ + wmb(); + + if (enable) + config = (frame_count & MISR_FRAME_COUNT_MASK) | + MISR_CTRL_ENABLE | INTF_MISR_CTRL_FREE_RUN_MASK; + + DPU_REG_WRITE(c, LM_MISR_CTRL, config); +} + +static u32 dpu_hw_lm_collect_misr(struct dpu_hw_mixer *ctx) +{ + struct dpu_hw_blk_reg_map *c = &ctx->hw; + + return DPU_REG_READ(c, LM_MISR_SIGNATURE); +} + +static void _setup_mixer_ops(struct dpu_mdss_cfg *m, + struct dpu_hw_lm_ops *ops, + unsigned long features) +{ + ops->setup_mixer_out = dpu_hw_lm_setup_out; + if (IS_SDM845_TARGET(m->hwversion) || IS_SDM670_TARGET(m->hwversion)) + ops->setup_blend_config = dpu_hw_lm_setup_blend_config_sdm845; + else + ops->setup_blend_config = dpu_hw_lm_setup_blend_config; + ops->setup_alpha_out = dpu_hw_lm_setup_color3; + ops->setup_border_color = dpu_hw_lm_setup_border_color; + ops->setup_gc = dpu_hw_lm_gc; + ops->setup_misr = dpu_hw_lm_setup_misr; + ops->collect_misr = dpu_hw_lm_collect_misr; +}; + +static struct dpu_hw_blk_ops dpu_hw_ops = { + .start = NULL, + .stop = NULL, +}; + +struct dpu_hw_mixer *dpu_hw_lm_init(enum dpu_lm idx, + void __iomem *addr, + struct dpu_mdss_cfg *m) +{ + struct dpu_hw_mixer *c; + struct dpu_lm_cfg *cfg; + int rc; + + c = kzalloc(sizeof(*c), GFP_KERNEL); + if (!c) + return ERR_PTR(-ENOMEM); + + cfg = _lm_offset(idx, m, addr, &c->hw); + if (IS_ERR_OR_NULL(cfg)) { + kfree(c); + return ERR_PTR(-EINVAL); + } + + /* Assign ops */ + c->idx = idx; + c->cap = cfg; + _setup_mixer_ops(m, &c->ops, c->cap->features); + + rc = dpu_hw_blk_init(&c->base, DPU_HW_BLK_LM, idx, &dpu_hw_ops); + if (rc) { + DPU_ERROR("failed to init hw blk %d\n", rc); + goto blk_init_error; + } + + return c; + +blk_init_error: + kzfree(c); + + return ERR_PTR(rc); +} + +void dpu_hw_lm_destroy(struct dpu_hw_mixer *lm) +{ + if (lm) + dpu_hw_blk_destroy(&lm->base); + kfree(lm); +} diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_lm.h b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_lm.h new file mode 100644 index 000000000000..e29e5dab31bf --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_lm.h @@ -0,0 +1,122 @@ +/* Copyright (c) 2015-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#ifndef _DPU_HW_LM_H +#define _DPU_HW_LM_H + +#include "dpu_hw_mdss.h" +#include "dpu_hw_util.h" +#include "dpu_hw_blk.h" + +struct dpu_hw_mixer; + +struct dpu_hw_mixer_cfg { + u32 out_width; + u32 out_height; + bool right_mixer; + int flags; +}; + +struct dpu_hw_color3_cfg { + u8 keep_fg[DPU_STAGE_MAX]; +}; + +/** + * + * struct dpu_hw_lm_ops : Interface to the mixer Hw driver functions + * Assumption is these functions will be called after clocks are enabled + */ +struct dpu_hw_lm_ops { + /* + * Sets up mixer output width and height + * and border color if enabled + */ + void (*setup_mixer_out)(struct dpu_hw_mixer *ctx, + struct dpu_hw_mixer_cfg *cfg); + + /* + * Alpha blending configuration + * for the specified stage + */ + void (*setup_blend_config)(struct dpu_hw_mixer *ctx, uint32_t stage, + uint32_t fg_alpha, uint32_t bg_alpha, uint32_t blend_op); + + /* + * Alpha color component selection from either fg or bg + */ + void (*setup_alpha_out)(struct dpu_hw_mixer *ctx, uint32_t mixer_op); + + /** + * setup_border_color : enable/disable border color + */ + void (*setup_border_color)(struct dpu_hw_mixer *ctx, + struct dpu_mdss_color *color, + u8 border_en); + /** + * setup_gc : enable/disable gamma correction feature + */ + void (*setup_gc)(struct dpu_hw_mixer *mixer, + void *cfg); + + /* setup_misr: enables/disables MISR in HW register */ + void (*setup_misr)(struct dpu_hw_mixer *ctx, + bool enable, u32 frame_count); + + /* collect_misr: reads and stores MISR data from HW register */ + u32 (*collect_misr)(struct dpu_hw_mixer *ctx); +}; + +struct dpu_hw_mixer { + struct dpu_hw_blk base; + struct dpu_hw_blk_reg_map hw; + + /* lm */ + enum dpu_lm idx; + const struct dpu_lm_cfg *cap; + const struct dpu_mdp_cfg *mdp; + const struct dpu_ctl_cfg *ctl; + + /* ops */ + struct dpu_hw_lm_ops ops; + + /* store mixer info specific to display */ + struct dpu_hw_mixer_cfg cfg; +}; + +/** + * to_dpu_hw_mixer - convert base object dpu_hw_base to container + * @hw: Pointer to base hardware block + * return: Pointer to hardware block container + */ +static inline struct dpu_hw_mixer *to_dpu_hw_mixer(struct dpu_hw_blk *hw) +{ + return container_of(hw, struct dpu_hw_mixer, base); +} + +/** + * dpu_hw_lm_init(): Initializes the mixer hw driver object. + * should be called once before accessing every mixer. + * @idx: mixer index for which driver object is required + * @addr: mapped register io address of MDP + * @m : pointer to mdss catalog data + */ +struct dpu_hw_mixer *dpu_hw_lm_init(enum dpu_lm idx, + void __iomem *addr, + struct dpu_mdss_cfg *m); + +/** + * dpu_hw_lm_destroy(): Destroys layer mixer driver context + * @lm: Pointer to LM driver context + */ +void dpu_hw_lm_destroy(struct dpu_hw_mixer *lm); + +#endif /*_DPU_HW_LM_H */ diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_mdss.h b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_mdss.h new file mode 100644 index 000000000000..35e6bf930924 --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_mdss.h @@ -0,0 +1,465 @@ +/* Copyright (c) 2015-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#ifndef _DPU_HW_MDSS_H +#define _DPU_HW_MDSS_H + +#include <linux/kernel.h> +#include <linux/err.h> + +#include "msm_drv.h" + +#define DPU_DBG_NAME "dpu" + +#define DPU_NONE 0 + +#ifndef DPU_CSC_MATRIX_COEFF_SIZE +#define DPU_CSC_MATRIX_COEFF_SIZE 9 +#endif + +#ifndef DPU_CSC_CLAMP_SIZE +#define DPU_CSC_CLAMP_SIZE 6 +#endif + +#ifndef DPU_CSC_BIAS_SIZE +#define DPU_CSC_BIAS_SIZE 3 +#endif + +#ifndef DPU_MAX_PLANES +#define DPU_MAX_PLANES 4 +#endif + +#define PIPES_PER_STAGE 2 +#ifndef DPU_MAX_DE_CURVES +#define DPU_MAX_DE_CURVES 3 +#endif + +enum dpu_format_flags { + DPU_FORMAT_FLAG_YUV_BIT, + DPU_FORMAT_FLAG_DX_BIT, + DPU_FORMAT_FLAG_COMPRESSED_BIT, + DPU_FORMAT_FLAG_BIT_MAX, +}; + +#define DPU_FORMAT_FLAG_YUV BIT(DPU_FORMAT_FLAG_YUV_BIT) +#define DPU_FORMAT_FLAG_DX BIT(DPU_FORMAT_FLAG_DX_BIT) +#define DPU_FORMAT_FLAG_COMPRESSED BIT(DPU_FORMAT_FLAG_COMPRESSED_BIT) +#define DPU_FORMAT_IS_YUV(X) \ + (test_bit(DPU_FORMAT_FLAG_YUV_BIT, (X)->flag)) +#define DPU_FORMAT_IS_DX(X) \ + (test_bit(DPU_FORMAT_FLAG_DX_BIT, (X)->flag)) +#define DPU_FORMAT_IS_LINEAR(X) ((X)->fetch_mode == DPU_FETCH_LINEAR) +#define DPU_FORMAT_IS_TILE(X) \ + (((X)->fetch_mode == DPU_FETCH_UBWC) && \ + !test_bit(DPU_FORMAT_FLAG_COMPRESSED_BIT, (X)->flag)) +#define DPU_FORMAT_IS_UBWC(X) \ + (((X)->fetch_mode == DPU_FETCH_UBWC) && \ + test_bit(DPU_FORMAT_FLAG_COMPRESSED_BIT, (X)->flag)) + +#define DPU_BLEND_FG_ALPHA_FG_CONST (0 << 0) +#define DPU_BLEND_FG_ALPHA_BG_CONST (1 << 0) +#define DPU_BLEND_FG_ALPHA_FG_PIXEL (2 << 0) +#define DPU_BLEND_FG_ALPHA_BG_PIXEL (3 << 0) +#define DPU_BLEND_FG_INV_ALPHA (1 << 2) +#define DPU_BLEND_FG_MOD_ALPHA (1 << 3) +#define DPU_BLEND_FG_INV_MOD_ALPHA (1 << 4) +#define DPU_BLEND_FG_TRANSP_EN (1 << 5) +#define DPU_BLEND_BG_ALPHA_FG_CONST (0 << 8) +#define DPU_BLEND_BG_ALPHA_BG_CONST (1 << 8) +#define DPU_BLEND_BG_ALPHA_FG_PIXEL (2 << 8) +#define DPU_BLEND_BG_ALPHA_BG_PIXEL (3 << 8) +#define DPU_BLEND_BG_INV_ALPHA (1 << 10) +#define DPU_BLEND_BG_MOD_ALPHA (1 << 11) +#define DPU_BLEND_BG_INV_MOD_ALPHA (1 << 12) +#define DPU_BLEND_BG_TRANSP_EN (1 << 13) + +#define DPU_VSYNC0_SOURCE_GPIO 0 +#define DPU_VSYNC1_SOURCE_GPIO 1 +#define DPU_VSYNC2_SOURCE_GPIO 2 +#define DPU_VSYNC_SOURCE_INTF_0 3 +#define DPU_VSYNC_SOURCE_INTF_1 4 +#define DPU_VSYNC_SOURCE_INTF_2 5 +#define DPU_VSYNC_SOURCE_INTF_3 6 +#define DPU_VSYNC_SOURCE_WD_TIMER_4 11 +#define DPU_VSYNC_SOURCE_WD_TIMER_3 12 +#define DPU_VSYNC_SOURCE_WD_TIMER_2 13 +#define DPU_VSYNC_SOURCE_WD_TIMER_1 14 +#define DPU_VSYNC_SOURCE_WD_TIMER_0 15 + +enum dpu_hw_blk_type { + DPU_HW_BLK_TOP = 0, + DPU_HW_BLK_SSPP, + DPU_HW_BLK_LM, + DPU_HW_BLK_CTL, + DPU_HW_BLK_CDM, + DPU_HW_BLK_PINGPONG, + DPU_HW_BLK_INTF, + DPU_HW_BLK_WB, + DPU_HW_BLK_MAX, +}; + +enum dpu_mdp { + MDP_TOP = 0x1, + MDP_MAX, +}; + +enum dpu_sspp { + SSPP_NONE, + SSPP_VIG0, + SSPP_VIG1, + SSPP_VIG2, + SSPP_VIG3, + SSPP_RGB0, + SSPP_RGB1, + SSPP_RGB2, + SSPP_RGB3, + SSPP_DMA0, + SSPP_DMA1, + SSPP_DMA2, + SSPP_DMA3, + SSPP_CURSOR0, + SSPP_CURSOR1, + SSPP_MAX +}; + +enum dpu_sspp_type { + SSPP_TYPE_VIG, + SSPP_TYPE_RGB, + SSPP_TYPE_DMA, + SSPP_TYPE_CURSOR, + SSPP_TYPE_MAX +}; + +enum dpu_lm { + LM_0 = 1, + LM_1, + LM_2, + LM_3, + LM_4, + LM_5, + LM_6, + LM_MAX +}; + +enum dpu_stage { + DPU_STAGE_BASE = 0, + DPU_STAGE_0, + DPU_STAGE_1, + DPU_STAGE_2, + DPU_STAGE_3, + DPU_STAGE_4, + DPU_STAGE_5, + DPU_STAGE_6, + DPU_STAGE_7, + DPU_STAGE_8, + DPU_STAGE_9, + DPU_STAGE_10, + DPU_STAGE_MAX +}; +enum dpu_dspp { + DSPP_0 = 1, + DSPP_1, + DSPP_2, + DSPP_3, + DSPP_MAX +}; + +enum dpu_ds { + DS_TOP, + DS_0, + DS_1, + DS_MAX +}; + +enum dpu_ctl { + CTL_0 = 1, + CTL_1, + CTL_2, + CTL_3, + CTL_4, + CTL_MAX +}; + +enum dpu_cdm { + CDM_0 = 1, + CDM_1, + CDM_MAX +}; + +enum dpu_pingpong { + PINGPONG_0 = 1, + PINGPONG_1, + PINGPONG_2, + PINGPONG_3, + PINGPONG_4, + PINGPONG_S0, + PINGPONG_MAX +}; + +enum dpu_intf { + INTF_0 = 1, + INTF_1, + INTF_2, + INTF_3, + INTF_4, + INTF_5, + INTF_6, + INTF_MAX +}; + +enum dpu_intf_type { + INTF_NONE = 0x0, + INTF_DSI = 0x1, + INTF_HDMI = 0x3, + INTF_LCDC = 0x5, + INTF_EDP = 0x9, + INTF_DP = 0xa, + INTF_TYPE_MAX, + + /* virtual interfaces */ + INTF_WB = 0x100, +}; + +enum dpu_intf_mode { + INTF_MODE_NONE = 0, + INTF_MODE_CMD, + INTF_MODE_VIDEO, + INTF_MODE_WB_BLOCK, + INTF_MODE_WB_LINE, + INTF_MODE_MAX +}; + +enum dpu_wb { + WB_0 = 1, + WB_1, + WB_2, + WB_3, + WB_MAX +}; + +enum dpu_ad { + AD_0 = 0x1, + AD_1, + AD_MAX +}; + +enum dpu_cwb { + CWB_0 = 0x1, + CWB_1, + CWB_2, + CWB_3, + CWB_MAX +}; + +enum dpu_wd_timer { + WD_TIMER_0 = 0x1, + WD_TIMER_1, + WD_TIMER_2, + WD_TIMER_3, + WD_TIMER_4, + WD_TIMER_5, + WD_TIMER_MAX +}; + +enum dpu_vbif { + VBIF_0, + VBIF_1, + VBIF_MAX, + VBIF_RT = VBIF_0, + VBIF_NRT = VBIF_1 +}; + +enum dpu_iommu_domain { + DPU_IOMMU_DOMAIN_UNSECURE, + DPU_IOMMU_DOMAIN_SECURE, + DPU_IOMMU_DOMAIN_MAX +}; + +/** + * DPU HW,Component order color map + */ +enum { + C0_G_Y = 0, + C1_B_Cb = 1, + C2_R_Cr = 2, + C3_ALPHA = 3 +}; + +/** + * enum dpu_plane_type - defines how the color component pixel packing + * @DPU_PLANE_INTERLEAVED : Color components in single plane + * @DPU_PLANE_PLANAR : Color component in separate planes + * @DPU_PLANE_PSEUDO_PLANAR : Chroma components interleaved in separate plane + */ +enum dpu_plane_type { + DPU_PLANE_INTERLEAVED, + DPU_PLANE_PLANAR, + DPU_PLANE_PSEUDO_PLANAR, +}; + +/** + * enum dpu_chroma_samp_type - chroma sub-samplng type + * @DPU_CHROMA_RGB : No chroma subsampling + * @DPU_CHROMA_H2V1 : Chroma pixels are horizontally subsampled + * @DPU_CHROMA_H1V2 : Chroma pixels are vertically subsampled + * @DPU_CHROMA_420 : 420 subsampling + */ +enum dpu_chroma_samp_type { + DPU_CHROMA_RGB, + DPU_CHROMA_H2V1, + DPU_CHROMA_H1V2, + DPU_CHROMA_420 +}; + +/** + * dpu_fetch_type - Defines How DPU HW fetches data + * @DPU_FETCH_LINEAR : fetch is line by line + * @DPU_FETCH_TILE : fetches data in Z order from a tile + * @DPU_FETCH_UBWC : fetch and decompress data + */ +enum dpu_fetch_type { + DPU_FETCH_LINEAR, + DPU_FETCH_TILE, + DPU_FETCH_UBWC +}; + +/** + * Value of enum chosen to fit the number of bits + * expected by the HW programming. + */ +enum { + COLOR_ALPHA_1BIT = 0, + COLOR_ALPHA_4BIT = 1, + COLOR_4BIT = 0, + COLOR_5BIT = 1, /* No 5-bit Alpha */ + COLOR_6BIT = 2, /* 6-Bit Alpha also = 2 */ + COLOR_8BIT = 3, /* 8-Bit Alpha also = 3 */ +}; + +/** + * enum dpu_3d_blend_mode + * Desribes how the 3d data is blended + * @BLEND_3D_NONE : 3d blending not enabled + * @BLEND_3D_FRAME_INT : Frame interleaving + * @BLEND_3D_H_ROW_INT : Horizontal row interleaving + * @BLEND_3D_V_ROW_INT : vertical row interleaving + * @BLEND_3D_COL_INT : column interleaving + * @BLEND_3D_MAX : + */ +enum dpu_3d_blend_mode { + BLEND_3D_NONE = 0, + BLEND_3D_FRAME_INT, + BLEND_3D_H_ROW_INT, + BLEND_3D_V_ROW_INT, + BLEND_3D_COL_INT, + BLEND_3D_MAX +}; + +/** struct dpu_format - defines the format configuration which + * allows DPU HW to correctly fetch and decode the format + * @base: base msm_format struture containing fourcc code + * @fetch_planes: how the color components are packed in pixel format + * @element: element color ordering + * @bits: element bit widths + * @chroma_sample: chroma sub-samplng type + * @unpack_align_msb: unpack aligned, 0 to LSB, 1 to MSB + * @unpack_tight: 0 for loose, 1 for tight + * @unpack_count: 0 = 1 component, 1 = 2 component + * @bpp: bytes per pixel + * @alpha_enable: whether the format has an alpha channel + * @num_planes: number of planes (including meta data planes) + * @fetch_mode: linear, tiled, or ubwc hw fetch behavior + * @is_yuv: is format a yuv variant + * @flag: usage bit flags + * @tile_width: format tile width + * @tile_height: format tile height + */ +struct dpu_format { + struct msm_format base; + enum dpu_plane_type fetch_planes; + u8 element[DPU_MAX_PLANES]; + u8 bits[DPU_MAX_PLANES]; + enum dpu_chroma_samp_type chroma_sample; + u8 unpack_align_msb; + u8 unpack_tight; + u8 unpack_count; + u8 bpp; + u8 alpha_enable; + u8 num_planes; + enum dpu_fetch_type fetch_mode; + DECLARE_BITMAP(flag, DPU_FORMAT_FLAG_BIT_MAX); + u16 tile_width; + u16 tile_height; +}; +#define to_dpu_format(x) container_of(x, struct dpu_format, base) + +/** + * struct dpu_hw_fmt_layout - format information of the source pixel data + * @format: pixel format parameters + * @num_planes: number of planes (including meta data planes) + * @width: image width + * @height: image height + * @total_size: total size in bytes + * @plane_addr: address of each plane + * @plane_size: length of each plane + * @plane_pitch: pitch of each plane + */ +struct dpu_hw_fmt_layout { + const struct dpu_format *format; + uint32_t num_planes; + uint32_t width; + uint32_t height; + uint32_t total_size; + uint32_t plane_addr[DPU_MAX_PLANES]; + uint32_t plane_size[DPU_MAX_PLANES]; + uint32_t plane_pitch[DPU_MAX_PLANES]; +}; + +struct dpu_csc_cfg { + /* matrix coefficients in S15.16 format */ + uint32_t csc_mv[DPU_CSC_MATRIX_COEFF_SIZE]; + uint32_t csc_pre_bv[DPU_CSC_BIAS_SIZE]; + uint32_t csc_post_bv[DPU_CSC_BIAS_SIZE]; + uint32_t csc_pre_lv[DPU_CSC_CLAMP_SIZE]; + uint32_t csc_post_lv[DPU_CSC_CLAMP_SIZE]; +}; + +/** + * struct dpu_mdss_color - mdss color description + * color 0 : green + * color 1 : blue + * color 2 : red + * color 3 : alpha + */ +struct dpu_mdss_color { + u32 color_0; + u32 color_1; + u32 color_2; + u32 color_3; +}; + +/* + * Define bit masks for h/w logging. + */ +#define DPU_DBG_MASK_NONE (1 << 0) +#define DPU_DBG_MASK_CDM (1 << 1) +#define DPU_DBG_MASK_INTF (1 << 2) +#define DPU_DBG_MASK_LM (1 << 3) +#define DPU_DBG_MASK_CTL (1 << 4) +#define DPU_DBG_MASK_PINGPONG (1 << 5) +#define DPU_DBG_MASK_SSPP (1 << 6) +#define DPU_DBG_MASK_WB (1 << 7) +#define DPU_DBG_MASK_TOP (1 << 8) +#define DPU_DBG_MASK_VBIF (1 << 9) +#define DPU_DBG_MASK_ROT (1 << 10) + +#endif /* _DPU_HW_MDSS_H */ diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_pingpong.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_pingpong.c new file mode 100644 index 000000000000..cc3a623903f4 --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_pingpong.c @@ -0,0 +1,250 @@ +/* Copyright (c) 2015-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#include <linux/iopoll.h> + +#include "dpu_hw_mdss.h" +#include "dpu_hwio.h" +#include "dpu_hw_catalog.h" +#include "dpu_hw_pingpong.h" +#include "dpu_dbg.h" +#include "dpu_kms.h" +#include "dpu_trace.h" + +#define PP_TEAR_CHECK_EN 0x000 +#define PP_SYNC_CONFIG_VSYNC 0x004 +#define PP_SYNC_CONFIG_HEIGHT 0x008 +#define PP_SYNC_WRCOUNT 0x00C +#define PP_VSYNC_INIT_VAL 0x010 +#define PP_INT_COUNT_VAL 0x014 +#define PP_SYNC_THRESH 0x018 +#define PP_START_POS 0x01C +#define PP_RD_PTR_IRQ 0x020 +#define PP_WR_PTR_IRQ 0x024 +#define PP_OUT_LINE_COUNT 0x028 +#define PP_LINE_COUNT 0x02C + +#define PP_FBC_MODE 0x034 +#define PP_FBC_BUDGET_CTL 0x038 +#define PP_FBC_LOSSY_MODE 0x03C + +static struct dpu_pingpong_cfg *_pingpong_offset(enum dpu_pingpong pp, + struct dpu_mdss_cfg *m, + void __iomem *addr, + struct dpu_hw_blk_reg_map *b) +{ + int i; + + for (i = 0; i < m->pingpong_count; i++) { + if (pp == m->pingpong[i].id) { + b->base_off = addr; + b->blk_off = m->pingpong[i].base; + b->length = m->pingpong[i].len; + b->hwversion = m->hwversion; + b->log_mask = DPU_DBG_MASK_PINGPONG; + return &m->pingpong[i]; + } + } + + return ERR_PTR(-EINVAL); +} + +static int dpu_hw_pp_setup_te_config(struct dpu_hw_pingpong *pp, + struct dpu_hw_tear_check *te) +{ + struct dpu_hw_blk_reg_map *c; + int cfg; + + if (!pp || !te) + return -EINVAL; + c = &pp->hw; + + cfg = BIT(19); /*VSYNC_COUNTER_EN */ + if (te->hw_vsync_mode) + cfg |= BIT(20); + + cfg |= te->vsync_count; + + DPU_REG_WRITE(c, PP_SYNC_CONFIG_VSYNC, cfg); + DPU_REG_WRITE(c, PP_SYNC_CONFIG_HEIGHT, te->sync_cfg_height); + DPU_REG_WRITE(c, PP_VSYNC_INIT_VAL, te->vsync_init_val); + DPU_REG_WRITE(c, PP_RD_PTR_IRQ, te->rd_ptr_irq); + DPU_REG_WRITE(c, PP_START_POS, te->start_pos); + DPU_REG_WRITE(c, PP_SYNC_THRESH, + ((te->sync_threshold_continue << 16) | + te->sync_threshold_start)); + DPU_REG_WRITE(c, PP_SYNC_WRCOUNT, + (te->start_pos + te->sync_threshold_start + 1)); + + return 0; +} + +static int dpu_hw_pp_poll_timeout_wr_ptr(struct dpu_hw_pingpong *pp, + u32 timeout_us) +{ + struct dpu_hw_blk_reg_map *c; + u32 val; + int rc; + + if (!pp) + return -EINVAL; + + c = &pp->hw; + rc = readl_poll_timeout(c->base_off + c->blk_off + PP_LINE_COUNT, + val, (val & 0xffff) >= 1, 10, timeout_us); + + return rc; +} + +static int dpu_hw_pp_enable_te(struct dpu_hw_pingpong *pp, bool enable) +{ + struct dpu_hw_blk_reg_map *c; + + if (!pp) + return -EINVAL; + c = &pp->hw; + + DPU_REG_WRITE(c, PP_TEAR_CHECK_EN, enable); + return 0; +} + +static int dpu_hw_pp_connect_external_te(struct dpu_hw_pingpong *pp, + bool enable_external_te) +{ + struct dpu_hw_blk_reg_map *c = &pp->hw; + u32 cfg; + int orig; + + if (!pp) + return -EINVAL; + + c = &pp->hw; + cfg = DPU_REG_READ(c, PP_SYNC_CONFIG_VSYNC); + orig = (bool)(cfg & BIT(20)); + if (enable_external_te) + cfg |= BIT(20); + else + cfg &= ~BIT(20); + DPU_REG_WRITE(c, PP_SYNC_CONFIG_VSYNC, cfg); + trace_dpu_pp_connect_ext_te(pp->idx - PINGPONG_0, cfg); + + return orig; +} + +static int dpu_hw_pp_get_vsync_info(struct dpu_hw_pingpong *pp, + struct dpu_hw_pp_vsync_info *info) +{ + struct dpu_hw_blk_reg_map *c; + u32 val; + + if (!pp || !info) + return -EINVAL; + c = &pp->hw; + + val = DPU_REG_READ(c, PP_VSYNC_INIT_VAL); + info->rd_ptr_init_val = val & 0xffff; + + val = DPU_REG_READ(c, PP_INT_COUNT_VAL); + info->rd_ptr_frame_count = (val & 0xffff0000) >> 16; + info->rd_ptr_line_count = val & 0xffff; + + val = DPU_REG_READ(c, PP_LINE_COUNT); + info->wr_ptr_line_count = val & 0xffff; + + return 0; +} + +static u32 dpu_hw_pp_get_line_count(struct dpu_hw_pingpong *pp) +{ + struct dpu_hw_blk_reg_map *c = &pp->hw; + u32 height, init; + u32 line = 0xFFFF; + + if (!pp) + return 0; + c = &pp->hw; + + init = DPU_REG_READ(c, PP_VSYNC_INIT_VAL) & 0xFFFF; + height = DPU_REG_READ(c, PP_SYNC_CONFIG_HEIGHT) & 0xFFFF; + + if (height < init) + goto line_count_exit; + + line = DPU_REG_READ(c, PP_INT_COUNT_VAL) & 0xFFFF; + + if (line < init) + line += (0xFFFF - init); + else + line -= init; + +line_count_exit: + return line; +} + +static void _setup_pingpong_ops(struct dpu_hw_pingpong_ops *ops, + const struct dpu_pingpong_cfg *hw_cap) +{ + ops->setup_tearcheck = dpu_hw_pp_setup_te_config; + ops->enable_tearcheck = dpu_hw_pp_enable_te; + ops->connect_external_te = dpu_hw_pp_connect_external_te; + ops->get_vsync_info = dpu_hw_pp_get_vsync_info; + ops->poll_timeout_wr_ptr = dpu_hw_pp_poll_timeout_wr_ptr; + ops->get_line_count = dpu_hw_pp_get_line_count; +}; + +static struct dpu_hw_blk_ops dpu_hw_ops = { + .start = NULL, + .stop = NULL, +}; + +struct dpu_hw_pingpong *dpu_hw_pingpong_init(enum dpu_pingpong idx, + void __iomem *addr, + struct dpu_mdss_cfg *m) +{ + struct dpu_hw_pingpong *c; + struct dpu_pingpong_cfg *cfg; + int rc; + + c = kzalloc(sizeof(*c), GFP_KERNEL); + if (!c) + return ERR_PTR(-ENOMEM); + + cfg = _pingpong_offset(idx, m, addr, &c->hw); + if (IS_ERR_OR_NULL(cfg)) { + kfree(c); + return ERR_PTR(-EINVAL); + } + + c->idx = idx; + c->caps = cfg; + _setup_pingpong_ops(&c->ops, c->caps); + + rc = dpu_hw_blk_init(&c->base, DPU_HW_BLK_PINGPONG, idx, &dpu_hw_ops); + if (rc) { + DPU_ERROR("failed to init hw blk %d\n", rc); + goto blk_init_error; + } + + return c; + +blk_init_error: + kzfree(c); + + return ERR_PTR(rc); +} + +void dpu_hw_pingpong_destroy(struct dpu_hw_pingpong *pp) +{ + if (pp) + dpu_hw_blk_destroy(&pp->base); + kfree(pp); +} diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_pingpong.h b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_pingpong.h new file mode 100644 index 000000000000..3caccd7d6a3e --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_pingpong.h @@ -0,0 +1,136 @@ +/* Copyright (c) 2015-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#ifndef _DPU_HW_PINGPONG_H +#define _DPU_HW_PINGPONG_H + +#include "dpu_hw_catalog.h" +#include "dpu_hw_mdss.h" +#include "dpu_hw_util.h" +#include "dpu_hw_blk.h" + +struct dpu_hw_pingpong; + +struct dpu_hw_tear_check { + /* + * This is ratio of MDP VSYNC clk freq(Hz) to + * refresh rate divided by no of lines + */ + u32 vsync_count; + u32 sync_cfg_height; + u32 vsync_init_val; + u32 sync_threshold_start; + u32 sync_threshold_continue; + u32 start_pos; + u32 rd_ptr_irq; + u8 hw_vsync_mode; +}; + +struct dpu_hw_pp_vsync_info { + u32 rd_ptr_init_val; /* value of rd pointer at vsync edge */ + u32 rd_ptr_frame_count; /* num frames sent since enabling interface */ + u32 rd_ptr_line_count; /* current line on panel (rd ptr) */ + u32 wr_ptr_line_count; /* current line within pp fifo (wr ptr) */ +}; + +/** + * + * struct dpu_hw_pingpong_ops : Interface to the pingpong Hw driver functions + * Assumption is these functions will be called after clocks are enabled + * @setup_tearcheck : program tear check values + * @enable_tearcheck : enables tear check + * @get_vsync_info : retries timing info of the panel + * @setup_dither : function to program the dither hw block + * @get_line_count: obtain current vertical line counter + */ +struct dpu_hw_pingpong_ops { + /** + * enables vysnc generation and sets up init value of + * read pointer and programs the tear check cofiguration + */ + int (*setup_tearcheck)(struct dpu_hw_pingpong *pp, + struct dpu_hw_tear_check *cfg); + + /** + * enables tear check block + */ + int (*enable_tearcheck)(struct dpu_hw_pingpong *pp, + bool enable); + + /** + * read, modify, write to either set or clear listening to external TE + * @Return: 1 if TE was originally connected, 0 if not, or -ERROR + */ + int (*connect_external_te)(struct dpu_hw_pingpong *pp, + bool enable_external_te); + + /** + * provides the programmed and current + * line_count + */ + int (*get_vsync_info)(struct dpu_hw_pingpong *pp, + struct dpu_hw_pp_vsync_info *info); + + /** + * poll until write pointer transmission starts + * @Return: 0 on success, -ETIMEDOUT on timeout + */ + int (*poll_timeout_wr_ptr)(struct dpu_hw_pingpong *pp, u32 timeout_us); + + /** + * Obtain current vertical line counter + */ + u32 (*get_line_count)(struct dpu_hw_pingpong *pp); +}; + +struct dpu_hw_pingpong { + struct dpu_hw_blk base; + struct dpu_hw_blk_reg_map hw; + + /* pingpong */ + enum dpu_pingpong idx; + const struct dpu_pingpong_cfg *caps; + + /* ops */ + struct dpu_hw_pingpong_ops ops; +}; + +/** + * dpu_hw_pingpong - convert base object dpu_hw_base to container + * @hw: Pointer to base hardware block + * return: Pointer to hardware block container + */ +static inline struct dpu_hw_pingpong *to_dpu_hw_pingpong(struct dpu_hw_blk *hw) +{ + return container_of(hw, struct dpu_hw_pingpong, base); +} + +/** + * dpu_hw_pingpong_init - initializes the pingpong driver for the passed + * pingpong idx. + * @idx: Pingpong index for which driver object is required + * @addr: Mapped register io address of MDP + * @m: Pointer to mdss catalog data + * Returns: Error code or allocated dpu_hw_pingpong context + */ +struct dpu_hw_pingpong *dpu_hw_pingpong_init(enum dpu_pingpong idx, + void __iomem *addr, + struct dpu_mdss_cfg *m); + +/** + * dpu_hw_pingpong_destroy - destroys pingpong driver context + * should be called to free the context + * @pp: Pointer to PP driver context returned by dpu_hw_pingpong_init + */ +void dpu_hw_pingpong_destroy(struct dpu_hw_pingpong *pp); + +#endif /*_DPU_HW_PINGPONG_H */ diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_sspp.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_sspp.c new file mode 100644 index 000000000000..c25b52a6b219 --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_sspp.c @@ -0,0 +1,753 @@ +/* Copyright (c) 2015-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#include "dpu_hwio.h" +#include "dpu_hw_catalog.h" +#include "dpu_hw_lm.h" +#include "dpu_hw_sspp.h" +#include "dpu_dbg.h" +#include "dpu_kms.h" + +#define DPU_FETCH_CONFIG_RESET_VALUE 0x00000087 + +/* DPU_SSPP_SRC */ +#define SSPP_SRC_SIZE 0x00 +#define SSPP_SRC_XY 0x08 +#define SSPP_OUT_SIZE 0x0c +#define SSPP_OUT_XY 0x10 +#define SSPP_SRC0_ADDR 0x14 +#define SSPP_SRC1_ADDR 0x18 +#define SSPP_SRC2_ADDR 0x1C +#define SSPP_SRC3_ADDR 0x20 +#define SSPP_SRC_YSTRIDE0 0x24 +#define SSPP_SRC_YSTRIDE1 0x28 +#define SSPP_SRC_FORMAT 0x30 +#define SSPP_SRC_UNPACK_PATTERN 0x34 +#define SSPP_SRC_OP_MODE 0x38 + +/* SSPP_MULTIRECT*/ +#define SSPP_SRC_SIZE_REC1 0x16C +#define SSPP_SRC_XY_REC1 0x168 +#define SSPP_OUT_SIZE_REC1 0x160 +#define SSPP_OUT_XY_REC1 0x164 +#define SSPP_SRC_FORMAT_REC1 0x174 +#define SSPP_SRC_UNPACK_PATTERN_REC1 0x178 +#define SSPP_SRC_OP_MODE_REC1 0x17C +#define SSPP_MULTIRECT_OPMODE 0x170 +#define SSPP_SRC_CONSTANT_COLOR_REC1 0x180 +#define SSPP_EXCL_REC_SIZE_REC1 0x184 +#define SSPP_EXCL_REC_XY_REC1 0x188 + +#define MDSS_MDP_OP_DEINTERLACE BIT(22) +#define MDSS_MDP_OP_DEINTERLACE_ODD BIT(23) +#define MDSS_MDP_OP_IGC_ROM_1 BIT(18) +#define MDSS_MDP_OP_IGC_ROM_0 BIT(17) +#define MDSS_MDP_OP_IGC_EN BIT(16) +#define MDSS_MDP_OP_FLIP_UD BIT(14) +#define MDSS_MDP_OP_FLIP_LR BIT(13) +#define MDSS_MDP_OP_BWC_EN BIT(0) +#define MDSS_MDP_OP_PE_OVERRIDE BIT(31) +#define MDSS_MDP_OP_BWC_LOSSLESS (0 << 1) +#define MDSS_MDP_OP_BWC_Q_HIGH (1 << 1) +#define MDSS_MDP_OP_BWC_Q_MED (2 << 1) + +#define SSPP_SRC_CONSTANT_COLOR 0x3c +#define SSPP_EXCL_REC_CTL 0x40 +#define SSPP_UBWC_STATIC_CTRL 0x44 +#define SSPP_FETCH_CONFIG 0x048 +#define SSPP_DANGER_LUT 0x60 +#define SSPP_SAFE_LUT 0x64 +#define SSPP_CREQ_LUT 0x68 +#define SSPP_QOS_CTRL 0x6C +#define SSPP_DECIMATION_CONFIG 0xB4 +#define SSPP_SRC_ADDR_SW_STATUS 0x70 +#define SSPP_CREQ_LUT_0 0x74 +#define SSPP_CREQ_LUT_1 0x78 +#define SSPP_SW_PIX_EXT_C0_LR 0x100 +#define SSPP_SW_PIX_EXT_C0_TB 0x104 +#define SSPP_SW_PIX_EXT_C0_REQ_PIXELS 0x108 +#define SSPP_SW_PIX_EXT_C1C2_LR 0x110 +#define SSPP_SW_PIX_EXT_C1C2_TB 0x114 +#define SSPP_SW_PIX_EXT_C1C2_REQ_PIXELS 0x118 +#define SSPP_SW_PIX_EXT_C3_LR 0x120 +#define SSPP_SW_PIX_EXT_C3_TB 0x124 +#define SSPP_SW_PIX_EXT_C3_REQ_PIXELS 0x128 +#define SSPP_TRAFFIC_SHAPER 0x130 +#define SSPP_CDP_CNTL 0x134 +#define SSPP_UBWC_ERROR_STATUS 0x138 +#define SSPP_TRAFFIC_SHAPER_PREFILL 0x150 +#define SSPP_TRAFFIC_SHAPER_REC1_PREFILL 0x154 +#define SSPP_TRAFFIC_SHAPER_REC1 0x158 +#define SSPP_EXCL_REC_SIZE 0x1B4 +#define SSPP_EXCL_REC_XY 0x1B8 +#define SSPP_VIG_OP_MODE 0x0 +#define SSPP_VIG_CSC_10_OP_MODE 0x0 +#define SSPP_TRAFFIC_SHAPER_BPC_MAX 0xFF + +/* SSPP_QOS_CTRL */ +#define SSPP_QOS_CTRL_VBLANK_EN BIT(16) +#define SSPP_QOS_CTRL_DANGER_SAFE_EN BIT(0) +#define SSPP_QOS_CTRL_DANGER_VBLANK_MASK 0x3 +#define SSPP_QOS_CTRL_DANGER_VBLANK_OFF 4 +#define SSPP_QOS_CTRL_CREQ_VBLANK_MASK 0x3 +#define SSPP_QOS_CTRL_CREQ_VBLANK_OFF 20 + +/* DPU_SSPP_SCALER_QSEED2 */ +#define SCALE_CONFIG 0x04 +#define COMP0_3_PHASE_STEP_X 0x10 +#define COMP0_3_PHASE_STEP_Y 0x14 +#define COMP1_2_PHASE_STEP_X 0x18 +#define COMP1_2_PHASE_STEP_Y 0x1c +#define COMP0_3_INIT_PHASE_X 0x20 +#define COMP0_3_INIT_PHASE_Y 0x24 +#define COMP1_2_INIT_PHASE_X 0x28 +#define COMP1_2_INIT_PHASE_Y 0x2C +#define VIG_0_QSEED2_SHARP 0x30 + +/* + * Definitions for ViG op modes + */ +#define VIG_OP_CSC_DST_DATAFMT BIT(19) +#define VIG_OP_CSC_SRC_DATAFMT BIT(18) +#define VIG_OP_CSC_EN BIT(17) +#define VIG_OP_MEM_PROT_CONT BIT(15) +#define VIG_OP_MEM_PROT_VAL BIT(14) +#define VIG_OP_MEM_PROT_SAT BIT(13) +#define VIG_OP_MEM_PROT_HUE BIT(12) +#define VIG_OP_HIST BIT(8) +#define VIG_OP_SKY_COL BIT(7) +#define VIG_OP_FOIL BIT(6) +#define VIG_OP_SKIN_COL BIT(5) +#define VIG_OP_PA_EN BIT(4) +#define VIG_OP_PA_SAT_ZERO_EXP BIT(2) +#define VIG_OP_MEM_PROT_BLEND BIT(1) + +/* + * Definitions for CSC 10 op modes + */ +#define VIG_CSC_10_SRC_DATAFMT BIT(1) +#define VIG_CSC_10_EN BIT(0) +#define CSC_10BIT_OFFSET 4 + +/* traffic shaper clock in Hz */ +#define TS_CLK 19200000 + +static inline int _sspp_subblk_offset(struct dpu_hw_pipe *ctx, + int s_id, + u32 *idx) +{ + int rc = 0; + const struct dpu_sspp_sub_blks *sblk = ctx->cap->sblk; + + if (!ctx) + return -EINVAL; + + switch (s_id) { + case DPU_SSPP_SRC: + *idx = sblk->src_blk.base; + break; + case DPU_SSPP_SCALER_QSEED2: + case DPU_SSPP_SCALER_QSEED3: + case DPU_SSPP_SCALER_RGB: + *idx = sblk->scaler_blk.base; + break; + case DPU_SSPP_CSC: + case DPU_SSPP_CSC_10BIT: + *idx = sblk->csc_blk.base; + break; + default: + rc = -EINVAL; + } + + return rc; +} + +static void dpu_hw_sspp_setup_multirect(struct dpu_hw_pipe *ctx, + enum dpu_sspp_multirect_index index, + enum dpu_sspp_multirect_mode mode) +{ + u32 mode_mask; + u32 idx; + + if (_sspp_subblk_offset(ctx, DPU_SSPP_SRC, &idx)) + return; + + if (index == DPU_SSPP_RECT_SOLO) { + /** + * if rect index is RECT_SOLO, we cannot expect a + * virtual plane sharing the same SSPP id. So we go + * and disable multirect + */ + mode_mask = 0; + } else { + mode_mask = DPU_REG_READ(&ctx->hw, SSPP_MULTIRECT_OPMODE + idx); + mode_mask |= index; + if (mode == DPU_SSPP_MULTIRECT_TIME_MX) + mode_mask |= BIT(2); + else + mode_mask &= ~BIT(2); + } + + DPU_REG_WRITE(&ctx->hw, SSPP_MULTIRECT_OPMODE + idx, mode_mask); +} + +static void _sspp_setup_opmode(struct dpu_hw_pipe *ctx, + u32 mask, u8 en) +{ + u32 idx; + u32 opmode; + + if (!test_bit(DPU_SSPP_SCALER_QSEED2, &ctx->cap->features) || + _sspp_subblk_offset(ctx, DPU_SSPP_SCALER_QSEED2, &idx) || + !test_bit(DPU_SSPP_CSC, &ctx->cap->features)) + return; + + opmode = DPU_REG_READ(&ctx->hw, SSPP_VIG_OP_MODE + idx); + + if (en) + opmode |= mask; + else + opmode &= ~mask; + + DPU_REG_WRITE(&ctx->hw, SSPP_VIG_OP_MODE + idx, opmode); +} + +static void _sspp_setup_csc10_opmode(struct dpu_hw_pipe *ctx, + u32 mask, u8 en) +{ + u32 idx; + u32 opmode; + + if (_sspp_subblk_offset(ctx, DPU_SSPP_CSC_10BIT, &idx)) + return; + + opmode = DPU_REG_READ(&ctx->hw, SSPP_VIG_CSC_10_OP_MODE + idx); + if (en) + opmode |= mask; + else + opmode &= ~mask; + + DPU_REG_WRITE(&ctx->hw, SSPP_VIG_CSC_10_OP_MODE + idx, opmode); +} + +/** + * Setup source pixel format, flip, + */ +static void dpu_hw_sspp_setup_format(struct dpu_hw_pipe *ctx, + const struct dpu_format *fmt, u32 flags, + enum dpu_sspp_multirect_index rect_mode) +{ + struct dpu_hw_blk_reg_map *c; + u32 chroma_samp, unpack, src_format; + u32 opmode = 0; + u32 fast_clear = 0; + u32 op_mode_off, unpack_pat_off, format_off; + u32 idx; + + if (_sspp_subblk_offset(ctx, DPU_SSPP_SRC, &idx) || !fmt) + return; + + if (rect_mode == DPU_SSPP_RECT_SOLO || rect_mode == DPU_SSPP_RECT_0) { + op_mode_off = SSPP_SRC_OP_MODE; + unpack_pat_off = SSPP_SRC_UNPACK_PATTERN; + format_off = SSPP_SRC_FORMAT; + } else { + op_mode_off = SSPP_SRC_OP_MODE_REC1; + unpack_pat_off = SSPP_SRC_UNPACK_PATTERN_REC1; + format_off = SSPP_SRC_FORMAT_REC1; + } + + c = &ctx->hw; + opmode = DPU_REG_READ(c, op_mode_off + idx); + opmode &= ~(MDSS_MDP_OP_FLIP_LR | MDSS_MDP_OP_FLIP_UD | + MDSS_MDP_OP_BWC_EN | MDSS_MDP_OP_PE_OVERRIDE); + + if (flags & DPU_SSPP_FLIP_LR) + opmode |= MDSS_MDP_OP_FLIP_LR; + if (flags & DPU_SSPP_FLIP_UD) + opmode |= MDSS_MDP_OP_FLIP_UD; + + chroma_samp = fmt->chroma_sample; + if (flags & DPU_SSPP_SOURCE_ROTATED_90) { + if (chroma_samp == DPU_CHROMA_H2V1) + chroma_samp = DPU_CHROMA_H1V2; + else if (chroma_samp == DPU_CHROMA_H1V2) + chroma_samp = DPU_CHROMA_H2V1; + } + + src_format = (chroma_samp << 23) | (fmt->fetch_planes << 19) | + (fmt->bits[C3_ALPHA] << 6) | (fmt->bits[C2_R_Cr] << 4) | + (fmt->bits[C1_B_Cb] << 2) | (fmt->bits[C0_G_Y] << 0); + + if (flags & DPU_SSPP_ROT_90) + src_format |= BIT(11); /* ROT90 */ + + if (fmt->alpha_enable && fmt->fetch_planes == DPU_PLANE_INTERLEAVED) + src_format |= BIT(8); /* SRCC3_EN */ + + if (flags & DPU_SSPP_SOLID_FILL) + src_format |= BIT(22); + + unpack = (fmt->element[3] << 24) | (fmt->element[2] << 16) | + (fmt->element[1] << 8) | (fmt->element[0] << 0); + src_format |= ((fmt->unpack_count - 1) << 12) | + (fmt->unpack_tight << 17) | + (fmt->unpack_align_msb << 18) | + ((fmt->bpp - 1) << 9); + + if (fmt->fetch_mode != DPU_FETCH_LINEAR) { + if (DPU_FORMAT_IS_UBWC(fmt)) + opmode |= MDSS_MDP_OP_BWC_EN; + src_format |= (fmt->fetch_mode & 3) << 30; /*FRAME_FORMAT */ + DPU_REG_WRITE(c, SSPP_FETCH_CONFIG, + DPU_FETCH_CONFIG_RESET_VALUE | + ctx->mdp->highest_bank_bit << 18); + if (IS_UBWC_20_SUPPORTED(ctx->catalog->caps->ubwc_version)) { + fast_clear = fmt->alpha_enable ? BIT(31) : 0; + DPU_REG_WRITE(c, SSPP_UBWC_STATIC_CTRL, + fast_clear | (ctx->mdp->ubwc_swizzle) | + (ctx->mdp->highest_bank_bit << 4)); + } + } + + opmode |= MDSS_MDP_OP_PE_OVERRIDE; + + /* if this is YUV pixel format, enable CSC */ + if (DPU_FORMAT_IS_YUV(fmt)) + src_format |= BIT(15); + + if (DPU_FORMAT_IS_DX(fmt)) + src_format |= BIT(14); + + /* update scaler opmode, if appropriate */ + if (test_bit(DPU_SSPP_CSC, &ctx->cap->features)) + _sspp_setup_opmode(ctx, VIG_OP_CSC_EN | VIG_OP_CSC_SRC_DATAFMT, + DPU_FORMAT_IS_YUV(fmt)); + else if (test_bit(DPU_SSPP_CSC_10BIT, &ctx->cap->features)) + _sspp_setup_csc10_opmode(ctx, + VIG_CSC_10_EN | VIG_CSC_10_SRC_DATAFMT, + DPU_FORMAT_IS_YUV(fmt)); + + DPU_REG_WRITE(c, format_off + idx, src_format); + DPU_REG_WRITE(c, unpack_pat_off + idx, unpack); + DPU_REG_WRITE(c, op_mode_off + idx, opmode); + + /* clear previous UBWC error */ + DPU_REG_WRITE(c, SSPP_UBWC_ERROR_STATUS + idx, BIT(31)); +} + +static void dpu_hw_sspp_setup_pe_config(struct dpu_hw_pipe *ctx, + struct dpu_hw_pixel_ext *pe_ext) +{ + struct dpu_hw_blk_reg_map *c; + u8 color; + u32 lr_pe[4], tb_pe[4], tot_req_pixels[4]; + const u32 bytemask = 0xff; + const u32 shortmask = 0xffff; + u32 idx; + + if (_sspp_subblk_offset(ctx, DPU_SSPP_SRC, &idx) || !pe_ext) + return; + + c = &ctx->hw; + + /* program SW pixel extension override for all pipes*/ + for (color = 0; color < DPU_MAX_PLANES; color++) { + /* color 2 has the same set of registers as color 1 */ + if (color == 2) + continue; + + lr_pe[color] = ((pe_ext->right_ftch[color] & bytemask) << 24)| + ((pe_ext->right_rpt[color] & bytemask) << 16)| + ((pe_ext->left_ftch[color] & bytemask) << 8)| + (pe_ext->left_rpt[color] & bytemask); + + tb_pe[color] = ((pe_ext->btm_ftch[color] & bytemask) << 24)| + ((pe_ext->btm_rpt[color] & bytemask) << 16)| + ((pe_ext->top_ftch[color] & bytemask) << 8)| + (pe_ext->top_rpt[color] & bytemask); + + tot_req_pixels[color] = (((pe_ext->roi_h[color] + + pe_ext->num_ext_pxls_top[color] + + pe_ext->num_ext_pxls_btm[color]) & shortmask) << 16) | + ((pe_ext->roi_w[color] + + pe_ext->num_ext_pxls_left[color] + + pe_ext->num_ext_pxls_right[color]) & shortmask); + } + + /* color 0 */ + DPU_REG_WRITE(c, SSPP_SW_PIX_EXT_C0_LR + idx, lr_pe[0]); + DPU_REG_WRITE(c, SSPP_SW_PIX_EXT_C0_TB + idx, tb_pe[0]); + DPU_REG_WRITE(c, SSPP_SW_PIX_EXT_C0_REQ_PIXELS + idx, + tot_req_pixels[0]); + + /* color 1 and color 2 */ + DPU_REG_WRITE(c, SSPP_SW_PIX_EXT_C1C2_LR + idx, lr_pe[1]); + DPU_REG_WRITE(c, SSPP_SW_PIX_EXT_C1C2_TB + idx, tb_pe[1]); + DPU_REG_WRITE(c, SSPP_SW_PIX_EXT_C1C2_REQ_PIXELS + idx, + tot_req_pixels[1]); + + /* color 3 */ + DPU_REG_WRITE(c, SSPP_SW_PIX_EXT_C3_LR + idx, lr_pe[3]); + DPU_REG_WRITE(c, SSPP_SW_PIX_EXT_C3_TB + idx, lr_pe[3]); + DPU_REG_WRITE(c, SSPP_SW_PIX_EXT_C3_REQ_PIXELS + idx, + tot_req_pixels[3]); +} + +static void _dpu_hw_sspp_setup_scaler3(struct dpu_hw_pipe *ctx, + struct dpu_hw_pipe_cfg *sspp, + struct dpu_hw_pixel_ext *pe, + void *scaler_cfg) +{ + u32 idx; + struct dpu_hw_scaler3_cfg *scaler3_cfg = scaler_cfg; + + (void)pe; + if (_sspp_subblk_offset(ctx, DPU_SSPP_SCALER_QSEED3, &idx) || !sspp + || !scaler3_cfg || !ctx || !ctx->cap || !ctx->cap->sblk) + return; + + dpu_hw_setup_scaler3(&ctx->hw, scaler3_cfg, idx, + ctx->cap->sblk->scaler_blk.version, + sspp->layout.format); +} + +static u32 _dpu_hw_sspp_get_scaler3_ver(struct dpu_hw_pipe *ctx) +{ + u32 idx; + + if (!ctx || _sspp_subblk_offset(ctx, DPU_SSPP_SCALER_QSEED3, &idx)) + return 0; + + return dpu_hw_get_scaler3_ver(&ctx->hw, idx); +} + +/** + * dpu_hw_sspp_setup_rects() + */ +static void dpu_hw_sspp_setup_rects(struct dpu_hw_pipe *ctx, + struct dpu_hw_pipe_cfg *cfg, + enum dpu_sspp_multirect_index rect_index) +{ + struct dpu_hw_blk_reg_map *c; + u32 src_size, src_xy, dst_size, dst_xy, ystride0, ystride1; + u32 src_size_off, src_xy_off, out_size_off, out_xy_off; + u32 idx; + + if (_sspp_subblk_offset(ctx, DPU_SSPP_SRC, &idx) || !cfg) + return; + + c = &ctx->hw; + + if (rect_index == DPU_SSPP_RECT_SOLO || rect_index == DPU_SSPP_RECT_0) { + src_size_off = SSPP_SRC_SIZE; + src_xy_off = SSPP_SRC_XY; + out_size_off = SSPP_OUT_SIZE; + out_xy_off = SSPP_OUT_XY; + } else { + src_size_off = SSPP_SRC_SIZE_REC1; + src_xy_off = SSPP_SRC_XY_REC1; + out_size_off = SSPP_OUT_SIZE_REC1; + out_xy_off = SSPP_OUT_XY_REC1; + } + + + /* src and dest rect programming */ + src_xy = (cfg->src_rect.y1 << 16) | cfg->src_rect.x1; + src_size = (drm_rect_height(&cfg->src_rect) << 16) | + drm_rect_width(&cfg->src_rect); + dst_xy = (cfg->dst_rect.y1 << 16) | cfg->dst_rect.x1; + dst_size = (drm_rect_height(&cfg->dst_rect) << 16) | + drm_rect_width(&cfg->dst_rect); + + if (rect_index == DPU_SSPP_RECT_SOLO) { + ystride0 = (cfg->layout.plane_pitch[0]) | + (cfg->layout.plane_pitch[1] << 16); + ystride1 = (cfg->layout.plane_pitch[2]) | + (cfg->layout.plane_pitch[3] << 16); + } else { + ystride0 = DPU_REG_READ(c, SSPP_SRC_YSTRIDE0 + idx); + ystride1 = DPU_REG_READ(c, SSPP_SRC_YSTRIDE1 + idx); + + if (rect_index == DPU_SSPP_RECT_0) { + ystride0 = (ystride0 & 0xFFFF0000) | + (cfg->layout.plane_pitch[0] & 0x0000FFFF); + ystride1 = (ystride1 & 0xFFFF0000)| + (cfg->layout.plane_pitch[2] & 0x0000FFFF); + } else { + ystride0 = (ystride0 & 0x0000FFFF) | + ((cfg->layout.plane_pitch[0] << 16) & + 0xFFFF0000); + ystride1 = (ystride1 & 0x0000FFFF) | + ((cfg->layout.plane_pitch[2] << 16) & + 0xFFFF0000); + } + } + + /* rectangle register programming */ + DPU_REG_WRITE(c, src_size_off + idx, src_size); + DPU_REG_WRITE(c, src_xy_off + idx, src_xy); + DPU_REG_WRITE(c, out_size_off + idx, dst_size); + DPU_REG_WRITE(c, out_xy_off + idx, dst_xy); + + DPU_REG_WRITE(c, SSPP_SRC_YSTRIDE0 + idx, ystride0); + DPU_REG_WRITE(c, SSPP_SRC_YSTRIDE1 + idx, ystride1); +} + +static void dpu_hw_sspp_setup_sourceaddress(struct dpu_hw_pipe *ctx, + struct dpu_hw_pipe_cfg *cfg, + enum dpu_sspp_multirect_index rect_mode) +{ + int i; + u32 idx; + + if (_sspp_subblk_offset(ctx, DPU_SSPP_SRC, &idx)) + return; + + if (rect_mode == DPU_SSPP_RECT_SOLO) { + for (i = 0; i < ARRAY_SIZE(cfg->layout.plane_addr); i++) + DPU_REG_WRITE(&ctx->hw, SSPP_SRC0_ADDR + idx + i * 0x4, + cfg->layout.plane_addr[i]); + } else if (rect_mode == DPU_SSPP_RECT_0) { + DPU_REG_WRITE(&ctx->hw, SSPP_SRC0_ADDR + idx, + cfg->layout.plane_addr[0]); + DPU_REG_WRITE(&ctx->hw, SSPP_SRC2_ADDR + idx, + cfg->layout.plane_addr[2]); + } else { + DPU_REG_WRITE(&ctx->hw, SSPP_SRC1_ADDR + idx, + cfg->layout.plane_addr[0]); + DPU_REG_WRITE(&ctx->hw, SSPP_SRC3_ADDR + idx, + cfg->layout.plane_addr[2]); + } +} + +static void dpu_hw_sspp_setup_csc(struct dpu_hw_pipe *ctx, + struct dpu_csc_cfg *data) +{ + u32 idx; + bool csc10 = false; + + if (_sspp_subblk_offset(ctx, DPU_SSPP_CSC, &idx) || !data) + return; + + if (test_bit(DPU_SSPP_CSC_10BIT, &ctx->cap->features)) { + idx += CSC_10BIT_OFFSET; + csc10 = true; + } + + dpu_hw_csc_setup(&ctx->hw, idx, data, csc10); +} + +static void dpu_hw_sspp_setup_solidfill(struct dpu_hw_pipe *ctx, u32 color, enum + dpu_sspp_multirect_index rect_index) +{ + u32 idx; + + if (_sspp_subblk_offset(ctx, DPU_SSPP_SRC, &idx)) + return; + + if (rect_index == DPU_SSPP_RECT_SOLO || rect_index == DPU_SSPP_RECT_0) + DPU_REG_WRITE(&ctx->hw, SSPP_SRC_CONSTANT_COLOR + idx, color); + else + DPU_REG_WRITE(&ctx->hw, SSPP_SRC_CONSTANT_COLOR_REC1 + idx, + color); +} + +static void dpu_hw_sspp_setup_danger_safe_lut(struct dpu_hw_pipe *ctx, + struct dpu_hw_pipe_qos_cfg *cfg) +{ + u32 idx; + + if (_sspp_subblk_offset(ctx, DPU_SSPP_SRC, &idx)) + return; + + DPU_REG_WRITE(&ctx->hw, SSPP_DANGER_LUT + idx, cfg->danger_lut); + DPU_REG_WRITE(&ctx->hw, SSPP_SAFE_LUT + idx, cfg->safe_lut); +} + +static void dpu_hw_sspp_setup_creq_lut(struct dpu_hw_pipe *ctx, + struct dpu_hw_pipe_qos_cfg *cfg) +{ + u32 idx; + + if (_sspp_subblk_offset(ctx, DPU_SSPP_SRC, &idx)) + return; + + if (ctx->cap && test_bit(DPU_SSPP_QOS_8LVL, &ctx->cap->features)) { + DPU_REG_WRITE(&ctx->hw, SSPP_CREQ_LUT_0 + idx, cfg->creq_lut); + DPU_REG_WRITE(&ctx->hw, SSPP_CREQ_LUT_1 + idx, + cfg->creq_lut >> 32); + } else { + DPU_REG_WRITE(&ctx->hw, SSPP_CREQ_LUT + idx, cfg->creq_lut); + } +} + +static void dpu_hw_sspp_setup_qos_ctrl(struct dpu_hw_pipe *ctx, + struct dpu_hw_pipe_qos_cfg *cfg) +{ + u32 idx; + u32 qos_ctrl = 0; + + if (_sspp_subblk_offset(ctx, DPU_SSPP_SRC, &idx)) + return; + + if (cfg->vblank_en) { + qos_ctrl |= ((cfg->creq_vblank & + SSPP_QOS_CTRL_CREQ_VBLANK_MASK) << + SSPP_QOS_CTRL_CREQ_VBLANK_OFF); + qos_ctrl |= ((cfg->danger_vblank & + SSPP_QOS_CTRL_DANGER_VBLANK_MASK) << + SSPP_QOS_CTRL_DANGER_VBLANK_OFF); + qos_ctrl |= SSPP_QOS_CTRL_VBLANK_EN; + } + + if (cfg->danger_safe_en) + qos_ctrl |= SSPP_QOS_CTRL_DANGER_SAFE_EN; + + DPU_REG_WRITE(&ctx->hw, SSPP_QOS_CTRL + idx, qos_ctrl); +} + +static void dpu_hw_sspp_setup_cdp(struct dpu_hw_pipe *ctx, + struct dpu_hw_pipe_cdp_cfg *cfg) +{ + u32 idx; + u32 cdp_cntl = 0; + + if (!ctx || !cfg) + return; + + if (_sspp_subblk_offset(ctx, DPU_SSPP_SRC, &idx)) + return; + + if (cfg->enable) + cdp_cntl |= BIT(0); + if (cfg->ubwc_meta_enable) + cdp_cntl |= BIT(1); + if (cfg->tile_amortize_enable) + cdp_cntl |= BIT(2); + if (cfg->preload_ahead == DPU_SSPP_CDP_PRELOAD_AHEAD_64) + cdp_cntl |= BIT(3); + + DPU_REG_WRITE(&ctx->hw, SSPP_CDP_CNTL, cdp_cntl); +} + +static void _setup_layer_ops(struct dpu_hw_pipe *c, + unsigned long features) +{ + if (test_bit(DPU_SSPP_SRC, &features)) { + c->ops.setup_format = dpu_hw_sspp_setup_format; + c->ops.setup_rects = dpu_hw_sspp_setup_rects; + c->ops.setup_sourceaddress = dpu_hw_sspp_setup_sourceaddress; + c->ops.setup_solidfill = dpu_hw_sspp_setup_solidfill; + c->ops.setup_pe = dpu_hw_sspp_setup_pe_config; + } + + if (test_bit(DPU_SSPP_QOS, &features)) { + c->ops.setup_danger_safe_lut = + dpu_hw_sspp_setup_danger_safe_lut; + c->ops.setup_creq_lut = dpu_hw_sspp_setup_creq_lut; + c->ops.setup_qos_ctrl = dpu_hw_sspp_setup_qos_ctrl; + } + + if (test_bit(DPU_SSPP_CSC, &features) || + test_bit(DPU_SSPP_CSC_10BIT, &features)) + c->ops.setup_csc = dpu_hw_sspp_setup_csc; + + if (dpu_hw_sspp_multirect_enabled(c->cap)) + c->ops.setup_multirect = dpu_hw_sspp_setup_multirect; + + if (test_bit(DPU_SSPP_SCALER_QSEED3, &features)) { + c->ops.setup_scaler = _dpu_hw_sspp_setup_scaler3; + c->ops.get_scaler_ver = _dpu_hw_sspp_get_scaler3_ver; + } + + if (test_bit(DPU_SSPP_CDP, &features)) + c->ops.setup_cdp = dpu_hw_sspp_setup_cdp; +} + +static struct dpu_sspp_cfg *_sspp_offset(enum dpu_sspp sspp, + void __iomem *addr, + struct dpu_mdss_cfg *catalog, + struct dpu_hw_blk_reg_map *b) +{ + int i; + + if ((sspp < SSPP_MAX) && catalog && addr && b) { + for (i = 0; i < catalog->sspp_count; i++) { + if (sspp == catalog->sspp[i].id) { + b->base_off = addr; + b->blk_off = catalog->sspp[i].base; + b->length = catalog->sspp[i].len; + b->hwversion = catalog->hwversion; + b->log_mask = DPU_DBG_MASK_SSPP; + return &catalog->sspp[i]; + } + } + } + + return ERR_PTR(-ENOMEM); +} + +static struct dpu_hw_blk_ops dpu_hw_ops = { + .start = NULL, + .stop = NULL, +}; + +struct dpu_hw_pipe *dpu_hw_sspp_init(enum dpu_sspp idx, + void __iomem *addr, struct dpu_mdss_cfg *catalog, + bool is_virtual_pipe) +{ + struct dpu_hw_pipe *hw_pipe; + struct dpu_sspp_cfg *cfg; + int rc; + + if (!addr || !catalog) + return ERR_PTR(-EINVAL); + + hw_pipe = kzalloc(sizeof(*hw_pipe), GFP_KERNEL); + if (!hw_pipe) + return ERR_PTR(-ENOMEM); + + cfg = _sspp_offset(idx, addr, catalog, &hw_pipe->hw); + if (IS_ERR_OR_NULL(cfg)) { + kfree(hw_pipe); + return ERR_PTR(-EINVAL); + } + + /* Assign ops */ + hw_pipe->catalog = catalog; + hw_pipe->mdp = &catalog->mdp[0]; + hw_pipe->idx = idx; + hw_pipe->cap = cfg; + _setup_layer_ops(hw_pipe, hw_pipe->cap->features); + + rc = dpu_hw_blk_init(&hw_pipe->base, DPU_HW_BLK_SSPP, idx, &dpu_hw_ops); + if (rc) { + DPU_ERROR("failed to init hw blk %d\n", rc); + goto blk_init_error; + } + + return hw_pipe; + +blk_init_error: + kzfree(hw_pipe); + + return ERR_PTR(rc); +} + +void dpu_hw_sspp_destroy(struct dpu_hw_pipe *ctx) +{ + if (ctx) + dpu_hw_blk_destroy(&ctx->base); + kfree(ctx); +} + diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_sspp.h b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_sspp.h new file mode 100644 index 000000000000..4d81e5f5ce1b --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_sspp.h @@ -0,0 +1,424 @@ +/* Copyright (c) 2015-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#ifndef _DPU_HW_SSPP_H +#define _DPU_HW_SSPP_H + +#include "dpu_hw_catalog.h" +#include "dpu_hw_mdss.h" +#include "dpu_hw_util.h" +#include "dpu_hw_blk.h" +#include "dpu_formats.h" + +struct dpu_hw_pipe; + +/** + * Flags + */ +#define DPU_SSPP_FLIP_LR BIT(0) +#define DPU_SSPP_FLIP_UD BIT(1) +#define DPU_SSPP_SOURCE_ROTATED_90 BIT(2) +#define DPU_SSPP_ROT_90 BIT(3) +#define DPU_SSPP_SOLID_FILL BIT(4) + +/** + * Define all scaler feature bits in catalog + */ +#define DPU_SSPP_SCALER ((1UL << DPU_SSPP_SCALER_RGB) | \ + (1UL << DPU_SSPP_SCALER_QSEED2) | \ + (1UL << DPU_SSPP_SCALER_QSEED3)) + +/** + * Component indices + */ +enum { + DPU_SSPP_COMP_0, + DPU_SSPP_COMP_1_2, + DPU_SSPP_COMP_2, + DPU_SSPP_COMP_3, + + DPU_SSPP_COMP_MAX +}; + +/** + * DPU_SSPP_RECT_SOLO - multirect disabled + * DPU_SSPP_RECT_0 - rect0 of a multirect pipe + * DPU_SSPP_RECT_1 - rect1 of a multirect pipe + * + * Note: HW supports multirect with either RECT0 or + * RECT1. Considering no benefit of such configs over + * SOLO mode and to keep the plane management simple, + * we dont support single rect multirect configs. + */ +enum dpu_sspp_multirect_index { + DPU_SSPP_RECT_SOLO = 0, + DPU_SSPP_RECT_0, + DPU_SSPP_RECT_1, +}; + +enum dpu_sspp_multirect_mode { + DPU_SSPP_MULTIRECT_NONE = 0, + DPU_SSPP_MULTIRECT_PARALLEL, + DPU_SSPP_MULTIRECT_TIME_MX, +}; + +enum { + DPU_FRAME_LINEAR, + DPU_FRAME_TILE_A4X, + DPU_FRAME_TILE_A5X, +}; + +enum dpu_hw_filter { + DPU_SCALE_FILTER_NEAREST = 0, + DPU_SCALE_FILTER_BIL, + DPU_SCALE_FILTER_PCMN, + DPU_SCALE_FILTER_CA, + DPU_SCALE_FILTER_MAX +}; + +enum dpu_hw_filter_alpa { + DPU_SCALE_ALPHA_PIXEL_REP, + DPU_SCALE_ALPHA_BIL +}; + +enum dpu_hw_filter_yuv { + DPU_SCALE_2D_4X4, + DPU_SCALE_2D_CIR, + DPU_SCALE_1D_SEP, + DPU_SCALE_BIL +}; + +struct dpu_hw_sharp_cfg { + u32 strength; + u32 edge_thr; + u32 smooth_thr; + u32 noise_thr; +}; + +struct dpu_hw_pixel_ext { + /* scaling factors are enabled for this input layer */ + uint8_t enable_pxl_ext; + + int init_phase_x[DPU_MAX_PLANES]; + int phase_step_x[DPU_MAX_PLANES]; + int init_phase_y[DPU_MAX_PLANES]; + int phase_step_y[DPU_MAX_PLANES]; + + /* + * Number of pixels extension in left, right, top and bottom direction + * for all color components. This pixel value for each color component + * should be sum of fetch + repeat pixels. + */ + int num_ext_pxls_left[DPU_MAX_PLANES]; + int num_ext_pxls_right[DPU_MAX_PLANES]; + int num_ext_pxls_top[DPU_MAX_PLANES]; + int num_ext_pxls_btm[DPU_MAX_PLANES]; + + /* + * Number of pixels needs to be overfetched in left, right, top and + * bottom directions from source image for scaling. + */ + int left_ftch[DPU_MAX_PLANES]; + int right_ftch[DPU_MAX_PLANES]; + int top_ftch[DPU_MAX_PLANES]; + int btm_ftch[DPU_MAX_PLANES]; + + /* + * Number of pixels needs to be repeated in left, right, top and + * bottom directions for scaling. + */ + int left_rpt[DPU_MAX_PLANES]; + int right_rpt[DPU_MAX_PLANES]; + int top_rpt[DPU_MAX_PLANES]; + int btm_rpt[DPU_MAX_PLANES]; + + uint32_t roi_w[DPU_MAX_PLANES]; + uint32_t roi_h[DPU_MAX_PLANES]; + + /* + * Filter type to be used for scaling in horizontal and vertical + * directions + */ + enum dpu_hw_filter horz_filter[DPU_MAX_PLANES]; + enum dpu_hw_filter vert_filter[DPU_MAX_PLANES]; + +}; + +/** + * struct dpu_hw_pipe_cfg : Pipe description + * @layout: format layout information for programming buffer to hardware + * @src_rect: src ROI, caller takes into account the different operations + * such as decimation, flip etc to program this field + * @dest_rect: destination ROI. + * @index: index of the rectangle of SSPP + * @mode: parallel or time multiplex multirect mode + */ +struct dpu_hw_pipe_cfg { + struct dpu_hw_fmt_layout layout; + struct drm_rect src_rect; + struct drm_rect dst_rect; + enum dpu_sspp_multirect_index index; + enum dpu_sspp_multirect_mode mode; +}; + +/** + * struct dpu_hw_pipe_qos_cfg : Source pipe QoS configuration + * @danger_lut: LUT for generate danger level based on fill level + * @safe_lut: LUT for generate safe level based on fill level + * @creq_lut: LUT for generate creq level based on fill level + * @creq_vblank: creq value generated to vbif during vertical blanking + * @danger_vblank: danger value generated during vertical blanking + * @vblank_en: enable creq_vblank and danger_vblank during vblank + * @danger_safe_en: enable danger safe generation + */ +struct dpu_hw_pipe_qos_cfg { + u32 danger_lut; + u32 safe_lut; + u64 creq_lut; + u32 creq_vblank; + u32 danger_vblank; + bool vblank_en; + bool danger_safe_en; +}; + +/** + * enum CDP preload ahead address size + */ +enum { + DPU_SSPP_CDP_PRELOAD_AHEAD_32, + DPU_SSPP_CDP_PRELOAD_AHEAD_64 +}; + +/** + * struct dpu_hw_pipe_cdp_cfg : CDP configuration + * @enable: true to enable CDP + * @ubwc_meta_enable: true to enable ubwc metadata preload + * @tile_amortize_enable: true to enable amortization control for tile format + * @preload_ahead: number of request to preload ahead + * DPU_SSPP_CDP_PRELOAD_AHEAD_32, + * DPU_SSPP_CDP_PRELOAD_AHEAD_64 + */ +struct dpu_hw_pipe_cdp_cfg { + bool enable; + bool ubwc_meta_enable; + bool tile_amortize_enable; + u32 preload_ahead; +}; + +/** + * struct dpu_hw_pipe_ts_cfg - traffic shaper configuration + * @size: size to prefill in bytes, or zero to disable + * @time: time to prefill in usec, or zero to disable + */ +struct dpu_hw_pipe_ts_cfg { + u64 size; + u64 time; +}; + +/** + * struct dpu_hw_sspp_ops - interface to the SSPP Hw driver functions + * Caller must call the init function to get the pipe context for each pipe + * Assumption is these functions will be called after clocks are enabled + */ +struct dpu_hw_sspp_ops { + /** + * setup_format - setup pixel format cropping rectangle, flip + * @ctx: Pointer to pipe context + * @cfg: Pointer to pipe config structure + * @flags: Extra flags for format config + * @index: rectangle index in multirect + */ + void (*setup_format)(struct dpu_hw_pipe *ctx, + const struct dpu_format *fmt, u32 flags, + enum dpu_sspp_multirect_index index); + + /** + * setup_rects - setup pipe ROI rectangles + * @ctx: Pointer to pipe context + * @cfg: Pointer to pipe config structure + * @index: rectangle index in multirect + */ + void (*setup_rects)(struct dpu_hw_pipe *ctx, + struct dpu_hw_pipe_cfg *cfg, + enum dpu_sspp_multirect_index index); + + /** + * setup_pe - setup pipe pixel extension + * @ctx: Pointer to pipe context + * @pe_ext: Pointer to pixel ext settings + */ + void (*setup_pe)(struct dpu_hw_pipe *ctx, + struct dpu_hw_pixel_ext *pe_ext); + + /** + * setup_sourceaddress - setup pipe source addresses + * @ctx: Pointer to pipe context + * @cfg: Pointer to pipe config structure + * @index: rectangle index in multirect + */ + void (*setup_sourceaddress)(struct dpu_hw_pipe *ctx, + struct dpu_hw_pipe_cfg *cfg, + enum dpu_sspp_multirect_index index); + + /** + * setup_csc - setup color space coversion + * @ctx: Pointer to pipe context + * @data: Pointer to config structure + */ + void (*setup_csc)(struct dpu_hw_pipe *ctx, struct dpu_csc_cfg *data); + + /** + * setup_solidfill - enable/disable colorfill + * @ctx: Pointer to pipe context + * @const_color: Fill color value + * @flags: Pipe flags + * @index: rectangle index in multirect + */ + void (*setup_solidfill)(struct dpu_hw_pipe *ctx, u32 color, + enum dpu_sspp_multirect_index index); + + /** + * setup_multirect - setup multirect configuration + * @ctx: Pointer to pipe context + * @index: rectangle index in multirect + * @mode: parallel fetch / time multiplex multirect mode + */ + + void (*setup_multirect)(struct dpu_hw_pipe *ctx, + enum dpu_sspp_multirect_index index, + enum dpu_sspp_multirect_mode mode); + + /** + * setup_sharpening - setup sharpening + * @ctx: Pointer to pipe context + * @cfg: Pointer to config structure + */ + void (*setup_sharpening)(struct dpu_hw_pipe *ctx, + struct dpu_hw_sharp_cfg *cfg); + + /** + * setup_danger_safe_lut - setup danger safe LUTs + * @ctx: Pointer to pipe context + * @cfg: Pointer to pipe QoS configuration + * + */ + void (*setup_danger_safe_lut)(struct dpu_hw_pipe *ctx, + struct dpu_hw_pipe_qos_cfg *cfg); + + /** + * setup_creq_lut - setup CREQ LUT + * @ctx: Pointer to pipe context + * @cfg: Pointer to pipe QoS configuration + * + */ + void (*setup_creq_lut)(struct dpu_hw_pipe *ctx, + struct dpu_hw_pipe_qos_cfg *cfg); + + /** + * setup_qos_ctrl - setup QoS control + * @ctx: Pointer to pipe context + * @cfg: Pointer to pipe QoS configuration + * + */ + void (*setup_qos_ctrl)(struct dpu_hw_pipe *ctx, + struct dpu_hw_pipe_qos_cfg *cfg); + + /** + * setup_histogram - setup histograms + * @ctx: Pointer to pipe context + * @cfg: Pointer to histogram configuration + */ + void (*setup_histogram)(struct dpu_hw_pipe *ctx, + void *cfg); + + /** + * setup_scaler - setup scaler + * @ctx: Pointer to pipe context + * @pipe_cfg: Pointer to pipe configuration + * @pe_cfg: Pointer to pixel extension configuration + * @scaler_cfg: Pointer to scaler configuration + */ + void (*setup_scaler)(struct dpu_hw_pipe *ctx, + struct dpu_hw_pipe_cfg *pipe_cfg, + struct dpu_hw_pixel_ext *pe_cfg, + void *scaler_cfg); + + /** + * get_scaler_ver - get scaler h/w version + * @ctx: Pointer to pipe context + */ + u32 (*get_scaler_ver)(struct dpu_hw_pipe *ctx); + + /** + * setup_cdp - setup client driven prefetch + * @ctx: Pointer to pipe context + * @cfg: Pointer to cdp configuration + */ + void (*setup_cdp)(struct dpu_hw_pipe *ctx, + struct dpu_hw_pipe_cdp_cfg *cfg); +}; + +/** + * struct dpu_hw_pipe - pipe description + * @base: hardware block base structure + * @hw: block hardware details + * @catalog: back pointer to catalog + * @mdp: pointer to associated mdp portion of the catalog + * @idx: pipe index + * @cap: pointer to layer_cfg + * @ops: pointer to operations possible for this pipe + */ +struct dpu_hw_pipe { + struct dpu_hw_blk base; + struct dpu_hw_blk_reg_map hw; + struct dpu_mdss_cfg *catalog; + struct dpu_mdp_cfg *mdp; + + /* Pipe */ + enum dpu_sspp idx; + const struct dpu_sspp_cfg *cap; + + /* Ops */ + struct dpu_hw_sspp_ops ops; +}; + +/** + * dpu_hw_pipe - convert base object dpu_hw_base to container + * @hw: Pointer to base hardware block + * return: Pointer to hardware block container + */ +static inline struct dpu_hw_pipe *to_dpu_hw_pipe(struct dpu_hw_blk *hw) +{ + return container_of(hw, struct dpu_hw_pipe, base); +} + +/** + * dpu_hw_sspp_init - initializes the sspp hw driver object. + * Should be called once before accessing every pipe. + * @idx: Pipe index for which driver object is required + * @addr: Mapped register io address of MDP + * @catalog : Pointer to mdss catalog data + * @is_virtual_pipe: is this pipe virtual pipe + */ +struct dpu_hw_pipe *dpu_hw_sspp_init(enum dpu_sspp idx, + void __iomem *addr, struct dpu_mdss_cfg *catalog, + bool is_virtual_pipe); + +/** + * dpu_hw_sspp_destroy(): Destroys SSPP driver context + * should be called during Hw pipe cleanup. + * @ctx: Pointer to SSPP driver context returned by dpu_hw_sspp_init + */ +void dpu_hw_sspp_destroy(struct dpu_hw_pipe *ctx); + +#endif /*_DPU_HW_SSPP_H */ + diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_top.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_top.c new file mode 100644 index 000000000000..db2798e862fc --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_top.c @@ -0,0 +1,398 @@ +/* Copyright (c) 2015-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#include "dpu_hwio.h" +#include "dpu_hw_catalog.h" +#include "dpu_hw_top.h" +#include "dpu_dbg.h" +#include "dpu_kms.h" + +#define SSPP_SPARE 0x28 +#define UBWC_STATIC 0x144 + +#define FLD_SPLIT_DISPLAY_CMD BIT(1) +#define FLD_SMART_PANEL_FREE_RUN BIT(2) +#define FLD_INTF_1_SW_TRG_MUX BIT(4) +#define FLD_INTF_2_SW_TRG_MUX BIT(8) +#define FLD_TE_LINE_INTER_WATERLEVEL_MASK 0xFFFF + +#define DANGER_STATUS 0x360 +#define SAFE_STATUS 0x364 + +#define TE_LINE_INTERVAL 0x3F4 + +#define TRAFFIC_SHAPER_EN BIT(31) +#define TRAFFIC_SHAPER_RD_CLIENT(num) (0x030 + (num * 4)) +#define TRAFFIC_SHAPER_WR_CLIENT(num) (0x060 + (num * 4)) +#define TRAFFIC_SHAPER_FIXPOINT_FACTOR 4 + +#define MDP_WD_TIMER_0_CTL 0x380 +#define MDP_WD_TIMER_0_CTL2 0x384 +#define MDP_WD_TIMER_0_LOAD_VALUE 0x388 +#define MDP_WD_TIMER_1_CTL 0x390 +#define MDP_WD_TIMER_1_CTL2 0x394 +#define MDP_WD_TIMER_1_LOAD_VALUE 0x398 +#define MDP_WD_TIMER_2_CTL 0x420 +#define MDP_WD_TIMER_2_CTL2 0x424 +#define MDP_WD_TIMER_2_LOAD_VALUE 0x428 +#define MDP_WD_TIMER_3_CTL 0x430 +#define MDP_WD_TIMER_3_CTL2 0x434 +#define MDP_WD_TIMER_3_LOAD_VALUE 0x438 +#define MDP_WD_TIMER_4_CTL 0x440 +#define MDP_WD_TIMER_4_CTL2 0x444 +#define MDP_WD_TIMER_4_LOAD_VALUE 0x448 + +#define MDP_TICK_COUNT 16 +#define XO_CLK_RATE 19200 +#define MS_TICKS_IN_SEC 1000 + +#define CALCULATE_WD_LOAD_VALUE(fps) \ + ((uint32_t)((MS_TICKS_IN_SEC * XO_CLK_RATE)/(MDP_TICK_COUNT * fps))) + +#define DCE_SEL 0x450 + +static void dpu_hw_setup_split_pipe(struct dpu_hw_mdp *mdp, + struct split_pipe_cfg *cfg) +{ + struct dpu_hw_blk_reg_map *c; + u32 upper_pipe = 0; + u32 lower_pipe = 0; + + if (!mdp || !cfg) + return; + + c = &mdp->hw; + + if (cfg->en) { + if (cfg->mode == INTF_MODE_CMD) { + lower_pipe = FLD_SPLIT_DISPLAY_CMD; + /* interface controlling sw trigger */ + if (cfg->intf == INTF_2) + lower_pipe |= FLD_INTF_1_SW_TRG_MUX; + else + lower_pipe |= FLD_INTF_2_SW_TRG_MUX; + upper_pipe = lower_pipe; + } else { + if (cfg->intf == INTF_2) { + lower_pipe = FLD_INTF_1_SW_TRG_MUX; + upper_pipe = FLD_INTF_2_SW_TRG_MUX; + } else { + lower_pipe = FLD_INTF_2_SW_TRG_MUX; + upper_pipe = FLD_INTF_1_SW_TRG_MUX; + } + } + } + + DPU_REG_WRITE(c, SSPP_SPARE, cfg->split_flush_en ? 