Skip to content

Commit

Permalink
Update select_compute_arch.cmake (#113)
Browse files Browse the repository at this point in the history 
Added complier flags for ampere gpus.
  • Loading branch information
@Subhajit-Roy-Partho
Subhajit-Roy-Partho authored Aug 6, 2024
1 parent 2583813 commit 21d7e48
Showing 1 changed file with 72 additions and 54 deletions.
126 changes: 72 additions & 54 deletions select_compute_arch.cmake
View file
Original file line number Diff line number Diff line change
Expand Up @@ -5,9 +5,9 @@
# - "Auto" detects local machine GPU compute arch at runtime.
# - "Common" and "All" cover common and entire subsets of architectures
# ARCH_AND_PTX : NAME | NUM.NUM | NUM.NUM(NUM.NUM) | NUM.NUM+PTX
# NAME: Fermi Kepler Maxwell Kepler+Tegra Kepler+Tesla Maxwell+Tegra Pascal Volta Turing
# NAME: Fermi Kepler Maxwell Kepler+Tegra Kepler+Tesla Maxwell+Tegra Pascal Volta Turing Ampere
# NUM: Any number. Only those pairs are currently accepted by NVCC though:
# 2.0 2.1 3.0 3.2 3.5 3.7 5.0 5.2 5.3 6.0 6.2 7.0 7.2 7.5
# 2.0 2.1 3.0 3.2 3.5 3.7 5.0 5.2 5.3 6.0 6.2 7.0 7.2 7.5 8.0 8.6
# Returns LIST of flags to be added to CUDA_NVCC_FLAGS in ${out_variable}
# Additionally, sets ${out_variable}_readable to the resulting numeric list
# Example:
Expand All @@ -25,70 +25,85 @@ if(CMAKE_CUDA_COMPILER_LOADED) # CUDA as a language
endif()

# See: https://docs.nvidia.com/cuda/cuda-compiler-driver-nvcc/index.html#gpu-feature-list
# Additions, deprecations, and removals can be found in the release notes:
# https://developer.nvidia.com/cuda-toolkit-archive

if(CUDA_VERSION VERSION_LESS "11.0")
# This list will be used for CUDA_ARCH_NAME = All option
set(CUDA_KNOWN_GPU_ARCHITECTURES "Fermi" "Kepler" "Maxwell")

# This list will be used for CUDA_ARCH_NAME = Common option (enabled by default)
set(CUDA_COMMON_GPU_ARCHITECTURES "3.0" "3.5")
else()
set(CUDA_KNOWN_GPU_ARCHITECTURES "")
set(CUDA_COMMON_GPU_ARCHITECTURES "")
endif()

# as of now (Sept 2020), all CUDA versions support Maxwell GPUs
list(APPEND CUDA_KNOWN_GPU_ARCHITECTURES "Maxwell")
list(APPEND CUDA_COMMON_GPU_ARCHITECTURES "5.0")

if(CUDA_VERSION VERSION_LESS "7.0")
set(CUDA_LIMIT_GPU_ARCHITECTURE "4.9")
endif()
# The initial status here is for CUDA 7.0
set(CUDA_KNOWN_GPU_ARCHITECTURES "Fermi" "Kepler" "Maxwell" "Kepler+Tegra" "Kepler+Tesla" "Maxwell+Tegra")
set(CUDA_COMMON_GPU_ARCHITECTURES "2.0" "2.1" "3.0" "3.5" "5.0" "5.3")
set(CUDA_LIMIT_GPU_ARCHITECTURE "6.0")
set(CUDA_ALL_GPU_ARCHITECTURES "2.0" "2.1" "3.0" "3.2" "3.5" "3.7" "5.0" "5.2" "5.3")
set(_CUDA_MAX_COMMON_ARCHITECTURE "5.2+PTX")

# This list is used to filter CUDA archs when autodetecting
set(CUDA_ALL_GPU_ARCHITECTURES "3.0" "3.2" "3.5" "5.0")

if(CUDA_VERSION VERSION_GREATER "6.9")
list(APPEND CUDA_KNOWN_GPU_ARCHITECTURES "Kepler+Tegra" "Kepler+Tesla" "Maxwell+Tegra")
list(APPEND CUDA_COMMON_GPU_ARCHITECTURES "5.2")

if(CUDA_VERSION VERSION_LESS "8.0")
list(APPEND CUDA_COMMON_GPU_ARCHITECTURES "5.2+PTX")
set(CUDA_LIMIT_GPU_ARCHITECTURE "5.9")
endif()
endif()

