CMakeLists.txt 10.2 KB
Newer Older
1
cmake_minimum_required(VERSION 3.14)
JD's avatar
JD committed
2
3

# Check support for CUDA/HIP in Cmake
Chao Liu's avatar
Chao Liu committed
4
project(composable_kernel)
Chao Liu's avatar
Chao Liu committed
5

6
7
list(APPEND CMAKE_MODULE_PATH "${PROJECT_SOURCE_DIR}/cmake")

JD's avatar
JD committed
8
9
enable_testing()

10
set(ROCM_SYMLINK_LIBS OFF)
Anthony Chang's avatar
Anthony Chang committed
11
find_package(ROCM REQUIRED PATHS /opt/rocm)
JD's avatar
JD committed
12
13
14
15
16
17

include(ROCMInstallTargets)
include(ROCMPackageConfigHelpers)
include(ROCMSetupVersion)
include(ROCMInstallSymlinks)
include(ROCMCreatePackage)
Chao Liu's avatar
Chao Liu committed
18
19
include(CheckCXXCompilerFlag)

20
rocm_setup_version(VERSION 0.2.0)
JD's avatar
JD committed
21
22
23
include(TargetFlags)
list(APPEND CMAKE_PREFIX_PATH ${CMAKE_INSTALL_PREFIX} ${CMAKE_INSTALL_PREFIX}/llvm ${CMAKE_INSTALL_PREFIX}/hip /opt/rocm /opt/rocm/llvm /opt/rocm/hip)

Adam Osewski's avatar
Adam Osewski committed
24
25
26
27
28
29
30
31
option(USE_BITINT_EXTENSION_INT4, "Whether to enable clang's BitInt extension to provide int4 data type." OFF)

if(USE_BITINT_EXTENSION_INT4)
    add_compile_definitions(CK_EXPERIMENTAL_BIT_INT_EXTENSION_INT4)
    add_compile_options(-Wno-bit-int-extension)
    message("CK compiled with USE_BITINT_EXTENSION_INT4 set to ${USE_BITINT_EXTENSION_INT4}")
endif()

32
33
34
35
36
## Threads
set(THREADS_PREFER_PTHREAD_FLAG ON)
find_package(Threads REQUIRED)
link_libraries(Threads::Threads)

37
## C++
Chao Liu's avatar
Chao Liu committed
38
enable_language(CXX)
Chao Liu's avatar
Chao Liu committed
39
set(CMAKE_CXX_STANDARD 17)
Chao Liu's avatar
Chao Liu committed
40
set(CMAKE_CXX_STANDARD_REQUIRED ON)
Chao Liu's avatar
Chao Liu committed
41
set(CMAKE_CXX_EXTENSIONS OFF)
Chao Liu's avatar
Chao Liu committed
42
43
message("CMAKE_CXX_COMPILER_ID: ${CMAKE_CXX_COMPILER_ID}")

44
## OpenMP
Chao Liu's avatar
Chao Liu committed
45
46
47
48
49
50
51
52
53
54
55
if(CMAKE_CXX_COMPILER_ID MATCHES "Clang")
	# workaround issue hipcc in rocm3.5 cannot find openmp
	set(OpenMP_CXX "${CMAKE_CXX_COMPILER}")
	set(OpenMP_CXX_FLAGS "-fopenmp=libomp -Wno-unused-command-line-argument")
	set(OpenMP_CXX_LIB_NAMES "libomp" "libgomp" "libiomp5")
	set(OpenMP_libomp_LIBRARY ${OpenMP_CXX_LIB_NAMES})
	set(OpenMP_libgomp_LIBRARY ${OpenMP_CXX_LIB_NAMES})
	set(OpenMP_libiomp5_LIBRARY ${OpenMP_CXX_LIB_NAMES})
else()
	find_package(OpenMP REQUIRED)
endif()
Chao Liu's avatar
Chao Liu committed
56

Chao Liu's avatar
Chao Liu committed
57
58
59
60
message("OpenMP_CXX_LIB_NAMES: ${OpenMP_CXX_LIB_NAMES}")
message("OpenMP_gomp_LIBRARY: ${OpenMP_gomp_LIBRARY}")
message("OpenMP_pthread_LIBRARY: ${OpenMP_pthread_LIBRARY}")
message("OpenMP_CXX_FLAGS: ${OpenMP_CXX_FLAGS}")
Chao Liu's avatar
Chao Liu committed
61