0x1 : 0x0); + DPU_REG_WRITE(c, SPLIT_DISPLAY_LOWER_PIPE_CTRL, lower_pipe); + DPU_REG_WRITE(c, SPLIT_DISPLAY_UPPER_PIPE_CTRL, upper_pipe); + DPU_REG_WRITE(c, SPLIT_DISPLAY_EN, cfg->en & 0x1); +} + +static void dpu_hw_setup_cdm_output(struct dpu_hw_mdp *mdp, + struct cdm_output_cfg *cfg) +{ + struct dpu_hw_blk_reg_map *c; + u32 out_ctl = 0; + + if (!mdp || !cfg) + return; + + c = &mdp->hw; + + if (cfg->intf_en) + out_ctl |= BIT(19); + + DPU_REG_WRITE(c, MDP_OUT_CTL_0, out_ctl); +} + +static bool dpu_hw_setup_clk_force_ctrl(struct dpu_hw_mdp *mdp, + enum dpu_clk_ctrl_type clk_ctrl, bool enable) +{ + struct dpu_hw_blk_reg_map *c; + u32 reg_off, bit_off; + u32 reg_val, new_val; + bool clk_forced_on; + + if (!mdp) + return false; + + c = &mdp->hw; + + if (clk_ctrl <= DPU_CLK_CTRL_NONE || clk_ctrl >= DPU_CLK_CTRL_MAX) + return false; + + reg_off = mdp->caps->clk_ctrls[clk_ctrl].reg_off; + bit_off = mdp->caps->clk_ctrls[clk_ctrl].bit_off; + + reg_val = DPU_REG_READ(c, reg_off); + + if (enable) + new_val = reg_val | BIT(bit_off); + else + new_val = reg_val & ~BIT(bit_off); + + DPU_REG_WRITE(c, reg_off, new_val); + + clk_forced_on = !(reg_val & BIT(bit_off)); + + return clk_forced_on; +} + + +static void dpu_hw_get_danger_status(struct dpu_hw_mdp *mdp, + struct dpu_danger_safe_status *status) +{ + struct dpu_hw_blk_reg_map *c; + u32 value; + + if (!mdp || !status) + return; + + c = &mdp->hw; + + value = DPU_REG_READ(c, DANGER_STATUS); + status->mdp = (value >> 0) & 0x3; + status->sspp[SSPP_VIG0] = (value >> 4) & 0x3; + status->sspp[SSPP_VIG1] = (value >> 6) & 0x3; + status->sspp[SSPP_VIG2] = (value >> 8) & 0x3; + status->sspp[SSPP_VIG3] = (value >> 10) & 0x3; + status->sspp[SSPP_RGB0] = (value >> 12) & 0x3; + status->sspp[SSPP_RGB1] = (value >> 14) & 0x3; + status->sspp[SSPP_RGB2] = (value >> 16) & 0x3; + status->sspp[SSPP_RGB3] = (value >> 18) & 0x3; + status->sspp[SSPP_DMA0] = (value >> 20) & 0x3; + status->sspp[SSPP_DMA1] = (value >> 22) & 0x3; + status->sspp[SSPP_DMA2] = (value >> 28) & 0x3; + status->sspp[SSPP_DMA3] = (value >> 30) & 0x3; + status->sspp[SSPP_CURSOR0] = (value >> 24) & 0x3; + status->sspp[SSPP_CURSOR1] = (value >> 26) & 0x3; +} + +static void dpu_hw_setup_vsync_source(struct dpu_hw_mdp *mdp, + struct dpu_vsync_source_cfg *cfg) +{ + struct dpu_hw_blk_reg_map *c; + u32 reg, wd_load_value, wd_ctl, wd_ctl2, i; + static const u32 pp_offset[PINGPONG_MAX] = {0xC, 0x8, 0x4, 0x13, 0x18}; + + if (!mdp || !cfg || (cfg->pp_count > ARRAY_SIZE(cfg->ppnumber))) + return; + + c = &mdp->hw; + reg = DPU_REG_READ(c, MDP_VSYNC_SEL); + for (i = 0; i < cfg->pp_count; i++) { + int pp_idx = cfg->ppnumber[i] - PINGPONG_0; + + if (pp_idx >= ARRAY_SIZE(pp_offset)) + continue; + + reg &= ~(0xf << pp_offset[pp_idx]); + reg |= (cfg->vsync_source & 0xf) << pp_offset[pp_idx]; + } + DPU_REG_WRITE(c, MDP_VSYNC_SEL, reg); + + if (cfg->vsync_source >= DPU_VSYNC_SOURCE_WD_TIMER_4 && + cfg->vsync_source <= DPU_VSYNC_SOURCE_WD_TIMER_0) { + switch (cfg->vsync_source) { + case DPU_VSYNC_SOURCE_WD_TIMER_4: + wd_load_value = MDP_WD_TIMER_4_LOAD_VALUE; + wd_ctl = MDP_WD_TIMER_4_CTL; + wd_ctl2 = MDP_WD_TIMER_4_CTL2; + break; + case DPU_VSYNC_SOURCE_WD_TIMER_3: + wd_load_value = MDP_WD_TIMER_3_LOAD_VALUE; + wd_ctl = MDP_WD_TIMER_3_CTL; + wd_ctl2 = MDP_WD_TIMER_3_CTL2; + break; + case DPU_VSYNC_SOURCE_WD_TIMER_2: + wd_load_value = MDP_WD_TIMER_2_LOAD_VALUE; + wd_ctl = MDP_WD_TIMER_2_CTL; + wd_ctl2 = MDP_WD_TIMER_2_CTL2; + break; + case DPU_VSYNC_SOURCE_WD_TIMER_1: + wd_load_value = MDP_WD_TIMER_1_LOAD_VALUE; + wd_ctl = MDP_WD_TIMER_1_CTL; + wd_ctl2 = MDP_WD_TIMER_1_CTL2; + break; + case DPU_VSYNC_SOURCE_WD_TIMER_0: + default: + wd_load_value = MDP_WD_TIMER_0_LOAD_VALUE; + wd_ctl = MDP_WD_TIMER_0_CTL; + wd_ctl2 = MDP_WD_TIMER_0_CTL2; + break; + } + + DPU_REG_WRITE(c, wd_load_value, + CALCULATE_WD_LOAD_VALUE(cfg->frame_rate)); + + DPU_REG_WRITE(c, wd_ctl, BIT(0)); /* clear timer */ + reg = DPU_REG_READ(c, wd_ctl2); + reg |= BIT(8); /* enable heartbeat timer */ + reg |= BIT(0); /* enable WD timer */ + DPU_REG_WRITE(c, wd_ctl2, reg); + + /* make sure that timers are enabled/disabled for vsync state */ + wmb(); + } +} + +static void dpu_hw_get_safe_status(struct dpu_hw_mdp *mdp, + struct dpu_danger_safe_status *status) +{ + struct dpu_hw_blk_reg_map *c; + u32 value; + + if (!mdp || !status) + return; + + c = &mdp->hw; + + value = DPU_REG_READ(c, SAFE_STATUS); + status->mdp = (value >> 0) & 0x1; + status->sspp[SSPP_VIG0] = (value >> 4) & 0x1; + status->sspp[SSPP_VIG1] = (value >> 6) & 0x1; + status->sspp[SSPP_VIG2] = (value >> 8) & 0x1; + status->sspp[SSPP_VIG3] = (value >> 10) & 0x1; + status->sspp[SSPP_RGB0] = (value >> 12) & 0x1; + status->sspp[SSPP_RGB1] = (value >> 14) & 0x1; + status->sspp[SSPP_RGB2] = (value >> 16) & 0x1; + status->sspp[SSPP_RGB3] = (value >> 18) & 0x1; + status->sspp[SSPP_DMA0] = (value >> 20) & 0x1; + status->sspp[SSPP_DMA1] = (value >> 22) & 0x1; + status->sspp[SSPP_DMA2] = (value >> 28) & 0x1; + status->sspp[SSPP_DMA3] = (value >> 30) & 0x1; + status->sspp[SSPP_CURSOR0] = (value >> 24) & 0x1; + status->sspp[SSPP_CURSOR1] = (value >> 26) & 0x1; +} + +static void dpu_hw_reset_ubwc(struct dpu_hw_mdp *mdp, struct dpu_mdss_cfg *m) +{ + struct dpu_hw_blk_reg_map c; + + if (!mdp || !m) + return; + + if (!IS_UBWC_20_SUPPORTED(m->caps->ubwc_version)) + return; + + /* force blk offset to zero to access beginning of register region */ + c = mdp->hw; + c.blk_off = 0x0; + DPU_REG_WRITE(&c, UBWC_STATIC, m->mdp[0].ubwc_static); +} + +static void dpu_hw_intf_audio_select(struct dpu_hw_mdp *mdp) +{ + struct dpu_hw_blk_reg_map *c; + + if (!mdp) + return; + + c = &mdp->hw; + + DPU_REG_WRITE(c, HDMI_DP_CORE_SELECT, 0x1); +} + +static void _setup_mdp_ops(struct dpu_hw_mdp_ops *ops, + unsigned long cap) +{ + ops->setup_split_pipe = dpu_hw_setup_split_pipe; + ops->setup_cdm_output = dpu_hw_setup_cdm_output; + ops->setup_clk_force_ctrl = dpu_hw_setup_clk_force_ctrl; + ops->get_danger_status = dpu_hw_get_danger_status; + ops->setup_vsync_source = dpu_hw_setup_vsync_source; + ops->get_safe_status = dpu_hw_get_safe_status; + ops->reset_ubwc = dpu_hw_reset_ubwc; + ops->intf_audio_select = dpu_hw_intf_audio_select; +} + +static const struct dpu_mdp_cfg *_top_offset(enum dpu_mdp mdp, + const struct dpu_mdss_cfg *m, + void __iomem *addr, + struct dpu_hw_blk_reg_map *b) +{ + int i; + + if (!m || !addr || !b) + return ERR_PTR(-EINVAL); + + for (i = 0; i < m->mdp_count; i++) { + if (mdp == m->mdp[i].id) { + b->base_off = addr; + b->blk_off = m->mdp[i].base; + b->length = m->mdp[i].len; + b->hwversion = m->hwversion; + b->log_mask = DPU_DBG_MASK_TOP; + return &m->mdp[i]; + } + } + + return ERR_PTR(-EINVAL); +} + +static struct dpu_hw_blk_ops dpu_hw_ops = { + .start = NULL, + .stop = NULL, +}; + +struct dpu_hw_mdp *dpu_hw_mdptop_init(enum dpu_mdp idx, + void __iomem *addr, + const struct dpu_mdss_cfg *m) +{ + struct dpu_hw_mdp *mdp; + const struct dpu_mdp_cfg *cfg; + int rc; + + if (!addr || !m) + return ERR_PTR(-EINVAL); + + mdp = kzalloc(sizeof(*mdp), GFP_KERNEL); + if (!mdp) + return ERR_PTR(-ENOMEM); + + cfg = _top_offset(idx, m, addr, &mdp->hw); + if (IS_ERR_OR_NULL(cfg)) { + kfree(mdp); + return ERR_PTR(-EINVAL); + } + + /* + * Assign ops + */ + mdp->idx = idx; + mdp->caps = cfg; + _setup_mdp_ops(&mdp->ops, mdp->caps->features); + + rc = dpu_hw_blk_init(&mdp->base, DPU_HW_BLK_TOP, idx, &dpu_hw_ops); + if (rc) { + DPU_ERROR("failed to init hw blk %d\n", rc); + goto blk_init_error; + } + + dpu_dbg_set_dpu_top_offset(mdp->hw.blk_off); + + return mdp; + +blk_init_error: + kzfree(mdp); + + return ERR_PTR(rc); +} + +void dpu_hw_mdp_destroy(struct dpu_hw_mdp *mdp) +{ + if (mdp) + dpu_hw_blk_destroy(&mdp->base); + kfree(mdp); +} + diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_top.h b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_top.h new file mode 100644 index 000000000000..899925aaa6d7 --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_top.h @@ -0,0 +1,202 @@ +/* Copyright (c) 2015-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#ifndef _DPU_HW_TOP_H +#define _DPU_HW_TOP_H + +#include "dpu_hw_catalog.h" +#include "dpu_hw_mdss.h" +#include "dpu_hw_util.h" +#include "dpu_hw_blk.h" + +struct dpu_hw_mdp; + +/** + * struct traffic_shaper_cfg: traffic shaper configuration + * @en : enable/disable traffic shaper + * @rd_client : true if read client; false if write client + * @client_id : client identifier + * @bpc_denom : denominator of byte per clk + * @bpc_numer : numerator of byte per clk + */ +struct traffic_shaper_cfg { + bool en; + bool rd_client; + u32 client_id; + u32 bpc_denom; + u64 bpc_numer; +}; + +/** + * struct split_pipe_cfg - pipe configuration for dual display panels + * @en : Enable/disable dual pipe confguration + * @mode : Panel interface mode + * @intf : Interface id for main control path + * @split_flush_en: Allows both the paths to be flushed when master path is + * flushed + */ +struct split_pipe_cfg { + bool en; + enum dpu_intf_mode mode; + enum dpu_intf intf; + bool split_flush_en; +}; + +/** + * struct cdm_output_cfg: output configuration for cdm + * @intf_en : enable/disable interface output + */ +struct cdm_output_cfg { + bool intf_en; +}; + +/** + * struct dpu_danger_safe_status: danger and safe status signals + * @mdp: top level status + * @sspp: source pipe status + */ +struct dpu_danger_safe_status { + u8 mdp; + u8 sspp[SSPP_MAX]; +}; + +/** + * struct dpu_vsync_source_cfg - configure vsync source and configure the + * watchdog timers if required. + * @pp_count: number of ping pongs active + * @frame_rate: Display frame rate + * @ppnumber: ping pong index array + * @vsync_source: vsync source selection + */ +struct dpu_vsync_source_cfg { + u32 pp_count; + u32 frame_rate; + u32 ppnumber[PINGPONG_MAX]; + u32 vsync_source; +}; + +/** + * struct dpu_hw_mdp_ops - interface to the MDP TOP Hw driver functions + * Assumption is these functions will be called after clocks are enabled. + * @setup_split_pipe : Programs the pipe control registers + * @setup_pp_split : Programs the pp split control registers + * @setup_cdm_output : programs cdm control + * @setup_traffic_shaper : programs traffic shaper control + */ +struct dpu_hw_mdp_ops { + /** setup_split_pipe() : Regsiters are not double buffered, thisk + * function should be called before timing control enable + * @mdp : mdp top context driver + * @cfg : upper and lower part of pipe configuration + */ + void (*setup_split_pipe)(struct dpu_hw_mdp *mdp, + struct split_pipe_cfg *p); + + /** + * setup_cdm_output() : Setup selection control of the cdm data path + * @mdp : mdp top context driver + * @cfg : cdm output configuration + */ + void (*setup_cdm_output)(struct dpu_hw_mdp *mdp, + struct cdm_output_cfg *cfg); + + /** + * setup_traffic_shaper() : Setup traffic shaper control + * @mdp : mdp top context driver + * @cfg : traffic shaper configuration + */ + void (*setup_traffic_shaper)(struct dpu_hw_mdp *mdp, + struct traffic_shaper_cfg *cfg); + + /** + * setup_clk_force_ctrl - set clock force control + * @mdp: mdp top context driver + * @clk_ctrl: clock to be controlled + * @enable: force on enable + * @return: if the clock is forced-on by this function + */ + bool (*setup_clk_force_ctrl)(struct dpu_hw_mdp *mdp, + enum dpu_clk_ctrl_type clk_ctrl, bool enable); + + /** + * get_danger_status - get danger status + * @mdp: mdp top context driver + * @status: Pointer to danger safe status + */ + void (*get_danger_status)(struct dpu_hw_mdp *mdp, + struct dpu_danger_safe_status *status); + + /** + * setup_vsync_source - setup vsync source configuration details + * @mdp: mdp top context driver + * @cfg: vsync source selection configuration + */ + void (*setup_vsync_source)(struct dpu_hw_mdp *mdp, + struct dpu_vsync_source_cfg *cfg); + + /** + * get_safe_status - get safe status + * @mdp: mdp top context driver + * @status: Pointer to danger safe status + */ + void (*get_safe_status)(struct dpu_hw_mdp *mdp, + struct dpu_danger_safe_status *status); + + /** + * reset_ubwc - reset top level UBWC configuration + * @mdp: mdp top context driver + * @m: pointer to mdss catalog data + */ + void (*reset_ubwc)(struct dpu_hw_mdp *mdp, struct dpu_mdss_cfg *m); + + /** + * intf_audio_select - select the external interface for audio + * @mdp: mdp top context driver + */ + void (*intf_audio_select)(struct dpu_hw_mdp *mdp); +}; + +struct dpu_hw_mdp { + struct dpu_hw_blk base; + struct dpu_hw_blk_reg_map hw; + + /* top */ + enum dpu_mdp idx; + const struct dpu_mdp_cfg *caps; + + /* ops */ + struct dpu_hw_mdp_ops ops; +}; + +/** + * to_dpu_hw_mdp - convert base object dpu_hw_base to container + * @hw: Pointer to base hardware block + * return: Pointer to hardware block container + */ +static inline struct dpu_hw_mdp *to_dpu_hw_mdp(struct dpu_hw_blk *hw) +{ + return container_of(hw, struct dpu_hw_mdp, base); +} + +/** + * dpu_hw_mdptop_init - initializes the top driver for the passed idx + * @idx: Interface index for which driver object is required + * @addr: Mapped register io address of MDP + * @m: Pointer to mdss catalog data + */ +struct dpu_hw_mdp *dpu_hw_mdptop_init(enum dpu_mdp idx, + void __iomem *addr, + const struct dpu_mdss_cfg *m); + +void dpu_hw_mdp_destroy(struct dpu_hw_mdp *mdp); + +#endif /*_DPU_HW_TOP_H */ diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_util.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_util.c new file mode 100644 index 000000000000..4cabae480a7b --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_util.c @@ -0,0 +1,368 @@ +/* Copyright (c) 2015-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ +#define pr_fmt(fmt) "[drm:%s:%d] " fmt, __func__, __LINE__ + +#include "msm_drv.h" +#include "dpu_kms.h" +#include "dpu_hw_mdss.h" +#include "dpu_hw_util.h" + +/* using a file static variables for debugfs access */ +static u32 dpu_hw_util_log_mask = DPU_DBG_MASK_NONE; + +/* DPU_SCALER_QSEED3 */ +#define QSEED3_HW_VERSION 0x00 +#define QSEED3_OP_MODE 0x04 +#define QSEED3_RGB2Y_COEFF 0x08 +#define QSEED3_PHASE_INIT 0x0C +#define QSEED3_PHASE_STEP_Y_H 0x10 +#define QSEED3_PHASE_STEP_Y_V 0x14 +#define QSEED3_PHASE_STEP_UV_H 0x18 +#define QSEED3_PHASE_STEP_UV_V 0x1C +#define QSEED3_PRELOAD 0x20 +#define QSEED3_DE_SHARPEN 0x24 +#define QSEED3_DE_SHARPEN_CTL 0x28 +#define QSEED3_DE_SHAPE_CTL 0x2C +#define QSEED3_DE_THRESHOLD 0x30 +#define QSEED3_DE_ADJUST_DATA_0 0x34 +#define QSEED3_DE_ADJUST_DATA_1 0x38 +#define QSEED3_DE_ADJUST_DATA_2 0x3C +#define QSEED3_SRC_SIZE_Y_RGB_A 0x40 +#define QSEED3_SRC_SIZE_UV 0x44 +#define QSEED3_DST_SIZE 0x48 +#define QSEED3_COEF_LUT_CTRL 0x4C +#define QSEED3_COEF_LUT_SWAP_BIT 0 +#define QSEED3_COEF_LUT_DIR_BIT 1 +#define QSEED3_COEF_LUT_Y_CIR_BIT 2 +#define QSEED3_COEF_LUT_UV_CIR_BIT 3 +#define QSEED3_COEF_LUT_Y_SEP_BIT 4 +#define QSEED3_COEF_LUT_UV_SEP_BIT 5 +#define QSEED3_BUFFER_CTRL 0x50 +#define QSEED3_CLK_CTRL0 0x54 +#define QSEED3_CLK_CTRL1 0x58 +#define QSEED3_CLK_STATUS 0x5C +#define QSEED3_MISR_CTRL 0x70 +#define QSEED3_MISR_SIGNATURE_0 0x74 +#define QSEED3_MISR_SIGNATURE_1 0x78 +#define QSEED3_PHASE_INIT_Y_H 0x90 +#define QSEED3_PHASE_INIT_Y_V 0x94 +#define QSEED3_PHASE_INIT_UV_H 0x98 +#define QSEED3_PHASE_INIT_UV_V 0x9C +#define QSEED3_COEF_LUT 0x100 +#define QSEED3_FILTERS 5 +#define QSEED3_LUT_REGIONS 4 +#define QSEED3_CIRCULAR_LUTS 9 +#define QSEED3_SEPARABLE_LUTS 10 +#define QSEED3_LUT_SIZE 60 +#define QSEED3_ENABLE 2 +#define QSEED3_DIR_LUT_SIZE (200 * sizeof(u32)) +#define QSEED3_CIR_LUT_SIZE \ + (QSEED3_LUT_SIZE * QSEED3_CIRCULAR_LUTS * sizeof(u32)) +#define QSEED3_SEP_LUT_SIZE \ + (QSEED3_LUT_SIZE * QSEED3_SEPARABLE_LUTS * sizeof(u32)) + +void dpu_reg_write(struct dpu_hw_blk_reg_map *c, + u32 reg_off, + u32 val, + const char *name) +{ + /* don't need to mutex protect this */ + if (c->log_mask & dpu_hw_util_log_mask) + DPU_DEBUG_DRIVER("[%s:0x%X] <= 0x%X\n", + name, c->blk_off + reg_off, val); + writel_relaxed(val, c->base_off + c->blk_off + reg_off); +} + +int dpu_reg_read(struct dpu_hw_blk_reg_map *c, u32 reg_off) +{ + return readl_relaxed(c->base_off + c->blk_off + reg_off); +} + +u32 *dpu_hw_util_get_log_mask_ptr(void) +{ + return &dpu_hw_util_log_mask; +} + +static void _dpu_hw_setup_scaler3_lut(struct dpu_hw_blk_reg_map *c, + struct dpu_hw_scaler3_cfg *scaler3_cfg, u32 offset) +{ + int i, j, filter; + int config_lut = 0x0; + unsigned long lut_flags; + u32 lut_addr, lut_offset, lut_len; + u32 *lut[QSEED3_FILTERS] = {NULL, NULL, NULL, NULL, NULL}; + static const uint32_t off_tbl[QSEED3_FILTERS][QSEED3_LUT_REGIONS][2] = { + {{18, 0x000}, {12, 0x120}, {12, 0x1E0}, {8, 0x2A0} }, + {{6, 0x320}, {3, 0x3E0}, {3, 0x440}, {3, 0x4A0} }, + {{6, 0x500}, {3, 0x5c0}, {3, 0x620}, {3, 0x680} }, + {{6, 0x380}, {3, 0x410}, {3, 0x470}, {3, 0x4d0} }, + {{6, 0x560}, {3, 0x5f0}, {3, 0x650}, {3, 0x6b0} }, + }; + + lut_flags = (unsigned long) scaler3_cfg->lut_flag; + if (test_bit(QSEED3_COEF_LUT_DIR_BIT, &lut_flags) && + (scaler3_cfg->dir_len == QSEED3_DIR_LUT_SIZE)) { + lut[0] = scaler3_cfg->dir_lut; + config_lut = 1; + } + if (test_bit(QSEED3_COEF_LUT_Y_CIR_BIT, &lut_flags) && + (scaler3_cfg->y_rgb_cir_lut_idx < QSEED3_CIRCULAR_LUTS) && + (scaler3_cfg->cir_len == QSEED3_CIR_LUT_SIZE)) { + lut[1] = scaler3_cfg->cir_lut + + scaler3_cfg->y_rgb_cir_lut_idx * QSEED3_LUT_SIZE; + config_lut = 1; + } + if (test_bit(QSEED3_COEF_LUT_UV_CIR_BIT, &lut_flags) && + (scaler3_cfg->uv_cir_lut_idx < QSEED3_CIRCULAR_LUTS) && + (scaler3_cfg->cir_len == QSEED3_CIR_LUT_SIZE)) { + lut[2] = scaler3_cfg->cir_lut + + scaler3_cfg->uv_cir_lut_idx * QSEED3_LUT_SIZE; + config_lut = 1; + } + if (test_bit(QSEED3_COEF_LUT_Y_SEP_BIT, &lut_flags) && + (scaler3_cfg->y_rgb_sep_lut_idx < QSEED3_SEPARABLE_LUTS) && + (scaler3_cfg->sep_len == QSEED3_SEP_LUT_SIZE)) { + lut[3] = scaler3_cfg->sep_lut + + scaler3_cfg->y_rgb_sep_lut_idx * QSEED3_LUT_SIZE; + config_lut = 1; + } + if (test_bit(QSEED3_COEF_LUT_UV_SEP_BIT, &lut_flags) && + (scaler3_cfg->uv_sep_lut_idx < QSEED3_SEPARABLE_LUTS) && + (scaler3_cfg->sep_len == QSEED3_SEP_LUT_SIZE)) { + lut[4] = scaler3_cfg->sep_lut + + scaler3_cfg->uv_sep_lut_idx * QSEED3_LUT_SIZE; + config_lut = 1; + } + + if (config_lut) { + for (filter = 0; filter < QSEED3_FILTERS; filter++) { + if (!lut[filter]) + continue; + lut_offset = 0; + for (i = 0; i < QSEED3_LUT_REGIONS; i++) { + lut_addr = QSEED3_COEF_LUT + offset + + off_tbl[filter][i][1]; + lut_len = off_tbl[filter][i][0] << 2; + for (j = 0; j < lut_len; j++) { + DPU_REG_WRITE(c, + lut_addr, + (lut[filter])[lut_offset++]); + lut_addr += 4; + } + } + } + } + + if (test_bit(QSEED3_COEF_LUT_SWAP_BIT, &lut_flags)) + DPU_REG_WRITE(c, QSEED3_COEF_LUT_CTRL + offset, BIT(0)); + +} + +static void _dpu_hw_setup_scaler3_de(struct dpu_hw_blk_reg_map *c, + struct dpu_hw_scaler3_de_cfg *de_cfg, u32 offset) +{ + u32 sharp_lvl, sharp_ctl, shape_ctl, de_thr; + u32 adjust_a, adjust_b, adjust_c; + + if (!de_cfg->enable) + return; + + sharp_lvl = (de_cfg->sharpen_level1 & 0x1FF) | + ((de_cfg->sharpen_level2 & 0x1FF) << 16); + + sharp_ctl = ((de_cfg->limit & 0xF) << 9) | + ((de_cfg->prec_shift & 0x7) << 13) | + ((de_cfg->clip & 0x7) << 16); + + shape_ctl = (de_cfg->thr_quiet & 0xFF) | + ((de_cfg->thr_dieout & 0x3FF) << 16); + + de_thr = (de_cfg->thr_low & 0x3FF) | + ((de_cfg->thr_high & 0x3FF) << 16); + + adjust_a = (de_cfg->adjust_a[0] & 0x3FF) | + ((de_cfg->adjust_a[1] & 0x3FF) << 10) | + ((de_cfg->adjust_a[2] & 0x3FF) << 20); + + adjust_b = (de_cfg->adjust_b[0] & 0x3FF) | + ((de_cfg->adjust_b[1] & 0x3FF) << 10) | + ((de_cfg->adjust_b[2] & 0x3FF) << 20); + + adjust_c = (de_cfg->adjust_c[0] & 0x3FF) | + ((de_cfg->adjust_c[1] & 0x3FF) << 10) | + ((de_cfg->adjust_c[2] & 0x3FF) << 20); + + DPU_REG_WRITE(c, QSEED3_DE_SHARPEN + offset, sharp_lvl); + DPU_REG_WRITE(c, QSEED3_DE_SHARPEN_CTL + offset, sharp_ctl); + DPU_REG_WRITE(c, QSEED3_DE_SHAPE_CTL + offset, shape_ctl); + DPU_REG_WRITE(c, QSEED3_DE_THRESHOLD + offset, de_thr); + DPU_REG_WRITE(c, QSEED3_DE_ADJUST_DATA_0 + offset, adjust_a); + DPU_REG_WRITE(c, QSEED3_DE_ADJUST_DATA_1 + offset, adjust_b); + DPU_REG_WRITE(c, QSEED3_DE_ADJUST_DATA_2 + offset, adjust_c); + +} + +void dpu_hw_setup_scaler3(struct dpu_hw_blk_reg_map *c, + struct dpu_hw_scaler3_cfg *scaler3_cfg, + u32 scaler_offset, u32 scaler_version, + const struct dpu_format *format) +{ + u32 op_mode = 0; + u32 phase_init, preload, src_y_rgb, src_uv, dst; + + if (!scaler3_cfg->enable) + goto end; + + op_mode |= BIT(0); + op_mode |= (scaler3_cfg->y_rgb_filter_cfg & 0x3) << 16; + + if (format && DPU_FORMAT_IS_YUV(format)) { + op_mode |= BIT(12); + op_mode |= (scaler3_cfg->uv_filter_cfg & 0x3) << 24; + } + + op_mode |= (scaler3_cfg->blend_cfg & 1) << 31; + op_mode |= (scaler3_cfg->dir_en) ? BIT(4) : 0; + + preload = + ((scaler3_cfg->preload_x[0] & 0x7F) << 0) | + ((scaler3_cfg->preload_y[0] & 0x7F) << 8) | + ((scaler3_cfg->preload_x[1] & 0x7F) << 16) | + ((scaler3_cfg->preload_y[1] & 0x7F) << 24); + + src_y_rgb = (scaler3_cfg->src_width[0] & 0x1FFFF) | + ((scaler3_cfg->src_height[0] & 0x1FFFF) << 16); + + src_uv = (scaler3_cfg->src_width[1] & 0x1FFFF) | + ((scaler3_cfg->src_height[1] & 0x1FFFF) << 16); + + dst = (scaler3_cfg->dst_width & 0x1FFFF) | + ((scaler3_cfg->dst_height & 0x1FFFF) << 16); + + if (scaler3_cfg->de.enable) { + _dpu_hw_setup_scaler3_de(c, &scaler3_cfg->de, scaler_offset); + op_mode |= BIT(8); + } + + if (scaler3_cfg->lut_flag) + _dpu_hw_setup_scaler3_lut(c, scaler3_cfg, + scaler_offset); + + if (scaler_version == 0x1002) { + phase_init = + ((scaler3_cfg->init_phase_x[0] & 0x3F) << 0) | + ((scaler3_cfg->init_phase_y[0] & 0x3F) << 8) | + ((scaler3_cfg->init_phase_x[1] & 0x3F) << 16) | + ((scaler3_cfg->init_phase_y[1] & 0x3F) << 24); + DPU_REG_WRITE(c, QSEED3_PHASE_INIT + scaler_offset, phase_init); + } else { + DPU_REG_WRITE(c, QSEED3_PHASE_INIT_Y_H + scaler_offset, + scaler3_cfg->init_phase_x[0] & 0x1FFFFF); + DPU_REG_WRITE(c, QSEED3_PHASE_INIT_Y_V + scaler_offset, + scaler3_cfg->init_phase_y[0] & 0x1FFFFF); + DPU_REG_WRITE(c, QSEED3_PHASE_INIT_UV_H + scaler_offset, + scaler3_cfg->init_phase_x[1] & 0x1FFFFF); + DPU_REG_WRITE(c, QSEED3_PHASE_INIT_UV_V + scaler_offset, + scaler3_cfg->init_phase_y[1] & 0x1FFFFF); + } + + DPU_REG_WRITE(c, QSEED3_PHASE_STEP_Y_H + scaler_offset, + scaler3_cfg->phase_step_x[0] & 0xFFFFFF); + + DPU_REG_WRITE(c, QSEED3_PHASE_STEP_Y_V + scaler_offset, + scaler3_cfg->phase_step_y[0] & 0xFFFFFF); + + DPU_REG_WRITE(c, QSEED3_PHASE_STEP_UV_H + scaler_offset, + scaler3_cfg->phase_step_x[1] & 0xFFFFFF); + + DPU_REG_WRITE(c, QSEED3_PHASE_STEP_UV_V + scaler_offset, + scaler3_cfg->phase_step_y[1] & 0xFFFFFF); + + DPU_REG_WRITE(c, QSEED3_PRELOAD + scaler_offset, preload); + + DPU_REG_WRITE(c, QSEED3_SRC_SIZE_Y_RGB_A + scaler_offset, src_y_rgb); + + DPU_REG_WRITE(c, QSEED3_SRC_SIZE_UV + scaler_offset, src_uv); + + DPU_REG_WRITE(c, QSEED3_DST_SIZE + scaler_offset, dst); + +end: + if (format && !DPU_FORMAT_IS_DX(format)) + op_mode |= BIT(14); + + if (format && format->alpha_enable) { + op_mode |= BIT(10); + if (scaler_version == 0x1002) + op_mode |= (scaler3_cfg->alpha_filter_cfg & 0x1) << 30; + else + op_mode |= (scaler3_cfg->alpha_filter_cfg & 0x3) << 29; + } + + DPU_REG_WRITE(c, QSEED3_OP_MODE + scaler_offset, op_mode); +} + +u32 dpu_hw_get_scaler3_ver(struct dpu_hw_blk_reg_map *c, + u32 scaler_offset) +{ + return DPU_REG_READ(c, QSEED3_HW_VERSION + scaler_offset); +} + +void dpu_hw_csc_setup(struct dpu_hw_blk_reg_map *c, + u32 csc_reg_off, + struct dpu_csc_cfg *data, bool csc10) +{ + static const u32 matrix_shift = 7; + u32 clamp_shift = csc10 ? 16 : 8; + u32 val; + + /* matrix coeff - convert S15.16 to S4.9 */ + val = ((data->csc_mv[0] >> matrix_shift) & 0x1FFF) | + (((data->csc_mv[1] >> matrix_shift) & 0x1FFF) << 16); + DPU_REG_WRITE(c, csc_reg_off, val); + val = ((data->csc_mv[2] >> matrix_shift) & 0x1FFF) | + (((data->csc_mv[3] >> matrix_shift) & 0x1FFF) << 16); + DPU_REG_WRITE(c, csc_reg_off + 0x4, val); + val = ((data->csc_mv[4] >> matrix_shift) & 0x1FFF) | + (((data->csc_mv[5] >> matrix_shift) & 0x1FFF) << 16); + DPU_REG_WRITE(c, csc_reg_off + 0x8, val); + val = ((data->csc_mv[6] >> matrix_shift) & 0x1FFF) | + (((data->csc_mv[7] >> matrix_shift) & 0x1FFF) << 16); + DPU_REG_WRITE(c, csc_reg_off + 0xc, val); + val = (data->csc_mv[8] >> matrix_shift) & 0x1FFF; + DPU_REG_WRITE(c, csc_reg_off + 0x10, val); + + /* Pre clamp */ + val = (data->csc_pre_lv[0] << clamp_shift) | data->csc_pre_lv[1]; + DPU_REG_WRITE(c, csc_reg_off + 0x14, val); + val = (data->csc_pre_lv[2] << clamp_shift) | data->csc_pre_lv[3]; + DPU_REG_WRITE(c, csc_reg_off + 0x18, val); + val = (data->csc_pre_lv[4] << clamp_shift) | data->csc_pre_lv[5]; + DPU_REG_WRITE(c, csc_reg_off + 0x1c, val); + + /* Post clamp */ + val = (data->csc_post_lv[0] << clamp_shift) | data->csc_post_lv[1]; + DPU_REG_WRITE(c, csc_reg_off + 0x20, val); + val = (data->csc_post_lv[2] << clamp_shift) | data->csc_post_lv[3]; + DPU_REG_WRITE(c, csc_reg_off + 0x24, val); + val = (data->csc_post_lv[4] << clamp_shift) | data->csc_post_lv[5]; + DPU_REG_WRITE(c, csc_reg_off + 0x28, val); + + /* Pre-Bias */ + DPU_REG_WRITE(c, csc_reg_off + 0x2c, data->csc_pre_bv[0]); + DPU_REG_WRITE(c, csc_reg_off + 0x30, data->csc_pre_bv[1]); + DPU_REG_WRITE(c, csc_reg_off + 0x34, data->csc_pre_bv[2]); + + /* Post-Bias */ + DPU_REG_WRITE(c, csc_reg_off + 0x38, data->csc_post_bv[0]); + DPU_REG_WRITE(c, csc_reg_off + 0x3c, data->csc_post_bv[1]); + DPU_REG_WRITE(c, csc_reg_off + 0x40, data->csc_post_bv[2]); +} diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_util.h b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_util.h new file mode 100644 index 000000000000..1240f505ca53 --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_util.h @@ -0,0 +1,348 @@ +/* Copyright (c) 2015-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#ifndef _DPU_HW_UTIL_H +#define _DPU_HW_UTIL_H + +#include <linux/io.h> +#include <linux/slab.h> +#include "dpu_hw_mdss.h" + +#define REG_MASK(n) ((BIT(n)) - 1) +struct dpu_format_extended; + +/* + * This is the common struct maintained by each sub block + * for mapping the register offsets in this block to the + * absoulute IO address + * @base_off: mdp register mapped offset + * @blk_off: pipe offset relative to mdss offset + * @length length of register block offset + * @xin_id xin id + * @hwversion mdss hw version number + */ +struct dpu_hw_blk_reg_map { + void __iomem *base_off; + u32 blk_off; + u32 length; + u32 xin_id; + u32 hwversion; + u32 log_mask; +}; + +/** + * struct dpu_hw_scaler3_de_cfg : QSEEDv3 detail enhancer configuration + * @enable: detail enhancer enable/disable + * @sharpen_level1: sharpening strength for noise + * @sharpen_level2: sharpening strength for signal + * @ clip: clip shift + * @ limit: limit value + * @ thr_quiet: quiet threshold + * @ thr_dieout: dieout threshold + * @ thr_high: low threshold + * @ thr_high: high threshold + * @ prec_shift: precision shift + * @ adjust_a: A-coefficients for mapping curve + * @ adjust_b: B-coefficients for mapping curve + * @ adjust_c: C-coefficients for mapping curve + */ +struct dpu_hw_scaler3_de_cfg { + u32 enable; + int16_t sharpen_level1; + int16_t sharpen_level2; + uint16_t clip; + uint16_t limit; + uint16_t thr_quiet; + uint16_t thr_dieout; + uint16_t thr_low; + uint16_t thr_high; + uint16_t prec_shift; + int16_t adjust_a[DPU_MAX_DE_CURVES]; + int16_t adjust_b[DPU_MAX_DE_CURVES]; + int16_t adjust_c[DPU_MAX_DE_CURVES]; +}; + + +/** + * struct dpu_hw_scaler3_cfg : QSEEDv3 configuration + * @enable: scaler enable + * @dir_en: direction detection block enable + * @ init_phase_x: horizontal initial phase + * @ phase_step_x: horizontal phase step + * @ init_phase_y: vertical initial phase + * @ phase_step_y: vertical phase step + * @ preload_x: horizontal preload value + * @ preload_y: vertical preload value + * @ src_width: source width + * @ src_height: source height + * @ dst_width: destination width + * @ dst_height: destination height + * @ y_rgb_filter_cfg: y/rgb plane filter configuration + * @ uv_filter_cfg: uv plane filter configuration + * @ alpha_filter_cfg: alpha filter configuration + * @ blend_cfg: blend coefficients configuration + * @ lut_flag: scaler LUT update flags + * 0x1 swap LUT bank + * 0x2 update 2D filter LUT + * 0x4 update y circular filter LUT + * 0x8 update uv circular filter LUT + * 0x10 update y separable filter LUT + * 0x20 update uv separable filter LUT + * @ dir_lut_idx: 2D filter LUT index + * @ y_rgb_cir_lut_idx: y circular filter LUT index + * @ uv_cir_lut_idx: uv circular filter LUT index + * @ y_rgb_sep_lut_idx: y circular filter LUT index + * @ uv_sep_lut_idx: uv separable filter LUT index + * @ dir_lut: pointer to 2D LUT + * @ cir_lut: pointer to circular filter LUT + * @ sep_lut: pointer to separable filter LUT + * @ de: detail enhancer configuration + */ +struct dpu_hw_scaler3_cfg { + u32 enable; + u32 dir_en; + int32_t init_phase_x[DPU_MAX_PLANES]; + int32_t phase_step_x[DPU_MAX_PLANES]; + int32_t init_phase_y[DPU_MAX_PLANES]; + int32_t phase_step_y[DPU_MAX_PLANES]; + + u32 preload_x[DPU_MAX_PLANES]; + u32 preload_y[DPU_MAX_PLANES]; + u32 src_width[DPU_MAX_PLANES]; + u32 src_height[DPU_MAX_PLANES]; + + u32 dst_width; + u32 dst_height; + + u32 y_rgb_filter_cfg; + u32 uv_filter_cfg; + u32 alpha_filter_cfg; + u32 blend_cfg; + + u32 lut_flag; + u32 dir_lut_idx; + + u32 y_rgb_cir_lut_idx; + u32 uv_cir_lut_idx; + u32 y_rgb_sep_lut_idx; + u32 uv_sep_lut_idx; + u32 *dir_lut; + size_t dir_len; + u32 *cir_lut; + size_t cir_len; + u32 *sep_lut; + size_t sep_len; + + /* + * Detail enhancer settings + */ + struct dpu_hw_scaler3_de_cfg de; +}; + +struct dpu_hw_scaler3_lut_cfg { + bool is_configured; + u32 *dir_lut; + size_t dir_len; + u32 *cir_lut; + size_t cir_len; + u32 *sep_lut; + size_t sep_len; +}; + +/** + * struct dpu_drm_pix_ext_v1 - version 1 of pixel ext structure + * @num_ext_pxls_lr: Number of total horizontal pixels + * @num_ext_pxls_tb: Number of total vertical lines + * @left_ftch: Number of extra pixels to overfetch from left + * @right_ftch: Number of extra pixels to overfetch from right + * @top_ftch: Number of extra lines to overfetch from top + * @btm_ftch: Number of extra lines to overfetch from bottom + * @left_rpt: Number of extra pixels to repeat from left + * @right_rpt: Number of extra pixels to repeat from right + * @top_rpt: Number of extra lines to repeat from top + * @btm_rpt: Number of extra lines to repeat from bottom + */ +struct dpu_drm_pix_ext_v1 { + /* + * Number of pixels ext in left, right, top and bottom direction + * for all color components. + */ + int32_t num_ext_pxls_lr[DPU_MAX_PLANES]; + int32_t num_ext_pxls_tb[DPU_MAX_PLANES]; + + /* + * Number of pixels needs to be overfetched in left, right, top + * and bottom directions from source image for scaling. + */ + int32_t left_ftch[DPU_MAX_PLANES]; + int32_t right_ftch[DPU_MAX_PLANES]; + int32_t top_ftch[DPU_MAX_PLANES]; + int32_t btm_ftch[DPU_MAX_PLANES]; + /* + * Number of pixels needs to be repeated in left, right, top and + * bottom directions for scaling. + */ + int32_t left_rpt[DPU_MAX_PLANES]; + int32_t right_rpt[DPU_MAX_PLANES]; + int32_t top_rpt[DPU_MAX_PLANES]; + int32_t btm_rpt[DPU_MAX_PLANES]; + +}; + +/** + * struct dpu_drm_de_v1 - version 1 of detail enhancer structure + * @enable: Enables/disables detail enhancer + * @sharpen_level1: Sharpening strength for noise + * @sharpen_level2: Sharpening strength for context + * @clip: Clip coefficient + * @limit: Detail enhancer limit factor + * @thr_quiet: Quite zone threshold + * @thr_dieout: Die-out zone threshold + * @thr_low: Linear zone left threshold + * @thr_high: Linear zone right threshold + * @prec_shift: Detail enhancer precision + * @adjust_a: Mapping curves A coefficients + * @adjust_b: Mapping curves B coefficients + * @adjust_c: Mapping curves C coefficients + */ +struct dpu_drm_de_v1 { + uint32_t enable; + int16_t sharpen_level1; + int16_t sharpen_level2; + uint16_t clip; + uint16_t limit; + uint16_t thr_quiet; + uint16_t thr_dieout; + uint16_t thr_low; + uint16_t thr_high; + uint16_t prec_shift; + int16_t adjust_a[DPU_MAX_DE_CURVES]; + int16_t adjust_b[DPU_MAX_DE_CURVES]; + int16_t adjust_c[DPU_MAX_DE_CURVES]; +}; + +/** + * struct dpu_drm_scaler_v2 - version 2 of struct dpu_drm_scaler + * @enable: Scaler enable + * @dir_en: Detail enhancer enable + * @pe: Pixel extension settings + * @horz_decimate: Horizontal decimation factor + * @vert_decimate: Vertical decimation factor + * @init_phase_x: Initial scaler phase values for x + * @phase_step_x: Phase step values for x + * @init_phase_y: Initial scaler phase values for y + * @phase_step_y: Phase step values for y + * @preload_x: Horizontal preload value + * @preload_y: Vertical preload value + * @src_width: Source width + * @src_height: Source height + * @dst_width: Destination width + * @dst_height: Destination height + * @y_rgb_filter_cfg: Y/RGB plane filter configuration + * @uv_filter_cfg: UV plane filter configuration + * @alpha_filter_cfg: Alpha filter configuration + * @blend_cfg: Selection of blend coefficients + * @lut_flag: LUT configuration flags + * @dir_lut_idx: 2d 4x4 LUT index + * @y_rgb_cir_lut_idx: Y/RGB circular LUT index + * @uv_cir_lut_idx: UV circular LUT index + * @y_rgb_sep_lut_idx: Y/RGB separable LUT index + * @uv_sep_lut_idx: UV separable LUT index + * @de: Detail enhancer settings + */ +struct dpu_drm_scaler_v2 { + /* + * General definitions + */ + uint32_t enable; + uint32_t dir_en; + + /* + * Pix ext settings + */ + struct dpu_drm_pix_ext_v1 pe; + + /* + * Decimation settings + */ + uint32_t horz_decimate; + uint32_t vert_decimate; + + /* + * Phase settings + */ + int32_t init_phase_x[DPU_MAX_PLANES]; + int32_t phase_step_x[DPU_MAX_PLANES]; + int32_t init_phase_y[DPU_MAX_PLANES]; + int32_t phase_step_y[DPU_MAX_PLANES]; + + uint32_t preload_x[DPU_MAX_PLANES]; + uint32_t preload_y[DPU_MAX_PLANES]; + uint32_t src_width[DPU_MAX_PLANES]; + uint32_t src_height[DPU_MAX_PLANES]; + + uint32_t dst_width; + uint32_t dst_height; + + uint32_t y_rgb_filter_cfg; + uint32_t uv_filter_cfg; + uint32_t alpha_filter_cfg; + uint32_t blend_cfg; + + uint32_t lut_flag; + uint32_t dir_lut_idx; + + /* for Y(RGB) and UV planes*/ + uint32_t y_rgb_cir_lut_idx; + uint32_t uv_cir_lut_idx; + uint32_t y_rgb_sep_lut_idx; + uint32_t uv_sep_lut_idx; + + /* + * Detail enhancer settings + */ + struct dpu_drm_de_v1 de; +}; + + +u32 *dpu_hw_util_get_log_mask_ptr(void); + +void dpu_reg_write(struct dpu_hw_blk_reg_map *c, + u32 reg_off, + u32 val, + const char *name); +int dpu_reg_read(struct dpu_hw_blk_reg_map *c, u32 reg_off); + +#define DPU_REG_WRITE(c, off, val) dpu_reg_write(c, off, val, #off) +#define DPU_REG_READ(c, off) dpu_reg_read(c, off) + +#define MISR_FRAME_COUNT_MASK 0xFF +#define MISR_CTRL_ENABLE BIT(8) +#define MISR_CTRL_STATUS BIT(9) +#define MISR_CTRL_STATUS_CLEAR BIT(10) +#define INTF_MISR_CTRL_FREE_RUN_MASK BIT(31) + +void *dpu_hw_util_get_dir(void); + +void dpu_hw_setup_scaler3(struct dpu_hw_blk_reg_map *c, + struct dpu_hw_scaler3_cfg *scaler3_cfg, + u32 scaler_offset, u32 scaler_version, + const struct dpu_format *format); + +u32 dpu_hw_get_scaler3_ver(struct dpu_hw_blk_reg_map *c, + u32 scaler_offset); + +void dpu_hw_csc_setup(struct dpu_hw_blk_reg_map *c, + u32 csc_reg_off, + struct dpu_csc_cfg *data, bool csc10); + +#endif /* _DPU_HW_UTIL_H */ diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_vbif.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_vbif.c new file mode 100644 index 000000000000..d43905525f92 --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_vbif.c @@ -0,0 +1,275 @@ +/* Copyright (c) 2015-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#include "dpu_hwio.h" +#include "dpu_hw_catalog.h" +#include "dpu_hw_vbif.h" +#include "dpu_dbg.h" + +#define VBIF_VERSION 0x0000 +#define VBIF_CLK_FORCE_CTRL0 0x0008 +#define VBIF_CLK_FORCE_CTRL1 0x000C +#define VBIF_QOS_REMAP_00 0x0020 +#define VBIF_QOS_REMAP_01 0x0024 +#define VBIF_QOS_REMAP_10 0x0028 +#define VBIF_QOS_REMAP_11 0x002C +#define VBIF_WRITE_GATHER_EN 0x00AC +#define VBIF_IN_RD_LIM_CONF0 0x00B0 +#define VBIF_IN_RD_LIM_CONF1 0x00B4 +#define VBIF_IN_RD_LIM_CONF2 0x00B8 +#define VBIF_IN_WR_LIM_CONF0 0x00C0 +#define VBIF_IN_WR_LIM_CONF1 0x00C4 +#define VBIF_IN_WR_LIM_CONF2 0x00C8 +#define VBIF_OUT_RD_LIM_CONF0 0x00D0 +#define VBIF_OUT_WR_LIM_CONF0 0x00D4 +#define VBIF_OUT_AXI_AMEMTYPE_CONF0 0x0160 +#define VBIF_OUT_AXI_AMEMTYPE_CONF1 0x0164 +#define VBIF_XIN_PND_ERR 0x0190 +#define VBIF_XIN_SRC_ERR 0x0194 +#define VBIF_XIN_CLR_ERR 0x019C +#define VBIF_XIN_HALT_CTRL0 0x0200 +#define VBIF_XIN_HALT_CTRL1 0x0204 +#define VBIF_XINL_QOS_RP_REMAP_000 0x0550 +#define VBIF_XINL_QOS_LVL_REMAP_000 0x0590 + +static void dpu_hw_clear_errors(struct dpu_hw_vbif *vbif, + u32 *pnd_errors, u32 *src_errors) +{ + struct dpu_hw_blk_reg_map *c; + u32 pnd, src; + + if (!vbif) + return; + c = &vbif->hw; + pnd = DPU_REG_READ(c, VBIF_XIN_PND_ERR); + src = DPU_REG_READ(c, VBIF_XIN_SRC_ERR); + + if (pnd_errors) + *pnd_errors = pnd; + if (src_errors) + *src_errors = src; + + DPU_REG_WRITE(c, VBIF_XIN_CLR_ERR, pnd | src); +} + +static void dpu_hw_set_mem_type(struct dpu_hw_vbif *vbif, + u32 xin_id, u32 value) +{ + struct dpu_hw_blk_reg_map *c; + u32 reg_off; + u32 bit_off; + u32 reg_val; + + /* + * Assume 4 bits per bit field, 8 fields per 32-bit register so + * 16 bit fields maximum across two registers + */ + if (!vbif || xin_id >= MAX_XIN_COUNT || xin_id >= 16) + return; + + c = &vbif->hw; + + if (xin_id >= 8) { + xin_id -= 8; + reg_off = VBIF_OUT_AXI_AMEMTYPE_CONF1; + } else { + reg_off = VBIF_OUT_AXI_AMEMTYPE_CONF0; + } + bit_off = (xin_id & 0x7) * 4; + reg_val = DPU_REG_READ(c, reg_off); + reg_val &= ~(0x7 << bit_off); + reg_val |= (value & 0x7) << bit_off; + DPU_REG_WRITE(c, reg_off, reg_val); +} + +static void dpu_hw_set_limit_conf(struct dpu_hw_vbif *vbif, + u32 xin_id, bool rd, u32 limit) +{ + struct dpu_hw_blk_reg_map *c = &vbif->hw; + u32 reg_val; + u32 reg_off; + u32 bit_off; + + if (rd) + reg_off = VBIF_IN_RD_LIM_CONF0; + else + reg_off = VBIF_IN_WR_LIM_CONF0; + + reg_off += (xin_id / 4) * 4; + bit_off = (xin_id % 4) * 8; + reg_val = DPU_REG_READ(c, reg_off); + reg_val &= ~(0xFF << bit_off); + reg_val |= (limit) << bit_off; + DPU_REG_WRITE(c, reg_off, reg_val); +} + +static u32 dpu_hw_get_limit_conf(struct dpu_hw_vbif *vbif, + u32 xin_id, bool rd) +{ + struct dpu_hw_blk_reg_map *c = &vbif->hw; + u32 reg_val; + u32 reg_off; + u32 bit_off; + u32 limit; + + if (rd) + reg_off = VBIF_IN_RD_LIM_CONF0; + else + reg_off = VBIF_IN_WR_LIM_CONF0; + + reg_off += (xin_id / 4) * 4; + bit_off = (xin_id % 4) * 8; + reg_val = DPU_REG_READ(c, reg_off); + limit = (reg_val >> bit_off) & 0xFF; + + return limit; +} + +static void dpu_hw_set_halt_ctrl(struct dpu_hw_vbif *vbif, + u32 xin_id, bool enable) +{ + struct dpu_hw_blk_reg_map *c = &vbif->hw; + u32 reg_val; + + reg_val = DPU_REG_READ(c, VBIF_XIN_HALT_CTRL0); + + if (enable) + reg_val |= BIT(xin_id); + else + reg_val &= ~BIT(xin_id); + + DPU_REG_WRITE(c, VBIF_XIN_HALT_CTRL0, reg_val); +} + +static bool dpu_hw_get_halt_ctrl(struct dpu_hw_vbif *vbif, + u32 xin_id) +{ + struct dpu_hw_blk_reg_map *c = &vbif->hw; + u32 reg_val; + + reg_val = DPU_REG_READ(c, VBIF_XIN_HALT_CTRL1); + + return (reg_val & BIT(xin_id)) ? true : false; +} + +static void dpu_hw_set_qos_remap(struct dpu_hw_vbif *vbif, + u32 xin_id, u32 level, u32 remap_level) +{ + struct dpu_hw_blk_reg_map *c; + u32 reg_val, reg_val_lvl, mask, reg_high, reg_shift; + + if (!vbif) + return; + + c = &vbif->hw; + + reg_high = ((xin_id & 0x8) >> 3) * 4 + (level * 8); + reg_shift = (xin_id & 0x7) * 4; + + reg_val = DPU_REG_READ(c, VBIF_XINL_QOS_RP_REMAP_000 + reg_high); + reg_val_lvl = DPU_REG_READ(c, VBIF_XINL_QOS_LVL_REMAP_000 + reg_high); + + mask = 0x7 << reg_shift; + + reg_val &= ~mask; + reg_val |= (remap_level << reg_shift) & mask; + + reg_val_lvl &= ~mask; + reg_val_lvl |= (remap_level << reg_shift) & mask; + + DPU_REG_WRITE(c, VBIF_XINL_QOS_RP_REMAP_000 + reg_high, reg_val); + DPU_REG_WRITE(c, VBIF_XINL_QOS_LVL_REMAP_000 + reg_high, reg_val_lvl); +} + +static void dpu_hw_set_write_gather_en(struct dpu_hw_vbif *vbif, u32 xin_id) +{ + struct dpu_hw_blk_reg_map *c; + u32 reg_val; + + if (!vbif || xin_id >= MAX_XIN_COUNT) + return; + + c = &vbif->hw; + + reg_val = DPU_REG_READ(c, VBIF_WRITE_GATHER_EN); + reg_val |= BIT(xin_id); + DPU_REG_WRITE(c, VBIF_WRITE_GATHER_EN, reg_val); +} + +static void _setup_vbif_ops(struct dpu_hw_vbif_ops *ops, + unsigned long cap) +{ + ops->set_limit_conf = dpu_hw_set_limit_conf; + ops->get_limit_conf = dpu_hw_get_limit_conf; + ops->set_halt_ctrl = dpu_hw_set_halt_ctrl; + ops->get_halt_ctrl = dpu_hw_get_halt_ctrl; + if (test_bit(DPU_VBIF_QOS_REMAP, &cap)) + ops->set_qos_remap = dpu_hw_set_qos_remap; + ops->set_mem_type = dpu_hw_set_mem_type; + ops->clear_errors = dpu_hw_clear_errors; + ops->set_write_gather_en = dpu_hw_set_write_gather_en; +} + +static const struct dpu_vbif_cfg *_top_offset(enum dpu_vbif vbif, + const struct dpu_mdss_cfg *m, + void __iomem *addr, + struct dpu_hw_blk_reg_map *b) +{ + int i; + + for (i = 0; i < m->vbif_count; i++) { + if (vbif == m->vbif[i].id) { + b->base_off = addr; + b->blk_off = m->vbif[i].base; + b->length = m->vbif[i].len; + b->hwversion = m->hwversion; + b->log_mask = DPU_DBG_MASK_VBIF; + return &m->vbif[i]; + } + } + + return ERR_PTR(-EINVAL); +} + +struct dpu_hw_vbif *dpu_hw_vbif_init(enum dpu_vbif idx, + void __iomem *addr, + const struct dpu_mdss_cfg *m) +{ + struct dpu_hw_vbif *c; + const struct dpu_vbif_cfg *cfg; + + c = kzalloc(sizeof(*c), GFP_KERNEL); + if (!c) + return ERR_PTR(-ENOMEM); + + cfg = _top_offset(idx, m, addr, &c->hw); + if (IS_ERR_OR_NULL(cfg)) { + kfree(c); + return ERR_PTR(-EINVAL); + } + + /* + * Assign ops + */ + c->idx = idx; + c->cap = cfg; + _setup_vbif_ops(&c->ops, c->cap->features); + + /* no need to register sub-range in dpu dbg, dump entire vbif io base */ + + return c; +} + +void dpu_hw_vbif_destroy(struct dpu_hw_vbif *vbif) +{ + kfree(vbif); +} diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_vbif.h b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_vbif.h new file mode 100644 index 000000000000..471ff673c045 --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_vbif.h @@ -0,0 +1,128 @@ +/* Copyright (c) 2015-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#ifndef _DPU_HW_VBIF_H +#define _DPU_HW_VBIF_H + +#include "dpu_hw_catalog.h" +#include "dpu_hw_mdss.h" +#include "dpu_hw_util.h" + +struct dpu_hw_vbif; + +/** + * struct dpu_hw_vbif_ops : Interface to the VBIF hardware driver functions + * Assumption is these functions will be called after clocks are enabled + */ +struct dpu_hw_vbif_ops { + /** + * set_limit_conf - set transaction limit config + * @vbif: vbif context driver + * @xin_id: client interface identifier + * @rd: true for read limit; false for write limit + * @limit: outstanding transaction limit + */ + void (*set_limit_conf)(struct dpu_hw_vbif *vbif, + u32 xin_id, bool rd, u32 limit); + + /** + * get_limit_conf - get transaction limit config + * @vbif: vbif context driver + * @xin_id: client interface identifier + * @rd: true for read limit; false for write limit + * @return: outstanding transaction limit + */ + u32 (*get_limit_conf)(struct dpu_hw_vbif *vbif, + u32 xin_id, bool rd); + + /** + * set_halt_ctrl - set halt control + * @vbif: vbif context driver + * @xin_id: client interface identifier + * @enable: halt control enable + */ + void (*set_halt_ctrl)(struct dpu_hw_vbif *vbif, + u32 xin_id, bool enable); + + /** + * get_halt_ctrl - get halt control + * @vbif: vbif context driver + * @xin_id: client interface identifier + * @return: halt control enable + */ + bool (*get_halt_ctrl)(struct dpu_hw_vbif *vbif, + u32 xin_id); + + /** + * set_qos_remap - set QoS priority remap + * @vbif: vbif context driver + * @xin_id: client interface identifier + * @level: priority level + * @remap_level: remapped level + */ + void (*set_qos_remap)(struct dpu_hw_vbif *vbif, + u32 xin_id, u32 level, u32 remap_level); + + /** + * set_mem_type - set memory type + * @vbif: vbif context driver + * @xin_id: client interface identifier + * @value: memory type value + */ + void (*set_mem_type)(struct dpu_hw_vbif *vbif, + u32 xin_id, u32 value); + + /** + * clear_errors - clear any vbif errors + * This function clears any detected pending/source errors + * on the VBIF interface, and optionally returns the detected + * error mask(s). + * @vbif: vbif context driver + * @pnd_errors: pointer to pending error reporting variable + * @src_errors: pointer to source error reporting variable + */ + void (*clear_errors)(struct dpu_hw_vbif *vbif, + u32 *pnd_errors, u32 *src_errors); + + /** + * set_write_gather_en - set write_gather enable + * @vbif: vbif context driver + * @xin_id: client interface identifier + */ + void (*set_write_gather_en)(struct dpu_hw_vbif *vbif, u32 xin_id); +}; + +struct dpu_hw_vbif { + /* base */ + struct dpu_hw_blk_reg_map hw; + + /* vbif */ + enum dpu_vbif idx; + const struct dpu_vbif_cfg *cap; + + /* ops */ + struct dpu_hw_vbif_ops ops; +}; + +/** + * dpu_hw_vbif_init - initializes the vbif driver for the passed interface idx + * @idx: Interface index for which driver object is required + * @addr: Mapped register io address of MDSS + * @m: Pointer to mdss catalog data + */ +struct dpu_hw_vbif *dpu_hw_vbif_init(enum dpu_vbif idx, + void __iomem *addr, + const struct dpu_mdss_cfg *m); + +void dpu_hw_vbif_destroy(struct dpu_hw_vbif *vbif); + +#endif /*_DPU_HW_VBIF_H */ diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_hwio.h b/drivers/gpu/drm/msm/disp/dpu1/dpu_hwio.h new file mode 100644 index 000000000000..5b2bc9b65b15 --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_hwio.h @@ -0,0 +1,56 @@ +/* Copyright (c) 2015-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#ifndef _DPU_HWIO_H +#define _DPU_HWIO_H + +#include "dpu_hw_util.h" + +/** + * MDP TOP block Register and bit fields and defines + */ +#define DISP_INTF_SEL 0x004 +#define INTR_EN 0x010 +#define INTR_STATUS 0x014 +#define INTR_CLEAR 0x018 +#define INTR2_EN 0x008 +#define INTR2_STATUS 0x00c +#define INTR2_CLEAR 0x02c +#define HIST_INTR_EN 0x01c +#define HIST_INTR_STATUS 0x020 +#define HIST_INTR_CLEAR 0x024 +#define INTF_INTR_EN 0x1C0 +#define INTF_INTR_STATUS 0x1C4 +#define INTF_INTR_CLEAR 0x1C8 +#define SPLIT_DISPLAY_EN 0x2F4 +#define SPLIT_DISPLAY_UPPER_PIPE_CTRL 0x2F8 +#define DSPP_IGC_COLOR0_RAM_LUTN 0x300 +#define DSPP_IGC_COLOR1_RAM_LUTN 0x304 +#define DSPP_IGC_COLOR2_RAM_LUTN 0x308 +#define HW_EVENTS_CTL 0x37C +#define CLK_CTRL3 0x3A8 +#define CLK_STATUS3 0x3AC +#define CLK_CTRL4 0x3B0 +#define CLK_STATUS4 0x3B4 +#define CLK_CTRL5 0x3B8 +#define CLK_STATUS5 0x3BC +#define CLK_CTRL7 0x3D0 +#define CLK_STATUS7 0x3D4 +#define SPLIT_DISPLAY_LOWER_PIPE_CTRL 0x3F0 +#define SPLIT_DISPLAY_TE_LINE_INTERVAL 0x3F4 +#define INTF_SW_RESET_MASK 0x3FC +#define HDMI_DP_CORE_SELECT 0x408 +#define MDP_OUT_CTL_0 0x410 +#define MDP_VSYNC_SEL 0x414 +#define DCE_SEL 0x450 + +#endif /*_DPU_HWIO_H */ diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_io_util.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_io_util.c new file mode 100644 index 000000000000..790d39f816dc --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_io_util.c @@ -0,0 +1,203 @@ +/* Copyright (c) 2012-2015, 2017-2018, The Linux Foundation. + * All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#include <linux/clk.h> +#include <linux/clk/clk-conf.h> +#include <linux/err.h> +#include <linux/delay.h> + +#include "dpu_io_util.h" + +void msm_dss_put_clk(struct dss_clk *clk_arry, int num_clk) +{ + int i; + + for (i = num_clk - 1; i >= 0; i--) { + if (clk_arry[i].clk) + clk_put(clk_arry[i].clk); + clk_arry[i].clk = NULL; + } +} + +int msm_dss_get_clk(struct device *dev, struct dss_clk *clk_arry, int num_clk) +{ + int i, rc = 0; + + for (i = 0; i < num_clk; i++) { + clk_arry[i].clk = clk_get(dev, clk_arry[i].clk_name); + rc = PTR_ERR_OR_ZERO(clk_arry[i].clk); + if (rc) { + DEV_ERR("%pS->%s: '%s' get failed. rc=%d\n", + __builtin_return_address(0), __func__, + clk_arry[i].clk_name, rc); + goto error; + } + } + + return rc; + +error: + for (i--; i >= 0; i--) { + if (clk_arry[i].clk) + clk_put(clk_arry[i].clk); + clk_arry[i].clk = NULL; + } + + return rc; +} + +int msm_dss_clk_set_rate(struct dss_clk *clk_arry, int num_clk) +{ + int i, rc = 0; + + for (i = 0; i < num_clk; i++) { + if (clk_arry[i].clk) { + if (clk_arry[i].type != DSS_CLK_AHB) { + DEV_DBG("%pS->%s: '%s' rate %ld\n", + __builtin_return_address(0), __func__, + clk_arry[i].clk_name, + clk_arry[i].rate); + rc = clk_set_rate(clk_arry[i].clk, + clk_arry[i].rate); + if (rc) { + DEV_ERR("%pS->%s: %s failed. rc=%d\n", + __builtin_return_address(0), + __func__, + clk_arry[i].clk_name, rc); + break; + } + } + } else { + DEV_ERR("%pS->%s: '%s' is not available\n", + __builtin_return_address(0), __func__, + clk_arry[i].clk_name); + rc = -EPERM; + break; + } + } + + return rc; +} + +int msm_dss_enable_clk(struct dss_clk *clk_arry, int num_clk, int enable) +{ + int i, rc = 0; + + if (enable) { + for (i = 0; i < num_clk; i++) { + DEV_DBG("%pS->%s: enable '%s'\n", + __builtin_return_address(0), __func__, + clk_arry[i].clk_name); + if (clk_arry[i].clk) { + rc = clk_prepare_enable(clk_arry[i].clk); + if (rc) + DEV_ERR("%pS->%s: %s en fail. rc=%d\n", + __builtin_return_address(0), + __func__, + clk_arry[i].clk_name, rc); + } else { + DEV_ERR("%pS->%s: '%s' is not available\n", + __builtin_return_address(0), __func__, + clk_arry[i].clk_name); + rc = -EPERM; + } + + if (rc) { + msm_dss_enable_clk(&clk_arry[i], + i, false); + break; + } + } + } else { + for (i = num_clk - 1; i >= 0; i--) { + DEV_DBG("%pS->%s: disable '%s'\n", + __builtin_return_address(0), __func__, + clk_arry[i].clk_name); + + if (clk_arry[i].clk) + clk_disable_unprepare(clk_arry[i].clk); + else + DEV_ERR("%pS->%s: '%s' is not available\n", + __builtin_return_address(0), __func__, + clk_arry[i].clk_name); + } + } + + return rc; +} + +int msm_dss_parse_clock(struct platform_device *pdev, + struct dss_module_power *mp) +{ + u32 i, rc = 0; + const char *clock_name; + int num_clk = 0; + + if (!pdev || !mp) + return -EINVAL; + + mp->num_clk = 0; + num_clk = of_property_count_strings(pdev->dev.of_node, "clock-names"); + if (num_clk <= 0) { + pr_debug("clocks are not defined\n"); + return 0; + } + + mp->clk_config = devm_kzalloc(&pdev->dev, + sizeof(struct dss_clk) * num_clk, + GFP_KERNEL); + if (!mp->clk_config) + return -ENOMEM; + + for (i = 0; i < num_clk; i++) { + rc = of_property_read_string_index(pdev->dev.of_node, + "clock-names", i, + &clock_name); + if (rc) { + dev_err(&pdev->dev, "Failed to get clock name for %d\n", + i); + break; + } + strlcpy(mp->clk_config[i].clk_name, clock_name, + sizeof(mp->clk_config[i].clk_name)); + + mp->clk_config[i].type = DSS_CLK_AHB; + } + + rc = msm_dss_get_clk(&pdev->dev, mp->clk_config, num_clk); + if (rc) { + dev_err(&pdev->dev, "Failed to get clock refs %d\n", rc); + goto err; + } + + rc = of_clk_set_defaults(pdev->dev.of_node, false); + if (rc) { + dev_err(&pdev->dev, "Failed to set clock defaults %d\n", rc); + goto err; + } + + for (i = 0; i < num_clk; i++) { + u32 rate = clk_get_rate(mp->clk_config[i].clk); + if (!rate) + continue; + mp->clk_config[i].rate = rate; + mp->clk_config[i].type = DSS_CLK_PCLK; + } + + mp->num_clk = num_clk; + return 0; + +err: + msm_dss_put_clk(mp->clk_config, num_clk); + return rc; +} diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_io_util.h b/drivers/gpu/drm/msm/disp/dpu1/dpu_io_util.h new file mode 100644 index 000000000000..bc07381d7429 --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_io_util.h @@ -0,0 +1,57 @@ +/* Copyright (c) 2012, 2017-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#ifndef __DPU_IO_UTIL_H__ +#define __DPU_IO_UTIL_H__ + +#include <linux/gpio.h> +#include <linux/platform_device.h> +#include <linux/types.h> + +#define DEV_DBG(fmt, args...) pr_debug(fmt, ##args) +#define DEV_INFO(fmt, args...) pr_info(fmt, ##args) +#define DEV_WARN(fmt, args...) pr_warn(fmt, ##args) +#define DEV_ERR(fmt, args...) pr_err(fmt, ##args) + +struct dss_gpio { + unsigned int gpio; + unsigned int value; + char gpio_name[32]; +}; + +enum dss_clk_type { + DSS_CLK_AHB, /* no set rate. rate controlled through rpm */ + DSS_CLK_PCLK, +}; + +struct dss_clk { + struct clk *clk; /* clk handle */ + char clk_name[32]; + enum dss_clk_type type; + unsigned long rate; + unsigned long max_rate; +}; + +struct dss_module_power { + unsigned int num_gpio; + struct dss_gpio *gpio_config; + unsigned int num_clk; + struct dss_clk *clk_config; +}; + +int msm_dss_get_clk(struct device *dev, struct dss_clk *clk_arry, int num_clk); +void msm_dss_put_clk(struct dss_clk *clk_arry, int num_clk); +int msm_dss_clk_set_rate(struct dss_clk *clk_arry, int num_clk); +int msm_dss_enable_clk(struct dss_clk *clk_arry, int num_clk, int enable); +int msm_dss_parse_clock(struct platform_device *pdev, + struct dss_module_power *mp); +#endif /* __DPU_IO_UTIL_H__ */ diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_irq.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_irq.c new file mode 100644 index 000000000000..d5e6ce0140cf --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_irq.c @@ -0,0 +1,66 @@ +/* Copyright (c) 2015-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#define pr_fmt(fmt) "[drm:%s:%d] " fmt, __func__, __LINE__ + +#include <linux/irqdomain.h> +#include <linux/irq.h> +#include <linux/kthread.h> + +#include "dpu_irq.h" +#include "dpu_core_irq.h" + +irqreturn_t dpu_irq(struct msm_kms *kms) +{ + struct dpu_kms *dpu_kms = to_dpu_kms(kms); + + return dpu_core_irq(dpu_kms); +} + +void dpu_irq_preinstall(struct msm_kms *kms) +{ + struct dpu_kms *dpu_kms = to_dpu_kms(kms); + + if (!dpu_kms->dev || !dpu_kms->dev->dev) { + pr_err("invalid device handles\n"); + return; + } + + dpu_core_irq_preinstall(dpu_kms); +} + +int dpu_irq_postinstall(struct msm_kms *kms) +{ + struct dpu_kms *dpu_kms = to_dpu_kms(kms); + int rc; + + if (!kms) { + DPU_ERROR("invalid parameters\n"); + return -EINVAL; + } + + rc = dpu_core_irq_postinstall(dpu_kms); + + return rc; +} + +void dpu_irq_uninstall(struct msm_kms *kms) +{ + struct dpu_kms *dpu_kms = to_dpu_kms(kms); + + if (!kms) { + DPU_ERROR("invalid parameters\n"); + return; + } + + dpu_core_irq_uninstall(dpu_kms); +} diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_irq.h b/drivers/gpu/drm/msm/disp/dpu1/dpu_irq.h new file mode 100644 index 000000000000..3e147f7176e2 --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_irq.h @@ -0,0 +1,59 @@ +/* Copyright (c) 2015-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#ifndef __DPU_IRQ_H__ +#define __DPU_IRQ_H__ + +#include <linux/kernel.h> +#include <linux/irqdomain.h> + +#include "msm_kms.h" + +/** + * dpu_irq_controller - define MDSS level interrupt controller context + * @enabled_mask: enable status of MDSS level interrupt + * @domain: interrupt domain of this controller + */ +struct dpu_irq_controller { + unsigned long enabled_mask; + struct irq_domain *domain; +}; + +/** + * dpu_irq_preinstall - perform pre-installation of MDSS IRQ handler + * @kms: pointer to kms context + * @return: none + */ +void dpu_irq_preinstall(struct msm_kms *kms); + +/** + * dpu_irq_postinstall - perform post-installation of MDSS IRQ handler + * @kms: pointer to kms context + * @return: 0 if success; error code otherwise + */ +int dpu_irq_postinstall(struct msm_kms *kms); + +/** + * dpu_irq_uninstall - uninstall MDSS IRQ handler + * @drm_dev: pointer to kms context + * @return: none + */ +void dpu_irq_uninstall(struct msm_kms *kms); + +/** + * dpu_irq - MDSS level IRQ handler + * @kms: pointer to kms context + * @return: interrupt handling status + */ +irqreturn_t dpu_irq(struct msm_kms *kms); + +#endif /* __DPU_IRQ_H__ */ diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_kms.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_kms.c new file mode 100644 index 000000000000..7dd6bd2d6d37 --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_kms.c @@ -0,0 +1,1345 @@ +/* + * Copyright (c) 2014-2018, The Linux Foundation. All rights reserved. + * Copyright (C) 2013 Red Hat + * Author: Rob Clark <robdclark@gmail.