CMakeLists.txt 22 KB
Newer Older
1
cmake_minimum_required(VERSION 3.14)
2
3
4
5
if(POLICY CMP0140)
  # policies CMP0140 not known to CMake until 3.25
  cmake_policy(SET CMP0140 NEW)
endif()
JD's avatar
JD committed
6

7
8
get_property(_GENERATOR_IS_MULTI_CONFIG GLOBAL PROPERTY GENERATOR_IS_MULTI_CONFIG)

9
10
# This has to be initialized before the project() command appears
# Set the default of CMAKE_BUILD_TYPE to be release, unless user specifies with -D.  MSVC_IDE does not use CMAKE_BUILD_TYPE
11
12
13
14
15
16
if(_GENERATOR_IS_MULTI_CONFIG)
    set(CMAKE_CONFIGURATION_TYPES "Debug;Release;RelWithDebInfo;MinSizeRel" CACHE STRING
            "Available build types (configurations) on multi-config generators")
else()
    set(CMAKE_BUILD_TYPE Release CACHE STRING
            "Choose the type of build, options are: None Debug Release RelWithDebInfo MinSizeRel.")
17
18
19
endif()

# Default installation path
20
if(NOT WIN32)
21
22
23
    set(CMAKE_INSTALL_PREFIX "/opt/rocm" CACHE PATH "")
endif()

24
set(version 1.1.0)
JD's avatar
JD committed
25
# Check support for CUDA/HIP in Cmake
26
project(composable_kernel VERSION ${version} LANGUAGES CXX HIP)
27
include(CTest)
Chao Liu's avatar
Chao Liu committed
28

29
30
31
# Usage: for customized Python location cmake -DCK_USE_ALTERNATIVE_PYTHON="/opt/Python-3.8.13/bin/python3.8"
# CK Codegen requires dataclass which is added in Python 3.7
# Python version 3.8 is required for general good practice as it is default for Ubuntu 20.04
32
if(NOT CK_USE_ALTERNATIVE_PYTHON)
33
   find_package(Python3 3.8 COMPONENTS Interpreter REQUIRED)
34
35
36
else()
   message("Using alternative python version")
   set(EXTRA_PYTHON_PATH)
37
   # this is overly restrictive, we may need to be more flexible on the following
38
39
40
41
42
43
44
45
   string(REPLACE "/bin/python3.8" "" EXTRA_PYTHON_PATH "${CK_USE_ALTERNATIVE_PYTHON}")
   message("alternative python path is: ${EXTRA_PYTHON_PATH}")
   find_package(Python3 3.6 COMPONENTS Interpreter REQUIRED)
   add_definitions(-DPython3_EXECUTABLE="${CK_USE_ALTERNATIVE_PYTHON}")
   set(Python3_EXECUTABLE "${CK_USE_ALTERNATIVE_PYTHON}")
   set(PYTHON_EXECUTABLE "${CK_USE_ALTERNATIVE_PYTHON}")
   set(ENV{LD_LIBRARY_PATH} "${EXTRA_PYTHON_PATH}/lib:$ENV{LD_LIBRARY_PATH}")
endif()
carlushuang's avatar
carlushuang committed
46

47
48
list(APPEND CMAKE_MODULE_PATH "${PROJECT_SOURCE_DIR}/cmake")

