CMakeLists.txt 9.7 KB
Newer Older
Chao Liu's avatar
Chao Liu committed
1
cmake_minimum_required(VERSION 3.5)
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
10
11
12
13
14
15
16
enable_testing()

find_package(ROCM REQUIRED PATHS /opt/rocm)

include(ROCMInstallTargets)
include(ROCMPackageConfigHelpers)
include(ROCMSetupVersion)
include(ROCMInstallSymlinks)
include(ROCMCreatePackage)
Chao Liu's avatar
Chao Liu committed
17
18
include(CheckCXXCompilerFlag)

JD's avatar
JD committed
19
20
21
22
rocm_setup_version(VERSION 1.0.0)
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)

23
## C++
Chao Liu's avatar
Chao Liu committed
24
enable_language(CXX)
Chao Liu's avatar
Chao Liu committed
25
set(CMAKE_CXX_STANDARD 17)
Chao Liu's avatar
Chao Liu committed
26
set(CMAKE_CXX_STANDARD_REQUIRED ON)
Chao Liu's avatar
Chao Liu committed
27
set(CMAKE_CXX_EXTENSIONS OFF)
Chao Liu's avatar
Chao Liu committed
28
29
message("CMAKE_CXX_COMPILER_ID: ${CMAKE_CXX_COMPILER_ID}")

30
## OpenMP
Chao Liu's avatar
Chao Liu committed
31
32
33
34
35
36
37
38
39
40
41
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
42

Chao Liu's avatar
Chao Liu committed
43
44
45
46
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
47

JD's avatar
JD committed
48
# set (CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} ${OpenMP_CXX_FLAGS}")
Chao Liu's avatar
Chao Liu committed
49
50
link_libraries(${OpenMP_gomp_LIBRARY})
link_libraries(${OpenMP_pthread_LIBRARY})
Chao Liu's avatar
Chao Liu committed
51

52
53
## HIP
find_package(HIP REQUIRED)
JD's avatar
JD committed
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
# 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}")
73

Chao Liu's avatar
tidy  
Chao Liu committed
74
75
## half
#find_path(HALF_INCLUDE_DIR half.hpp)
76
set(HALF_INCLUDE_DIR "${PROJECT_SOURCE_DIR}/external/half/include")
Chao Liu's avatar
tidy  
Chao Liu committed
77
message("HALF_INCLUDE_DIR: ${HALF_INCLUDE_DIR}")
Chao Liu's avatar
tidy  
Chao Liu committed
78

Chao Liu's avatar
Chao Liu committed
79

JD's avatar
JD committed
80
81
82
83
84
rocm_create_package(
    NAME CK-${CK_BACKEND}
    DESCRIPTION "High Performance Composable Kernels for AMD GPUs"
    LDCONFIG
)
Chao Liu's avatar
Chao Liu committed
85
86
## tidy
include(EnableCompilerWarnings)
JD's avatar
JD committed
87
set(CK_TIDY_ERRORS ERRORS * -readability-inconsistent-declaration-parameter-name)
Chao Liu's avatar
Chao Liu committed
88
if(CMAKE_CXX_COMPILER MATCHES ".*hcc" OR CMAKE_CXX_COMPILER MATCHES ".*clang\\+\\+")
JD's avatar
JD committed
89
    set(CK_TIDY_CHECKS -modernize-use-override -readability-non-const-parameter)
Chao Liu's avatar
Chao Liu committed
90
# Enable tidy on hip
JD's avatar
JD committed
91
92
elseif(CK_BACKEND STREQUAL "HIP" OR CK_BACKEND STREQUAL "HIPNOGPU")
    set(CK_TIDY_ERRORS ALL)
Chao Liu's avatar
Chao Liu committed
93
94
endif()

