diff options
Diffstat (limited to 'drivers/gpu/drm/amd/amdgpu')
22 files changed, 227 insertions, 126 deletions
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c index 8a707d8bbb1c..f92597c292fe 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c @@ -206,11 +206,9 @@ static int amdgpu_amdkfd_remove_eviction_fence(struct amdgpu_bo *bo, struct amdgpu_amdkfd_fence ***ef_list, unsigned int *ef_count) { - struct reservation_object_list *fobj; - struct reservation_object *resv; - unsigned int i = 0, j = 0, k = 0, shared_count; - unsigned int count = 0; - struct amdgpu_amdkfd_fence **fence_list; + struct reservation_object *resv = bo->tbo.resv; + struct reservation_object_list *old, *new; + unsigned int i, j, k; if (!ef && !ef_list) return -EINVAL; @@ -220,76 +218,67 @@ static int amdgpu_amdkfd_remove_eviction_fence(struct amdgpu_bo *bo, *ef_count = 0; } - resv = bo->tbo.resv; - fobj = reservation_object_get_list(resv); - - if (!fobj) + old = reservation_object_get_list(resv); + if (!old) return 0; - preempt_disable(); - write_seqcount_begin(&resv->seq); + new = kmalloc(offsetof(typeof(*new), shared[old->shared_max]), + GFP_KERNEL); + if (!new) + return -ENOMEM; - /* Go through all the shared fences in the resevation object. If - * ef is specified and it exists in the list, remove it and reduce the - * count. If ef is not specified, then get the count of eviction fences - * present. + /* Go through all the shared fences in the resevation object and sort + * the interesting ones to the end of the list. */ - shared_count = fobj->shared_count; - for (i = 0; i < shared_count; ++i) { + for (i = 0, j = old->shared_count, k = 0; i < old->shared_count; ++i) { struct dma_fence *f; - f = rcu_dereference_protected(fobj->shared[i], + f = rcu_dereference_protected(old->shared[i], reservation_object_held(resv)); - if (ef) { - if (f->context == ef->base.context) { - dma_fence_put(f); - fobj->shared_count--; - } else { - RCU_INIT_POINTER(fobj->shared[j++], f); - } - } else if (to_amdgpu_amdkfd_fence(f)) - count++; + if ((ef && f->context == ef->base.context) || + (!ef && to_amdgpu_amdkfd_fence(f))) + RCU_INIT_POINTER(new->shared[--j], f); + else + RCU_INIT_POINTER(new->shared[k++], f); } - write_seqcount_end(&resv->seq); - preempt_enable(); - - if (ef || !count) - return 0; - - /* Alloc memory for count number of eviction fence pointers. Fill the - * ef_list array and ef_count - */ - fence_list = kcalloc(count, sizeof(struct amdgpu_amdkfd_fence *), - GFP_KERNEL); - if (!fence_list) - return -ENOMEM; + new->shared_max = old->shared_max; + new->shared_count = k; - preempt_disable(); - write_seqcount_begin(&resv->seq); + if (!ef) { + unsigned int count = old->shared_count - j; - j = 0; - for (i = 0; i < shared_count; ++i) { - struct dma_fence *f; - struct amdgpu_amdkfd_fence *efence; - - f = rcu_dereference_protected(fobj->shared[i], - reservation_object_held(resv)); + /* Alloc memory for count number of eviction fence pointers. + * Fill the ef_list array and ef_count + */ + *ef_list = kcalloc(count, sizeof(**ef_list), GFP_KERNEL); + *ef_count = count; - efence = to_amdgpu_amdkfd_fence(f); - if (efence) { - fence_list[k++] = efence; - fobj->shared_count--; - } else { - RCU_INIT_POINTER(fobj->shared[j++], f); + if (!