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published by + * the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program. If not, see <http://www.gnu.org/licenses/>. + */ + +#define pr_fmt(fmt) "[drm:%s:%d] " fmt, __func__, __LINE__ + +#include <drm/drm_crtc.h> +#include <linux/debugfs.h> +#include <linux/of_irq.h> +#include <linux/dma-buf.h> + +#include "msm_drv.h" +#include "msm_mmu.h" +#include "msm_gem.h" + +#include "dpu_kms.h" +#include "dpu_core_irq.h" +#include "dpu_formats.h" +#include "dpu_hw_vbif.h" +#include "dpu_vbif.h" +#include "dpu_encoder.h" +#include "dpu_plane.h" +#include "dpu_crtc.h" + +#define CREATE_TRACE_POINTS +#include "dpu_trace.h" + +static const char * const iommu_ports[] = { + "mdp_0", +}; + +/* + * To enable overall DRM driver logging + * # echo 0x2 > /sys/module/drm/parameters/debug + * + * To enable DRM driver h/w logging + * # echo <mask> > /sys/kernel/debug/dri/0/debug/hw_log_mask + * + * See dpu_hw_mdss.h for h/w logging mask definitions (search for DPU_DBG_MASK_) + */ +#define DPU_DEBUGFS_DIR "msm_dpu" +#define DPU_DEBUGFS_HWMASKNAME "hw_log_mask" + +static int dpu_kms_hw_init(struct msm_kms *kms); +static int _dpu_kms_mmu_destroy(struct dpu_kms *dpu_kms); + +static unsigned long dpu_iomap_size(struct platform_device *pdev, + const char *name) +{ + struct resource *res; + + res = platform_get_resource_byname(pdev, IORESOURCE_MEM, name); + if (!res) { + DRM_ERROR("failed to get memory resource: %s\n", name); + return 0; + } + + return resource_size(res); +} + +#ifdef CONFIG_DEBUG_FS +static int _dpu_danger_signal_status(struct seq_file *s, + bool danger_status) +{ + struct dpu_kms *kms = (struct dpu_kms *)s->private; + struct msm_drm_private *priv; + struct dpu_danger_safe_status status; + int i; + + if (!kms || !kms->dev || !kms->dev->dev_private || !kms->hw_mdp) { + DPU_ERROR("invalid arg(s)\n"); + return 0; + } + + priv = kms->dev->dev_private; + memset(&status, 0, sizeof(struct dpu_danger_safe_status)); + + pm_runtime_get_sync(&kms->pdev->dev); + if (danger_status) { + seq_puts(s, "\nDanger signal status:\n"); + if (kms->hw_mdp->ops.get_danger_status) + kms->hw_mdp->ops.get_danger_status(kms->hw_mdp, + &status); + } else { + seq_puts(s, "\nSafe signal status:\n"); + if (kms->hw_mdp->ops.get_danger_status) + kms->hw_mdp->ops.get_danger_status(kms->hw_mdp, + &status); + } + pm_runtime_put_sync(&kms->pdev->dev); + + seq_printf(s, "MDP : 0x%x\n", status.mdp); + + for (i = SSPP_VIG0; i < SSPP_MAX; i++) + seq_printf(s, "SSPP%d : 0x%x \t", i - SSPP_VIG0, + status.sspp[i]); + seq_puts(s, "\n"); + + return 0; +} + +#define DEFINE_DPU_DEBUGFS_SEQ_FOPS(__prefix) \ +static int __prefix ## _open(struct inode *inode, struct file *file) \ +{ \ + return single_open(file, __prefix ## _show, inode->i_private); \ +} \ +static const struct file_operations __prefix ## _fops = { \ + .owner = THIS_MODULE, \ + .open = __prefix ## _open, \ + .release = single_release, \ + .read = seq_read, \ + .llseek = seq_lseek, \ +} + +static int dpu_debugfs_danger_stats_show(struct seq_file *s, void *v) +{ + return _dpu_danger_signal_status(s, true); +} +DEFINE_DPU_DEBUGFS_SEQ_FOPS(dpu_debugfs_danger_stats); + +static int dpu_debugfs_safe_stats_show(struct seq_file *s, void *v) +{ + return _dpu_danger_signal_status(s, false); +} +DEFINE_DPU_DEBUGFS_SEQ_FOPS(dpu_debugfs_safe_stats); + +static void dpu_debugfs_danger_destroy(struct dpu_kms *dpu_kms) +{ + debugfs_remove_recursive(dpu_kms->debugfs_danger); + dpu_kms->debugfs_danger = NULL; +} + +static int dpu_debugfs_danger_init(struct dpu_kms *dpu_kms, + struct dentry *parent) +{ + dpu_kms->debugfs_danger = debugfs_create_dir("danger", + parent); + if (!dpu_kms->debugfs_danger) { + DPU_ERROR("failed to create danger debugfs\n"); + return -EINVAL; + } + + debugfs_create_file("danger_status", 0600, dpu_kms->debugfs_danger, + dpu_kms, &dpu_debugfs_danger_stats_fops); + debugfs_create_file("safe_status", 0600, dpu_kms->debugfs_danger, + dpu_kms, &dpu_debugfs_safe_stats_fops); + + return 0; +} + +static int _dpu_debugfs_show_regset32(struct seq_file *s, void *data) +{ + struct dpu_debugfs_regset32 *regset; + struct dpu_kms *dpu_kms; + struct drm_device *dev; + struct msm_drm_private *priv; + void __iomem *base; + uint32_t i, addr; + + if (!s || !s->private) + return 0; + + regset = s->private; + + dpu_kms = regset->dpu_kms; + if (!dpu_kms || !dpu_kms->mmio) + return 0; + + dev = dpu_kms->dev; + if (!dev) + return 0; + + priv = dev->dev_private; + if (!priv) + return 0; + + base = dpu_kms->mmio + regset->offset; + + /* insert padding spaces, if needed */ + if (regset->offset & 0xF) { + seq_printf(s, "[%x]", regset->offset & ~0xF); + for (i = 0; i < (regset->offset & 0xF); i += 4) + seq_puts(s, " "); + } + + pm_runtime_get_sync(&dpu_kms->pdev->dev); + + /* main register output */ + for (i = 0; i < regset->blk_len; i += 4) { + addr = regset->offset + i; + if ((addr & 0xF) == 0x0) + seq_printf(s, i ? "\n[%x]" : "[%x]", addr); + seq_printf(s, " %08x", readl_relaxed(base + i)); + } + seq_puts(s, "\n"); + pm_runtime_put_sync(&dpu_kms->pdev->dev); + + return 0; +} + +static int dpu_debugfs_open_regset32(struct inode *inode, + struct file *file) +{ + return single_open(file, _dpu_debugfs_show_regset32, inode->i_private); +} + +static const struct file_operations dpu_fops_regset32 = { + .open = dpu_debugfs_open_regset32, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; + +void dpu_debugfs_setup_regset32(struct dpu_debugfs_regset32 *regset, + uint32_t offset, uint32_t length, struct dpu_kms *dpu_kms) +{ + if (regset) { + regset->offset = offset; + regset->blk_len = length; + regset->dpu_kms = dpu_kms; + } +} + +void *dpu_debugfs_create_regset32(const char *name, umode_t mode, + void *parent, struct dpu_debugfs_regset32 *regset) +{ + if (!name || !regset || !regset->dpu_kms || !regset->blk_len) + return NULL; + + /* make sure offset is a multiple of 4 */ + regset->offset = round_down(regset->offset, 4); + + return debugfs_create_file(name, mode, parent, + regset, &dpu_fops_regset32); +} + +static int _dpu_debugfs_init(struct dpu_kms *dpu_kms) +{ + void *p; + int rc; + + p = dpu_hw_util_get_log_mask_ptr(); + + if (!dpu_kms || !p) + return -EINVAL; + + dpu_kms->debugfs_root = debugfs_create_dir("debug", + dpu_kms->dev->primary->debugfs_root); + if (IS_ERR_OR_NULL(dpu_kms->debugfs_root)) { + DRM_ERROR("debugfs create_dir failed %ld\n", + PTR_ERR(dpu_kms->debugfs_root)); + return PTR_ERR(dpu_kms->debugfs_root); + } + + rc = dpu_dbg_debugfs_register(dpu_kms->debugfs_root); + if (rc) { + DRM_ERROR("failed to reg dpu dbg debugfs: %d\n", rc); + return rc; + } + + /* allow root to be NULL */ + debugfs_create_x32(DPU_DEBUGFS_HWMASKNAME, 0600, dpu_kms->debugfs_root, p); + + (void) dpu_debugfs_danger_init(dpu_kms, dpu_kms->debugfs_root); + (void) dpu_debugfs_vbif_init(dpu_kms, dpu_kms->debugfs_root); + (void) dpu_debugfs_core_irq_init(dpu_kms, dpu_kms->debugfs_root); + + rc = dpu_core_perf_debugfs_init(&dpu_kms->perf, dpu_kms->debugfs_root); + if (rc) { + DPU_ERROR("failed to init perf %d\n", rc); + return rc; + } + + return 0; +} + +static void _dpu_debugfs_destroy(struct dpu_kms *dpu_kms) +{ + /* don't need to NULL check debugfs_root */ + if (dpu_kms) { + dpu_debugfs_vbif_destroy(dpu_kms); + dpu_debugfs_danger_destroy(dpu_kms); + dpu_debugfs_core_irq_destroy(dpu_kms); + debugfs_remove_recursive(dpu_kms->debugfs_root); + } +} +#else +static void _dpu_debugfs_destroy(struct dpu_kms *dpu_kms) +{ +} +#endif + +static int dpu_kms_enable_vblank(struct msm_kms *kms, struct drm_crtc *crtc) +{ + return dpu_crtc_vblank(crtc, true); +} + +static void dpu_kms_disable_vblank(struct msm_kms *kms, struct drm_crtc *crtc) +{ + dpu_crtc_vblank(crtc, false); +} + +static void dpu_kms_prepare_commit(struct msm_kms *kms, + struct drm_atomic_state *state) +{ + struct dpu_kms *dpu_kms; + struct msm_drm_private *priv; + struct drm_device *dev; + struct drm_encoder *encoder; + + if (!kms) + return; + dpu_kms = to_dpu_kms(kms); + dev = dpu_kms->dev; + + if (!dev || !dev->dev_private) + return; + priv = dev->dev_private; + pm_runtime_get_sync(&dpu_kms->pdev->dev); + + list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) + if (encoder->crtc != NULL) + dpu_encoder_prepare_commit(encoder); +} + +/* + * Override the encoder enable since we need to setup the inline rotator and do + * some crtc magic before enabling any bridge that might be present. + */ +void dpu_kms_encoder_enable(struct drm_encoder *encoder) +{ + const struct drm_encoder_helper_funcs *funcs = encoder->helper_private; + struct drm_crtc *crtc = encoder->crtc; + + /* Forward this enable call to the commit hook */ + if (funcs && funcs->commit) + funcs->commit(encoder); + + if (crtc && crtc->state->active) { + trace_dpu_kms_enc_enable(DRMID(crtc)); + dpu_crtc_commit_kickoff(crtc); + } +} + +static void dpu_kms_commit(struct msm_kms *kms, struct drm_atomic_state *state) +{ + struct drm_crtc *crtc; + struct drm_crtc_state *crtc_state; + int i; + + for_each_new_crtc_in_state(state, crtc, crtc_state, i) { + /* If modeset is required, kickoff is run in encoder_enable */ + if (drm_atomic_crtc_needs_modeset(crtc_state)) + continue; + + if (crtc->state->active) { + trace_dpu_kms_commit(DRMID(crtc)); + dpu_crtc_commit_kickoff(crtc); + } + } +} + +static void dpu_kms_complete_commit(struct msm_kms *kms, + struct drm_atomic_state *old_state) +{ + struct dpu_kms *dpu_kms; + struct msm_drm_private *priv; + struct drm_crtc *crtc; + struct drm_crtc_state *old_crtc_state; + int i; + + if (!kms || !old_state) + return; + dpu_kms = to_dpu_kms(kms); + + if (!dpu_kms->dev || !dpu_kms->dev->dev_private) + return; + priv = dpu_kms->dev->dev_private; + + DPU_ATRACE_BEGIN("kms_complete_commit"); + + for_each_old_crtc_in_state(old_state, crtc, old_crtc_state, i) + dpu_crtc_complete_commit(crtc, old_crtc_state); + + pm_runtime_put_sync(&dpu_kms->pdev->dev); + + DPU_ATRACE_END("kms_complete_commit"); +} + +static void dpu_kms_wait_for_commit_done(struct msm_kms *kms, + struct drm_crtc *crtc) +{ + struct drm_encoder *encoder; + struct drm_device *dev; + int ret; + + if (!kms || !crtc || !crtc->state) { + DPU_ERROR("invalid params\n"); + return; + } + + dev = crtc->dev; + + if (!crtc->state->enable) { + DPU_DEBUG("[crtc:%d] not enable\n", crtc->base.id); + return; + } + + if (!crtc->state->active) { + DPU_DEBUG("[crtc:%d] not active\n", crtc->base.id); + return; + } + + list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) { + if (encoder->crtc != crtc) + continue; + /* + * Wait for post-flush if necessary to delay before + * plane_cleanup. For example, wait for vsync in case of video + * mode panels. This may be a no-op for command mode panels. + */ + trace_dpu_kms_wait_for_commit_done(DRMID(crtc)); + ret = dpu_encoder_wait_for_event(encoder, MSM_ENC_COMMIT_DONE); + if (ret && ret != -EWOULDBLOCK) { + DPU_ERROR("wait for commit done returned %d\n", ret); + break; + } + } +} + +static void _dpu_kms_initialize_dsi(struct drm_device *dev, + struct msm_drm_private *priv, + struct dpu_kms *dpu_kms) +{ + struct drm_encoder *encoder = NULL; + int i, rc; + + /*TODO: Support two independent DSI connectors */ + encoder = dpu_encoder_init(dev, DRM_MODE_CONNECTOR_DSI); + if (IS_ERR_OR_NULL(encoder)) { + DPU_ERROR("encoder init failed for dsi display\n"); + return; + } + + priv->encoders[priv->num_encoders++] = encoder; + + for (i = 0; i < ARRAY_SIZE(priv->dsi); i++) { + if (!priv->dsi[i]) { + DPU_DEBUG("invalid msm_dsi for ctrl %d\n", i); + return; + } + + rc = msm_dsi_modeset_init(priv->dsi[i], dev, encoder); + if (rc) { + DPU_ERROR("modeset_init failed for dsi[%d], rc = %d\n", + i, rc); + continue; + } + } +} + +/** + * _dpu_kms_setup_displays - create encoders, bridges and connectors + * for underlying displays + * @dev: Pointer to drm device structure + * @priv: Pointer to private drm device data + * @dpu_kms: Pointer to dpu kms structure + * Returns: Zero on success + */ +static void _dpu_kms_setup_displays(struct drm_device *dev, + struct msm_drm_private *priv, + struct dpu_kms *dpu_kms) +{ + _dpu_kms_initialize_dsi(dev, priv, dpu_kms); + + /** + * Extend this function to initialize other + * types of displays + */ +} + +static void _dpu_kms_drm_obj_destroy(struct dpu_kms *dpu_kms) +{ + struct msm_drm_private *priv; + int i; + + if (!dpu_kms) { + DPU_ERROR("invalid dpu_kms\n"); + return; + } else if (!dpu_kms->dev) { + DPU_ERROR("invalid dev\n"); + return; + } else if (!dpu_kms->dev->dev_private) { + DPU_ERROR("invalid dev_private\n"); + return; + } + priv = dpu_kms->dev->dev_private; + + for (i = 0; i < priv->num_crtcs; i++) + priv->crtcs[i]->funcs->destroy(priv->crtcs[i]); + priv->num_crtcs = 0; + + for (i = 0; i < priv->num_planes; i++) + priv->planes[i]->funcs->destroy(priv->planes[i]); + priv->num_planes = 0; + + for (i = 0; i < priv->num_connectors; i++) + priv->connectors[i]->funcs->destroy(priv->connectors[i]); + priv->num_connectors = 0; + + for (i = 0; i < priv->num_encoders; i++) + priv->encoders[i]->funcs->destroy(priv->encoders[i]); + priv->num_encoders = 0; +} + +static int _dpu_kms_drm_obj_init(struct dpu_kms *dpu_kms) +{ + struct drm_device *dev; + struct drm_plane *primary_planes[MAX_PLANES], *plane; + struct drm_crtc *crtc; + + struct msm_drm_private *priv; + struct dpu_mdss_cfg *catalog; + + int primary_planes_idx = 0, i, ret; + int max_crtc_count; + + if (!dpu_kms || !dpu_kms->dev || !dpu_kms->dev->dev) { + DPU_ERROR("invalid dpu_kms\n"); + return -EINVAL; + } + + dev = dpu_kms->dev; + priv = dev->dev_private; + catalog = dpu_kms->catalog; + + /* + * Create encoder and query display drivers to create + * bridges and connectors + */ + _dpu_kms_setup_displays(dev, priv, dpu_kms); + + max_crtc_count = min(catalog->mixer_count, priv->num_encoders); + + /* Create the planes */ + for (i = 0; i < catalog->sspp_count; i++) { + bool primary = true; + + if (catalog->sspp[i].features & BIT(DPU_SSPP_CURSOR) + || primary_planes_idx >= max_crtc_count) + primary = false; + + plane = dpu_plane_init(dev, catalog->sspp[i].id, primary, + (1UL << max_crtc_count) - 1, 0); + if (IS_ERR(plane)) { + DPU_ERROR("dpu_plane_init failed\n"); + ret = PTR_ERR(plane); + goto fail; + } + priv->planes[priv->num_planes++] = plane; + + if (primary) + primary_planes[primary_planes_idx++] = plane; + } + + max_crtc_count = min(max_crtc_count, primary_planes_idx); + + /* Create one CRTC per encoder */ + for (i = 0; i < max_crtc_count; i++) { + crtc = dpu_crtc_init(dev, primary_planes[i]); + if (IS_ERR(crtc)) { + ret = PTR_ERR(crtc); + goto fail; + } + priv->crtcs[priv->num_crtcs++] = crtc; + } + + /* All CRTCs are compatible with all encoders */ + for (i = 0; i < priv->num_encoders; i++) + priv->encoders[i]->possible_crtcs = (1 << priv->num_crtcs) - 1; + + return 0; +fail: + _dpu_kms_drm_obj_destroy(dpu_kms); + return ret; +} + +#ifdef CONFIG_DEBUG_FS +static int dpu_kms_debugfs_init(struct msm_kms *kms, struct drm_minor *minor) +{ + struct dpu_kms *dpu_kms = to_dpu_kms(kms); + struct drm_device *dev; + int rc; + + if (!dpu_kms || !dpu_kms->dev || !dpu_kms->dev->dev) { + DPU_ERROR("invalid dpu_kms\n"); + return -EINVAL; + } + + dev = dpu_kms->dev; + + rc = _dpu_debugfs_init(dpu_kms); + if (rc) + DPU_ERROR("dpu_debugfs init failed: %d\n", rc); + + return rc; +} +#endif + +static long dpu_kms_round_pixclk(struct msm_kms *kms, unsigned long rate, + struct drm_encoder *encoder) +{ + return rate; +} + +static void _dpu_kms_hw_destroy(struct dpu_kms *dpu_kms) +{ + struct drm_device *dev; + int i; + + dev = dpu_kms->dev; + if (!dev) + return; + + if (dpu_kms->hw_intr) + dpu_hw_intr_destroy(dpu_kms->hw_intr); + dpu_kms->hw_intr = NULL; + + if (dpu_kms->power_event) + dpu_power_handle_unregister_event( + &dpu_kms->phandle, dpu_kms->power_event); + + /* safe to call these more than once during shutdown */ + _dpu_debugfs_destroy(dpu_kms); + _dpu_kms_mmu_destroy(dpu_kms); + + if (dpu_kms->catalog) { + for (i = 0; i < dpu_kms->catalog->vbif_count; i++) { + u32 vbif_idx = dpu_kms->catalog->vbif[i].id; + + if ((vbif_idx < VBIF_MAX) && dpu_kms->hw_vbif[vbif_idx]) + dpu_hw_vbif_destroy(dpu_kms->hw_vbif[vbif_idx]); + } + } + + if (dpu_kms->rm_init) + dpu_rm_destroy(&dpu_kms->rm); + dpu_kms->rm_init = false; + + if (dpu_kms->catalog) + dpu_hw_catalog_deinit(dpu_kms->catalog); + dpu_kms->catalog = NULL; + + if (dpu_kms->core_client) + dpu_power_client_destroy(&dpu_kms->phandle, + dpu_kms->core_client); + dpu_kms->core_client = NULL; + + if (dpu_kms->vbif[VBIF_NRT]) + devm_iounmap(&dpu_kms->pdev->dev, dpu_kms->vbif[VBIF_NRT]); + dpu_kms->vbif[VBIF_NRT] = NULL; + + if (dpu_kms->vbif[VBIF_RT]) + devm_iounmap(&dpu_kms->pdev->dev, dpu_kms->vbif[VBIF_RT]); + dpu_kms->vbif[VBIF_RT] = NULL; + + if (dpu_kms->mmio) + devm_iounmap(&dpu_kms->pdev->dev, dpu_kms->mmio); + dpu_kms->mmio = NULL; +} + +static void dpu_kms_destroy(struct msm_kms *kms) +{ + struct dpu_kms *dpu_kms; + + if (!kms) { + DPU_ERROR("invalid kms\n"); + return; + } + + dpu_kms = to_dpu_kms(kms); + + dpu_dbg_destroy(); + _dpu_kms_hw_destroy(dpu_kms); +} + +static int dpu_kms_pm_suspend(struct device *dev) +{ + struct drm_device *ddev; + struct drm_modeset_acquire_ctx ctx; + struct drm_atomic_state *state; + struct dpu_kms *dpu_kms; + int ret = 0, num_crtcs = 0; + + if (!dev) + return -EINVAL; + + ddev = dev_get_drvdata(dev); + if (!ddev || !ddev_to_msm_kms(ddev)) + return -EINVAL; + + dpu_kms = to_dpu_kms(ddev_to_msm_kms(ddev)); + + /* disable hot-plug polling */ + drm_kms_helper_poll_disable(ddev); + + /* acquire modeset lock(s) */ + drm_modeset_acquire_init(&ctx, 0); + +retry: + DPU_ATRACE_BEGIN("kms_pm_suspend"); + + ret = drm_modeset_lock_all_ctx(ddev, &ctx); + if (ret) + goto unlock; + + /* save current state for resume */ + if (dpu_kms->suspend_state) + drm_atomic_state_put(dpu_kms->suspend_state); + dpu_kms->suspend_state = drm_atomic_helper_duplicate_state(ddev, &ctx); + if (IS_ERR_OR_NULL(dpu_kms->suspend_state)) { + DRM_ERROR("failed to back up suspend state\n"); + dpu_kms->suspend_state = NULL; + goto unlock; + } + + /* create atomic state to disable all CRTCs */ + state = drm_atomic_state_alloc(ddev); + if (IS_ERR_OR_NULL(state)) { + DRM_ERROR("failed to allocate crtc disable state\n"); + goto unlock; + } + + state->acquire_ctx = &ctx; + + /* check for nothing to do */ + if (num_crtcs == 0) { + DRM_DEBUG("all crtcs are already in the off state\n"); + drm_atomic_state_put(state); + goto suspended; + } + + /* commit the "disable all" state */ + ret = drm_atomic_commit(state); + if (ret < 0) { + DRM_ERROR("failed to disable crtcs, %d\n", ret); + drm_atomic_state_put(state); + goto unlock; + } + +suspended: + dpu_kms->suspend_block = true; + +unlock: + if (ret == -EDEADLK) { + drm_modeset_backoff(&ctx); + goto retry; + } + drm_modeset_drop_locks(&ctx); + drm_modeset_acquire_fini(&ctx); + + DPU_ATRACE_END("kms_pm_suspend"); + return 0; +} + +static int dpu_kms_pm_resume(struct device *dev) +{ + struct drm_device *ddev; + struct dpu_kms *dpu_kms; + int ret; + + if (!dev) + return -EINVAL; + + ddev = dev_get_drvdata(dev); + if (!ddev || !ddev_to_msm_kms(ddev)) + return -EINVAL; + + dpu_kms = to_dpu_kms(ddev_to_msm_kms(ddev)); + + DPU_ATRACE_BEGIN("kms_pm_resume"); + + drm_mode_config_reset(ddev); + + drm_modeset_lock_all(ddev); + + dpu_kms->suspend_block = false; + + if (dpu_kms->suspend_state) { + dpu_kms->suspend_state->acquire_ctx = + ddev->mode_config.acquire_ctx; + ret = drm_atomic_commit(dpu_kms->suspend_state); + if (ret < 0) { + DRM_ERROR("failed to restore state, %d\n", ret); + drm_atomic_state_put(dpu_kms->suspend_state); + } + dpu_kms->suspend_state = NULL; + } + drm_modeset_unlock_all(ddev); + + /* enable hot-plug polling */ + drm_kms_helper_poll_enable(ddev); + + DPU_ATRACE_END("kms_pm_resume"); + return 0; +} + +static void _dpu_kms_set_encoder_mode(struct msm_kms *kms, + struct drm_encoder *encoder, + bool cmd_mode) +{ + struct msm_display_info info; + struct msm_drm_private *priv = encoder->dev->dev_private; + int i, rc = 0; + + memset(&info, 0, sizeof(info)); + + info.intf_type = encoder->encoder_type; + info.capabilities = cmd_mode ? MSM_DISPLAY_CAP_CMD_MODE : + MSM_DISPLAY_CAP_VID_MODE; + + /* TODO: No support for DSI swap */ + for (i = 0; i < ARRAY_SIZE(priv->dsi); i++) { + if (priv->dsi[i]) { + info.h_tile_instance[info.num_of_h_tiles] = i; + info.num_of_h_tiles++; + } + } + + rc = dpu_encoder_setup(encoder->dev, encoder, &info); + if (rc) + DPU_ERROR("failed to setup DPU encoder %d: rc:%d\n", + encoder->base.id, rc); +} + +static const struct msm_kms_funcs kms_funcs = { + .hw_init = dpu_kms_hw_init, + .irq_preinstall = dpu_irq_preinstall, + .irq_postinstall = dpu_irq_postinstall, + .irq_uninstall = dpu_irq_uninstall, + .irq = dpu_irq, + .prepare_commit = dpu_kms_prepare_commit, + .commit = dpu_kms_commit, + .complete_commit = dpu_kms_complete_commit, + .wait_for_crtc_commit_done = dpu_kms_wait_for_commit_done, + .enable_vblank = dpu_kms_enable_vblank, + .disable_vblank = dpu_kms_disable_vblank, + .check_modified_format = dpu_format_check_modified_format, + .get_format = dpu_get_msm_format, + .round_pixclk = dpu_kms_round_pixclk, + .pm_suspend = dpu_kms_pm_suspend, + .pm_resume = dpu_kms_pm_resume, + .destroy = dpu_kms_destroy, + .set_encoder_mode = _dpu_kms_set_encoder_mode, +#ifdef CONFIG_DEBUG_FS + .debugfs_init = dpu_kms_debugfs_init, +#endif +}; + +/* the caller api needs to turn on clock before calling it */ +static inline void _dpu_kms_core_hw_rev_init(struct dpu_kms *dpu_kms) +{ + dpu_kms->core_rev = readl_relaxed(dpu_kms->mmio + 0x0); +} + +static int _dpu_kms_mmu_destroy(struct dpu_kms *dpu_kms) +{ + struct msm_mmu *mmu; + + mmu = dpu_kms->base.aspace->mmu; + + mmu->funcs->detach(mmu, (const char **)iommu_ports, + ARRAY_SIZE(iommu_ports)); + msm_gem_address_space_put(dpu_kms->base.aspace); + + return 0; +} + +static int _dpu_kms_mmu_init(struct dpu_kms *dpu_kms) +{ + struct iommu_domain *domain; + struct msm_gem_address_space *aspace; + int ret; + + domain = iommu_domain_alloc(&platform_bus_type); + if (!domain) + return 0; + + aspace = msm_gem_address_space_create(dpu_kms->dev->dev, + domain, "dpu1"); + if (IS_ERR(aspace)) { + ret = PTR_ERR(aspace); + goto fail; + } + + dpu_kms->base.aspace = aspace; + + ret = aspace->mmu->funcs->attach(aspace->mmu, iommu_ports, + ARRAY_SIZE(iommu_ports)); + if (ret) { + DPU_ERROR("failed to attach iommu %d\n", ret); + msm_gem_address_space_put(aspace); + goto fail; + } + + return 0; +fail: + _dpu_kms_mmu_destroy(dpu_kms); + + return ret; +} + +static struct dss_clk *_dpu_kms_get_clk(struct dpu_kms *dpu_kms, + char *clock_name) +{ + struct dss_module_power *mp = &dpu_kms->mp; + int i; + + for (i = 0; i < mp->num_clk; i++) { + if (!strcmp(mp->clk_config[i].clk_name, clock_name)) + return &mp->clk_config[i]; + } + + return NULL; +} + +u64 dpu_kms_get_clk_rate(struct dpu_kms *dpu_kms, char *clock_name) +{ + struct dss_clk *clk; + + clk = _dpu_kms_get_clk(dpu_kms, clock_name); + if (!clk) + return -EINVAL; + + return clk_get_rate(clk->clk); +} + +static void dpu_kms_handle_power_event(u32 event_type, void *usr) +{ + struct dpu_kms *dpu_kms = usr; + + if (!dpu_kms) + return; + + if (event_type == DPU_POWER_EVENT_POST_ENABLE) + dpu_vbif_init_memtypes(dpu_kms); +} + +static int dpu_kms_hw_init(struct msm_kms *kms) +{ + struct dpu_kms *dpu_kms; + struct drm_device *dev; + struct msm_drm_private *priv; + int i, rc = -EINVAL; + + if (!kms) { + DPU_ERROR("invalid kms\n"); + goto end; + } + + dpu_kms = to_dpu_kms(kms); + dev = dpu_kms->dev; + if (!dev) { + DPU_ERROR("invalid device\n"); + goto end; + } + + rc = dpu_dbg_init(&dpu_kms->pdev->dev); + if (rc) { + DRM_ERROR("failed to init dpu dbg: %d\n", rc); + goto end; + } + + priv = dev->dev_private; + if (!priv) { + DPU_ERROR("invalid private data\n"); + goto dbg_destroy; + } + + dpu_kms->mmio = msm_ioremap(dpu_kms->pdev, "mdp", "mdp"); + if (IS_ERR(dpu_kms->mmio)) { + rc = PTR_ERR(dpu_kms->mmio); + DPU_ERROR("mdp register memory map failed: %d\n", rc); + dpu_kms->mmio = NULL; + goto error; + } + DRM_DEBUG("mapped dpu address space @%pK\n", dpu_kms->mmio); + dpu_kms->mmio_len = dpu_iomap_size(dpu_kms->pdev, "mdp"); + + dpu_kms->vbif[VBIF_RT] = msm_ioremap(dpu_kms->pdev, "vbif", "vbif"); + if (IS_ERR(dpu_kms->vbif[VBIF_RT])) { + rc = PTR_ERR(dpu_kms->vbif[VBIF_RT]); + DPU_ERROR("vbif register memory map failed: %d\n", rc); + dpu_kms->vbif[VBIF_RT] = NULL; + goto error; + } + dpu_kms->vbif_len[VBIF_RT] = dpu_iomap_size(dpu_kms->pdev, "vbif"); + dpu_kms->vbif[VBIF_NRT] = msm_ioremap(dpu_kms->pdev, "vbif_nrt", "vbif_nrt"); + if (IS_ERR(dpu_kms->vbif[VBIF_NRT])) { + dpu_kms->vbif[VBIF_NRT] = NULL; + DPU_DEBUG("VBIF NRT is not defined"); + } else { + dpu_kms->vbif_len[VBIF_NRT] = dpu_iomap_size(dpu_kms->pdev, + "vbif_nrt"); + } + + dpu_kms->reg_dma = msm_ioremap(dpu_kms->pdev, "regdma", "regdma"); + if (IS_ERR(dpu_kms->reg_dma)) { + dpu_kms->reg_dma = NULL; + DPU_DEBUG("REG_DMA is not defined"); + } else { + dpu_kms->reg_dma_len = dpu_iomap_size(dpu_kms->pdev, "regdma"); + } + + dpu_kms->core_client = dpu_power_client_create(&dpu_kms->phandle, + "core"); + if (IS_ERR_OR_NULL(dpu_kms->core_client)) { + rc = PTR_ERR(dpu_kms->core_client); + if (!dpu_kms->core_client) + rc = -EINVAL; + DPU_ERROR("dpu power client create failed: %d\n", rc); + dpu_kms->core_client = NULL; + goto error; + } + + pm_runtime_get_sync(&dpu_kms->pdev->dev); + + _dpu_kms_core_hw_rev_init(dpu_kms); + + pr_info("dpu hardware revision:0x%x\n", dpu_kms->core_rev); + + dpu_kms->catalog = dpu_hw_catalog_init(dpu_kms->core_rev); + if (IS_ERR_OR_NULL(dpu_kms->catalog)) { + rc = PTR_ERR(dpu_kms->catalog); + if (!dpu_kms->catalog) + rc = -EINVAL; + DPU_ERROR("catalog init failed: %d\n", rc); + dpu_kms->catalog = NULL; + goto power_error; + } + + dpu_dbg_init_dbg_buses(dpu_kms->core_rev); + + /* + * Now we need to read the HW catalog and initialize resources such as + * clocks, regulators, GDSC/MMAGIC, ioremap the register ranges etc + */ + rc = _dpu_kms_mmu_init(dpu_kms); + if (rc) { + DPU_ERROR("dpu_kms_mmu_init failed: %d\n", rc); + goto power_error; + } + + rc = dpu_rm_init(&dpu_kms->rm, dpu_kms->catalog, dpu_kms->mmio, + dpu_kms->dev); + if (rc) { + DPU_ERROR("rm init failed: %d\n", rc); + goto power_error; + } + + dpu_kms->rm_init = true; + + dpu_kms->hw_mdp = dpu_rm_get_mdp(&dpu_kms->rm); + if (IS_ERR_OR_NULL(dpu_kms->hw_mdp)) { + rc = PTR_ERR(dpu_kms->hw_mdp); + if (!dpu_kms->hw_mdp) + rc = -EINVAL; + DPU_ERROR("failed to get hw_mdp: %d\n", rc); + dpu_kms->hw_mdp = NULL; + goto power_error; + } + + for (i = 0; i < dpu_kms->catalog->vbif_count; i++) { + u32 vbif_idx = dpu_kms->catalog->vbif[i].id; + + dpu_kms->hw_vbif[i] = dpu_hw_vbif_init(vbif_idx, + dpu_kms->vbif[vbif_idx], dpu_kms->catalog); + if (IS_ERR_OR_NULL(dpu_kms->hw_vbif[vbif_idx])) { + rc = PTR_ERR(dpu_kms->hw_vbif[vbif_idx]); + if (!dpu_kms->hw_vbif[vbif_idx]) + rc = -EINVAL; + DPU_ERROR("failed to init vbif %d: %d\n", vbif_idx, rc); + dpu_kms->hw_vbif[vbif_idx] = NULL; + goto power_error; + } + } + + rc = dpu_core_perf_init(&dpu_kms->perf, dev, dpu_kms->catalog, + &dpu_kms->phandle, + _dpu_kms_get_clk(dpu_kms, "core")); + if (rc) { + DPU_ERROR("failed to init perf %d\n", rc); + goto perf_err; + } + + dpu_kms->hw_intr = dpu_hw_intr_init(dpu_kms->mmio, dpu_kms->catalog); + if (IS_ERR_OR_NULL(dpu_kms->hw_intr)) { + rc = PTR_ERR(dpu_kms->hw_intr); + DPU_ERROR("hw_intr init failed: %d\n", rc); + dpu_kms->hw_intr = NULL; + goto hw_intr_init_err; + } + + /* + * _dpu_kms_drm_obj_init should create the DRM related objects + * i.e. CRTCs, planes, encoders, connectors and so forth + */ + rc = _dpu_kms_drm_obj_init(dpu_kms); + if (rc) { + DPU_ERROR("modeset init failed: %d\n", rc); + goto drm_obj_init_err; + } + + dev->mode_config.min_width = 0; + dev->mode_config.min_height = 0; + + /* + * max crtc width is equal to the max mixer width * 2 and max height is + * is 4K + */ + dev->mode_config.max_width = + dpu_kms->catalog->caps->max_mixer_width * 2; + dev->mode_config.max_height = 4096; + + /* + * Support format modifiers for compression etc. + */ + dev->mode_config.allow_fb_modifiers = true; + + /* + * Handle (re)initializations during power enable + */ + dpu_kms_handle_power_event(DPU_POWER_EVENT_POST_ENABLE, dpu_kms); + dpu_kms->power_event = dpu_power_handle_register_event( + &dpu_kms->phandle, + DPU_POWER_EVENT_POST_ENABLE, + dpu_kms_handle_power_event, dpu_kms, "kms"); + + pm_runtime_put_sync(&dpu_kms->pdev->dev); + + return 0; + +drm_obj_init_err: + dpu_core_perf_destroy(&dpu_kms->perf); +hw_intr_init_err: +perf_err: +power_error: + pm_runtime_put_sync(&dpu_kms->pdev->dev); +error: + _dpu_kms_hw_destroy(dpu_kms); +dbg_destroy: + dpu_dbg_destroy(); +end: + return rc; +} + +struct msm_kms *dpu_kms_init(struct drm_device *dev) +{ + struct msm_drm_private *priv; + struct dpu_kms *dpu_kms; + int irq; + + if (!dev || !dev->dev_private) { + DPU_ERROR("drm device node invalid\n"); + return ERR_PTR(-EINVAL); + } + + priv = dev->dev_private; + dpu_kms = to_dpu_kms(priv->kms); + + irq = irq_of_parse_and_map(dpu_kms->pdev->dev.of_node, 0); + if (irq < 0) { + DPU_ERROR("failed to get irq: %d\n", irq); + return ERR_PTR(irq); + } + dpu_kms->base.irq = irq; + + return &dpu_kms->base; +} + +static int dpu_bind(struct device *dev, struct device *master, void *data) +{ + struct drm_device *ddev = dev_get_drvdata(master); + struct platform_device *pdev = to_platform_device(dev); + struct msm_drm_private *priv = ddev->dev_private; + struct dpu_kms *dpu_kms; + struct dss_module_power *mp; + int ret = 0; + + dpu_kms = devm_kzalloc(&pdev->dev, sizeof(*dpu_kms), GFP_KERNEL); + if (!dpu_kms) + return -ENOMEM; + + mp = &dpu_kms->mp; + ret = msm_dss_parse_clock(pdev, mp); + if (ret) { + DPU_ERROR("failed to parse clocks, ret=%d\n", ret); + return ret; + } + + dpu_power_resource_init(pdev, &dpu_kms->phandle); + + platform_set_drvdata(pdev, dpu_kms); + + msm_kms_init(&dpu_kms->base, &kms_funcs); + dpu_kms->dev = ddev; + dpu_kms->pdev = pdev; + + pm_runtime_enable(&pdev->dev); + dpu_kms->rpm_enabled = true; + + priv->kms = &dpu_kms->base; + return ret; +} + +static void dpu_unbind(struct device *dev, struct device *master, void *data) +{ + struct platform_device *pdev = to_platform_device(dev); + struct dpu_kms *dpu_kms = platform_get_drvdata(pdev); + struct dss_module_power *mp = &dpu_kms->mp; + + dpu_power_resource_deinit(pdev, &dpu_kms->phandle); + msm_dss_put_clk(mp->clk_config, mp->num_clk); + devm_kfree(&pdev->dev, mp->clk_config); + mp->num_clk = 0; + + if (dpu_kms->rpm_enabled) + pm_runtime_disable(&pdev->dev); +} + +static const struct component_ops dpu_ops = { + .bind = dpu_bind, + .unbind = dpu_unbind, +}; + +static int dpu_dev_probe(struct platform_device *pdev) +{ + return component_add(&pdev->dev, &dpu_ops); +} + +static int dpu_dev_remove(struct platform_device *pdev) +{ + component_del(&pdev->dev, &dpu_ops); + return 0; +} + +static int __maybe_unused dpu_runtime_suspend(struct device *dev) +{ + int rc = -1; + struct platform_device *pdev = to_platform_device(dev); + struct dpu_kms *dpu_kms = platform_get_drvdata(pdev); + struct drm_device *ddev; + struct dss_module_power *mp = &dpu_kms->mp; + + ddev = dpu_kms->dev; + if (!ddev) { + DPU_ERROR("invalid drm_device\n"); + goto exit; + } + + rc = dpu_power_resource_enable(&dpu_kms->phandle, + dpu_kms->core_client, false); + if (rc) + DPU_ERROR("resource disable failed: %d\n", rc); + + rc = msm_dss_enable_clk(mp->clk_config, mp->num_clk, false); + if (rc) + DPU_ERROR("clock disable failed rc:%d\n", rc); + +exit: + return rc; +} + +static int __maybe_unused dpu_runtime_resume(struct device *dev) +{ + int rc = -1; + struct platform_device *pdev = to_platform_device(dev); + struct dpu_kms *dpu_kms = platform_get_drvdata(pdev); + struct drm_device *ddev; + struct dss_module_power *mp = &dpu_kms->mp; + + ddev = dpu_kms->dev; + if (!ddev) { + DPU_ERROR("invalid drm_device\n"); + goto exit; + } + + rc = msm_dss_enable_clk(mp->clk_config, mp->num_clk, true); + if (rc) { + DPU_ERROR("clock enable failed rc:%d\n", rc); + goto exit; + } + + rc = dpu_power_resource_enable(&dpu_kms->phandle, + dpu_kms->core_client, true); + if (rc) + DPU_ERROR("resource enable failed: %d\n", rc); + +exit: + return rc; +} + +static const struct dev_pm_ops dpu_pm_ops = { + SET_RUNTIME_PM_OPS(dpu_runtime_suspend, dpu_runtime_resume, NULL) +}; + +static const struct of_device_id dpu_dt_match[] = { + { .compatible = "qcom,sdm845-dpu", }, + {} +}; +MODULE_DEVICE_TABLE(of, dpu_dt_match); + +static struct platform_driver dpu_driver = { + .probe = dpu_dev_probe, + .remove = dpu_dev_remove, + .driver = { + .name = "msm_dpu", + .of_match_table = dpu_dt_match, + .pm = &dpu_pm_ops, + }, +}; + +void __init msm_dpu_register(void) +{ + platform_driver_register(&dpu_driver); +} + +void __exit msm_dpu_unregister(void) +{ + platform_driver_unregister(&dpu_driver); +} diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_kms.h b/drivers/gpu/drm/msm/disp/dpu1/dpu_kms.h new file mode 100644 index 000000000000..66d466628e2b --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_kms.h @@ -0,0 +1,290 @@ +/* + * Copyright (c) 2015-2018, The Linux Foundation. All rights reserved. + * Copyright (C) 2013 Red Hat + * Author: Rob Clark <robdclark@gmail.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published by + * the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program. If not, see <http://www.gnu.org/licenses/>. + */ + +#ifndef __DPU_KMS_H__ +#define __DPU_KMS_H__ + +#include "msm_drv.h" +#include "msm_kms.h" +#include "msm_mmu.h" +#include "msm_gem.h" +#include "dpu_dbg.h" +#include "dpu_hw_catalog.h" +#include "dpu_hw_ctl.h" +#include "dpu_hw_lm.h" +#include "dpu_hw_interrupts.h" +#include "dpu_hw_top.h" +#include "dpu_rm.h" +#include "dpu_power_handle.h" +#include "dpu_irq.h" +#include "dpu_core_perf.h" + +#define DRMID(x) ((x) ? (x)->base.id : -1) + +/** + * DPU_DEBUG - macro for kms/plane/crtc/encoder/connector logs + * @fmt: Pointer to format string + */ +#define DPU_DEBUG(fmt, ...) \ + do { \ + if (unlikely(drm_debug & DRM_UT_KMS)) \ + DRM_DEBUG(fmt, ##__VA_ARGS__); \ + else \ + pr_debug(fmt, ##__VA_ARGS__); \ + } while (0) + +/** + * DPU_DEBUG_DRIVER - macro for hardware driver logging + * @fmt: Pointer to format string + */ +#define DPU_DEBUG_DRIVER(fmt, ...) \ + do { \ + if (unlikely(drm_debug & DRM_UT_DRIVER)) \ + DRM_ERROR(fmt, ##__VA_ARGS__); \ + else \ + pr_debug(fmt, ##__VA_ARGS__); \ + } while (0) + +#define DPU_ERROR(fmt, ...) pr_err("[dpu error]" fmt, ##__VA_ARGS__) + +/** + * ktime_compare_safe - compare two ktime structures + * This macro is similar to the standard ktime_compare() function, but + * attempts to also handle ktime overflows. + * @A: First ktime value + * @B: Second ktime value + * Returns: -1 if A < B, 0 if A == B, 1 if A > B + */ +#define ktime_compare_safe(A, B) \ + ktime_compare(ktime_sub((A), (B)), ktime_set(0, 0)) + +#define DPU_NAME_SIZE 12 + +/* timeout in frames waiting for frame done */ +#define DPU_FRAME_DONE_TIMEOUT 60 + +/* + * struct dpu_irq_callback - IRQ callback handlers + * @list: list to callback + * @func: intr handler + * @arg: argument for the handler + */ +struct dpu_irq_callback { + struct list_head list; + void (*func)(void *arg, int irq_idx); + void *arg; +}; + +/** + * struct dpu_irq: IRQ structure contains callback registration info + * @total_irq: total number of irq_idx obtained from HW interrupts mapping + * @irq_cb_tbl: array of IRQ callbacks setting + * @enable_counts array of IRQ enable counts + * @cb_lock: callback lock + * @debugfs_file: debugfs file for irq statistics + */ +struct dpu_irq { + u32 total_irqs; + struct list_head *irq_cb_tbl; + atomic_t *enable_counts; + atomic_t *irq_counts; + spinlock_t cb_lock; + struct dentry *debugfs_file; +}; + +struct dpu_kms { + struct msm_kms base; + struct drm_device *dev; + int core_rev; + struct dpu_mdss_cfg *catalog; + + struct dpu_power_handle phandle; + struct dpu_power_client *core_client; + struct dpu_power_event *power_event; + + /* directory entry for debugfs */ + struct dentry *debugfs_root; + struct dentry *debugfs_danger; + struct dentry *debugfs_vbif; + + /* io/register spaces: */ + void __iomem *mmio, *vbif[VBIF_MAX], *reg_dma; + unsigned long mmio_len, vbif_len[VBIF_MAX], reg_dma_len; + + struct regulator *vdd; + struct regulator *mmagic; + struct regulator *venus; + + struct dpu_hw_intr *hw_intr; + struct dpu_irq irq_obj; + + struct dpu_core_perf perf; + + /* saved atomic state during system suspend */ + struct drm_atomic_state *suspend_state; + bool suspend_block; + + struct dpu_rm rm; + bool rm_init; + + struct dpu_hw_vbif *hw_vbif[VBIF_MAX]; + struct dpu_hw_mdp *hw_mdp; + + bool has_danger_ctrl; + + struct platform_device *pdev; + bool rpm_enabled; + struct dss_module_power mp; +}; + +struct vsync_info { + u32 frame_count; + u32 line_count; +}; + +#define to_dpu_kms(x) container_of(x, struct dpu_kms, base) + +/* get struct msm_kms * from drm_device * */ +#define ddev_to_msm_kms(D) ((D) && (D)->dev_private ? \ + ((struct msm_drm_private *)((D)->dev_private))->kms : NULL) + +/** + * dpu_kms_is_suspend_state - whether or not the system is pm suspended + * @dev: Pointer to drm device + * Return: Suspend status + */ +static inline bool dpu_kms_is_suspend_state(struct drm_device *dev) +{ + if (!ddev_to_msm_kms(dev)) + return false; + + return to_dpu_kms(ddev_to_msm_kms(dev))->suspend_state != NULL; +} + +/** + * dpu_kms_is_suspend_blocked - whether or not commits are blocked due to pm + * suspend status + * @dev: Pointer to drm device + * Return: True if commits should be rejected due to pm suspend + */ +static inline bool dpu_kms_is_suspend_blocked(struct drm_device *dev) +{ + if (!dpu_kms_is_suspend_state(dev)) + return false; + + return to_dpu_kms(ddev_to_msm_kms(dev))->suspend_block; +} + +/** + * Debugfs functions - extra helper functions for debugfs support + * + * Main debugfs documentation is located at, + * + * Documentation/filesystems/debugfs.txt + * + * @dpu_debugfs_setup_regset32: Initialize data for dpu_debugfs_create_regset32 + * @dpu_debugfs_create_regset32: Create 32-bit register dump file + * @dpu_debugfs_get_root: Get root dentry for DPU_KMS's debugfs node + */ + +/** + * Companion structure for dpu_debugfs_create_regset32. Do not initialize the + * members of this structure explicitly; use dpu_debugfs_setup_regset32 instead. + */ +struct dpu_debugfs_regset32 { + uint32_t offset; + uint32_t blk_len; + struct dpu_kms *dpu_kms; +}; + +/** + * dpu_debugfs_setup_regset32 - Initialize register block definition for debugfs + * This function is meant to initialize dpu_debugfs_regset32 structures for use + * with dpu_debugfs_create_regset32. + * @regset: opaque register definition structure + * @offset: sub-block offset + * @length: sub-block length, in bytes + * @dpu_kms: pointer to dpu kms structure + */ +void dpu_debugfs_setup_regset32(struct dpu_debugfs_regset32 *regset, + uint32_t offset, uint32_t length, struct dpu_kms *dpu_kms); + +/** + * dpu_debugfs_create_regset32 - Create register read back file for debugfs + * + * This function is almost identical to the standard debugfs_create_regset32() + * function, with the main difference being that a list of register + * names/offsets do not need to be provided. The 'read' function simply outputs + * sequential register values over a specified range. + * + * Similar to the related debugfs_create_regset32 API, the structure pointed to + * by regset needs to persist for the lifetime of the created file. The calling + * code is responsible for initialization/management of this structure. + * + * The structure pointed to by regset is meant to be opaque. Please use + * dpu_debugfs_setup_regset32 to initialize it. + * + * @name: File name within debugfs + * @mode: File mode within debugfs + * @parent: Parent directory entry within debugfs, can be NULL + * @regset: Pointer to persistent register block definition + * + * Return: dentry pointer for newly created file, use either debugfs_remove() + * or debugfs_remove_recursive() (on a parent directory) to remove the + * file + */ +void *dpu_debugfs_create_regset32(const char *name, umode_t mode, + void *parent, struct dpu_debugfs_regset32 *regset); + +/** + * dpu_debugfs_get_root - Return root directory entry for KMS's debugfs + * + * The return value should be passed as the 'parent' argument to subsequent + * debugfs create calls. + * + * @dpu_kms: Pointer to DPU's KMS structure + * + * Return: dentry pointer for DPU's debugfs location + */ +void *dpu_debugfs_get_root(struct dpu_kms *dpu_kms); + +/** + * DPU info management functions + * These functions/definitions allow for building up a 'dpu_info' structure + * containing one or more "key=value\n" entries. + */ +#define DPU_KMS_INFO_MAX_SIZE 4096 + +/** + * Vblank enable/disable functions + */ +int dpu_enable_vblank(struct msm_kms *kms, struct drm_crtc *crtc); +void dpu_disable_vblank(struct msm_kms *kms, struct drm_crtc *crtc); + +void dpu_kms_encoder_enable(struct drm_encoder *encoder); + +/** + * dpu_kms_get_clk_rate() - get the clock rate + * @dpu_kms: poiner to dpu_kms structure + * @clock_name: clock name to get the rate + * + * Return: current clock rate + */ +u64 dpu_kms_get_clk_rate(struct dpu_kms *dpu_kms, char *clock_name); + +#endif /* __dpu_kms_H__ */ diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_mdss.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_mdss.c new file mode 100644 index 000000000000..9e533b86682c --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_mdss.c @@ -0,0 +1,245 @@ +/* + * SPDX-License-Identifier: GPL-2.0 + * Copyright (c) 2018, The Linux Foundation + */ + +#include "dpu_kms.h" + +#define to_dpu_mdss(x) container_of(x, struct dpu_mdss, base) + +#define HW_INTR_STATUS 0x0010 + +struct dpu_mdss { + struct msm_mdss base; + void __iomem *mmio; + unsigned long mmio_len; + u32 hwversion; + struct dss_module_power mp; + struct dpu_irq_controller irq_controller; +}; + +static irqreturn_t dpu_mdss_irq(int irq, void *arg) +{ + struct dpu_mdss *dpu_mdss = arg; + u32 interrupts; + + interrupts = readl_relaxed(dpu_mdss->mmio + HW_INTR_STATUS); + + while (interrupts) { + irq_hw_number_t hwirq = fls(interrupts) - 1; + unsigned int mapping; + int rc; + + mapping = irq_find_mapping(dpu_mdss->irq_controller.domain, + hwirq); + if (mapping == 0) { + DRM_ERROR("couldn't find irq mapping for %lu\n", hwirq); + return IRQ_NONE; + } + + rc = generic_handle_irq(mapping); + if (rc < 0) { + DRM_ERROR("handle irq fail: irq=%lu mapping=%u rc=%d\n", + hwirq, mapping, rc); + return IRQ_NONE; + } + + interrupts &= ~(1 << hwirq); + } + + return IRQ_HANDLED; +} + +static void dpu_mdss_irq_mask(struct irq_data *irqd) +{ + struct dpu_mdss *dpu_mdss = irq_data_get_irq_chip_data(irqd); + + /* memory barrier */ + smp_mb__before_atomic(); + clear_bit(irqd->hwirq, &dpu_mdss->irq_controller.enabled_mask); + /* memory barrier */ + smp_mb__after_atomic(); +} + +static void dpu_mdss_irq_unmask(struct irq_data *irqd) +{ + struct dpu_mdss *dpu_mdss = irq_data_get_irq_chip_data(irqd); + + /* memory barrier */ + smp_mb__before_atomic(); + set_bit(irqd->hwirq, &dpu_mdss->irq_controller.enabled_mask); + /* memory barrier */ + smp_mb__after_atomic(); +} + +static struct irq_chip dpu_mdss_irq_chip = { + .name = "dpu_mdss", + .irq_mask = dpu_mdss_irq_mask, + .irq_unmask = dpu_mdss_irq_unmask, +}; + +static int dpu_mdss_irqdomain_map(struct irq_domain *domain, + unsigned int irq, irq_hw_number_t hwirq) +{ + struct dpu_mdss *dpu_mdss = domain->host_data; + int ret; + + irq_set_chip_and_handler(irq, &dpu_mdss_irq_chip, handle_level_irq); + ret = irq_set_chip_data(irq, dpu_mdss); + + return ret; +} + +static const struct irq_domain_ops dpu_mdss_irqdomain_ops = { + .map = dpu_mdss_irqdomain_map, + .xlate = irq_domain_xlate_onecell, +}; + +static int _dpu_mdss_irq_domain_add(struct dpu_mdss *dpu_mdss) +{ + struct device *dev; + struct irq_domain *domain; + + dev = dpu_mdss->base.dev->dev; + + domain = irq_domain_add_linear(dev->of_node, 32, + &dpu_mdss_irqdomain_ops, dpu_mdss); + if (!domain) { + DPU_ERROR("failed to add irq_domain\n"); + return -EINVAL; + } + + dpu_mdss->irq_controller.enabled_mask = 0; + dpu_mdss->irq_controller.domain = domain; + + return 0; +} + +static int _dpu_mdss_irq_domain_fini(struct dpu_mdss *dpu_mdss) +{ + if (dpu_mdss->irq_controller.domain) { + irq_domain_remove(dpu_mdss->irq_controller.domain); + dpu_mdss->irq_controller.domain = NULL; + } + return 0; +} +static int dpu_mdss_enable(struct msm_mdss *mdss) +{ + struct dpu_mdss *dpu_mdss = to_dpu_mdss(mdss); + struct dss_module_power *mp = &dpu_mdss->mp; + int ret; + + ret = msm_dss_enable_clk(mp->clk_config, mp->num_clk, true); + if (ret) + DPU_ERROR("clock enable failed, ret:%d\n", ret); + + return ret; +} + +static int dpu_mdss_disable(struct msm_mdss *mdss) +{ + struct dpu_mdss *dpu_mdss = to_dpu_mdss(mdss); + struct dss_module_power *mp = &dpu_mdss->mp; + int ret; + + ret = msm_dss_enable_clk(mp->clk_config, mp->num_clk, false); + if (ret) + DPU_ERROR("clock disable failed, ret:%d\n", ret); + + return ret; +} + +static void dpu_mdss_destroy(struct drm_device *dev) +{ + struct platform_device *pdev = to_platform_device(dev->dev); + struct msm_drm_private *priv = dev->dev_private; + struct dpu_mdss *dpu_mdss = to_dpu_mdss(priv->mdss); + struct dss_module_power *mp = &dpu_mdss->mp; + + _dpu_mdss_irq_domain_fini(dpu_mdss); + + msm_dss_put_clk(mp->clk_config, mp->num_clk); + devm_kfree(&pdev->dev, mp->clk_config); + + if (dpu_mdss->mmio) + devm_iounmap(&pdev->dev, dpu_mdss->mmio); + dpu_mdss->mmio = NULL; + + pm_runtime_disable(dev->dev); + priv->mdss = NULL; +} + +static const struct msm_mdss_funcs mdss_funcs = { + .enable = dpu_mdss_enable, + .disable = dpu_mdss_disable, + .destroy = dpu_mdss_destroy, +}; + +int dpu_mdss_init(struct drm_device *dev) +{ + struct platform_device *pdev = to_platform_device(dev->dev); + struct msm_drm_private *priv = dev->dev_private; + struct resource *res; + struct dpu_mdss *dpu_mdss; + struct dss_module_power *mp; + int ret = 0; + + dpu_mdss = devm_kzalloc(dev->dev, sizeof(*dpu_mdss), GFP_KERNEL); + if (!dpu_mdss) + return -ENOMEM; + + dpu_mdss->mmio = msm_ioremap(pdev, "mdss", "mdss"); + if (IS_ERR(dpu_mdss->mmio)) + return PTR_ERR(dpu_mdss->mmio); + + DRM_DEBUG("mapped mdss address space @%pK\n", dpu_mdss->mmio); + + res = platform_get_resource_byname(pdev, IORESOURCE_MEM, "mdss"); + if (!res) { + DRM_ERROR("failed to get memory resource for mdss\n"); + return -ENOMEM; + } + dpu_mdss->mmio_len = resource_size(res); + + mp = &dpu_mdss->mp; + ret = msm_dss_parse_clock(pdev, mp); + if (ret) { + DPU_ERROR("failed to parse clocks, ret=%d\n", ret); + goto clk_parse_err; + } + + dpu_mdss->base.dev = dev; + dpu_mdss->base.funcs = &mdss_funcs; + + ret = _dpu_mdss_irq_domain_add(dpu_mdss); + if (ret) + goto irq_domain_error; + + ret = devm_request_irq(dev->dev, platform_get_irq(pdev, 0), + dpu_mdss_irq, 0, "dpu_mdss_isr", dpu_mdss); + if (ret) { + DPU_ERROR("failed to init irq: %d\n", ret); + goto irq_error; + } + + pm_runtime_enable(dev->dev); + + pm_runtime_get_sync(dev->dev); + dpu_mdss->hwversion = readl_relaxed(dpu_mdss->mmio); + pm_runtime_put_sync(dev->dev); + + priv->mdss = &dpu_mdss->base; + + return ret; + +irq_error: + _dpu_mdss_irq_domain_fini(dpu_mdss); +irq_domain_error: + msm_dss_put_clk(mp->clk_config, mp->num_clk); +clk_parse_err: + devm_kfree(&pdev->dev, mp->clk_config); + if (dpu_mdss->mmio) + devm_iounmap(&pdev->dev, dpu_mdss->mmio); + dpu_mdss->mmio = NULL; + return ret; +} diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_plane.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_plane.c new file mode 100644 index 000000000000..b640e39ebaca --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_plane.c @@ -0,0 +1,1963 @@ +/* + * Copyright (C) 2014-2018 The Linux Foundation. All rights reserved. + * Copyright (C) 2013 Red Hat + * Author: Rob Clark <robdclark@gmail.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published by + * the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program. If not, see <http://www.gnu.org/licenses/>. + */ + +#define pr_fmt(fmt) "[drm:%s:%d] " fmt, __func__, __LINE__ + +#include <linux/debugfs.h> +#include <linux/dma-buf.h> + +#include "msm_drv.h" +#include "dpu_kms.h" +#include "dpu_formats.h" +#include "dpu_hw_sspp.h" +#include "dpu_hw_catalog_format.h" +#include "dpu_trace.h" +#include "dpu_crtc.h" +#include "dpu_vbif.h" +#include "dpu_plane.h" + +#define DPU_DEBUG_PLANE(pl, fmt, ...) DPU_DEBUG("plane%d " fmt,\ + (pl) ? (pl)->base.base.id : -1, ##__VA_ARGS__) + +#define DPU_ERROR_PLANE(pl, fmt, ...) DPU_ERROR("plane%d " fmt,\ + (pl) ? (pl)->base.base.id : -1, ##__VA_ARGS__) + +#define DECIMATED_DIMENSION(dim, deci) (((dim) + ((1 << (deci)) - 1)) >> (deci)) +#define PHASE_STEP_SHIFT 21 +#define PHASE_STEP_UNIT_SCALE ((int) (1 << PHASE_STEP_SHIFT)) +#define PHASE_RESIDUAL 15 + +#define SHARP_STRENGTH_DEFAULT 32 +#define SHARP_EDGE_THR_DEFAULT 112 +#define SHARP_SMOOTH_THR_DEFAULT 8 +#define SHARP_NOISE_THR_DEFAULT 2 + +#define DPU_NAME_SIZE 12 + +#define DPU_PLANE_COLOR_FILL_FLAG BIT(31) +#define DPU_ZPOS_MAX 255 + +/* multirect rect index */ +enum { + R0, + R1, + R_MAX +}; + +#define DPU_QSEED3_DEFAULT_PRELOAD_H 0x4 +#define DPU_QSEED3_DEFAULT_PRELOAD_V 0x3 + +#define DEFAULT_REFRESH_RATE 60 + +/** + * enum dpu_plane_qos - Different qos configurations for each pipe + * + * @DPU_PLANE_QOS_VBLANK_CTRL: Setup VBLANK qos for the pipe. + * @DPU_PLANE_QOS_VBLANK_AMORTIZE: Enables Amortization within pipe. + * this configuration is mutually exclusive from VBLANK_CTRL. + * @DPU_PLANE_QOS_PANIC_CTRL: Setup panic for the pipe. + */ +enum dpu_plane_qos { + DPU_PLANE_QOS_VBLANK_CTRL = BIT(0), + DPU_PLANE_QOS_VBLANK_AMORTIZE = BIT(1), + DPU_PLANE_QOS_PANIC_CTRL = BIT(2), +}; + +/* + * struct dpu_plane - local dpu plane structure + * @aspace: address space pointer + * @csc_ptr: Points to dpu_csc_cfg structure to use for current + * @mplane_list: List of multirect planes of the same pipe + * @catalog: Points to dpu catalog structure + * @revalidate: force revalidation of all the plane properties + */ +struct dpu_plane { + struct drm_plane base; + + struct mutex lock; + + enum dpu_sspp pipe; + uint32_t features; /* capabilities from catalog */ + uint32_t nformats; + uint32_t formats[64]; + + struct dpu_hw_pipe *pipe_hw; + struct dpu_hw_pipe_cfg pipe_cfg; + struct dpu_hw_pipe_qos_cfg pipe_qos_cfg; + uint32_t color_fill; + bool is_error; + bool is_rt_pipe; + bool is_virtual; + struct list_head mplane_list; + struct dpu_mdss_cfg *catalog; + + struct dpu_csc_cfg *csc_ptr; + + const struct dpu_sspp_sub_blks *pipe_sblk; + char pipe_name[DPU_NAME_SIZE]; + + /* debugfs related stuff */ + struct dentry *debugfs_root; + struct dpu_debugfs_regset32 debugfs_src; + struct dpu_debugfs_regset32 debugfs_scaler; + struct dpu_debugfs_regset32 debugfs_csc; + bool debugfs_default_scale; +}; + +#define to_dpu_plane(x) container_of(x, struct dpu_plane, base) + +static struct dpu_kms *_dpu_plane_get_kms(struct drm_plane *plane) +{ + struct msm_drm_private *priv; + + if (!plane || !plane->dev) + return NULL; + priv = plane->dev->dev_private; + if (!priv) + return NULL; + return to_dpu_kms(priv->kms); +} + +static bool dpu_plane_enabled(struct drm_plane_state *state) +{ + return state && state->fb && state->crtc; +} + +static bool dpu_plane_sspp_enabled(struct drm_plane_state *state) +{ + return state && state->crtc; +} + +/** + * _dpu_plane_calc_fill_level - calculate fill level of the given source format + * @plane: Pointer to drm plane + * @fmt: Pointer to source buffer format + * @src_wdith: width of source buffer + * Return: fill level corresponding to the source buffer/format or 0 if error + */ +static inline int _dpu_plane_calc_fill_level(struct drm_plane *plane, + const struct dpu_format *fmt, u32 src_width) +{ + struct dpu_plane *pdpu, *tmp; + struct dpu_plane_state *pstate; + u32 fixed_buff_size; + u32 total_fl; + + if (!plane || !fmt || !plane->state || !src_width || !fmt->bpp) { + DPU_ERROR("invalid arguments\n"); + return 0; + } + + pdpu = to_dpu_plane(plane); + pstate = to_dpu_plane_state(plane->state); + fixed_buff_size = pdpu->pipe_sblk->common->pixel_ram_size; + + list_for_each_entry(tmp, &pdpu->mplane_list, mplane_list) { + if (!dpu_plane_enabled(tmp->base.state)) + continue; + DPU_DEBUG("plane%d/%d src_width:%d/%d\n", + pdpu->base.base.id, tmp->base.base.id, + src_width, + drm_rect_width(&tmp->pipe_cfg.src_rect)); + src_width = max_t(u32, src_width, + drm_rect_width(&tmp->pipe_cfg.src_rect)); + } + + if (fmt->fetch_planes == DPU_PLANE_PSEUDO_PLANAR) { + if (fmt->chroma_sample == DPU_CHROMA_420) { + /* NV12 */ + total_fl = (fixed_buff_size / 2) / + ((src_width + 32) * fmt->bpp); + } else { + /* non NV12 */ + total_fl = (fixed_buff_size / 2) * 2 / + ((src_width + 32) * fmt->bpp); + } + } else { + if (pstate->multirect_mode == DPU_SSPP_MULTIRECT_PARALLEL) { + total_fl = (fixed_buff_size / 2) * 2 / + ((src_width + 32) * fmt->bpp); + } else { + total_fl = (fixed_buff_size) * 2 / + ((src_width + 32) * fmt->bpp); + } + } + + DPU_DEBUG("plane%u: pnum:%d fmt: %4.4s w:%u fl:%u\n", + plane->base.id, pdpu->pipe - SSPP_VIG0, + (char *)&fmt->base.pixel_format, + src_width, total_fl); + + return total_fl; +} + +/** + * _dpu_plane_get_qos_lut - get LUT mapping based on fill level + * @tbl: Pointer to LUT table + * @total_fl: fill level + * Return: LUT setting corresponding to the fill level + */ +static u64 _dpu_plane_get_qos_lut(const struct dpu_qos_lut_tbl *tbl, + u32 total_fl) +{ + int i; + + if (!tbl || !tbl->nentry || !tbl->entries) + return 0; + + for (i = 0; i < tbl->nentry; i++) + if (total_fl <= tbl->entries[i].fl) + return tbl->entries[i].lut; + + /* if last fl is zero, use as default */ + if (!tbl->entries[i-1].fl) + return tbl->entries[i-1].lut; + + return 0; +} + +/** + * _dpu_plane_set_qos_lut - set QoS LUT of the given plane + * @plane: Pointer to drm plane + * @fb: Pointer to framebuffer associated with the given plane + */ +static void _dpu_plane_set_qos_lut(struct drm_plane *plane, + struct drm_framebuffer *fb) +{ + struct dpu_plane *pdpu; + const struct dpu_format *fmt = NULL; + u64 qos_lut; + u32 total_fl = 0, lut_usage; + + if (!plane || !fb) { + DPU_ERROR("invalid arguments plane %d fb %d\n", + plane != 0, fb != 0); + return; + } + + pdpu = to_dpu_plane(plane); + + if (!pdpu->pipe_hw || !pdpu->pipe_sblk || !pdpu->catalog) { + DPU_ERROR("invalid arguments\n"); + return; + } else if (!pdpu->pipe_hw->ops.setup_creq_lut) { + return; + } + + if (!pdpu->is_rt_pipe) { + lut_usage = DPU_QOS_LUT_USAGE_NRT; + } else { + fmt = dpu_get_dpu_format_ext( + fb->format->format, + fb->modifier); + total_fl = _dpu_plane_calc_fill_level(plane, fmt, + drm_rect_width(&pdpu->pipe_cfg.src_rect)); + + if (fmt && DPU_FORMAT_IS_LINEAR(fmt)) + lut_usage = DPU_QOS_LUT_USAGE_LINEAR; + else + lut_usage = DPU_QOS_LUT_USAGE_MACROTILE; + } + + qos_lut = _dpu_plane_get_qos_lut( + &pdpu->catalog->perf.qos_lut_tbl[lut_usage], total_fl); + + pdpu->pipe_qos_cfg.creq_lut = qos_lut; + + trace_dpu_perf_set_qos_luts(pdpu->pipe - SSPP_VIG0, + (fmt) ? fmt->base.pixel_format : 0, + pdpu->is_rt_pipe, total_fl, qos_lut, lut_usage); + + DPU_DEBUG("plane%u: pnum:%d fmt: %4.4s rt:%d fl:%u lut:0x%llx\n", + plane->base.id, + pdpu->pipe - SSPP_VIG0, + fmt ? (char *)&fmt->base.pixel_format : NULL, + pdpu->is_rt_pipe, total_fl, qos_lut); + + pdpu->pipe_hw->ops.setup_creq_lut(pdpu->pipe_hw, &pdpu->pipe_qos_cfg); +} + +/** + * _dpu_plane_set_panic_lut - set danger/safe LUT of the given plane + * @plane: Pointer to drm plane + * @fb: Pointer to framebuffer associated with the given plane + */ +static void _dpu_plane_set_danger_lut(struct drm_plane *plane, + struct drm_framebuffer *fb) +{ + struct dpu_plane *pdpu; + const struct dpu_format *fmt = NULL; + u32 danger_lut, safe_lut; + + if (!plane || !fb) { + DPU_ERROR("invalid arguments\n"); + return; + } + + pdpu = to_dpu_plane(plane); + + if (!pdpu->pipe_hw || !pdpu->pipe_sblk || !pdpu->catalog) { + DPU_ERROR("invalid arguments\n"); + return; + } else if (!pdpu->pipe_hw->ops.setup_danger_safe_lut) { + return; + } + + if (!pdpu->is_rt_pipe) { + danger_lut = pdpu->catalog->perf.danger_lut_tbl + [DPU_QOS_LUT_USAGE_NRT]; + safe_lut = pdpu->catalog->perf.safe_lut_tbl + [DPU_QOS_LUT_USAGE_NRT]; + } else { + fmt = dpu_get_dpu_format_ext( + fb->format->format, + fb->modifier); + + if (fmt && DPU_FORMAT_IS_LINEAR(fmt)) { + danger_lut = pdpu->catalog->perf.danger_lut_tbl + [DPU_QOS_LUT_USAGE_LINEAR]; + safe_lut = pdpu->catalog->perf.safe_lut_tbl + [DPU_QOS_LUT_USAGE_LINEAR]; + } else { + danger_lut = pdpu->catalog->perf.danger_lut_tbl + [DPU_QOS_LUT_USAGE_MACROTILE]; + safe_lut = pdpu->catalog->perf.safe_lut_tbl + [DPU_QOS_LUT_USAGE_MACROTILE]; + } + } + + pdpu->pipe_qos_cfg.danger_lut = danger_lut; + pdpu->pipe_qos_cfg.safe_lut = safe_lut; + + trace_dpu_perf_set_danger_luts(pdpu->pipe - SSPP_VIG0, + (fmt) ? fmt->base.pixel_format : 0, + (fmt) ? fmt->fetch_mode : 0, + pdpu->pipe_qos_cfg.danger_lut, + pdpu->pipe_qos_cfg.safe_lut); + + DPU_DEBUG("plane%u: pnum:%d fmt: %4.4s mode:%d luts[0x%x, 0x%x]\n", + plane->base.id, + pdpu->pipe - SSPP_VIG0, + fmt ? (char *)&fmt->base.pixel_format : NULL, + fmt ? fmt->fetch_mode : -1, + pdpu->pipe_qos_cfg.danger_lut, + pdpu->pipe_qos_cfg.safe_lut); + + pdpu->pipe_hw->ops.setup_danger_safe_lut(pdpu->pipe_hw, + &pdpu->pipe_qos_cfg); +} + +/** + * _dpu_plane_set_qos_ctrl - set QoS control of the given plane + * @plane: Pointer to drm plane + * @enable: true to enable QoS control + * @flags: QoS control mode (enum dpu_plane_qos) + */ +static void _dpu_plane_set_qos_ctrl(struct drm_plane *plane, + bool enable, u32 flags) +{ + struct dpu_plane *pdpu; + + if (!plane) { + DPU_ERROR("invalid arguments\n"); + return; + } + + pdpu = to_dpu_plane(plane); + + if (!pdpu->pipe_hw || !pdpu->pipe_sblk) { + DPU_ERROR("invalid arguments\n"); + return; + } else if (!pdpu->pipe_hw->ops.setup_qos_ctrl) { + return; + } + + if (flags & DPU_PLANE_QOS_VBLANK_CTRL) { + pdpu->pipe_qos_cfg.creq_vblank = pdpu->pipe_sblk->creq_vblank; + pdpu->pipe_qos_cfg.danger_vblank = + pdpu->pipe_sblk->danger_vblank; + pdpu->pipe_qos_cfg.vblank_en = enable; + } + + if (flags & DPU_PLANE_QOS_VBLANK_AMORTIZE) { + /* this feature overrules previous VBLANK_CTRL */ + pdpu->pipe_qos_cfg.vblank_en = false; + pdpu->pipe_qos_cfg.creq_vblank = 0; /* clear vblank bits */ + } + + if (flags & DPU_PLANE_QOS_PANIC_CTRL) + pdpu->pipe_qos_cfg.danger_safe_en = enable; + + if (!pdpu->is_rt_pipe) { + pdpu->pipe_qos_cfg.vblank_en = false; + pdpu->pipe_qos_cfg.danger_safe_en = false; + } + + DPU_DEBUG("plane%u: pnum:%d ds:%d vb:%d pri[0x%x, 0x%x] is_rt:%d\n", + plane->base.id, + pdpu->pipe - SSPP_VIG0, + pdpu->pipe_qos_cfg.danger_safe_en, + pdpu->pipe_qos_cfg.vblank_en, + pdpu->pipe_qos_cfg.creq_vblank, + pdpu->pipe_qos_cfg.danger_vblank, + pdpu->is_rt_pipe); + + pdpu->pipe_hw->ops.setup_qos_ctrl(pdpu->pipe_hw, + &pdpu->pipe_qos_cfg); +} + +int dpu_plane_danger_signal_ctrl(struct drm_plane *plane, bool enable) +{ + struct dpu_plane *pdpu; + struct msm_drm_private *priv; + struct dpu_kms *dpu_kms; + + if (!plane || !plane->dev) { + DPU_ERROR("invalid arguments\n"); + return -EINVAL; + } + + priv = plane->dev->dev_private; + if (!priv || !priv->kms) { + DPU_ERROR("invalid KMS reference\n"); + return -EINVAL; + } + + dpu_kms = to_dpu_kms(priv->kms); + pdpu = to_dpu_plane(plane); + + if (!pdpu->is_rt_pipe) + goto end; + + pm_runtime_get_sync(&dpu_kms->pdev->dev); + _dpu_plane_set_qos_ctrl(plane, enable, DPU_PLANE_QOS_PANIC_CTRL); + pm_runtime_put_sync(&dpu_kms->pdev->dev); + +end: + return 0; +} + +/** + * _dpu_plane_set_ot_limit - set OT limit for the given plane + * @plane: Pointer to drm plane + * @crtc: Pointer to drm crtc + */ +static void _dpu_plane_set_ot_limit(struct drm_plane *plane, + struct drm_crtc *crtc) +{ + struct dpu_plane *pdpu; + struct dpu_vbif_set_ot_params ot_params; + struct msm_drm_private *priv; + struct dpu_kms *dpu_kms; + + if (!plane || !plane->dev || !crtc) { + DPU_ERROR("invalid arguments plane %d crtc %d\n", + plane != 0, crtc != 0); + return; + } + + priv = plane->dev->dev_private; + if (!priv || !priv->kms) { + DPU_ERROR("invalid KMS reference\n"); + return; + } + + dpu_kms = to_dpu_kms(priv->kms); + pdpu = to_dpu_plane(plane); + if (!pdpu->pipe_hw) { + DPU_ERROR("invalid pipe reference\n"); + return; + } + + memset(&ot_params, 0, sizeof(ot_params)); + ot_params.xin_id = pdpu->pipe_hw->cap->xin_id; + ot_params.num = pdpu->pipe_hw->idx - SSPP_NONE; + ot_params.width = drm_rect_width(&pdpu->pipe_cfg.src_rect); + ot_params.height = drm_rect_height(&pdpu->pipe_cfg.src_rect); + ot_params.is_wfd = !pdpu->is_rt_pipe; + ot_params.frame_rate = crtc->mode.vrefresh; + ot_params.vbif_idx = VBIF_RT; + ot_params.clk_ctrl = pdpu->pipe_hw->cap->clk_ctrl; + ot_params.rd = true; + + dpu_vbif_set_ot_limit(dpu_kms, &ot_params); +} + +/** + * _dpu_plane_set_vbif_qos - set vbif QoS for the given plane + * @plane: Pointer to drm plane + */ +static void _dpu_plane_set_qos_remap(struct drm_plane *plane) +{ + struct dpu_plane *pdpu; + struct dpu_vbif_set_qos_params qos_params; + struct msm_drm_private *priv; + struct dpu_kms *dpu_kms; + + if (!plane || !plane->dev) { + DPU_ERROR("invalid arguments\n"); + return; + } + + priv = plane->dev->dev_private; + if (!priv || !priv->kms) { + DPU_ERROR("invalid KMS reference\n"); + return; + } + + dpu_kms = to_dpu_kms(priv->kms); + pdpu = to_dpu_plane(plane); + if (!pdpu->pipe_hw) { + DPU_ERROR("invalid pipe reference\n"); + return; + } + + memset(&qos_params, 0, sizeof(qos_params)); + qos_params.vbif_idx = VBIF_RT; + qos_params.clk_ctrl = pdpu->pipe_hw->cap->clk_ctrl; + qos_params.xin_id = pdpu->pipe_hw->cap->xin_id; + qos_params.num = pdpu->pipe_hw->idx - SSPP_VIG0; + qos_params.is_rt = pdpu->is_rt_pipe; + + DPU_DEBUG("plane%d pipe:%d vbif:%d xin:%d rt:%d, clk_ctrl:%d\n", + plane->base.id, qos_params.num, + qos_params.vbif_idx, + qos_params.xin_id, qos_params.is_rt, + qos_params.clk_ctrl); + + dpu_vbif_set_qos_remap(dpu_kms, &qos_params); +} + +/** + * _dpu_plane_get_aspace: gets the address space + */ +static int _dpu_plane_get_aspace( + struct dpu_plane *pdpu, + struct dpu_plane_state *pstate, + struct msm_gem_address_space **aspace) +{ + struct dpu_kms *kms; + + if (!pdpu || !pstate || !aspace) { + DPU_ERROR("invalid parameters\n"); + return -EINVAL; + } + + kms = _dpu_plane_get_kms(&pdpu->base); + if (!kms) { + DPU_ERROR("invalid kms\n"); + return -EINVAL; + } + + *aspace = kms->base.aspace; + + return 0; +} + +static inline void _dpu_plane_set_scanout(struct drm_plane *plane, + struct dpu_plane_state *pstate, + struct dpu_hw_pipe_cfg *pipe_cfg, + struct drm_framebuffer *fb) +{ + struct dpu_plane *pdpu; + struct msm_gem_address_space *aspace = NULL; + int ret; + + if (!plane || !pstate || !pipe_cfg || !fb) { + DPU_ERROR( + "invalid arg(s), plane %d state %d cfg %d fb %d\n", + plane != 0, pstate != 0, pipe_cfg != 0, fb != 0); + return; + } + + pdpu = to_dpu_plane(plane); + if (!pdpu->pipe_hw) { + DPU_ERROR_PLANE(pdpu, "invalid pipe_hw\n"); + return; + } + + ret = _dpu_plane_get_aspace(pdpu, pstate, &aspace); + if (ret) { + DPU_ERROR_PLANE(pdpu, "Failed to get aspace %d\n", ret); + return; + } + + ret = dpu_format_populate_layout(aspace, fb, &pipe_cfg->layout); + if (ret == -EAGAIN) + DPU_DEBUG_PLANE(pdpu, "not updating same src addrs\n"); + else if (ret) + DPU_ERROR_PLANE(pdpu, "failed to get format layout, %d\n", ret); + else if (pdpu->pipe_hw->ops.setup_sourceaddress) { + trace_dpu_plane_set_scanout(pdpu->pipe_hw->idx, + &pipe_cfg->layout, + pstate->multirect_index); + pdpu->pipe_hw->ops.setup_sourceaddress(pdpu->pipe_hw, pipe_cfg, + pstate->multirect_index); + } +} + +static void _dpu_plane_setup_scaler3(struct dpu_plane *pdpu, + struct dpu_plane_state *pstate, + uint32_t src_w, uint32_t src_h, uint32_t dst_w, uint32_t dst_h, + struct dpu_hw_scaler3_cfg *scale_cfg, + const struct dpu_format *fmt, + uint32_t chroma_subsmpl_h, uint32_t chroma_subsmpl_v) +{ + uint32_t i; + + if (!pdpu || !pstate || !scale_cfg || !fmt || !chroma_subsmpl_h || + !chroma_subsmpl_v) { + DPU_ERROR( + "pdpu %d pstate %d scale_cfg %d fmt %d smp_h %d smp_v %d\n", + !!pdpu, !!pstate, !!scale_cfg, !!fmt, chroma_subsmpl_h, + chroma_subsmpl_v); + return; + } + + memset(scale_cfg, 0, sizeof(*scale_cfg)); + memset(&pstate->pixel_ext, 0, sizeof(struct dpu_hw_pixel_ext)); + + scale_cfg->phase_step_x[DPU_SSPP_COMP_0] = + mult_frac((1 << PHASE_STEP_SHIFT), src_w, dst_w); + scale_cfg->phase_step_y[DPU_SSPP_COMP_0] = + mult_frac((1 << PHASE_STEP_SHIFT), src_h, dst_h); + + + scale_cfg->phase_step_y[DPU_SSPP_COMP_1_2] = + scale_cfg->phase_step_y[DPU_SSPP_COMP_0] / chroma_subsmpl_v; + scale_cfg->phase_step_x[DPU_SSPP_COMP_1_2] = + scale_cfg->phase_step_x[DPU_SSPP_COMP_0] / chroma_subsmpl_h; + + scale_cfg->phase_step_x[DPU_SSPP_COMP_2] = + scale_cfg->phase_step_x[DPU_SSPP_COMP_1_2]; + scale_cfg->phase_step_y[DPU_SSPP_COMP_2] = + scale_cfg->phase_step_y[DPU_SSPP_COMP_1_2]; + + scale_cfg->phase_step_x[DPU_SSPP_COMP_3] = + scale_cfg->phase_step_x[DPU_SSPP_COMP_0]; + scale_cfg->phase_step_y[DPU_SSPP_COMP_3] = + scale_cfg->phase_step_y[DPU_SSPP_COMP_0]; + + for (i = 0; i < DPU_MAX_PLANES; i++) { + scale_cfg->src_width[i] = src_w; + scale_cfg->src_height[i] = src_h; + if (i == DPU_SSPP_COMP_1_2 || i == DPU_SSPP_COMP_2) { + scale_cfg->src_width[i] /= chroma_subsmpl_h; + scale_cfg->src_height[i] /= chroma_subsmpl_v; + } + scale_cfg->preload_x[i] = DPU_QSEED3_DEFAULT_PRELOAD_H; + scale_cfg->preload_y[i] = DPU_QSEED3_DEFAULT_PRELOAD_V; + pstate->pixel_ext.num_ext_pxls_top[i] = + scale_cfg->src_height[i]; + pstate->pixel_ext.num_ext_pxls_left[i] = + scale_cfg->src_width[i]; + } + if (!