/* SPDX-License-Identifier: BSD-2-Clause OR GPL-2.0-only */ /* SPDX-FileCopyrightText: Copyright Amazon.com, Inc. or its affiliates. All rights reserved. */ #include "efa_unit_tests.h" static void test_efa_mr_impl(struct efa_domain *efa_domain, struct fid_mr *mr, int mr_reg_count, int mr_reg_size, bool gdrcopy_flag) { struct efa_mr *efa_mr; assert_int_equal(efa_domain->ibv_mr_reg_ct, mr_reg_count); assert_int_equal(efa_domain->ibv_mr_reg_sz, mr_reg_size); if (mr) { efa_mr = container_of(mr, struct efa_mr, mr_fid); if (cuda_is_gdrcopy_enabled()) { if (gdrcopy_flag) assert_true(efa_mr->peer.flags & OFI_HMEM_DATA_DEV_REG_HANDLE); else assert_false(efa_mr->peer.flags & OFI_HMEM_DATA_DEV_REG_HANDLE); } } } void test_efa_rdm_mr_reg_host_memory(struct efa_resource **state) { struct efa_resource *resource = *state; struct efa_domain *efa_domain; size_t mr_size = 64; void *buf; struct fid_mr *mr = NULL; efa_unit_test_resource_construct(resource, FI_EP_RDM, EFA_FABRIC_NAME); efa_domain = container_of(resource->domain, struct efa_domain, util_domain.domain_fid); buf = malloc(mr_size); assert_non_null(buf); test_efa_mr_impl(efa_domain, mr, 0, 0, false); assert_int_equal(fi_mr_reg(resource->domain, buf, mr_size, FI_SEND | FI_RECV, 0, 0, 0, &mr, NULL), 0); /* No GDRCopy registration for host memory */ test_efa_mr_impl(efa_domain, mr, 1, mr_size, false); assert_int_equal(fi_close(&mr->fid), 0); test_efa_mr_impl(efa_domain, NULL, 0, 0, false); free(buf); } #if HAVE_CUDA void test_efa_rdm_mr_reg_cuda_memory(struct efa_resource **state) { struct efa_resource *resource = *state; struct efa_domain *efa_domain; size_t mr_size = 64; void *buf; struct fid_mr *mr = NULL; struct fi_mr_attr mr_reg_attr = { 0 }; struct iovec iovec; int err; if (hmem_ops[FI_HMEM_CUDA].initialized && g_efa_hmem_info[FI_HMEM_CUDA].p2p_supported_by_device) { resource->hints = efa_unit_test_alloc_hints_hmem( FI_EP_RDM, EFA_FABRIC_NAME); efa_unit_test_resource_construct_with_hints(resource, FI_EP_RDM, FI_VERSION(2, 0), resource->hints, true, true); efa_domain = container_of(resource->domain, struct efa_domain, util_domain.domain_fid); test_efa_mr_impl(efa_domain, mr, 0, 0, false); err = ofi_cudaMalloc(&buf, mr_size); assert_int_equal(err, 0); assert_non_null(buf); mr_reg_attr.access = FI_SEND | FI_RECV; mr_reg_attr.iface = FI_HMEM_CUDA; iovec.iov_base = buf; iovec.iov_len = mr_size; mr_reg_attr.mr_iov = &iovec; mr_reg_attr.iov_count = 1; err = fi_mr_regattr(resource->domain, &mr_reg_attr, 0, &mr); assert_int_equal(err, 0); /* FI_MR_DMABUF flag was not set, so GDRCopy should be registered if available */ test_efa_mr_impl(efa_domain, mr, 1, mr_size, true); assert_int_equal(fi_close(&mr->fid), 0); test_efa_mr_impl(efa_domain, NULL, 0, 0, false); err = ofi_cudaFree(buf); assert_int_equal(err, 0); } } #else void test_efa_rdm_mr_reg_cuda_memory(struct efa_resource **state) { skip(); } #endif #if HAVE_CUDA void test_efa_direct_mr_reg_no_gdrcopy(struct efa_resource **state) { struct efa_resource *resource = *state; struct efa_domain *efa_domain; size_t mr_size = 64; void *buf; struct fid_mr *mr = NULL; struct fi_mr_attr mr_reg_attr = { 0 }; struct iovec iovec; int err; if (g_efa_hmem_info[FI_HMEM_CUDA].initialized && g_efa_hmem_info[FI_HMEM_CUDA].p2p_supported_by_device) { resource->hints = efa_unit_test_alloc_hints_hmem( FI_EP_RDM, EFA_DIRECT_FABRIC_NAME); efa_unit_test_resource_construct_with_hints(resource, FI_EP_RDM, FI_VERSION(2, 0), resource->hints, true, true); efa_domain = container_of(resource->domain, struct efa_domain, util_domain.domain_fid); test_efa_mr_impl(efa_domain, mr, 0, 0, false); err = ofi_cudaMalloc(&buf, mr_size); assert_int_equal(err, 0); assert_non_null(buf); mr_reg_attr.access = FI_SEND | FI_RECV; mr_reg_attr.iface = FI_HMEM_CUDA; iovec.iov_base = buf; iovec.iov_len = mr_size; mr_reg_attr.mr_iov = &iovec; mr_reg_attr.iov_count = 1; err = fi_mr_regattr(resource->domain, &mr_reg_attr, 0, &mr); assert_int_equal(err, 0); /* no GDRCopy in the efa-direct path */ test_efa_mr_impl(efa_domain, mr, 1, mr_size, false); assert_int_equal(fi_close(&mr->fid), 0); test_efa_mr_impl(efa_domain, NULL, 0, 0, false); err = ofi_cudaFree(buf); assert_int_equal(err, 0); } } #else void test_efa_direct_mr_reg_no_gdrcopy(struct efa_resource **state) { skip(); } #endif