*ef_list) { + kfree(new); + return -ENOMEM; } } + /* Install the new fence list, seqcount provides the barriers */ + preempt_disable(); + write_seqcount_begin(&resv->seq); + RCU_INIT_POINTER(resv->fence, new); write_seqcount_end(&resv->seq); preempt_enable(); - *ef_list = fence_list; - *ef_count = k; + /* Drop the references to the removed fences or move them to ef_list */ + for (i = j, k = 0; i < old->shared_count; ++i) { + struct dma_fence *f; + + f = rcu_dereference_protected(new->shared[i], + reservation_object_held(resv)); + if (!ef) + (*ef_list)[k++] = to_amdgpu_amdkfd_fence(f); + else + dma_fence_put(f); + } + kfree_rcu(old, rcu); return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index 1e66dfd0e39c..8ab5ccbc14ac 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -2274,7 +2274,7 @@ bool amdgpu_device_asic_has_dc_support(enum amd_asic_type asic_type) case CHIP_VEGA10: case CHIP_VEGA12: case CHIP_VEGA20: -#ifdef CONFIG_X86 +#if defined(CONFIG_DRM_AMD_DC_DCN1_0) case CHIP_RAVEN: #endif return amdgpu_dc != 0; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c index a365ea2383d1..e55508b39496 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c @@ -178,12 +178,18 @@ void amdgpu_mn_unlock(struct amdgpu_mn *mn) * * @amn: our notifier */ -static void amdgpu_mn_read_lock(struct amdgpu_mn *amn) +static int amdgpu_mn_read_lock(struct amdgpu_mn *amn, bool blockable) { - mutex_lock(&amn->read_lock); + if (blockable) + mutex_lock(&amn->read_lock); + else if (!mutex_trylock(&amn->read_lock)) + return -EAGAIN; + if (atomic_inc_return(&amn->recursion) == 1) down_read_non_owner(&amn->lock); mutex_unlock(&amn->read_lock); + + return 0; } /** @@ -239,10 +245,11 @@ static void amdgpu_mn_invalidate_node(struct amdgpu_mn_node *node, * Block for operations on BOs to finish and mark pages as accessed and * potentially dirty. */ -static void amdgpu_mn_invalidate_range_start_gfx(struct mmu_notifier *mn, +static int amdgpu_mn_invalidate_range_start_gfx(struct mmu_notifier *mn, struct mm_struct *mm, unsigned long start, - unsigned long end) + unsigned long end, + bool blockable) { struct amdgpu_mn *amn = container_of(mn, struct amdgpu_mn, mn); struct interval_tree_node *it; @@ -250,17 +257,28 @@ static void amdgpu_mn_invalidate_range_start_gfx(struct mmu_notifier *mn, /* notification is exclusive, but interval is inclusive */ end -= 1; - amdgpu_mn_read_lock(amn); + /* TODO we should be able to split locking for interval tree and + * amdgpu_mn_invalidate_node + */ + if (amdgpu_mn_read_lock(amn, blockable)) + return -EAGAIN; it = interval_tree_iter_first(&amn->objects, start, end); while (it) { struct amdgpu_mn_node *node; + if (!blockable) { + amdgpu_mn_read_unlock(amn); + return -EAGAIN; + } + node = container_of(it, struct amdgpu_mn_node, it); it = interval_tree_iter_next(it, start, end); amdgpu_mn_invalidate_node(node, start, end); } + + return 0; } /** @@ -275,10 +293,11 @@ static void amdgpu_mn_invalidate_range_start_gfx(struct mmu_notifier *mn, * necessitates evicting all user-mode queues of the process. The BOs * are restorted in amdgpu_mn_invalidate_range_end_hsa. */ -static void