49
if (DTYPES)
50
51
52
53
54
55
56
57
    add_definitions(-DDTYPES)
    if (DTYPES MATCHES "int8")
        add_definitions(-DCK_ENABLE_INT8)
        set(CK_ENABLE_INT8 "ON")
    endif()
    if (DTYPES MATCHES "fp8")
        add_definitions(-DCK_ENABLE_FP8)
        set(CK_ENABLE_FP8 "ON")
58
59
60
61
    endif()
    if (DTYPES MATCHES "bf8")
        add_definitions(-DCK_ENABLE_BF8)
        set(CK_ENABLE_BF8 "ON")
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
    endif()
    if (DTYPES MATCHES "fp16")
        add_definitions(-DCK_ENABLE_FP16)
        set(CK_ENABLE_FP16 "ON")
    endif()
    if (DTYPES MATCHES "fp32")
        add_definitions(-DCK_ENABLE_FP32)
        set(CK_ENABLE_FP32 "ON")
    endif()
    if (DTYPES MATCHES "fp64")
        add_definitions(-DCK_ENABLE_FP64)
        set(CK_ENABLE_FP64 "ON")
    endif()
    if (DTYPES MATCHES "bf16")
        add_definitions(-DCK_ENABLE_BF16)
        set(CK_ENABLE_BF16 "ON")
    endif()
    message("DTYPES macro set to ${DTYPES}")
80
else()
81
    add_definitions(-DCK_ENABLE_INT8 -DCK_ENABLE_FP16 -DCK_ENABLE_FP32 -DCK_ENABLE_FP64 -DCK_ENABLE_BF16 -DCK_ENABLE_FP8 -DCK_ENABLE_BF8)
82
83
84
85
86
    set(CK_ENABLE_INT8 "ON")
    set(CK_ENABLE_FP16 "ON")
    set(CK_ENABLE_FP32 "ON")
    set(CK_ENABLE_FP64 "ON")
    set(CK_ENABLE_BF16 "ON")
87
88
    set(CK_ENABLE_FP8 "ON")
    set(CK_ENABLE_BF8 "ON")
89
90
endif()

91
92
#for f8/bf8_t type
add_compile_options(-Wno-bit-int-extension)
93
add_compile_options(-Wno-pass-failed)
94
add_compile_options(-Wno-switch-default)
95

96
97
if(DL_KERNELS)
    add_definitions(-DDL_KERNELS)
98
    set(CK_ENABLE_DL_KERNELS "ON")
99
endif()
100
101
102
103
option(CK_USE_CODEGEN "Enable codegen library" OFF)
if(CK_USE_CODEGEN)
    add_definitions(-DCK_USE_CODEGEN)
endif()
104

105
106
include(getopt)

107
108
109
# CK version file to record release version as well as git commit hash
find_package(Git REQUIRED)
execute_process(COMMAND "${GIT_EXECUTABLE}" rev-parse HEAD OUTPUT_VARIABLE COMMIT_ID OUTPUT_STRIP_TRAILING_WHITESPACE)
110
configure_file(include/ck/version.h.in ${CMAKE_CURRENT_BINARY_DIR}/include/ck/version.h)
JD's avatar
JD committed
111

112
set(ROCM_SYMLINK_LIBS OFF)
Anthony Chang's avatar
Anthony Chang committed
113
find_package(ROCM REQUIRED PATHS /opt/rocm)
JD's avatar
JD committed
114
115
116
117
118
119

include(ROCMInstallTargets)
include(ROCMPackageConfigHelpers)
include(ROCMSetupVersion)
include(ROCMInstallSymlinks)
include(ROCMCreatePackage)
Chao Liu's avatar
Chao Liu committed
120
include(CheckCXXCompilerFlag)
121
include(ROCMCheckTargetIds)
JD's avatar
JD committed
122
include(TargetFlags)
123
124
125

rocm_setup_version(VERSION ${version})

126
list(APPEND CMAKE_PREFIX_PATH ${CMAKE_INSTALL_PREFIX} ${CMAKE_INSTALL_PREFIX}/llvm ${CMAKE_INSTALL_PREFIX}/hip /opt/rocm /opt/rocm/llvm /opt/rocm/hip "$ENV{ROCM_PATH}" "$ENV{HIP_PATH}")
JD's avatar
JD committed
127

128
message("GPU_TARGETS= ${GPU_TARGETS}")
129
130
131
132
133
134
message("GPU_ARCHS= ${GPU_ARCHS}")
if(GPU_ARCHS)
    #disable GPU_TARGETS to avoid conflicts, this needs to happen before we call hip package
    unset(GPU_TARGETS CACHE)
    unset(AMDGPU_TARGETS CACHE)
endif()
135