Chao Liu's avatar
Chao Liu committed
62
63
link_libraries(${OpenMP_gomp_LIBRARY})
link_libraries(${OpenMP_pthread_LIBRARY})
Chao Liu's avatar
Chao Liu committed
64

65
66
## HIP
find_package(HIP REQUIRED)
JD's avatar
JD committed
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
# Override HIP version in config.h, if necessary.
# The variables set by find_package() can't be overwritten,
# therefore let's use intermediate variables.
set(CK_HIP_VERSION_MAJOR "${HIP_VERSION_MAJOR}")
set(CK_HIP_VERSION_MINOR "${HIP_VERSION_MINOR}")
set(CK_HIP_VERSION_PATCH "${HIP_VERSION_PATCH}")
if( DEFINED CK_OVERRIDE_HIP_VERSION_MAJOR )
    set(CK_HIP_VERSION_MAJOR "${CK_OVERRIDE_HIP_VERSION_MAJOR}")
    message(STATUS "CK_HIP_VERSION_MAJOR overriden with ${CK_OVERRIDE_HIP_VERSION_MAJOR}")
endif()
if( DEFINED CK_OVERRIDE_HIP_VERSION_MINOR )
    set(CK_HIP_VERSION_MINOR "${CK_OVERRIDE_HIP_VERSION_MINOR}")
    message(STATUS "CK_HIP_VERSION_MINOR overriden with ${CK_OVERRIDE_HIP_VERSION_MINOR}")
endif()
if( DEFINED CK_OVERRIDE_HIP_VERSION_PATCH )
    set(CK_HIP_VERSION_PATCH "${CK_OVERRIDE_HIP_VERSION_PATCH}")
    message(STATUS "CK_HIP_VERSION_PATCH overriden with ${CK_OVERRIDE_HIP_VERSION_PATCH}")
endif()
message(STATUS "Build with HIP ${HIP_VERSION}")
86
87
link_libraries(hip::device)
add_compile_definitions(__HIP_PLATFORM_HCC__=1)
88

Chao Liu's avatar
Chao Liu committed
89
90
## tidy
include(EnableCompilerWarnings)
JD's avatar
JD committed
91
set(CK_TIDY_ERRORS ERRORS * -readability-inconsistent-declaration-parameter-name)
Chao Liu's avatar
Chao Liu committed
92
if(CMAKE_CXX_COMPILER MATCHES ".*hcc" OR CMAKE_CXX_COMPILER MATCHES ".*clang\\+\\+")
JD's avatar
JD committed
93
    set(CK_TIDY_CHECKS -modernize-use-override -readability-non-const-parameter)
Chao Liu's avatar
Chao Liu committed
94
# Enable tidy on hip
JD's avatar
JD committed
95
96
elseif(CK_BACKEND STREQUAL "HIP" OR CK_BACKEND STREQUAL "HIPNOGPU")
    set(CK_TIDY_ERRORS ALL)
Chao Liu's avatar
Chao Liu committed
97
98
endif()