if(CUDA_VERSION VERSION_GREATER "7.9")
if(CUDA_VERSION VERSION_GREATER_EQUAL "8.0")
list(APPEND CUDA_KNOWN_GPU_ARCHITECTURES "Pascal")
list(APPEND CUDA_COMMON_GPU_ARCHITECTURES "6.0" "6.1")
list(APPEND CUDA_ALL_GPU_ARCHITECTURES "6.0" "6.1" "6.2")

if(CUDA_VERSION VERSION_LESS "9.0")
list(APPEND CUDA_COMMON_GPU_ARCHITECTURES "6.1+PTX")
set(CUDA_LIMIT_GPU_ARCHITECTURE "6.9")
endif()
set(_CUDA_MAX_COMMON_ARCHITECTURE "6.2+PTX")
set(CUDA_LIMIT_GPU_ARCHITECTURE "7.0")

list(REMOVE_ITEM CUDA_COMMON_GPU_ARCHITECTURES "2.0" "2.1")
endif ()

if(CUDA_VERSION VERSION_GREATER "8.9")
if(CUDA_VERSION VERSION_GREATER_EQUAL "9.0")
list(APPEND CUDA_KNOWN_GPU_ARCHITECTURES "Volta")
list(APPEND CUDA_COMMON_GPU_ARCHITECTURES "7.0" "7.0+PTX")
list(APPEND CUDA_ALL_GPU_ARCHITECTURES "7.0" "7.0+PTX" "7.2" "7.2+PTX")
list(APPEND CUDA_COMMON_GPU_ARCHITECTURES "7.0")
list(APPEND CUDA_ALL_GPU_ARCHITECTURES "7.0" "7.2")

if(CUDA_VERSION VERSION_LESS "10.0")
set(CUDA_LIMIT_GPU_ARCHITECTURE "7.9")
endif()
set(_CUDA_MAX_COMMON_ARCHITECTURE "7.2+PTX")
set(CUDA_LIMIT_GPU_ARCHITECTURE "8.0")

list(REMOVE_ITEM CUDA_KNOWN_GPU_ARCHITECTURES "Fermi")
list(REMOVE_ITEM CUDA_ALL_GPU_ARCHITECTURES "2.0" "2.1")
endif()

if(CUDA_VERSION VERSION_GREATER "9.9")
if(CUDA_VERSION VERSION_GREATER_EQUAL "10.0")
list(APPEND CUDA_KNOWN_GPU_ARCHITECTURES "Turing")
list(APPEND CUDA_COMMON_GPU_ARCHITECTURES "7.5" "7.5+PTX")
list(APPEND CUDA_ALL_GPU_ARCHITECTURES "7.5" "7.5+PTX")
list(APPEND CUDA_COMMON_GPU_ARCHITECTURES "7.5")
list(APPEND CUDA_ALL_GPU_ARCHITECTURES "7.5")

if(CUDA_VERSION VERSION_LESS "11.0")
set(CUDA_LIMIT_GPU_ARCHITECTURE "8.9")
endif()
set(_CUDA_MAX_COMMON_ARCHITECTURE "7.5+PTX")
set(CUDA_LIMIT_GPU_ARCHITECTURE "8.0")

list(REMOVE_ITEM CUDA_COMMON_GPU_ARCHITECTURES "3.0")
endif()

# https://docs.nvidia.com/cuda/archive/11.0/cuda-toolkit-release-notes/index.html#cuda-general-new-features
# https://docs.nvidia.com/cuda/archive/11.0/cuda-toolkit-release-notes/index.html#deprecated-features
if(CUDA_VERSION VERSION_GREATER_EQUAL "11.0")
list(APPEND CUDA_KNOWN_GPU_ARCHITECTURES "Ampere")
list(APPEND CUDA_COMMON_GPU_ARCHITECTURES "8.0")
list(APPEND CUDA_ALL_GPU_ARCHITECTURES "8.0")

set(_CUDA_MAX_COMMON_ARCHITECTURE "8.0+PTX")
set(CUDA_LIMIT_GPU_ARCHITECTURE "8.6")

list(REMOVE_ITEM CUDA_COMMON_GPU_ARCHITECTURES "3.5" "5.0")
list(REMOVE_ITEM CUDA_ALL_GPU_ARCHITECTURES "3.0" "3.2")
endif()

if(CUDA_VERSION VERSION_GREATER_EQUAL "11.1")
list(APPEND CUDA_COMMON_GPU_ARCHITECTURES "8.6")
list(APPEND CUDA_ALL_GPU_ARCHITECTURES "8.6")

set(_CUDA_MAX_COMMON_ARCHITECTURE "8.6+PTX")
set(CUDA_LIMIT_GPU_ARCHITECTURE "9.0")
endif()

list(APPEND CUDA_COMMON_GPU_ARCHITECTURES "${_CUDA_MAX_COMMON_ARCHITECTURE}")