136
137
138
139
140
141
find_package(hip)
# No assumption that HIP kernels are launched with uniform block size for backward compatibility
# SWDEV-413293 and https://reviews.llvm.org/D155213
math(EXPR hip_VERSION_FLAT "(${hip_VERSION_MAJOR} * 1000 + ${hip_VERSION_MINOR}) * 100000 + ${hip_VERSION_PATCH}")
message("hip_version_flat=${hip_VERSION_FLAT}")

142
message("checking which targets are supported")
143
144
145
146
147
148
149
150
151
#In order to build just the CK library (without tests and examples) for all supported GPU targets
#use -D GPU_ARCHS="gfx908;gfx90a;gfx940;gfx941;gfx942;gfx1030;gfx1100;gfx1101;gfx1102;gfx1200;gfx1201" 
#the GPU_TARGETS flag will be reset in this case in order to avoid conflicts.
#
#In order to build CK along with all tests and examples it should be OK to set GPU_TARGETS to just 1 or 2 similar architectures.
if(NOT ENABLE_ASAN_PACKAGING)
    if(NOT WIN32 AND ${hip_VERSION_FLAT} LESS 600300000)
        # WORKAROUND: compiler does not yet fully support gfx12 targets, need to fix version above
        set(CK_GPU_TARGETS "gfx908;gfx90a;gfx940;gfx941;gfx942;gfx1030;gfx1100;gfx1101;gfx1102")
152
    else()
153
        set(CK_GPU_TARGETS "gfx908;gfx90a;gfx940;gfx941;gfx942;gfx1030;gfx1100;gfx1101;gfx1102;gfx1200;gfx1201")
154
    endif()
155
else()
156
157
158
159
160
161
162
163
164
    #build CK only for xnack-supported targets when using ASAN
    set(CK_GPU_TARGETS "gfx908:xnack+;gfx90a:xnack+;gfx940:xnack+;gfx941:xnack+;gfx942:xnack+")
endif()

#if user set GPU_ARCHS on the cmake command line, overwrite default target list with user's list
#otherwise, if user set GPU_TARGETS, use that set of targets
if(GPU_ARCHS)
    set(CK_GPU_TARGETS ${GPU_ARCHS})
else()
165
    if(GPU_TARGETS)
166
        set(CK_GPU_TARGETS ${GPU_TARGETS})
167
168
169
    endif()
endif()

170
171
172
#make sure all the targets on the list are actually supported by the current compiler
rocm_check_target_ids(SUPPORTED_GPU_TARGETS
        TARGETS ${CK_GPU_TARGETS})
173

174
message("Building CK for the following targets: ${SUPPORTED_GPU_TARGETS}")
175
176
177
178
179
180

if (GPU_TARGETS)
    if (GPU_TARGETS MATCHES "gfx9")
        add_definitions(-DCK_USE_XDL)
        set(CK_USE_XDL "ON")
    endif()
181
    if (GPU_TARGETS MATCHES "gfx11" OR GPU_TARGETS MATCHES "gfx12")
182
183
184
185
186
187
188
189
190
        add_definitions(-DCK_USE_WMMA)
        set(CK_USE_WMMA "ON")
    endif()
else()
    add_definitions(-DCK_USE_WMMA -DCK_USE_XDL)
    set(CK_USE_XDL "ON")
    set(CK_USE_WMMA "ON")
endif()

191
192
193
# CK config file to record supported datatypes, etc.
configure_file(include/ck/config.h.in ${CMAKE_CURRENT_BINARY_DIR}/include/ck/config.h)