JD's avatar
JD committed
95

Chao Liu's avatar
Chao Liu committed
96
97
98
99
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
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
188
189
        ${CK_TIDY_CHECKS}
        ${CK_TIDY_ERRORS}
Chao Liu's avatar
Chao Liu committed
190
191
192
    HEADER_FILTER
        "\.hpp$"
    EXTRA_ARGS
JD's avatar
JD committed
193
        -DCK_USE_CLANG_TIDY
Chao Liu's avatar
Chao Liu committed
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
)

include(CppCheck)
enable_cppcheck(
    CHECKS
        warning
        style
        performance
        portability
    SUPPRESS
        ConfigurationNotChecked
        constStatement
        duplicateCondition
        noExplicitConstructor
        passedByValue
Chao Liu's avatar
Chao Liu committed
209
        preprocessorErrorDirective
Chao Liu's avatar
Chao Liu committed
210
211
212
213
214
215
216
217
        shadowVariable
        unusedFunction
        unusedPrivateFunction
        unusedStructMember
        unmatchedSuppression
    FORCE
    SOURCES
        host/host_tensor/src
Chao Liu's avatar
tidy  
Chao Liu committed
218
219
        host/driver_offline/src
        composable_kernel/src/kernel_wrapper
Chao Liu's avatar
Chao Liu committed
220
    INCLUDE
Chao Liu's avatar
tidy  
Chao Liu committed
221
        host/host_tensor/include
222
        host/device/include
Chao Liu's avatar
tidy  
Chao Liu committed
223
224
225
        host/solver/include
        host/driver_offline/include
        composable_kernel/include/*
Chao Liu's avatar
Chao Liu committed
226
227
228
229
230
231
        ${CMAKE_CURRENT_SOURCE_DIR}/include
        ${CMAKE_CURRENT_BINARY_DIR}/include
    DEFINE
        CPPCHECK=1
        __linux__=1
)
JD's avatar
JD committed
232
233
234
235
236
237
238
239
240
241
242
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)

file(GLOB_RECURSE COMPOSABLE_KERNEL_HEADERS "composable_kernel/include/*/*.hpp")
file(GLOB_RECURSE DEVICE_OPS_HEADERS "device_operation/include/*.hpp")

file(GLOB_RECURSE DEVICE_OPS_SOURCE "device_operation/*.cpp")

set(CK_HEADERS ${COMPOSABLE_KERNEL_HEADERS} ${DEVICE_OPS_HEADERS})
set(CK_SOURCE ${DEVICE_OPS_SOURCE})
243
244
add_library(composable_kernel ${CK_SOURCE})

JD's avatar
JD committed
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276

target_include_directories(composable_kernel PUBLIC
    $<BUILD_INTERFACE:${PROJECT_SOURCE_DIR}/composable_kernel/include>
)
target_include_directories(composable_kernel PUBLIC 
    $<BUILD_INTERFACE:${PROJECT_SOURCE_DIR}/device_operation/include>
)
target_include_directories(composable_kernel PUBLIC 
    $<BUILD_INTERFACE:${PROJECT_SOURCE_DIR}/host/include>
)
target_include_directories(composable_kernel PUBLIC 
    $<BUILD_INTERFACE:${PROJECT_SOURCE_DIR}/host/host_tensor/include>
)
# The following should eventually be removed
target_include_directories(composable_kernel PUBLIC
    $<BUILD_INTERFACE:${PROJECT_SOURCE_DIR}/composable_kernel/include/utility>
)
target_include_directories(composable_kernel PUBLIC
    $<BUILD_INTERFACE:${PROJECT_SOURCE_DIR}/composable_kernel/include/tensor_operation>
)
target_include_directories(composable_kernel PUBLIC
    $<BUILD_INTERFACE:${PROJECT_SOURCE_DIR}/composable_kernel/include/tensor_description>
)
# clang_tidy_check(composable_kernel)
SET(BUILD_DEV ON CACHE BOOL "BUILD_DEV")
if(BUILD_DEV)
    target_compile_options(composable_kernel PRIVATE -Werror)
    target_compile_options(composable_kernel PRIVATE -Weverything)
endif()
message("CMAKE_CXX_FLAGS: ${CMAKE_CXX_FLAGS}")

configure_file("${PROJECT_SOURCE_DIR}/composable_kernel/include/hip_version.hpp.in" "${PROJECT_BINARY_DIR}/composable_kernel/include/hip_version.hpp")
Chao Liu's avatar
Chao Liu committed
277

278
add_subdirectory(host)
Chao Liu's avatar
Chao Liu committed
279
add_subdirectory(device_operation)
280
281
add_subdirectory(example)
add_subdirectory(profiler)
282
add_subdirectory(test)