(DPU_FORMAT_IS_YUV(fmt)) && (src_h == dst_h) + && (src_w == dst_w)) + return; + + scale_cfg->dst_width = dst_w; + scale_cfg->dst_height = dst_h; + scale_cfg->y_rgb_filter_cfg = DPU_SCALE_BIL; + scale_cfg->uv_filter_cfg = DPU_SCALE_BIL; + scale_cfg->alpha_filter_cfg = DPU_SCALE_ALPHA_BIL; + scale_cfg->lut_flag = 0; + scale_cfg->blend_cfg = 1; + scale_cfg->enable = 1; +} + +static inline void _dpu_plane_setup_csc(struct dpu_plane *pdpu) +{ + static const struct dpu_csc_cfg dpu_csc_YUV2RGB_601L = { + { + /* S15.16 format */ + 0x00012A00, 0x00000000, 0x00019880, + 0x00012A00, 0xFFFF9B80, 0xFFFF3000, + 0x00012A00, 0x00020480, 0x00000000, + }, + /* signed bias */ + { 0xfff0, 0xff80, 0xff80,}, + { 0x0, 0x0, 0x0,}, + /* unsigned clamp */ + { 0x10, 0xeb, 0x10, 0xf0, 0x10, 0xf0,}, + { 0x00, 0xff, 0x00, 0xff, 0x00, 0xff,}, + }; + static const struct dpu_csc_cfg dpu_csc10_YUV2RGB_601L = { + { + /* S15.16 format */ + 0x00012A00, 0x00000000, 0x00019880, + 0x00012A00, 0xFFFF9B80, 0xFFFF3000, + 0x00012A00, 0x00020480, 0x00000000, + }, + /* signed bias */ + { 0xffc0, 0xfe00, 0xfe00,}, + { 0x0, 0x0, 0x0,}, + /* unsigned clamp */ + { 0x40, 0x3ac, 0x40, 0x3c0, 0x40, 0x3c0,}, + { 0x00, 0x3ff, 0x00, 0x3ff, 0x00, 0x3ff,}, + }; + + if (!pdpu) { + DPU_ERROR("invalid plane\n"); + return; + } + + if (BIT(DPU_SSPP_CSC_10BIT) & pdpu->features) + pdpu->csc_ptr = (struct dpu_csc_cfg *)&dpu_csc10_YUV2RGB_601L; + else + pdpu->csc_ptr = (struct dpu_csc_cfg *)&dpu_csc_YUV2RGB_601L; + + DPU_DEBUG_PLANE(pdpu, "using 0x%X 0x%X 0x%X...\n", + pdpu->csc_ptr->csc_mv[0], + pdpu->csc_ptr->csc_mv[1], + pdpu->csc_ptr->csc_mv[2]); +} + +static void _dpu_plane_setup_scaler(struct dpu_plane *pdpu, + struct dpu_plane_state *pstate, + const struct dpu_format *fmt, bool color_fill) +{ + struct dpu_hw_pixel_ext *pe; + uint32_t chroma_subsmpl_h, chroma_subsmpl_v; + + if (!pdpu || !fmt || !pstate) { + DPU_ERROR("invalid arg(s), plane %d fmt %d state %d\n", + pdpu != 0, fmt != 0, pstate != 0); + return; + } + + pe = &pstate->pixel_ext; + + /* don't chroma subsample if decimating */ + chroma_subsmpl_h = + drm_format_horz_chroma_subsampling(fmt->base.pixel_format); + chroma_subsmpl_v = + drm_format_vert_chroma_subsampling(fmt->base.pixel_format); + + /* update scaler. calculate default config for QSEED3 */ + _dpu_plane_setup_scaler3(pdpu, pstate, + drm_rect_width(&pdpu->pipe_cfg.src_rect), + drm_rect_height(&pdpu->pipe_cfg.src_rect), + drm_rect_width(&pdpu->pipe_cfg.dst_rect), + drm_rect_height(&pdpu->pipe_cfg.dst_rect), + &pstate->scaler3_cfg, fmt, + chroma_subsmpl_h, chroma_subsmpl_v); +} + +/** + * _dpu_plane_color_fill - enables color fill on plane + * @pdpu: Pointer to DPU plane object + * @color: RGB fill color value, [23..16] Blue, [15..8] Green, [7..0] Red + * @alpha: 8-bit fill alpha value, 255 selects 100% alpha + * Returns: 0 on success + */ +static int _dpu_plane_color_fill(struct dpu_plane *pdpu, + uint32_t color, uint32_t alpha) +{ + const struct dpu_format *fmt; + const struct drm_plane *plane; + struct dpu_plane_state *pstate; + + if (!pdpu || !pdpu->base.state) { + DPU_ERROR("invalid plane\n"); + return -EINVAL; + } + + if (!pdpu->pipe_hw) { + DPU_ERROR_PLANE(pdpu, "invalid plane h/w pointer\n"); + return -EINVAL; + } + + plane = &pdpu->base; + pstate = to_dpu_plane_state(plane->state); + + DPU_DEBUG_PLANE(pdpu, "\n"); + + /* + * select fill format to match user property expectation, + * h/w only supports RGB variants + */ + fmt = dpu_get_dpu_format(DRM_FORMAT_ABGR8888); + + /* update sspp */ + if (fmt && pdpu->pipe_hw->ops.setup_solidfill) { + pdpu->pipe_hw->ops.setup_solidfill(pdpu->pipe_hw, + (color & 0xFFFFFF) | ((alpha & 0xFF) << 24), + pstate->multirect_index); + + /* override scaler/decimation if solid fill */ + pdpu->pipe_cfg.src_rect.x1 = 0; + pdpu->pipe_cfg.src_rect.y1 = 0; + pdpu->pipe_cfg.src_rect.x2 = + drm_rect_width(&pdpu->pipe_cfg.dst_rect); + pdpu->pipe_cfg.src_rect.y2 = + drm_rect_height(&pdpu->pipe_cfg.dst_rect); + _dpu_plane_setup_scaler(pdpu, pstate, fmt, true); + + if (pdpu->pipe_hw->ops.setup_format) + pdpu->pipe_hw->ops.setup_format(pdpu->pipe_hw, + fmt, DPU_SSPP_SOLID_FILL, + pstate->multirect_index); + + if (pdpu->pipe_hw->ops.setup_rects) + pdpu->pipe_hw->ops.setup_rects(pdpu->pipe_hw, + &pdpu->pipe_cfg, + pstate->multirect_index); + + if (pdpu->pipe_hw->ops.setup_pe) + pdpu->pipe_hw->ops.setup_pe(pdpu->pipe_hw, + &pstate->pixel_ext); + + if (pdpu->pipe_hw->ops.setup_scaler && + pstate->multirect_index != DPU_SSPP_RECT_1) + pdpu->pipe_hw->ops.setup_scaler(pdpu->pipe_hw, + &pdpu->pipe_cfg, &pstate->pixel_ext, + &pstate->scaler3_cfg); + } + + return 0; +} + +void dpu_plane_clear_multirect(const struct drm_plane_state *drm_state) +{ + struct dpu_plane_state *pstate; + + if (!drm_state) + return; + + pstate = to_dpu_plane_state(drm_state); + + pstate->multirect_index = DPU_SSPP_RECT_SOLO; + pstate->multirect_mode = DPU_SSPP_MULTIRECT_NONE; +} + +int dpu_plane_validate_multirect_v2(struct dpu_multirect_plane_states *plane) +{ + struct dpu_plane_state *pstate[R_MAX]; + const struct drm_plane_state *drm_state[R_MAX]; + struct drm_rect src[R_MAX], dst[R_MAX]; + struct dpu_plane *dpu_plane[R_MAX]; + const struct dpu_format *fmt[R_MAX]; + int i, buffer_lines; + unsigned int max_tile_height = 1; + bool parallel_fetch_qualified = true; + bool has_tiled_rect = false; + + for (i = 0; i < R_MAX; i++) { + const struct msm_format *msm_fmt; + + drm_state[i] = i ? plane->r1 : plane->r0; + msm_fmt = msm_framebuffer_format(drm_state[i]->fb); + fmt[i] = to_dpu_format(msm_fmt); + + if (DPU_FORMAT_IS_UBWC(fmt[i])) { + has_tiled_rect = true; + if (fmt[i]->tile_height > max_tile_height) + max_tile_height = fmt[i]->tile_height; + } + } + + for (i = 0; i < R_MAX; i++) { + int width_threshold; + + pstate[i] = to_dpu_plane_state(drm_state[i]); + dpu_plane[i] = to_dpu_plane(drm_state[i]->plane); + + if (pstate[i] == NULL) { + DPU_ERROR("DPU plane state of plane id %d is NULL\n", + drm_state[i]->plane->base.id); + return -EINVAL; + } + + src[i].x1 = drm_state[i]->src_x >> 16; + src[i].y1 = drm_state[i]->src_y >> 16; + src[i].x2 = src[i].x1 + (drm_state[i]->src_w >> 16); + src[i].y2 = src[i].y1 + (drm_state[i]->src_h >> 16); + + dst[i] = drm_plane_state_dest(drm_state[i]); + + if (drm_rect_calc_hscale(&src[i], &dst[i], 1, 1) != 1 || + drm_rect_calc_vscale(&src[i], &dst[i], 1, 1) != 1) { + DPU_ERROR_PLANE(dpu_plane[i], + "scaling is not supported in multirect mode\n"); + return -EINVAL; + } + + if (DPU_FORMAT_IS_YUV(fmt[i])) { + DPU_ERROR_PLANE(dpu_plane[i], + "Unsupported format for multirect mode\n"); + return -EINVAL; + } + + /** + * SSPP PD_MEM is split half - one for each RECT. + * Tiled formats need 5 lines of buffering while fetching + * whereas linear formats need only 2 lines. + * So we cannot support more than half of the supported SSPP + * width for tiled formats. + */ + width_threshold = dpu_plane[i]->pipe_sblk->common->maxlinewidth; + if (has_tiled_rect) + width_threshold /= 2; + + if (parallel_fetch_qualified && + drm_rect_width(&src[i]) > width_threshold) + parallel_fetch_qualified = false; + + } + + /* Validate RECT's and set the mode */ + + /* Prefer PARALLEL FETCH Mode over TIME_MX Mode */ + if (parallel_fetch_qualified) { + pstate[R0]->multirect_mode = DPU_SSPP_MULTIRECT_PARALLEL; + pstate[R1]->multirect_mode = DPU_SSPP_MULTIRECT_PARALLEL; + + goto done; + } + + /* TIME_MX Mode */ + buffer_lines = 2 * max_tile_height; + + if (dst[R1].y1 >= dst[R0].y2 + buffer_lines || + dst[R0].y1 >= dst[R1].y2 + buffer_lines) { + pstate[R0]->multirect_mode = DPU_SSPP_MULTIRECT_TIME_MX; + pstate[R1]->multirect_mode = DPU_SSPP_MULTIRECT_TIME_MX; + } else { + DPU_ERROR( + "No multirect mode possible for the planes (%d - %d)\n", + drm_state[R0]->plane->base.id, + drm_state[R1]->plane->base.id); + return -EINVAL; + } + +done: + if (dpu_plane[R0]->is_virtual) { + pstate[R0]->multirect_index = DPU_SSPP_RECT_1; + pstate[R1]->multirect_index = DPU_SSPP_RECT_0; + } else { + pstate[R0]->multirect_index = DPU_SSPP_RECT_0; + pstate[R1]->multirect_index = DPU_SSPP_RECT_1; + }; + + DPU_DEBUG_PLANE(dpu_plane[R0], "R0: %d - %d\n", + pstate[R0]->multirect_mode, pstate[R0]->multirect_index); + DPU_DEBUG_PLANE(dpu_plane[R1], "R1: %d - %d\n", + pstate[R1]->multirect_mode, pstate[R1]->multirect_index); + return 0; +} + +/** + * dpu_plane_get_ctl_flush - get control flush for the given plane + * @plane: Pointer to drm plane structure + * @ctl: Pointer to hardware control driver + * @flush_sspp: Pointer to sspp flush control word + */ +void dpu_plane_get_ctl_flush(struct drm_plane *plane, struct dpu_hw_ctl *ctl, + u32 *flush_sspp) +{ + struct dpu_plane_state *pstate; + + if (!plane || !flush_sspp) { + DPU_ERROR("invalid parameters\n"); + return; + } + + pstate = to_dpu_plane_state(plane->state); + + *flush_sspp = ctl->ops.get_bitmask_sspp(ctl, dpu_plane_pipe(plane)); +} + +static int dpu_plane_prepare_fb(struct drm_plane *plane, + struct drm_plane_state *new_state) +{ + struct drm_framebuffer *fb = new_state->fb; + struct dpu_plane *pdpu = to_dpu_plane(plane); + struct dpu_plane_state *pstate = to_dpu_plane_state(new_state); + struct dpu_hw_fmt_layout layout; + struct drm_gem_object *obj; + struct msm_gem_object *msm_obj; + struct dma_fence *fence; + struct msm_gem_address_space *aspace; + int ret; + + if (!new_state->fb) + return 0; + + DPU_DEBUG_PLANE(pdpu, "FB[%u]\n", fb->base.id); + + ret = _dpu_plane_get_aspace(pdpu, pstate, &aspace); + if (ret) { + DPU_ERROR_PLANE(pdpu, "Failed to get aspace\n"); + return ret; + } + + /* cache aspace */ + pstate->aspace = aspace; + + /* + * TODO: Need to sort out the msm_framebuffer_prepare() call below so + * we can use msm_atomic_prepare_fb() instead of doing the + * implicit fence and fb prepare by hand here. + */ + obj = msm_framebuffer_bo(new_state->fb, 0); + msm_obj = to_msm_bo(obj); + fence = reservation_object_get_excl_rcu(msm_obj->resv); + if (fence) + drm_atomic_set_fence_for_plane(new_state, fence); + + if (pstate->aspace) { + ret = msm_framebuffer_prepare(new_state->fb, + pstate->aspace); + if (ret) { + DPU_ERROR("failed to prepare framebuffer\n"); + return ret; + } + } + + /* validate framebuffer layout before commit */ + ret = dpu_format_populate_layout(pstate->aspace, + new_state->fb, &layout); + if (ret) { + DPU_ERROR_PLANE(pdpu, "failed to get format layout, %d\n", ret); + return ret; + } + + return 0; +} + +static void dpu_plane_cleanup_fb(struct drm_plane *plane, + struct drm_plane_state *old_state) +{ + struct dpu_plane *pdpu = to_dpu_plane(plane); + struct dpu_plane_state *old_pstate; + + if (!old_state || !old_state->fb) + return; + + old_pstate = to_dpu_plane_state(old_state); + + DPU_DEBUG_PLANE(pdpu, "FB[%u]\n", old_state->fb->base.id); + + msm_framebuffer_cleanup(old_state->fb, old_pstate->aspace); +} + +static bool dpu_plane_validate_src(struct drm_rect *src, + struct drm_rect *fb_rect, + uint32_t min_src_size) +{ + /* Ensure fb size is supported */ + if (drm_rect_width(fb_rect) > MAX_IMG_WIDTH || + drm_rect_height(fb_rect) > MAX_IMG_HEIGHT) + return false; + + /* Ensure src rect is above the minimum size */ + if (drm_rect_width(src) < min_src_size || + drm_rect_height(src) < min_src_size) + return false; + + /* Ensure src is fully encapsulated in fb */ + return drm_rect_intersect(fb_rect, src) && + drm_rect_equals(fb_rect, src); +} + +static int dpu_plane_sspp_atomic_check(struct drm_plane *plane, + struct drm_plane_state *state) +{ + int ret = 0; + struct dpu_plane *pdpu; + struct dpu_plane_state *pstate; + const struct dpu_format *fmt; + struct drm_rect src, dst, fb_rect = { 0 }; + uint32_t max_upscale = 1, max_downscale = 1; + uint32_t min_src_size, max_linewidth; + int hscale = 1, vscale = 1; + + if (!plane || !state) { + DPU_ERROR("invalid arg(s), plane %d state %d\n", + plane != 0, state != 0); + ret = -EINVAL; + goto exit; + } + + pdpu = to_dpu_plane(plane); + pstate = to_dpu_plane_state(state); + + if (!pdpu->pipe_sblk) { + DPU_ERROR_PLANE(pdpu, "invalid catalog\n"); + ret = -EINVAL; + goto exit; + } + + src.x1 = state->src_x >> 16; + src.y1 = state->src_y >> 16; + src.x2 = src.x1 + (state->src_w >> 16); + src.y2 = src.y1 + (state->src_h >> 16); + + dst = drm_plane_state_dest(state); + + fb_rect.x2 = state->fb->width; + fb_rect.y2 = state->fb->height; + + max_linewidth = pdpu->pipe_sblk->common->maxlinewidth; + + if (pdpu->features & DPU_SSPP_SCALER) { + max_downscale = pdpu->pipe_sblk->maxdwnscale; + max_upscale = pdpu->pipe_sblk->maxupscale; + } + if (drm_rect_width(&src) < drm_rect_width(&dst)) + hscale = drm_rect_calc_hscale(&src, &dst, 1, max_upscale); + else + hscale = drm_rect_calc_hscale(&dst, &src, 1, max_downscale); + if (drm_rect_height(&src) < drm_rect_height(&dst)) + vscale = drm_rect_calc_vscale(&src, &dst, 1, max_upscale); + else + vscale = drm_rect_calc_vscale(&dst, &src, 1, max_downscale); + + DPU_DEBUG_PLANE(pdpu, "check %d -> %d\n", + dpu_plane_enabled(plane->state), dpu_plane_enabled(state)); + + if (!dpu_plane_enabled(state)) + goto exit; + + fmt = to_dpu_format(msm_framebuffer_format(state->fb)); + + min_src_size = DPU_FORMAT_IS_YUV(fmt) ? 2 : 1; + + if (DPU_FORMAT_IS_YUV(fmt) && + (!(pdpu->features & DPU_SSPP_SCALER) || + !(pdpu->features & (BIT(DPU_SSPP_CSC) + | BIT(DPU_SSPP_CSC_10BIT))))) { + DPU_ERROR_PLANE(pdpu, + "plane doesn't have scaler/csc for yuv\n"); + ret = -EINVAL; + + /* check src bounds */ + } else if (!dpu_plane_validate_src(&src, &fb_rect, min_src_size)) { + DPU_ERROR_PLANE(pdpu, "invalid source " DRM_RECT_FMT "\n", + DRM_RECT_ARG(&src)); + ret = -E2BIG; + + /* valid yuv image */ + } else if (DPU_FORMAT_IS_YUV(fmt) && + (src.x1 & 0x1 || src.y1 & 0x1 || + drm_rect_width(&src) & 0x1 || + drm_rect_height(&src) & 0x1)) { + DPU_ERROR_PLANE(pdpu, "invalid yuv source " DRM_RECT_FMT "\n", + DRM_RECT_ARG(&src)); + ret = -EINVAL; + + /* min dst support */ + } else if (drm_rect_width(&dst) < 0x1 || drm_rect_height(&dst) < 0x1) { + DPU_ERROR_PLANE(pdpu, "invalid dest rect " DRM_RECT_FMT "\n", + DRM_RECT_ARG(&dst)); + ret = -EINVAL; + + /* check decimated source width */ + } else if (drm_rect_width(&src) > max_linewidth) { + DPU_ERROR_PLANE(pdpu, "invalid src " DRM_RECT_FMT " line:%u\n", + DRM_RECT_ARG(&src), max_linewidth); + ret = -E2BIG; + + /* check scaler capability */ + } else if (hscale < 0 || vscale < 0) { + DPU_ERROR_PLANE(pdpu, "invalid scaling requested src=" + DRM_RECT_FMT " dst=" DRM_RECT_FMT "\n", + DRM_RECT_ARG(&src), DRM_RECT_ARG(&dst)); + ret = -E2BIG; + } + +exit: + return ret; +} + +static int dpu_plane_atomic_check(struct drm_plane *plane, + struct drm_plane_state *state) +{ + if (!state->fb) + return 0; + + DPU_DEBUG_PLANE(to_dpu_plane(plane), "\n"); + + return dpu_plane_sspp_atomic_check(plane, state); +} + +void dpu_plane_flush(struct drm_plane *plane) +{ + struct dpu_plane *pdpu; + struct dpu_plane_state *pstate; + + if (!plane || !plane->state) { + DPU_ERROR("invalid plane\n"); + return; + } + + pdpu = to_dpu_plane(plane); + pstate = to_dpu_plane_state(plane->state); + + /* + * These updates have to be done immediately before the plane flush + * timing, and may not be moved to the atomic_update/mode_set functions. + */ + if (pdpu->is_error) + /* force white frame with 100% alpha pipe output on error */ + _dpu_plane_color_fill(pdpu, 0xFFFFFF, 0xFF); + else if (pdpu->color_fill & DPU_PLANE_COLOR_FILL_FLAG) + /* force 100% alpha */ + _dpu_plane_color_fill(pdpu, pdpu->color_fill, 0xFF); + else if (pdpu->pipe_hw && pdpu->csc_ptr && pdpu->pipe_hw->ops.setup_csc) + pdpu->pipe_hw->ops.setup_csc(pdpu->pipe_hw, pdpu->csc_ptr); + + /* flag h/w flush complete */ + if (plane->state) + pstate->pending = false; +} + +/** + * dpu_plane_set_error: enable/disable error condition + * @plane: pointer to drm_plane structure + */ +void dpu_plane_set_error(struct drm_plane *plane, bool error) +{ + struct dpu_plane *pdpu; + + if (!plane) + return; + + pdpu = to_dpu_plane(plane); + pdpu->is_error = error; +} + +static int dpu_plane_sspp_atomic_update(struct drm_plane *plane, + struct drm_plane_state *old_state) +{ + uint32_t nplanes, src_flags; + struct dpu_plane *pdpu; + struct drm_plane_state *state; + struct dpu_plane_state *pstate; + struct dpu_plane_state *old_pstate; + const struct dpu_format *fmt; + struct drm_crtc *crtc; + struct drm_framebuffer *fb; + struct drm_rect src, dst; + + if (!plane) { + DPU_ERROR("invalid plane\n"); + return -EINVAL; + } else if (!plane->state) { + DPU_ERROR("invalid plane state\n"); + return -EINVAL; + } else if (!old_state) { + DPU_ERROR("invalid old state\n"); + return -EINVAL; + } + + pdpu = to_dpu_plane(plane); + state = plane->state; + + pstate = to_dpu_plane_state(state); + + old_pstate = to_dpu_plane_state(old_state); + + crtc = state->crtc; + fb = state->fb; + if (!crtc || !fb) { + DPU_ERROR_PLANE(pdpu, "invalid crtc %d or fb %d\n", + crtc != 0, fb != 0); + return -EINVAL; + } + fmt = to_dpu_format(msm_framebuffer_format(fb)); + nplanes = fmt->num_planes; + + memset(&(pdpu->pipe_cfg), 0, sizeof(struct dpu_hw_pipe_cfg)); + + _dpu_plane_set_scanout(plane, pstate, &pdpu->pipe_cfg, fb); + + pstate->pending = true; + + pdpu->is_rt_pipe = (dpu_crtc_get_client_type(crtc) != NRT_CLIENT); + _dpu_plane_set_qos_ctrl(plane, false, DPU_PLANE_QOS_PANIC_CTRL); + + src.x1 = state->src_x >> 16; + src.y1 = state->src_y >> 16; + src.x2 = src.x1 + (state->src_w >> 16); + src.y2 = src.y1 + (state->src_h >> 16); + + dst = drm_plane_state_dest(state); + + DPU_DEBUG_PLANE(pdpu, "FB[%u] " DRM_RECT_FMT "->crtc%u " DRM_RECT_FMT + ", %4.4s ubwc %d\n", fb->base.id, DRM_RECT_ARG(&src), + crtc->base.id, DRM_RECT_ARG(&dst), + (char *)&fmt->base.pixel_format, + DPU_FORMAT_IS_UBWC(fmt)); + + pdpu->pipe_cfg.src_rect = src; + pdpu->pipe_cfg.dst_rect = dst; + + _dpu_plane_setup_scaler(pdpu, pstate, fmt, false); + + /* override for color fill */ + if (pdpu->color_fill & DPU_PLANE_COLOR_FILL_FLAG) { + /* skip remaining processing on color fill */ + return 0; + } + + if (pdpu->pipe_hw->ops.setup_rects) { + pdpu->pipe_hw->ops.setup_rects(pdpu->pipe_hw, + &pdpu->pipe_cfg, + pstate->multirect_index); + } + + if (pdpu->pipe_hw->ops.setup_pe && + (pstate->multirect_index != DPU_SSPP_RECT_1)) + pdpu->pipe_hw->ops.setup_pe(pdpu->pipe_hw, + &pstate->pixel_ext); + + /** + * when programmed in multirect mode, scalar block will be + * bypassed. Still we need to update alpha and bitwidth + * ONLY for RECT0 + */ + if (pdpu->pipe_hw->ops.setup_scaler && + pstate->multirect_index != DPU_SSPP_RECT_1) + pdpu->pipe_hw->ops.setup_scaler(pdpu->pipe_hw, + &pdpu->pipe_cfg, &pstate->pixel_ext, + &pstate->scaler3_cfg); + + if (pdpu->pipe_hw->ops.setup_multirect) + pdpu->pipe_hw->ops.setup_multirect( + pdpu->pipe_hw, + pstate->multirect_index, + pstate->multirect_mode); + + if (pdpu->pipe_hw->ops.setup_format) { + src_flags = 0x0; + + /* update format */ + pdpu->pipe_hw->ops.setup_format(pdpu->pipe_hw, fmt, src_flags, + pstate->multirect_index); + + if (pdpu->pipe_hw->ops.setup_cdp) { + struct dpu_hw_pipe_cdp_cfg *cdp_cfg = &pstate->cdp_cfg; + + memset(cdp_cfg, 0, sizeof(struct dpu_hw_pipe_cdp_cfg)); + + cdp_cfg->enable = pdpu->catalog->perf.cdp_cfg + [DPU_PERF_CDP_USAGE_RT].rd_enable; + cdp_cfg->ubwc_meta_enable = + DPU_FORMAT_IS_UBWC(fmt); + cdp_cfg->tile_amortize_enable = + DPU_FORMAT_IS_UBWC(fmt) || + DPU_FORMAT_IS_TILE(fmt); + cdp_cfg->preload_ahead = DPU_SSPP_CDP_PRELOAD_AHEAD_64; + + pdpu->pipe_hw->ops.setup_cdp(pdpu->pipe_hw, cdp_cfg); + } + + /* update csc */ + if (DPU_FORMAT_IS_YUV(fmt)) + _dpu_plane_setup_csc(pdpu); + else + pdpu->csc_ptr = 0; + } + + _dpu_plane_set_qos_lut(plane, fb); + _dpu_plane_set_danger_lut(plane, fb); + + if (plane->type != DRM_PLANE_TYPE_CURSOR) { + _dpu_plane_set_qos_ctrl(plane, true, DPU_PLANE_QOS_PANIC_CTRL); + _dpu_plane_set_ot_limit(plane, crtc); + } + + _dpu_plane_set_qos_remap(plane); + return 0; +} + +static void _dpu_plane_atomic_disable(struct drm_plane *plane, + struct drm_plane_state *old_state) +{ + struct dpu_plane *pdpu; + struct drm_plane_state *state; + struct dpu_plane_state *pstate; + + if (!plane) { + DPU_ERROR("invalid plane\n"); + return; + } else if (!plane->state) { + DPU_ERROR("invalid plane state\n"); + return; + } else if (!old_state) { + DPU_ERROR("invalid old state\n"); + return; + } + + pdpu = to_dpu_plane(plane); + state = plane->state; + pstate = to_dpu_plane_state(state); + + trace_dpu_plane_disable(DRMID(plane), is_dpu_plane_virtual(plane), + pstate->multirect_mode); + + pstate->pending = true; + + if (is_dpu_plane_virtual(plane) && + pdpu->pipe_hw && pdpu->pipe_hw->ops.setup_multirect) + pdpu->pipe_hw->ops.setup_multirect(pdpu->pipe_hw, + DPU_SSPP_RECT_SOLO, DPU_SSPP_MULTIRECT_NONE); +} + +static void dpu_plane_atomic_update(struct drm_plane *plane, + struct drm_plane_state *old_state) +{ + struct dpu_plane *pdpu; + struct drm_plane_state *state; + + if (!plane) { + DPU_ERROR("invalid plane\n"); + return; + } else if (!plane->state) { + DPU_ERROR("invalid plane state\n"); + return; + } + + pdpu = to_dpu_plane(plane); + pdpu->is_error = false; + state = plane->state; + + DPU_DEBUG_PLANE(pdpu, "\n"); + + if (!dpu_plane_sspp_enabled(state)) { + _dpu_plane_atomic_disable(plane, old_state); + } else { + int ret; + + ret = dpu_plane_sspp_atomic_update(plane, old_state); + /* atomic_check should have ensured that this doesn't fail */ + WARN_ON(ret < 0); + } +} + +void dpu_plane_restore(struct drm_plane *plane) +{ + struct dpu_plane *pdpu; + + if (!plane || !plane->state) { + DPU_ERROR("invalid plane\n"); + return; + } + + pdpu = to_dpu_plane(plane); + + DPU_DEBUG_PLANE(pdpu, "\n"); + + /* last plane state is same as current state */ + dpu_plane_atomic_update(plane, plane->state); +} + +static void dpu_plane_destroy(struct drm_plane *plane) +{ + struct dpu_plane *pdpu = plane ? to_dpu_plane(plane) : NULL; + + DPU_DEBUG_PLANE(pdpu, "\n"); + + if (pdpu) { + _dpu_plane_set_qos_ctrl(plane, false, DPU_PLANE_QOS_PANIC_CTRL); + + mutex_destroy(&pdpu->lock); + + drm_plane_helper_disable(plane, NULL); + + /* this will destroy the states as well */ + drm_plane_cleanup(plane); + + if (pdpu->pipe_hw) + dpu_hw_sspp_destroy(pdpu->pipe_hw); + + kfree(pdpu); + } +} + +static void dpu_plane_destroy_state(struct drm_plane *plane, + struct drm_plane_state *state) +{ + struct dpu_plane_state *pstate; + + if (!plane || !state) { + DPU_ERROR("invalid arg(s), plane %d state %d\n", + plane != 0, state != 0); + return; + } + + pstate = to_dpu_plane_state(state); + + /* remove ref count for frame buffers */ + if (state->fb) + drm_framebuffer_put(state->fb); + + kfree(pstate); +} + +static struct drm_plane_state * +dpu_plane_duplicate_state(struct drm_plane *plane) +{ + struct dpu_plane *pdpu; + struct dpu_plane_state *pstate; + struct dpu_plane_state *old_state; + + if (!plane) { + DPU_ERROR("invalid plane\n"); + return NULL; + } else if (!plane->state) { + DPU_ERROR("invalid plane state\n"); + return NULL; + } + + old_state = to_dpu_plane_state(plane->state); + pdpu = to_dpu_plane(plane); + pstate = kmemdup(old_state, sizeof(*old_state), GFP_KERNEL); + if (!pstate) { + DPU_ERROR_PLANE(pdpu, "failed to allocate state\n"); + return NULL; + } + + DPU_DEBUG_PLANE(pdpu, "\n"); + + pstate->pending = false; + + __drm_atomic_helper_plane_duplicate_state(plane, &pstate->base); + + return &pstate->base; +} + +static void dpu_plane_reset(struct drm_plane *plane) +{ + struct dpu_plane *pdpu; + struct dpu_plane_state *pstate; + + if (!plane) { + DPU_ERROR("invalid plane\n"); + return; + } + + pdpu = to_dpu_plane(plane); + DPU_DEBUG_PLANE(pdpu, "\n"); + + /* remove previous state, if present */ + if (plane->state) { + dpu_plane_destroy_state(plane, plane->state); + plane->state = 0; + } + + pstate = kzalloc(sizeof(*pstate), GFP_KERNEL); + if (!pstate) { + DPU_ERROR_PLANE(pdpu, "failed to allocate state\n"); + return; + } + + pstate->base.plane = plane; + + plane->state = &pstate->base; +} + +#ifdef CONFIG_DEBUG_FS +static ssize_t _dpu_plane_danger_read(struct file *file, + char __user *buff, size_t count, loff_t *ppos) +{ + struct dpu_kms *kms = file->private_data; + struct dpu_mdss_cfg *cfg = kms->catalog; + int len = 0; + char buf[40] = {'\0'}; + + if (!cfg) + return -ENODEV; + + if (*ppos) + return 0; /* the end */ + + len = snprintf(buf, sizeof(buf), "%d\n", !kms->has_danger_ctrl); + if (len < 0 || len >= sizeof(buf)) + return 0; + + if ((count < sizeof(buf)) || copy_to_user(buff, buf, len)) + return -EFAULT; + + *ppos += len; /* increase offset */ + + return len; +} + +static void _dpu_plane_set_danger_state(struct dpu_kms *kms, bool enable) +{ + struct drm_plane *plane; + + drm_for_each_plane(plane, kms->dev) { + if (plane->fb && plane->state) { + dpu_plane_danger_signal_ctrl(plane, enable); + DPU_DEBUG("plane:%d img:%dx%d ", + plane->base.id, plane->fb->width, + plane->fb->height); + DPU_DEBUG("src[%d,%d,%d,%d] dst[%d,%d,%d,%d]\n", + plane->state->src_x >> 16, + plane->state->src_y >> 16, + plane->state->src_w >> 16, + plane->state->src_h >> 16, + plane->state->crtc_x, plane->state->crtc_y, + plane->state->crtc_w, plane->state->crtc_h); + } else { + DPU_DEBUG("Inactive plane:%d\n", plane->base.id); + } + } +} + +static ssize_t _dpu_plane_danger_write(struct file *file, + const char __user *user_buf, size_t count, loff_t *ppos) +{ + struct dpu_kms *kms = file->private_data; + struct dpu_mdss_cfg *cfg = kms->catalog; + int disable_panic; + char buf[10]; + + if (!cfg) + return -EFAULT; + + if (count >= sizeof(buf)) + return -EFAULT; + + if (copy_from_user(buf, user_buf, count)) + return -EFAULT; + + buf[count] = 0; /* end of string */ + + if (kstrtoint(buf, 0, &disable_panic)) + return -EFAULT; + + if (disable_panic) { + /* Disable panic signal for all active pipes */ + DPU_DEBUG("Disabling danger:\n"); + _dpu_plane_set_danger_state(kms, false); + kms->has_danger_ctrl = false; + } else { + /* Enable panic signal for all active pipes */ + DPU_DEBUG("Enabling danger:\n"); + kms->has_danger_ctrl = true; + _dpu_plane_set_danger_state(kms, true); + } + + return count; +} + +static const struct file_operations dpu_plane_danger_enable = { + .open = simple_open, + .read = _dpu_plane_danger_read, + .write = _dpu_plane_danger_write, +}; + +static int _dpu_plane_init_debugfs(struct drm_plane *plane) +{ + struct dpu_plane *pdpu; + struct dpu_kms *kms; + struct msm_drm_private *priv; + const struct dpu_sspp_sub_blks *sblk = 0; + const struct dpu_sspp_cfg *cfg = 0; + + if (!plane || !plane->dev) { + DPU_ERROR("invalid arguments\n"); + return -EINVAL; + } + + priv = plane->dev->dev_private; + if (!priv || !priv->kms) { + DPU_ERROR("invalid KMS reference\n"); + return -EINVAL; + } + + kms = to_dpu_kms(priv->kms); + pdpu = to_dpu_plane(plane); + + if (pdpu && pdpu->pipe_hw) + cfg = pdpu->pipe_hw->cap; + if (cfg) + sblk = cfg->sblk; + + if (!sblk) + return 0; + + /* create overall sub-directory for the pipe */ + pdpu->debugfs_root = + debugfs_create_dir(pdpu->pipe_name, + plane->dev->primary->debugfs_root); + + if (!pdpu->debugfs_root) + return -ENOMEM; + + /* don't error check these */ + debugfs_create_x32("features", 0600, + pdpu->debugfs_root, &pdpu->features); + + /* add register dump support */ + dpu_debugfs_setup_regset32(&pdpu->debugfs_src, + sblk->src_blk.base + cfg->base, + sblk->src_blk.len, + kms); + dpu_debugfs_create_regset32("src_blk", 0400, + pdpu->debugfs_root, &pdpu->debugfs_src); + + if (cfg->features & BIT(DPU_SSPP_SCALER_QSEED3) || + cfg->features & BIT(DPU_SSPP_SCALER_QSEED2)) { + dpu_debugfs_setup_regset32(&pdpu->debugfs_scaler, + sblk->scaler_blk.base + cfg->base, + sblk->scaler_blk.len, + kms); + dpu_debugfs_create_regset32("scaler_blk", 0400, + pdpu->debugfs_root, + &pdpu->debugfs_scaler); + debugfs_create_bool("default_scaling", + 0600, + pdpu->debugfs_root, + &pdpu->debugfs_default_scale); + } + + if (cfg->features & BIT(DPU_SSPP_CSC) || + cfg->features & BIT(DPU_SSPP_CSC_10BIT)) { + dpu_debugfs_setup_regset32(&pdpu->debugfs_csc, + sblk->csc_blk.base + cfg->base, + sblk->csc_blk.len, + kms); + dpu_debugfs_create_regset32("csc_blk", 0400, + pdpu->debugfs_root, &pdpu->debugfs_csc); + } + + debugfs_create_u32("xin_id", + 0400, + pdpu->debugfs_root, + (u32 *) &cfg->xin_id); + debugfs_create_u32("clk_ctrl", + 0400, + pdpu->debugfs_root, + (u32 *) &cfg->clk_ctrl); + debugfs_create_x32("creq_vblank", + 0600, + pdpu->debugfs_root, + (u32 *) &sblk->creq_vblank); + debugfs_create_x32("danger_vblank", + 0600, + pdpu->debugfs_root, + (u32 *) &sblk->danger_vblank); + + debugfs_create_file("disable_danger", + 0600, + pdpu->debugfs_root, + kms, &dpu_plane_danger_enable); + + return 0; +} + +static void _dpu_plane_destroy_debugfs(struct drm_plane *plane) +{ + struct dpu_plane *pdpu; + + if (!plane) + return; + pdpu = to_dpu_plane(plane); + + debugfs_remove_recursive(pdpu->debugfs_root); +} +#else +static int _dpu_plane_init_debugfs(struct drm_plane *plane) +{ + return 0; +} +static void _dpu_plane_destroy_debugfs(struct drm_plane *plane) +{ +} +#endif + +static int dpu_plane_late_register(struct drm_plane *plane) +{ + return _dpu_plane_init_debugfs(plane); +} + +static void dpu_plane_early_unregister(struct drm_plane *plane) +{ + _dpu_plane_destroy_debugfs(plane); +} + +static const struct drm_plane_funcs dpu_plane_funcs = { + .update_plane = drm_atomic_helper_update_plane, + .disable_plane = drm_atomic_helper_disable_plane, + .destroy = dpu_plane_destroy, + .reset = dpu_plane_reset, + .atomic_duplicate_state = dpu_plane_duplicate_state, + .atomic_destroy_state = dpu_plane_destroy_state, + .late_register = dpu_plane_late_register, + .early_unregister = dpu_plane_early_unregister, +}; + +static const struct drm_plane_helper_funcs dpu_plane_helper_funcs = { + .prepare_fb = dpu_plane_prepare_fb, + .cleanup_fb = dpu_plane_cleanup_fb, + .atomic_check = dpu_plane_atomic_check, + .atomic_update = dpu_plane_atomic_update, +}; + +enum dpu_sspp dpu_plane_pipe(struct drm_plane *plane) +{ + return plane ? to_dpu_plane(plane)->pipe : SSPP_NONE; +} + +bool is_dpu_plane_virtual(struct drm_plane *plane) +{ + return plane ? to_dpu_plane(plane)->is_virtual : false; +} + +/* initialize plane */ +struct drm_plane *dpu_plane_init(struct drm_device *dev, + uint32_t pipe, bool primary_plane, + unsigned long possible_crtcs, u32 master_plane_id) +{ + struct drm_plane *plane = NULL, *master_plane = NULL; + const struct dpu_format_extended *format_list; + struct dpu_plane *pdpu; + struct msm_drm_private *priv; + struct dpu_kms *kms; + enum drm_plane_type type; + int zpos_max = DPU_ZPOS_MAX; + int ret = -EINVAL; + + if (!dev) { + DPU_ERROR("[%u]device is NULL\n", pipe); + goto exit; + } + + priv = dev->dev_private; + if (!priv) { + DPU_ERROR("[%u]private data is NULL\n", pipe); + goto exit; + } + + if (!priv->kms) { + DPU_ERROR("[%u]invalid KMS reference\n", pipe); + goto exit; + } + kms = to_dpu_kms(priv->kms); + + if (!kms->catalog) { + DPU_ERROR("[%u]invalid catalog reference\n", pipe); + goto exit; + } + + /* create and zero local structure */ + pdpu = kzalloc(sizeof(*pdpu), GFP_KERNEL); + if (!pdpu) { + DPU_ERROR("[%u]failed to allocate local plane struct\n", pipe); + ret = -ENOMEM; + goto exit; + } + + /* cache local stuff for later */ + plane = &pdpu->base; + pdpu->pipe = pipe; + pdpu->is_virtual = (master_plane_id != 0); + INIT_LIST_HEAD(&pdpu->mplane_list); + master_plane = drm_plane_find(dev, NULL, master_plane_id); + if (master_plane) { + struct dpu_plane *mpdpu = to_dpu_plane(master_plane); + + list_add_tail(&pdpu->mplane_list, &mpdpu->mplane_list); + } + + /* initialize underlying h/w driver */ + pdpu->pipe_hw = dpu_hw_sspp_init(pipe, kms->mmio, kms->catalog, + master_plane_id != 0); + if (IS_ERR(pdpu->pipe_hw)) { + DPU_ERROR("[%u]SSPP init failed\n", pipe); + ret = PTR_ERR(pdpu->pipe_hw); + goto clean_plane; + } else if (!pdpu->pipe_hw->cap || !pdpu->pipe_hw->cap->sblk) { + DPU_ERROR("[%u]SSPP init returned invalid cfg\n", pipe); + goto clean_sspp; + } + + /* cache features mask for later */ + pdpu->features = pdpu->pipe_hw->cap->features; + pdpu->pipe_sblk = pdpu->pipe_hw->cap->sblk; + if (!pdpu->pipe_sblk) { + DPU_ERROR("[%u]invalid sblk\n", pipe); + goto clean_sspp; + } + + if (!master_plane_id) + format_list = pdpu->pipe_sblk->format_list; + else + format_list = pdpu->pipe_sblk->virt_format_list; + + pdpu->nformats = dpu_populate_formats(format_list, + pdpu->formats, + 0, + ARRAY_SIZE(pdpu->formats)); + + if (!pdpu->nformats) { + DPU_ERROR("[%u]no valid formats for plane\n", pipe); + goto clean_sspp; + } + + if (pdpu->features & BIT(DPU_SSPP_CURSOR)) + type = DRM_PLANE_TYPE_CURSOR; + else if (primary_plane) + type = DRM_PLANE_TYPE_PRIMARY; + else + type = DRM_PLANE_TYPE_OVERLAY; + ret = drm_universal_plane_init(dev, plane, 0xff, &dpu_plane_funcs, + pdpu->formats, pdpu->nformats, + NULL, type, NULL); + if (ret) + goto clean_sspp; + + pdpu->catalog = kms->catalog; + + if (kms->catalog->mixer_count && + kms->catalog->mixer[0].sblk->maxblendstages) { + zpos_max = kms->catalog->mixer[0].sblk->maxblendstages - 1; + if (zpos_max > DPU_STAGE_MAX - DPU_STAGE_0 - 1) + zpos_max = DPU_STAGE_MAX - DPU_STAGE_0 - 1; + } + + ret = drm_plane_create_zpos_property(plane, 0, 0, zpos_max); + if (ret) + DPU_ERROR("failed to install zpos property, rc = %d\n", ret); + + /* success! finalize initialization */ + drm_plane_helper_add(plane, &dpu_plane_helper_funcs); + + /* save user friendly pipe name for later */ + snprintf(pdpu->pipe_name, DPU_NAME_SIZE, "plane%u", plane->base.id); + + mutex_init(&pdpu->lock); + + DPU_DEBUG("%s created for pipe:%u id:%u virtual:%u\n", pdpu->pipe_name, + pipe, plane->base.id, master_plane_id); + return plane; + +clean_sspp: + if (pdpu && pdpu->pipe_hw) + dpu_hw_sspp_destroy(pdpu->pipe_hw); +clean_plane: + kfree(pdpu); +exit: + return ERR_PTR(ret); +} diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_plane.h b/drivers/gpu/drm/msm/disp/dpu1/dpu_plane.h new file mode 100644 index 000000000000..f6fe6ddc7a3a --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_plane.h @@ -0,0 +1,175 @@ +/* + * Copyright (c) 2015-2018, The Linux Foundation. All rights reserved. + * Copyright (C) 2013 Red Hat + * Author: Rob Clark <robdclark@gmail.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published by + * the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program. If not, see <http://www.gnu.org/licenses/>. + */ + +#ifndef _DPU_PLANE_H_ +#define _DPU_PLANE_H_ + +#include <drm/drm_crtc.h> + +#include "dpu_kms.h" +#include "dpu_hw_mdss.h" +#include "dpu_hw_sspp.h" + +/** + * struct dpu_plane_state: Define dpu extension of drm plane state object + * @base: base drm plane state object + * @property_state: Local storage for msm_prop properties + * @property_values: cached plane property values + * @aspace: pointer to address space for input/output buffers + * @input_fence: dereferenced input fence pointer + * @stage: assigned by crtc blender + * @multirect_index: index of the rectangle of SSPP + * @multirect_mode: parallel or time multiplex multirect mode + * @pending: whether the current update is still pending + * @scaler3_cfg: configuration data for scaler3 + * @pixel_ext: configuration data for pixel extensions + * @scaler_check_state: indicates status of user provided pixel extension data + * @cdp_cfg: CDP configuration + */ +struct dpu_plane_state { + struct drm_plane_state base; + struct msm_gem_address_space *aspace; + void *input_fence; + enum dpu_stage stage; + uint32_t multirect_index; + uint32_t multirect_mode; + bool pending; + + /* scaler configuration */ + struct dpu_hw_scaler3_cfg scaler3_cfg; + struct dpu_hw_pixel_ext pixel_ext; + + struct dpu_hw_pipe_cdp_cfg cdp_cfg; +}; + +/** + * struct dpu_multirect_plane_states: Defines multirect pair of drm plane states + * @r0: drm plane configured on rect 0 + * @r1: drm plane configured on rect 1 + */ +struct dpu_multirect_plane_states { + const struct drm_plane_state *r0; + const struct drm_plane_state *r1; +}; + +#define to_dpu_plane_state(x) \ + container_of(x, struct dpu_plane_state, base) + +/** + * dpu_plane_pipe - return sspp identifier for the given plane + * @plane: Pointer to DRM plane object + * Returns: sspp identifier of the given plane + */ +enum dpu_sspp dpu_plane_pipe(struct drm_plane *plane); + +/** + * is_dpu_plane_virtual - check for virtual plane + * @plane: Pointer to DRM plane object + * returns: true - if the plane is virtual + * false - if the plane is primary + */ +bool is_dpu_plane_virtual(struct drm_plane *plane); + +/** + * dpu_plane_get_ctl_flush - get control flush mask + * @plane: Pointer to DRM plane object + * @ctl: Pointer to control hardware + * @flush_sspp: Pointer to sspp flush control word + */ +void dpu_plane_get_ctl_flush(struct drm_plane *plane, struct dpu_hw_ctl *ctl, + u32 *flush_sspp); + +/** + * dpu_plane_restore - restore hw state if previously power collapsed + * @plane: Pointer to drm plane structure + */ +void dpu_plane_restore(struct drm_plane *plane); + +/** + * dpu_plane_flush - final plane operations before commit flush + * @plane: Pointer to drm plane structure + */ +void dpu_plane_flush(struct drm_plane *plane); + +/** + * dpu_plane_kickoff - final plane operations before commit kickoff + * @plane: Pointer to drm plane structure + */ +void dpu_plane_kickoff(struct drm_plane *plane); + +/** + * dpu_plane_set_error: enable/disable error condition + * @plane: pointer to drm_plane structure + */ +void dpu_plane_set_error(struct drm_plane *plane, bool error); + +/** + * dpu_plane_init - create new dpu plane for the given pipe + * @dev: Pointer to DRM device + * @pipe: dpu hardware pipe identifier + * @primary_plane: true if this pipe is primary plane for crtc + * @possible_crtcs: bitmask of crtc that can be attached to the given pipe + * @master_plane_id: primary plane id of a multirect pipe. 0 value passed for + * a regular plane initialization. A non-zero primary plane + * id will be passed for a virtual pipe initialization. + * + */ +struct drm_plane *dpu_plane_init(struct drm_device *dev, + uint32_t pipe, bool primary_plane, + unsigned long possible_crtcs, u32 master_plane_id); + +/** + * dpu_plane_validate_multirecti_v2 - validate the multirect planes + * against hw limitations + * @plane: drm plate states of the multirect pair + */ +int dpu_plane_validate_multirect_v2(struct dpu_multirect_plane_states *plane); + +/** + * dpu_plane_clear_multirect - clear multirect bits for the given pipe + * @drm_state: Pointer to DRM plane state + */ +void dpu_plane_clear_multirect(const struct drm_plane_state *drm_state); + +/** + * dpu_plane_wait_input_fence - wait for input fence object + * @plane: Pointer to DRM plane object + * @wait_ms: Wait timeout value + * Returns: Zero on success + */ +int dpu_plane_wait_input_fence(struct drm_plane *plane, uint32_t wait_ms); + +/** + * dpu_plane_color_fill - enables color fill on plane + * @plane: Pointer to DRM plane object + * @color: RGB fill color value, [23..16] Blue, [15..8] Green, [7..0] Red + * @alpha: 8-bit fill alpha value, 255 selects 100% alpha + * Returns: 0 on success + */ +int dpu_plane_color_fill(struct drm_plane *plane, + uint32_t color, uint32_t alpha); + +/** + * dpu_plane_set_revalidate - sets revalidate flag which forces a full + * validation of the plane properties in the next atomic check + * @plane: Pointer to DRM plane object + * @enable: Boolean to set/unset the flag + */ +void dpu_plane_set_revalidate(struct drm_plane *plane, bool enable); + +#endif /* _DPU_PLANE_H_ */ diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_power_handle.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_power_handle.c new file mode 100644 index 000000000000..a68f1249388c --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_power_handle.c @@ -0,0 +1,249 @@ +/* Copyright (c) 2014-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + */ + +#define pr_fmt(fmt) "[drm:%s:%d]: " fmt, __func__, __LINE__ + +#include <linux/kernel.h> +#include <linux/of.h> +#include <linux/string.h> +#include <linux/of_address.h> +#include <linux/slab.h> +#include <linux/mutex.h> +#include <linux/of_platform.h> + +#include "dpu_power_handle.h" +#include "dpu_trace.h" + +static const char *data_bus_name[DPU_POWER_HANDLE_DBUS_ID_MAX] = { + [DPU_POWER_HANDLE_DBUS_ID_MNOC] = "qcom,dpu-data-bus", + [DPU_POWER_HANDLE_DBUS_ID_LLCC] = "qcom,dpu-llcc-bus", + [DPU_POWER_HANDLE_DBUS_ID_EBI] = "qcom,dpu-ebi-bus", +}; + +const char *dpu_power_handle_get_dbus_name(u32 bus_id) +{ + if (bus_id < DPU_POWER_HANDLE_DBUS_ID_MAX) + return data_bus_name[bus_id]; + + return NULL; +} + +static void dpu_power_event_trigger_locked(struct dpu_power_handle *phandle, + u32 event_type) +{ + struct dpu_power_event *event; + + list_for_each_entry(event, &phandle->event_list, list) { + if (event->event_type & event_type) + event->cb_fnc(event_type, event->usr); + } +} + +struct dpu_power_client *dpu_power_client_create( + struct dpu_power_handle *phandle, char *client_name) +{ + struct dpu_power_client *client; + static u32 id; + + if (!client_name || !phandle) { + pr_err("client name is null or invalid power data\n"); + return ERR_PTR(-EINVAL); + } + + client = kzalloc(sizeof(struct dpu_power_client), GFP_KERNEL); + if (!client) + return ERR_PTR(-ENOMEM); + + mutex_lock(&phandle->phandle_lock); + strlcpy(client->name, client_name, MAX_CLIENT_NAME_LEN); + client->usecase_ndx = VOTE_INDEX_DISABLE; + client->id = id; + client->active = true; + pr_debug("client %s created:%pK id :%d\n", client_name, + client, id); + id++; + list_add(&client->list, &phandle->power_client_clist); + mutex_unlock(&phandle->phandle_lock); + + return client; +} + +void dpu_power_client_destroy(struct dpu_power_handle *phandle, + struct dpu_power_client *client) +{ + if (!client || !phandle) { + pr_err("reg bus vote: invalid client handle\n"); + } else if (!client->active) { + pr_err("dpu power deinit already done\n"); + kfree(client); + } else { + pr_debug("bus vote client %s destroyed:%pK id:%u\n", + client->name, client, client->id); + mutex_lock(&phandle->phandle_lock); + list_del_init(&client->list); + mutex_unlock(&phandle->phandle_lock); + kfree(client); + } +} + +void dpu_power_resource_init(struct platform_device *pdev, + struct dpu_power_handle *phandle) +{ + phandle->dev = &pdev->dev; + + INIT_LIST_HEAD(&phandle->power_client_clist); + INIT_LIST_HEAD(&phandle->event_list); + + mutex_init(&phandle->phandle_lock); +} + +void dpu_power_resource_deinit(struct platform_device *pdev, + struct dpu_power_handle *phandle) +{ + struct dpu_power_client *curr_client, *next_client; + struct dpu_power_event *curr_event, *next_event; + + if (!phandle || !pdev) { + pr_err("invalid input param\n"); + return; + } + + mutex_lock(&phandle->phandle_lock); + list_for_each_entry_safe(curr_client, next_client, + &phandle->power_client_clist, list) { + pr_err("cliend:%s-%d still registered with refcount:%d\n", + curr_client->name, curr_client->id, + curr_client->refcount); + curr_client->active = false; + list_del(&curr_client->list); + } + + list_for_each_entry_safe(curr_event, next_event, + &phandle->event_list, list) { + pr_err("event:%d, client:%s still registered\n", + curr_event->event_type, + curr_event->client_name); + curr_event->active = false; + list_del(&curr_event->list); + } + mutex_unlock(&phandle->phandle_lock); +} + +int dpu_power_resource_enable(struct dpu_power_handle *phandle, + struct dpu_power_client *pclient, bool enable) +{ + bool changed = false; + u32 max_usecase_ndx = VOTE_INDEX_DISABLE, prev_usecase_ndx; + struct dpu_power_client *client; + + if (!phandle || !pclient) { + pr_err("invalid input argument\n"); + return -EINVAL; + } + + mutex_lock(&phandle->phandle_lock); + if (enable) + pclient->refcount++; + else if (pclient->refcount) + pclient->refcount--; + + if (pclient->refcount) + pclient->usecase_ndx = VOTE_INDEX_LOW; + else + pclient->usecase_ndx = VOTE_INDEX_DISABLE; + + list_for_each_entry(client, &phandle->power_client_clist, list) { + if (client->usecase_ndx < VOTE_INDEX_MAX && + client->usecase_ndx > max_usecase_ndx) + max_usecase_ndx = client->usecase_ndx; + } + + if (phandle->current_usecase_ndx != max_usecase_ndx) { + changed = true; + prev_usecase_ndx = phandle->current_usecase_ndx; + phandle->current_usecase_ndx = max_usecase_ndx; + } + + pr_debug("%pS: changed=%d current idx=%d request client %s id:%u enable:%d refcount:%d\n", + __builtin_return_address(0), changed, max_usecase_ndx, + pclient->name, pclient->id, enable, pclient->refcount); + + if (!changed) + goto end; + + if (enable) { + dpu_power_event_trigger_locked(phandle, + DPU_POWER_EVENT_PRE_ENABLE); + dpu_power_event_trigger_locked(phandle, + DPU_POWER_EVENT_POST_ENABLE); + + } else { + dpu_power_event_trigger_locked(phandle, + DPU_POWER_EVENT_PRE_DISABLE); + dpu_power_event_trigger_locked(phandle, + DPU_POWER_EVENT_POST_DISABLE); + } + +end: + mutex_unlock(&phandle->phandle_lock); + return 0; +} + +struct dpu_power_event *dpu_power_handle_register_event( + struct dpu_power_handle *phandle, + u32 event_type, void (*cb_fnc)(u32 event_type, void *usr), + void *usr, char *client_name) +{ + struct dpu_power_event *event; + + if (!phandle) { + pr_err("invalid power handle\n"); + return ERR_PTR(-EINVAL); + } else if (!cb_fnc || !event_type) { + pr_err("no callback fnc or event type\n"); + return ERR_PTR(-EINVAL); + } + + event = kzalloc(sizeof(struct dpu_power_event), GFP_KERNEL); + if (!event) + return ERR_PTR(-ENOMEM); + + event->event_type = event_type; + event->cb_fnc = cb_fnc; + event->usr = usr; + strlcpy(event->client_name, client_name, MAX_CLIENT_NAME_LEN); + event->active = true; + + mutex_lock(&phandle->phandle_lock); + list_add(&event->list, &phandle->event_list); + mutex_unlock(&phandle->phandle_lock); + + return event; +} + +void dpu_power_handle_unregister_event( + struct dpu_power_handle *phandle, + struct dpu_power_event *event) +{ + if (!phandle || !event) { + pr_err("invalid phandle or event\n"); + } else if (!event->active) { + pr_err("power handle deinit already done\n"); + kfree(event); + } else { + mutex_lock(&phandle->phandle_lock); + list_del_init(&event->list); + mutex_unlock(&phandle->phandle_lock); + kfree(event); + } +} diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_power_handle.h b/drivers/gpu/drm/msm/disp/dpu1/dpu_power_handle.h new file mode 100644 index 000000000000..344f74464eca --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_power_handle.h @@ -0,0 +1,225 @@ +/* Copyright (c) 2016-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + */ + +#ifndef _DPU_POWER_HANDLE_H_ +#define _DPU_POWER_HANDLE_H_ + +#define MAX_CLIENT_NAME_LEN 128 + +#define DPU_POWER_HANDLE_ENABLE_BUS_AB_QUOTA 0 +#define DPU_POWER_HANDLE_DISABLE_BUS_AB_QUOTA 0 +#define DPU_POWER_HANDLE_ENABLE_BUS_IB_QUOTA 1600000000 +#define DPU_POWER_HANDLE_DISABLE_BUS_IB_QUOTA 0 + +#include "dpu_io_util.h" + +/* event will be triggered before power handler disable */ +#define DPU_POWER_EVENT_PRE_DISABLE 0x1 + +/* event will be triggered after power handler disable */ +#define DPU_POWER_EVENT_POST_DISABLE 0x2 + +/* event will be triggered before power handler enable */ +#define DPU_POWER_EVENT_PRE_ENABLE 0x4 + +/* event will be triggered after power handler enable */ +#define DPU_POWER_EVENT_POST_ENABLE 0x8 + +/** + * mdss_bus_vote_type: register bus vote type + * VOTE_INDEX_DISABLE: removes the client vote + * VOTE_INDEX_LOW: keeps the lowest vote for register bus + * VOTE_INDEX_MAX: invalid + */ +enum mdss_bus_vote_type { + VOTE_INDEX_DISABLE, + VOTE_INDEX_LOW, + VOTE_INDEX_MAX, +}; + +/** + * enum dpu_power_handle_data_bus_client - type of axi bus clients + * @DPU_POWER_HANDLE_DATA_BUS_CLIENT_RT: core real-time bus client + * @DPU_POWER_HANDLE_DATA_BUS_CLIENT_NRT: core non-real-time bus client + * @DPU_POWER_HANDLE_DATA_BUS_CLIENT_MAX: maximum number of bus client type + */ +enum dpu_power_handle_data_bus_client { + DPU_POWER_HANDLE_DATA_BUS_CLIENT_RT, + DPU_POWER_HANDLE_DATA_BUS_CLIENT_NRT, + DPU_POWER_HANDLE_DATA_BUS_CLIENT_MAX +}; + +/** + * enum DPU_POWER_HANDLE_DBUS_ID - data bus identifier + * @DPU_POWER_HANDLE_DBUS_ID_MNOC: DPU/MNOC data bus + * @DPU_POWER_HANDLE_DBUS_ID_LLCC: MNOC/LLCC data bus + * @DPU_POWER_HANDLE_DBUS_ID_EBI: LLCC/EBI data bus + */ +enum DPU_POWER_HANDLE_DBUS_ID { + DPU_POWER_HANDLE_DBUS_ID_MNOC, + DPU_POWER_HANDLE_DBUS_ID_LLCC, + DPU_POWER_HANDLE_DBUS_ID_EBI, + DPU_POWER_HANDLE_DBUS_ID_MAX, +}; + +/** + * struct dpu_power_client: stores the power client for dpu driver + * @name: name of the client + * @usecase_ndx: current regs bus vote type + * @refcount: current refcount if multiple modules are using same + * same client for enable/disable. Power module will + * aggregate the refcount and vote accordingly for this + * client. + * @id: assigned during create. helps for debugging. + * @list: list to attach power handle master list + * @ab: arbitrated bandwidth for each bus client + * @ib: instantaneous bandwidth for each bus client + * @active: inidcates the state of dpu power handle + */ +struct dpu_power_client { + char name[MAX_CLIENT_NAME_LEN]; + short usecase_ndx; + short refcount; + u32 id; + struct list_head list; + u64 ab[DPU_POWER_HANDLE_DATA_BUS_CLIENT_MAX]; + u64 ib[DPU_POWER_HANDLE_DATA_BUS_CLIENT_MAX]; + bool active; +}; + +/* + * struct dpu_power_event - local event registration structure + * @client_name: name of the client registering + * @cb_fnc: pointer to desired callback function + * @usr: user pointer to pass to callback event trigger + * @event: refer to DPU_POWER_HANDLE_EVENT_* + * @list: list to attach event master list + * @active: indicates the state of dpu power handle + */ +struct dpu_power_event { + char client_name[MAX_CLIENT_NAME_LEN]; + void (*cb_fnc)(u32 event_type, void *usr); + void *usr; + u32 event_type; + struct list_head list; + bool active; +}; + +/** + * struct dpu_power_handle: power handle main struct + * @client_clist: master list to store all clients + * @phandle_lock: lock to synchronize the enable/disable + * @dev: pointer to device structure + * @usecase_ndx: current usecase index + * @event_list: current power handle event list + */ +struct dpu_power_handle { + struct list_head power_client_clist; + struct mutex phandle_lock; + struct device *dev; + u32 current_usecase_ndx; + struct list_head event_list; +}; + +/** + * dpu_power_resource_init() - initializes the dpu power handle + * @pdev: platform device to search the power resources + * @pdata: power handle to store the power resources + */ +void dpu_power_resource_init(struct platform_device *pdev, + struct dpu_power_handle *pdata); + +/** + * dpu_power_resource_deinit() - release the dpu power handle + * @pdev: platform device for power resources + * @pdata: power handle containing the resources + * + * Return: error code. + */ +void dpu_power_resource_deinit(struct platform_device *pdev, + struct dpu_power_handle *pdata); + +/** + * dpu_power_client_create() - create the client on power handle + * @pdata: power handle containing the resources + * @client_name: new client name for registration + * + * Return: error code. + */ +struct dpu_power_client *dpu_power_client_create(struct dpu_power_handle *pdata, + char *client_name); + +/** + * dpu_power_client_destroy() - destroy the client on power handle + * @pdata: power handle containing the resources + * @client_name: new client name for registration + * + * Return: none + */ +void dpu_power_client_destroy(struct dpu_power_handle *phandle, + struct dpu_power_client *client); + +/** + * dpu_power_resource_enable() - enable/disable the power resources + * @pdata: power handle containing the resources + * @client: client information to enable/disable its vote + * @enable: boolean request for enable/disable + * + * Return: error code. + */ +int dpu_power_resource_enable(struct dpu_power_handle *pdata, + struct dpu_power_client *pclient, bool enable); + +/** + * dpu_power_data_bus_bandwidth_ctrl() - control data bus bandwidth enable + * @phandle: power handle containing the resources + * @client: client information to bandwidth control + * @enable: true to enable bandwidth for data base + * + * Return: none + */ +void dpu_power_data_bus_bandwidth_ctrl(struct dpu_power_handle *phandle, + struct dpu_power_client *pclient, int enable); + +/** + * dpu_power_handle_register_event - register a callback function for an event. + * Clients can register for multiple events with a single register. + * Any block with access to phandle can register for the event + * notification. + * @phandle: power handle containing the resources + * @event_type: event type to register; refer DPU_POWER_HANDLE_EVENT_* + * @cb_fnc: pointer to desired callback function + * @usr: user pointer to pass to callback on event trigger + * + * Return: event pointer if success, or error code otherwise + */ +struct dpu_power_event *dpu_power_handle_register_event( + struct dpu_power_handle *phandle, + u32 event_type, void (*cb_fnc)(u32 event_type, void *usr), + void *usr, char *client_name); +/** + * dpu_power_handle_unregister_event - unregister callback for event(s) + * @phandle: power handle containing the resources + * @event: event pointer returned after power handle register + */ +void dpu_power_handle_unregister_event(struct dpu_power_handle *phandle, + struct dpu_power_event *event); + +/** + * dpu_power_handle_get_dbus_name - get name of given data bus identifier + * @bus_id: data bus identifier + * Return: Pointer to name string if success; NULL otherwise + */ +const char *dpu_power_handle_get_dbus_name(u32 bus_id); + +#endif /* _DPU_POWER_HANDLE_H_ */ diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_rm.