56fd68e7dd
CCCL (CUDA C++ Core Libraries) provides: - CUB: device/block/warp-level GPU primitives (reduce, scan, sort, topk) - Thrust: high-level parallel algorithms (transform_reduce, sort, scan) - libcudacxx: CUDA C++ standard library (atomics, barriers, memory) - cudax: experimental features (memory resources, allocators) - Tuning policies: per-SM hardware-specific algorithm parameters Competition optimization vectors mapped to CCCL: - Output TPS (83% weight): warp_reduce, block_reduce, device_topk - Input TPS (14% weight): device_scan, block_load, prefetch - Cache TPS (3% weight): prefix caching strategy patterns - Memory (0.9 util): pooled/cached/buddy allocators Source: https://github.com/NVIDIA/cccl (shallow clone, HEAD only) License: Apache-2.0
55 lines
1.8 KiB
CMake
55 lines
1.8 KiB
CMake
# For every public header, build a translation unit containing `#include <header>`
|
|
# to let the compiler try to figure out warnings in that header if it is not otherwise
|
|
# included in tests, and also to verify if the headers are modular enough.
|
|
# .inl files are not globbed for, because they are not supposed to be used as public
|
|
# entrypoints.
|
|
|
|
if (NOT CUB_ENABLE_LAUNCH_NO_LAUNCHER)
|
|
# Header tests are treated as core-only artifacts.
|
|
return()
|
|
endif()
|
|
|
|
function(cub_add_header_test label definitions)
|
|
set(headertest_target cub.headers.${label})
|
|
|
|
cccl_generate_header_tests(${headertest_target} cub GLOBS "cub/*.cuh")
|
|
cub_configure_cuda_target(${headertest_target} RDC ${CUB_FORCE_RDC})
|
|
target_link_libraries(${headertest_target} PUBLIC cub.compiler_interface)
|
|
target_compile_definitions(${headertest_target} PRIVATE ${definitions})
|
|
endfunction()
|
|
|
|
# Wrap Thrust/CUB in a custom namespace to check proper use of ns macros:
|
|
set(
|
|
header_definitions
|
|
"THRUST_WRAPPED_NAMESPACE=wrapped_thrust"
|
|
"CUB_WRAPPED_NAMESPACE=wrapped_cub"
|
|
)
|
|
cub_add_header_test(base "${header_definitions}")
|
|
|
|
# Check that BF16 support can be disabled
|
|
set(
|
|
header_definitions
|
|
"THRUST_WRAPPED_NAMESPACE=wrapped_thrust"
|
|
"CUB_WRAPPED_NAMESPACE=wrapped_cub"
|
|
"CCCL_DISABLE_BF16_SUPPORT"
|
|
)
|
|
cub_add_header_test(no_bf16 "${header_definitions}")
|
|
|
|
# Check that half support can be disabled
|
|
set(
|
|
header_definitions
|
|
"THRUST_WRAPPED_NAMESPACE=wrapped_thrust"
|
|
"CUB_WRAPPED_NAMESPACE=wrapped_cub"
|
|
"CCCL_DISABLE_FP16_SUPPORT"
|
|
)
|
|
cub_add_header_test(no_half "${header_definitions}")
|
|
|
|
# Check that half support can be disabled
|
|
set(
|
|
header_definitions
|
|
"THRUST_WRAPPED_NAMESPACE=wrapped_thrust"
|
|
"CUB_WRAPPED_NAMESPACE=wrapped_cub"
|
|
"CCCL_DISABLE_FP8_SUPPORT"
|
|
)
|
|
cub_add_header_test(no_fp8 "${header_definitions}")
|