194
if(NOT WIN32 AND ${hip_VERSION_FLAT} GREATER 500723302)
195
196
197
198
199
  check_cxx_compiler_flag("-fno-offload-uniform-block" HAS_NO_OFFLOAD_UNIFORM_BLOCK)
  if(HAS_NO_OFFLOAD_UNIFORM_BLOCK)
    message("Adding the fno-offload-uniform-block compiler flag")
    add_compile_options(-fno-offload-uniform-block)
  endif()
200
endif()
201
if(NOT WIN32 AND ${hip_VERSION_FLAT} GREATER 600140090)
202
203
204
205
206
  check_cxx_compiler_flag("-mllvm -enable-post-misched=0" HAS_ENABLE_POST_MISCHED)
  if(HAS_ENABLE_POST_MISCHED)
    message("Adding the enable-post-misched=0 compiler flag")
    add_compile_options("SHELL: -mllvm -enable-post-misched=0")
  endif()
207
endif()
208
209
210
set(check-coerce)
check_cxx_compiler_flag(" -mllvm -amdgpu-coerce-illegal-types=1" check-coerce)
if(NOT WIN32 AND check-coerce AND ${hip_VERSION_FLAT} GREATER 600241132 AND ${hip_VERSION_FLAT} LESS 600300000)
211
212
213
214
215
216
217
   message("Adding the amdgpu-coerce-illegal-types=1")
   add_compile_options("SHELL: -mllvm -amdgpu-coerce-illegal-types=1")
endif()
if(NOT WIN32 AND ${hip_VERSION_FLAT} GREATER 600241132)
   message("Adding -amdgpu-early-inline-all=true and -amdgpu-function-calls=false")
   add_compile_options("SHELL: -mllvm -amdgpu-early-inline-all=true")
   add_compile_options("SHELL: -mllvm -amdgpu-function-calls=false")
218
endif()
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
#
# Seperate linking jobs from compiling
# Too many concurrent linking jobs can break the build
# Copied from LLVM
set(CK_PARALLEL_LINK_JOBS "" CACHE STRING
  "Define the maximum number of concurrent link jobs (Ninja only).")
if(CMAKE_GENERATOR MATCHES "Ninja")
  if(CK_PARALLEL_LINK_JOBS)
    set_property(GLOBAL APPEND PROPERTY JOB_POOLS link_job_pool=${CK_PARALLEL_LINK_JOBS})
    set(CMAKE_JOB_POOL_LINK link_job_pool)
  endif()
elseif(CK_PARALLEL_LINK_JOBS)
  message(WARNING "Job pooling is only available with Ninja generators.")
endif()
# Similar for compiling
set(CK_PARALLEL_COMPILE_JOBS "" CACHE STRING
  "Define the maximum number of concurrent compile jobs (Ninja only).")
if(CMAKE_GENERATOR MATCHES "Ninja")
  if(CK_PARALLEL_COMPILE_JOBS)
    set_property(GLOBAL APPEND PROPERTY JOB_POOLS compile_job_pool=${CK_PARALLEL_COMPILE_JOBS})
    set(CMAKE_JOB_POOL_COMPILE compile_job_pool)
  endif()
elseif(CK_PARALLEL_COMPILE_JOBS)
  message(WARNING "Job pooling is only available with Ninja generators.")
endif()


246
option(USE_BITINT_EXTENSION_INT4 "Whether to enable clang's BitInt extension to provide int4 data type." OFF)
Illia Silin's avatar
Illia Silin committed
247
option(USE_OPT_GFX11 "Whether to enable LDS cumode and Wavefront32 mode for GFX11 silicons." OFF)
Adam Osewski's avatar
Adam Osewski committed
248
249
250
251
252
253
254

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()

Illia Silin's avatar
Illia Silin committed
255
if(USE_OPT_GFX11)
256
257
    add_compile_options(-mcumode)
    add_compile_options(-mno-wavefrontsize64)
Illia Silin's avatar
Illia Silin committed
258
    message("CK compiled with USE_OPT_GFX11 set to ${USE_OPT_GFX11}")
259
260
endif()

261
262
263
264
265
## Threads
set(THREADS_PREFER_PTHREAD_FLAG ON)
find_package(Threads REQUIRED)
link_libraries(Threads::Threads)