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_rm.c new file mode 100644 index 000000000000..13c0a36d4ef9 --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_rm.c @@ -0,0 +1,1079 @@ +/* + * Copyright (c) 2016-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + */ + +#define pr_fmt(fmt) "[drm:%s] " fmt, __func__ +#include "dpu_kms.h" +#include "dpu_hw_lm.h" +#include "dpu_hw_ctl.h" +#include "dpu_hw_cdm.h" +#include "dpu_hw_pingpong.h" +#include "dpu_hw_intf.h" +#include "dpu_encoder.h" +#include "dpu_trace.h" + +#define RESERVED_BY_OTHER(h, r) \ + ((h)->rsvp && ((h)->rsvp->enc_id != (r)->enc_id)) + +#define RM_RQ_LOCK(r) ((r)->top_ctrl & BIT(DPU_RM_TOPCTL_RESERVE_LOCK)) +#define RM_RQ_CLEAR(r) ((r)->top_ctrl & BIT(DPU_RM_TOPCTL_RESERVE_CLEAR)) +#define RM_RQ_DS(r) ((r)->top_ctrl & BIT(DPU_RM_TOPCTL_DS)) +#define RM_IS_TOPOLOGY_MATCH(t, r) ((t).num_lm == (r).num_lm && \ + (t).num_comp_enc == (r).num_enc && \ + (t).num_intf == (r).num_intf) + +struct dpu_rm_topology_def { + enum dpu_rm_topology_name top_name; + int num_lm; + int num_comp_enc; + int num_intf; + int num_ctl; + int needs_split_display; +}; + +static const struct dpu_rm_topology_def g_top_table[] = { + { DPU_RM_TOPOLOGY_NONE, 0, 0, 0, 0, false }, + { DPU_RM_TOPOLOGY_SINGLEPIPE, 1, 0, 1, 1, false }, + { DPU_RM_TOPOLOGY_DUALPIPE, 2, 0, 2, 2, true }, + { DPU_RM_TOPOLOGY_DUALPIPE_3DMERGE, 2, 0, 1, 1, false }, +}; + +/** + * struct dpu_rm_requirements - Reservation requirements parameter bundle + * @top_ctrl: topology control preference from kernel client + * @top: selected topology for the display + * @hw_res: Hardware resources required as reported by the encoders + */ +struct dpu_rm_requirements { + uint64_t top_ctrl; + const struct dpu_rm_topology_def *topology; + struct dpu_encoder_hw_resources hw_res; +}; + +/** + * struct dpu_rm_rsvp - Use Case Reservation tagging structure + * Used to tag HW blocks as reserved by a CRTC->Encoder->Connector chain + * By using as a tag, rather than lists of pointers to HW blocks used + * we can avoid some list management since we don't know how many blocks + * of each type a given use case may require. + * @list: List head for list of all reservations + * @seq: Global RSVP sequence number for debugging, especially for + * differentiating differenct allocations for same encoder. + * @enc_id: Reservations are tracked by Encoder DRM object ID. + * CRTCs may be connected to multiple Encoders. + * An encoder or connector id identifies the display path. + * @topology DRM<->HW topology use case + */ +struct dpu_rm_rsvp { + struct list_head list; + uint32_t seq; + uint32_t enc_id; + enum dpu_rm_topology_name topology; +}; + +/** + * struct dpu_rm_hw_blk - hardware block tracking list member + * @list: List head for list of all hardware blocks tracking items + * @rsvp: Pointer to use case reservation if reserved by a client + * @rsvp_nxt: Temporary pointer used during reservation to the incoming + * request. Will be swapped into rsvp if proposal is accepted + * @type: Type of hardware block this structure tracks + * @id: Hardware ID number, within it's own space, ie. LM_X + * @catalog: Pointer to the hardware catalog entry for this block + * @hw: Pointer to the hardware register access object for this block + */ +struct dpu_rm_hw_blk { + struct list_head list; + struct dpu_rm_rsvp *rsvp; + struct dpu_rm_rsvp *rsvp_nxt; + enum dpu_hw_blk_type type; + uint32_t id; + struct dpu_hw_blk *hw; +}; + +/** + * dpu_rm_dbg_rsvp_stage - enum of steps in making reservation for event logging + */ +enum dpu_rm_dbg_rsvp_stage { + DPU_RM_STAGE_BEGIN, + DPU_RM_STAGE_AFTER_CLEAR, + DPU_RM_STAGE_AFTER_RSVPNEXT, + DPU_RM_STAGE_FINAL +}; + +static void _dpu_rm_print_rsvps( + struct dpu_rm *rm, + enum dpu_rm_dbg_rsvp_stage stage) +{ + struct dpu_rm_rsvp *rsvp; + struct dpu_rm_hw_blk *blk; + enum dpu_hw_blk_type type; + + DPU_DEBUG("%d\n", stage); + + list_for_each_entry(rsvp, &rm->rsvps, list) { + DRM_DEBUG_KMS("%d rsvp[s%ue%u] topology %d\n", stage, rsvp->seq, + rsvp->enc_id, rsvp->topology); + } + + for (type = 0; type < DPU_HW_BLK_MAX; type++) { + list_for_each_entry(blk, &rm->hw_blks[type], list) { + if (!blk->rsvp && !blk->rsvp_nxt) + continue; + + DRM_DEBUG_KMS("%d rsvp[s%ue%u->s%ue%u] %d %d\n", stage, + (blk->rsvp) ? blk->rsvp->seq : 0, + (blk->rsvp) ? blk->rsvp->enc_id : 0, + (blk->rsvp_nxt) ? blk->rsvp_nxt->seq : 0, + (blk->rsvp_nxt) ? blk->rsvp_nxt->enc_id : 0, + blk->type, blk->id); + } + } +} + +struct dpu_hw_mdp *dpu_rm_get_mdp(struct dpu_rm *rm) +{ + return rm->hw_mdp; +} + +enum dpu_rm_topology_name +dpu_rm_get_topology_name(struct msm_display_topology topology) +{ + int i; + + for (i = 0; i < DPU_RM_TOPOLOGY_MAX; i++) + if (RM_IS_TOPOLOGY_MATCH(g_top_table[i], topology)) + return g_top_table[i].top_name; + + return DPU_RM_TOPOLOGY_NONE; +} + +void dpu_rm_init_hw_iter( + struct dpu_rm_hw_iter *iter, + uint32_t enc_id, + enum dpu_hw_blk_type type) +{ + memset(iter, 0, sizeof(*iter)); + iter->enc_id = enc_id; + iter->type = type; +} + +static bool _dpu_rm_get_hw_locked(struct dpu_rm *rm, struct dpu_rm_hw_iter *i) +{ + struct list_head *blk_list; + + if (!rm || !i || i->type >= DPU_HW_BLK_MAX) { + DPU_ERROR("invalid rm\n"); + return false; + } + + i->hw = NULL; + blk_list = &rm->hw_blks[i->type]; + + if (i->blk && (&i->blk->list == blk_list)) { + DPU_DEBUG("attempt resume iteration past last\n"); + return false; + } + + i->blk = list_prepare_entry(i->blk, blk_list, list); + + list_for_each_entry_continue(i->blk, blk_list, list) { + struct dpu_rm_rsvp *rsvp = i->blk->rsvp; + + if (i->blk->type != i->type) { + DPU_ERROR("found incorrect block type %d on %d list\n", + i->blk->type, i->type); + return false; + } + + if ((i->enc_id == 0) || (rsvp && rsvp->enc_id == i->enc_id)) { + i->hw = i->blk->hw; + DPU_DEBUG("found type %d id %d for enc %d\n", + i->type, i->blk->id, i->enc_id); + return true; + } + } + + DPU_DEBUG("no match, type %d for enc %d\n", i->type, i->enc_id); + + return false; +} + +bool dpu_rm_get_hw(struct dpu_rm *rm, struct dpu_rm_hw_iter *i) +{ + bool ret; + + mutex_lock(&rm->rm_lock); + ret = _dpu_rm_get_hw_locked(rm, i); + mutex_unlock(&rm->rm_lock); + + return ret; +} + +static void _dpu_rm_hw_destroy(enum dpu_hw_blk_type type, void *hw) +{ + switch (type) { + case DPU_HW_BLK_LM: + dpu_hw_lm_destroy(hw); + break; + case DPU_HW_BLK_CTL: + dpu_hw_ctl_destroy(hw); + break; + case DPU_HW_BLK_CDM: + dpu_hw_cdm_destroy(hw); + break; + case DPU_HW_BLK_PINGPONG: + dpu_hw_pingpong_destroy(hw); + break; + case DPU_HW_BLK_INTF: + dpu_hw_intf_destroy(hw); + break; + case DPU_HW_BLK_SSPP: + /* SSPPs are not managed by the resource manager */ + case DPU_HW_BLK_TOP: + /* Top is a singleton, not managed in hw_blks list */ + case DPU_HW_BLK_MAX: + default: + DPU_ERROR("unsupported block type %d\n", type); + break; + } +} + +int dpu_rm_destroy(struct dpu_rm *rm) +{ + + struct dpu_rm_rsvp *rsvp_cur, *rsvp_nxt; + struct dpu_rm_hw_blk *hw_cur, *hw_nxt; + enum dpu_hw_blk_type type; + + if (!rm) { + DPU_ERROR("invalid rm\n"); + return -EINVAL; + } + + list_for_each_entry_safe(rsvp_cur, rsvp_nxt, &rm->rsvps, list) { + list_del(&rsvp_cur->list); + kfree(rsvp_cur); + } + + + for (type = 0; type < DPU_HW_BLK_MAX; type++) { + list_for_each_entry_safe(hw_cur, hw_nxt, &rm->hw_blks[type], + list) { + list_del(&hw_cur->list); + _dpu_rm_hw_destroy(hw_cur->type, hw_cur->hw); + kfree(hw_cur); + } + } + + dpu_hw_mdp_destroy(rm->hw_mdp); + rm->hw_mdp = NULL; + + mutex_destroy(&rm->rm_lock); + + return 0; +} + +static int _dpu_rm_hw_blk_create( + struct dpu_rm *rm, + struct dpu_mdss_cfg *cat, + void __iomem *mmio, + enum dpu_hw_blk_type type, + uint32_t id, + void *hw_catalog_info) +{ + struct dpu_rm_hw_blk *blk; + struct dpu_hw_mdp *hw_mdp; + void *hw; + + hw_mdp = rm->hw_mdp; + + switch (type) { + case DPU_HW_BLK_LM: + hw = dpu_hw_lm_init(id, mmio, cat); + break; + case DPU_HW_BLK_CTL: + hw = dpu_hw_ctl_init(id, mmio, cat); + break; + case DPU_HW_BLK_CDM: + hw = dpu_hw_cdm_init(id, mmio, cat, hw_mdp); + break; + case DPU_HW_BLK_PINGPONG: + hw = dpu_hw_pingpong_init(id, mmio, cat); + break; + case DPU_HW_BLK_INTF: + hw = dpu_hw_intf_init(id, mmio, cat); + break; + case DPU_HW_BLK_SSPP: + /* SSPPs are not managed by the resource manager */ + case DPU_HW_BLK_TOP: + /* Top is a singleton, not managed in hw_blks list */ + case DPU_HW_BLK_MAX: + default: + DPU_ERROR("unsupported block type %d\n", type); + return -EINVAL; + } + + if (IS_ERR_OR_NULL(hw)) { + DPU_ERROR("failed hw object creation: type %d, err %ld\n", + type, PTR_ERR(hw)); + return -EFAULT; + } + + blk = kzalloc(sizeof(*blk), GFP_KERNEL); + if (!blk) { + _dpu_rm_hw_destroy(type, hw); + return -ENOMEM; + } + + blk->type = type; + blk->id = id; + blk->hw = hw; + list_add_tail(&blk->list, &rm->hw_blks[type]); + + return 0; +} + +int dpu_rm_init(struct dpu_rm *rm, + struct dpu_mdss_cfg *cat, + void __iomem *mmio, + struct drm_device *dev) +{ + int rc, i; + enum dpu_hw_blk_type type; + + if (!rm || !cat || !mmio || !dev) { + DPU_ERROR("invalid kms\n"); + return -EINVAL; + } + + /* Clear, setup lists */ + memset(rm, 0, sizeof(*rm)); + + mutex_init(&rm->rm_lock); + + INIT_LIST_HEAD(&rm->rsvps); + for (type = 0; type < DPU_HW_BLK_MAX; type++) + INIT_LIST_HEAD(&rm->hw_blks[type]); + + rm->dev = dev; + + /* Some of the sub-blocks require an mdptop to be created */ + rm->hw_mdp = dpu_hw_mdptop_init(MDP_TOP, mmio, cat); + if (IS_ERR_OR_NULL(rm->hw_mdp)) { + rc = PTR_ERR(rm->hw_mdp); + rm->hw_mdp = NULL; + DPU_ERROR("failed: mdp hw not available\n"); + goto fail; + } + + /* Interrogate HW catalog and create tracking items for hw blocks */ + for (i = 0; i < cat->mixer_count; i++) { + struct dpu_lm_cfg *lm = &cat->mixer[i]; + + if (lm->pingpong == PINGPONG_MAX) { + DPU_DEBUG("skip mixer %d without pingpong\n", lm->id); + continue; + } + + rc = _dpu_rm_hw_blk_create(rm, cat, mmio, DPU_HW_BLK_LM, + cat->mixer[i].id, &cat->mixer[i]); + if (rc) { + DPU_ERROR("failed: lm hw not available\n"); + goto fail; + } + + if (!rm->lm_max_width) { + rm->lm_max_width = lm->sblk->maxwidth; + } else if (rm->lm_max_width != lm->sblk->maxwidth) { + /* + * Don't expect to have hw where lm max widths differ. + * If found, take the min. + */ + DPU_ERROR("unsupported: lm maxwidth differs\n"); + if (rm->lm_max_width > lm->sblk->maxwidth) + rm->lm_max_width = lm->sblk->maxwidth; + } + } + + for (i = 0; i < cat->pingpong_count; i++) { + rc = _dpu_rm_hw_blk_create(rm, cat, mmio, DPU_HW_BLK_PINGPONG, + cat->pingpong[i].id, &cat->pingpong[i]); + if (rc) { + DPU_ERROR("failed: pp hw not available\n"); + goto fail; + } + } + + for (i = 0; i < cat->intf_count; i++) { + if (cat->intf[i].type == INTF_NONE) { + DPU_DEBUG("skip intf %d with type none\n", i); + continue; + } + + rc = _dpu_rm_hw_blk_create(rm, cat, mmio, DPU_HW_BLK_INTF, + cat->intf[i].id, &cat->intf[i]); + if (rc) { + DPU_ERROR("failed: intf hw not available\n"); + goto fail; + } + } + + for (i = 0; i < cat->ctl_count; i++) { + rc = _dpu_rm_hw_blk_create(rm, cat, mmio, DPU_HW_BLK_CTL, + cat->ctl[i].id, &cat->ctl[i]); + if (rc) { + DPU_ERROR("failed: ctl hw not available\n"); + goto fail; + } + } + + for (i = 0; i < cat->cdm_count; i++) { + rc = _dpu_rm_hw_blk_create(rm, cat, mmio, DPU_HW_BLK_CDM, + cat->cdm[i].id, &cat->cdm[i]); + if (rc) { + DPU_ERROR("failed: cdm hw not available\n"); + goto fail; + } + } + + return 0; + +fail: + dpu_rm_destroy(rm); + + return rc; +} + +/** + * _dpu_rm_check_lm_and_get_connected_blks - check if proposed layer mixer meets + * proposed use case requirements, incl. hardwired dependent blocks like + * pingpong + * @rm: dpu resource manager handle + * @rsvp: reservation currently being created + * @reqs: proposed use case requirements + * @lm: proposed layer mixer, function checks if lm, and all other hardwired + * blocks connected to the lm (pp) is available and appropriate + * @pp: output parameter, pingpong block attached to the layer mixer. + * NULL if pp was not available, or not matching requirements. + * @primary_lm: if non-null, this function check if lm is compatible primary_lm + * as well as satisfying all other requirements + * @Return: true if lm matches all requirements, false otherwise + */ +static bool _dpu_rm_check_lm_and_get_connected_blks( + struct dpu_rm *rm, + struct dpu_rm_rsvp *rsvp, + struct dpu_rm_requirements *reqs, + struct dpu_rm_hw_blk *lm, + struct dpu_rm_hw_blk **pp, + struct dpu_rm_hw_blk *primary_lm) +{ + const struct dpu_lm_cfg *lm_cfg = to_dpu_hw_mixer(lm->hw)->cap; + struct dpu_rm_hw_iter iter; + + *pp = NULL; + + DPU_DEBUG("check lm %d pp %d\n", + lm_cfg->id, lm_cfg->pingpong); + + /* Check if this layer mixer is a peer of the proposed primary LM */ + if (primary_lm) { + const struct dpu_lm_cfg *prim_lm_cfg = + to_dpu_hw_mixer(primary_lm->hw)->cap; + + if (!test_bit(lm_cfg->id, &prim_lm_cfg->lm_pair_mask)) { + DPU_DEBUG("lm %d not peer of lm %d\n", lm_cfg->id, + prim_lm_cfg->id); + return false; + } + } + + /* Already reserved? */ + if (RESERVED_BY_OTHER(lm, rsvp)) { + DPU_DEBUG("lm %d already reserved\n", lm_cfg->id); + return false; + } + + dpu_rm_init_hw_iter(&iter, 0, DPU_HW_BLK_PINGPONG); + while (_dpu_rm_get_hw_locked(rm, &iter)) { + if (iter.blk->id == lm_cfg->pingpong) { + *pp = iter.blk; + break; + } + } + + if (!*pp) { + DPU_ERROR("failed to get pp on lm %d\n", lm_cfg->pingpong); + return false; + } + + if (RESERVED_BY_OTHER(*pp, rsvp)) { + DPU_DEBUG("lm %d pp %d already reserved\n", lm->id, + (*pp)->id); + return false; + } + + return true; +} + +static int _dpu_rm_reserve_lms( + struct dpu_rm *rm, + struct dpu_rm_rsvp *rsvp, + struct dpu_rm_requirements *reqs) + +{ + struct dpu_rm_hw_blk *lm[MAX_BLOCKS]; + struct dpu_rm_hw_blk *pp[MAX_BLOCKS]; + struct dpu_rm_hw_iter iter_i, iter_j; + int lm_count = 0; + int i, rc = 0; + + if (!reqs->topology->num_lm) { + DPU_ERROR("invalid number of lm: %d\n", reqs->topology->num_lm); + return -EINVAL; + } + + /* Find a primary mixer */ + dpu_rm_init_hw_iter(&iter_i, 0, DPU_HW_BLK_LM); + while (lm_count != reqs->topology->num_lm && + _dpu_rm_get_hw_locked(rm, &iter_i)) { + memset(&lm, 0, sizeof(lm)); + memset(&pp, 0, sizeof(pp)); + + lm_count = 0; + lm[lm_count] = iter_i.blk; + + if (!_dpu_rm_check_lm_and_get_connected_blks( + rm, rsvp, reqs, lm[lm_count], + &pp[lm_count], NULL)) + continue; + + ++lm_count; + + /* Valid primary mixer found, find matching peers */ + dpu_rm_init_hw_iter(&iter_j, 0, DPU_HW_BLK_LM); + + while (lm_count != reqs->topology->num_lm && + _dpu_rm_get_hw_locked(rm, &iter_j)) { + if (iter_i.blk == iter_j.blk) + continue; + + if (!_dpu_rm_check_lm_and_get_connected_blks( + rm, rsvp, reqs, iter_j.blk, + &pp[lm_count], iter_i.blk)) + continue; + + lm[lm_count] = iter_j.blk; + ++lm_count; + } + } + + if (lm_count != reqs->topology->num_lm) { + DPU_DEBUG("unable to find appropriate mixers\n"); + return -ENAVAIL; + } + + for (i = 0; i < ARRAY_SIZE(lm); i++) { + if (!lm[i]) + break; + + lm[i]->rsvp_nxt = rsvp; + pp[i]->rsvp_nxt = rsvp; + + trace_dpu_rm_reserve_lms(lm[i]->id, lm[i]->type, rsvp->enc_id, + pp[i]->id); + } + + return rc; +} + +static int _dpu_rm_reserve_ctls( + struct dpu_rm *rm, + struct dpu_rm_rsvp *rsvp, + const struct dpu_rm_topology_def *top) +{ + struct dpu_rm_hw_blk *ctls[MAX_BLOCKS]; + struct dpu_rm_hw_iter iter; + int i = 0; + + memset(&ctls, 0, sizeof(ctls)); + + dpu_rm_init_hw_iter(&iter, 0, DPU_HW_BLK_CTL); + while (_dpu_rm_get_hw_locked(rm, &iter)) { + const struct dpu_hw_ctl *ctl = to_dpu_hw_ctl(iter.blk->hw); + unsigned long features = ctl->caps->features; + bool has_split_display; + + if (RESERVED_BY_OTHER(iter.blk, rsvp)) + continue; + + has_split_display = BIT(DPU_CTL_SPLIT_DISPLAY) & features; + + DPU_DEBUG("ctl %d caps 0x%lX\n", iter.blk->id, features); + + if (top->needs_split_display != has_split_display) + continue; + + ctls[i] = iter.blk; + DPU_DEBUG("ctl %d match\n", iter.blk->id); + + if (++i == top->num_ctl) + break; + } + + if (i != top->num_ctl) + return -ENAVAIL; + + for (i = 0; i < ARRAY_SIZE(ctls) && i < top->num_ctl; i++) { + ctls[i]->rsvp_nxt = rsvp; + trace_dpu_rm_reserve_ctls(ctls[i]->id, ctls[i]->type, + rsvp->enc_id); + } + + return 0; +} + +static int _dpu_rm_reserve_cdm( + struct dpu_rm *rm, + struct dpu_rm_rsvp *rsvp, + uint32_t id, + enum dpu_hw_blk_type type) +{ + struct dpu_rm_hw_iter iter; + + DRM_DEBUG_KMS("type %d id %d\n", type, id); + + dpu_rm_init_hw_iter(&iter, 0, DPU_HW_BLK_CDM); + while (_dpu_rm_get_hw_locked(rm, &iter)) { + const struct dpu_hw_cdm *cdm = to_dpu_hw_cdm(iter.blk->hw); + const struct dpu_cdm_cfg *caps = cdm->caps; + bool match = false; + + if (RESERVED_BY_OTHER(iter.blk, rsvp)) + continue; + + if (type == DPU_HW_BLK_INTF && id != INTF_MAX) + match = test_bit(id, &caps->intf_connect); + + DRM_DEBUG_KMS("iter: type:%d id:%d enc:%d cdm:%lu match:%d\n", + iter.blk->type, iter.blk->id, rsvp->enc_id, + caps->intf_connect, match); + + if (!match) + continue; + + trace_dpu_rm_reserve_cdm(iter.blk->id, iter.blk->type, + rsvp->enc_id); + iter.blk->rsvp_nxt = rsvp; + break; + } + + if (!iter.hw) { + DPU_ERROR("couldn't reserve cdm for type %d id %d\n", type, id); + return -ENAVAIL; + } + + return 0; +} + +static int _dpu_rm_reserve_intf( + struct dpu_rm *rm, + struct dpu_rm_rsvp *rsvp, + uint32_t id, + enum dpu_hw_blk_type type, + bool needs_cdm) +{ + struct dpu_rm_hw_iter iter; + int ret = 0; + + /* Find the block entry in the rm, and note the reservation */ + dpu_rm_init_hw_iter(&iter, 0, type); + while (_dpu_rm_get_hw_locked(rm, &iter)) { + if (iter.blk->id != id) + continue; + + if (RESERVED_BY_OTHER(iter.blk, rsvp)) { + DPU_ERROR("type %d id %d already reserved\n", type, id); + return -ENAVAIL; + } + + iter.blk->rsvp_nxt = rsvp; + trace_dpu_rm_reserve_intf(iter.blk->id, iter.blk->type, + rsvp->enc_id); + break; + } + + /* Shouldn't happen since intfs are fixed at probe */ + if (!iter.hw) { + DPU_ERROR("couldn't find type %d id %d\n", type, id); + return -EINVAL; + } + + if (needs_cdm) + ret = _dpu_rm_reserve_cdm(rm, rsvp, id, type); + + return ret; +} + +static int _dpu_rm_reserve_intf_related_hw( + struct dpu_rm *rm, + struct dpu_rm_rsvp *rsvp, + struct dpu_encoder_hw_resources *hw_res) +{ + int i, ret = 0; + u32 id; + + for (i = 0; i < ARRAY_SIZE(hw_res->intfs); i++) { + if (hw_res->intfs[i] == INTF_MODE_NONE) + continue; + id = i + INTF_0; + ret = _dpu_rm_reserve_intf(rm, rsvp, id, + DPU_HW_BLK_INTF, hw_res->needs_cdm); + if (ret) + return ret; + } + + return ret; +} + +static int _dpu_rm_make_next_rsvp( + struct dpu_rm *rm, + struct drm_encoder *enc, + struct drm_crtc_state *crtc_state, + struct drm_connector_state *conn_state, + struct dpu_rm_rsvp *rsvp, + struct dpu_rm_requirements *reqs) +{ + int ret; + struct dpu_rm_topology_def topology; + + /* Create reservation info, tag reserved blocks with it as we go */ + rsvp->seq = ++rm->rsvp_next_seq; + rsvp->enc_id = enc->base.id; + rsvp->topology = reqs->topology->top_name; + list_add_tail(&rsvp->list, &rm->rsvps); + + ret = _dpu_rm_reserve_lms(rm, rsvp, reqs); + if (ret) { + DPU_ERROR("unable to find appropriate mixers\n"); + return ret; + } + + /* + * Do assignment preferring to give away low-resource CTLs first: + * - Check mixers without Split Display + * - Only then allow to grab from CTLs with split display capability + */ + _dpu_rm_reserve_ctls(rm, rsvp, reqs->topology); + if (ret && !reqs->topology->needs_split_display) { + memcpy(&topology, reqs->topology, sizeof(topology)); + topology.needs_split_display = true; + _dpu_rm_reserve_ctls(rm, rsvp, &topology); + } + if (ret) { + DPU_ERROR("unable to find appropriate CTL\n"); + return ret; + } + + /* Assign INTFs and blks whose usage is tied to them: CTL & CDM */ + ret = _dpu_rm_reserve_intf_related_hw(rm, rsvp, &reqs->hw_res); + if (ret) + return ret; + + return ret; +} + +static int _dpu_rm_populate_requirements( + struct dpu_rm *rm, + struct drm_encoder *enc, + struct drm_crtc_state *crtc_state, + struct drm_connector_state *conn_state, + struct dpu_rm_requirements *reqs, + struct msm_display_topology req_topology) +{ + int i; + + memset(reqs, 0, sizeof(*reqs)); + + dpu_encoder_get_hw_resources(enc, &reqs->hw_res, conn_state); + + for (i = 0; i < DPU_RM_TOPOLOGY_MAX; i++) { + if (RM_IS_TOPOLOGY_MATCH(g_top_table[i], + req_topology)) { + reqs->topology = &g_top_table[i]; + break; + } + } + + if (!reqs->topology) { + DPU_ERROR("invalid topology for the display\n"); + return -EINVAL; + } + + /** + * Set the requirement based on caps if not set from user space + * This will ensure to select LM tied with DS blocks + * Currently, DS blocks are tied with LM 0 and LM 1 (primary display) + */ + if (!RM_RQ_DS(reqs) && rm->hw_mdp->caps->has_dest_scaler && + conn_state->connector->connector_type == DRM_MODE_CONNECTOR_DSI) + reqs->top_ctrl |= BIT(DPU_RM_TOPCTL_DS); + + DRM_DEBUG_KMS("top_ctrl: 0x%llX num_h_tiles: %d\n", reqs->top_ctrl, + reqs->hw_res.display_num_of_h_tiles); + DRM_DEBUG_KMS("num_lm: %d num_ctl: %d topology: %d split_display: %d\n", + reqs->topology->num_lm, reqs->topology->num_ctl, + reqs->topology->top_name, + reqs->topology->needs_split_display); + + return 0; +} + +static struct dpu_rm_rsvp *_dpu_rm_get_rsvp( + struct dpu_rm *rm, + struct drm_encoder *enc) +{ + struct dpu_rm_rsvp *i; + + if (!rm || !enc) { + DPU_ERROR("invalid params\n"); + return NULL; + } + + if (list_empty(&rm->rsvps)) + return NULL; + + list_for_each_entry(i, &rm->rsvps, list) + if (i->enc_id == enc->base.id) + return i; + + return NULL; +} + +static struct drm_connector *_dpu_rm_get_connector( + struct drm_encoder *enc) +{ + struct drm_connector *conn = NULL; + struct list_head *connector_list = + &enc->dev->mode_config.connector_list; + + list_for_each_entry(conn, connector_list, head) + if (conn->encoder == enc) + return conn; + + return NULL; +} + +/** + * _dpu_rm_release_rsvp - release resources and release a reservation + * @rm: KMS handle + * @rsvp: RSVP pointer to release and release resources for + */ +static void _dpu_rm_release_rsvp( + struct dpu_rm *rm, + struct dpu_rm_rsvp *rsvp, + struct drm_connector *conn) +{ + struct dpu_rm_rsvp *rsvp_c, *rsvp_n; + struct dpu_rm_hw_blk *blk; + enum dpu_hw_blk_type type; + + if (!rsvp) + return; + + DPU_DEBUG("rel rsvp %d enc %d\n", rsvp->seq, rsvp->enc_id); + + list_for_each_entry_safe(rsvp_c, rsvp_n, &rm->rsvps, list) { + if (rsvp == rsvp_c) { + list_del(&rsvp_c->list); + break; + } + } + + for (type = 0; type < DPU_HW_BLK_MAX; type++) { + list_for_each_entry(blk, &rm->hw_blks[type], list) { + if (blk->rsvp == rsvp) { + blk->rsvp = NULL; + DPU_DEBUG("rel rsvp %d enc %d %d %d\n", + rsvp->seq, rsvp->enc_id, + blk->type, blk->id); + } + if (blk->rsvp_nxt == rsvp) { + blk->rsvp_nxt = NULL; + DPU_DEBUG("rel rsvp_nxt %d enc %d %d %d\n", + rsvp->seq, rsvp->enc_id, + blk->type, blk->id); + } + } + } + + kfree(rsvp); +} + +void dpu_rm_release(struct dpu_rm *rm, struct drm_encoder *enc) +{ + struct dpu_rm_rsvp *rsvp; + struct drm_connector *conn; + + if (!rm || !enc) { + DPU_ERROR("invalid params\n"); + return; + } + + mutex_lock(&rm->rm_lock); + + rsvp = _dpu_rm_get_rsvp(rm, enc); + if (!rsvp) { + DPU_ERROR("failed to find rsvp for enc %d\n", enc->base.id); + goto end; + } + + conn = _dpu_rm_get_connector(enc); + if (!conn) { + DPU_ERROR("failed to get connector for enc %d\n", enc->base.id); + goto end; + } + + _dpu_rm_release_rsvp(rm, rsvp, conn); +end: + mutex_unlock(&rm->rm_lock); +} + +static int _dpu_rm_commit_rsvp( + struct dpu_rm *rm, + struct dpu_rm_rsvp *rsvp, + struct drm_connector_state *conn_state) +{ + struct dpu_rm_hw_blk *blk; + enum dpu_hw_blk_type type; + int ret = 0; + + /* Swap next rsvp to be the active */ + for (type = 0; type < DPU_HW_BLK_MAX; type++) { + list_for_each_entry(blk, &rm->hw_blks[type], list) { + if (blk->rsvp_nxt) { + blk->rsvp = blk->rsvp_nxt; + blk->rsvp_nxt = NULL; + } + } + } + + if (!ret) + DRM_DEBUG_KMS("rsrv enc %d topology %d\n", rsvp->enc_id, + rsvp->topology); + + return ret; +} + +int dpu_rm_reserve( + struct dpu_rm *rm, + struct drm_encoder *enc, + struct drm_crtc_state *crtc_state, + struct drm_connector_state *conn_state, + struct msm_display_topology topology, + bool test_only) +{ + struct dpu_rm_rsvp *rsvp_cur, *rsvp_nxt; + struct dpu_rm_requirements reqs; + int ret; + + if (!rm || !enc || !crtc_state || !conn_state) { + DPU_ERROR("invalid arguments\n"); + return -EINVAL; + } + + /* Check if this is just a page-flip */ + if (!drm_atomic_crtc_needs_modeset(crtc_state)) + return 0; + + DRM_DEBUG_KMS("reserving hw for conn %d enc %d crtc %d test_only %d\n", + conn_state->connector->base.id, enc->base.id, + crtc_state->crtc->base.id, test_only); + + mutex_lock(&rm->rm_lock); + + _dpu_rm_print_rsvps(rm, DPU_RM_STAGE_BEGIN); + + ret = _dpu_rm_populate_requirements(rm, enc, crtc_state, + conn_state, &reqs, topology); + if (ret) { + DPU_ERROR("failed to populate hw requirements\n"); + goto end; + } + + /* + * We only support one active reservation per-hw-block. But to implement + * transactional semantics for test-only, and for allowing failure while + * modifying your existing reservation, over the course of this + * function we can have two reservations: + * Current: Existing reservation + * Next: Proposed reservation. The proposed reservation may fail, or may + * be discarded if in test-only mode. + * If reservation is successful, and we're not in test-only, then we + * replace the current with the next. + */ + rsvp_nxt = kzalloc(sizeof(*rsvp_nxt), GFP_KERNEL); + if (!rsvp_nxt) { + ret = -ENOMEM; + goto end; + } + + rsvp_cur = _dpu_rm_get_rsvp(rm, enc); + + /* + * User can request that we clear out any reservation during the + * atomic_check phase by using this CLEAR bit + */ + if (rsvp_cur && test_only && RM_RQ_CLEAR(&reqs)) { + DPU_DEBUG("test_only & CLEAR: clear rsvp[s%de%d]\n", + rsvp_cur->seq, rsvp_cur->enc_id); + _dpu_rm_release_rsvp(rm, rsvp_cur, conn_state->connector); + rsvp_cur = NULL; + _dpu_rm_print_rsvps(rm, DPU_RM_STAGE_AFTER_CLEAR); + } + + /* Check the proposed reservation, store it in hw's "next" field */ + ret = _dpu_rm_make_next_rsvp(rm, enc, crtc_state, conn_state, + rsvp_nxt, &reqs); + + _dpu_rm_print_rsvps(rm, DPU_RM_STAGE_AFTER_RSVPNEXT); + + if (ret) { + DPU_ERROR("failed to reserve hw resources: %d\n", ret); + _dpu_rm_release_rsvp(rm, rsvp_nxt, conn_state->connector); + } else if (test_only && !RM_RQ_LOCK(&reqs)) { + /* + * Normally, if test_only, test the reservation and then undo + * However, if the user requests LOCK, then keep the reservation + * made during the atomic_check phase. + */ + DPU_DEBUG("test_only: discard test rsvp[s%de%d]\n", + rsvp_nxt->seq, rsvp_nxt->enc_id); + _dpu_rm_release_rsvp(rm, rsvp_nxt, conn_state->connector); + } else { + if (test_only && RM_RQ_LOCK(&reqs)) + DPU_DEBUG("test_only & LOCK: lock rsvp[s%de%d]\n", + rsvp_nxt->seq, rsvp_nxt->enc_id); + + _dpu_rm_release_rsvp(rm, rsvp_cur, conn_state->connector); + + ret = _dpu_rm_commit_rsvp(rm, rsvp_nxt, conn_state); + } + + _dpu_rm_print_rsvps(rm, DPU_RM_STAGE_FINAL); + +end: + mutex_unlock(&rm->rm_lock); + + return ret; +} diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_rm.h b/drivers/gpu/drm/msm/disp/dpu1/dpu_rm.h new file mode 100644 index 000000000000..ffd1841a6067 --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_rm.h @@ -0,0 +1,199 @@ +/* + * Copyright (c) 2016-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + */ + +#ifndef __DPU_RM_H__ +#define __DPU_RM_H__ + +#include <linux/list.h> + +#include "msm_kms.h" +#include "dpu_hw_top.h" + +/** + * enum dpu_rm_topology_name - HW resource use case in use by connector + * @DPU_RM_TOPOLOGY_NONE: No topology in use currently + * @DPU_RM_TOPOLOGY_SINGLEPIPE: 1 LM, 1 PP, 1 INTF/WB + * @DPU_RM_TOPOLOGY_DUALPIPE: 2 LM, 2 PP, 2 INTF/WB + * @DPU_RM_TOPOLOGY_DUALPIPE_3DMERGE: 2 LM, 2 PP, 3DMux, 1 INTF/WB + */ +enum dpu_rm_topology_name { + DPU_RM_TOPOLOGY_NONE = 0, + DPU_RM_TOPOLOGY_SINGLEPIPE, + DPU_RM_TOPOLOGY_DUALPIPE, + DPU_RM_TOPOLOGY_DUALPIPE_3DMERGE, + DPU_RM_TOPOLOGY_MAX, +}; + +/** + * enum dpu_rm_topology_control - HW resource use case in use by connector + * @DPU_RM_TOPCTL_RESERVE_LOCK: If set, in AtomicTest phase, after a successful + * test, reserve the resources for this display. + * Normal behavior would not impact the reservation + * list during the AtomicTest phase. + * @DPU_RM_TOPCTL_RESERVE_CLEAR: If set, in AtomicTest phase, before testing, + * release any reservation held by this display. + * Normal behavior would not impact the + * reservation list during the AtomicTest phase. + * @DPU_RM_TOPCTL_DS : Require layer mixers with DS capabilities + */ +enum dpu_rm_topology_control { + DPU_RM_TOPCTL_RESERVE_LOCK, + DPU_RM_TOPCTL_RESERVE_CLEAR, + DPU_RM_TOPCTL_DS, +}; + +/** + * struct dpu_rm - DPU dynamic hardware resource manager + * @dev: device handle for event logging purposes + * @rsvps: list of hardware reservations by each crtc->encoder->connector + * @hw_blks: array of lists of hardware resources present in the system, one + * list per type of hardware block + * @hw_mdp: hardware object for mdp_top + * @lm_max_width: cached layer mixer maximum width + * @rsvp_next_seq: sequence number for next reservation for debugging purposes + * @rm_lock: resource manager mutex + */ +struct dpu_rm { + struct drm_device *dev; + struct list_head rsvps; + struct list_head hw_blks[DPU_HW_BLK_MAX]; + struct dpu_hw_mdp *hw_mdp; + uint32_t lm_max_width; + uint32_t rsvp_next_seq; + struct mutex rm_lock; +}; + +/** + * struct dpu_rm_hw_blk - resource manager internal structure + * forward declaration for single iterator definition without void pointer + */ +struct dpu_rm_hw_blk; + +/** + * struct dpu_rm_hw_iter - iterator for use with dpu_rm + * @hw: dpu_hw object requested, or NULL on failure + * @blk: dpu_rm internal block representation. Clients ignore. Used as iterator. + * @enc_id: DRM ID of Encoder client wishes to search for, or 0 for Any Encoder + * @type: Hardware Block Type client wishes to search for. + */ +struct dpu_rm_hw_iter { + void *hw; + struct dpu_rm_hw_blk *blk; + uint32_t enc_id; + enum dpu_hw_blk_type type; +}; + +/** + * dpu_rm_init - Read hardware catalog and create reservation tracking objects + * for all HW blocks. + * @rm: DPU Resource Manager handle + * @cat: Pointer to hardware catalog + * @mmio: mapped register io address of MDP + * @dev: device handle for event logging purposes + * @Return: 0 on Success otherwise -ERROR + */ +int dpu_rm_init(struct dpu_rm *rm, + struct dpu_mdss_cfg *cat, + void __iomem *mmio, + struct drm_device *dev); + +/** + * dpu_rm_destroy - Free all memory allocated by dpu_rm_init + * @rm: DPU Resource Manager handle + * @Return: 0 on Success otherwise -ERROR + */ +int dpu_rm_destroy(struct dpu_rm *rm); + +/** + * dpu_rm_reserve - Given a CRTC->Encoder->Connector display chain, analyze + * the use connections and user requirements, specified through related + * topology control properties, and reserve hardware blocks to that + * display chain. + * HW blocks can then be accessed through dpu_rm_get_* functions. + * HW Reservations should be released via dpu_rm_release_hw. + * @rm: DPU Resource Manager handle + * @drm_enc: DRM Encoder handle + * @crtc_state: Proposed Atomic DRM CRTC State handle + * @conn_state: Proposed Atomic DRM Connector State handle + * @topology: Pointer to topology info for the display + * @test_only: Atomic-Test phase, discard results (unless property overrides) + * @Return: 0 on Success otherwise -ERROR + */ +int dpu_rm_reserve(struct dpu_rm *rm, + struct drm_encoder *drm_enc, + struct drm_crtc_state *crtc_state, + struct drm_connector_state *conn_state, + struct msm_display_topology topology, + bool test_only); + +/** + * dpu_rm_reserve - Given the encoder for the display chain, release any + * HW blocks previously reserved for that use case. + * @rm: DPU Resource Manager handle + * @enc: DRM Encoder handle + * @Return: 0 on Success otherwise -ERROR + */ +void dpu_rm_release(struct dpu_rm *rm, struct drm_encoder *enc); + +/** + * dpu_rm_get_mdp - Retrieve HW block for MDP TOP. + * This is never reserved, and is usable by any display. + * @rm: DPU Resource Manager handle + * @Return: Pointer to hw block or NULL + */ +struct dpu_hw_mdp *dpu_rm_get_mdp(struct dpu_rm *rm); + +/** + * dpu_rm_init_hw_iter - setup given iterator for new iteration over hw list + * using dpu_rm_get_hw + * @iter: iter object to initialize + * @enc_id: DRM ID of Encoder client wishes to search for, or 0 for Any Encoder + * @type: Hardware Block Type client wishes to search for. + */ +void dpu_rm_init_hw_iter( + struct dpu_rm_hw_iter *iter, + uint32_t enc_id, + enum dpu_hw_blk_type type); +/** + * dpu_rm_get_hw - retrieve reserved hw object given encoder and hw type + * Meant to do a single pass through the hardware list to iteratively + * retrieve hardware blocks of a given type for a given encoder. + * Initialize an iterator object. + * Set hw block type of interest. Set encoder id of interest, 0 for any. + * Function returns first hw of type for that encoder. + * Subsequent calls will return the next reserved hw of that type in-order. + * Iterator HW pointer will be null on failure to find hw. + * @rm: DPU Resource Manager handle + * @iter: iterator object + * @Return: true on match found, false on no match found + */ +bool dpu_rm_get_hw(struct dpu_rm *rm, struct dpu_rm_hw_iter *iter); + +/** + * dpu_rm_check_property_topctl - validate property bitmask before it is set + * @val: user's proposed topology control bitmask + * @Return: 0 on success or error + */ +int dpu_rm_check_property_topctl(uint64_t val); + +/** + * dpu_rm_get_topology_name - returns the name of the the given topology + * definition + * @topology: topology definition + * @Return: name of the topology + */ +enum dpu_rm_topology_name +dpu_rm_get_topology_name(struct msm_display_topology topology); + +#endif /* __DPU_RM_H__ */ diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_trace.h b/drivers/gpu/drm/msm/disp/dpu1/dpu_trace.h new file mode 100644 index 000000000000..ae0ca5076238 --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_trace.h @@ -0,0 +1,1007 @@ +/* Copyright (c) 2014-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#if !defined(_DPU_TRACE_H_) || defined(TRACE_HEADER_MULTI_READ) +#define _DPU_TRACE_H_ + +#include <linux/stringify.h> +#include <linux/types.h> +#include <linux/tracepoint.h> + +#include <drm/drm_rect.h> +#include "dpu_crtc.h" +#include "dpu_encoder_phys.h" +#include "dpu_hw_mdss.h" +#include "dpu_hw_vbif.h" +#include "dpu_plane.h" + +#undef TRACE_SYSTEM +#define TRACE_SYSTEM dpu +#undef TRACE_INCLUDE_FILE +#define TRACE_INCLUDE_FILE dpu_trace + +TRACE_EVENT(dpu_perf_set_qos_luts, + TP_PROTO(u32 pnum, u32 fmt, bool rt, u32 fl, + u32 lut, u32 lut_usage), + TP_ARGS(pnum, fmt, rt, fl, lut, lut_usage), + TP_STRUCT__entry( + __field(u32, pnum) + __field(u32, fmt) + __field(bool, rt) + __field(u32, fl) + __field(u64, lut) + __field(u32, lut_usage) + ), + TP_fast_assign( + __entry->pnum = pnum; + __entry->fmt = fmt; + __entry->rt = rt; + __entry->fl = fl; + __entry->lut = lut; + __entry->lut_usage = lut_usage; + ), + TP_printk("pnum=%d fmt=%x rt=%d fl=%d lut=0x%llx lut_usage=%d", + __entry->pnum, __entry->fmt, + __entry->rt, __entry->fl, + __entry->lut, __entry->lut_usage) +); + +TRACE_EVENT(dpu_perf_set_danger_luts, + TP_PROTO(u32 pnum, u32 fmt, u32 mode, u32 danger_lut, + u32 safe_lut), + TP_ARGS(pnum, fmt, mode, danger_lut, safe_lut), + TP_STRUCT__entry( + __field(u32, pnum) + __field(u32, fmt) + __field(u32, mode) + __field(u32, danger_lut) + __field(u32, safe_lut) + ), + TP_fast_assign( + __entry->pnum = pnum; + __entry->fmt = fmt; + __entry->mode = mode; + __entry->danger_lut = danger_lut; + __entry->safe_lut = safe_lut; + ), + TP_printk("pnum=%d fmt=%x mode=%d luts[0x%x, 0x%x]", + __entry->pnum, __entry->fmt, + __entry->mode, __entry->danger_lut, + __entry->safe_lut) +); + +TRACE_EVENT(dpu_perf_set_ot, + TP_PROTO(u32 pnum, u32 xin_id, u32 rd_lim, u32 vbif_idx), + TP_ARGS(pnum, xin_id, rd_lim, vbif_idx), + TP_STRUCT__entry( + __field(u32, pnum) + __field(u32, xin_id) + __field(u32, rd_lim) + __field(u32, vbif_idx) + ), + TP_fast_assign( + __entry->pnum = pnum; + __entry->xin_id = xin_id; + __entry->rd_lim = rd_lim; + __entry->vbif_idx = vbif_idx; + ), + TP_printk("pnum:%d xin_id:%d ot:%d vbif:%d", + __entry->pnum, __entry->xin_id, __entry->rd_lim, + __entry->vbif_idx) +) + +TRACE_EVENT(dpu_perf_update_bus, + TP_PROTO(int client, unsigned long long ab_quota, + unsigned long long ib_quota), + TP_ARGS(client, ab_quota, ib_quota), + TP_STRUCT__entry( + __field(int, client) + __field(u64, ab_quota) + __field(u64, ib_quota) + ), + TP_fast_assign( + __entry->client = client; + __entry->ab_quota = ab_quota; + __entry->ib_quota = ib_quota; + ), + TP_printk("Request client:%d ab=%llu ib=%llu", + __entry->client, + __entry->ab_quota, + __entry->ib_quota) +) + + +TRACE_EVENT(dpu_cmd_release_bw, + TP_PROTO(u32 crtc_id), + TP_ARGS(crtc_id), + TP_STRUCT__entry( + __field(u32, crtc_id) + ), + TP_fast_assign( + __entry->crtc_id = crtc_id; + ), + TP_printk("crtc:%d", __entry->crtc_id) +); + +TRACE_EVENT(tracing_mark_write, + TP_PROTO(int pid, const char *name, bool trace_begin), + TP_ARGS(pid, name, trace_begin), + TP_STRUCT__entry( + __field(int, pid) + __string(trace_name, name) + __field(bool, trace_begin) + ), + TP_fast_assign( + __entry->pid = pid; + __assign_str(trace_name, name); + __entry->trace_begin = trace_begin; + ), + TP_printk("%s|%d|%s", __entry->trace_begin ? "B" : "E", + __entry->pid, __get_str(trace_name)) +) + +TRACE_EVENT(dpu_trace_counter, + TP_PROTO(int pid, char *name, int value), + TP_ARGS(pid, name, value), + TP_STRUCT__entry( + __field(int, pid) + __string(counter_name, name) + __field(int, value) + ), + TP_fast_assign( + __entry->pid = current->tgid; + __assign_str(counter_name, name); + __entry->value = value; + ), + TP_printk("%d|%s|%d", __entry->pid, + __get_str(counter_name), __entry->value) +) + +TRACE_EVENT(dpu_perf_crtc_update, + TP_PROTO(u32 crtc, u64 bw_ctl_mnoc, u64 bw_ctl_llcc, + u64 bw_ctl_ebi, u32 core_clk_rate, + bool stop_req, u32 update_bus, u32 update_clk), + TP_ARGS(crtc, bw_ctl_mnoc, bw_ctl_llcc, bw_ctl_ebi, core_clk_rate, + stop_req, update_bus, update_clk), + TP_STRUCT__entry( + __field(u32, crtc) + __field(u64, bw_ctl_mnoc) + __field(u64, bw_ctl_llcc) + __field(u64, bw_ctl_ebi) + __field(u32, core_clk_rate) + __field(bool, stop_req) + __field(u32, update_bus) + __field(u32, update_clk) + ), + TP_fast_assign( + __entry->crtc = crtc; + __entry->bw_ctl_mnoc = bw_ctl_mnoc; + __entry->bw_ctl_llcc = bw_ctl_llcc; + __entry->bw_ctl_ebi = bw_ctl_ebi; + __entry->core_clk_rate = core_clk_rate; + __entry->stop_req = stop_req; + __entry->update_bus = update_bus; + __entry->update_clk = update_clk; + ), + TP_printk( + "crtc=%d bw_mnoc=%llu bw_llcc=%llu bw_ebi=%llu clk_rate=%u stop_req=%d u_bus=%d u_clk=%d", + __entry->crtc, + __entry->bw_ctl_mnoc, + __entry->bw_ctl_llcc, + __entry->bw_ctl_ebi, + __entry->core_clk_rate, + __entry->stop_req, + __entry->update_bus, + __entry->update_clk) +); + +DECLARE_EVENT_CLASS(dpu_enc_irq_template, + TP_PROTO(uint32_t drm_id, enum dpu_intr_idx intr_idx, int hw_idx, + int irq_idx), + TP_ARGS(drm_id, intr_idx, hw_idx, irq_idx), + TP_STRUCT__entry( + __field( uint32_t, drm_id ) + __field( enum dpu_intr_idx, intr_idx ) + __field( int, hw_idx ) + __field( int, irq_idx ) + ), + TP_fast_assign( + __entry->drm_id = drm_id; + __entry->intr_idx = intr_idx; + __entry->hw_idx = hw_idx; + __entry->irq_idx = irq_idx; + ), + TP_printk("id=%u, intr=%d, hw=%d, irq=%d", + __entry->drm_id, __entry->intr_idx, __entry->hw_idx, + __entry->irq_idx) +); +DEFINE_EVENT(dpu_enc_irq_template, dpu_enc_irq_register_success, + TP_PROTO(uint32_t drm_id, enum dpu_intr_idx intr_idx, int hw_idx, + int irq_idx), + TP_ARGS(drm_id, intr_idx, hw_idx, irq_idx) +); +DEFINE_EVENT(dpu_enc_irq_template, dpu_enc_irq_unregister_success, + TP_PROTO(uint32_t drm_id, enum dpu_intr_idx intr_idx, int hw_idx, + int irq_idx), + TP_ARGS(drm_id, intr_idx, hw_idx, irq_idx) +); + +TRACE_EVENT(dpu_enc_irq_wait_success, + TP_PROTO(uint32_t drm_id, enum dpu_intr_idx intr_idx, int hw_idx, + int irq_idx, enum dpu_pingpong pp_idx, int atomic_cnt), + TP_ARGS(drm_id, intr_idx, hw_idx, irq_idx, pp_idx, atomic_cnt), + TP_STRUCT__entry( + __field( uint32_t, drm_id ) + __field( enum dpu_intr_idx, intr_idx ) + __field( int, hw_idx ) + __field( int, irq_idx ) + __field( enum dpu_pingpong, pp_idx ) + __field( int, atomic_cnt ) + ), + TP_fast_assign( + __entry->drm_id = drm_id; + __entry->intr_idx = intr_idx; + __entry->hw_idx = hw_idx; + __entry->irq_idx = irq_idx; + __entry->pp_idx = pp_idx; + __entry->atomic_cnt = atomic_cnt; + ), + TP_printk("id=%u, intr=%d, hw=%d, irq=%d, pp=%d, atomic_cnt=%d", + __entry->drm_id, __entry->intr_idx, __entry->hw_idx, + __entry->irq_idx, __entry->pp_idx, __entry->atomic_cnt) +); + +DECLARE_EVENT_CLASS(dpu_drm_obj_template, + TP_PROTO(uint32_t drm_id), + TP_ARGS(drm_id), + TP_STRUCT__entry( + __field( uint32_t, drm_id ) + ), + TP_fast_assign( + __entry->drm_id = drm_id; + ), + TP_printk("id=%u", __entry->drm_id) +); +DEFINE_EVENT(dpu_drm_obj_template, dpu_enc_atomic_check, + TP_PROTO(uint32_t drm_id), + TP_ARGS(drm_id) +); +DEFINE_EVENT(dpu_drm_obj_template, dpu_enc_mode_set, + TP_PROTO(uint32_t drm_id), + TP_ARGS(drm_id) +); +DEFINE_EVENT(dpu_drm_obj_template, dpu_enc_disable, + TP_PROTO(uint32_t drm_id), + TP_ARGS(drm_id) +); +DEFINE_EVENT(dpu_drm_obj_template, dpu_enc_kickoff, + TP_PROTO(uint32_t drm_id), + TP_ARGS(drm_id) +); +DEFINE_EVENT(dpu_drm_obj_template, dpu_enc_prepare_kickoff, + TP_PROTO(uint32_t drm_id), + TP_ARGS(drm_id) +); +DEFINE_EVENT(dpu_drm_obj_template, dpu_enc_prepare_kickoff_reset, + TP_PROTO(uint32_t drm_id), + TP_ARGS(drm_id) +); +DEFINE_EVENT(dpu_drm_obj_template, dpu_crtc_complete_flip, + TP_PROTO(uint32_t drm_id), + TP_ARGS(drm_id) +); +DEFINE_EVENT(dpu_drm_obj_template, dpu_crtc_vblank_cb, + TP_PROTO(uint32_t drm_id), + TP_ARGS(drm_id) +); +DEFINE_EVENT(dpu_drm_obj_template, dpu_crtc_complete_commit, + TP_PROTO(uint32_t drm_id), + TP_ARGS(drm_id) +); +DEFINE_EVENT(dpu_drm_obj_template, dpu_kms_enc_enable, + TP_PROTO(uint32_t drm_id), + TP_ARGS(drm_id) +); +DEFINE_EVENT(dpu_drm_obj_template, dpu_kms_commit, + TP_PROTO(uint32_t drm_id), + TP_ARGS(drm_id) +); +DEFINE_EVENT(dpu_drm_obj_template, dpu_kms_wait_for_commit_done, + TP_PROTO(uint32_t drm_id), + TP_ARGS(drm_id) +); + +TRACE_EVENT(dpu_enc_enable, + TP_PROTO(uint32_t drm_id, int hdisplay, int vdisplay), + TP_ARGS(drm_id, hdisplay, vdisplay), + TP_STRUCT__entry( + __field( uint32_t, drm_id ) + __field( int, hdisplay ) + __field( int, vdisplay ) + ), + TP_fast_assign( + __entry->drm_id = drm_id; + __entry->hdisplay = hdisplay; + __entry->vdisplay = vdisplay; + ), + TP_printk("id=%u, mode=%dx%d", + __entry->drm_id, __entry->hdisplay, __entry->vdisplay) +); + +DECLARE_EVENT_CLASS(dpu_enc_keyval_template, + TP_PROTO(uint32_t drm_id, int val), + TP_ARGS(drm_id, val), + TP_STRUCT__entry( + __field( uint32_t, drm_id ) + __field( int, val ) + ), + TP_fast_assign( + __entry->drm_id = drm_id; + __entry->val = val; + ), + TP_printk("id=%u, val=%d", __entry->drm_id, __entry->val) +); +DEFINE_EVENT(dpu_enc_keyval_template, dpu_enc_underrun_cb, + TP_PROTO(uint32_t drm_id, int count), + TP_ARGS(drm_id, count) +); +DEFINE_EVENT(dpu_enc_keyval_template, dpu_enc_trigger_start, + TP_PROTO(uint32_t drm_id, int ctl_idx), + TP_ARGS(drm_id, ctl_idx) +); + +TRACE_EVENT(dpu_enc_atomic_check_flags, + TP_PROTO(uint32_t drm_id, unsigned int flags, int private_flags), + TP_ARGS(drm_id, flags, private_flags), + TP_STRUCT__entry( + __field( uint32_t, drm_id ) + __field( unsigned int, flags ) + __field( int, private_flags ) + ), + TP_fast_assign( + __entry->drm_id = drm_id; + __entry->flags = flags; + __entry->private_flags = private_flags; + ), + TP_printk("id=%u, flags=%u, private_flags=%d", + __entry->drm_id, __entry->flags, __entry->private_flags) +); + +DECLARE_EVENT_CLASS(dpu_enc_id_enable_template, + TP_PROTO(uint32_t drm_id, bool enable), + TP_ARGS(drm_id, enable), + TP_STRUCT__entry( + __field( uint32_t, drm_id ) + __field( bool, enable ) + ), + TP_fast_assign( + __entry->drm_id = drm_id; + __entry->enable = enable; + ), + TP_printk("id=%u, enable=%s", + __entry->drm_id, __entry->enable ? "true" : "false") +); +DEFINE_EVENT(dpu_enc_id_enable_template, dpu_enc_rc_helper, + TP_PROTO(uint32_t drm_id, bool enable), + TP_ARGS(drm_id, enable) +); +DEFINE_EVENT(dpu_enc_id_enable_template, dpu_enc_vblank_cb, + TP_PROTO(uint32_t drm_id, bool enable), + TP_ARGS(drm_id, enable) +); +DEFINE_EVENT(dpu_enc_id_enable_template, dpu_enc_frame_event_cb, + TP_PROTO(uint32_t drm_id, bool enable), + TP_ARGS(drm_id, enable) +); +DEFINE_EVENT(dpu_enc_id_enable_template, dpu_enc_phys_cmd_connect_te, + TP_PROTO(uint32_t drm_id, bool enable), + TP_ARGS(drm_id, enable) +); + +TRACE_EVENT(dpu_enc_rc, + TP_PROTO(uint32_t drm_id, u32 sw_event, bool idle_pc_supported, + int rc_state, const char *stage), + TP_ARGS(drm_id, sw_event, idle_pc_supported, rc_state, stage), + TP_STRUCT__entry( + __field( uint32_t, drm_id ) + __field( u32, sw_event ) + __field( bool, idle_pc_supported ) + __field( int, rc_state ) + __string( stage_str, stage ) + ), + TP_fast_assign( + __entry->drm_id = drm_id; + __entry->sw_event = sw_event; + __entry->idle_pc_supported = idle_pc_supported; + __entry->rc_state = rc_state; + __assign_str(stage_str, stage); + ), + TP_printk("%s: id:%u, sw_event:%d, idle_pc_supported:%s, rc_state:%d\n", + __get_str(stage_str), __entry->drm_id, __entry->sw_event, + __entry->idle_pc_supported ? "true" : "false", + __entry->rc_state) +); + +TRACE_EVENT(dpu_enc_frame_done_cb_not_busy, + TP_PROTO(uint32_t drm_id, u32 event, enum dpu_intf intf_idx), + TP_ARGS(drm_id, event, intf_idx), + TP_STRUCT__entry( + __field( uint32_t, drm_id ) + __field( u32, event ) + __field( enum dpu_intf, intf_idx ) + ), + TP_fast_assign( + __entry->drm_id = drm_id; + __entry->event = event; + __entry->intf_idx = intf_idx; + ), + TP_printk("id=%u, event=%u, intf=%d", __entry->drm_id, __entry->event, + __entry->intf_idx) +); + +TRACE_EVENT(dpu_enc_frame_done_cb, + TP_PROTO(uint32_t drm_id, unsigned int idx, + unsigned long frame_busy_mask), + TP_ARGS(drm_id, idx, frame_busy_mask), + TP_STRUCT__entry( + __field( uint32_t, drm_id ) + __field( unsigned int, idx ) + __field( unsigned long, frame_busy_mask ) + ), + TP_fast_assign( + __entry->drm_id = drm_id; + __entry->idx = idx; + __entry->frame_busy_mask = frame_busy_mask; + ), + TP_printk("id=%u, idx=%u, frame_busy_mask=%lx", __entry->drm_id, + __entry->idx, __entry->frame_busy_mask) +); + +TRACE_EVENT(dpu_enc_trigger_flush, + TP_PROTO(uint32_t drm_id, enum dpu_intf intf_idx, + int pending_kickoff_cnt, int ctl_idx, u32 pending_flush_ret), + TP_ARGS(drm_id, intf_idx, pending_kickoff_cnt, ctl_idx, + pending_flush_ret), + TP_STRUCT__entry( + __field( uint32_t, drm_id ) + __field( enum dpu_intf, intf_idx ) + __field( int, pending_kickoff_cnt ) + __field( int, ctl_idx ) + __field( u32, pending_flush_ret ) + ), + TP_fast_assign( + __entry->drm_id = drm_id; + __entry->intf_idx = intf_idx; + __entry->pending_kickoff_cnt = pending_kickoff_cnt; + __entry->ctl_idx = ctl_idx; + __entry->pending_flush_ret = pending_flush_ret; + ), + TP_printk("id=%u, intf_idx=%d, pending_kickoff_cnt=%d ctl_idx=%d " + "pending_flush_ret=%u", __entry->drm_id, + __entry->intf_idx, __entry->pending_kickoff_cnt, + __entry->ctl_idx, __entry->pending_flush_ret) +); + +DECLARE_EVENT_CLASS(dpu_enc_ktime_template, + TP_PROTO(uint32_t drm_id, ktime_t time), + TP_ARGS(drm_id, time), + TP_STRUCT__entry( + __field( uint32_t, drm_id ) + __field( ktime_t, time ) + ), + TP_fast_assign( + __entry->drm_id = drm_id; + __entry->time = time; + ), + TP_printk("id=%u, time=%lld", __entry->drm_id, + ktime_to_ms(__entry->time)) +); +DEFINE_EVENT(dpu_enc_ktime_template, dpu_enc_vsync_event_work, + TP_PROTO(uint32_t drm_id, ktime_t time), + TP_ARGS(drm_id, time) +); +DEFINE_EVENT(dpu_enc_ktime_template, dpu_enc_early_kickoff, + TP_PROTO(uint32_t drm_id, ktime_t time), + TP_ARGS(drm_id, time) +); + +DECLARE_EVENT_CLASS(dpu_id_event_template, + TP_PROTO(uint32_t drm_id, u32 event), + TP_ARGS(drm_id, event), + TP_STRUCT__entry( + __field( uint32_t, drm_id ) + __field( u32, event ) + ), + TP_fast_assign( + __entry->drm_id = drm_id; + __entry->event = event; + ), + TP_printk("id=%u, event=%u", __entry->drm_id, __entry->event) +); +DEFINE_EVENT(dpu_id_event_template, dpu_enc_frame_done_timeout, + TP_PROTO(uint32_t drm_id, u32 event), + TP_ARGS(drm_id, event) +); +DEFINE_EVENT(dpu_id_event_template, dpu_crtc_frame_event_cb, + TP_PROTO(uint32_t drm_id, u32 event), + TP_ARGS(drm_id, event) +); +DEFINE_EVENT(dpu_id_event_template, dpu_crtc_handle_power_event, + TP_PROTO(uint32_t drm_id, u32 event), + TP_ARGS(drm_id, event) +); +DEFINE_EVENT(dpu_id_event_template, dpu_crtc_frame_event_done, + TP_PROTO(uint32_t drm_id, u32 event), + TP_ARGS(drm_id, event) +); +DEFINE_EVENT(dpu_id_event_template, dpu_crtc_frame_event_more_pending, + TP_PROTO(uint32_t drm_id, u32 event), + TP_ARGS(drm_id, event) +); + +TRACE_EVENT(dpu_enc_wait_event_timeout, + TP_PROTO(uint32_t drm_id, int32_t hw_id, int rc, s64 time, + s64 expected_time, int atomic_cnt), + TP_ARGS(drm_id, hw_id, rc, time, expected_time, atomic_cnt), + TP_STRUCT__entry( + __field( uint32_t, drm_id ) + __field( int32_t, hw_id ) + __field( int, rc ) + __field( s64, time ) + __field( s64, expected_time ) + __field( int, atomic_cnt ) + ), + TP_fast_assign( + __entry->drm_id = drm_id; + __entry->hw_id = hw_id; + __entry->rc = rc; + __entry->time = time; + __entry->expected_time = expected_time; + __entry->atomic_cnt = atomic_cnt; + ), + TP_printk("id=%u, hw_id=%d, rc=%d, time=%lld, expected=%lld cnt=%d", + __entry->drm_id, __entry->hw_id, __entry->rc, __entry->time, + __entry->expected_time, __entry->atomic_cnt) +); + +TRACE_EVENT(dpu_enc_phys_cmd_irq_ctrl, + TP_PROTO(uint32_t drm_id, enum dpu_pingpong pp, bool enable, + int refcnt), + TP_ARGS(drm_id, pp, enable, refcnt), + TP_STRUCT__entry( + __field( uint32_t, drm_id ) + __field( enum dpu_pingpong, pp ) + __field( bool, enable ) + __field( int, refcnt ) + ), + TP_fast_assign( + __entry->drm_id = drm_id; + __entry->pp = pp; + __entry->enable = enable; + __entry->refcnt = refcnt; + ), + TP_printk("id=%u, pp=%d, enable=%s, refcnt=%d", __entry->drm_id, + __entry->pp, __entry->enable ? "true" : "false", + __entry->refcnt) +); + +TRACE_EVENT(dpu_enc_phys_cmd_pp_tx_done, + TP_PROTO(uint32_t drm_id, enum dpu_pingpong pp, int new_count, + u32 event), + TP_ARGS(drm_id, pp, new_count, event), + TP_STRUCT__entry( + __field( uint32_t, drm_id ) + __field( enum dpu_pingpong, pp ) + __field( int, new_count ) + __field( u32, event ) + ), + TP_fast_assign( + __entry->drm_id = drm_id; + __entry->pp = pp; + __entry->new_count = new_count; + __entry->event = event; + ), + TP_printk("id=%u, pp=%d, new_count=%d, event=%u", __entry->drm_id, + __entry->pp, __entry->new_count, __entry->event) +); + +TRACE_EVENT(dpu_enc_phys_cmd_pdone_timeout, + TP_PROTO(uint32_t drm_id, enum dpu_pingpong pp, int timeout_count, + int kickoff_count, u32 event), + TP_ARGS(drm_id, pp, timeout_count, kickoff_count, event), + TP_STRUCT__entry( + __field( uint32_t, drm_id ) + __field( enum dpu_pingpong, pp ) + __field( int, timeout_count ) + __field( int, kickoff_count ) + __field( u32, event ) + ), + TP_fast_assign( + __entry->drm_id = drm_id; + __entry->pp = pp; + __entry->timeout_count = timeout_count; + __entry->kickoff_count = kickoff_count; + __entry->event = event; + ), + TP_printk("id=%u, pp=%d, timeout_count=%d, kickoff_count=%d, event=%u", + __entry->drm_id, __entry->pp, __entry->timeout_count, + __entry->kickoff_count, __entry->event) +); + +TRACE_EVENT(dpu_enc_phys_vid_post_kickoff, + TP_PROTO(uint32_t drm_id, enum dpu_intf intf_idx), + TP_ARGS(drm_id, intf_idx), + TP_STRUCT__entry( + __field( uint32_t, drm_id ) + __field( enum dpu_intf, intf_idx ) + ), + TP_fast_assign( + __entry->drm_id = drm_id; + __entry->intf_idx = intf_idx; + ), + TP_printk("id=%u, intf_idx=%d", __entry->drm_id, __entry->intf_idx) +); + +TRACE_EVENT(dpu_enc_phys_vid_irq_ctrl, + TP_PROTO(uint32_t drm_id, enum dpu_intf intf_idx, bool enable, + int refcnt), + TP_ARGS(drm_id, intf_idx, enable, refcnt), + TP_STRUCT__entry( + __field( uint32_t, drm_id ) + __field( enum dpu_intf, intf_idx ) + __field( bool, enable ) + __field( int, refcnt ) + ), + TP_fast_assign( + __entry->drm_id = drm_id; + __entry->intf_idx = intf_idx; + __entry->enable = enable; + __entry->refcnt = refcnt; + ), + TP_printk("id=%u, intf_idx=%d enable=%s refcnt=%d", __entry->drm_id, + __entry->intf_idx, __entry->enable ? "true" : "false", + __entry->drm_id) +); + +TRACE_EVENT(dpu_crtc_setup_mixer, + TP_PROTO(uint32_t crtc_id, uint32_t plane_id, + struct drm_plane_state *state, struct dpu_plane_state *pstate, + uint32_t stage_idx, enum dpu_sspp sspp, uint32_t pixel_format, + uint64_t modifier), + TP_ARGS(crtc_id, plane_id, state, pstate, stage_idx, sspp, + pixel_format, modifier), + TP_STRUCT__entry( + __field( uint32_t, crtc_id ) + __field( uint32_t, plane_id ) + __field( struct drm_plane_state*,state ) + __field( struct dpu_plane_state*,pstate ) + __field( uint32_t, stage_idx ) + __field( enum dpu_sspp, sspp ) + __field( uint32_t, pixel_format ) + __field( uint64_t, modifier ) + ), + TP_fast_assign( + __entry->crtc_id = crtc_id; + __entry->plane_id = plane_id; + __entry->state = state; + __entry->pstate = pstate; + __entry->stage_idx = stage_idx; + __entry->sspp = sspp; + __entry->pixel_format = pixel_format; + __entry->modifier = modifier; + ), + TP_printk("crtc_id:%u plane_id:%u fb_id:%u src:{%ux%u+%ux%u} " + "dst:{%ux%u+%ux%u} stage_idx:%u stage:%d, sspp:%d " + "multirect_index:%d multirect_mode:%u pix_format:%u " + "modifier:%llu", + __entry->crtc_id, __entry->plane_id, + __entry->state->fb ? __entry->state->fb->base.id : -1, + __entry->state->src_w >> 16, __entry->state->src_h >> 16, + __entry->state->src_x >> 16, __entry->state->src_y >> 16, + __entry->state->crtc_w, __entry->state->crtc_h, + __entry->state->crtc_x, __entry->state->crtc_y, + __entry->stage_idx, __entry->pstate->stage, __entry->sspp, + __entry->pstate->multirect_index, + __entry->pstate->multirect_mode, __entry->pixel_format, + __entry->modifier) +); + +TRACE_EVENT(dpu_crtc_setup_lm_bounds, + TP_PROTO(uint32_t drm_id, int mixer, struct drm_rect *bounds), + TP_ARGS(drm_id, mixer, bounds), + TP_STRUCT__entry( + __field( uint32_t, drm_id ) + __field( int, mixer ) + __field( struct drm_rect *, bounds ) + ), + TP_fast_assign( + __entry->drm_id = drm_id; + __entry->mixer = mixer; + __entry->bounds = bounds; + ), + TP_printk("id:%u mixer:%d bounds:" DRM_RECT_FMT, __entry->drm_id, + __entry->mixer, DRM_RECT_ARG(__entry->bounds)) +); + +TRACE_EVENT(dpu_crtc_vblank_enable, + TP_PROTO(uint32_t drm_id, uint32_t enc_id, bool enable, + struct dpu_crtc *crtc), + TP_ARGS(drm_id, enc_id, enable, crtc), + TP_STRUCT__entry( + __field( uint32_t, drm_id ) + __field( uint32_t, enc_id ) + __field( bool, enable ) + __field( struct dpu_crtc *, crtc ) + ), + TP_fast_assign( + __entry->drm_id = drm_id; + __entry->enc_id = enc_id; + __entry->enable = enable; + __entry->crtc = crtc; + ), + TP_printk("id:%u encoder:%u enable:%s state{enabled:%s suspend:%s " + "vblank_req:%s}", + __entry->drm_id, __entry->enc_id, + __entry->enable ? "true" : "false", + __entry->crtc->enabled ? "true" : "false", + __entry->crtc->suspend ? "true" : "false", + __entry->crtc->vblank_requested ? "true" : "false") +); + +DECLARE_EVENT_CLASS(dpu_crtc_enable_template, + TP_PROTO(uint32_t drm_id, bool enable, struct dpu_crtc *crtc), + TP_ARGS(drm_id, enable, crtc), + TP_STRUCT__entry( + __field( uint32_t, drm_id ) + __field( bool, enable ) + __field( struct dpu_crtc *, crtc ) + ), + TP_fast_assign( + __entry->drm_id = drm_id; + __entry->enable = enable; + __entry->crtc = crtc; + ), + TP_printk("id:%u enable:%s state{enabled:%s suspend:%s vblank_req:%s}", + __entry->drm_id, __entry->enable ? "true" : "false", + __entry->crtc->enabled ? "true" : "false", + __entry->crtc->suspend ? "true" : "false", + __entry->crtc->vblank_requested ? "true" : "false") +); +DEFINE_EVENT(dpu_crtc_enable_template, dpu_crtc_set_suspend, + TP_PROTO(uint32_t drm_id, bool enable, struct dpu_crtc *crtc), + TP_ARGS(drm_id, enable, crtc) +); +DEFINE_EVENT(dpu_crtc_enable_template, dpu_crtc_enable, + TP_PROTO(uint32_t drm_id, bool enable, struct dpu_crtc *crtc), + TP_ARGS(drm_id, enable, crtc) +); +DEFINE_EVENT(dpu_crtc_enable_template, dpu_crtc_disable, + TP_PROTO(uint32_t drm_id, bool enable, struct dpu_crtc *crtc), + TP_ARGS(drm_id, enable, crtc) +); +DEFINE_EVENT(dpu_crtc_enable_template, dpu_crtc_vblank, + TP_PROTO(uint32_t drm_id, bool enable, struct dpu_crtc *crtc), + TP_ARGS(drm_id, enable, crtc) +); + +TRACE_EVENT(dpu_crtc_disable_frame_pending, + TP_PROTO(uint32_t drm_id, int frame_pending), + TP_ARGS(drm_id, frame_pending), + TP_STRUCT__entry( + __field( uint32_t, drm_id ) + __field( int, frame_pending ) + ), + TP_fast_assign( + __entry->drm_id = drm_id; + __entry->frame_pending = frame_pending; + ), + TP_printk("id:%u frame_pending:%d", __entry->drm_id, + __entry->frame_pending) +); + +TRACE_EVENT(dpu_plane_set_scanout, + TP_PROTO(enum dpu_sspp index, struct dpu_hw_fmt_layout *layout, + enum dpu_sspp_multirect_index multirect_index), + TP_ARGS(index, layout, multirect_index), + TP_STRUCT__entry( + __field( enum dpu_sspp, index ) + __field( struct dpu_hw_fmt_layout*, layout ) + __field( enum dpu_sspp_multirect_index, multirect_index) + ), + TP_fast_assign( + __entry->index = index; + __entry->layout = layout; + __entry->multirect_index = multirect_index; + ), + TP_printk("index:%d layout:{%ux%u @ [%u/%u, %u/%u, %u/%u, %u/%u]} " + "multirect_index:%d", __entry->index, __entry->layout->width, + __entry->layout->height, __entry->layout->plane_addr[0], + __entry->layout->plane_size[0], + __entry->layout->plane_addr[1], + __entry->layout->plane_size[1], + __entry->layout->plane_addr[2], + __entry->layout->plane_size[2], + __entry->layout->plane_addr[3], + __entry->layout->plane_size[3], __entry->multirect_index) +); + +TRACE_EVENT(dpu_plane_disable, + TP_PROTO(uint32_t drm_id, bool is_virtual, uint32_t multirect_mode), + TP_ARGS(drm_id, is_virtual, multirect_mode), + TP_STRUCT__entry( + __field( uint32_t, drm_id ) + __field( bool, is_virtual ) + __field( uint32_t, multirect_mode ) + ), + TP_fast_assign( + __entry->drm_id = drm_id; + __entry->is_virtual = is_virtual; + __entry->multirect_mode = multirect_mode; + ), + TP_printk("id:%u is_virtual:%s multirect_mode:%u", __entry->drm_id, + __entry->is_virtual ? "true" : "false", + __entry->multirect_mode) +); + +DECLARE_EVENT_CLASS(dpu_rm_iter_template, + TP_PROTO(uint32_t id, enum dpu_hw_blk_type type, uint32_t enc_id), + TP_ARGS(id, type, enc_id), + TP_STRUCT__entry( + __field( uint32_t, id ) + __field( enum dpu_hw_blk_type, type ) + __field( uint32_t, enc_id ) + ), + TP_fast_assign( + __entry->id = id; + __entry->type = type; + __entry->enc_id = enc_id; + ), + TP_printk("id:%d type:%d enc_id:%u", __entry->id, __entry->type, + __entry->enc_id) +); +DEFINE_EVENT(dpu_rm_iter_template, dpu_rm_reserve_cdm, + TP_PROTO(uint32_t id, enum dpu_hw_blk_type type, uint32_t enc_id), + TP_ARGS(id, type, enc_id) +); +DEFINE_EVENT(dpu_rm_iter_template, dpu_rm_reserve_intf, + TP_PROTO(uint32_t id, enum dpu_hw_blk_type type, uint32_t enc_id), + TP_ARGS(id, type, enc_id) +); +DEFINE_EVENT(dpu_rm_iter_template, dpu_rm_reserve_ctls, + TP_PROTO(uint32_t id, enum dpu_hw_blk_type type, uint32_t enc_id), + TP_ARGS(id, type, enc_id) +); + +TRACE_EVENT(dpu_rm_reserve_lms, + TP_PROTO(uint32_t id, enum dpu_hw_blk_type type, uint32_t enc_id, + uint32_t pp_id), + TP_ARGS(id, type, enc_id, pp_id), + TP_STRUCT__entry( + __field( uint32_t, id ) + __field( enum dpu_hw_blk_type, type ) + __field( uint32_t, enc_id ) + __field( uint32_t, pp_id ) + ), + TP_fast_assign( + __entry->id = id; + __entry->type = type; + __entry->enc_id = enc_id; + __entry->pp_id = pp_id; + ), + TP_printk("id:%d type:%d enc_id:%u pp_id:%u", __entry->id, + __entry->type, __entry->enc_id, __entry->pp_id) +); + +TRACE_EVENT(dpu_vbif_wait_xin_halt_fail, + TP_PROTO(enum dpu_vbif index, u32 xin_id), + TP_ARGS(index, xin_id), + TP_STRUCT__entry( + __field( enum dpu_vbif, index ) + __field( u32, xin_id ) + ), + TP_fast_assign( + __entry->index = index; + __entry->xin_id = xin_id; + ), + TP_printk("index:%d xin_id:%u", __entry->index, __entry->xin_id) +); + +TRACE_EVENT(dpu_pp_connect_ext_te, + TP_PROTO(enum dpu_pingpong pp, u32 cfg), + TP_ARGS(pp, cfg), + TP_STRUCT__entry( + __field( enum dpu_pingpong, pp ) + __field( u32, cfg ) + ), + TP_fast_assign( + __entry->pp = pp; + __entry->cfg = cfg; + ), + TP_printk("pp:%d cfg:%u", __entry->pp, __entry->cfg) +); + +DECLARE_EVENT_CLASS(dpu_core_irq_idx_cnt_template, + TP_PROTO(int irq_idx, int enable_count), + TP_ARGS(irq_idx, enable_count), + TP_STRUCT__entry( + __field( int, irq_idx ) + __field( int, enable_count ) + ), + TP_fast_assign( + __entry->irq_idx = irq_idx; + __entry->enable_count = enable_count; + ), + TP_printk("irq_idx:%d enable_count:%u", __entry->irq_idx, + __entry->enable_count) +); +DEFINE_EVENT(dpu_core_irq_idx_cnt_template, dpu_core_irq_enable_idx, + TP_PROTO(int irq_idx, int enable_count), + TP_ARGS(irq_idx, enable_count) +); +DEFINE_EVENT(dpu_core_irq_idx_cnt_template, dpu_core_irq_disable_idx, + TP_PROTO(int irq_idx, int enable_count), + TP_ARGS(irq_idx, enable_count) +); + +DECLARE_EVENT_CLASS(dpu_core_irq_callback_template, + TP_PROTO(int irq_idx, struct dpu_irq_callback *callback), + TP_ARGS(irq_idx, callback), + TP_STRUCT__entry( + __field( int, irq_idx ) + __field( struct dpu_irq_callback *, callback) + ), + TP_fast_assign( + __entry->irq_idx = irq_idx; + __entry->callback = callback; + ), + TP_printk("irq_idx:%d callback:%pK", __entry->irq_idx, + __entry->callback) +); +DEFINE_EVENT(dpu_core_irq_callback_template, dpu_core_irq_register_callback, + TP_PROTO(int irq_idx, struct dpu_irq_callback *callback), + TP_ARGS(irq_idx, callback) +); +DEFINE_EVENT(dpu_core_irq_callback_template, dpu_core_irq_unregister_callback, + TP_PROTO(int irq_idx, struct dpu_irq_callback *callback), + TP_ARGS(irq_idx, callback) +); + +TRACE_EVENT(dpu_core_perf_update_clk, + TP_PROTO(struct drm_device *dev, bool stop_req, u64 clk_rate), + TP_ARGS(dev, stop_req, clk_rate), + TP_STRUCT__entry( + __field( struct drm_device *, dev ) + __field( bool, stop_req ) + __field( u64, clk_rate ) + ), + TP_fast_assign( + __entry->dev = dev; + __entry->stop_req = stop_req; + __entry->clk_rate = clk_rate; + ), + TP_printk("dev:%s stop_req:%s clk_rate:%llu", __entry->dev->unique, + __entry->stop_req ? "true" : "false", __entry->clk_rate) +); + +#define DPU_ATRACE_END(name) trace_tracing_mark_write(current->tgid, name, 0) +#define DPU_ATRACE_BEGIN(name) trace_tracing_mark_write(current->tgid, name, 1) +#define DPU_ATRACE_FUNC() DPU_ATRACE_BEGIN(__func__) + +#define DPU_ATRACE_INT(name, value) \ + trace_dpu_trace_counter(current->tgid, name, value) + +#endif /* _DPU_TRACE_H_ */ + +/* This part must be outside protection */ +#undef TRACE_INCLUDE_PATH +#define TRACE_INCLUDE_PATH . +#include <trace/define_trace.h> diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_vbif.