mmdeploy/cmake/cuda.cmake

58 lines
2.3 KiB
CMake
Raw Normal View History

Merge sdk (#251) * check in cmake * move backend_ops to csrc/backend_ops * check in preprocess, model, some codebase and their c-apis * check in CMakeLists.txt * check in parts of test_csrc * commit everything else * add readme * update core's BUILD_INTERFACE directory * skip codespell on third_party * update trt_net and ort_net's CMakeLists * ignore clion's build directory * check in pybind11 * add onnx.proto. Remove MMDeploy's dependency on ncnn's source code * export MMDeployTargets only when MMDEPLOY_BUILD_SDK is ON * remove useless message * target include directory is wrong * change target name from mmdeploy_ppl_net to mmdeploy_pplnn_net * skip install directory * update project's cmake * remove useless code * set CMAKE_BUILD_TYPE to Release by force if it isn't set by user * update custom ops CMakeLists * pass object target's source lists * fix lint end-of-file * fix lint: trailing whitespace * fix codespell hook * remove bicubic_interpolate to csrc/backend_ops/ * set MMDEPLOY_BUILD_SDK OFF * change custom ops build command * add spdlog installation command * update docs on how to checkout pybind11 * move bicubic_interpolate to backend_ops/tensorrt directory * remove useless code * correct cmake * fix typo * fix typo * fix install directory * correct sdk's readme * set cub dir when cuda version < 11.0 * change directory where clang-format will apply to * fix build command * add .clang-format * change clang-format style from google to file * reformat csrc/backend_ops * format sdk's code * turn off clang-format for some files * add -Xcompiler=-fno-gnu-unique * fix trt topk initialize * check in config for sdk demo * update cmake script and csrc's readme * correct config's path * add cuda include directory, otherwise compile failed in case of tensorrt8.2 * clang-format onnx2ncnn.cpp Co-authored-by: zhangli <lzhang329@gmail.com> Co-authored-by: grimoire <yaoqian@sensetime.com>
2021-12-07 10:57:55 +08:00
# Copyright (c) OpenMMLab. All rights reserved.
#message(STATUS "CMAKE_VERSION: ${CMAKE_VERSION}")
if (${CMAKE_VERSION} VERSION_GREATER_EQUAL "3.18.0")
# suppress 'CMAKE_CUDA_ARCHITECTURES' warning
cmake_policy(SET CMP0104 OLD)
endif ()
# nvcc compiler settings
find_package(CUDA REQUIRED)
#message(STATUS "CUDA VERSION: ${CUDA_VERSION_STRING}")
if (MSVC)
set(CMAKE_CUDA_COMPILER ${CUDA_TOOLKIT_ROOT_DIR}/bin/nvcc.exe)
set(CUDA_NVCC_FLAGS "${CUDA_NVCC_FLAGS} -Xcompiler=/wd4819,/wd4828")
if (HAVE_CXX_FLAG_UTF_8)
set(CUDA_NVCC_FLAGS "${CUDA_NVCC_FLAGS} -Xcompiler=/utf-8")
endif ()
else ()
set(CMAKE_CUDA_COMPILER ${CUDA_TOOLKIT_ROOT_DIR}/bin/nvcc)
# Explicitly set the cuda host compiler. Because the default host compiler #
# selected by cmake maybe wrong.
set(CMAKE_CUDA_HOST_COMPILER ${CMAKE_CXX_COMPILER})
set(CUDA_NVCC_FLAGS
"${CUDA_NVCC_FLAGS} -Xcompiler=-fPIC,-Wall,-fvisibility=hidden")
set(CUDA_NVCC_FLAGS "${CUDA_NVCC_FLAGS} -Xcompiler=-fno-gnu-unique")
endif ()
enable_language(CUDA)
# set virtual compute architecture and real ones
set(_NVCC_FLAGS)
set(_NVCC_FLAGS "${_NVCC_FLAGS} -gencode arch=compute_52,code=sm_52")
if (CUDA_VERSION_MAJOR VERSION_GREATER_EQUAL "8")
set(_NVCC_FLAGS "${_NVCC_FLAGS} -gencode arch=compute_60,code=sm_60")
set(_NVCC_FLAGS "${_NVCC_FLAGS} -gencode arch=compute_61,code=sm_61")
endif ()
if (CUDA_VERSION_MAJOR VERSION_GREATER_EQUAL "9")
set(_NVCC_FLAGS "${_NVCC_FLAGS} -gencode arch=compute_70,code=sm_70")
endif ()
if (CUDA_VERSION_MAJOR VERSION_GREATER_EQUAL "10")
set(_NVCC_FLAGS "${_NVCC_FLAGS} -gencode arch=compute_72,code=sm_72")
set(_NVCC_FLAGS "${_NVCC_FLAGS} -gencode arch=compute_75,code=sm_75")
endif ()
if (CUDA_VERSION_MAJOR VERSION_GREATER_EQUAL "11")
set(_NVCC_FLAGS "${_NVCC_FLAGS} -gencode arch=compute_80,code=sm_80")
if (CUDA_VERSION_MINOR VERSION_GREATER_EQUAL "1")
# cuda doesn't support `sm_86` until version 11.1
set(_NVCC_FLAGS "${_NVCC_FLAGS} -gencode arch=compute_86,code=sm_86")
endif ()
endif ()
set(CUDA_NVCC_FLAGS_DEBUG "-g -O0")
set(CUDA_NVCC_FLAGS_RELEASE "-O3")
set(CUDA_NVCC_FLAGS "${CUDA_NVCC_FLAGS}")
set(CMAKE_CUDA_STANDARD 14)
set(CMAKE_CUDA_FLAGS "${CMAKE_CUDA_FLAGS} ${CUDA_NVCC_FLAGS} ${_NVCC_FLAGS}")