amdgpu_mn_invalidate_range_start_hsa(struct mmu_notifier *mn, +static int amdgpu_mn_invalidate_range_start_hsa(struct mmu_notifier *mn, struct mm_struct *mm, unsigned long start, - unsigned long end) + unsigned long end, + bool blockable) { struct amdgpu_mn *amn = container_of(mn, struct amdgpu_mn, mn); struct interval_tree_node *it; @@ -286,13 +305,19 @@ static void amdgpu_mn_invalidate_range_start_hsa(struct mmu_notifier *mn, /* notification is exclusive, but interval is inclusive */ end -= 1; - amdgpu_mn_read_lock(amn); + if (amdgpu_mn_read_lock(amn, blockable)) + return -EAGAIN; it = interval_tree_iter_first(&amn->objects, start, end); while (it) { struct amdgpu_mn_node *node; struct amdgpu_bo *bo; + if (!blockable) { + amdgpu_mn_read_unlock(amn); + return -EAGAIN; + } + node = container_of(it, struct amdgpu_mn_node, it); it = interval_tree_iter_next(it, start, end); @@ -304,6 +329,8 @@ static void amdgpu_mn_invalidate_range_start_hsa(struct mmu_notifier *mn, amdgpu_amdkfd_evict_userptr(mem, mm); } } + + return 0; } /** diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c index 9f1a5bd39ae8..5b39d1399630 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c @@ -131,6 +131,11 @@ psp_cmd_submit_buf(struct psp_context *psp, msleep(1); } + if (ucode) { + ucode->tmr_mc_addr_lo = psp->cmd_buf_mem->resp.fw_addr_lo; + ucode->tmr_mc_addr_hi = psp->cmd_buf_mem->resp.fw_addr_hi; + } + return ret; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_sched.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_sched.c index 86a0715d9431..1cafe8d83a4d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_sched.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_sched.c @@ -53,9 +53,8 @@ static int amdgpu_sched_process_priority_override(struct amdgpu_device *adev, int fd, enum drm_sched_priority priority) { - struct file *filp = fcheck(fd); + struct file *filp = fget(fd); struct drm_file *file; - struct pid *pid; struct amdgpu_fpriv *fpriv; struct amdgpu_ctx *ctx; uint32_t id; @@ -63,20 +62,12 @@ static int amdgpu_sched_process_priority_override(struct amdgpu_device *adev, if (!filp) return -EINVAL; - pid = get_pid(((struct drm_file *)filp->private_data)->pid); + file = filp->private_data; + fpriv = file->driver_priv; + idr_for_each_entry(&fpriv->ctx_mgr.ctx_handles, ctx, id) + amdgpu_ctx_priority_override(ctx, priority); - mutex_lock(&adev->ddev->filelist_mutex); - list_for_each_entry(file, &adev->ddev->filelist, lhead) { - if (file->pid != pid) - continue; - - fpriv = file->driver_priv; - idr_for_each_entry(&fpriv->ctx_mgr.ctx_handles, ctx, id) - amdgpu_ctx_priority_override(ctx, priority); - } - mutex_unlock(&adev->ddev->filelist_mutex); - - put_pid(pid); + fput(filp); return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h index 08e38579af24..bdc472b6e641 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h @@ -194,6 +194,7 @@ enum AMDGPU_UCODE_ID { AMDGPU_UCODE_ID_SMC, AMDGPU_UCODE_ID_UVD, AMDGPU_UCODE_ID_VCE, + AMDGPU_UCODE_ID_VCN, AMDGPU_UCODE_ID_MAXIMUM, }; @@ -226,6 +227,9 @@ struct amdgpu_firmware_info { void *kaddr; /* ucode_size_bytes */ uint32_t ucode_size; + /* starting tmr mc address */ + uint32_t tmr_mc_addr_lo; + uint32_t tmr_mc_addr_hi; }; void