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_vbif.c new file mode 100644 index 000000000000..295528292296 --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_vbif.c @@ -0,0 +1,384 @@ +/* Copyright (c) 2015-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#define pr_fmt(fmt) "[drm:%s:%d] " fmt, __func__, __LINE__ + +#include <linux/debugfs.h> + +#include "dpu_vbif.h" +#include "dpu_hw_vbif.h" +#include "dpu_trace.h" + +/** + * _dpu_vbif_wait_for_xin_halt - wait for the xin to halt + * @vbif: Pointer to hardware vbif driver + * @xin_id: Client interface identifier + * @return: 0 if success; error code otherwise + */ +static int _dpu_vbif_wait_for_xin_halt(struct dpu_hw_vbif *vbif, u32 xin_id) +{ + ktime_t timeout; + bool status; + int rc; + + if (!vbif || !vbif->cap || !vbif->ops.get_halt_ctrl) { + DPU_ERROR("invalid arguments vbif %d\n", vbif != 0); + return -EINVAL; + } + + timeout = ktime_add_us(ktime_get(), vbif->cap->xin_halt_timeout); + for (;;) { + status = vbif->ops.get_halt_ctrl(vbif, xin_id); + if (status) + break; + if (ktime_compare_safe(ktime_get(), timeout) > 0) { + status = vbif->ops.get_halt_ctrl(vbif, xin_id); + break; + } + usleep_range(501, 1000); + } + + if (!status) { + rc = -ETIMEDOUT; + DPU_ERROR("VBIF %d client %d not halting. TIMEDOUT.\n", + vbif->idx - VBIF_0, xin_id); + } else { + rc = 0; + DPU_DEBUG("VBIF %d client %d is halted\n", + vbif->idx - VBIF_0, xin_id); + } + + return rc; +} + +/** + * _dpu_vbif_apply_dynamic_ot_limit - determine OT based on usecase parameters + * @vbif: Pointer to hardware vbif driver + * @ot_lim: Pointer to OT limit to be modified + * @params: Pointer to usecase parameters + */ +static void _dpu_vbif_apply_dynamic_ot_limit(struct dpu_hw_vbif *vbif, + u32 *ot_lim, struct dpu_vbif_set_ot_params *params) +{ + u64 pps; + const struct dpu_vbif_dynamic_ot_tbl *tbl; + u32 i; + + if (!vbif || !(vbif->cap->features & BIT(DPU_VBIF_QOS_OTLIM))) + return; + + /* Dynamic OT setting done only for WFD */ + if (!params->is_wfd) + return; + + pps = params->frame_rate; + pps *= params->width; + pps *= params->height; + + tbl = params->rd ? &vbif->cap->dynamic_ot_rd_tbl : + &vbif->cap->dynamic_ot_wr_tbl; + + for (i = 0; i < tbl->count; i++) { + if (pps <= tbl->cfg[i].pps) { + *ot_lim = tbl->cfg[i].ot_limit; + break; + } + } + + DPU_DEBUG("vbif:%d xin:%d w:%d h:%d fps:%d pps:%llu ot:%u\n", + vbif->idx - VBIF_0, params->xin_id, + params->width, params->height, params->frame_rate, + pps, *ot_lim); +} + +/** + * _dpu_vbif_get_ot_limit - get OT based on usecase & configuration parameters + * @vbif: Pointer to hardware vbif driver + * @params: Pointer to usecase parameters + * @return: OT limit + */ +static u32 _dpu_vbif_get_ot_limit(struct dpu_hw_vbif *vbif, + struct dpu_vbif_set_ot_params *params) +{ + u32 ot_lim = 0; + u32 val; + + if (!vbif || !vbif->cap) { + DPU_ERROR("invalid arguments vbif %d\n", vbif != 0); + return -EINVAL; + } + + if (vbif->cap->default_ot_wr_limit && !params->rd) + ot_lim = vbif->cap->default_ot_wr_limit; + else if (vbif->cap->default_ot_rd_limit && params->rd) + ot_lim = vbif->cap->default_ot_rd_limit; + + /* + * If default ot is not set from dt/catalog, + * then do not configure it. + */ + if (ot_lim == 0) + goto exit; + + /* Modify the limits if the target and the use case requires it */ + _dpu_vbif_apply_dynamic_ot_limit(vbif, &ot_lim, params); + + if (vbif && vbif->ops.get_limit_conf) { + val = vbif->ops.get_limit_conf(vbif, + params->xin_id, params->rd); + if (val == ot_lim) + ot_lim = 0; + } + +exit: + DPU_DEBUG("vbif:%d xin:%d ot_lim:%d\n", + vbif->idx - VBIF_0, params->xin_id, ot_lim); + return ot_lim; +} + +/** + * dpu_vbif_set_ot_limit - set OT based on usecase & configuration parameters + * @vbif: Pointer to hardware vbif driver + * @params: Pointer to usecase parameters + * + * Note this function would block waiting for bus halt. + */ +void dpu_vbif_set_ot_limit(struct dpu_kms *dpu_kms, + struct dpu_vbif_set_ot_params *params) +{ + struct dpu_hw_vbif *vbif = NULL; + struct dpu_hw_mdp *mdp; + bool forced_on = false; + u32 ot_lim; + int ret, i; + + if (!dpu_kms) { + DPU_ERROR("invalid arguments\n"); + return; + } + mdp = dpu_kms->hw_mdp; + + for (i = 0; i < ARRAY_SIZE(dpu_kms->hw_vbif); i++) { + if (dpu_kms->hw_vbif[i] && + dpu_kms->hw_vbif[i]->idx == params->vbif_idx) + vbif = dpu_kms->hw_vbif[i]; + } + + if (!vbif || !mdp) { + DPU_DEBUG("invalid arguments vbif %d mdp %d\n", + vbif != 0, mdp != 0); + return; + } + + if (!mdp->ops.setup_clk_force_ctrl || + !vbif->ops.set_limit_conf || + !vbif->ops.set_halt_ctrl) + return; + + /* set write_gather_en for all write clients */ + if (vbif->ops.set_write_gather_en && !params->rd) + vbif->ops.set_write_gather_en(vbif, params->xin_id); + + ot_lim = _dpu_vbif_get_ot_limit(vbif, params) & 0xFF; + + if (ot_lim == 0) + goto exit; + + trace_dpu_perf_set_ot(params->num, params->xin_id, ot_lim, + params->vbif_idx); + + forced_on = mdp->ops.setup_clk_force_ctrl(mdp, params->clk_ctrl, true); + + vbif->ops.set_limit_conf(vbif, params->xin_id, params->rd, ot_lim); + + vbif->ops.set_halt_ctrl(vbif, params->xin_id, true); + + ret = _dpu_vbif_wait_for_xin_halt(vbif, params->xin_id); + if (ret) + trace_dpu_vbif_wait_xin_halt_fail(vbif->idx, params->xin_id); + + vbif->ops.set_halt_ctrl(vbif, params->xin_id, false); + + if (forced_on) + mdp->ops.setup_clk_force_ctrl(mdp, params->clk_ctrl, false); +exit: + return; +} + +void dpu_vbif_set_qos_remap(struct dpu_kms *dpu_kms, + struct dpu_vbif_set_qos_params *params) +{ + struct dpu_hw_vbif *vbif = NULL; + struct dpu_hw_mdp *mdp; + bool forced_on = false; + const struct dpu_vbif_qos_tbl *qos_tbl; + int i; + + if (!dpu_kms || !params || !dpu_kms->hw_mdp) { + DPU_ERROR("invalid arguments\n"); + return; + } + mdp = dpu_kms->hw_mdp; + + for (i = 0; i < ARRAY_SIZE(dpu_kms->hw_vbif); i++) { + if (dpu_kms->hw_vbif[i] && + dpu_kms->hw_vbif[i]->idx == params->vbif_idx) { + vbif = dpu_kms->hw_vbif[i]; + break; + } + } + + if (!vbif || !vbif->cap) { + DPU_ERROR("invalid vbif %d\n", params->vbif_idx); + return; + } + + if (!vbif->ops.set_qos_remap || !mdp->ops.setup_clk_force_ctrl) { + DPU_DEBUG("qos remap not supported\n"); + return; + } + + qos_tbl = params->is_rt ? &vbif->cap->qos_rt_tbl : + &vbif->cap->qos_nrt_tbl; + + if (!qos_tbl->npriority_lvl || !qos_tbl->priority_lvl) { + DPU_DEBUG("qos tbl not defined\n"); + return; + } + + forced_on = mdp->ops.setup_clk_force_ctrl(mdp, params->clk_ctrl, true); + + for (i = 0; i < qos_tbl->npriority_lvl; i++) { + DPU_DEBUG("vbif:%d xin:%d lvl:%d/%d\n", + params->vbif_idx, params->xin_id, i, + qos_tbl->priority_lvl[i]); + vbif->ops.set_qos_remap(vbif, params->xin_id, i, + qos_tbl->priority_lvl[i]); + } + + if (forced_on) + mdp->ops.setup_clk_force_ctrl(mdp, params->clk_ctrl, false); +} + +void dpu_vbif_clear_errors(struct dpu_kms *dpu_kms) +{ + struct dpu_hw_vbif *vbif; + u32 i, pnd, src; + + if (!dpu_kms) { + DPU_ERROR("invalid argument\n"); + return; + } + + for (i = 0; i < ARRAY_SIZE(dpu_kms->hw_vbif); i++) { + vbif = dpu_kms->hw_vbif[i]; + if (vbif && vbif->ops.clear_errors) { + vbif->ops.clear_errors(vbif, &pnd, &src); + if (pnd || src) { + DRM_DEBUG_KMS("VBIF %d: pnd 0x%X, src 0x%X\n", + vbif->idx - VBIF_0, pnd, src); + } + } + } +} + +void dpu_vbif_init_memtypes(struct dpu_kms *dpu_kms) +{ + struct dpu_hw_vbif *vbif; + int i, j; + + if (!dpu_kms) { + DPU_ERROR("invalid argument\n"); + return; + } + + for (i = 0; i < ARRAY_SIZE(dpu_kms->hw_vbif); i++) { + vbif = dpu_kms->hw_vbif[i]; + if (vbif && vbif->cap && vbif->ops.set_mem_type) { + for (j = 0; j < vbif->cap->memtype_count; j++) + vbif->ops.set_mem_type( + vbif, j, vbif->cap->memtype[j]); + } + } +} + +#ifdef CONFIG_DEBUG_FS +void dpu_debugfs_vbif_destroy(struct dpu_kms *dpu_kms) +{ + debugfs_remove_recursive(dpu_kms->debugfs_vbif); + dpu_kms->debugfs_vbif = NULL; +} + +int dpu_debugfs_vbif_init(struct dpu_kms *dpu_kms, struct dentry *debugfs_root) +{ + char vbif_name[32]; + struct dentry *debugfs_vbif; + int i, j; + + dpu_kms->debugfs_vbif = debugfs_create_dir("vbif", debugfs_root); + if (!dpu_kms->debugfs_vbif) { + DPU_ERROR("failed to create vbif debugfs\n"); + return -EINVAL; + } + + for (i = 0; i < dpu_kms->catalog->vbif_count; i++) { + struct dpu_vbif_cfg *vbif = &dpu_kms->catalog->vbif[i]; + + snprintf(vbif_name, sizeof(vbif_name), "%d", vbif->id); + + debugfs_vbif = debugfs_create_dir(vbif_name, + dpu_kms->debugfs_vbif); + + debugfs_create_u32("features", 0600, debugfs_vbif, + (u32 *)&vbif->features); + + debugfs_create_u32("xin_halt_timeout", 0400, debugfs_vbif, + (u32 *)&vbif->xin_halt_timeout); + + debugfs_create_u32("default_rd_ot_limit", 0400, debugfs_vbif, + (u32 *)&vbif->default_ot_rd_limit); + + debugfs_create_u32("default_wr_ot_limit", 0400, debugfs_vbif, + (u32 *)&vbif->default_ot_wr_limit); + + for (j = 0; j < vbif->dynamic_ot_rd_tbl.count; j++) { + struct dpu_vbif_dynamic_ot_cfg *cfg = + &vbif->dynamic_ot_rd_tbl.cfg[j]; + + snprintf(vbif_name, sizeof(vbif_name), + "dynamic_ot_rd_%d_pps", j); + debugfs_create_u64(vbif_name, 0400, debugfs_vbif, + (u64 *)&cfg->pps); + snprintf(vbif_name, sizeof(vbif_name), + "dynamic_ot_rd_%d_ot_limit", j); + debugfs_create_u32(vbif_name, 0400, debugfs_vbif, + (u32 *)&cfg->ot_limit); + } + + for (j = 0; j < vbif->dynamic_ot_wr_tbl.count; j++) { + struct dpu_vbif_dynamic_ot_cfg *cfg = + &vbif->dynamic_ot_wr_tbl.cfg[j]; + + snprintf(vbif_name, sizeof(vbif_name), + "dynamic_ot_wr_%d_pps", j); + debugfs_create_u64(vbif_name, 0400, debugfs_vbif, + (u64 *)&cfg->pps); + snprintf(vbif_name, sizeof(vbif_name), + "dynamic_ot_wr_%d_ot_limit", j); + debugfs_create_u32(vbif_name, 0400, debugfs_vbif, + (u32 *)&cfg->ot_limit); + } + } + + return 0; +} +#endif diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_vbif.h b/drivers/gpu/drm/msm/disp/dpu1/dpu_vbif.h new file mode 100644 index 000000000000..f17af52dbbd5 --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_vbif.h @@ -0,0 +1,94 @@ +/* Copyright (c) 2016-2018, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#ifndef __DPU_VBIF_H__ +#define __DPU_VBIF_H__ + +#include "dpu_kms.h" + +struct dpu_vbif_set_ot_params { + u32 xin_id; + u32 num; + u32 width; + u32 height; + u32 frame_rate; + bool rd; + bool is_wfd; + u32 vbif_idx; + u32 clk_ctrl; +}; + +struct dpu_vbif_set_memtype_params { + u32 xin_id; + u32 vbif_idx; + u32 clk_ctrl; + bool is_cacheable; +}; + +/** + * struct dpu_vbif_set_qos_params - QoS remapper parameter + * @vbif_idx: vbif identifier + * @xin_id: client interface identifier + * @clk_ctrl: clock control identifier of the xin + * @num: pipe identifier (debug only) + * @is_rt: true if pipe is used in real-time use case + */ +struct dpu_vbif_set_qos_params { + u32 vbif_idx; + u32 xin_id; + u32 clk_ctrl; + u32 num; + bool is_rt; +}; + +/** + * dpu_vbif_set_ot_limit - set OT limit for vbif client + * @dpu_kms: DPU handler + * @params: Pointer to OT configuration parameters + */ +void dpu_vbif_set_ot_limit(struct dpu_kms *dpu_kms, + struct dpu_vbif_set_ot_params *params); + +/** + * dpu_vbif_set_qos_remap - set QoS priority level remap + * @dpu_kms: DPU handler + * @params: Pointer to QoS configuration parameters + */ +void dpu_vbif_set_qos_remap(struct dpu_kms *dpu_kms, + struct dpu_vbif_set_qos_params *params); + +/** + * dpu_vbif_clear_errors - clear any vbif errors + * @dpu_kms: DPU handler + */ +void dpu_vbif_clear_errors(struct dpu_kms *dpu_kms); + +/** + * dpu_vbif_init_memtypes - initialize xin memory types for vbif + * @dpu_kms: DPU handler + */ +void dpu_vbif_init_memtypes(struct dpu_kms *dpu_kms); + +#ifdef CONFIG_DEBUG_FS +int dpu_debugfs_vbif_init(struct dpu_kms *dpu_kms, struct dentry *debugfs_root); +void dpu_debugfs_vbif_destroy(struct dpu_kms *dpu_kms); +#else +static inline int dpu_debugfs_vbif_init(struct dpu_kms *dpu_kms, + struct dentry *debugfs_root) +{ + return 0; +} +static inline void dpu_debugfs_vbif_destroy(struct dpu_kms *dpu_kms) +{ +} +#endif +#endif /* __DPU_VBIF_H__ */ diff --git a/drivers/gpu/drm/msm/disp/dpu1/msm_media_info.h b/drivers/gpu/drm/msm/disp/dpu1/msm_media_info.h new file mode 100644 index 000000000000..4f12e5c534c8 --- /dev/null +++ b/drivers/gpu/drm/msm/disp/dpu1/msm_media_info.h @@ -0,0 +1,1376 @@ +#ifndef __MEDIA_INFO_H__ +#define __MEDIA_INFO_H__ + +#ifndef MSM_MEDIA_ALIGN +#define MSM_MEDIA_ALIGN(__sz, __align) (((__align) & ((__align) - 1)) ?\ + ((((__sz) + (__align) - 1) / (__align)) * (__align)) :\ + (((__sz) + (__align) - 1) & (~((__align) - 1)))) +#endif + +#ifndef MSM_MEDIA_ROUNDUP +#define MSM_MEDIA_ROUNDUP(__sz, __r) (((__sz) + ((__r) - 1)) / (__r)) +#endif + +#ifndef MSM_MEDIA_MAX +#define MSM_MEDIA_MAX(__a, __b) ((__a) > (__b)?(__a):(__b)) +#endif + +enum color_fmts { + /* Venus NV12: + * YUV 4:2:0 image with a plane of 8 bit Y samples followed + * by an interleaved U/V plane containing 8 bit 2x2 subsampled + * colour difference samples. + * + * <-------- Y/UV_Stride --------> + * <------- Width -------> + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . ^ ^ + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . | | + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . Height | + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . | Y_Scanlines + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . | | + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . | | + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . | | + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . V | + * . . . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . . . V + * U V U V U V U V U V U V . . . . ^ + * U V U V U V U V U V U V . . . . | + * U V U V U V U V U V U V . . . . | + * U V U V U V U V U V U V . . . . UV_Scanlines + * . . . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . . . V + * . . . . . . . . . . . . . . . . --> Buffer size alignment + * + * Y_Stride : Width aligned to 128 + * UV_Stride : Width aligned to 128 + * Y_Scanlines: Height aligned to 32 + * UV_Scanlines: Height/2 aligned to 16 + * Extradata: Arbitrary (software-imposed) padding + * Total size = align((Y_Stride * Y_Scanlines + * + UV_Stride * UV_Scanlines + * + max(Extradata, Y_Stride * 8), 4096) + */ + COLOR_FMT_NV12, + + /* Venus NV21: + * YUV 4:2:0 image with a plane of 8 bit Y samples followed + * by an interleaved V/U plane containing 8 bit 2x2 subsampled + * colour difference samples. + * + * <-------- Y/UV_Stride --------> + * <------- Width -------> + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . ^ ^ + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . | | + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . Height | + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . | Y_Scanlines + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . | | + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . | | + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . | | + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . V | + * . . . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . . . V + * V U V U V U V U V U V U . . . . ^ + * V U V U V U V U V U V U . . . . | + * V U V U V U V U V U V U . . . . | + * V U V U V U V U V U V U . . . . UV_Scanlines + * . . . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . . . V + * . . . . . . . . . . . . . . . . --> Padding & Buffer size alignment + * + * Y_Stride : Width aligned to 128 + * UV_Stride : Width aligned to 128 + * Y_Scanlines: Height aligned to 32 + * UV_Scanlines: Height/2 aligned to 16 + * Extradata: Arbitrary (software-imposed) padding + * Total size = align((Y_Stride * Y_Scanlines + * + UV_Stride * UV_Scanlines + * + max(Extradata, Y_Stride * 8), 4096) + */ + COLOR_FMT_NV21, + /* Venus NV12_MVTB: + * Two YUV 4:2:0 images/views one after the other + * in a top-bottom layout, same as NV12 + * with a plane of 8 bit Y samples followed + * by an interleaved U/V plane containing 8 bit 2x2 subsampled + * colour difference samples. + * + * + * <-------- Y/UV_Stride --------> + * <------- Width -------> + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . ^ ^ ^ + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . | | | + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . Height | | + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . | Y_Scanlines | + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . | | | + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . | | | + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . | | | + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . V | | + * . . . . . . . . . . . . . . . . | View_1 + * . . . . . . . . . . . . . . . . | | + * . . . . . . . . . . . . . . . . | | + * . . . . . . . . . . . . . . . . V | + * U V U V U V U V U V U V . . . . ^ | + * U V U V U V U V U V U V . . . . | | + * U V U V U V U V U V U V . . . . | | + * U V U V U V U V U V U V . . . . UV_Scanlines | + * . . . . . . . . . . . . . . . . | | + * . . . . . . . . . . . . . . . . V V + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . ^ ^ ^ + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . | | | + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . Height | | + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . | Y_Scanlines | + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . | | | + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . | | | + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . | | | + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . V | | + * . . . . . . . . . . . . . . . . | View_2 + * . . . . . . . . . . . . . . . . | | + * . . . . . . . . . . . . . . . . | | + * . . . . . . . . . . . . . . . . V | + * U V U V U V U V U V U V . . . . ^ | + * U V U V U V U V U V U V . . . . | | + * U V U V U V U V U V U V . . . . | | + * U V U V U V U V U V U V . . . . UV_Scanlines | + * . . . . . . . . . . . . . . . . | | + * . . . . . . . . . . . . . . . . V V + * . . . . . . . . . . . . . . . . --> Buffer size alignment + * + * Y_Stride : Width aligned to 128 + * UV_Stride : Width aligned to 128 + * Y_Scanlines: Height aligned to 32 + * UV_Scanlines: Height/2 aligned to 16 + * View_1 begin at: 0 (zero) + * View_2 begin at: Y_Stride * Y_Scanlines + UV_Stride * UV_Scanlines + * Extradata: Arbitrary (software-imposed) padding + * Total size = align((2*(Y_Stride * Y_Scanlines) + * + 2*(UV_Stride * UV_Scanlines) + Extradata), 4096) + */ + COLOR_FMT_NV12_MVTB, + /* + * The buffer can be of 2 types: + * (1) Venus NV12 UBWC Progressive + * (2) Venus NV12 UBWC Interlaced + * + * (1) Venus NV12 UBWC Progressive Buffer Format: + * Compressed Macro-tile format for NV12. + * Contains 4 planes in the following order - + * (A) Y_Meta_Plane + * (B) Y_UBWC_Plane + * (C) UV_Meta_Plane + * (D) UV_UBWC_Plane + * + * Y_Meta_Plane consists of meta information to decode compressed + * tile data in Y_UBWC_Plane. + * Y_UBWC_Plane consists of Y data in compressed macro-tile format. + * UBWC decoder block will use the Y_Meta_Plane data together with + * Y_UBWC_Plane data to produce loss-less uncompressed 8 bit Y samples. + * + * UV_Meta_Plane consists of meta information to decode compressed + * tile data in UV_UBWC_Plane. + * UV_UBWC_Plane consists of UV data in compressed macro-tile format. + * UBWC decoder block will use UV_Meta_Plane data together with + * UV_UBWC_Plane data to produce loss-less uncompressed 8 bit 2x2 + * subsampled color difference samples. + * + * Each tile in Y_UBWC_Plane/UV_UBWC_Plane is independently decodable + * and randomly accessible. There is no dependency between tiles. + * + * <----- Y_Meta_Stride ----> + * <-------- Width ------> + * M M M M M M M M M M M M . . ^ ^ + * M M M M M M M M M M M M . . | | + * M M M M M M M M M M M M . . Height | + * M M M M M M M M M M M M . . | Meta_Y_Scanlines + * M M M M M M M M M M M M . . | | + * M M M M M M M M M M M M . . | | + * M M M M M M M M M M M M . . | | + * M M M M M M M M M M M M . . V | + * . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . -------> Buffer size aligned to 4k + * . . . . . . . . . . . . . . V + * <--Compressed tile Y Stride---> + * <------- Width -------> + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . ^ ^ + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . | | + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . Height | + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . | Macro_tile_Y_Scanlines + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . | | + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . | | + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . | | + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . V | + * . . . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . . . -------> Buffer size aligned to 4k + * . . . . . . . . . . . . . . . . V + * <----- UV_Meta_Stride ----> + * M M M M M M M M M M M M . . ^ + * M M M M M M M M M M M M . . | + * M M M M M M M M M M M M . . | + * M M M M M M M M M M M M . . M_UV_Scanlines + * . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . V + * . . . . . . . . . . . . . . -------> Buffer size aligned to 4k + * <--Compressed tile UV Stride---> + * U* V* U* V* U* V* U* V* . . . . ^ + * U* V* U* V* U* V* U* V* . . . . | + * U* V* U* V* U* V* U* V* . . . . | + * U* V* U* V* U* V* U* V* . . . . UV_Scanlines + * . . . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . . . V + * . . . . . . . . . . . . . . . . -------> Buffer size aligned to 4k + * + * Y_Stride = align(Width, 128) + * UV_Stride = align(Width, 128) + * Y_Scanlines = align(Height, 32) + * UV_Scanlines = align(Height/2, 16) + * Y_UBWC_Plane_size = align(Y_Stride * Y_Scanlines, 4096) + * UV_UBWC_Plane_size = align(UV_Stride * UV_Scanlines, 4096) + * Y_Meta_Stride = align(roundup(Width, Y_TileWidth), 64) + * Y_Meta_Scanlines = align(roundup(Height, Y_TileHeight), 16) + * Y_Meta_Plane_size = align(Y_Meta_Stride * Y_Meta_Scanlines, 4096) + * UV_Meta_Stride = align(roundup(Width, UV_TileWidth), 64) + * UV_Meta_Scanlines = align(roundup(Height, UV_TileHeight), 16) + * UV_Meta_Plane_size = align(UV_Meta_Stride * UV_Meta_Scanlines, 4096) + * Extradata = 8k + * + * Total size = align( Y_UBWC_Plane_size + UV_UBWC_Plane_size + + * Y_Meta_Plane_size + UV_Meta_Plane_size + * + max(Extradata, Y_Stride * 48), 4096) + * + * + * (2) Venus NV12 UBWC Interlaced Buffer Format: + * Compressed Macro-tile format for NV12 interlaced. + * Contains 8 planes in the following order - + * (A) Y_Meta_Top_Field_Plane + * (B) Y_UBWC_Top_Field_Plane + * (C) UV_Meta_Top_Field_Plane + * (D) UV_UBWC_Top_Field_Plane + * (E) Y_Meta_Bottom_Field_Plane + * (F) Y_UBWC_Bottom_Field_Plane + * (G) UV_Meta_Bottom_Field_Plane + * (H) UV_UBWC_Bottom_Field_Plane + * Y_Meta_Top_Field_Plane consists of meta information to decode + * compressed tile data for Y_UBWC_Top_Field_Plane. + * Y_UBWC_Top_Field_Plane consists of Y data in compressed macro-tile + * format for top field of an interlaced frame. + * UBWC decoder block will use the Y_Meta_Top_Field_Plane data together + * with Y_UBWC_Top_Field_Plane data to produce loss-less uncompressed + * 8 bit Y samples for top field of an interlaced frame. + * + * UV_Meta_Top_Field_Plane consists of meta information to decode + * compressed tile data in UV_UBWC_Top_Field_Plane. + * UV_UBWC_Top_Field_Plane consists of UV data in compressed macro-tile + * format for top field of an interlaced frame. + * UBWC decoder block will use UV_Meta_Top_Field_Plane data together + * with UV_UBWC_Top_Field_Plane data to produce loss-less uncompressed + * 8 bit subsampled color difference samples for top field of an + * interlaced frame. + * + * Each tile in Y_UBWC_Top_Field_Plane/UV_UBWC_Top_Field_Plane is + * independently decodable and randomly accessible. There is no + * dependency between tiles. + * + * Y_Meta_Bottom_Field_Plane consists of meta information to decode + * compressed tile data for Y_UBWC_Bottom_Field_Plane. + * Y_UBWC_Bottom_Field_Plane consists of Y data in compressed macro-tile + * format for bottom field of an interlaced frame. + * UBWC decoder block will use the Y_Meta_Bottom_Field_Plane data + * together with Y_UBWC_Bottom_Field_Plane data to produce loss-less + * uncompressed 8 bit Y samples for bottom field of an interlaced frame. + * + * UV_Meta_Bottom_Field_Plane consists of meta information to decode + * compressed tile data in UV_UBWC_Bottom_Field_Plane. + * UV_UBWC_Bottom_Field_Plane consists of UV data in compressed + * macro-tile format for bottom field of an interlaced frame. + * UBWC decoder block will use UV_Meta_Bottom_Field_Plane data together + * with UV_UBWC_Bottom_Field_Plane data to produce loss-less + * uncompressed 8 bit subsampled color difference samples for bottom + * field of an interlaced frame. + * + * Each tile in Y_UBWC_Bottom_Field_Plane/UV_UBWC_Bottom_Field_Plane is + * independently decodable and randomly accessible. There is no + * dependency between tiles. + * + * <-----Y_TF_Meta_Stride----> + * <-------- Width ------> + * M M M M M M M M M M M M . . ^ ^ + * M M M M M M M M M M M M . . | | + * M M M M M M M M M M M M . . Half_height | + * M M M M M M M M M M M M . . | Meta_Y_TF_Scanlines + * M M M M M M M M M M M M . . | | + * M M M M M M M M M M M M . . | | + * M M M M M M M M M M M M . . | | + * M M M M M M M M M M M M . . V | + * . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . -------> Buffer size aligned to 4k + * . . . . . . . . . . . . . . V + * <-Compressed tile Y_TF Stride-> + * <------- Width -------> + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . ^ ^ + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . | | + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . Half_height | + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . | Macro_tile_Y_TF_Scanlines + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . | | + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . | | + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . | | + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . V | + * . . . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . . . -------> Buffer size aligned to 4k + * . . . . . . . . . . . . . . . . V + * <----UV_TF_Meta_Stride----> + * M M M M M M M M M M M M . . ^ + * M M M M M M M M M M M M . . | + * M M M M M M M M M M M M . . | + * M M M M M M M M M M M M . . M_UV_TF_Scanlines + * . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . V + * . . . . . . . . . . . . . . -------> Buffer size aligned to 4k + * <-Compressed tile UV_TF Stride-> + * U* V* U* V* U* V* U* V* . . . . ^ + * U* V* U* V* U* V* U* V* . . . . | + * U* V* U* V* U* V* U* V* . . . . | + * U* V* U* V* U* V* U* V* . . . . UV_TF_Scanlines + * . . . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . . . V + * . . . . . . . . . . . . . . . . -------> Buffer size aligned to 4k + * <-----Y_BF_Meta_Stride----> + * <-------- Width ------> + * M M M M M M M M M M M M . . ^ ^ + * M M M M M M M M M M M M . . | | + * M M M M M M M M M M M M . . Half_height | + * M M M M M M M M M M M M . . | Meta_Y_BF_Scanlines + * M M M M M M M M M M M M . . | | + * M M M M M M M M M M M M . . | | + * M M M M M M M M M M M M . . | | + * M M M M M M M M M M M M . . V | + * . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . -------> Buffer size aligned to 4k + * . . . . . . . . . . . . . . V + * <-Compressed tile Y_BF Stride-> + * <------- Width -------> + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . ^ ^ + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . | | + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . Half_height | + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . | Macro_tile_Y_BF_Scanlines + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . | | + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . | | + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . | | + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . V | + * . . . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . . . -------> Buffer size aligned to 4k + * . . . . . . . . . . . . . . . . V + * <----UV_BF_Meta_Stride----> + * M M M M M M M M M M M M . . ^ + * M M M M M M M M M M M M . . | + * M M M M M M M M M M M M . . | + * M M M M M M M M M M M M . . M_UV_BF_Scanlines + * . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . V + * . . . . . . . . . . . . . . -------> Buffer size aligned to 4k + * <-Compressed tile UV_BF Stride-> + * U* V* U* V* U* V* U* V* . . . . ^ + * U* V* U* V* U* V* U* V* . . . . | + * U* V* U* V* U* V* U* V* . . . . | + * U* V* U* V* U* V* U* V* . . . . UV_BF_Scanlines + * . . . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . . . V + * . . . . . . . . . . . . . . . . -------> Buffer size aligned to 4k + * + * Half_height = (Height+1)>>1 + * Y_TF_Stride = align(Width, 128) + * UV_TF_Stride = align(Width, 128) + * Y_TF_Scanlines = align(Half_height, 32) + * UV_TF_Scanlines = align((Half_height+1)/2, 32) + * Y_UBWC_TF_Plane_size = align(Y_TF_Stride * Y_TF_Scanlines, 4096) + * UV_UBWC_TF_Plane_size = align(UV_TF_Stride * UV_TF_Scanlines, 4096) + * Y_TF_Meta_Stride = align(roundup(Width, Y_TileWidth), 64) + * Y_TF_Meta_Scanlines = align(roundup(Half_height, Y_TileHeight), 16) + * Y_TF_Meta_Plane_size = + * align(Y_TF_Meta_Stride * Y_TF_Meta_Scanlines, 4096) + * UV_TF_Meta_Stride = align(roundup(Width, UV_TileWidth), 64) + * UV_TF_Meta_Scanlines = align(roundup(Half_height, UV_TileHeight), 16) + * UV_TF_Meta_Plane_size = + * align(UV_TF_Meta_Stride * UV_TF_Meta_Scanlines, 4096) + * Y_BF_Stride = align(Width, 128) + * UV_BF_Stride = align(Width, 128) + * Y_BF_Scanlines = align(Half_height, 32) + * UV_BF_Scanlines = align((Half_height+1)/2, 32) + * Y_UBWC_BF_Plane_size = align(Y_BF_Stride * Y_BF_Scanlines, 4096) + * UV_UBWC_BF_Plane_size = align(UV_BF_Stride * UV_BF_Scanlines, 4096) + * Y_BF_Meta_Stride = align(roundup(Width, Y_TileWidth), 64) + * Y_BF_Meta_Scanlines = align(roundup(Half_height, Y_TileHeight), 16) + * Y_BF_Meta_Plane_size = + * align(Y_BF_Meta_Stride * Y_BF_Meta_Scanlines, 4096) + * UV_BF_Meta_Stride = align(roundup(Width, UV_TileWidth), 64) + * UV_BF_Meta_Scanlines = align(roundup(Half_height, UV_TileHeight), 16) + * UV_BF_Meta_Plane_size = + * align(UV_BF_Meta_Stride * UV_BF_Meta_Scanlines, 4096) + * Extradata = 8k + * + * Total size = align( Y_UBWC_TF_Plane_size + UV_UBWC_TF_Plane_size + + * Y_TF_Meta_Plane_size + UV_TF_Meta_Plane_size + + * Y_UBWC_BF_Plane_size + UV_UBWC_BF_Plane_size + + * Y_BF_Meta_Plane_size + UV_BF_Meta_Plane_size + + * + max(Extradata, Y_TF_Stride * 48), 4096) + */ + COLOR_FMT_NV12_UBWC, + /* Venus NV12 10-bit UBWC: + * Compressed Macro-tile format for NV12. + * Contains 4 planes in the following order - + * (A) Y_Meta_Plane + * (B) Y_UBWC_Plane + * (C) UV_Meta_Plane + * (D) UV_UBWC_Plane + * + * Y_Meta_Plane consists of meta information to decode compressed + * tile data in Y_UBWC_Plane. + * Y_UBWC_Plane consists of Y data in compressed macro-tile format. + * UBWC decoder block will use the Y_Meta_Plane data together with + * Y_UBWC_Plane data to produce loss-less uncompressed 10 bit Y samples. + * + * UV_Meta_Plane consists of meta information to decode compressed + * tile data in UV_UBWC_Plane. + * UV_UBWC_Plane consists of UV data in compressed macro-tile format. + * UBWC decoder block will use UV_Meta_Plane data together with + * UV_UBWC_Plane data to produce loss-less uncompressed 10 bit 2x2 + * subsampled color difference samples. + * + * Each tile in Y_UBWC_Plane/UV_UBWC_Plane is independently decodable + * and randomly accessible. There is no dependency between tiles. + * + * <----- Y_Meta_Stride -----> + * <-------- Width ------> + * M M M M M M M M M M M M . . ^ ^ + * M M M M M M M M M M M M . . | | + * M M M M M M M M M M M M . . Height | + * M M M M M M M M M M M M . . | Meta_Y_Scanlines + * M M M M M M M M M M M M . . | | + * M M M M M M M M M M M M . . | | + * M M M M M M M M M M M M . . | | + * M M M M M M M M M M M M . . V | + * . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . -------> Buffer size aligned to 4k + * . . . . . . . . . . . . . . V + * <--Compressed tile Y Stride---> + * <------- Width -------> + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . ^ ^ + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . | | + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . Height | + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . | Macro_tile_Y_Scanlines + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . | | + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . | | + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . | | + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . V | + * . . . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . . . -------> Buffer size aligned to 4k + * . . . . . . . . . . . . . . . . V + * <----- UV_Meta_Stride ----> + * M M M M M M M M M M M M . . ^ + * M M M M M M M M M M M M . . | + * M M M M M M M M M M M M . . | + * M M M M M M M M M M M M . . M_UV_Scanlines + * . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . V + * . . . . . . . . . . . . . . -------> Buffer size aligned to 4k + * <--Compressed tile UV Stride---> + * U* V* U* V* U* V* U* V* . . . . ^ + * U* V* U* V* U* V* U* V* . . . . | + * U* V* U* V* U* V* U* V* . . . . | + * U* V* U* V* U* V* U* V* . . . . UV_Scanlines + * . . . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . . . V + * . . . . . . . . . . . . . . . . -------> Buffer size aligned to 4k + * + * + * Y_Stride = align(Width * 4/3, 128) + * UV_Stride = align(Width * 4/3, 128) + * Y_Scanlines = align(Height, 32) + * UV_Scanlines = align(Height/2, 16) + * Y_UBWC_Plane_Size = align(Y_Stride * Y_Scanlines, 4096) + * UV_UBWC_Plane_Size = align(UV_Stride * UV_Scanlines, 4096) + * Y_Meta_Stride = align(roundup(Width, Y_TileWidth), 64) + * Y_Meta_Scanlines = align(roundup(Height, Y_TileHeight), 16) + * Y_Meta_Plane_size = align(Y_Meta_Stride * Y_Meta_Scanlines, 4096) + * UV_Meta_Stride = align(roundup(Width, UV_TileWidth), 64) + * UV_Meta_Scanlines = align(roundup(Height, UV_TileHeight), 16) + * UV_Meta_Plane_size = align(UV_Meta_Stride * UV_Meta_Scanlines, 4096) + * Extradata = 8k + * + * Total size = align(Y_UBWC_Plane_size + UV_UBWC_Plane_size + + * Y_Meta_Plane_size + UV_Meta_Plane_size + * + max(Extradata, Y_Stride * 48), 4096) + */ + COLOR_FMT_NV12_BPP10_UBWC, + /* Venus RGBA8888 format: + * Contains 1 plane in the following order - + * (A) RGBA plane + * + * <-------- RGB_Stride --------> + * <------- Width -------> + * R R R R R R R R R R R R . . . . ^ ^ + * R R R R R R R R R R R R . . . . | | + * R R R R R R R R R R R R . . . . Height | + * R R R R R R R R R R R R . . . . | RGB_Scanlines + * R R R R R R R R R R R R . . . . | | + * R R R R R R R R R R R R . . . . | | + * R R R R R R R R R R R R . . . . | | + * R R R R R R R R R R R R . . . . V | + * . . . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . . . V + * + * RGB_Stride = align(Width * 4, 128) + * RGB_Scanlines = align(Height, 32) + * RGB_Plane_size = align(RGB_Stride * RGB_Scanlines, 4096) + * Extradata = 8k + * + * Total size = align(RGB_Plane_size + Extradata, 4096) + */ + COLOR_FMT_RGBA8888, + /* Venus RGBA8888 UBWC format: + * Contains 2 planes in the following order - + * (A) Meta plane + * (B) RGBA plane + * + * <--- RGB_Meta_Stride ----> + * <-------- Width ------> + * M M M M M M M M M M M M . . ^ ^ + * M M M M M M M M M M M M . . | | + * M M M M M M M M M M M M . . Height | + * M M M M M M M M M M M M . . | Meta_RGB_Scanlines + * M M M M M M M M M M M M . . | | + * M M M M M M M M M M M M . . | | + * M M M M M M M M M M M M . . | | + * M M M M M M M M M M M M . . V | + * . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . -------> Buffer size aligned to 4k + * . . . . . . . . . . . . . . V + * <-------- RGB_Stride --------> + * <------- Width -------> + * R R R R R R R R R R R R . . . . ^ ^ + * R R R R R R R R R R R R . . . . | | + * R R R R R R R R R R R R . . . . Height | + * R R R R R R R R R R R R . . . . | RGB_Scanlines + * R R R R R R R R R R R R . . . . | | + * R R R R R R R R R R R R . . . . | | + * R R R R R R R R R R R R . . . . | | + * R R R R R R R R R R R R . . . . V | + * . . . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . . . -------> Buffer size aligned to 4k + * . . . . . . . . . . . . . . . . V + * + * RGB_Stride = align(Width * 4, 128) + * RGB_Scanlines = align(Height, 32) + * RGB_Plane_size = align(RGB_Stride * RGB_Scanlines, 4096) + * RGB_Meta_Stride = align(roundup(Width, RGB_TileWidth), 64) + * RGB_Meta_Scanline = align(roundup(Height, RGB_TileHeight), 16) + * RGB_Meta_Plane_size = align(RGB_Meta_Stride * + * RGB_Meta_Scanlines, 4096) + * Extradata = 8k + * + * Total size = align(RGB_Meta_Plane_size + RGB_Plane_size + + * Extradata, 4096) + */ + COLOR_FMT_RGBA8888_UBWC, + /* Venus RGBA1010102 UBWC format: + * Contains 2 planes in the following order - + * (A) Meta plane + * (B) RGBA plane + * + * <--- RGB_Meta_Stride ----> + * <-------- Width ------> + * M M M M M M M M M M M M . . ^ ^ + * M M M M M M M M M M M M . . | | + * M M M M M M M M M M M M . . Height | + * M M M M M M M M M M M M . . | Meta_RGB_Scanlines + * M M M M M M M M M M M M . . | | + * M M M M M M M M M M M M . . | | + * M M M M M M M M M M M M . . | | + * M M M M M M M M M M M M . . V | + * . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . -------> Buffer size aligned to 4k + * . . . . . . . . . . . . . . V + * <-------- RGB_Stride --------> + * <------- Width -------> + * R R R R R R R R R R R R . . . . ^ ^ + * R R R R R R R R R R R R . . . . | | + * R R R R R R R R R R R R . . . . Height | + * R R R R R R R R R R R R . . . . | RGB_Scanlines + * R R R R R R R R R R R R . . . . | | + * R R R R R R R R R R R R . . . . | | + * R R R R R R R R R R R R . . . . | | + * R R R R R R R R R R R R . . . . V | + * . . . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . . . -------> Buffer size aligned to 4k + * . . . . . . . . . . . . . . . . V + * + * RGB_Stride = align(Width * 4, 256) + * RGB_Scanlines = align(Height, 16) + * RGB_Plane_size = align(RGB_Stride * RGB_Scanlines, 4096) + * RGB_Meta_Stride = align(roundup(Width, RGB_TileWidth), 64) + * RGB_Meta_Scanline = align(roundup(Height, RGB_TileHeight), 16) + * RGB_Meta_Plane_size = align(RGB_Meta_Stride * + * RGB_Meta_Scanlines, 4096) + * Extradata = 8k + * + * Total size = align(RGB_Meta_Plane_size + RGB_Plane_size + + * Extradata, 4096) + */ + COLOR_FMT_RGBA1010102_UBWC, + /* Venus RGB565 UBWC format: + * Contains 2 planes in the following order - + * (A) Meta plane + * (B) RGB plane + * + * <--- RGB_Meta_Stride ----> + * <-------- Width ------> + * M M M M M M M M M M M M . . ^ ^ + * M M M M M M M M M M M M . . | | + * M M M M M M M M M M M M . . Height | + * M M M M M M M M M M M M . . | Meta_RGB_Scanlines + * M M M M M M M M M M M M . . | | + * M M M M M M M M M M M M . . | | + * M M M M M M M M M M M M . . | | + * M M M M M M M M M M M M . . V | + * . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . -------> Buffer size aligned to 4k + * . . . . . . . . . . . . . . V + * <-------- RGB_Stride --------> + * <------- Width -------> + * R R R R R R R R R R R R . . . . ^ ^ + * R R R R R R R R R R R R . . . . | | + * R R R R R R R R R R R R . . . . Height | + * R R R R R R R R R R R R . . . . | RGB_Scanlines + * R R R R R R R R R R R R . . . . | | + * R R R R R R R R R R R R . . . . | | + * R R R R R R R R R R R R . . . . | | + * R R R R R R R R R R R R . . . . V | + * . . . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . . . -------> Buffer size aligned to 4k + * . . . . . . . . . . . . . . . . V + * + * RGB_Stride = align(Width * 2, 128) + * RGB_Scanlines = align(Height, 16) + * RGB_Plane_size = align(RGB_Stride * RGB_Scanlines, 4096) + * RGB_Meta_Stride = align(roundup(Width, RGB_TileWidth), 64) + * RGB_Meta_Scanline = align(roundup(Height, RGB_TileHeight), 16) + * RGB_Meta_Plane_size = align(RGB_Meta_Stride * + * RGB_Meta_Scanlines, 4096) + * Extradata = 8k + * + * Total size = align(RGB_Meta_Plane_size + RGB_Plane_size + + * Extradata, 4096) + */ + COLOR_FMT_RGB565_UBWC, + /* P010 UBWC: + * Compressed Macro-tile format for NV12. + * Contains 4 planes in the following order - + * (A) Y_Meta_Plane + * (B) Y_UBWC_Plane + * (C) UV_Meta_Plane + * (D) UV_UBWC_Plane + * + * Y_Meta_Plane consists of meta information to decode compressed + * tile data in Y_UBWC_Plane. + * Y_UBWC_Plane consists of Y data in compressed macro-tile format. + * UBWC decoder block will use the Y_Meta_Plane data together with + * Y_UBWC_Plane data to produce loss-less uncompressed 10 bit Y samples. + * + * UV_Meta_Plane consists of meta information to decode compressed + * tile data in UV_UBWC_Plane. + * UV_UBWC_Plane consists of UV data in compressed macro-tile format. + * UBWC decoder block will use UV_Meta_Plane data together with + * UV_UBWC_Plane data to produce loss-less uncompressed 10 bit 2x2 + * subsampled color difference samples. + * + * Each tile in Y_UBWC_Plane/UV_UBWC_Plane is independently decodable + * and randomly accessible. There is no dependency between tiles. + * + * <----- Y_Meta_Stride -----> + * <-------- Width ------> + * M M M M M M M M M M M M . . ^ ^ + * M M M M M M M M M M M M . . | | + * M M M M M M M M M M M M . . Height | + * M M M M M M M M M M M M . . | Meta_Y_Scanlines + * M M M M M M M M M M M M . . | | + * M M M M M M M M M M M M . . | | + * M M M M M M M M M M M M . . | | + * M M M M M M M M M M M M . . V | + * . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . -------> Buffer size aligned to 4k + * . . . . . . . . . . . . . . V + * <--Compressed tile Y Stride---> + * <------- Width -------> + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . ^ ^ + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . | | + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . Height | + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . | Macro_tile_Y_Scanlines + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . | | + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . | | + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . | | + * Y* Y* Y* Y* Y* Y* Y* Y* . . . . V | + * . . . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . . . -------> Buffer size aligned to 4k + * . . . . . . . . . . . . . . . . V + * <----- UV_Meta_Stride ----> + * M M M M M M M M M M M M . . ^ + * M M M M M M M M M M M M . . | + * M M M M M M M M M M M M . . | + * M M M M M M M M M M M M . . M_UV_Scanlines + * . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . V + * . . . . . . . . . . . . . . -------> Buffer size aligned to 4k + * <--Compressed tile UV Stride---> + * U* V* U* V* U* V* U* V* . . . . ^ + * U* V* U* V* U* V* U* V* . . . . | + * U* V* U* V* U* V* U* V* . . . . | + * U* V* U* V* U* V* U* V* . . . . UV_Scanlines + * . . . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . . . V + * . . . . . . . . . . . . . . . . -------> Buffer size aligned to 4k + * + * + * Y_Stride = align(Width * 2, 256) + * UV_Stride = align(Width * 2, 256) + * Y_Scanlines = align(Height, 16) + * UV_Scanlines = align(Height/2, 16) + * Y_UBWC_Plane_Size = align(Y_Stride * Y_Scanlines, 4096) + * UV_UBWC_Plane_Size = align(UV_Stride * UV_Scanlines, 4096) + * Y_Meta_Stride = align(roundup(Width, Y_TileWidth), 64) + * Y_Meta_Scanlines = align(roundup(Height, Y_TileHeight), 16) + * Y_Meta_Plane_size = align(Y_Meta_Stride * Y_Meta_Scanlines, 4096) + * UV_Meta_Stride = align(roundup(Width, UV_TileWidth), 64) + * UV_Meta_Scanlines = align(roundup(Height, UV_TileHeight), 16) + * UV_Meta_Plane_size = align(UV_Meta_Stride * UV_Meta_Scanlines, 4096) + * Extradata = 8k + * + * Total size = align(Y_UBWC_Plane_size + UV_UBWC_Plane_size + + * Y_Meta_Plane_size + UV_Meta_Plane_size + * + max(Extradata, Y_Stride * 48), 4096) + */ + COLOR_FMT_P010_UBWC, + /* Venus P010: + * YUV 4:2:0 image with a plane of 10 bit Y samples followed + * by an interleaved U/V plane containing 10 bit 2x2 subsampled + * colour difference samples. + * + * <-------- Y/UV_Stride --------> + * <------- Width -------> + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . ^ ^ + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . | | + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . Height | + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . | Y_Scanlines + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . | | + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . | | + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . | | + * Y Y Y Y Y Y Y Y Y Y Y Y . . . . V | + * . . . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . . . V + * U V U V U V U V U V U V . . . . ^ + * U V U V U V U V U V U V . . . . | + * U V U V U V U V U V U V . . . . | + * U V U V U V U V U V U V . . . . UV_Scanlines + * . . . . . . . . . . . . . . . . | + * . . . . . . . . . . . . . . . . V + * . . . . . . . . . . . . . . . . --> Buffer size alignment + * + * Y_Stride : Width * 2 aligned to 128 + * UV_Stride : Width * 2 aligned to 128 + * Y_Scanlines: Height aligned to 32 + * UV_Scanlines: Height/2 aligned to 16 + * Extradata: Arbitrary (software-imposed) padding + * Total size = align((Y_Stride * Y_Scanlines + * + UV_Stride * UV_Scanlines + * + max(Extradata, Y_Stride * 8), 4096) + */ + COLOR_FMT_P010, +}; + +#define COLOR_FMT_RGBA1010102_UBWC COLOR_FMT_RGBA1010102_UBWC +#define COLOR_FMT_RGB565_UBWC COLOR_FMT_RGB565_UBWC +#define COLOR_FMT_P010_UBWC COLOR_FMT_P010_UBWC +#define COLOR_FMT_P010 COLOR_FMT_P010 + +static inline unsigned int VENUS_EXTRADATA_SIZE(int width, int height) +{ + (void)height; + (void)width; + + /* + * In the future, calculate the size based on the w/h but just + * hardcode it for now since 16K satisfies all current usecases. + */ + return 16 * 1024; +} + +/* + * Function arguments: + * @color_fmt + * @width + * Progressive: width + * Interlaced: width + */ +static inline unsigned int VENUS_Y_STRIDE(int color_fmt, int width) +{ + unsigned int alignment, stride = 0; + + if (!width) + goto invalid_input; + + switch (color_fmt) { + case COLOR_FMT_NV21: + case COLOR_FMT_NV12: + case COLOR_FMT_NV12_MVTB: + case COLOR_FMT_NV12_UBWC: + alignment = 128; + stride = MSM_MEDIA_ALIGN(width, alignment); + break; + case COLOR_FMT_NV12_BPP10_UBWC: + alignment = 256; + stride = MSM_MEDIA_ALIGN(width, 192); + stride = MSM_MEDIA_ALIGN(stride * 4/3, alignment); + break; + case COLOR_FMT_P010_UBWC: + alignment = 256; + stride = MSM_MEDIA_ALIGN(width * 2, alignment); + break; + case COLOR_FMT_P010: + alignment = 128; + stride = MSM_MEDIA_ALIGN(width*2, alignment); + break; + default: + break; + } +invalid_input: + return stride; +} + +/* + * Function arguments: + * @color_fmt + * @width + * Progressive: width + * Interlaced: width + */ +static inline unsigned int VENUS_UV_STRIDE(int color_fmt, int width) +{ + unsigned int alignment, stride = 0; + + if (!width) + goto invalid_input; + + switch (color_fmt) { + case COLOR_FMT_NV21: + case COLOR_FMT_NV12: + case COLOR_FMT_NV12_MVTB: + case COLOR_FMT_NV12_UBWC: + alignment = 128; + stride = MSM_MEDIA_ALIGN(width, alignment); + break; + case COLOR_FMT_NV12_BPP10_UBWC: + alignment = 256; + stride = MSM_MEDIA_ALIGN(width, 192); + stride = MSM_MEDIA_ALIGN(stride * 4/3, alignment); + break; + case COLOR_FMT_P010_UBWC: + alignment = 256; + stride = MSM_MEDIA_ALIGN(width * 2, alignment); + break; + case COLOR_FMT_P010: + alignment = 128; + stride = MSM_MEDIA_ALIGN(width*2, alignment); + break; + default: + break; + } +invalid_input: + return stride; +} + +/* + * Function arguments: + * @color_fmt + * @height + * Progressive: height + * Interlaced: (height+1)>>1 + */ +static inline unsigned int VENUS_Y_SCANLINES(int color_fmt, int height) +{ + unsigned int alignment, sclines = 0; + + if (!height) + goto invalid_input; + + switch (color_fmt) { + case COLOR_FMT_NV21: + case COLOR_FMT_NV12: + case COLOR_FMT_NV12_MVTB: + case COLOR_FMT_NV12_UBWC: + case COLOR_FMT_P010: + alignment = 32; + break; + case COLOR_FMT_NV12_BPP10_UBWC: + case COLOR_FMT_P010_UBWC: + alignment = 16; + break; + default: + return 0; + } + sclines = MSM_MEDIA_ALIGN(height, alignment); +invalid_input: + return sclines; +} + +/* + * Function arguments: + * @color_fmt + * @height + * Progressive: height + * Interlaced: (height+1)>>1 + */ +static inline unsigned int VENUS_UV_SCANLINES(int color_fmt, int height) +{ + unsigned int alignment, sclines = 0; + + if (!height) + goto invalid_input; + + switch (color_fmt) { + case COLOR_FMT_NV21: + case COLOR_FMT_NV12: + case COLOR_FMT_NV12_MVTB: + case COLOR_FMT_NV12_BPP10_UBWC: + case COLOR_FMT_P010_UBWC: + case COLOR_FMT_P010: + alignment = 16; + break; + case COLOR_FMT_NV12_UBWC: + alignment = 32; + break; + default: + goto invalid_input; + } + + sclines = MSM_MEDIA_ALIGN((height+1)>>1, alignment); + +invalid_input: + return sclines; +} + +/* + * Function arguments: + * @color_fmt + * @width + * Progressive: width + * Interlaced: width + */ +static inline unsigned int VENUS_Y_META_STRIDE(int color_fmt, int width) +{ + int y_tile_width = 0, y_meta_stride = 0; + + if (!width) + goto invalid_input; + + switch (color_fmt) { + case COLOR_FMT_NV12_UBWC: + case COLOR_FMT_P010_UBWC: + y_tile_width = 32; + break; + case COLOR_FMT_NV12_BPP10_UBWC: + y_tile_width = 48; + break; + default: + goto invalid_input; + } + + y_meta_stride = MSM_MEDIA_ROUNDUP(width, y_tile_width); + y_meta_stride = MSM_MEDIA_ALIGN(y_meta_stride, 64); + +invalid_input: + return y_meta_stride; +} + +/* + * Function arguments: + * @color_fmt + * @height + * Progressive: height + * Interlaced: (height+1)>>1 + */ +static inline unsigned int VENUS_Y_META_SCANLINES(int color_fmt, int height) +{ + int y_tile_height = 0, y_meta_scanlines = 0; + + if (!height) + goto invalid_input; + + switch (color_fmt) { + case COLOR_FMT_NV12_UBWC: + y_tile_height = 8; + break; + case COLOR_FMT_NV12_BPP10_UBWC: + case COLOR_FMT_P010_UBWC: + y_tile_height = 4; + break; + default: + goto invalid_input; + } + + y_meta_scanlines = MSM_MEDIA_ROUNDUP(height, y_tile_height); + y_meta_scanlines = MSM_MEDIA_ALIGN(y_meta_scanlines, 16); + +invalid_input: + return y_meta_scanlines; +} + +/* + * Function arguments: + * @color_fmt + * @width + * Progressive: width + * Interlaced: width + */ +static inline unsigned int VENUS_UV_META_STRIDE(int color_fmt, int width) +{ + int uv_tile_width = 0, uv_meta_stride = 0; + + if (!width) + goto invalid_input; + + switch (color_fmt) { + case COLOR_FMT_NV12_UBWC: + case COLOR_FMT_P010_UBWC: + uv_tile_width = 16; + break; + case COLOR_FMT_NV12_BPP10_UBWC: + uv_tile_width = 24; + break; + default: + goto invalid_input; + } + + uv_meta_stride = MSM_MEDIA_ROUNDUP((width+1)>>1, uv_tile_width); + uv_meta_stride = MSM_MEDIA_ALIGN(uv_meta_stride, 64); + +invalid_input: + return uv_meta_stride; +} + +/* + * Function arguments: + * @color_fmt + * @height + * Progressive: height + * Interlaced: (height+1)>>1 + */ +static inline unsigned int VENUS_UV_META_SCANLINES(int color_fmt, int height) +{ + int uv_tile_height = 0, uv_meta_scanlines = 0; + + if (!height) + goto invalid_input; + + switch (color_fmt) { + case COLOR_FMT_NV12_UBWC: + uv_tile_height = 8; + break; + case COLOR_FMT_NV12_BPP10_UBWC: + case COLOR_FMT_P010_UBWC: + uv_tile_height = 4; + break; + default: + goto invalid_input; + } + + uv_meta_scanlines = MSM_MEDIA_ROUNDUP((height+1)>>1, uv_tile_height); + uv_meta_scanlines = MSM_MEDIA_ALIGN(uv_meta_scanlines, 16); + +invalid_input: + return uv_meta_scanlines; +} + +static inline unsigned int VENUS_RGB_STRIDE(int color_fmt, int width) +{ + unsigned int alignment = 0, stride = 0, bpp = 4; + + if (!width) + goto invalid_input; + + switch (color_fmt) { + case COLOR_FMT_RGBA8888: + alignment = 128; + break; + case COLOR_FMT_RGB565_UBWC: + alignment = 256; + bpp = 2; + break; + case COLOR_FMT_RGBA8888_UBWC: + case COLOR_FMT_RGBA1010102_UBWC: + alignment = 256; + break; + default: + goto invalid_input; + } + + stride = MSM_MEDIA_ALIGN(width * bpp, alignment); + +invalid_input: + return stride; +} + +static inline unsigned int VENUS_RGB_SCANLINES(int color_fmt, int height) +{ + unsigned int alignment = 0, scanlines = 0; + + if (!height) + goto invalid_input; + + switch (color_fmt) { + case COLOR_FMT_RGBA8888: + alignment = 32; + break; + case COLOR_FMT_RGBA8888_UBWC: + case COLOR_FMT_RGBA1010102_UBWC: + case COLOR_FMT_RGB565_UBWC: + alignment = 16; + break; + default: + goto invalid_input; + } + + scanlines = MSM_MEDIA_ALIGN(height, alignment); + +invalid_input: + return scanlines; +} + +static inline unsigned int VENUS_RGB_META_STRIDE(int color_fmt, int width) +{ + int rgb_tile_width = 0, rgb_meta_stride = 0; + + if (!width) + goto invalid_input; + + switch (color_fmt) { + case COLOR_FMT_RGBA8888_UBWC: + case COLOR_FMT_RGBA1010102_UBWC: + case COLOR_FMT_RGB565_UBWC: + rgb_tile_width = 16; + break; + default: + goto invalid_input; + } + + rgb_meta_stride = MSM_MEDIA_ROUNDUP(width, rgb_tile_width); + rgb_meta_stride = MSM_MEDIA_ALIGN(rgb_meta_stride, 64); + +invalid_input: + return rgb_meta_stride; +} + +static inline unsigned int VENUS_RGB_META_SCANLINES(int color_fmt, int height) +{ + int rgb_tile_height = 0, rgb_meta_scanlines = 0; + + if (!height) + goto invalid_input; + + switch (color_fmt) { + case COLOR_FMT_RGBA8888_UBWC: + case COLOR_FMT_RGBA1010102_UBWC: + case COLOR_FMT_RGB565_UBWC: + rgb_tile_height = 4; + break; + default: + goto invalid_input; + } + + rgb_meta_scanlines = MSM_MEDIA_ROUNDUP(height, rgb_tile_height); + rgb_meta_scanlines = MSM_MEDIA_ALIGN(rgb_meta_scanlines, 16); + +invalid_input: + return rgb_meta_scanlines; +} + +/* + * Function arguments: + * @color_fmt + * @width + * Progressive: width + * Interlaced: width + * @height + * Progressive: height + * Interlaced: height + */ +static inline unsigned int VENUS_BUFFER_SIZE( + int color_fmt, int width, int height) +{ + const unsigned int extra_size = VENUS_EXTRADATA_SIZE(width, height); + unsigned int uv_alignment = 0, size = 0; + unsigned int y_plane, uv_plane, y_stride, + uv_stride, y_sclines, uv_sclines; + unsigned int y_ubwc_plane = 0, uv_ubwc_plane = 0; + unsigned int y_meta_stride = 0, y_meta_scanlines = 0; + unsigned int uv_meta_stride = 0, uv_meta_scanlines = 0; + unsigned int y_meta_plane = 0, uv_meta_plane = 0; + unsigned int rgb_stride = 0, rgb_scanlines = 0; + unsigned int rgb_plane = 0, rgb_ubwc_plane = 0, rgb_meta_plane = 0; + unsigned int rgb_meta_stride = 0, rgb_meta_scanlines = 0; + + if (!width || !height) + goto invalid_input; + + y_stride = VENUS_Y_STRIDE(color_fmt, width); + uv_stride = VENUS_UV_STRIDE(color_fmt, width); + y_sclines = VENUS_Y_SCANLINES(color_fmt, height); + uv_sclines = VENUS_UV_SCANLINES(color_fmt, height); + rgb_stride = VENUS_RGB_STRIDE(color_fmt, width); + rgb_scanlines = VENUS_RGB_SCANLINES(color_fmt, height); + + switch (color_fmt) { + case COLOR_FMT_NV21: + case COLOR_FMT_NV12: + case COLOR_FMT_P010: + uv_alignment = 4096; + y_plane = y_stride * y_sclines; + uv_plane = uv_stride * uv_sclines + uv_alignment; + size = y_plane + uv_plane + + MSM_MEDIA_MAX(extra_size, 8 * y_stride); + size = MSM_MEDIA_ALIGN(size, 4096); + break; + case COLOR_FMT_NV12_MVTB: + uv_alignment = 4096; + y_plane = y_stride * y_sclines; + uv_plane = uv_stride * uv_sclines + uv_alignment; + size = y_plane + uv_plane; + size = 2 * size + extra_size; + size = MSM_MEDIA_ALIGN(size, 4096); + break; + case COLOR_FMT_NV12_UBWC: + y_sclines = VENUS_Y_SCANLINES(color_fmt, (height+1)>>1); + y_ubwc_plane = MSM_MEDIA_ALIGN(y_stride * y_sclines, 4096); + uv_sclines = VENUS_UV_SCANLINES(color_fmt, (height+1)>>1); + uv_ubwc_plane = MSM_MEDIA_ALIGN(uv_stride * uv_sclines, 4096); + y_meta_stride = VENUS_Y_META_STRIDE(color_fmt, width); + y_meta_scanlines = + VENUS_Y_META_SCANLINES(color_fmt, (height+1)>>1); + y_meta_plane = MSM_MEDIA_ALIGN( + y_meta_stride * y_meta_scanlines, 4096); + uv_meta_stride = VENUS_UV_META_STRIDE(color_fmt, width); + uv_meta_scanlines = + VENUS_UV_META_SCANLINES(color_fmt, (height+1)>>1); + uv_meta_plane = MSM_MEDIA_ALIGN(uv_meta_stride * + uv_meta_scanlines, 4096); + + size = (y_ubwc_plane + uv_ubwc_plane + y_meta_plane + + uv_meta_plane)*2 + + MSM_MEDIA_MAX(extra_size + 8192, 48 * y_stride); + size = MSM_MEDIA_ALIGN(size, 4096); + break; + case COLOR_FMT_NV12_BPP10_UBWC: + y_ubwc_plane = MSM_MEDIA_ALIGN(y_stride * y_sclines, 4096); + uv_ubwc_plane = MSM_MEDIA_ALIGN(uv_stride * uv_sclines, 4096); + y_meta_stride = VENUS_Y_META_STRIDE(color_fmt, width); + y_meta_scanlines = VENUS_Y_META_SCANLINES(color_fmt, height); + y_meta_plane = MSM_MEDIA_ALIGN( + y_meta_stride * y_meta_scanlines, 4096); + uv_meta_stride = VENUS_UV_META_STRIDE(color_fmt, width); + uv_meta_scanlines = VENUS_UV_META_SCANLINES(color_fmt, height); + uv_meta_plane = MSM_MEDIA_ALIGN(uv_meta_stride * + uv_meta_scanlines, 4096); + + size = y_ubwc_plane + uv_ubwc_plane + y_meta_plane + + uv_meta_plane + + MSM_MEDIA_MAX(extra_size + 8192, 48 * y_stride); + size = MSM_MEDIA_ALIGN(size, 4096); + break; + case COLOR_FMT_P010_UBWC: + y_ubwc_plane = MSM_MEDIA_ALIGN(y_stride * y_sclines, 4096); + uv_ubwc_plane = MSM_MEDIA_ALIGN(uv_stride * uv_sclines, 4096); + y_meta_stride = VENUS_Y_META_STRIDE(color_fmt, width); + y_meta_scanlines = VENUS_Y_META_SCANLINES(color_fmt, height); + y_meta_plane = MSM_MEDIA_ALIGN( + y_meta_stride * y_meta_scanlines, 4096); + uv_meta_stride = VENUS_UV_META_STRIDE(color_fmt, width); + uv_meta_scanlines = VENUS_UV_META_SCANLINES(color_fmt, height); + uv_meta_plane = MSM_MEDIA_ALIGN(uv_meta_stride * + uv_meta_scanlines, 4096); + + size = y_ubwc_plane + uv_ubwc_plane + y_meta_plane + + uv_meta_plane; + size = MSM_MEDIA_ALIGN(size, 4096); + break; + case COLOR_FMT_RGBA8888: + rgb_plane = MSM_MEDIA_ALIGN(rgb_stride * rgb_scanlines, 4096); + size = rgb_plane; + size = MSM_MEDIA_ALIGN(size, 4096); + break; + case COLOR_FMT_RGBA8888_UBWC: + case COLOR_FMT_RGBA1010102_UBWC: + case COLOR_FMT_RGB565_UBWC: + rgb_ubwc_plane = MSM_MEDIA_ALIGN(rgb_stride * rgb_scanlines, + 4096); + rgb_meta_stride = VENUS_RGB_META_STRIDE(color_fmt, width); + rgb_meta_scanlines = VENUS_RGB_META_SCANLINES(color_fmt, + height); + rgb_meta_plane = MSM_MEDIA_ALIGN(rgb_meta_stride * + rgb_meta_scanlines, 4096); + size = rgb_ubwc_plane + rgb_meta_plane; + size = MSM_MEDIA_ALIGN(size, 4096); + break; + default: + break; + } +invalid_input: + return size; +} + +static inline unsigned int VENUS_VIEW2_OFFSET( + int color_fmt, int width, int height) +{ + unsigned int offset = 0; + unsigned int y_plane, uv_plane, y_stride, + uv_stride, y_sclines, uv_sclines; + if (!width || !height) + goto invalid_input; + + y_stride = VENUS_Y_STRIDE(color_fmt, width); + uv_stride = VENUS_UV_STRIDE(color_fmt, width); + y_sclines = VENUS_Y_SCANLINES(color_fmt, height); + uv_sclines = VENUS_UV_SCANLINES(color_fmt, height); + switch (color_fmt) { + case COLOR_FMT_NV12_MVTB: + y_plane = y_stride * y_sclines; + uv_plane = uv_stride * uv_sclines; + offset = y_plane + uv_plane; + break; + default: + break; + } +invalid_input: + return offset; +} + +#endif diff --git a/drivers/gpu/drm/msm/disp/mdp4/mdp4_kms.c b/drivers/gpu/drm/msm/disp/mdp4/mdp4_kms.c index 4b646bf9c214..44d1cda56974 100644 --- a/drivers/gpu/drm/msm/disp/mdp4/mdp4_kms.c +++ b/drivers/gpu/drm/msm/disp/mdp4/mdp4_kms.c @@ -125,6 +125,8 @@ static void mdp4_complete_commit(struct msm_kms *kms, struct drm_atomic_state *s struct drm_crtc *crtc; struct drm_crtc_state *crtc_state; + drm_atomic_helper_wait_for_vblanks(mdp4_kms->dev, state); + /* see 119ecb7fd */ for_each_new_crtc_in_state(state, crtc, crtc_state, i) drm_crtc_vblank_put(crtc); diff --git a/drivers/gpu/drm/msm/disp/mdp5/mdp5_crtc.c b/drivers/gpu/drm/msm/disp/mdp5/mdp5_crtc.c index 24e00274844b..b1da9ce54379 100644 --- a/drivers/gpu/drm/msm/disp/mdp5/mdp5_crtc.c +++ b/drivers/gpu/drm/msm/disp/mdp5/mdp5_crtc.c @@ -65,7 +65,7 @@ struct mdp5_crtc { struct drm_gem_object *scanout_bo; uint64_t iova; uint32_t width, height; - uint32_t x, y; + int x, y; } cursor; }; #define to_mdp5_crtc(x) container_of(x, struct mdp5_crtc, base) @@ -760,20 +760,31 @@ static void get_roi(struct drm_crtc *crtc, uint32_t *roi_w, uint32_t *roi_h) * Cursor Region Of Interest (ROI) is a plane read from cursor * buffer to render. The ROI region is determined by the visibility of * the cursor point. In the default Cursor image the cursor point will - * be at the top left of the cursor image, unless it is specified - * otherwise using hotspot feature. + * be at the top left of the cursor image. * + * Without rotation: * If the cursor point reaches the right (xres - x < cursor.width) or * bottom (yres - y < cursor.height) boundary of the screen, then ROI * width and ROI height need to be evaluated to crop the cursor image * accordingly. * (xres-x) will be new cursor width when x > (xres - cursor.width) * (yres-y) will be new cursor height when y > (yres - cursor.height) + * + * With rotation: + * We get negative x and/or y coordinates. + * (cursor.width - abs(x)) will be new cursor width when x < 0 + * (cursor.height - abs(y)) will be new cursor width when y < 0 */ - *roi_w = min(mdp5_crtc->cursor.width, xres - + if (mdp5_crtc->cursor.x >= 0) + *roi_w = min(mdp5_crtc->cursor.width, xres - mdp5_crtc->cursor.x); - *roi_h = min(mdp5_crtc->cursor.height, yres - + else + *roi_w = mdp5_crtc->cursor.width - abs(mdp5_crtc->cursor.x); + if (mdp5_crtc->cursor.y >= 0) + *roi_h = min(mdp5_crtc->cursor.height, yres - mdp5_crtc->cursor.y); + else + *roi_h = mdp5_crtc->cursor.height - abs(mdp5_crtc->cursor.y); } static void mdp5_crtc_restore_cursor(struct drm_crtc *crtc) @@ -783,7 +794,7 @@ static void mdp5_crtc_restore_cursor(struct drm_crtc *crtc) struct mdp5_kms *mdp5_kms = get_kms(crtc); const enum mdp5_cursor_alpha cur_alpha = CURSOR_ALPHA_PER_PIXEL; uint32_t blendcfg, stride; - uint32_t x, y, width, height; + uint32_t x, y, src_x, src_y, width, height; uint32_t roi_w, roi_h; int lm; @@ -800,6 +811,26 @@ static void mdp5_crtc_restore_cursor(struct drm_crtc *crtc) get_roi(crtc, &roi_w, &roi_h); + /* If cusror buffer overlaps due to rotation on the + * upper or left screen border the pixel offset inside + * the cursor buffer of the ROI is the positive overlap + * distance. + */ + if (mdp5_crtc->cursor.x < 0) { + src_x = abs(mdp5_crtc->cursor.x); + x = 0; + } else { + src_x = 0; + } + if (mdp5_crtc->cursor.y < 0) { + src_y = abs(mdp5_crtc->cursor.y); + y = 0; + } else { + src_y = 0; + } + DBG("%s: x=%d, y=%d roi_w=%d roi_h=%d src_x=%d src_y=%d", + crtc->name, x, y, roi_w, roi_h, src_x, src_y); + mdp5_write(mdp5_kms, REG_MDP5_LM_CURSOR_STRIDE(lm), stride); mdp5_write(mdp5_kms, REG_MDP5_LM_CURSOR_FORMAT(lm), MDP5_LM_CURSOR_FORMAT_FORMAT(CURSOR_FMT_ARGB8888)); @@ -812,6 +843,9 @@ static void mdp5_crtc_restore_cursor(struct drm_crtc *crtc) mdp5_write(mdp5_kms, REG_MDP5_LM_CURSOR_START_XY(lm), MDP5_LM_CURSOR_START_XY_Y_START(y) | MDP5_LM_CURSOR_START_XY_X_START(x)); + mdp5_write(mdp5_kms, REG_MDP5_LM_CURSOR_XY(lm), + MDP5_LM_CURSOR_XY_SRC_Y(src_y) | + MDP5_LM_CURSOR_XY_SRC_X(src_x)); mdp5_write(mdp5_kms, REG_MDP5_LM_CURSOR_BASE_ADDR(lm), mdp5_crtc->cursor.iova); @@ -932,8 +966,9 @@ static int mdp5_crtc_cursor_move(struct drm_crtc *crtc, int x, int y) if (unlikely(!crtc->state->enable)) return 0; - mdp5_crtc->cursor.x = x = max(x, 0); - mdp5_crtc->cursor.y = y = max(y, 0); + /* accept negative x/y coordinates up to maximum cursor overlap */ + mdp5_crtc->cursor.x = x = max(x, -(int)mdp5_crtc->cursor.width); + mdp5_crtc->cursor.y = y = max(y, -(int)mdp5_crtc->cursor.height); get_roi(crtc, &roi_w, &roi_h); diff --git a/drivers/gpu/drm/msm/disp/mdp5/mdp5_encoder.c b/drivers/gpu/drm/msm/disp/mdp5/mdp5_encoder.c index 9af94e35f678..fcd44d1d1068 100644 --- a/drivers/gpu/drm/msm/disp/mdp5/mdp5_encoder.c +++ b/drivers/gpu/drm/msm/disp/mdp5/mdp5_encoder.c @@ -319,7 +319,17 @@ static int mdp5_encoder_atomic_check(struct drm_encoder *encoder, mdp5_cstate->ctl = ctl; mdp5_cstate->pipeline.intf = intf; - mdp5_cstate->defer_start = true; + + /* + * This is a bit awkward, but we want to flush the CTL and hit the + * START bit at most once for an atomic update. In the non-full- + * modeset case, this is done from crtc->atomic_flush(), but that + * is too early in the case of full modeset, in which case we + * defer to encoder->enable(). But we need to *know* whether + * encoder->enable() will be called to do this: + */ + if (drm_atomic_crtc_needs_modeset(crtc_state)) + mdp5_cstate->defer_start = true; return 0; } diff --git a/drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.c b/drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.c index 6e12e275deba..bddd625ab91b 100644 --- a/drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.c +++ b/drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.c @@ -170,6 +170,8 @@ static void mdp5_complete_commit(struct msm_kms *kms, struct drm_atomic_state *s struct device *dev = &mdp5_kms->pdev->dev; struct mdp5_global_state *global_state; + drm_atomic_helper_wait_for_vblanks(mdp5_kms->dev, state); + global_state = mdp5_get_existing_global_state(mdp5_kms); if (mdp5_kms->smp) diff --git a/drivers/gpu/drm/msm/disp/mdp5/mdp5_mdss.c b/drivers/gpu/drm/msm/disp/mdp5/mdp5_mdss.c index f2a0db7a8a03..1cc4e57f0226 100644 --- a/drivers/gpu/drm/msm/disp/mdp5/mdp5_mdss.c +++ b/drivers/gpu/drm/msm/disp/mdp5/mdp5_mdss.c @@ -20,12 +20,10 @@ #include "msm_drv.h" #include "mdp5_kms.h" -/* - * If needed, this can become more specific: something like struct mdp5_mdss, - * which contains a 'struct msm_mdss base' member. - */ -struct msm_mdss { - struct drm_device *dev; +#define to_mdp5_mdss(x) container_of(x, struct mdp5_mdss, base) + +struct mdp5_mdss { + struct msm_mdss base; void __iomem *mmio, *vbif; @@ -41,22 +39,22 @@ struct msm_mdss { } irqcontroller; }; -static inline void mdss_write(struct msm_mdss *mdss, u32 reg, u32 data) +static inline void mdss_write(struct mdp5_mdss *mdp5_mdss, u32 reg, u32 data) { - msm_writel(data, mdss->mmio + reg); + msm_writel(data, mdp5_mdss->mmio + reg); } -static inline u32 mdss_read(struct msm_mdss *mdss, u32 reg) +static inline u32 mdss_read(struct mdp5_mdss *mdp5_mdss, u32 reg) { - return msm_readl(mdss->mmio + reg); + return msm_readl(mdp5_mdss->mmio + reg); } static irqreturn_t mdss_irq(int irq, void *arg) { - struct msm_mdss *mdss = arg; + struct mdp5_mdss *mdp5_mdss = arg; u32 intr; - intr = mdss_read(mdss, REG_MDSS_HW_INTR_STATUS); + intr = mdss_read(mdp5_mdss, REG_MDSS_HW_INTR_STATUS); VERB("intr=%08x", intr); @@ -64,7 +62,7 @@ static irqreturn_t mdss_irq(int irq, void *arg) irq_hw_number_t hwirq = fls(intr) - 1; generic_handle_irq(irq_find_mapping( - mdss->irqcontroller.domain, hwirq)); + mdp5_mdss->irqcontroller.domain, hwirq)); intr &= ~(1 << hwirq); } @@ -84,19 +82,19 @@ static irqreturn_t mdss_irq(int irq, void *arg) static void mdss_hw_mask_irq(struct irq_data *irqd) { - struct msm_mdss *mdss = irq_data_get_irq_chip_data(irqd); + struct mdp5_mdss *mdp5_mdss = irq_data_get_irq_chip_data(irqd); smp_mb__before_atomic(); - clear_bit(irqd->hwirq, &mdss->irqcontroller.enabled_mask); + clear_bit(irqd->hwirq, &mdp5_mdss->irqcontroller.enabled_mask); smp_mb__after_atomic(); } static void mdss_hw_unmask_irq(struct irq_data *irqd) { - struct msm_mdss *mdss = irq_data_get_irq_chip_data(irqd); + struct mdp5_mdss *mdp5_mdss = irq_data_get_irq_chip_data(irqd); smp_mb__before_atomic(); - set_bit(irqd->hwirq, &mdss->irqcontroller.enabled_mask); + set_bit(irqd->hwirq, &mdp5_mdss->irqcontroller.enabled_mask); smp_mb__after_atomic(); } @@ -109,13 +107,13 @@ static struct irq_chip mdss_hw_irq_chip = { static int mdss_hw_irqdomain_map(struct irq_domain *d, unsigned int irq, irq_hw_number_t hwirq) { - struct msm_mdss *mdss = d->host_data; + struct mdp5_mdss *mdp5_mdss = d->host_data; if (!(VALID_IRQS & (1 << hwirq))) return -EPERM; irq_set_chip_and_handler(irq, &mdss_hw_irq_chip, handle_level_irq); - irq_set_chip_data(irq, mdss); + irq_set_chip_data(irq, mdp5_mdss); return 0; } @@ -126,90 +124,99 @@ static const struct irq_domain_ops mdss_hw_irqdomain_ops = { }; -static int mdss_irq_domain_init(struct msm_mdss *mdss) +static int mdss_irq_domain_init(struct mdp5_mdss *mdp5_mdss) { - struct device *dev = mdss->dev->dev; + struct device *dev = mdp5_mdss->base.dev->dev; struct irq_domain *d; d = irq_domain_add_linear(dev->of_node, 32, &mdss_hw_irqdomain_ops, - mdss); + mdp5_mdss); if (!d) { dev_err(dev, "mdss irq domain add failed\n"); return -ENXIO; } - mdss->irqcontroller.enabled_mask = 0; - mdss->irqcontroller.domain = d; + mdp5_mdss->irqcontroller.enabled_mask = 0; + mdp5_mdss->irqcontroller.domain = d; return 0; } -int msm_mdss_enable(struct msm_mdss *mdss) +static int mdp5_mdss_enable(struct msm_mdss *mdss) { + struct mdp5_mdss *mdp5_mdss = to_mdp5_mdss(mdss); DBG(""); - clk_prepare_enable(mdss->ahb_clk); - if (mdss->axi_clk) - clk_prepare_enable(mdss->axi_clk); - if (mdss->vsync_clk) - clk_prepare_enable(mdss->vsync_clk); + clk_prepare_enable(mdp5_mdss->ahb_clk); + if (mdp5_mdss->axi_clk) + clk_prepare_enable(mdp5_mdss->axi_clk); + if (mdp5_mdss->vsync_clk) + clk_prepare_enable(mdp5_mdss->vsync_clk); return 0; } -int msm_mdss_disable(struct msm_mdss *mdss) +static int mdp5_mdss_disable(struct msm_mdss *mdss) { + struct mdp5_mdss *mdp5_mdss = to_mdp5_mdss(mdss); DBG(""); - if (mdss->vsync_clk) - clk_disable_unprepare(mdss->vsync_clk); - if (mdss->axi_clk) - clk_disable_unprepare(mdss->axi_clk); - clk_disable_unprepare(mdss->ahb_clk); + if (mdp5_mdss->vsync_clk) + clk_disable_unprepare(mdp5_mdss->vsync_clk); + if (mdp5_mdss->axi_clk) + clk_disable_unprepare(mdp5_mdss->axi_clk); + clk_disable_unprepare(mdp5_mdss->ahb_clk); return 0; } -static int msm_mdss_get_clocks(struct msm_mdss *mdss) +static int msm_mdss_get_clocks(struct mdp5_mdss *mdp5_mdss) { - struct platform_device *pdev = to_platform_device(mdss->dev->dev); + struct platform_device *pdev = + to_platform_device(mdp5_mdss->base.dev->dev); - mdss->ahb_clk = msm_clk_get(pdev, "iface"); - if (IS_ERR(mdss->ahb_clk)) - mdss->ahb_clk = NULL; + mdp5_mdss->ahb_clk = msm_clk_get(pdev, "iface"); + if (IS_ERR(mdp5_mdss->ahb_clk)) + mdp5_mdss->ahb_clk = NULL; - mdss->axi_clk = msm_clk_get(pdev, "bus"); - if (IS_ERR(mdss->axi_clk)) - mdss->axi_clk = NULL; + mdp5_mdss->axi_clk = msm_clk_get(pdev, "bus"); + if (IS_ERR(mdp5_mdss->axi_clk)) + mdp5_mdss->axi_clk = NULL; - mdss->vsync_clk = msm_clk_get(pdev, "vsync"); - if (IS_ERR(mdss->vsync_clk)) - mdss->vsync_clk = NULL; + mdp5_mdss->vsync_clk = msm_clk_get(pdev, "vsync"); + if (IS_ERR(mdp5_mdss->vsync_clk)) + mdp5_mdss->vsync_clk = NULL; return 0; } -void msm_mdss_destroy(struct drm_device *dev) +static void mdp5_mdss_destroy(struct drm_device *dev) { struct msm_drm_private *priv = dev->dev_private; - struct msm_mdss *mdss = priv->mdss; + struct mdp5_mdss *mdp5_mdss = to_mdp5_mdss(priv->mdss); - if (!mdss) + if (!mdp5_mdss) return; - irq_domain_remove(mdss->irqcontroller.domain); - mdss->irqcontroller.domain = NULL; + irq_domain_remove(mdp5_mdss->irqcontroller.domain); + mdp5_mdss->irqcontroller.domain = NULL; - regulator_disable(mdss->vdd); + regulator_disable(mdp5_mdss->vdd); pm_runtime_disable(dev->dev); } -int msm_mdss_init(struct drm_device *dev) +static const struct msm_mdss_funcs mdss_funcs = { + .enable = mdp5_mdss_enable, + .disable = mdp5_mdss_disable, + .destroy = mdp5_mdss_destroy, +}; + +int mdp5_mdss_init(struct drm_device *dev) { struct platform_device *pdev = to_platform_device(dev->dev); struct msm_drm_private *priv = dev->dev_private; - struct msm_mdss *mdss; + struct mdp5_mdss *mdp5_mdss; int ret; DBG(""); @@ -217,40 +224,40 @@ int msm_mdss_init(struct drm_device *dev) if (!of_device_is_compatible(dev->dev->of_node, "qcom,mdss")) return 0; - mdss = devm_kzalloc(dev->dev, sizeof(*mdss), GFP_KERNEL); - if (!mdss) { + mdp5_mdss = devm_kzalloc(dev->dev, sizeof(*mdp5_mdss), GFP_KERNEL); + if (!mdp5_mdss) { ret = -ENOMEM; goto fail; } - mdss->dev = dev; + mdp5_mdss->base.dev = dev; - mdss->mmio = msm_ioremap(pdev, "mdss_phys", "MDSS"); - if (IS_ERR(mdss->mmio)) { - ret = PTR_ERR(mdss->mmio); + mdp5_mdss->mmio = msm_ioremap(pdev, "mdss_phys", "MDSS"); + if (IS_ERR(mdp5_mdss->mmio)) { + ret = PTR_ERR(mdp5_mdss->mmio); goto fail; } - mdss->vbif = msm_ioremap(pdev, "vbif_phys", "VBIF"); - if (IS_ERR(mdss->vbif)) { - ret = PTR_ERR(mdss->vbif); + mdp5_mdss->vbif = msm_ioremap(pdev, "vbif_phys", "VBIF"); + if (IS_ERR(mdp5_mdss->vbif)) { + ret = PTR_ERR(mdp5_mdss->vbif); goto fail; } - ret = msm_mdss_get_clocks(mdss); + ret = msm_mdss_get_clocks(mdp5_mdss); if (ret) { dev_err(dev->dev, "failed to get clocks: %d\n", ret); goto fail; } /* Regulator to enable GDSCs in downstream kernels */ - mdss->vdd = devm_regulator_get(dev->dev, "vdd"); - if (IS_ERR(mdss->vdd)) { - ret = PTR_ERR(mdss->vdd); + mdp5_mdss->vdd = devm_regulator_get(dev->dev, "vdd"); + if (IS_ERR(mdp5_mdss->vdd)) { + ret = PTR_ERR(mdp5_mdss->vdd); goto fail; } - ret = regulator_enable(mdss->vdd); + ret = regulator_enable(mdp5_mdss->vdd); if (ret) { dev_err(dev->dev, "failed to enable regulator vdd: %d\n", ret); @@ -258,25 +265,26 @@ int msm_mdss_init(struct drm_device *dev) } ret = devm_request_irq(dev->dev, platform_get_irq(pdev, 0), - mdss_irq, 0, "mdss_isr", mdss); + mdss_irq, 0, "mdss_isr", mdp5_mdss); if (ret) { dev_err(dev->dev, "failed to init irq: %d\n", ret); goto fail_irq; } - ret = mdss_irq_domain_init(mdss); + ret = mdss_irq_domain_init(mdp5_mdss); if (ret) { dev_err(dev->dev, "failed to init sub-block irqs: %d\n", ret); goto fail_irq; } - priv->mdss = mdss; + mdp5_mdss->base.funcs = &mdss_funcs; + priv->mdss = &mdp5_mdss->base; pm_runtime_enable(dev->dev); return 0; fail_irq: - regulator_disable(mdss->vdd); + regulator_disable(mdp5_mdss->vdd); fail: return ret; } diff --git a/drivers/gpu/drm/msm/dsi/dsi.c b/drivers/gpu/drm/msm/dsi/dsi.c index b744bcc7d8ad..ff8164cc6738 100644 --- a/drivers/gpu/drm/msm/dsi/dsi.c +++ b/drivers/gpu/drm/msm/dsi/dsi.c @@ -208,6 +208,9 @@ int msm_dsi_modeset_init(struct msm_dsi *msm_dsi, struct drm_device *dev, goto fail; } + if (!msm_dsi_manager_validate_current_config(msm_dsi->id)) + goto fail; + msm_dsi->encoder = encoder; msm_dsi->bridge = msm_dsi_manager_bridge_init(msm_dsi->id); diff --git a/drivers/gpu/drm/msm/dsi/dsi.h b/drivers/gpu/drm/msm/dsi/dsi.h index 70d9a9a47acd..08f3fc6771b7 100644 --- a/drivers/gpu/drm/msm/dsi/dsi.h +++ b/drivers/gpu/drm/msm/dsi/dsi.h @@ -100,6 +100,7 @@ bool msm_dsi_manager_cmd_xfer_trigger(int id, u32 dma_base, u32 len); void msm_dsi_manager_attach_dsi_device(int id, u32 device_flags); int msm_dsi_manager_register(struct msm_dsi *msm_dsi); void msm_dsi_manager_unregister(struct msm_dsi *msm_dsi); +bool msm_dsi_manager_validate_current_config(u8 id); /* msm dsi */ static inline bool msm_dsi_device_connected(struct msm_dsi *msm_dsi) @@ -149,6 +150,7 @@ static inline int msm_dsi_pll_set_usecase(struct msm_dsi_pll *pll, #endif /* dsi host */ +struct msm_dsi_host; int msm_dsi_host_xfer_prepare(struct mipi_dsi_host *host, const struct mipi_dsi_msg *msg); void msm_dsi_host_xfer_restore(struct mipi_dsi_host *host, @@ -162,7 +164,8 @@ void msm_dsi_host_cmd_xfer_commit(struct mipi_dsi_host *host, int msm_dsi_host_enable(struct mipi_dsi_host *host); int msm_dsi_host_disable(struct mipi_dsi_host *host); int msm_dsi_host_power_on(struct mipi_dsi_host *host, - struct msm_dsi_phy_shared_timings *phy_shared_timings); + struct msm_dsi_phy_shared_timings *phy_shared_timings, + bool is_dual_dsi); int msm_dsi_host_power_off(struct mipi_dsi_host *host); int msm_dsi_host_set_display_mode(struct mipi_dsi_host *host, struct drm_display_mode *mode); @@ -175,13 +178,29 @@ int msm_dsi_host_set_src_pll(struct mipi_dsi_host *host, struct msm_dsi_pll *src_pll); void msm_dsi_host_reset_phy(struct mipi_dsi_host *host); void msm_dsi_host_get_phy_clk_req(struct mipi_dsi_host *host, - struct msm_dsi_phy_clk_request *clk_req); + struct msm_dsi_phy_clk_request *clk_req, + bool is_dual_dsi); void msm_dsi_host_destroy(struct mipi_dsi_host *host); int msm_dsi_host_modeset_init(struct mipi_dsi_host *host, struct drm_device *dev); int msm_dsi_host_init(struct msm_dsi *msm_dsi); int msm_dsi_runtime_suspend(struct device *dev); int msm_dsi_runtime_resume(struct device *dev); +int dsi_link_clk_enable_6g(struct msm_dsi_host *msm_host); +int dsi_link_clk_enable_v2(struct msm_dsi_host *msm_host); +void dsi_link_clk_disable_6g(struct msm_dsi_host *msm_host); +void dsi_link_clk_disable_v2(struct msm_dsi_host *msm_host); +int dsi_tx_buf_alloc_6g(struct msm_dsi_host *msm_host, int size); +int dsi_tx_buf_alloc_v2(struct msm_dsi_host *msm_host, int size); +void *dsi_tx_buf_get_6g(struct msm_dsi_host *msm_host); +void *dsi_tx_buf_get_v2(struct msm_dsi_host *msm_host); +void dsi_tx_buf_put_6g(struct msm_dsi_host *msm_host); +int dsi_dma_base_get_6g(struct msm_dsi_host *msm_host, uint64_t *iova); +int dsi_dma_base_get_v2(struct msm_dsi_host *msm_host, uint64_t *iova); +int dsi_clk_init_v2(struct msm_dsi_host *msm_host); +int dsi_clk_init_6g_v2(struct msm_dsi_host *msm_host); +int dsi_calc_clk_rate_v2(struct msm_dsi_host *msm_host, bool is_dual_dsi); +int dsi_calc_clk_rate_6g(struct msm_dsi_host *msm_host, bool is_dual_dsi); /* dsi phy */ struct msm_dsi_phy; diff --git a/drivers/gpu/drm/msm/dsi/dsi_cfg.c b/drivers/gpu/drm/msm/dsi/dsi_cfg.c index 0327bb54b01b..dcdfb1bb54f9 100644 --- a/drivers/gpu/drm/msm/dsi/dsi_cfg.c +++ b/drivers/gpu/drm/msm/dsi/dsi_cfg.c @@ -136,20 +136,58 @@ static const struct msm_dsi_config sdm845_dsi_cfg = { .num_dsi = 2, }; +const static struct msm_dsi_host_cfg_ops msm_dsi_v2_host_ops = { + .link_clk_enable = dsi_link_clk_enable_v2, + .link_clk_disable = dsi_link_clk_disable_v2, + .clk_init_ver = dsi_clk_init_v2, + .tx_buf_alloc = dsi_tx_buf_alloc_v2, + .tx_buf_get = dsi_tx_buf_get_v2, + .tx_buf_put = NULL, + .dma_base_get = dsi_dma_base_get_v2, + .calc_clk_rate = dsi_calc_clk_rate_v2, +}; + +const static struct msm_dsi_host_cfg_ops msm_dsi_6g_host_ops = { + .link_clk_enable = dsi_link_clk_enable_6g, + .link_clk_disable = dsi_link_clk_disable_6g, + .clk_init_ver = NULL, + .tx_buf_alloc = dsi_tx_buf_alloc_6g, + .tx_buf_get = dsi_tx_buf_get_6g, + .tx_buf_put = dsi_tx_buf_put_6g, + .dma_base_get = dsi_dma_base_get_6g, + .calc_clk_rate = dsi_calc_clk_rate_6g, +}; + +const static struct msm_dsi_host_cfg_ops msm_dsi_6g_v2_host_ops = { + .link_clk_enable = dsi_link_clk_enable_6g, + .link_clk_disable = dsi_link_clk_disable_6g, + .clk_init_ver = dsi_clk_init_6g_v2, + .tx_buf_alloc = dsi_tx_buf_alloc_6g, + .tx_buf_get = dsi_tx_buf_get_6g, + .tx_buf_put = dsi_tx_buf_put_6g, + .dma_base_get = dsi_dma_base_get_6g, + .calc_clk_rate = dsi_calc_clk_rate_6g, +}; + static const struct msm_dsi_cfg_handler dsi_cfg_handlers[] = { - {MSM_DSI_VER_MAJOR_V2, MSM_DSI_V2_VER_MINOR_8064, &apq8064_dsi_cfg}, + {MSM_DSI_VER_MAJOR_V2, MSM_DSI_V2_VER_MINOR_8064, + &apq8064_dsi_cfg, &msm_dsi_v2_host_ops}, {MSM_DSI_VER_MAJOR_6G, MSM_DSI_6G_VER_MINOR_V1_0, - &msm8974_apq8084_dsi_cfg}, + &msm8974_apq8084_dsi_cfg, &msm_dsi_6g_host_ops}, {MSM_DSI_VER_MAJOR_6G, MSM_DSI_6G_VER_MINOR_V1_1, - &msm8974_apq8084_dsi_cfg}, + &msm8974_apq8084_dsi_cfg, &msm_dsi_6g_host_ops}, {MSM_DSI_VER_MAJOR_6G, MSM_DSI_6G_VER_MINOR_V1_1_1, - &msm8974_apq8084_dsi_cfg}, + &msm8974_apq8084_dsi_cfg, &msm_dsi_6g_host_ops}, {MSM_DSI_VER_MAJOR_6G, MSM_DSI_6G_VER_MINOR_V1_2, - &msm8974_apq8084_dsi_cfg}, - {MSM_DSI_VER_MAJOR_6G, MSM_DSI_6G_VER_MINOR_V1_3, &msm8994_dsi_cfg}, - {MSM_DSI_VER_MAJOR_6G, MSM_DSI_6G_VER_MINOR_V1_3_1, &msm8916_dsi_cfg}, - {MSM_DSI_VER_MAJOR_6G, MSM_DSI_6G_VER_MINOR_V1_4_1, &msm8996_dsi_cfg}, - {MSM_DSI_VER_MAJOR_6G, MSM_DSI_6G_VER_MINOR_V2_2_1, &sdm845_dsi_cfg}, + &msm8974_apq8084_dsi_cfg, &msm_dsi_6g_host_ops}, + {MSM_DSI_VER_MAJOR_6G, MSM_DSI_6G_VER_MINOR_V1_3, + &msm8994_dsi_cfg, &msm_dsi_6g_host_ops}, + {MSM_DSI_VER_MAJOR_6G, MSM_DSI_6G_VER_MINOR_V1_3_1, + &msm8916_dsi_cfg, &msm_dsi_6g_host_ops}, + {MSM_DSI_VER_MAJOR_6G, MSM_DSI_6G_VER_MINOR_V1_4_1, + &msm8996_dsi_cfg, &msm_dsi_6g_host_ops}, + {MSM_DSI_VER_MAJOR_6G, MSM_DSI_6G_VER_MINOR_V2_2_1, + &sdm845_dsi_cfg, &msm_dsi_6g_v2_host_ops}, }; const struct msm_dsi_cfg_handler *msm_dsi_cfg_get(u32 major, u32 minor) diff --git a/drivers/gpu/drm/msm/dsi/dsi_cfg.h b/drivers/gpu/drm/msm/dsi/dsi_cfg.h index 9cfdcf1c95d5..16c507911110 100644 --- a/drivers/gpu/drm/msm/dsi/dsi_cfg.h +++ b/drivers/gpu/drm/msm/dsi/dsi_cfg.h @@ -40,10 +40,22 @@ struct msm_dsi_config { const int num_dsi; }; +struct msm_dsi_host_cfg_ops { + int (*link_clk_enable)(struct msm_dsi_host *msm_host); + void (*link_clk_disable)(struct msm_dsi_host *msm_host); + int (*clk_init_ver)(struct msm_dsi_host *msm_host); + int (*tx_buf_alloc)(struct msm_dsi_host *msm_host, int size); + void* (*tx_buf_get)(struct msm_dsi_host *msm_host); + void (*tx_buf_put)(struct msm_dsi_host *msm_host); + int (*dma_base_get)(struct msm_dsi_host *msm_host, uint64_t *iova); + int (*calc_clk_rate)(struct msm_dsi_host *msm_host, bool is_dual_dsi); +}; + struct msm_dsi_cfg_handler { u32 major; u32 minor; const struct msm_dsi_config *cfg; + const struct msm_dsi_host_cfg_ops *ops; }; const struct msm_dsi_cfg_handler *msm_dsi_cfg_get(u32 major, u32 minor); diff --git a/drivers/gpu/drm/msm/dsi/dsi_host.c b/drivers/gpu/drm/msm/dsi/dsi_host.c index 29841f440111..96fb5f635314 100644 --- a/drivers/gpu/drm/msm/dsi/dsi_host.c +++ b/drivers/gpu/drm/msm/dsi/dsi_host.c @@ -118,6 +118,7 @@ struct msm_dsi_host { struct clk *byte_intf_clk; u32 byte_clk_rate; + u32 pixel_clk_rate; u32 esc_clk_rate; /* DSI v2 specific clocks */ @@ -332,6 +333,54 @@ static int dsi_regulator_init(struct msm_dsi_host *msm_host) return 0; } +int dsi_clk_init_v2(struct msm_dsi_host *msm_host) +{ + struct platform_device *pdev = msm_host->pdev; + int ret = 0; + + msm_host->src_clk = msm_clk_get(pdev, "src"); + + if (IS_ERR(msm_host->src_clk)) { + ret = PTR_ERR(msm_host->src_clk); + pr_err("%s: can't find src clock. ret=%d\n", + __func__, ret); + msm_host->src_clk = NULL; + return ret; + } + + msm_host->esc_clk_src = clk_get_parent(msm_host->esc_clk); + if (!msm_host->esc_clk_src) { + ret = -ENODEV; + pr_err("%s: can't get esc clock parent. ret=%d\n", + __func__, ret); + return ret; + } + + msm_host->dsi_clk_src = clk_get_parent(msm_host->src_clk); + if (!msm_host->dsi_clk_src) { + ret = -ENODEV; + pr_err("%s: can't get src clock parent. ret=%d\n", + __func__, ret); + } + + return ret; +} + +int dsi_clk_init_6g_v2(struct msm_dsi_host *msm_host) +{ + struct platform_device *pdev = msm_host->pdev; + int ret = 0; + + msm_host->byte_intf_clk = msm_clk_get(pdev, "byte_intf"); + if (IS_ERR(msm_host->byte_intf_clk)) { + ret = PTR_ERR(msm_host->byte_intf_clk); + pr_err("%s: can't find byte_intf clock. ret=%d\n", + __func__, ret); + } + + return ret; +} + static int dsi_clk_init(struct msm_dsi_host *msm_host) { struct platform_device *pdev = msm_host->pdev; @@ -379,19 +428,6 @@ static int dsi_clk_init(struct msm_dsi_host *msm_host) goto exit; } - if (cfg_hnd->major == MSM_DSI_VER_MAJOR_6G && - cfg_hnd->minor >= MSM_DSI_6G_VER_MINOR_V2_2_1) { - msm_host->byte_intf_clk = msm_clk_get(pdev, "byte_intf"); - if (IS_ERR(msm_host->byte_intf_clk)) { - ret = PTR_ERR(msm_host->byte_intf_clk); - pr_err("%s: can't find byte_intf clock. ret=%d\n", - __func__, ret); - goto exit; - } - } else { - msm_host->byte_intf_clk = NULL; - } - msm_host->byte_clk_src = clk_get_parent(msm_host->byte_clk); if (!msm_host->byte_clk_src) { ret = -ENODEV; @@ -406,31 +442,8 @@ static int dsi_clk_init(struct msm_dsi_host *msm_host) goto exit; } - if (cfg_hnd->major == MSM_DSI_VER_MAJOR_V2) { - msm_host->src_clk = msm_clk_get(pdev, "src"); - if (IS_ERR(msm_host->src_clk)) { - ret = PTR_ERR(msm_host->src_clk); - pr_err("%s: can't find src clock. ret=%d\n", - __func__, ret); - msm_host->src_clk = NULL; - goto exit; - } - - msm_host->esc_clk_src = clk_get_parent(msm_host->esc_clk); - if (!msm_host->esc_clk_src) { - ret = -ENODEV; - pr_err("%s: can't get esc clock parent. ret=%d\n", - __func__, ret); - goto exit; - } - - msm_host->dsi_clk_src = clk_get_parent(msm_host->src_clk); - if (!msm_host->dsi_clk_src) { - ret = -ENODEV; - pr_err("%s: can't get src clock parent. ret=%d\n", - __func__, ret); - } - } + if (cfg_hnd->ops->clk_init_ver) + ret = cfg_hnd->ops->clk_init_ver(msm_host); exit: return ret; } @@ -498,7 +511,7 @@ int msm_dsi_runtime_resume(struct device *dev) return dsi_bus_clk_enable(msm_host); } -static int dsi_link_clk_enable_6g(struct msm_dsi_host *msm_host) +int dsi_link_clk_enable_6g(struct msm_dsi_host *msm_host) { int ret; @@ -511,7 +524,7 @@ static int dsi_link_clk_enable_6g(struct msm_dsi_host *msm_host) goto error; } - ret = clk_set_rate(msm_host->pixel_clk, msm_host->mode->clock * 1000); + ret = clk_set_rate(msm_host->pixel_clk, msm_host->pixel_clk_rate); if (ret) { pr_err("%s: Failed to set rate pixel clk, %d\n", __func__, ret); goto error; @@ -566,7 +579,7 @@ error: return ret; } -static int dsi_link_clk_enable_v2(struct msm_dsi_host *msm_host) +int dsi_link_clk_enable_v2(struct msm_dsi_host *msm_host) { int ret; @@ -592,7 +605,7 @@ static int dsi_link_clk_enable_v2(struct msm_dsi_host *msm_host) goto error; } - ret = clk_set_rate(msm_host->pixel_clk, msm_host->mode->clock * 1000); + ret = clk_set_rate(msm_host->pixel_clk, msm_host->pixel_clk_rate); if (ret) { pr_err("%s: Failed to set rate pixel clk, %d\n", __func__, ret); goto error; @@ -634,98 +647,121 @@ error: return ret; } -static int dsi_link_clk_enable(struct msm_dsi_host *msm_host) +void dsi_link_clk_disable_6g(struct msm_dsi_host *msm_host) { - const struct msm_dsi_cfg_handler *cfg_hnd = msm_host->cfg_hnd; + clk_disable_unprepare(msm_host->esc_clk); + clk_disable_unprepare(msm_host->pixel_clk); + if (msm_host->byte_intf_clk) + clk_disable_unprepare(msm_host->byte_intf_clk); + clk_disable_unprepare(msm_host->byte_clk); +} - if (cfg_hnd->major == MSM_DSI_VER_MAJOR_6G) - return dsi_link_clk_enable_6g(msm_host); - else - return dsi_link_clk_enable_v2(msm_host); +void dsi_link_clk_disable_v2(struct msm_dsi_host *msm_host) +{ + clk_disable_unprepare(msm_host->pixel_clk); + clk_disable_unprepare(msm_host->src_clk); + clk_disable_unprepare(msm_host->esc_clk); + clk_disable_unprepare(msm_host->byte_clk); } -static void dsi_link_clk_disable(struct msm_dsi_host *msm_host) +static u32 dsi_get_pclk_rate(struct msm_dsi_host *msm_host, bool is_dual_dsi) { - const struct msm_dsi_cfg_handler *cfg_hnd = msm_host->cfg_hnd; + struct drm_display_mode *mode = msm_host->mode; + u32 pclk_rate; - if (cfg_hnd->major == MSM_DSI_VER_MAJOR_6G) { - clk_disable_unprepare(msm_host->esc_clk); - clk_disable_unprepare(msm_host->pixel_clk); - if (msm_host->byte_intf_clk) - clk_disable_unprepare(msm_host->byte_intf_clk); - clk_disable_unprepare(msm_host->byte_clk); - } else { - clk_disable_unprepare(msm_host->pixel_clk); - clk_disable_unprepare(msm_host->src_clk); - clk_disable_unprepare(msm_host->esc_clk); - clk_disable_unprepare(msm_host->byte_clk); - } + pclk_rate = mode->clock * 1000; + + /* + * For dual DSI mode, the current DRM mode has the complete width of the + * panel. Since, the complete panel is driven by two DSI controllers, + * the clock rates have to be split between the two dsi controllers. + * Adjust the byte and pixel clock rates for each dsi host accordingly. + */ + if (is_dual_dsi) + pclk_rate /= 2; + + return pclk_rate; } -static int dsi_calc_clk_rate(struct msm_dsi_host *msm_host) +static void dsi_calc_pclk(struct msm_dsi_host *msm_host, bool is_dual_dsi) { - struct drm_display_mode *mode = msm_host->mode; - const struct msm_dsi_cfg_handler *cfg_hnd = msm_host->cfg_hnd; u8 lanes = msm_host->lanes; u32 bpp = dsi_get_bpp(msm_host->format); - u32 pclk_rate; + u32 pclk_rate = dsi_get_pclk_rate(msm_host, is_dual_dsi); + u64 pclk_bpp = (u64)pclk_rate * bpp; - if (!mode) { - pr_err("%s: mode not set\n", __func__); - return -EINVAL; - } - - pclk_rate = mode->clock * 1000; - if (lanes > 0) { - msm_host->byte_clk_rate = (pclk_rate * bpp) / (8 * lanes); - } else { + if (lanes == 0) { pr_err("%s: forcing mdss_dsi lanes to 1\n", __func__); - msm_host->byte_clk_rate = (pclk_rate * bpp) / 8; + lanes = 1; } - DBG("pclk=%d, bclk=%d", pclk_rate, msm_host->byte_clk_rate); + do_div(pclk_bpp, (8 * lanes)); - msm_host->esc_clk_rate = clk_get_rate(msm_host->esc_clk); + msm_host->pixel_clk_rate = pclk_rate; + msm_host->byte_clk_rate = pclk_bpp; - if (cfg_hnd->major == MSM_DSI_VER_MAJOR_V2) { - unsigned int esc_mhz, esc_div; - unsigned long byte_mhz; + DBG("pclk=%d, bclk=%d", msm_host->pixel_clk_rate, + msm_host->byte_clk_rate); - msm_host->src_clk_rate = (pclk_rate * bpp) / 8; +} - /* - * esc clock is byte clock followed by a 4 bit divider, - * we need to find an escape clock frequency within the - * mipi DSI spec range within the maximum divider limit - * We iterate here between an escape clock frequencey - * between 20 Mhz to 5 Mhz and pick up the first one - * that can be supported by our divider - */ +int dsi_calc_clk_rate_6g(struct msm_dsi_host *msm_host, bool is_dual_dsi) +{ + if (!msm_host->mode) { + pr_err("%s: mode not set\n", __func__); + return -EINVAL; + } + + dsi_calc_pclk(msm_host, is_dual_dsi); + msm_host->esc_clk_rate = clk_get_rate(msm_host->esc_clk); + return 0; +} - byte_mhz = msm_host->byte_clk_rate / 1000000; +int dsi_calc_clk_rate_v2(struct msm_dsi_host *msm_host, bool is_dual_dsi) +{ + u32 bpp = dsi_get_bpp(msm_host->format); + u64 pclk_bpp; + unsigned int esc_mhz, esc_div; + unsigned long byte_mhz; - for (esc_mhz = 20; esc_mhz >= 5; esc_mhz--) { - esc_div = DIV_ROUND_UP(byte_mhz, esc_mhz); + dsi_calc_pclk(msm_host, is_dual_dsi); - /* - * TODO: Ideally, we shouldn't know what sort of divider - * is available in mmss_cc, we're just assuming that - * it'll always be a 4 bit divider. Need to come up with - * a better way here. - */ - if (esc_div >= 1 && esc_div <= 16) - break; - } + pclk_bpp = (u64)dsi_get_pclk_rate(msm_host, is_dual_dsi) * bpp; + do_div(pclk_bpp, 8); + msm_host->src_clk_rate = pclk_bpp; - if (esc_mhz < 5) - return -EINVAL; + /* + * esc clock is byte clock followed by a 4 bit divider, + * we need to find an escape clock frequency within the + * mipi DSI spec range within the maximum divider limit + * We iterate here between an escape clock frequencey + * between 20 Mhz to 5 Mhz and pick up the first one + * that can be supported by our divider + */ + + byte_mhz = msm_host->byte_clk_rate / 1000000; - msm_host->esc_clk_rate = msm_host->byte_clk_rate / esc_div; + for (esc_mhz = 20; esc_mhz >= 5; esc_mhz--) { + esc_div = DIV_ROUND_UP(byte_mhz, esc_mhz); - DBG("esc=%d, src=%d", msm_host->esc_clk_rate, - msm_host->src_clk_rate); + /* + * TODO: Ideally, we shouldn't know what sort of divider + * is available in mmss_cc, we're just assuming that + * it'll always be a 4 bit divider. Need to come up with + * a better way here. + */ + if (esc_div >= 1 && esc_div <= 16) + break; } + if (esc_mhz < 5) + return -EINVAL; + + msm_host->esc_clk_rate = msm_host->byte_clk_rate / esc_div; + + DBG("esc=%d, src=%d", msm_host->esc_clk_rate, + msm_host->src_clk_rate); + return 0; } @@ -885,7 +921,7 @@ static void dsi_ctrl_config(struct msm_dsi_host *msm_host, bool enable, dsi_write(msm_host, REG_DSI_CTRL, data); } -static void dsi_timing_setup(struct msm_dsi_host *msm_host) +static void dsi_timing_setup(struct msm_dsi_host *msm_host, bool is_dual_dsi) { struct drm_display_mode *mode = msm_host->mode; u32 hs_start = 0, vs_start = 0; /* take sync start as 0 */ @@ -897,10 +933,26 @@ static void dsi_timing_setup(struct msm_dsi_host *msm_host) u32 ha_end = ha_start + mode->hdisplay; u32 va_start = v_total - mode->vsync_start; u32 va_end = va_start + mode->vdisplay; + u32 hdisplay = mode->hdisplay; u32 wc; DBG(""); + /* + * For dual DSI mode, the current DRM mode has + * the complete width of the panel. Since, the complete + * panel is driven by two DSI controllers, the horizontal + * timings have to be split between the two dsi controllers. + * Adjust the DSI host timing values accordingly. + */ + if (is_dual_dsi) { + h_total /= 2; + hs_end /= 2; + ha_start /= 2; + ha_end /= 2; + hdisplay /= 2; + } + if (msm_host->mode_flags & MIPI_DSI_MODE_VIDEO) { dsi_write(msm_host, REG_DSI_ACTIVE_H, DSI_ACTIVE_H_START(ha_start) | @@ -921,7 +973,7 @@ static void dsi_timing_setup(struct msm_dsi_host *msm_host) DSI_ACTIVE_VSYNC_VPOS_END(vs_end)); } else { /* command mode */ /* image data and 1 byte write_memory_start cmd */ - wc = mode->hdisplay * dsi_get_bpp(msm_host->format) / 8 + 1; + wc = hdisplay * dsi_get_bpp(msm_host->format) / 8 + 1; dsi_write(msm_host, REG_DSI_CMD_MDP_STREAM_CTRL, DSI_CMD_MDP_STREAM_CTRL_WORD_COUNT(wc) | @@ -931,7 +983,7 @@ static void dsi_timing_setup(struct msm_dsi_host *msm_host) MIPI_DSI_DCS_LONG_WRITE)); dsi_write(msm_host, REG_DSI_CMD_MDP_STREAM_TOTAL, - DSI_CMD_MDP_STREAM_TOTAL_H_TOTAL(mode->hdisplay) | + DSI_CMD_MDP_STREAM_TOTAL_H_TOTAL(hdisplay) | DSI_CMD_MDP_STREAM_TOTAL_V_TOTAL(mode->vdisplay)); } } @@ -1015,50 +1067,37 @@ static void dsi_wait4video_eng_busy(struct msm_dsi_host *msm_host) } } -/* dsi_cmd */ -static int dsi_tx_buf_alloc(struct msm_dsi_host *msm_host, int size) +int dsi_tx_buf_alloc_6g(struct msm_dsi_host *msm_host, int size) { struct drm_device *dev = msm_host->dev; struct msm_drm_private *priv = dev->dev_private; - const struct msm_dsi_cfg_handler *cfg_hnd = msm_host->cfg_hnd; - int ret; uint64_t iova; + u8 *data; - if (cfg_hnd->major == MSM_DSI_VER_MAJOR_6G) { - msm_host->tx_gem_obj = msm_gem_new(dev, size, MSM_BO_UNCACHED); - if (IS_ERR(msm_host->tx_gem_obj)) { - ret = PTR_ERR(msm_host->tx_gem_obj); - pr_err("%s: failed to allocate gem, %d\n", - __func__, ret); - msm_host->tx_gem_obj = NULL; - return ret; - } + data = msm_gem_kernel_new(dev, size, MSM_BO_UNCACHED, + priv->kms->aspace, + &msm_host->tx_gem_obj, &iova); - ret = msm_gem_get_iova(msm_host->tx_gem_obj, - priv->kms->aspace, &iova); - if (ret) { - pr_err("%s: failed to get iova, %d\n", __func__, ret); - return ret; - } + if (IS_ERR(data)) { + msm_host->tx_gem_obj = NULL; + return PTR_ERR(data); + } - if (iova & 0x07) { - pr_err("%s: buf NOT 8 bytes aligned\n", __func__); - return -EINVAL; - } + msm_host->tx_size = msm_host->tx_gem_obj->size; - msm_host->tx_size = msm_host->tx_gem_obj->size; - } else { - msm_host->tx_buf = dma_alloc_coherent(dev->dev, size, + return 0; +} + +int dsi_tx_buf_alloc_v2(struct msm_dsi_host *msm_host, int size) +{ + struct drm_device *dev = msm_host->dev; + + msm_host->tx_buf = dma_alloc_coherent(dev->dev, size, &msm_host->tx_buf_paddr, GFP_KERNEL); - if (!msm_host->tx_buf) { - ret = -ENOMEM; - pr_err("%s: failed to allocate tx buf, %d\n", - __func__, ret); - return ret; - } + if (!msm_host->tx_buf) + return -ENOMEM; - msm_host->tx_size = size; - } + msm_host->tx_size = size; return 0; } @@ -1089,6 +1128,21 @@ static void dsi_tx_buf_free(struct msm_dsi_host *msm_host) msm_host->tx_buf_paddr); } +void *dsi_tx_buf_get_6g(struct msm_dsi_host *msm_host) +{ + return msm_gem_get_vaddr(msm_host->tx_gem_obj); +} + +void *dsi_tx_buf_get_v2(struct msm_dsi_host *msm_host) +{ + return msm_host->tx_buf; +} + +void dsi_tx_buf_put_6g(struct msm_dsi_host *msm_host) +{ + msm_gem_put_vaddr(msm_host->tx_gem_obj); +} + /* * prepare cmd buffer to be txed */ @@ -1113,15 +1167,11 @@ static int dsi_cmd_dma_add(struct msm_dsi_host *msm_host, return -EINVAL; } - if (cfg_hnd->major == MSM_DSI_VER_MAJOR_6G) { - data = msm_gem_get_vaddr(msm_host->tx_gem_obj); - if (IS_ERR(data)) { - ret = PTR_ERR(data); - pr_err("%s: get vaddr failed, %d\n", __func__, ret); - return ret; - } - } else { - data = msm_host->tx_buf; + data = cfg_hnd->ops->tx_buf_get(msm_host); + if (IS_ERR(data)) { + ret = PTR_ERR(data); + pr_err("%s: get vaddr failed, %d\n", __func__, ret); + return ret; } /* MSM specific command format in memory */ @@ -1142,8 +1192,8 @@ static int dsi_cmd_dma_add(struct msm_dsi_host *msm_host, if (packet.size < len) memset(data + packet.size, 0xff, len - packet.size); - if (cfg_hnd->major == MSM_DSI_VER_MAJOR_6G) - msm_gem_put_vaddr(msm_host->tx_gem_obj); + if (cfg_hnd->ops->tx_buf_put) + cfg_hnd->ops->tx_buf_put(msm_host); return len; } @@ -1190,24 +1240,38 @@ static int dsi_long_read_resp(u8 *buf, const struct mipi_dsi_msg *msg) return msg->rx_len; } -static int dsi_cmd_dma_tx(struct msm_dsi_host *msm_host, int len) +int dsi_dma_base_get_6g(struct msm_dsi_host *msm_host, uint64_t *dma_base) { - const struct msm_dsi_cfg_handler *cfg_hnd = msm_host->cfg_hnd; struct drm_device *dev = msm_host->dev; struct msm_drm_private *priv = dev->dev_private; + + if (!dma_base) + return -EINVAL; + + return msm_gem_get_iova(msm_host->tx_gem_obj, + priv->kms->aspace, dma_base); +} + +int dsi_dma_base_get_v2(struct msm_dsi_host *msm_host, uint64_t *dma_base) +{ + if (!dma_base) + return -EINVAL; + + *dma_base = msm_host->tx_buf_paddr; + return 0; +} + +static int dsi_cmd_dma_tx(struct msm_dsi_host *msm_host, int len) +{ + const struct msm_dsi_cfg_handler *cfg_hnd = msm_host->cfg_hnd; int ret; uint64_t dma_base; bool triggered; - if (cfg_hnd->major == MSM_DSI_VER_MAJOR_6G) { - ret = msm_gem_get_iova(msm_host->tx_gem_obj, - priv->kms->aspace, &dma_base); - if (ret) { - pr_err("%s: failed to get iova: %d\n", __func__, ret); - return ret; - } - } else { - dma_base = msm_host->tx_buf_paddr; + ret = cfg_hnd->ops->dma_base_get(msm_host, &dma_base); + if (ret) { + pr_err("%s: failed to get iova: %d\n", __func__, ret); + return ret; } reinit_completion(&msm_host->dma_comp); @@ -1845,6 +1909,7 @@ int msm_dsi_host_modeset_init(struct mipi_dsi_host *host, struct drm_device *dev) { struct msm_dsi_host *msm_host = to_msm_dsi_host(host); + const struct msm_dsi_cfg_handler *cfg_hnd = msm_host->cfg_hnd; struct platform_device *pdev = msm_host->pdev; int ret; @@ -1865,7 +1930,7 @@ int msm_dsi_host_modeset_init(struct mipi_dsi_host *host, } msm_host->dev = dev; - ret = dsi_tx_buf_alloc(msm_host, SZ_4K); + ret = cfg_hnd->ops->tx_buf_alloc(msm_host, SZ_4K); if (ret) { pr_err("%s: alloc tx gem obj failed, %d\n", __func__, ret); return ret; @@ -1923,6 +1988,7 @@ int msm_dsi_host_xfer_prepare(struct mipi_dsi_host *host, const struct mipi_dsi_msg *msg) { struct msm_dsi_host *msm_host = to_msm_dsi_host(host); + const struct msm_dsi_cfg_handler *cfg_hnd = msm_host->cfg_hnd; /* TODO: make sure dsi_cmd_mdp is idle. * Since DSI6G v1.2.0, we can set DSI_TRIG_CTRL.BLOCK_DMA_WITHIN_FRAME @@ -1935,7 +2001,7 @@ int msm_dsi_host_xfer_prepare(struct mipi_dsi_host *host, * mdp clock need to be enabled to receive dsi interrupt */ pm_runtime_get_sync(&msm_host->pdev->dev); - dsi_link_clk_enable(msm_host); + cfg_hnd->ops->link_clk_enable(msm_host); /* TODO: vote for bus bandwidth */ @@ -1956,6 +2022,7 @@ void msm_dsi_host_xfer_restore(struct mipi_dsi_host *host, const struct mipi_dsi_msg *msg) { struct msm_dsi_host *msm_host = to_msm_dsi_host(host); + const struct msm_dsi_cfg_handler *cfg_hnd = msm_host->cfg_hnd; dsi_intr_ctrl(msm_host, DSI_IRQ_MASK_CMD_DMA_DONE, 0); dsi_write(msm_host, REG_DSI_CTRL, msm_host->dma_cmd_ctrl_restore); @@ -1965,7 +2032,7 @@ void msm_dsi_host_xfer_restore(struct mipi_dsi_host *host, /* TODO: unvote for bus bandwidth */ - dsi_link_clk_disable(msm_host); + cfg_hnd->ops->link_clk_disable(msm_host); pm_runtime_put_autosuspend(&msm_host->pdev->dev); } @@ -2129,7 +2196,6 @@ int msm_dsi_host_set_src_pll(struct mipi_dsi_host *host, struct msm_dsi_pll *src_pll) { struct msm_dsi_host *msm_host = to_msm_dsi_host(host); - const struct msm_dsi_cfg_handler *cfg_hnd = msm_host->cfg_hnd; struct clk *byte_clk_provider, *pixel_clk_provider; int ret; @@ -2155,14 +2221,16 @@ int msm_dsi_host_set_src_pll(struct mipi_dsi_host *host, goto exit; } - if (cfg_hnd->major == MSM_DSI_VER_MAJOR_V2) { + if (msm_host->dsi_clk_src) { ret = clk_set_parent(msm_host->dsi_clk_src, pixel_clk_provider); if (ret) { pr_err("%s: can't set parent to dsi_clk_src. ret=%d\n", __func__, ret); goto exit; } + } + if (msm_host->esc_clk_src) { ret = clk_set_parent(msm_host->esc_clk_src, byte_clk_provider); if (ret) { pr_err("%s: can't set parent to esc_clk_src. ret=%d\n", @@ -2189,12 +2257,14 @@ void msm_dsi_host_reset_phy(struct mipi_dsi_host *host) } void msm_dsi_host_get_phy_clk_req(struct mipi_dsi_host *host, - struct msm_dsi_phy_clk_request *clk_req) + struct msm_dsi_phy_clk_request *clk_req, + bool is_dual_dsi) { struct msm_dsi_host *msm_host = to_msm_dsi_host(host); + const struct msm_dsi_cfg_handler *cfg_hnd = msm_host->cfg_hnd; int ret; - ret = dsi_calc_clk_rate(msm_host); + ret = cfg_hnd->ops->calc_clk_rate(msm_host, is_dual_dsi); if (ret) { pr_err("%s: unable to calc clk rate, %d\n", __func__, ret); return; @@ -2256,9 +2326,11 @@ static void msm_dsi_sfpb_config(struct msm_dsi_host *msm_host, bool enable) } int msm_dsi_host_power_on(struct mipi_dsi_host *host, - struct msm_dsi_phy_shared_timings *phy_shared_timings) + struct msm_dsi_phy_shared_timings *phy_shared_timings, + bool is_dual_dsi) { struct msm_dsi_host *msm_host = to_msm_dsi_host(host); + const struct msm_dsi_cfg_handler *cfg_hnd = msm_host->cfg_hnd; int ret = 0; mutex_lock(&msm_host->dev_mutex); @@ -2277,7 +2349,7 @@ int msm_dsi_host_power_on(struct mipi_dsi_host *host, } pm_runtime_get_sync(&msm_host->pdev->dev); - ret = dsi_link_clk_enable(msm_host); + ret = cfg_hnd->ops->link_clk_enable(msm_host); if (ret) { pr_err("%s: failed to enable link clocks. ret=%d\n", __func__, ret); @@ -2291,7 +2363,7 @@ int msm_dsi_host_power_on(struct mipi_dsi_host *host, goto fail_disable_clk; } - dsi_timing_setup(msm_host); + dsi_timing_setup(msm_host, is_dual_dsi); dsi_sw_reset(msm_host); dsi_ctrl_config(msm_host, true, phy_shared_timings); @@ -2304,7 +2376,7 @@ int msm_dsi_host_power_on(struct mipi_dsi_host *host, return 0; fail_disable_clk: - dsi_link_clk_disable(msm_host); + cfg_hnd->ops->link_clk_disable(msm_host); pm_runtime_put_autosuspend(&msm_host->pdev->dev); fail_disable_reg: dsi_host_regulator_disable(msm_host); @@ -2316,6 +2388,7 @@ unlock_ret: int msm_dsi_host_power_off(struct mipi_dsi_host *host) { struct msm_dsi_host *msm_host = to_msm_dsi_host(host); + const struct msm_dsi_cfg_handler *cfg_hnd = msm_host->cfg_hnd; mutex_lock(&msm_host->dev_mutex); if (!msm_host->power_on) { @@ -2330,7 +2403,7 @@ int msm_dsi_host_power_off(struct mipi_dsi_host *host) pinctrl_pm_select_sleep_state(&msm_host->pdev->dev); - dsi_link_clk_disable(msm_host); + cfg_hnd->ops->link_clk_disable(msm_host); pm_runtime_put_autosuspend(&msm_host->pdev->dev); dsi_host_regulator_disable(msm_host); diff --git a/drivers/gpu/drm/msm/dsi/dsi_manager.c b/drivers/gpu/drm/msm/dsi/dsi_manager.c index d5006d6923e0..5224010d90e4 100644 --- a/drivers/gpu/drm/msm/dsi/dsi_manager.c +++ b/drivers/gpu/drm/msm/dsi/dsi_manager.c @@ -134,8 +134,9 @@ static int enable_phy(struct msm_dsi *msm_dsi, int src_pll_id, { struct msm_dsi_phy_clk_request clk_req; int ret; + bool is_dual_dsi = IS_DUAL_DSI(); - msm_dsi_host_get_phy_clk_req(msm_dsi->host, &clk_req); + msm_dsi_host_get_phy_clk_req(msm_dsi->host, &clk_req, is_dual_dsi); ret = msm_dsi_phy_enable(msm_dsi->phy, src_pll_id, &clk_req); msm_dsi_phy_get_shared_timings(msm_dsi->phy, shared_timings); @@ -305,102 +306,25 @@ static void dsi_mgr_connector_destroy(struct drm_connector *connector) kfree(dsi_connector); } -static void dsi_dual_connector_fix_modes(struct drm_connector *connector) -{ - struct drm_display_mode *mode, *m; - - /* Only support left-right mode */ - list_for_each_entry_safe(mode, m, &connector->probed_modes, head) { - mode->clock >>= 1; - mode->hdisplay >>= 1; - mode->hsync_start >>= 1; - mode->hsync_end >>= 1; - mode->htotal >>= 1; - drm_mode_set_name(mode); - } -} - -static int dsi_dual_connector_tile_init( - struct drm_connector *connector, int id) -{ - struct drm_display_mode *mode; - /* Fake topology id */ - char topo_id[8] = {'M', 'S', 'M', 'D', 'U', 'D', 'S', 'I'}; - - if (connector->tile_group) { - DBG("Tile property has been initialized"); - return 0; - } - - /* Use the first mode only for now */ - mode = list_first_entry(&connector->probed_modes, - struct drm_display_mode, - head); - if (!mode) - return -EINVAL; - - connector->tile_group = drm_mode_get_tile_group( - connector->dev, topo_id); - if (!connector->tile_group) - connector->tile_group = drm_mode_create_tile_group( - connector->dev, topo_id); - if (!connector->tile_group) { - pr_err("%s: failed to create tile group\n", __func__); - return -ENOMEM; - } - - connector->has_tile = true; - connector->tile_is_single_monitor = true; - - /* mode has been fixed */ - connector->tile_h_size = mode->hdisplay; - connector->tile_v_size = mode->vdisplay; - - /* Only support left-right mode */ - connector->num_h_tile = 2; - connector->num_v_tile = 1; - - connector->tile_v_loc = 0; - connector->tile_h_loc = (id == DSI_RIGHT) ? 1 : 0; - - return 0; -} - static int dsi_mgr_connector_get_modes(struct drm_connector *connector) { int id = dsi_mgr_connector_get_id(connector); struct msm_dsi *msm_dsi = dsi_mgr_get_dsi(id); struct drm_panel *panel = msm_dsi->panel; - int ret, num; + int num; if (!panel) return 0; - /* Since we have 2 connectors, but only 1 drm_panel in dual DSI mode, - * panel should not attach to any connector. - * Only temporarily attach panel to the current connector here, - * to let panel set mode to this connector. + /* + * In dual DSI mode, we have one connector that can be + * attached to the drm_panel. */ drm_panel_attach(panel, connector); num = drm_panel_get_modes(panel); - drm_panel_detach(panel); if (!num) return 0; - if (IS_DUAL_DSI()) { - /* report half resolution to user */ - dsi_dual_connector_fix_modes(connector); - ret = dsi_dual_connector_tile_init(connector, id); - if (ret) - return ret; - ret = drm_connector_set_tile_property(connector); - if (ret) { - pr_err("%s: set tile property failed, %d\n", - __func__, ret); - return ret; - } - } - return num; } @@ -454,11 +378,11 @@ static void dsi_mgr_bridge_pre_enable(struct drm_bridge *bridge) if (ret) goto phy_en_fail; - /* Do nothing with the host if it is DSI 1 in case of dual DSI */ - if (is_dual_dsi && (DSI_1 == id)) + /* Do nothing with the host if it is slave-DSI in case of dual DSI */ + if (is_dual_dsi && !IS_MASTER_DSI_LINK(id)) return; - ret = msm_dsi_host_power_on(host, &phy_shared_timings[id]); + ret = msm_dsi_host_power_on(host, &phy_shared_timings[id], is_dual_dsi); if (ret) { pr_err("%s: power on host %d failed, %d\n", __func__, id, ret); goto host_on_fail; @@ -466,7 +390,7 @@ static void dsi_mgr_bridge_pre_enable(struct drm_bridge *bridge) if (is_dual_dsi && msm_dsi1) { ret = msm_dsi_host_power_on(msm_dsi1->host, - &phy_shared_timings[DSI_1]); + &phy_shared_timings[DSI_1], is_dual_dsi); if (ret) { pr_err("%s: power on host1 failed, %d\n", __func__, ret); @@ -556,11 +480,11 @@ static void dsi_mgr_bridge_post_disable(struct drm_bridge *bridge) return; /* - * Do nothing with the host if it is DSI 1 in case of dual DSI. + * Do nothing with the host if it is slave-DSI in case of dual DSI. * It is safe to call dsi_mgr_phy_disable() here because a single PHY * won't be diabled until both PHYs request disable. */ - if (is_dual_dsi && (DSI_1 == id)) + if (is_dual_dsi && !IS_MASTER_DSI_LINK(id)) goto disable_phy; if (panel) { @@ -621,7 +545,7 @@ static void dsi_mgr_bridge_mode_set(struct drm_bridge *bridge, mode->vsync_end, mode->vtotal, mode->type, mode->flags); - if (is_dual_dsi && (DSI_1 == id)) + if (is_dual_dsi && !IS_MASTER_DSI_LINK(id)) return; msm_dsi_host_set_display_mode(host, adjusted_mode); @@ -689,6 +613,23 @@ struct drm_connector *msm_dsi_manager_connector_init(u8 id) return connector; } +bool msm_dsi_manager_validate_current_config(u8 id) +{ + bool is_dual_dsi = IS_DUAL_DSI(); + + /* + * For dual DSI, we only have one drm panel. For this + * use case, we register only one bridge/connector. + * Skip bridge/connector initialisation if it is + * slave-DSI for dual DSI configuration. + */ + if (is_dual_dsi && !IS_MASTER_DSI_LINK(id)) { + DBG("Skip bridge registration for slave DSI->id: %d\n", id); + return false; + } + return true; +} + /* initialize bridge */ struct drm_bridge *msm_dsi_manager_bridge_init(u8 id) { @@ -832,6 +773,7 @@ void msm_dsi_manager_attach_dsi_device(int id, u32 device_flags) struct msm_drm_private *priv; struct msm_kms *kms; struct drm_encoder *encoder; + bool cmd_mode; /* * drm_device pointer is assigned to msm_dsi only in the modeset_init @@ -846,10 +788,11 @@ void msm_dsi_manager_attach_dsi_device(int id, u32 device_flags) priv = dev->dev_private; kms = priv->kms; encoder = msm_dsi_get_encoder(msm_dsi); + cmd_mode = !(device_flags & + MIPI_DSI_MODE_VIDEO); if (encoder && kms->funcs->set_encoder_mode) - if (!