JD's avatar
JD committed
99

Chao Liu's avatar
Chao Liu committed
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
include(ClangTidy)
enable_clang_tidy(
    CHECKS
        *
        -abseil-*
        -android-cloexec-fopen
        # Yea we shouldn't be using rand()
        -cert-msc30-c
        -bugprone-exception-escape
        -bugprone-macro-parentheses
        -cert-env33-c
        -cert-msc32-c
        -cert-msc50-cpp
        -cert-msc51-cpp
        -cert-dcl37-c
        -cert-dcl51-cpp
        -clang-analyzer-alpha.core.CastToStruct
        -clang-analyzer-optin.performance.Padding
        -clang-diagnostic-deprecated-declarations
        -clang-diagnostic-extern-c-compat
        -clang-diagnostic-unused-command-line-argument
        -cppcoreguidelines-avoid-c-arrays
        -cppcoreguidelines-avoid-magic-numbers
        -cppcoreguidelines-explicit-virtual-functions
        -cppcoreguidelines-init-variables
        -cppcoreguidelines-macro-usage
        -cppcoreguidelines-non-private-member-variables-in-classes
        -cppcoreguidelines-pro-bounds-array-to-pointer-decay
        -cppcoreguidelines-pro-bounds-constant-array-index
        -cppcoreguidelines-pro-bounds-pointer-arithmetic
        -cppcoreguidelines-pro-type-member-init
        -cppcoreguidelines-pro-type-reinterpret-cast
        -cppcoreguidelines-pro-type-union-access
        -cppcoreguidelines-pro-type-vararg
        -cppcoreguidelines-special-member-functions
        -fuchsia-*
        -google-explicit-constructor
        -google-readability-braces-around-statements
        -google-readability-todo
        -google-runtime-int
        -google-runtime-references
        -hicpp-vararg
        -hicpp-braces-around-statements
        -hicpp-explicit-conversions
        -hicpp-named-parameter
        -hicpp-no-array-decay
        # We really shouldn't use bitwise operators with signed integers, but
        # opencl leaves us no choice
        -hicpp-avoid-c-arrays
        -hicpp-signed-bitwise
        -hicpp-special-member-functions
        -hicpp-uppercase-literal-suffix
        -hicpp-use-auto
        -hicpp-use-equals-default
        -hicpp-use-override
        -llvm-header-guard
        -llvm-include-order
        #-llvmlibc-*
        -llvmlibc-restrict-system-libc-headers
        -llvmlibc-callee-namespace
        -llvmlibc-implementation-in-namespace
        -llvm-else-after-return
        -llvm-qualified-auto
        -misc-misplaced-const
        -misc-non-private-member-variables-in-classes
        -misc-no-recursion
        -modernize-avoid-bind
        -modernize-avoid-c-arrays
        -modernize-pass-by-value
        -modernize-use-auto
        -modernize-use-default-member-init
        -modernize-use-equals-default
        -modernize-use-trailing-return-type
        -modernize-use-transparent-functors
        -performance-unnecessary-value-param
        -readability-braces-around-statements
        -readability-else-after-return
        # we are not ready to use it, but very useful
        -readability-function-cognitive-complexity
        -readability-isolate-declaration
        -readability-magic-numbers
        -readability-named-parameter
        -readability-uppercase-literal-suffix
        -readability-convert-member-functions-to-static
        -readability-qualified-auto
        -readability-redundant-string-init
        # too many narrowing conversions in our code
        -bugprone-narrowing-conversions
        -cppcoreguidelines-narrowing-conversions
        -altera-struct-pack-align
        -cppcoreguidelines-prefer-member-initializer
JD's avatar
JD committed
191
192
        ${CK_TIDY_CHECKS}
        ${CK_TIDY_ERRORS}
Chao Liu's avatar
Chao Liu committed
193
194
195
    HEADER_FILTER
        "\.hpp$"
    EXTRA_ARGS
JD's avatar
JD committed
196
        -DCK_USE_CLANG_TIDY
Chao Liu's avatar
Chao Liu committed
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
)

include(CppCheck)
enable_cppcheck(
    CHECKS
        warning
        style
        performance
        portability
    SUPPRESS
        ConfigurationNotChecked
        constStatement
        duplicateCondition
        noExplicitConstructor
        passedByValue
Chao Liu's avatar
Chao Liu committed
212
        preprocessorErrorDirective
Chao Liu's avatar
Chao Liu committed
213
214
215
216
217
218
219
        shadowVariable
        unusedFunction
        unusedPrivateFunction
        unusedStructMember
        unmatchedSuppression
    FORCE
    SOURCES
Chao Liu's avatar
Chao Liu committed
220
        library/src
Chao Liu's avatar
Chao Liu committed
221
222
223
    INCLUDE
        ${CMAKE_CURRENT_SOURCE_DIR}/include
        ${CMAKE_CURRENT_BINARY_DIR}/include
Chao Liu's avatar
Chao Liu committed
224
        ${CMAKE_CURRENT_SOURCE_DIR}/library/include
Chao Liu's avatar
Chao Liu committed
225
226
227
228
    DEFINE
        CPPCHECK=1
        __linux__=1
)
Chao Liu's avatar
Chao Liu committed
229

JD's avatar
JD committed
230
231
232
233
set(CMAKE_LIBRARY_OUTPUT_DIRECTORY ${CMAKE_CURRENT_BINARY_DIR}/lib)
set(CMAKE_ARCHIVE_OUTPUT_DIRECTORY ${CMAKE_CURRENT_BINARY_DIR}/lib)
set(CMAKE_RUNTIME_OUTPUT_DIRECTORY ${CMAKE_CURRENT_BINARY_DIR}/bin)