amdgpu_ucode_print_mc_hdr(const struct common_firmware_header *hdr); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c index 632fa5980ff4..e5a6db6beab7 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c @@ -122,8 +122,6 @@ static void amdgpu_uvd_idle_work_handler(struct work_struct *work); int amdgpu_uvd_sw_init(struct amdgpu_device *adev) { - struct amdgpu_ring *ring; - struct drm_sched_rq *rq; unsigned long bo_size; const char *fw_name; const struct common_firmware_header *hdr; @@ -266,13 +264,6 @@ int amdgpu_uvd_sw_init(struct amdgpu_device *adev) } } - ring = &adev->uvd.inst[0].ring; - rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL]; - r = drm_sched_entity_init(&adev->uvd.entity, &rq, 1, NULL); - if (r) { - DRM_ERROR("Failed setting up UVD kernel entity.\n"); - return r; - } for (i = 0; i < adev->uvd.max_handles; ++i) { atomic_set(&adev->uvd.handles[i], 0); adev->uvd.filp[i] = NULL; @@ -311,7 +302,7 @@ int amdgpu_uvd_sw_fini(struct amdgpu_device *adev) for (j = 0; j < adev->uvd.num_uvd_inst; ++j) { if (adev->uvd.harvest_config & (1 << j)) continue; - kfree(adev->uvd.inst[j].saved_bo); + kvfree(adev->uvd.inst[j].saved_bo); amdgpu_bo_free_kernel(&adev->uvd.inst[j].vcpu_bo, &adev->uvd.inst[j].gpu_addr, @@ -327,6 +318,29 @@ int amdgpu_uvd_sw_fini(struct amdgpu_device *adev) return 0; } +/** + * amdgpu_uvd_entity_init - init entity + * + * @adev: amdgpu_device pointer + * + */ +int amdgpu_uvd_entity_init(struct amdgpu_device *adev) +{ + struct amdgpu_ring *ring; + struct drm_sched_rq *rq; + int r; + + ring = &adev->uvd.inst[0].ring; + rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL]; + r = drm_sched_entity_init(&adev->uvd.entity, &rq, 1, NULL); + if (r) { + DRM_ERROR("Failed setting up UVD kernel entity.\n"); + return r; + } + + return 0; +} + int amdgpu_uvd_suspend(struct amdgpu_device *adev) { unsigned size; @@ -354,7 +368,7 @@ int amdgpu_uvd_suspend(struct amdgpu_device *adev) size = amdgpu_bo_size(adev->uvd.inst[j].vcpu_bo); ptr = adev->uvd.inst[j].cpu_addr; - adev->uvd.inst[j].saved_bo = kmalloc(size, GFP_KERNEL); + adev->uvd.inst[j].saved_bo = kvmalloc(size, GFP_KERNEL); if (!adev->uvd.inst[j].saved_bo) return -ENOMEM; @@ -380,7 +394,7 @@ int amdgpu_uvd_resume(struct amdgpu_device *adev) if (adev->uvd.inst[i].saved_bo != NULL) { memcpy_toio(ptr, adev->uvd.inst[i].saved_bo, size); - kfree(adev->uvd.inst[i].saved_bo); + kvfree(adev->uvd.inst[i].saved_bo); adev->uvd.inst[i].saved_bo = NULL; } else { const struct common_firmware_header *hdr; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.h index 33c5f806f925..a3ab1a41060f 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.h @@ -69,6 +69,7 @@ struct amdgpu_uvd { int amdgpu_uvd_sw_init(struct amdgpu_device *adev); int amdgpu_uvd_sw_fini(struct amdgpu_device *adev); +int amdgpu_uvd_entity_init(struct amdgpu_device *adev); int amdgpu_uvd_suspend(struct amdgpu_device *adev); int amdgpu_uvd_resume(struct amdgpu_device *adev); int amdgpu_uvd_get_create_msg(struct amdgpu_ring *ring, uint32_t handle, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c index b6ab4f5350c8..0cc5190f4f36 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c @@ -90,8 +90,6 @@ static void amdgpu_vce_idle_work_handler(struct work_struct *work); */ int amdgpu_vce_sw_init(struct amdgpu_device *adev, unsigned long size) { - struct amdgpu_ring *ring; - struct drm_sched_rq *rq; const char *fw_name; const struct common_firmware_header *hdr; unsigned ucode_version, version_major, version_minor, binary_id; @@ -188,14 +186,6 @@ int amdgpu_vce_sw_init(struct amdgpu_device *adev, unsigned long size) return r; } - ring = &adev->vce.ring[0]; - rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL]; - r = drm_sched_entity_init(&adev->vce.entity, &rq, 1, NULL); - if (r != 0) { - DRM_ERROR("Failed setting up VCE run queue.\n"); - return r; - } - for (i = 0; i < AMDGPU_MAX_VCE_HANDLES; ++i) { atomic_set(&adev->vce.handles[i], 0); adev->vce.filp[i] = NULL; @@ -236,6 +226,29 @@ int amdgpu_vce_sw_fini(struct amdgpu_device *adev) } /** + * amdgpu_vce_entity_init - init entity + * + * @adev: amdgpu_device pointer + * + */ +int amdgpu_vce_entity_init(struct amdgpu_device *adev) +{ + struct amdgpu_ring *ring; + struct drm_sched_rq *rq; + int r; + + ring = &adev->vce.ring[0]; + rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL]; + r = drm_sched_entity_init(&adev->vce.entity, &rq, 1, NULL); + if (r != 0) { + DRM_ERROR("Failed setting up VCE run queue.\n"); + return r; + } + + return 0; +} + +/** * amdgpu_vce_suspend - unpin VCE fw memory * * @adev: amdgpu_device pointer diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.h index 71781267ee4c..a1f209eed4c4 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.h @@ -55,6 +55,7 @@ struct amdgpu_vce { int amdgpu_vce_sw_init(struct amdgpu_device *adev, unsigned long size); int amdgpu_vce_sw_fini(struct amdgpu_device *adev); +int amdgpu_vce_entity_init(struct amdgpu_device *adev); int amdgpu_vce_suspend(struct amdgpu_device *adev); int amdgpu_vce_resume(struct amdgpu_device *adev); int amdgpu_vce_get_create_msg(struct amdgpu_ring *ring, uint32_t handle, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c index 798648a19710..fd654a4406db 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c @@ -111,9 +111,10 @@ int amdgpu_vcn_sw_init(struct amdgpu_device *adev) version_major, version_minor, family_id); } - bo_size = AMDGPU_GPU_PAGE_ALIGN(le32_to_cpu(hdr->ucode_size_bytes) + 8) - + AMDGPU_VCN_STACK_SIZE + AMDGPU_VCN_HEAP_SIZE + bo_size = AMDGPU_VCN_STACK_SIZE + AMDGPU_VCN_HEAP_SIZE + AMDGPU_VCN_SESSION_SIZE * 40; + if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP) + bo_size += AMDGPU_GPU_PAGE_ALIGN(le32_to_cpu(hdr->ucode_size_bytes) + 8); r = amdgpu_bo_create_kernel(adev, bo_size, PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM, &adev->vcn.vcpu_bo, &adev->vcn.gpu_addr, &adev->vcn.cpu_addr); @@ -129,7 +130,7 @@ int amdgpu_vcn_sw_fini(struct amdgpu_device *adev) { int i; - kfree(adev->vcn.saved_bo); + kvfree(adev->vcn.saved_bo); amdgpu_bo_free_kernel(&adev->vcn.vcpu_bo, &adev->vcn.gpu_addr, @@ -160,7 +161,7 @@ int amdgpu_vcn_suspend(struct amdgpu_device *adev) size = amdgpu_bo_size(adev->vcn.vcpu_bo); ptr = adev->vcn.cpu_addr; - adev->vcn.saved_bo = kmalloc(size, GFP_KERNEL); + adev->vcn.saved_bo = kvmalloc(size, GFP_KERNEL); if (!adev->vcn.saved_bo) return -ENOMEM; @@ -182,18 +183,20 @@ int amdgpu_vcn_resume(struct amdgpu_device *adev) if (adev->vcn.saved_bo != NULL) { memcpy_toio(ptr, adev->vcn.saved_bo, size); - kfree(adev->vcn.saved_bo); + kvfree(adev->vcn.saved_bo); adev->vcn.saved_bo = NULL; } else { const struct common_firmware_header *hdr; unsigned offset; hdr = (const struct common_firmware_header *)adev->vcn.fw->data; - offset = le32_to_cpu(hdr->ucode_array_offset_bytes); - memcpy_toio(adev->vcn.cpu_addr, adev->vcn.fw->data + offset, - le32_to_cpu(hdr->ucode_size_bytes)); - size -= le32_to_cpu(hdr->ucode_size_bytes); - ptr += le32_to_cpu(hdr->ucode_size_bytes); + if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP) { + offset = le32_to_cpu(hdr->ucode_array_offset_bytes); + memcpy_toio(adev->vcn.cpu_addr, adev->vcn.fw->data + offset, + le32_to_cpu(hdr->ucode_size_bytes)); + size -= le32_to_cpu(hdr->ucode_size_bytes); + ptr += le32_to_cpu(hdr->ucode_size_bytes); + } memset_io(ptr, 0, size); } diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c b/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c index 0ff136d02d9b..02be34e72ed9 100644 --- a/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c +++ b/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c @@ -88,6 +88,9 @@ psp_v10_0_get_fw_type(struct amdgpu_firmware_info *ucode, enum psp_gfx_fw_type * case AMDGPU_UCODE_ID_VCE: *type = GFX_FW_TYPE_VCE; break; + case AMDGPU_UCODE_ID_VCN: + *type = GFX_FW_TYPE_VCN; + break; case AMDGPU_UCODE_ID_MAXIMUM: default: return -EINVAL; diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c b/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c index 6fed3d7797a8..8a926d1df939 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c @@ -123,6 +123,10 @@ static int uvd_v4_2_sw_init(void *handle) ring = &adev->uvd.inst->ring; sprintf(ring->name, "uvd"); r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.inst->irq, 0); + if (r) + return r; + + r = amdgpu_uvd_entity_init(adev); return r; } diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c index aeaa1ca46a99..50248059412e 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c @@ -120,6 +120,10 @@ static int uvd_v5_0_sw_init(void *handle) ring = &adev->uvd.inst->ring; sprintf(ring->name, "uvd"); r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.inst->irq, 0); + if (r) + return r; + + r = amdgpu_uvd_entity_init(adev); return r; } diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c index 598dbeaba636..6ae82cc2e55e 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c @@ -440,6 +440,8 @@ static int uvd_v6_0_sw_init(void *handle) } } + r = amdgpu_uvd_entity_init(adev); + return r; } diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c index 5fab3560a71d..9b7f8469bc5c 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c @@ -410,6 +410,7 @@ static int uvd_v7_0_early_init(void *handle) static int uvd_v7_0_sw_init(void *handle) { struct amdgpu_ring *ring; + int i, j, r; struct amdgpu_device *adev = (struct amdgpu_device *)handle; @@ -478,6 +479,10 @@ static int uvd_v7_0_sw_init(void *handle) } } + r = amdgpu_uvd_entity_init(adev); + if (r) + return r; + r = amdgpu_virt_alloc_mm_table(adev); if (r) return r; diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c