266
## C++
Chao Liu's avatar
Chao Liu committed
267
set(CMAKE_CXX_STANDARD 17)
Chao Liu's avatar
Chao Liu committed
268
set(CMAKE_CXX_STANDARD_REQUIRED ON)
Chao Liu's avatar
Chao Liu committed
269
set(CMAKE_CXX_EXTENSIONS OFF)
270
271
message("CMAKE_CXX_COMPILER: ${CMAKE_CXX_COMPILER}")

272
273
274
275
276
277
278
279
280
281
# https://gcc.gnu.org/onlinedocs/libstdc++/manual/using_macros.html
# _GLIBCXX_ASSERTIONS
# Undefined by default. When defined, enables extra error checking in the form of
# precondition assertions, such as bounds checking in strings and null pointer
# checks when dereferencing smart pointers
option(USE_GLIBCXX_ASSERTIONS "Turn on additional c++ library checks." OFF)
if(USE_GLIBCXX_ASSERTIONS)
  add_compile_options(-Wp,-D_GLIBCXX_ASSERTIONS)
endif()

282
283
284
285
286
## HIP
set(CMAKE_HIP_PLATFORM amd)
set(CMAKE_HIP_COMPILER ${CMAKE_CXX_COMPILER})
set(CMAKE_HIP_EXTENSIONS ON)
message("CMAKE_HIP_COMPILER: ${CMAKE_HIP_COMPILER}")
Chao Liu's avatar
Chao Liu committed
287

288
## OpenMP
Chao Liu's avatar
Chao Liu committed
289
290
291
292
293
294
295
296
297
298
299
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
300

Chao Liu's avatar
Chao Liu committed
301
302
303
304
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
305

Chao Liu's avatar
Chao Liu committed
306
307
link_libraries(${OpenMP_gomp_LIBRARY})
link_libraries(${OpenMP_pthread_LIBRARY})
Chao Liu's avatar
Chao Liu committed
308

309
310
## HIP
find_package(HIP REQUIRED)
JD's avatar
JD committed
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
# 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}")
330
link_libraries(hip::device)
331
332
333
334
335
if(CK_hip_VERSION VERSION_GREATER_EQUAL 6.0.23494)
    add_compile_definitions(__HIP_PLATFORM_AMD__=1)
else()
    add_compile_definitions(__HIP_PLATFORM_HCC__=1)
endif()
336

Chao Liu's avatar
Chao Liu committed
337
338
## tidy
include(EnableCompilerWarnings)
JD's avatar
JD committed
339
set(CK_TIDY_ERRORS ERRORS * -readability-inconsistent-declaration-parameter-name)
Chao Liu's avatar
Chao Liu committed
340
if(CMAKE_CXX_COMPILER MATCHES ".*hcc" OR CMAKE_CXX_COMPILER MATCHES ".*clang\\+\\+")
JD's avatar
JD committed
341
    set(CK_TIDY_CHECKS -modernize-use-override -readability-non-const-parameter)
Chao Liu's avatar
Chao Liu committed
342
# Enable tidy on hip
JD's avatar
JD committed
343
344
elseif(CK_BACKEND STREQUAL "HIP" OR CK_BACKEND STREQUAL "HIPNOGPU")
    set(CK_TIDY_ERRORS ALL)
Chao Liu's avatar
Chao Liu committed
345
346
endif()

JD's avatar
JD committed
347

Chao Liu's avatar
Chao Liu committed
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
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
439
440
        ${CK_TIDY_CHECKS}
        ${CK_TIDY_ERRORS}
Chao Liu's avatar
Chao Liu committed
441
442
443
    HEADER_FILTER
        "\.hpp$"
    EXTRA_ARGS
JD's avatar
JD committed
444
        -DCK_USE_CLANG_TIDY
Chao Liu's avatar
Chao Liu committed
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
)