# Check with: cmake -DCUDA_VERSION=7.0 -P select_compute_arch.cmake
if(DEFINED CMAKE_SCRIPT_MODE_FILE)
include(CMakePrintHelpers)
cmake_print_variables(CUDA_KNOWN_GPU_ARCHITECTURES)
cmake_print_variables(CUDA_COMMON_GPU_ARCHITECTURES)
cmake_print_variables(CUDA_LIMIT_GPU_ARCHITECTURE)
cmake_print_variables(CUDA_ALL_GPU_ARCHITECTURES)
endif()


################################################################################################
# A function for automatic detection of GPUs installed (if autodetection is enabled)
# Usage:
Expand Down Expand Up @@ -120,10 +135,10 @@ function(CUDA_DETECT_INSTALLED_GPUS OUT_VARIABLE)
"}\n")

if(CMAKE_CUDA_COMPILER_LOADED) # CUDA as a language
try_run(run_result compile_result ${PROJECT_BINARY_DIR} ${file}
try_run(run_result compile_result SOURCES ${file}
RUN_OUTPUT_VARIABLE compute_capabilities)
else()
try_run(run_result compile_result ${PROJECT_BINARY_DIR} ${file}
try_run(run_result compile_result SOURCES ${file}
CMAKE_FLAGS "-DINCLUDE_DIRECTORIES=${CUDA_INCLUDE_DIRS}"
LINK_LIBRARIES ${CUDA_LIBRARIES}
RUN_OUTPUT_VARIABLE compute_capabilities)
Expand All @@ -147,7 +162,7 @@ function(CUDA_DETECT_INSTALLED_GPUS OUT_VARIABLE)
set(CUDA_GPU_DETECT_OUTPUT_FILTERED "")
separate_arguments(CUDA_GPU_DETECT_OUTPUT)
foreach(ITEM IN ITEMS ${CUDA_GPU_DETECT_OUTPUT})
if(CUDA_LIMIT_GPU_ARCHITECTURE AND ITEM VERSION_GREATER CUDA_LIMIT_GPU_ARCHITECTURE)
if(CUDA_LIMIT_GPU_ARCHITECTURE AND ITEM VERSION_GREATER_EQUAL CUDA_LIMIT_GPU_ARCHITECTURE)
list(GET CUDA_COMMON_GPU_ARCHITECTURES -1 NEWITEM)
string(APPEND CUDA_GPU_DETECT_OUTPUT_FILTERED " ${NEWITEM}")
else()
Expand All @@ -163,7 +178,7 @@ endfunction()
################################################################################################
# Function for selecting GPU arch flags for nvcc based on CUDA architectures from parameter list
# Usage:
# CUDA_SELECT_NVCC_ARCH_FLAGS(out_variable [list of CUDA compute archs])
# SELECT_NVCC_ARCH_FLAGS(out_variable [list of CUDA compute archs])
function(CUDA_SELECT_NVCC_ARCH_FLAGS out_variable)
set(CUDA_ARCH_LIST "${ARGN}")

Expand Down Expand Up @@ -223,6 +238,9 @@ function(CUDA_SELECT_NVCC_ARCH_FLAGS out_variable)
elseif(${arch_name} STREQUAL "Turing")
set(arch_bin 7.5)
set(arch_ptx 7.5)
elseif(${arch_name} STREQUAL "Ampere")
set(arch_bin 8.0)
set(arch_ptx 8.0)
else()
message(SEND_ERROR "Unknown CUDA Architecture Name ${arch_name} in CUDA_SELECT_NVCC_ARCH_FLAGS")
endif()
Expand Down Expand Up @@ -263,7 +281,7 @@ function(CUDA_SELECT_NVCC_ARCH_FLAGS out_variable)
list(APPEND nvcc_archs_readable sm_${CMAKE_MATCH_1})
else()
# User didn't explicitly specify ARCH for the concrete CODE, we assume ARCH=CODE
list(APPEND nvcc_flags "-gencode arch=compute_${arch},code=sm_${arch}")
list(APPEND nvcc_flags -gencode arch=compute_${arch},code=sm_${arch})
list(APPEND nvcc_archs_readable sm_${arch})
endif()
endforeach()
Expand Down

0 comments on commit 21d7e48

Please sign in to comment.