index d48e877b682e..7eaa54ba016b 100644 --- a/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c @@ -439,6 +439,8 @@ static int vce_v2_0_sw_init(void *handle) return r; } + r = amdgpu_vce_entity_init(adev); + return r; } diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c index cc6ce6cc03f4..c8390f9adfd6 100644 --- a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c @@ -448,6 +448,8 @@ static int vce_v3_0_sw_init(void *handle) return r; } + r = amdgpu_vce_entity_init(adev); + return r; } diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c index 65f8860169e9..2e4d1b5f6243 100644 --- a/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c @@ -419,6 +419,7 @@ static int vce_v4_0_sw_init(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; struct amdgpu_ring *ring; + unsigned size; int r, i; @@ -438,7 +439,7 @@ static int vce_v4_0_sw_init(void *handle) const struct common_firmware_header *hdr; unsigned size = amdgpu_bo_size(adev->vce.vcpu_bo); - adev->vce.saved_bo = kmalloc(size, GFP_KERNEL); + adev->vce.saved_bo = kvmalloc(size, GFP_KERNEL); if (!adev->vce.saved_bo) return -ENOMEM; @@ -474,6 +475,11 @@ static int vce_v4_0_sw_init(void *handle) return r; } + + r = amdgpu_vce_entity_init(adev); + if (r) + return r; + r = amdgpu_virt_alloc_mm_table(adev); if (r) return r; @@ -490,7 +496,7 @@ static int vce_v4_0_sw_fini(void *handle) amdgpu_virt_free_mm_table(adev); if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) { - kfree(adev->vce.saved_bo); + kvfree(adev->vce.saved_bo); adev->vce.saved_bo = NULL; } diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c index 2ce91a748c40..072371ef5975 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c @@ -100,6 +100,16 @@ static int vcn_v1_0_sw_init(void *handle) if (r) return r; + if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) { + const struct common_firmware_header *hdr; + hdr = (const struct common_firmware_header *)adev->vcn.fw->data; + adev->firmware.ucode[AMDGPU_UCODE_ID_VCN].ucode_id = AMDGPU_UCODE_ID_VCN; + adev->firmware.ucode[AMDGPU_UCODE_ID_VCN].fw = adev->vcn.fw; + adev->firmware.fw_size += + ALIGN(le32_to_cpu(hdr->ucode_size_bytes), PAGE_SIZE); + DRM_INFO("PSP loading VCN firmware\n"); + } + r = amdgpu_vcn_resume(adev); if (r) return r; @@ -265,26 +275,38 @@ static int vcn_v1_0_resume(void *handle) static void vcn_v1_0_mc_resume(struct amdgpu_device *adev) { uint32_t size = AMDGPU_GPU_PAGE_ALIGN(adev->vcn.fw->size + 4); - - WREG32_SOC15(UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW, + uint32_t offset; + + if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) { + WREG32_SOC15(UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW, + (adev->firmware.ucode[AMDGPU_UCODE_ID_VCN].tmr_mc_addr_lo)); + WREG32_SOC15(UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH, + (adev->firmware.ucode[AMDGPU_UCODE_ID_VCN].tmr_mc_addr_hi)); + WREG32_SOC15(UVD, 0, mmUVD_VCPU_CACHE_OFFSET0, 0); + offset = 0; + } else { + WREG32_SOC15(UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW, lower_32_bits(adev->vcn.gpu_addr)); - WREG32_SOC15(UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH, + WREG32_SOC15(UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH, upper_32_bits(adev->vcn.gpu_addr)); - WREG32_SOC15(UVD, 