include(CppCheck)
enable_cppcheck(
    CHECKS
        warning
        style
        performance
        portability
    SUPPRESS
        ConfigurationNotChecked
        constStatement
        duplicateCondition
        noExplicitConstructor
        passedByValue
Chao Liu's avatar
Chao Liu committed
460
        preprocessorErrorDirective
Chao Liu's avatar
Chao Liu committed
461
462
463
464
465
466
467
        shadowVariable
        unusedFunction
        unusedPrivateFunction
        unusedStructMember
        unmatchedSuppression
    FORCE
    SOURCES
Chao Liu's avatar
Chao Liu committed
468
        library/src
Chao Liu's avatar
Chao Liu committed
469
470
471
    INCLUDE
        ${CMAKE_CURRENT_SOURCE_DIR}/include
        ${CMAKE_CURRENT_BINARY_DIR}/include
Chao Liu's avatar
Chao Liu committed
472
        ${CMAKE_CURRENT_SOURCE_DIR}/library/include
Chao Liu's avatar
Chao Liu committed
473
474
475
476
    DEFINE
        CPPCHECK=1
        __linux__=1
)
Chao Liu's avatar
Chao Liu committed
477

JD's avatar
JD committed
478
479
480
481
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)

482
# set CK project include directories
Chao Liu's avatar
Chao Liu committed
483
include_directories(BEFORE
484
    ${PROJECT_BINARY_DIR}/include
Chao Liu's avatar
Chao Liu committed
485
486
    ${PROJECT_SOURCE_DIR}/include
    ${PROJECT_SOURCE_DIR}/library/include
487
    ${HIP_INCLUDE_DIRS}
JD's avatar
JD committed
488
)
Chao Liu's avatar
Chao Liu committed
489

JD's avatar
JD committed
490
491
SET(BUILD_DEV ON CACHE BOOL "BUILD_DEV")
if(BUILD_DEV)
492
493
    add_compile_options(-Werror)
    add_compile_options(-Weverything)
JD's avatar
JD committed
494
495
496
endif()
message("CMAKE_CXX_FLAGS: ${CMAKE_CXX_FLAGS}")

497
498
499
500
501
502
503
if("${CMAKE_CXX_COMPILER_ID}" MATCHES "Clang")
    add_compile_options(-fcolor-diagnostics)
endif()
if("${CMAKE_CXX_COMPILER_ID}" STREQUAL "GNU" AND CMAKE_CXX_COMPILER_VERSION VERSION_GREATER 4.9)
    add_compile_options(-fdiagnostics-color=always)
endif()

Anthony Chang's avatar
Anthony Chang committed
504
505
add_custom_target(check COMMAND ${CMAKE_CTEST_COMMAND} --output-on-failure -C ${CMAKE_CFG_INTDIR})

506
507
508
509
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 ${dir_list})
510
511
512
513
514
set(target_dir)
IF(IS_DIRECTORY "${PROJECT_SOURCE_DIR}/library/src/tensor_operation_instance/gpu/${subdir_path}")
    set(cmake_instance)
    file(READ "${PROJECT_SOURCE_DIR}/library/src/tensor_operation_instance/gpu/${subdir_path}/CMakeLists.txt" cmake_instance)
    set(add_inst 0)
515
    if(("${cmake_instance}" MATCHES "fp8" OR "${cmake_instance}" MATCHES "_f8") AND DTYPES MATCHES "fp8")
516
        set(add_inst 1)
517
    endif()
518
    if(("${cmake_instance}" MATCHES "bf8" OR "${cmake_instance}" MATCHES "_b8") AND DTYPES MATCHES "bf8")
519
520
        set(add_inst 1)
    endif()
521
    if(("${cmake_instance}" MATCHES "fp16" OR "${cmake_instance}" MATCHES "_f16") AND DTYPES MATCHES "fp16")
522
        set(add_inst 1)
523
    endif()
524
    if(("${cmake_instance}" MATCHES "fp32" OR "${cmake_instance}" MATCHES "_f32") AND DTYPES MATCHES "fp32")
525
        set(add_inst 1)
526
    endif()
527
    if(("${cmake_instance}" MATCHES "fp64" OR "${cmake_instance}" MATCHES "_f64") AND DTYPES MATCHES "fp64")
528
        set(add_inst 1)
529
    endif()
530
    if(("${cmake_instance}" MATCHES "bf16" OR "${cmake_instance}" MATCHES "_b16") AND DTYPES MATCHES "bf16")
531
        set(add_inst 1)
532
    endif()
533
    if(("${cmake_instance}" MATCHES "int8" OR "${cmake_instance}" MATCHES "_i8") AND DTYPES MATCHES "int8")
534
        set(add_inst 1)
535
536
    endif()
    if(NOT "${cmake_instance}" MATCHES "DTYPES")
537
        set(add_inst 1)
538
539
    endif()
    if(add_inst EQUAL 1 OR NOT DEFINED DTYPES)
540
        list(APPEND CK_DEVICE_INSTANCES device_${subdir_path}_instance)
541
542
    endif()
ENDIF()
543
ENDFOREACH()
544