(device_flags & MIPI_DSI_MODE_VIDEO)) - kms->funcs->set_encoder_mode(kms, encoder, true); + kms->funcs->set_encoder_mode(kms, encoder, cmd_mode); } int msm_dsi_manager_register(struct msm_dsi *msm_dsi) diff --git a/drivers/gpu/drm/msm/dsi/pll/dsi_pll_10nm.c b/drivers/gpu/drm/msm/dsi/pll/dsi_pll_10nm.c index c4c37a7df637..4c03f0b7343e 100644 --- a/drivers/gpu/drm/msm/dsi/pll/dsi_pll_10nm.c +++ b/drivers/gpu/drm/msm/dsi/pll/dsi_pll_10nm.c @@ -798,6 +798,8 @@ struct msm_dsi_pll *msm_dsi_pll_10nm_init(struct platform_device *pdev, int id) return ERR_PTR(-ENOMEM); } + spin_lock_init(&pll_10nm->postdiv_lock); + pll = &pll_10nm->base; pll->min_rate = 1000000000UL; pll->max_rate = 3500000000UL; diff --git a/drivers/gpu/drm/msm/msm_atomic.c b/drivers/gpu/drm/msm/msm_atomic.c index f0635c3da7f4..c1f1779c980f 100644 --- a/drivers/gpu/drm/msm/msm_atomic.c +++ b/drivers/gpu/drm/msm/msm_atomic.c @@ -71,12 +71,15 @@ void msm_atomic_commit_tail(struct drm_atomic_state *state) drm_atomic_helper_commit_modeset_enables(dev, state); + if (kms->funcs->commit) { + DRM_DEBUG_ATOMIC("triggering commit\n"); + kms->funcs->commit(kms, state); + } + msm_atomic_wait_for_commit_done(dev, state); kms->funcs->complete_commit(kms, state); - drm_atomic_helper_wait_for_vblanks(dev, state); - drm_atomic_helper_commit_hw_done(state); drm_atomic_helper_cleanup_planes(dev, state); diff --git a/drivers/gpu/drm/msm/msm_debugfs.c b/drivers/gpu/drm/msm/msm_debugfs.c index 1ff3fda245d1..f0da0d3c8a80 100644 --- a/drivers/gpu/drm/msm/msm_debugfs.c +++ b/drivers/gpu/drm/msm/msm_debugfs.c @@ -16,26 +16,101 @@ */ #ifdef CONFIG_DEBUG_FS +#include <linux/debugfs.h> #include "msm_drv.h" #include "msm_gpu.h" #include "msm_kms.h" #include "msm_debugfs.h" -static int msm_gpu_show(struct drm_device *dev, struct seq_file *m) +struct msm_gpu_show_priv { + struct msm_gpu_state *state; + struct drm_device *dev; +}; + +static int msm_gpu_show(struct seq_file *m, void *arg) +{ + struct drm_printer p = drm_seq_file_printer(m); + struct msm_gpu_show_priv *show_priv = m->private; + struct msm_drm_private *priv = show_priv->dev->dev_private; + struct msm_gpu *gpu = priv->gpu; + int ret; + + ret = mutex_lock_interruptible(&show_priv->dev->struct_mutex); + if (ret) + return ret; + + drm_printf(&p, "%s Status:\n", gpu->name); + gpu->funcs->show(gpu, show_priv->state, &p); + + mutex_unlock(&show_priv->dev->struct_mutex); + + return 0; +} + +static int msm_gpu_release(struct inode *inode, struct file *file) { + struct seq_file *m = file->private_data; + struct msm_gpu_show_priv *show_priv = m->private; + struct msm_drm_private *priv = show_priv->dev->dev_private; + struct msm_gpu *gpu = priv->gpu; + int ret; + + ret = mutex_lock_interruptible(&show_priv->dev->struct_mutex); + if (ret) + return ret; + + gpu->funcs->gpu_state_put(show_priv->state); + mutex_unlock(&show_priv->dev->struct_mutex); + + kfree(show_priv); + + return single_release(inode, file); +} + +static int msm_gpu_open(struct inode *inode, struct file *file) +{ + struct drm_device *dev = inode->i_private; struct msm_drm_private *priv = dev->dev_private; struct msm_gpu *gpu = priv->gpu; + struct msm_gpu_show_priv *show_priv; + int ret; - if (gpu) { - seq_printf(m, "%s Status:\n", gpu->name); - pm_runtime_get_sync(&gpu->pdev->dev); - gpu->funcs->show(gpu, m); - pm_runtime_put_sync(&gpu->pdev->dev); + if (!gpu) + return -ENODEV; + + show_priv = kmalloc(sizeof(*show_priv), GFP_KERNEL); + if (!show_priv) + return -ENOMEM; + + ret = mutex_lock_interruptible(&dev->struct_mutex); + if (ret) + return ret; + + pm_runtime_get_sync(&gpu->pdev->dev); + show_priv->state = gpu->funcs->gpu_state_get(gpu); + pm_runtime_put_sync(&gpu->pdev->dev); + + mutex_unlock(&dev->struct_mutex); + + if (IS_ERR(show_priv->state)) { + ret = PTR_ERR(show_priv->state); + kfree(show_priv); + return ret; } - return 0; + show_priv->dev = dev; + + return single_open(file, msm_gpu_show, show_priv); } +static const struct file_operations msm_gpu_fops = { + .owner = THIS_MODULE, + .open = msm_gpu_open, + .read = seq_read, + .llseek = seq_lseek, + .release = msm_gpu_release, +}; + static int msm_gem_show(struct drm_device *dev, struct seq_file *m) { struct msm_drm_private *priv = dev->dev_private; @@ -105,7 +180,6 @@ static int show_locked(struct seq_file *m, void *arg) } static struct drm_info_list msm_debugfs_list[] = { - {"gpu", show_locked, 0, msm_gpu_show}, {"gem", show_locked, 0, msm_gem_show}, { "mm", show_locked, 0, msm_mm_show }, { "fb", show_locked, 0, msm_fb_show }, @@ -158,6 +232,9 @@ int msm_debugfs_init(struct drm_minor *minor) return ret; } + debugfs_create_file("gpu", S_IRUSR, minor->debugfs_root, + dev, &msm_gpu_fops); + if (priv->kms->funcs->debugfs_init) { ret = priv->kms->funcs->debugfs_init(priv->kms, minor); if (ret) diff --git a/drivers/gpu/drm/msm/msm_drv.c b/drivers/gpu/drm/msm/msm_drv.c index 021a0b6f9a59..46876bc8b707 100644 --- a/drivers/gpu/drm/msm/msm_drv.c +++ b/drivers/gpu/drm/msm/msm_drv.c @@ -1,4 +1,5 @@ /* + * Copyright (c) 2016-2018, The Linux Foundation. All rights reserved. * Copyright (C) 2013 Red Hat * Author: Rob Clark <robdclark@gmail.com> * @@ -15,6 +16,8 @@ * this program. If not, see <http://www.gnu.org/licenses/>. */ +#include <linux/kthread.h> +#include <uapi/linux/sched/types.h> #include <drm/drm_of.h> #include "msm_drv.h" @@ -149,7 +152,7 @@ struct vblank_event { bool enable; }; -static void vblank_ctrl_worker(struct work_struct *work) +static void vblank_ctrl_worker(struct kthread_work *work) { struct msm_vblank_ctrl *vbl_ctrl = container_of(work, struct msm_vblank_ctrl, work); @@ -197,7 +200,8 @@ static int vblank_ctrl_queue_work(struct msm_drm_private *priv, list_add_tail(&vbl_ev->node, &vbl_ctrl->event_list); spin_unlock_irqrestore(&vbl_ctrl->lock, flags); - queue_work(priv->wq, &vbl_ctrl->work); + kthread_queue_work(&priv->disp_thread[crtc_id].worker, + &vbl_ctrl->work); return 0; } @@ -208,19 +212,36 @@ static int msm_drm_uninit(struct device *dev) struct drm_device *ddev = platform_get_drvdata(pdev); struct msm_drm_private *priv = ddev->dev_private; struct msm_kms *kms = priv->kms; + struct msm_mdss *mdss = priv->mdss; struct msm_vblank_ctrl *vbl_ctrl = &priv->vblank_ctrl; struct vblank_event *vbl_ev, *tmp; + int i; /* We must cancel and cleanup any pending vblank enable/disable * work before drm_irq_uninstall() to avoid work re-enabling an * irq after uninstall has disabled it. */ - cancel_work_sync(&vbl_ctrl->work); + kthread_flush_work(&vbl_ctrl->work); list_for_each_entry_safe(vbl_ev, tmp, &vbl_ctrl->event_list, node) { list_del(&vbl_ev->node); kfree(vbl_ev); } + /* clean up display commit/event worker threads */ + for (i = 0; i < priv->num_crtcs; i++) { + if (priv->disp_thread[i].thread) { + kthread_flush_worker(&priv->disp_thread[i].worker); + kthread_stop(priv->disp_thread[i].thread); + priv->disp_thread[i].thread = NULL; + } + + if (priv->event_thread[i].thread) { + kthread_flush_worker(&priv->event_thread[i].worker); + kthread_stop(priv->event_thread[i].thread); + priv->event_thread[i].thread = NULL; + } + } + msm_gem_shrinker_cleanup(ddev); drm_kms_helper_poll_fini(ddev); @@ -243,9 +264,6 @@ static int msm_drm_uninit(struct device *dev) flush_workqueue(priv->wq); destroy_workqueue(priv->wq); - flush_workqueue(priv->atomic_wq); - destroy_workqueue(priv->atomic_wq); - if (kms && kms->funcs) kms->funcs->destroy(kms); @@ -258,7 +276,8 @@ static int msm_drm_uninit(struct device *dev) component_unbind_all(dev, ddev); - msm_mdss_destroy(ddev); + if (mdss && mdss->funcs) + mdss->funcs->destroy(ddev); ddev->dev_private = NULL; drm_dev_unref(ddev); @@ -268,6 +287,10 @@ static int msm_drm_uninit(struct device *dev) return 0; } +#define KMS_MDP4 4 +#define KMS_MDP5 5 +#define KMS_DPU 3 + static int get_mdp_ver(struct platform_device *pdev) { struct device *dev = &pdev->dev; @@ -357,7 +380,9 @@ static int msm_drm_init(struct device *dev, struct drm_driver *drv) struct drm_device *ddev; struct msm_drm_private *priv; struct msm_kms *kms; - int ret; + struct msm_mdss *mdss; + int ret, i; + struct sched_param param; ddev = drm_dev_alloc(drv, dev); if (IS_ERR(ddev)) { @@ -369,53 +394,61 @@ static int msm_drm_init(struct device *dev, struct drm_driver *drv) priv = kzalloc(sizeof(*priv), GFP_KERNEL); if (!priv) { - drm_dev_unref(ddev); - return -ENOMEM; + ret = -ENOMEM; + goto err_unref_drm_dev; } ddev->dev_private = priv; priv->dev = ddev; - ret = msm_mdss_init(ddev); - if (ret) { - kfree(priv); - drm_dev_unref(ddev); - return ret; + switch (get_mdp_ver(pdev)) { + case KMS_MDP5: + ret = mdp5_mdss_init(ddev); + break; + case KMS_DPU: + ret = dpu_mdss_init(ddev); + break; + default: + ret = 0; + break; } + if (ret) + goto err_free_priv; + + mdss = priv->mdss; priv->wq = alloc_ordered_workqueue("msm", 0); - priv->atomic_wq = alloc_ordered_workqueue("msm:atomic", 0); INIT_LIST_HEAD(&priv->inactive_list); INIT_LIST_HEAD(&priv->vblank_ctrl.event_list); - INIT_WORK(&priv->vblank_ctrl.work, vblank_ctrl_worker); + kthread_init_work(&priv->vblank_ctrl.work, vblank_ctrl_worker); spin_lock_init(&priv->vblank_ctrl.lock); drm_mode_config_init(ddev); /* Bind all our sub-components: */ ret = component_bind_all(dev, ddev); - if (ret) { - msm_mdss_destroy(ddev); - kfree(priv); - drm_dev_unref(ddev); - return ret; - } + if (ret) + goto err_destroy_mdss; ret = msm_init_vram(ddev); if (ret) - goto fail; + goto err_msm_uninit; msm_gem_shrinker_init(ddev); switch (get_mdp_ver(pdev)) { - case 4: + case KMS_MDP4: kms = mdp4_kms_init(ddev); priv->kms = kms; break; - case 5: + case KMS_MDP5: kms = mdp5_kms_init(ddev); break; + case KMS_DPU: + kms = dpu_kms_init(ddev); + priv->kms = kms; + break; default: kms = ERR_PTR(-ENODEV); break; @@ -430,24 +463,100 @@ static int msm_drm_init(struct device *dev, struct drm_driver *drv) */ dev_err(dev, "failed to load kms\n"); ret = PTR_ERR(kms); - goto fail; + goto err_msm_uninit; } + /* Enable normalization of plane zpos */ + ddev->mode_config.normalize_zpos = true; + if (kms) { ret = kms->funcs->hw_init(kms); if (ret) { dev_err(dev, "kms hw init failed: %d\n", ret); - goto fail; + goto err_msm_uninit; } } ddev->mode_config.funcs = &mode_config_funcs; ddev->mode_config.helper_private = &mode_config_helper_funcs; + /** + * this priority was found during empiric testing to have appropriate + * realtime scheduling to process display updates and interact with + * other real time and normal priority task + */ + param.sched_priority = 16; + for (i = 0; i < priv->num_crtcs; i++) { + + /* initialize display thread */ + priv->disp_thread[i].crtc_id = priv->crtcs[i]->base.id; + kthread_init_worker(&priv->disp_thread[i].worker); + priv->disp_thread[i].dev = ddev; + priv->disp_thread[i].thread = + kthread_run(kthread_worker_fn, + &priv->disp_thread[i].worker, + "crtc_commit:%d", priv->disp_thread[i].crtc_id); + ret = sched_setscheduler(priv->disp_thread[i].thread, + SCHED_FIFO, ¶m); + if (ret) + pr_warn("display thread priority update failed: %d\n", + ret); + + if (IS_ERR(priv->disp_thread[i].thread)) { + dev_err(dev, "failed to create crtc_commit kthread\n"); + priv->disp_thread[i].thread = NULL; + } + + /* initialize event thread */ + priv->event_thread[i].crtc_id = priv->crtcs[i]->base.id; + kthread_init_worker(&priv->event_thread[i].worker); + priv->event_thread[i].dev = ddev; + priv->event_thread[i].thread = + kthread_run(kthread_worker_fn, + &priv->event_thread[i].worker, + "crtc_event:%d", priv->event_thread[i].crtc_id); + /** + * event thread should also run at same priority as disp_thread + * because it is handling frame_done events. A lower priority + * event thread and higher priority disp_thread can causes + * frame_pending counters beyond 2. This can lead to commit + * failure at crtc commit level. + */ + ret = sched_setscheduler(priv->event_thread[i].thread, + SCHED_FIFO, ¶m); + if (ret) + pr_warn("display event thread priority update failed: %d\n", + ret); + + if (IS_ERR(priv->event_thread[i].thread)) { + dev_err(dev, "failed to create crtc_event kthread\n"); + priv->event_thread[i].thread = NULL; + } + + if ((!priv->disp_thread[i].thread) || + !priv->event_thread[i].thread) { + /* clean up previously created threads if any */ + for ( ; i >= 0; i--) { + if (priv->disp_thread[i].thread) { + kthread_stop( + priv->disp_thread[i].thread); + priv->disp_thread[i].thread = NULL; + } + + if (priv->event_thread[i].thread) { + kthread_stop( + priv->event_thread[i].thread); + priv->event_thread[i].thread = NULL; + } + } + goto err_msm_uninit; + } + } + ret = drm_vblank_init(ddev, priv->num_crtcs); if (ret < 0) { dev_err(dev, "failed to initialize vblank\n"); - goto fail; + goto err_msm_uninit; } if (kms) { @@ -456,13 +565,13 @@ static int msm_drm_init(struct device *dev, struct drm_driver *drv) pm_runtime_put_sync(dev); if (ret < 0) { dev_err(dev, "failed to install IRQ handler\n"); - goto fail; + goto err_msm_uninit; } } ret = drm_dev_register(ddev, 0); if (ret) - goto fail; + goto err_msm_uninit; drm_mode_config_reset(ddev); @@ -473,15 +582,23 @@ static int msm_drm_init(struct device *dev, struct drm_driver *drv) ret = msm_debugfs_late_init(ddev); if (ret) - goto fail; + goto err_msm_uninit; drm_kms_helper_poll_init(ddev); return 0; -fail: +err_msm_uninit: msm_drm_uninit(dev); return ret; +err_destroy_mdss: + if (mdss && mdss->funcs) + mdss->funcs->destroy(ddev); +err_free_priv: + kfree(priv); +err_unref_drm_dev: + drm_dev_unref(ddev); + return ret; } /* @@ -894,16 +1011,35 @@ static struct drm_driver msm_driver = { static int msm_pm_suspend(struct device *dev) { struct drm_device *ddev = dev_get_drvdata(dev); + struct msm_drm_private *priv = ddev->dev_private; + struct msm_kms *kms = priv->kms; + + /* TODO: Use atomic helper suspend/resume */ + if (kms && kms->funcs && kms->funcs->pm_suspend) + return kms->funcs->pm_suspend(dev); drm_kms_helper_poll_disable(ddev); + priv->pm_state = drm_atomic_helper_suspend(ddev); + if (IS_ERR(priv->pm_state)) { + drm_kms_helper_poll_enable(ddev); + return PTR_ERR(priv->pm_state); + } + return 0; } static int msm_pm_resume(struct device *dev) { struct drm_device *ddev = dev_get_drvdata(dev); + struct msm_drm_private *priv = ddev->dev_private; + struct msm_kms *kms = priv->kms; + + /* TODO: Use atomic helper suspend/resume */ + if (kms && kms->funcs && kms->funcs->pm_resume) + return kms->funcs->pm_resume(dev); + drm_atomic_helper_resume(ddev, priv->pm_state); drm_kms_helper_poll_enable(ddev); return 0; @@ -915,11 +1051,12 @@ static int msm_runtime_suspend(struct device *dev) { struct drm_device *ddev = dev_get_drvdata(dev); struct msm_drm_private *priv = ddev->dev_private; + struct msm_mdss *mdss = priv->mdss; DBG(""); - if (priv->mdss) - return msm_mdss_disable(priv->mdss); + if (mdss && mdss->funcs) + return mdss->funcs->disable(mdss); return 0; } @@ -928,11 +1065,12 @@ static int msm_runtime_resume(struct device *dev) { struct drm_device *ddev = dev_get_drvdata(dev); struct msm_drm_private *priv = ddev->dev_private; + struct msm_mdss *mdss = priv->mdss; DBG(""); - if (priv->mdss) - return msm_mdss_enable(priv->mdss); + if (mdss && mdss->funcs) + return mdss->funcs->enable(mdss); return 0; } @@ -1031,12 +1169,13 @@ static int add_display_components(struct device *dev, int ret; /* - * MDP5 based devices don't have a flat hierarchy. There is a top level - * parent: MDSS, and children: MDP5, DSI, HDMI, eDP etc. Populate the - * children devices, find the MDP5 node, and then add the interfaces - * to our components list. + * MDP5/DPU based devices don't have a flat hierarchy. There is a top + * level parent: MDSS, and children: MDP5/DPU, DSI, HDMI, eDP etc. + * Populate the children devices, find the MDP5/DPU node, and then add + * the interfaces to our components list. */ - if (of_device_is_compatible(dev->of_node, "qcom,mdss")) { + if (of_device_is_compatible(dev->of_node, "qcom,mdss") || + of_device_is_compatible(dev->of_node, "qcom,sdm845-mdss")) { ret = of_platform_populate(dev->of_node, NULL, NULL, dev); if (ret) { dev_err(dev, "failed to populate children devices\n"); @@ -1146,8 +1285,9 @@ static int msm_pdev_remove(struct platform_device *pdev) } static const struct of_device_id dt_match[] = { - { .compatible = "qcom,mdp4", .data = (void *)4 }, /* MDP4 */ - { .compatible = "qcom,mdss", .data = (void *)5 }, /* MDP5 MDSS */ + { .compatible = "qcom,mdp4", .data = (void *)KMS_MDP4 }, + { .compatible = "qcom,mdss", .data = (void *)KMS_MDP5 }, + { .compatible = "qcom,sdm845-mdss", .data = (void *)KMS_DPU }, {} }; MODULE_DEVICE_TABLE(of, dt_match); @@ -1169,6 +1309,7 @@ static int __init msm_drm_register(void) DBG("init"); msm_mdp_register(); + msm_dpu_register(); msm_dsi_register(); msm_edp_register(); msm_hdmi_register(); @@ -1185,6 +1326,7 @@ static void __exit msm_drm_unregister(void) msm_edp_unregister(); msm_dsi_unregister(); msm_mdp_unregister(); + msm_dpu_unregister(); } module_init(msm_drm_register); diff --git a/drivers/gpu/drm/msm/msm_drv.h b/drivers/gpu/drm/msm/msm_drv.h index b2da1fbf81e0..0cba86ed3f54 100644 --- a/drivers/gpu/drm/msm/msm_drv.h +++ b/drivers/gpu/drm/msm/msm_drv.h @@ -1,4 +1,5 @@ /* + * Copyright (c) 2016-2018, The Linux Foundation. All rights reserved. * Copyright (C) 2013 Red Hat * Author: Rob Clark <robdclark@gmail.com> * @@ -33,6 +34,7 @@ #include <linux/of_graph.h> #include <linux/of_device.h> #include <asm/sizes.h> +#include <linux/kthread.h> #include <drm/drmP.h> #include <drm/drm_atomic.h> @@ -54,6 +56,12 @@ struct msm_fence_context; struct msm_gem_address_space; struct msm_gem_vma; +#define MAX_CRTCS 8 +#define MAX_PLANES 20 +#define MAX_ENCODERS 8 +#define MAX_BRIDGES 8 +#define MAX_CONNECTORS 8 + struct msm_file_private { rwlock_t queuelock; struct list_head submitqueues; @@ -68,12 +76,77 @@ enum msm_mdp_plane_property { }; struct msm_vblank_ctrl { - struct work_struct work; + struct kthread_work work; struct list_head event_list; spinlock_t lock; }; #define MSM_GPU_MAX_RINGS 4 +#define MAX_H_TILES_PER_DISPLAY 2 + +/** + * enum msm_display_caps - features/capabilities supported by displays + * @MSM_DISPLAY_CAP_VID_MODE: Video or "active" mode supported + * @MSM_DISPLAY_CAP_CMD_MODE: Command mode supported + * @MSM_DISPLAY_CAP_HOT_PLUG: Hot plug detection supported + * @MSM_DISPLAY_CAP_EDID: EDID supported + */ +enum msm_display_caps { + MSM_DISPLAY_CAP_VID_MODE = BIT(0), + MSM_DISPLAY_CAP_CMD_MODE = BIT(1), + MSM_DISPLAY_CAP_HOT_PLUG = BIT(2), + MSM_DISPLAY_CAP_EDID = BIT(3), +}; + +/** + * enum msm_event_wait - type of HW events to wait for + * @MSM_ENC_COMMIT_DONE - wait for the driver to flush the registers to HW + * @MSM_ENC_TX_COMPLETE - wait for the HW to transfer the frame to panel + * @MSM_ENC_VBLANK - wait for the HW VBLANK event (for driver-internal waiters) + */ +enum msm_event_wait { + MSM_ENC_COMMIT_DONE = 0, + MSM_ENC_TX_COMPLETE, + MSM_ENC_VBLANK, +}; + +/** + * struct msm_display_topology - defines a display topology pipeline + * @num_lm: number of layer mixers used + * @num_enc: number of compression encoder blocks used + * @num_intf: number of interfaces the panel is mounted on + */ +struct msm_display_topology { + u32 num_lm; + u32 num_enc; + u32 num_intf; +}; + +/** + * struct msm_display_info - defines display properties + * @intf_type: DRM_MODE_CONNECTOR_ display type + * @capabilities: Bitmask of display flags + * @num_of_h_tiles: Number of horizontal tiles in case of split interface + * @h_tile_instance: Controller instance used per tile. Number of elements is + * based on num_of_h_tiles + * @is_te_using_watchdog_timer: Boolean to indicate watchdog TE is + * used instead of panel TE in cmd mode panels + */ +struct msm_display_info { + int intf_type; + uint32_t capabilities; + uint32_t num_of_h_tiles; + uint32_t h_tile_instance[MAX_H_TILES_PER_DISPLAY]; + bool is_te_using_watchdog_timer; +}; + +/* Commit/Event thread specific structure */ +struct msm_drm_thread { + struct drm_device *dev; + struct task_struct *thread; + unsigned int crtc_id; + struct kthread_worker worker; +}; struct msm_drm_private { @@ -84,7 +157,7 @@ struct msm_drm_private { /* subordinate devices, if present: */ struct platform_device *gpu_pdev; - /* top level MDSS wrapper device (for MDP5 only) */ + /* top level MDSS wrapper device (for MDP5/DPU only) */ struct msm_mdss *mdss; /* possibly this should be in the kms component, but it is @@ -115,22 +188,24 @@ struct msm_drm_private { struct list_head inactive_list; struct workqueue_struct *wq; - struct workqueue_struct *atomic_wq; unsigned int num_planes; - struct drm_plane *planes[16]; + struct drm_plane *planes[MAX_PLANES]; unsigned int num_crtcs; - struct drm_crtc *crtcs[8]; + struct drm_crtc *crtcs[MAX_CRTCS]; + + struct msm_drm_thread disp_thread[MAX_CRTCS]; + struct msm_drm_thread event_thread[MAX_CRTCS]; unsigned int num_encoders; - struct drm_encoder *encoders[8]; + struct drm_encoder *encoders[MAX_ENCODERS]; unsigned int num_bridges; - struct drm_bridge *bridges[8]; + struct drm_bridge *bridges[MAX_BRIDGES]; unsigned int num_connectors; - struct drm_connector *connectors[8]; + struct drm_connector *connectors[MAX_CONNECTORS]; /* Properties */ struct drm_property *plane_property[PLANE_PROP_MAX_NUM]; @@ -150,6 +225,7 @@ struct msm_drm_private { struct shrinker shrinker; struct msm_vblank_ctrl vblank_ctrl; + struct drm_atomic_state *pm_state; }; struct msm_format { @@ -174,6 +250,9 @@ struct msm_gem_address_space * msm_gem_address_space_create(struct device *dev, struct iommu_domain *domain, const char *name); +int msm_register_mmu(struct drm_device *dev, struct msm_mmu *mmu); +void msm_unregister_mmu(struct drm_device *dev, struct msm_mmu *mmu); + void msm_gem_submit_free(struct msm_gem_submit *submit); int msm_ioctl_gem_submit(struct drm_device *dev, void *data, struct drm_file *file); @@ -285,6 +364,8 @@ static inline int msm_dsi_modeset_init(struct msm_dsi *msm_dsi, void __init msm_mdp_register(void); void __exit msm_mdp_unregister(void); +void __init msm_dpu_register(void); +void __exit msm_dpu_unregister(void); #ifdef CONFIG_DEBUG_FS void msm_gem_describe(struct drm_gem_object *obj, struct seq_file *m); diff --git a/drivers/gpu/drm/msm/msm_gpu.c b/drivers/gpu/drm/msm/msm_gpu.c index 1c09acfb4028..f388944c93e2 100644 --- a/drivers/gpu/drm/msm/msm_gpu.c +++ b/drivers/gpu/drm/msm/msm_gpu.c @@ -20,10 +20,11 @@ #include "msm_mmu.h" #include "msm_fence.h" +#include <generated/utsrelease.h> #include <linux/string_helpers.h> #include <linux/pm_opp.h> #include <linux/devfreq.h> - +#include <linux/devcoredump.h> /* * Power Management: @@ -273,6 +274,123 @@ int msm_gpu_hw_init(struct msm_gpu *gpu) return ret; } +#ifdef CONFIG_DEV_COREDUMP +static ssize_t msm_gpu_devcoredump_read(char *buffer, loff_t offset, + size_t count, void *data, size_t datalen) +{ + struct msm_gpu *gpu = data; + struct drm_print_iterator iter; + struct drm_printer p; + struct msm_gpu_state *state; + + state = msm_gpu_crashstate_get(gpu); + if (!state) + return 0; + + iter.data = buffer; + iter.offset = 0; + iter.start = offset; + iter.remain = count; + + p = drm_coredump_printer(&iter); + + drm_printf(&p, "---\n"); + drm_printf(&p, "kernel: " UTS_RELEASE "\n"); + drm_printf(&p, "module: " KBUILD_MODNAME "\n"); + drm_printf(&p, "time: %lld.%09ld\n", + state->time.tv_sec, state->time.tv_nsec); + if (state->comm) + drm_printf(&p, "comm: %s\n", state->comm); + if (state->cmd) + drm_printf(&p, "cmdline: %s\n", state->cmd); + + gpu->funcs->show(gpu, state, &p); + + msm_gpu_crashstate_put(gpu); + + return count - iter.remain; +} + +static void msm_gpu_devcoredump_free(void *data) +{ + struct msm_gpu *gpu = data; + + msm_gpu_crashstate_put(gpu); +} + +static void msm_gpu_crashstate_get_bo(struct msm_gpu_state *state, + struct msm_gem_object *obj, u64 iova, u32 flags) +{ + struct msm_gpu_state_bo *state_bo = &state->bos[state->nr_bos]; + + /* Don't record write only objects */ + + state_bo->size = obj->base.size; + state_bo->iova = iova; + + /* Only store the data for buffer objects marked for read */ + if ((flags & MSM_SUBMIT_BO_READ)) { + void *ptr; + + state_bo->data = kvmalloc(obj->base.size, GFP_KERNEL); + if (!state_bo->data) + return; + + ptr = msm_gem_get_vaddr_active(&obj->base); + if (IS_ERR(ptr)) { + kvfree(state_bo->data); + return; + } + + memcpy(state_bo->data, ptr, obj->base.size); + msm_gem_put_vaddr(&obj->base); + } + + state->nr_bos++; +} + +static void msm_gpu_crashstate_capture(struct msm_gpu *gpu, + struct msm_gem_submit *submit, char *comm, char *cmd) +{ + struct msm_gpu_state *state; + + /* Only save one crash state at a time */ + if (gpu->crashstate) + return; + + state = gpu->funcs->gpu_state_get(gpu); + if (IS_ERR_OR_NULL(state)) + return; + + /* Fill in the additional crash state information */ + state->comm = kstrdup(comm, GFP_KERNEL); + state->cmd = kstrdup(cmd, GFP_KERNEL); + + if (submit) { + int i; + + state->bos = kcalloc(submit->nr_bos, + sizeof(struct msm_gpu_state_bo), GFP_KERNEL); + + for (i = 0; state->bos && i < submit->nr_bos; i++) + msm_gpu_crashstate_get_bo(state, submit->bos[i].obj, + submit->bos[i].iova, submit->bos[i].flags); + } + + /* Set the active crash state to be dumped on failure */ + gpu->crashstate = state; + + /* FIXME: Release the crashstate if this errors out? */ + dev_coredumpm(gpu->dev->dev, THIS_MODULE, gpu, 0, GFP_KERNEL, + msm_gpu_devcoredump_read, msm_gpu_devcoredump_free); +} +#else +static void msm_gpu_crashstate_capture(struct msm_gpu *gpu, char *comm, + char *cmd) +{ +} +#endif + /* * Hangcheck detection for locked gpu: */ @@ -314,6 +432,7 @@ static void recover_worker(struct work_struct *work) struct msm_drm_private *priv = dev->dev_private; struct msm_gem_submit *submit; struct msm_ringbuffer *cur_ring = gpu->funcs->active_ring(gpu); + char *comm = NULL, *cmd = NULL; int i; mutex_lock(&dev->struct_mutex); @@ -327,7 +446,7 @@ static void recover_worker(struct work_struct *work) rcu_read_lock(); task = pid_task(submit->pid, PIDTYPE_PID); if (task) { - char *cmd; + comm = kstrdup(task->comm, GFP_ATOMIC); /* * So slightly annoying, in other paths like @@ -340,22 +459,28 @@ static void recover_worker(struct work_struct *work) * about the submit going away. */ mutex_unlock(&dev->struct_mutex); - cmd = kstrdup_quotable_cmdline(task, GFP_KERNEL); + cmd = kstrdup_quotable_cmdline(task, GFP_ATOMIC); mutex_lock(&dev->struct_mutex); + } + rcu_read_unlock(); + if (comm && cmd) { dev_err(dev->dev, "%s: offending task: %s (%s)\n", - gpu->name, task->comm, cmd); + gpu->name, comm, cmd); msm_rd_dump_submit(priv->hangrd, submit, - "offending task: %s (%s)", task->comm, cmd); - - kfree(cmd); - } else { + "offending task: %s (%s)", comm, cmd); + } else msm_rd_dump_submit(priv->hangrd, submit, NULL); - } - rcu_read_unlock(); } + /* Record the crash state */ + pm_runtime_get_sync(&gpu->pdev->dev); + msm_gpu_crashstate_capture(gpu, submit, comm, cmd); + pm_runtime_put_sync(&gpu->pdev->dev); + + kfree(cmd); + kfree(comm); /* * Update all the rings with the latest and greatest fence.. this diff --git a/drivers/gpu/drm/msm/msm_gpu.h b/drivers/gpu/drm/msm/msm_gpu.h index b8241179175a..1c6105bc55c7 100644 --- a/drivers/gpu/drm/msm/msm_gpu.h +++ b/drivers/gpu/drm/msm/msm_gpu.h @@ -27,6 +27,7 @@ struct msm_gem_submit; struct msm_gpu_perfcntr; +struct msm_gpu_state; struct msm_gpu_config { const char *ioname; @@ -64,11 +65,14 @@ struct msm_gpu_funcs { void (*destroy)(struct msm_gpu *gpu); #ifdef CONFIG_DEBUG_FS /* show GPU status in debugfs: */ - void (*show)(struct msm_gpu *gpu, struct seq_file *m); + void (*show)(struct msm_gpu *gpu, struct msm_gpu_state *state, + struct drm_printer *p); /* for generation specific debugfs: */ int (*debugfs_init)(struct msm_gpu *gpu, struct drm_minor *minor); #endif int (*gpu_busy)(struct msm_gpu *gpu, uint64_t *value); + struct msm_gpu_state *(*gpu_state_get)(struct msm_gpu *gpu); + int (*gpu_state_put)(struct msm_gpu_state *state); }; struct msm_gpu { @@ -129,6 +133,8 @@ struct msm_gpu { u64 busy_cycles; ktime_t time; } devfreq; + + struct msm_gpu_state *crashstate; }; /* It turns out that all targets use the same ringbuffer size */ @@ -175,6 +181,38 @@ struct msm_gpu_submitqueue { struct kref ref; }; +struct msm_gpu_state_bo { + u64 iova; + size_t size; + void *data; +}; + +struct msm_gpu_state { + struct kref ref; + struct timespec64 time; + + struct { + u64 iova; + u32 fence; + u32 seqno; + u32 rptr; + u32 wptr; + void *data; + int data_size; + } ring[MSM_GPU_MAX_RINGS]; + + int nr_registers; + u32 *registers; + + u32 rbbm_status; + + char *comm; + char *cmd; + + int nr_bos; + struct msm_gpu_state_bo *bos; +}; + static inline void gpu_write(struct msm_gpu *gpu, u32 reg, u32 data) { msm_writel(data, gpu->mmio + (reg << 2)); @@ -254,4 +292,32 @@ static inline void msm_submitqueue_put(struct msm_gpu_submitqueue *queue) kref_put(&queue->ref, msm_submitqueue_destroy); } +static inline struct msm_gpu_state *msm_gpu_crashstate_get(struct msm_gpu *gpu) +{ + struct msm_gpu_state *state = NULL; + + mutex_lock(&gpu->dev->struct_mutex); + + if (gpu->crashstate) { + kref_get(&gpu->crashstate->ref); + state = gpu->crashstate; + } + + mutex_unlock(&gpu->dev->struct_mutex); + + return state; +} + +static inline void msm_gpu_crashstate_put(struct msm_gpu *gpu) +{ + mutex_lock(&gpu->dev->struct_mutex); + + if (gpu->crashstate) { + if (gpu->funcs->gpu_state_put(gpu->crashstate)) + gpu->crashstate = NULL; + } + + mutex_unlock(&gpu->dev->struct_mutex); +} + #endif /* __MSM_GPU_H__ */ diff --git a/drivers/gpu/drm/msm/msm_kms.h b/drivers/gpu/drm/msm/msm_kms.h index dfd92947de2c..fd88cebb6adb 100644 --- a/drivers/gpu/drm/msm/msm_kms.h +++ b/drivers/gpu/drm/msm/msm_kms.h @@ -1,4 +1,5 @@ /* + * Copyright (c) 2016-2018, The Linux Foundation. All rights reserved. * Copyright (C) 2013 Red Hat * Author: Rob Clark <robdclark@gmail.com> * @@ -42,6 +43,7 @@ struct msm_kms_funcs { void (*disable_vblank)(struct msm_kms *kms, struct drm_crtc *crtc); /* modeset, bracketing atomic_commit(): */ void (*prepare_commit)(struct msm_kms *kms, struct drm_atomic_state *state); + void (*commit)(struct msm_kms *kms, struct drm_atomic_state *state); void (*complete_commit)(struct msm_kms *kms, struct drm_atomic_state *state); /* functions to wait for atomic commit completed on each CRTC */ void (*wait_for_crtc_commit_done)(struct msm_kms *kms, @@ -50,6 +52,11 @@ struct msm_kms_funcs { const struct msm_format *(*get_format)(struct msm_kms *kms, const uint32_t format, const uint64_t modifiers); + /* do format checking on format modified through fb_cmd2 modifiers */ + int (*check_modified_format)(const struct msm_kms *kms, + const struct msm_format *msm_fmt, + const struct drm_mode_fb_cmd2 *cmd, + struct drm_gem_object **bos); /* misc: */ long (*round_pixclk)(struct msm_kms *kms, unsigned long rate, struct drm_encoder *encoder); @@ -60,6 +67,9 @@ struct msm_kms_funcs { void (*set_encoder_mode)(struct msm_kms *kms, struct drm_encoder *encoder, bool cmd_mode); + /* pm suspend/resume hooks */ + int (*pm_suspend)(struct device *dev); + int (*pm_resume)(struct device *dev); /* cleanup: */ void (*destroy)(struct msm_kms *kms); #ifdef CONFIG_DEBUG_FS @@ -86,9 +96,20 @@ static inline void msm_kms_init(struct msm_kms *kms, struct msm_kms *mdp4_kms_init(struct drm_device *dev); struct msm_kms *mdp5_kms_init(struct drm_device *dev); -int msm_mdss_init(struct drm_device *dev); -void msm_mdss_destroy(struct drm_device *dev); -int msm_mdss_enable(struct msm_mdss *mdss); -int msm_mdss_disable(struct msm_mdss *mdss); +struct msm_kms *dpu_kms_init(struct drm_device *dev); + +struct msm_mdss_funcs { + int (*enable)(struct msm_mdss *mdss); + int (*disable)(struct msm_mdss *mdss); + void (*destroy)(struct drm_device *dev); +}; + +struct msm_mdss { + struct drm_device *dev; + const struct msm_mdss_funcs *funcs; +}; + +int mdp5_mdss_init(struct drm_device *dev); +int dpu_mdss_init(struct drm_device *dev); #endif /* __MSM_KMS_H__ */ diff --git a/drivers/gpu/drm/nouveau/dispnv04/disp.c b/drivers/gpu/drm/nouveau/dispnv04/disp.c index 501d2d290e9c..70dce544984e 100644 --- a/drivers/gpu/drm/nouveau/dispnv04/disp.c +++ b/drivers/gpu/drm/nouveau/dispnv04/disp.c @@ -55,6 +55,9 @@ nv04_display_create(struct drm_device *dev) nouveau_display(dev)->init = nv04_display_init; nouveau_display(dev)->fini = nv04_display_fini; + /* Pre-nv50 doesn't support atomic, so don't expose the ioctls */ + dev->driver->driver_features &= ~DRIVER_ATOMIC; + nouveau_hw_save_vga_fonts(dev, 1); nv04_crtc_create(dev, 0); diff --git a/drivers/gpu/drm/nouveau/dispnv50/disp.c b/drivers/gpu/drm/nouveau/dispnv50/disp.c index 0190377b02a6..8412119bd940 100644 --- a/drivers/gpu/drm/nouveau/dispnv50/disp.c +++ b/drivers/gpu/drm/nouveau/dispnv50/disp.c @@ -1610,8 +1610,9 @@ nv50_pior_create(struct drm_connector *connector, struct dcb_output *dcbe) *****************************************************************************/ static void -nv50_disp_atomic_commit_core(struct nouveau_drm *drm, u32 *interlock) +nv50_disp_atomic_commit_core(struct drm_atomic_state *state, u32 *interlock) { + struct nouveau_drm *drm = nouveau_drm(state->dev); struct nv50_disp *disp = nv50_disp(drm->dev); struct nv50_core *core = disp->core; struct nv50_mstm *mstm; @@ -1643,6 +1644,22 @@ nv50_disp_atomic_commit_core(struct nouveau_drm *drm, u32 *interlock) } static void +nv50_disp_atomic_commit_wndw(struct drm_atomic_state *state, u32 *interlock) +{ + struct drm_plane_state *new_plane_state; + struct drm_plane *plane; + int i; + + for_each_new_plane_in_state(state, plane, new_plane_state, i) { + struct nv50_wndw *wndw = nv50_wndw(plane); + if (interlock[wndw->interlock.type] & wndw->interlock.data) { + if (wndw->func->update) + wndw->func->update(wndw, interlock); + } + } +} + +static void nv50_disp_atomic_commit_tail(struct drm_atomic_state *state) { struct drm_device *dev = state->dev; @@ -1709,7 +1726,8 @@ nv50_disp_atomic_commit_tail(struct drm_atomic_state *state) help->disable(encoder); interlock[NV50_DISP_INTERLOCK_CORE] |= 1; if (outp->flush_disable) { - nv50_disp_atomic_commit_core(drm, interlock); + nv50_disp_atomic_commit_wndw(state, interlock); + nv50_disp_atomic_commit_core(state, interlock); memset(interlock, 0x00, sizeof(interlock)); } } @@ -1718,15 +1736,8 @@ nv50_disp_atomic_commit_tail(struct drm_atomic_state *state) /* Flush disable. */ if (interlock[NV50_DISP_INTERLOCK_CORE]) { if (atom->flush_disable) { - for_each_new_plane_in_state(state, plane, new_plane_state, i) { - struct nv50_wndw *wndw = nv50_wndw(plane); - if (interlock[wndw->interlock.type] & wndw->interlock.data) { - if (wndw->func->update) - wndw->func->update(wndw, interlock); - } - } - - nv50_disp_atomic_commit_core(drm, interlock); + nv50_disp_atomic_commit_wndw(state, interlock); + nv50_disp_atomic_commit_core(state, interlock); memset(interlock, 0x00, sizeof(interlock)); } } @@ -1787,18 +1798,14 @@ nv50_disp_atomic_commit_tail(struct drm_atomic_state *state) } /* Flush update. */ - for_each_new_plane_in_state(state, plane, new_plane_state, i) { - struct nv50_wndw *wndw = nv50_wndw(plane); - if (interlock[wndw->interlock.type] & wndw->interlock.data) { - if (wndw->func->update) - wndw->func->update(wndw, interlock); - } - } + nv50_disp_atomic_commit_wndw(state, interlock); if (interlock[NV50_DISP_INTERLOCK_CORE]) { if (interlock[NV50_DISP_INTERLOCK_BASE] || + interlock[NV50_DISP_INTERLOCK_OVLY] || + interlock[NV50_DISP_INTERLOCK_WNDW] || !atom->state.legacy_cursor_update) - nv50_disp_atomic_commit_core(drm, interlock); + nv50_disp_atomic_commit_core(state, interlock); else disp->core->func->update(disp->core, interlock, false); } @@ -1896,7 +1903,7 @@ nv50_disp_atomic_commit(struct drm_device *dev, nv50_disp_atomic_commit_tail(state); drm_for_each_crtc(crtc, dev) { - if (crtc->state->enable) { + if (crtc->state->active) { if (!drm->have_disp_power_ref) { drm->have_disp_power_ref = true; return 0; @@ -2144,10 +2151,6 @@ nv50_display_destroy(struct drm_device *dev) kfree(disp); } -MODULE_PARM_DESC(atomic, "Expose atomic ioctl (default: disabled)"); -static int nouveau_atomic = 0; -module_param_named(atomic, nouveau_atomic, int, 0400); - int nv50_display_create(struct drm_device *dev) { @@ -2172,8 +2175,6 @@ nv50_display_create(struct drm_device *dev) disp->disp = &nouveau_display(dev)->disp; dev->mode_config.funcs = &nv50_disp_func; dev->driver->driver_features |= DRIVER_PREFER_XBGR_30BPP; - if (nouveau_atomic) - dev->driver->driver_features |= DRIVER_ATOMIC; /* small shared memory area we use for notifiers and semaphores */ ret = nouveau_bo_new(&drm->client, 4096, 0x1000, TTM_PL_FLAG_VRAM, diff --git a/drivers/gpu/drm/nouveau/nouveau_backlight.c b/drivers/gpu/drm/nouveau/nouveau_backlight.c index debbbf0fd4bd..408b955e5c39 100644 --- a/drivers/gpu/drm/nouveau/nouveau_backlight.c +++ b/drivers/gpu/drm/nouveau/nouveau_backlight.c @@ -267,6 +267,7 @@ nouveau_backlight_init(struct drm_device *dev) struct nouveau_drm *drm = nouveau_drm(dev); struct nvif_device *device = &drm->client.device; struct drm_connector *connector; + struct drm_connector_list_iter conn_iter; INIT_LIST_HEAD(&drm->bl_connectors); @@ -275,7 +276,8 @@ nouveau_backlight_init(struct drm_device *dev) return 0; } - list_for_each_entry(connector, &dev->mode_config.connector_list, head) { + drm_connector_list_iter_begin(dev, &conn_iter); + drm_for_each_connector_iter(connector, &conn_iter) { if (connector->connector_type != DRM_MODE_CONNECTOR_LVDS && connector->connector_type != DRM_MODE_CONNECTOR_eDP) continue; @@ -292,7 +294,7 @@ nouveau_backlight_init(struct drm_device *dev) break; } } - + drm_connector_list_iter_end(&conn_iter); return 0; } diff --git a/drivers/gpu/drm/nouveau/nouveau_connector.c b/drivers/gpu/drm/nouveau/nouveau_connector.c index 22a15478d23d..51932c72334e 100644 --- a/drivers/gpu/drm/nouveau/nouveau_connector.c +++ b/drivers/gpu/drm/nouveau/nouveau_connector.c @@ -1193,14 +1193,19 @@ nouveau_connector_create(struct drm_device *dev, int index) struct nouveau_display *disp = nouveau_display(dev); struct nouveau_connector *nv_connector = NULL; struct drm_connector *connector; + struct drm_connector_list_iter conn_iter; int type, ret = 0; bool dummy; - list_for_each_entry(connector, &dev->mode_config.connector_list, head) { + drm_connector_list_iter_begin(dev, &conn_iter); + nouveau_for_each_non_mst_connector_iter(connector, &conn_iter) { nv_connector = nouveau_connector(connector); - if (nv_connector->index == index) + if (nv_connector->index == index) { + drm_connector_list_iter_end(&conn_iter); return connector; + } } + drm_connector_list_iter_end(&conn_iter); nv_connector = kzalloc(sizeof(*nv_connector), GFP_KERNEL); if (!nv_connector) diff --git a/drivers/gpu/drm/nouveau/nouveau_connector.h b/drivers/gpu/drm/nouveau/nouveau_connector.h index a4d1a059bd3d..dc7454e7f19a 100644 --- a/drivers/gpu/drm/nouveau/nouveau_connector.h +++ b/drivers/gpu/drm/nouveau/nouveau_connector.h @@ -33,6 +33,7 @@ #include <drm/drm_encoder.h> #include <drm/drm_dp_helper.h> #include "nouveau_crtc.h" +#include "nouveau_encoder.h" struct nvkm_i2c_port; @@ -60,19 +61,46 @@ static inline struct nouveau_connector *nouveau_connector( return container_of(con, struct nouveau_connector, base); } +static inline bool +nouveau_connector_is_mst(struct drm_connector *connector) +{ + const struct nouveau_encoder *nv_encoder; + const struct drm_encoder *encoder; + + if (connector->connector_type != DRM_MODE_CONNECTOR_DisplayPort) + return false; + + nv_encoder = find_encoder(connector, DCB_OUTPUT_ANY); + if (!nv_encoder) + return false; + + encoder = &nv_encoder->base.base; + return encoder->encoder_type == DRM_MODE_ENCODER_DPMST; +} + +#define nouveau_for_each_non_mst_connector_iter(connector, iter) \ + drm_for_each_connector_iter(connector, iter) \ + for_each_if(!nouveau_connector_is_mst(connector)) + static inline struct nouveau_connector * nouveau_crtc_connector_get(struct nouveau_crtc *nv_crtc) { struct drm_device *dev = nv_crtc->base.dev; struct drm_connector *connector; + struct drm_connector_list_iter conn_iter; + struct nouveau_connector *nv_connector = NULL; struct drm_crtc *crtc = to_drm_crtc(nv_crtc); - list_for_each_entry(connector, &dev->mode_config.connector_list, head) { - if (connector->encoder && connector->encoder->crtc == crtc) - return nouveau_connector(connector); + drm_connector_list_iter_begin(dev, &conn_iter); + nouveau_for_each_non_mst_connector_iter(connector, &conn_iter) { + if (connector->encoder && connector->encoder->crtc == crtc) { + nv_connector = nouveau_connector(connector); + break; + } } + drm_connector_list_iter_end(&conn_iter); - return NULL; + return nv_connector; } struct drm_connector * diff --git a/drivers/gpu/drm/nouveau/nouveau_display.c b/drivers/gpu/drm/nouveau/nouveau_display.c index dfa236370726..139368b31916 100644 --- a/drivers/gpu/drm/nouveau/nouveau_display.c +++ b/drivers/gpu/drm/nouveau/nouveau_display.c @@ -404,6 +404,7 @@ nouveau_display_init(struct drm_device *dev) struct nouveau_display *disp = nouveau_display(dev); struct nouveau_drm *drm = nouveau_drm(dev); struct drm_connector *connector; + struct drm_connector_list_iter conn_iter; int ret; ret = disp->init(dev); @@ -411,10 +412,12 @@ nouveau_display_init(struct drm_device *dev) return ret; /* enable hotplug interrupts */ - list_for_each_entry(connector, &dev->mode_config.connector_list, head) { + drm_connector_list_iter_begin(dev, &conn_iter); + nouveau_for_each_non_mst_connector_iter(connector, &conn_iter) { struct nouveau_connector *conn = nouveau_connector(connector); nvif_notify_get(&conn->hpd); } + drm_connector_list_iter_end(&conn_iter); /* enable flip completion events */ nvif_notify_get(&drm->flip); @@ -427,6 +430,7 @@ nouveau_display_fini(struct drm_device *dev, bool suspend) struct nouveau_display *disp = nouveau_display(dev); struct nouveau_drm *drm = nouveau_drm(dev); struct drm_connector *connector; + struct drm_connector_list_iter conn_iter; if (!suspend) { if (drm_drv_uses_atomic_modeset(dev)) @@ -439,10 +443,12 @@ nouveau_display_fini(struct drm_device *dev, bool suspend) nvif_notify_put(&drm->flip); /* disable hotplug interrupts */ - list_for_each_entry(connector, &dev->mode_config.connector_list, head) { + drm_connector_list_iter_begin(dev, &conn_iter); + nouveau_for_each_non_mst_connector_iter(connector, &conn_iter) { struct nouveau_connector *conn = nouveau_connector(connector); nvif_notify_put(&conn->hpd); } + drm_connector_list_iter_end(&conn_iter); drm_kms_helper_poll_disable(dev); disp->fini(dev); diff --git a/drivers/gpu/drm/nouveau/nouveau_drm.c b/drivers/gpu/drm/nouveau/nouveau_drm.c index c779ee3c665b..c7ec86d6c3c9 100644 --- a/drivers/gpu/drm/nouveau/nouveau_drm.c +++ b/drivers/gpu/drm/nouveau/nouveau_drm.c @@ -81,6 +81,10 @@ MODULE_PARM_DESC(modeset, "enable driver (default: auto, " int nouveau_modeset = -1; module_param_named(modeset, nouveau_modeset, int, 0400); +MODULE_PARM_DESC(atomic, "Expose atomic ioctl (default: disabled)"); +static int nouveau_atomic = 0; +module_param_named(atomic, nouveau_atomic, int, 0400); + MODULE_PARM_DESC(runpm, "disable (0), force enable (1), optimus only default (-1)"); static int nouveau_runtime_pm = -1; module_param_named(runpm, nouveau_runtime_pm, int, 0400); @@ -509,6 +513,9 @@ static int nouveau_drm_probe(struct pci_dev *pdev, pci_set_master(pdev); + if (nouveau_atomic) + driver_pci.driver_features |= DRIVER_ATOMIC; + ret = drm_get_pci_dev(pdev, pent, &driver_pci); if (ret) { nvkm_device_del(&device); @@ -874,22 +881,11 @@ nouveau_pmops_runtime_resume(struct device *dev) static int nouveau_pmops_runtime_idle(struct device *dev) { - struct pci_dev *pdev = to_pci_dev(dev); - struct drm_device *drm_dev = pci_get_drvdata(pdev); - struct nouveau_drm *drm = nouveau_drm(drm_dev); - struct drm_crtc *crtc; - if (!nouveau_pmops_runtime()) { pm_runtime_forbid(dev); return -EBUSY; } - list_for_each_entry(crtc, &drm->dev->mode_config.crtc_list, head) { - if (crtc->enabled) { - DRM_DEBUG_DRIVER("failing to power off - crtc active\n"); - return -EBUSY; - } - } pm_runtime_mark_last_busy(dev); pm_runtime_autosuspend(dev); /* we don't want the main rpm_idle to call suspend - we want to autosuspend */ diff --git a/drivers/gpu/drm/nouveau/nouveau_gem.c b/drivers/gpu/drm/nouveau/nouveau_gem.c index df73bec354e8..b56524d343c3 100644 --- a/drivers/gpu/drm/nouveau/nouveau_gem.c +++ b/drivers/gpu/drm/nouveau/nouveau_gem.c @@ -616,7 +616,7 @@ nouveau_gem_pushbuf_reloc_apply(struct nouveau_cli *cli, struct nouveau_bo *nvbo; uint32_t data; - if (unlikely(r->bo_index > req->nr_buffers)) { + if (unlikely(r->bo_index >= req->nr_buffers)) { NV_PRINTK(err, cli, "reloc bo index invalid\n"); ret = -EINVAL; break; @@ -626,7 +626,7 @@ nouveau_gem_pushbuf_reloc_apply(struct nouveau_cli *cli, if (b->presumed.valid) continue; - if (unlikely(r->reloc_bo_index > req->nr_buffers)) { + if (unlikely(r->reloc_bo_index >= req->nr_buffers)) { NV_PRINTK(err, cli, "reloc container bo index invalid\n"); ret = -EINVAL; break; diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/base.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/base.c index 73b5d46104bd..434d2fc5bb1c 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/base.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/base.c @@ -140,6 +140,9 @@ nvkm_fb_init(struct nvkm_subdev *subdev) if (fb->func->init) fb->func->init(fb); + if (fb->func->init_remapper) + fb->func->init_remapper(fb); + if (fb->func->init_page) { ret = fb->func->init_page(fb); if (WARN_ON(ret)) diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gp100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gp100.c index dffe1f5e1071..8205ce436b3e 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gp100.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gp100.c @@ -37,6 +37,14 @@ gp100_fb_init_unkn(struct nvkm_fb *base) } void +gp100_fb_init_remapper(struct nvkm_fb *fb) +{ + struct nvkm_device *device = fb->subdev.device; + /* Disable address remapper. */ + nvkm_mask(device, 0x100c14, 0x00040000, 0x00000000); +} + +void gp100_fb_init(struct nvkm_fb *base) { struct gf100_fb *fb = gf100_fb(base); @@ -56,6 +64,7 @@ gp100_fb = { .dtor = gf100_fb_dtor, .oneinit = gf100_fb_oneinit, .init = gp100_fb_init, + .init_remapper = gp100_fb_init_remapper, .init_page = gm200_fb_init_page, .init_unkn = gp100_fb_init_unkn, .ram_new = gp100_ram_new, diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gp102.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gp102.c index b84b9861ef26..b4d74e815674 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gp102.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gp102.c @@ -31,6 +31,7 @@ gp102_fb = { .dtor = gf100_fb_dtor, .oneinit = gf100_fb_oneinit, .init = gp100_fb_init, + .init_remapper = gp100_fb_init_remapper, .init_page = gm200_fb_init_page, .ram_new = gp100_ram_new, }; diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/priv.h b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/priv.h index 2857f31466bf..1e4ad61c19e1 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/priv.h +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/priv.h @@ -11,6 +11,7 @@ struct nvkm_fb_func { u32 (*tags)(struct nvkm_fb *); int (*oneinit)(struct nvkm_fb *); void (*init)(struct nvkm_fb *); + void (*init_remapper)(struct nvkm_fb *); int (*init_page)(struct nvkm_fb *); void (*init_unkn)(struct nvkm_fb *); void (*intr)(struct nvkm_fb *); @@ -69,5 +70,6 @@ int gf100_fb_init_page(struct nvkm_fb *); int gm200_fb_init_page(struct nvkm_fb *); +void gp100_fb_init_remapper(struct nvkm_fb *); void gp100_fb_init_unkn(struct nvkm_fb *); #endif diff --git a/drivers/gpu/drm/sun4i/Makefile b/drivers/gpu/drm/sun4i/Makefile index b04ea0f3da75..0eb38ac8e86e 100644 --- a/drivers/gpu/drm/sun4i/Makefile +++ b/drivers/gpu/drm/sun4i/Makefile @@ -32,7 +32,10 @@ obj-$(CONFIG_DRM_SUN4I) += sun4i-tcon.o obj-$(CONFIG_DRM_SUN4I) += sun4i_tv.o obj-$(CONFIG_DRM_SUN4I) += sun6i_drc.o -obj-$(CONFIG_DRM_SUN4I_BACKEND) += sun4i-backend.o sun4i-frontend.o +obj-$(CONFIG_DRM_SUN4I_BACKEND) += sun4i-backend.o +ifdef CONFIG_DRM_SUN4I_BACKEND +obj-$(CONFIG_DRM_SUN4I) += sun4i-frontend.o +endif obj-$(CONFIG_DRM_SUN4I_HDMI) += sun4i-drm-hdmi.o obj-$(CONFIG_DRM_SUN6I_DSI) += sun6i-dsi.o obj-$(CONFIG_DRM_SUN8I_DW_HDMI) += sun8i-drm-hdmi.o diff --git a/drivers/gpu/drm/tegra/drm.c b/drivers/gpu/drm/tegra/drm.c index 776c1513e582..a2bd5876c633 100644 --- a/drivers/gpu/drm/tegra/drm.c +++ b/drivers/gpu/drm/tegra/drm.c @@ -398,7 +398,7 @@ int tegra_drm_submit(struct tegra_drm_context *context, * unaligned offset is malformed and cause commands stream * corruption on the buffer address relocation. */ - if (offset & 3 || offset >= obj->gem.size) { + if (offset & 3 || offset > obj->gem.size) { err = -EINVAL; goto fail; } diff --git a/drivers/gpu/drm/udl/udl_drv.h b/drivers/gpu/drm/udl/udl_drv.h index 072582570a4f..e9e9b1ff678e 100644 --- a/drivers/gpu/drm/udl/udl_drv.h +++ b/drivers/gpu/drm/udl/udl_drv.h @@ -113,7 +113,7 @@ udl_fb_user_fb_create(struct drm_device *dev, struct drm_file *file, const struct drm_mode_fb_cmd2 *mode_cmd); -int udl_render_hline(struct drm_device *dev, int bpp, struct urb **urb_ptr, +int udl_render_hline(struct drm_device *dev, int log_bpp, struct urb **urb_ptr, const char *front, char **urb_buf_ptr, u32 byte_offset, u32 device_byte_offset, u32 byte_width, int *ident_ptr, int *sent_ptr); diff --git a/drivers/gpu/drm/udl/udl_fb.c b/drivers/gpu/drm/udl/udl_fb.c index 2ebdc6d5a76e..dbb62f6eb48a 100644 --- a/drivers/gpu/drm/udl/udl_fb.c +++ b/drivers/gpu/drm/udl/udl_fb.c @@ -90,7 +90,10 @@ int udl_handle_damage(struct udl_framebuffer *fb, int x, int y, int bytes_identical = 0; struct urb *urb; int aligned_x; - int bpp = fb->base.format->cpp[0]; + int log_bpp; + + BUG_ON(!is_power_of_2(fb->base.format->cpp[0])); + log_bpp = __ffs(fb->base.format->cpp[0]); if (!fb->active_16) return 0; @@ -125,19 +128,22 @@ int udl_handle_damage(struct udl_framebuffer *fb, int x, int y, for (i = y; i < y + height ; i++) { const int line_offset = fb->base.pitches[0] * i; - const int byte_offset = line_offset + (x * bpp); - const int dev_byte_offset = (fb->base.width * bpp * i) + (x * bpp); - if (udl_render_hline(dev, bpp, &urb, + const int byte_offset = line_offset + (x << log_bpp); + const int dev_byte_offset = (fb->base.width * i + x) << log_bpp; + if (udl_render_hline(dev, log_bpp, &urb, (char *) fb->obj->vmapping, &cmd, byte_offset, dev_byte_offset, - width * bpp, + width << log_bpp, &bytes_identical, &bytes_sent)) goto error; } if (cmd > (char *) urb->transfer_buffer) { /* Send partial buffer remaining before exiting */ - int len = cmd - (char *) urb->transfer_buffer; + int len; + if (cmd < (char *) urb->transfer_buffer + urb->transfer_buffer_length) + *cmd++ = 0xAF; + len = cmd - (char *) urb->transfer_buffer; ret = udl_submit_urb(dev, urb, len); bytes_sent += len; } else @@ -146,7 +152,7 @@ int udl_handle_damage(struct udl_framebuffer *fb, int x, int y, error: atomic_add(bytes_sent, &udl->bytes_sent); atomic_add(bytes_identical, &udl->bytes_identical); - atomic_add(width*height*bpp, &udl->bytes_rendered); + atomic_add((width * height) << log_bpp, &udl->bytes_rendered); end_cycles = get_cycles(); atomic_add(((unsigned int) ((end_cycles - start_cycles) >> 10)), /* Kcycles */ @@ -172,7 +178,7 @@ static int udl_fb_mmap(struct fb_info *info, struct vm_area_struct *vma) pos = (unsigned long)info->fix.smem_start + offset; - pr_notice("mmap() framebuffer addr:%lu size:%lu\n", + pr_debug("mmap() framebuffer addr:%lu size:%lu\n", pos, size); /* We don't want the framebuffer to be mapped encrypted */ @@ -218,7 +224,7 @@ static int udl_fb_open(struct fb_info *info, int user) struct fb_deferred_io *fbdefio; - fbdefio = kmalloc(sizeof(struct fb_deferred_io), GFP_KERNEL); + fbdefio = kzalloc(sizeof(struct fb_deferred_io), GFP_KERNEL); if (fbdefio) { fbdefio->delay = DL_DEFIO_WRITE_DELAY; @@ -230,7 +236,7 @@ static int udl_fb_open(struct fb_info *info, int user) } #endif - pr_notice("open /dev/fb%d user=%d fb_info=%p count=%d\n", + pr_debug("open /dev/fb%d user=%d fb_info=%p count=%d\n", info->node, user, info, ufbdev->fb_count); return 0; @@ -255,7 +261,7 @@ static int udl_fb_release(struct fb_info *info, int user) } #endif - pr_warn("released /dev/fb%d user=%d count=%d\n", + pr_debug("released /dev/fb%d user=%d count=%d\n", info->node, user, ufbdev->fb_count); return 0; diff --git a/drivers/gpu/drm/udl/udl_main.c b/drivers/gpu/drm/udl/udl_main.c index d518de8f496b..f455f095a146 100644 --- a/drivers/gpu/drm/udl/udl_main.c +++ b/drivers/gpu/drm/udl/udl_main.c @@ -170,25 +170,19 @@ static void udl_free_urb_list(struct drm_device *dev) struct list_head *node; struct urb_node *unode; struct urb *urb; - int ret; - unsigned long flags; DRM_DEBUG("Waiting for completes and freeing all render urbs\n"); /* keep waiting and freeing, until we've got 'em all */ while (count--) { + down(&udl->urbs.limit_sem); - /* Getting interrupted means a leak, but ok at shutdown*/ - ret = down_interruptible(&udl->urbs.limit_sem); - if (ret) - break; - - spin_lock_irqsave(&udl->urbs.lock, flags); + spin_lock_irq(&udl->urbs.lock); node = udl->urbs.list.next; /* have reserved one with sem */ list_del_init(node); - spin_unlock_irqrestore(&udl->urbs.lock, flags); + spin_unlock_irq(&udl->urbs.lock); unode = list_entry(node, struct urb_node, entry); urb = unode->urb; @@ -205,17 +199,22 @@ static void udl_free_urb_list(struct drm_device *dev) static int udl_alloc_urb_list(struct drm_device *dev, int count, size_t size) { struct udl_device *udl = dev->dev_private; - int i = 0; struct urb *urb; struct urb_node *unode; char *buf; + size_t wanted_size = count * size; spin_lock_init(&udl->urbs.lock); +retry: udl->urbs.size = size; INIT_LIST_HEAD(&udl->urbs.list); - while (i < count) { + sema_init(&udl->urbs.limit_sem, 0); + udl->urbs.count = 0; + udl->urbs.available = 0; + + while (udl->urbs.count * size < wanted_size) { unode = kzalloc(sizeof(struct urb_node), GFP_KERNEL); if (!unode) break; @@ -231,11 +230,16 @@ static int udl_alloc_urb_list(struct drm_device *dev, int count, size_t size) } unode->urb = urb; - buf = usb_alloc_coherent(udl->udev, MAX_TRANSFER, GFP_KERNEL, + buf = usb_alloc_coherent(udl->udev, size, GFP_KERNEL, &urb->transfer_dma); if (!buf) { kfree(unode); usb_free_urb(urb); + if (size > PAGE_SIZE) { + size /= 2; + udl_free_urb_list(dev); + goto retry; + } break; } @@ -246,16 +250,14 @@ static int udl_alloc_urb_list(struct drm_device *dev, int count, size_t size) list_add_tail(&unode->entry, &udl->urbs.list); - i++; + up(&udl->urbs.limit_sem); + udl->urbs.count++; + udl->urbs.available++; } - sema_init(&udl->urbs.limit_sem, i); - udl->urbs.count = i; - udl->urbs.available = i; - - DRM_DEBUG("allocated %d %d byte urbs\n", i, (int) size); + DRM_DEBUG("allocated %d %d byte urbs\n", udl->urbs.count, (int) size); - return i; + return udl->urbs.count; } struct urb *udl_get_urb(struct drm_device *dev) @@ -265,7 +267,6 @@ struct urb *udl_get_urb(struct drm_device *dev) struct list_head *entry; struct urb_node *unode; struct urb *urb = NULL; - unsigned long flags; /* Wait for an in-flight buffer to complete and get re-queued */ ret = down_timeout(&udl->urbs.limit_sem, GET_URB_TIMEOUT); @@ -276,14 +277,14 @@ struct urb *udl_get_urb(struct drm_device *dev) goto error; } - spin_lock_irqsave(&udl->urbs.lock, flags); + spin_lock_irq(&udl->urbs.lock); BUG_ON(list_empty(&udl->urbs.list)); /* reserved one with limit_sem */ entry = udl->urbs.list.next; list_del_init(entry); udl->urbs.available--; - spin_unlock_irqrestore(&udl->urbs.lock, flags); + spin_unlock_irq(&udl->urbs.lock); unode = list_entry(entry, struct urb_node, entry); urb = unode->urb; diff --git a/drivers/gpu/drm/udl/udl_modeset.c b/drivers/gpu/drm/udl/udl_modeset.c index 5bcae7649795..7e37765cf5ac 100644 --- a/drivers/gpu/drm/udl/udl_modeset.c +++ b/drivers/gpu/drm/udl/udl_modeset.c @@ -243,7 +243,7 @@ static int udl_crtc_write_mode_to_hw(struct drm_crtc *crtc) memcpy(buf, udl->mode_buf, udl->mode_buf_len); retval = udl_submit_urb(dev, urb, udl->mode_buf_len); - DRM_INFO("write mode info %d\n", udl->mode_buf_len); + DRM_DEBUG("write mode info %d\n", udl->mode_buf_len); return retval; } @@ -366,7 +366,6 @@ static int udl_crtc_page_flip(struct drm_crtc *crtc, { struct udl_framebuffer *ufb = to_udl_fb(fb); struct drm_device *dev = crtc->dev; - unsigned long flags; struct drm_framebuffer *old_fb = crtc->primary->fb; if (old_fb) { @@ -377,10 +376,10 @@ static int udl_crtc_page_flip(struct drm_crtc *crtc, udl_handle_damage(ufb, 0, 0, fb->width, fb->height); - spin_lock_irqsave(&dev->event_lock, flags); + spin_lock_irq(&dev->event_lock); if (event) drm_crtc_send_vblank_event(crtc, event); - spin_unlock_irqrestore(&dev->event_lock, flags); + spin_unlock_irq(&dev->event_lock); crtc->primary->fb = fb; return 0; diff --git a/drivers/gpu/drm/udl/udl_transfer.c b/drivers/gpu/drm/udl/udl_transfer.c index 0c87b1ac6b68..ce87661e544f 100644 --- a/drivers/gpu/drm/udl/udl_transfer.c +++ b/drivers/gpu/drm/udl/udl_transfer.c @@ -13,7 +13,6 @@ #include <linux/module.h> #include <linux/slab.h> #include <linux/fb.h> -#include <linux/prefetch.h> #include <asm/unaligned.h> #include <drm/drmP.h> @@ -51,9 +50,6 @@ static int udl_trim_hline(const u8 *bback, const u8 **bfront, int *width_bytes) int start = width; int end = width; - prefetch((void *) front); - prefetch((void *) back); - for (j = 0; j < width; j++) { if (back[j] != front[j]) { start = j; @@ -83,12 +79,12 @@ static inline u16 pixel32_to_be16(const uint32_t pixel) ((pixel >> 8) & 0xf800)); } -static inline u16 get_pixel_val16(const uint8_t *pixel, int bpp) +static inline u16 get_pixel_val16(const uint8_t *pixel, int log_bpp) { - u16 pixel_val16 = 0; - if (bpp == 2) + u16 pixel_val16; + if (log_bpp == 1) pixel_val16 = *(const uint16_t *)pixel; - else if (bpp == 4) + else pixel_val16 = pixel32_to_be16(*(const uint32_t *)pixel); return pixel_val16; } @@ -125,8 +121,9 @@ static void udl_compress_hline16( const u8 *const pixel_end, uint32_t *device_address_ptr, uint8_t **command_buffer_ptr, - const uint8_t *const cmd_buffer_end, int bpp) + const uint8_t *const cmd_buffer_end, int log_bpp) { + const int bpp = 1 << log_bpp; const u8 *pixel = *pixel_start_ptr; uint32_t dev_addr = *device_address_ptr; uint8_t *cmd = *command_buffer_ptr; @@ -139,8 +136,6 @@ static void udl_compress_hline16( const u8 *cmd_pixel_start, *cmd_pixel_end = NULL; uint16_t pixel_val16; - prefetchw((void *) cmd); /* pull in one cache line at least */ - *cmd++ = 0xaf; *cmd++ = 0x6b; *cmd++ = (uint8_t) ((dev_addr >> 16) & 0xFF); @@ -153,12 +148,11 @@ static void udl_compress_hline16( raw_pixels_count_byte = cmd++; /* we'll know this later */ raw_pixel_start = pixel; - cmd_pixel_end = pixel + (min(MAX_CMD_PIXELS + 1, - min((int)(pixel_end - pixel) / bpp, - (int)(cmd_buffer_end - cmd) / 2))) * bpp; + cmd_pixel_end = pixel + (min3(MAX_CMD_PIXELS + 1UL, + (unsigned long)(pixel_end - pixel) >> log_bpp, + (unsigned long)(cmd_buffer_end - 1 - cmd) / 2) << log_bpp); - prefetch_range((void *) pixel, (cmd_pixel_end - pixel) * bpp); - pixel_val16 = get_pixel_val16(pixel, bpp); + pixel_val16 = get_pixel_val16(pixel, log_bpp); while (pixel < cmd_pixel_end) { const u8 *const start = pixel; @@ -170,7 +164,7 @@ static void udl_compress_hline16( pixel += bpp; while (pixel < cmd_pixel_end) { - pixel_val16 = get_pixel_val16(pixel, bpp); + pixel_val16 = get_pixel_val16(pixel, log_bpp); if (pixel_val16 != repeating_pixel_val16) break; pixel += bpp; @@ -179,10 +173,10 @@ static void udl_compress_hline16( if (unlikely(pixel > start + bpp)) { /* go back and fill in raw pixel count */ *raw_pixels_count_byte = (((start - - raw_pixel_start) / bpp) + 1) & 0xFF; + raw_pixel_start) >> log_bpp) + 1) & 0xFF; /* immediately after raw data is repeat byte */ - *cmd++ = (((pixel - start) / bpp) - 1) & 0xFF; + *cmd++ = (((pixel - start) >> log_bpp) - 1) & 0xFF; /* Then start another raw pixel span */ raw_pixel_start = pixel; @@ -192,11 +186,14 @@ static void udl_compress_hline16( if (pixel > raw_pixel_start) { /* finalize last RAW span */ - *raw_pixels_count_byte = ((pixel-raw_pixel_start) / bpp) & 0xFF; + *raw_pixels_count_byte = ((pixel - raw_pixel_start) >> log_bpp) & 0xFF; + } else { + /* undo unused byte */ + cmd--; } - *cmd_pixels_count_byte = ((pixel - cmd_pixel_start) / bpp) & 0xFF; - dev_addr += ((pixel - cmd_pixel_start) / bpp) * 2; + *cmd_pixels_count_byte = ((pixel - cmd_pixel_start) >> log_bpp) & 0xFF; + dev_addr += ((pixel - cmd_pixel_start) >> log_bpp) * 2; } if (cmd_buffer_end <= MIN_RLX_CMD_BYTES + cmd) { @@ -219,19 +216,19 @@ static void udl_compress_hline16( * (that we can only write to, slowly, and can never read), and (optionally) * our shadow copy that tracks what's been sent to that hardware buffer. */ -int udl_render_hline(struct drm_device *dev, int bpp, struct urb **urb_ptr, +int udl_render_hline(struct drm_device *dev, int log_bpp, struct urb **urb_ptr, const char *front, char **urb_buf_ptr, u32 byte_offset, u32 device_byte_offset, u32 byte_width, int *ident_ptr, int *sent_ptr) { const u8 *line_start, *line_end, *next_pixel; - u32 base16 = 0 + (device_byte_offset / bpp) * 2; + u32 base16 = 0 + (device_byte_offset >> log_bpp) * 2; struct urb *urb = *urb_ptr; u8 *cmd = *urb_buf_ptr; u8 *cmd_end = (u8 *) urb->transfer_buffer + urb->transfer_buffer_length; - BUG_ON(!(bpp == 2 || bpp == 4)); + BUG_ON(!(log_bpp == 1 || log_bpp == 2)); line_start = (u8 *) (front + byte_offset); next_pixel = line_start; @@ -241,7 +238,7 @@ int udl_render_hline(struct drm_device *dev, int bpp, struct urb **urb_ptr, udl_compress_hline16(&next_pixel, line_end, &base16, - (u8 **) &cmd, (u8 *) cmd_end, bpp); + (u8 **) &cmd, (u8 *) cmd_end, log_bpp); if (cmd >= cmd_end) { int len = cmd - (u8 *) urb->transfer_buffer; |