0, mmUVD_VCPU_CACHE_OFFSET0, - AMDGPU_UVD_FIRMWARE_OFFSET >> 3); + offset = size; + WREG32_SOC15(UVD, 0, mmUVD_VCPU_CACHE_OFFSET0, + AMDGPU_UVD_FIRMWARE_OFFSET >> 3); + } + WREG32_SOC15(UVD, 0, mmUVD_VCPU_CACHE_SIZE0, size); WREG32_SOC15(UVD, 0, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_LOW, - lower_32_bits(adev->vcn.gpu_addr + size)); + lower_32_bits(adev->vcn.gpu_addr + offset)); WREG32_SOC15(UVD, 0, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_HIGH, - upper_32_bits(adev->vcn.gpu_addr + size)); + upper_32_bits(adev->vcn.gpu_addr + offset)); WREG32_SOC15(UVD, 0, mmUVD_VCPU_CACHE_OFFSET1, 0); WREG32_SOC15(UVD, 0, mmUVD_VCPU_CACHE_SIZE1, AMDGPU_VCN_HEAP_SIZE); WREG32_SOC15(UVD, 0, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_LOW, - lower_32_bits(adev->vcn.gpu_addr + size + AMDGPU_VCN_HEAP_SIZE)); + lower_32_bits(adev->vcn.gpu_addr + offset + AMDGPU_VCN_HEAP_SIZE)); WREG32_SOC15(UVD, 0, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_HIGH, - upper_32_bits(adev->vcn.gpu_addr + size + AMDGPU_VCN_HEAP_SIZE)); + upper_32_bits(adev->vcn.gpu_addr + offset + AMDGPU_VCN_HEAP_SIZE)); WREG32_SOC15(UVD, 0, mmUVD_VCPU_CACHE_OFFSET2, 0); WREG32_SOC15(UVD, 0, mmUVD_VCPU_CACHE_SIZE2, AMDGPU_VCN_STACK_SIZE + (AMDGPU_VCN_SESSION_SIZE * 40)); diff --git a/drivers/gpu/drm/amd/amdgpu/vega20_reg_init.c b/drivers/gpu/drm/amd/amdgpu/vega20_reg_init.c index 52778de93ab0..2d4473557b0d 100644 --- a/drivers/gpu/drm/amd/amdgpu/vega20_reg_init.c +++ b/drivers/gpu/drm/amd/amdgpu/vega20_reg_init.c @@ -38,6 +38,7 @@ int vega20_reg_base_init(struct amdgpu_device *adev) adev->reg_offset[ATHUB_HWIP][i] = (uint32_t *)(&(ATHUB_BASE.instance[i])); adev->reg_offset[NBIO_HWIP][i] = (uint32_t *)(&(NBIO_BASE.instance[i])); adev->reg_offset[MP0_HWIP][i] = (uint32_t *)(&(MP0_BASE.instance[i])); + adev->reg_offset[MP1_HWIP][i] = (uint32_t *)(&(MP1_BASE.instance[i])); adev->reg_offset[UVD_HWIP][i] = (uint32_t *)(&(UVD_BASE.instance[i])); adev->reg_offset[VCE_HWIP][i] = (uint32_t *)(&(VCE_BASE.instance[i])); adev->reg_offset[DF_HWIP][i] = (uint32_t *)(&(DF_BASE.instance[i])); @@ -46,6 +47,8 @@ int vega20_reg_base_init(struct amdgpu_device *adev) adev->reg_offset[SDMA0_HWIP][i] = (uint32_t *)(&(SDMA0_BASE.instance[i])); adev->reg_offset[SDMA1_HWIP][i] = (uint32_t *)(&(SDMA1_BASE.instance[i])); adev->reg_offset[SMUIO_HWIP][i] = (uint32_t *)(&(SMUIO_BASE.instance[i])); + adev->reg_offset[NBIF_HWIP][i] = (uint32_t *)(&(NBIO_BASE.instance[i])); + adev->reg_offset[THM_HWIP][i] = (uint32_t *)(&(THM_BASE.instance[i])); } return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/vi.c b/drivers/gpu/drm/amd/amdgpu/vi.c index 42c8ad105b05..88b57a5e9489 100644 --- a/drivers/gpu/drm/amd/amdgpu/vi.c +++ b/drivers/gpu/drm/amd/amdgpu/vi.c @@ -112,8 +112,8 @@ static u32 vi_smc_rreg(struct amdgpu_device *adev, u32 reg) u32 r; spin_lock_irqsave(&adev->smc_idx_lock, flags); - WREG32(mmSMC_IND_INDEX_11, (reg)); - r = RREG32(mmSMC_IND_DATA_11); + WREG32_NO_KIQ(mmSMC_IND_INDEX_11, (reg)); + r = RREG32_NO_KIQ(mmSMC_IND_DATA_11); spin_unlock_irqrestore(&adev->smc_idx_lock, flags); return r; } |