545
add_custom_target(instances DEPENDS utility;${CK_DEVICE_INSTANCES}  SOURCES ${INSTANCE_FILES})
Jing Zhang's avatar
Jing Zhang committed
546
#add_subdirectory(library)
547

548
if(NOT GPU_ARCHS)
549
   rocm_package_setup_component(tests
550
551
        LIBRARY_NAME composablekernel
        PACKAGE_NAME tests # Prevent -static suffix on package name
552
   )
553

554
   rocm_package_setup_component(examples
555
556
        LIBRARY_NAME composablekernel
        PACKAGE_NAME examples
557
   )
558
   add_subdirectory(example)
Jing Zhang's avatar
Jing Zhang committed
559
560
561
   #if(BUILD_TESTING)
	   #add_subdirectory(test)
   #endif()
562
endif()
JD's avatar
JD committed
563

564
565
566
567
rocm_package_setup_component(profiler
    LIBRARY_NAME composablekernel
    PACKAGE_NAME ckprofiler
)
Jing Zhang's avatar
Jing Zhang committed
568
#add_subdirectory(profiler)
569

Jing Zhang's avatar
Jing Zhang committed
570
571
572
#if(CK_USE_CODEGEN AND (GPU_TARGETS MATCHES "gfx9" OR GPU_ARCHS))
  #add_subdirectory(codegen)
#endif()
arai713's avatar
arai713 committed
573

JD's avatar
JD committed
574
575
576
577
578
579
580
581
582
#Create an interface target for the include only files and call it "composablekernels"
include(CMakePackageConfigHelpers)

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

Anthony Chang's avatar
Anthony Chang committed
583
configure_package_config_file(${CMAKE_CURRENT_SOURCE_DIR}/Config.cmake.in
584
585
586
    "${CMAKE_CURRENT_BINARY_DIR}/composable_kernelConfig.cmake"
    INSTALL_DESTINATION ${CMAKE_INSTALL_LIBDIR}/cmake/composable_kernel
    NO_CHECK_REQUIRED_COMPONENTS_MACRO
JD's avatar
JD committed
587
588
)

589
rocm_install(FILES
JD's avatar
JD committed
590
591
    "${CMAKE_CURRENT_BINARY_DIR}/composable_kernelConfig.cmake"
    "${CMAKE_CURRENT_BINARY_DIR}/composable_kernelConfigVersion.cmake"
Anthony Chang's avatar
Anthony Chang committed
592
    DESTINATION ${CMAKE_INSTALL_LIBDIR}/cmake/composable_kernel
JD's avatar
JD committed
593
)
594

595
# Install CK version and configuration files
596
rocm_install(FILES
597
598
599
600
601
    ${PROJECT_BINARY_DIR}/include/ck/version.h
    ${PROJECT_BINARY_DIR}/include/ck/config.h
    DESTINATION ${CMAKE_INSTALL_INCLUDEDIR}/ck/
)

602
603
604
605
606
607
608
609
610
611
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
)