Chao Liu's avatar
Chao Liu committed
234
235
236
include_directories(BEFORE
    ${PROJECT_SOURCE_DIR}/include
    ${PROJECT_SOURCE_DIR}/library/include
237
    ${HIP_INCLUDE_DIRS}
JD's avatar
JD committed
238
)
Chao Liu's avatar
Chao Liu committed
239

240

JD's avatar
JD committed
241
242
SET(BUILD_DEV ON CACHE BOOL "BUILD_DEV")
if(BUILD_DEV)
Chao Liu's avatar
Chao Liu committed
243
244
    add_compile_options(-Werror)
    add_compile_options(-Weverything)
JD's avatar
JD committed
245
246
247
endif()
message("CMAKE_CXX_FLAGS: ${CMAKE_CXX_FLAGS}")

Anthony Chang's avatar
Anthony Chang committed
248
249
add_custom_target(check COMMAND ${CMAKE_CTEST_COMMAND} --output-on-failure -C ${CMAKE_CFG_INTDIR})

250
251
252
253
254
255
256
257
258
259
file(GLOB_RECURSE INSTANCE_FILES "${PROJECT_SOURCE_DIR}/*/device_*_instance.cpp")
file(GLOB dir_list RELATIVE ${PROJECT_SOURCE_DIR}/library/src/tensor_operation_instance/gpu ${PROJECT_SOURCE_DIR}/library/src/tensor_operation_instance/gpu/*)
set(CK_DEVICE_INSTANCES)
FOREACH(subdir_path ${dir_list})
    IF(IS_DIRECTORY "${PROJECT_SOURCE_DIR}/library/src/tensor_operation_instance/gpu/${subdir_path}")
       list(APPEND CK_DEVICE_INSTANCES device_${subdir_path}_instance)
    ENDIF()
ENDFOREACH()
add_custom_target(instances DEPENDS utility;${CK_DEVICE_INSTANCES}  SOURCES ${INSTANCE_FILES})

260
261
262
263
264
rocm_package_setup_component(tests
        LIBRARY_NAME composablekernel
        PACKAGE_NAME tests # Prevent -static suffix on package name
)

Chao Liu's avatar
Chao Liu committed
265
add_subdirectory(library)
266
add_subdirectory(example)
267
add_subdirectory(test)
Chao Liu's avatar
Chao Liu committed
268
add_subdirectory(profiler)
JD's avatar
JD committed
269
270
271
272
273
274
275
276
277
278
279

#Create an interface target for the include only files and call it "composablekernels"
include(CMakePackageConfigHelpers)

set(version 1.0.0)
write_basic_package_version_file(
    "${CMAKE_CURRENT_BINARY_DIR}/composable_kernelConfigVersion.cmake"
    VERSION "${version}"
    COMPATIBILITY AnyNewerVersion
)

Anthony Chang's avatar
Anthony Chang committed
280
configure_package_config_file(${CMAKE_CURRENT_SOURCE_DIR}/Config.cmake.in
JD's avatar
JD committed
281
        "${CMAKE_CURRENT_BINARY_DIR}/composable_kernelConfig.cmake"
Anthony Chang's avatar
Anthony Chang committed
282
        INSTALL_DESTINATION ${CMAKE_INSTALL_LIBDIR}/cmake/composable_kernel
JD's avatar
JD committed
283
284
285
        NO_CHECK_REQUIRED_COMPONENTS_MACRO
)

286
rocm_install(FILES
JD's avatar
JD committed
287
288
    "${CMAKE_CURRENT_BINARY_DIR}/composable_kernelConfig.cmake"
    "${CMAKE_CURRENT_BINARY_DIR}/composable_kernelConfigVersion.cmake"
Anthony Chang's avatar
Anthony Chang committed
289
    DESTINATION ${CMAKE_INSTALL_LIBDIR}/cmake/composable_kernel
JD's avatar
JD committed
290
)
291
292
293
294
295
296
297
298
299
300
301

set(CPACK_RESOURCE_FILE_LICENSE "${CMAKE_CURRENT_SOURCE_DIR}/LICENSE")
set(CPACK_RPM_PACKAGE_LICENSE "MIT")

rocm_create_package(
    NAME composablekernel
    DESCRIPTION "High Performance Composable Kernel for AMD GPUs"
    MAINTAINER "MIOpen Kernels Dev Team <dl.MIOpen@amd.com>"
    LDCONFIG
    HEADER_ONLY
)