forked from jax-ml/jax
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy path.bazelrc
350 lines (292 loc) · 18 KB
/
.bazelrc
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
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
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
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
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
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
############################################################################
# All default build options below.
# Required by OpenXLA
# https://github.com/openxla/xla/issues/1323
build --nocheck_visibility
# Sets the default Apple platform to macOS.
build --apple_platform_type=macos
build --macos_minimum_os=10.14
# Make Bazel print out all options from rc files.
build --announce_rc
build --define open_source_build=true
build --spawn_strategy=standalone
build --enable_platform_specific_config
build --experimental_cc_shared_library
# Disable enabled-by-default TensorFlow features that we don't care about.
build --define=no_aws_support=true
build --define=no_gcp_support=true
build --define=no_hdfs_support=true
build --define=no_kafka_support=true
build --define=no_ignite_support=true
build --define=grpc_no_ares=true
build --define=tsl_link_protobuf=true
build -c opt
build --config=short_logs
build --copt=-DMLIR_PYTHON_PACKAGE_PREFIX=jaxlib.mlir.
# Later Bazel flag values override earlier values; if CUDA/ROCM/TPU are enabled,
# these values are overridden.
build --@xla//xla/python:enable_gpu=false
###########################################################################
build:posix --copt=-fvisibility=hidden
build:posix --copt=-Wno-sign-compare
build:posix --cxxopt=-std=c++17
build:posix --host_cxxopt=-std=c++17
build:avx_posix --copt=-mavx
build:avx_posix --host_copt=-mavx
build:avx_windows --copt=/arch=AVX
build:avx_linux --copt=-mavx
build:avx_linux --host_copt=-mavx
build:native_arch_posix --copt=-march=native
build:native_arch_posix --host_copt=-march=native
build:mkl_open_source_only --define=tensorflow_mkldnn_contraction_kernel=1
build:cuda --repo_env TF_NEED_CUDA=1
build:cuda --repo_env TF_NCCL_USE_STUB=1
# "sm" means we emit only cubin, which is forward compatible within a GPU generation.
# "compute" means we emit both cubin and PTX, which is larger but also forward compatible to future GPU generations.
build:cuda --action_env TF_CUDA_COMPUTE_CAPABILITIES="sm_50,sm_60,sm_70,sm_80,compute_90"
build:cuda --crosstool_top=@local_config_cuda//crosstool:toolchain
build:cuda --@local_config_cuda//:enable_cuda
build:cuda --@xla//xla/python:enable_gpu=true
build:cuda --@xla//xla/python:jax_cuda_pip_rpaths=true
build:cuda --define=xla_python_enable_gpu=true
# Build with nvcc for CUDA and clang for host
build:nvcc_clang --config=cuda
# Unfortunately, cuda_configure.bzl demands this for using nvcc + clang
build:nvcc_clang --action_env=TF_CUDA_CLANG="1"
build:nvcc_clang --action_env=TF_NVCC_CLANG="1"
build:nvcc_clang --@local_config_cuda//:cuda_compiler=nvcc
# Later Bazel flag values override earlier values.
# TODO(jieying): remove enable_gpu and xla_python_enable_gpu from build:cuda
# after the pluin is released.
build:cuda_plugin --@xla//xla/python:enable_gpu=false
build:cuda_plugin --define=xla_python_enable_gpu=false
# Force the linker to set RPATH, not RUNPATH. When resolving dynamic libraries,
# ld.so prefers in order: RPATH, LD_LIBRARY_PATH, RUNPATH. JAX sets RPATH to
# point to the $ORIGIN-relative location of the pip-installed NVIDIA CUDA
# packages.
# This has pros and cons:
# * pro: we'll ignore other CUDA installations, which has frequently confused
# users in the past. By setting RPATH, we'll always use the NVIDIA pip
# packages if they are installed.
# * con: the user cannot override the CUDA installation location
# via LD_LIBRARY_PATH, if the nvidia-... pip packages are installed. This is
# acceptable, because the workaround is "remove the nvidia-..." pip packages.
# The list of CUDA pip packages that JAX depends on are present in setup.py.
build:cuda --linkopt=-Wl,--disable-new-dtags
build:cuda_clang --@local_config_cuda//:cuda_compiler=clang
build:cuda_clang --action_env=CLANG_CUDA_COMPILER_PATH="/usr/lib/llvm-17/bin/clang"
build:cuda_clang --action_env=TF_CUDA_CLANG="1"
# Disable clang extention that rejects type definitions within offsetof.
# This was added in clang-16 by https://reviews.llvm.org/D133574.
# Can be removed once upb is updated, since a type definition is used within
# offset of in the current version of ubp.
# See https://github.com/protocolbuffers/upb/blob/9effcbcb27f0a665f9f345030188c0b291e32482/upb/upb.c#L183.
build:cuda_clang --copt=-Wno-gnu-offsetof-extensions
# Disable clang extention that rejects unknown arguments.
build:cuda_clang --copt=-Qunused-arguments
build:rocm --crosstool_top=@local_config_rocm//crosstool:toolchain
build:rocm --define=using_rocm=true --define=using_rocm_hipcc=true
build:rocm --@xla//xla/python:enable_gpu=true
build:rocm --define=xla_python_enable_gpu=true
build:rocm --repo_env TF_NEED_ROCM=1
build:rocm --action_env TF_ROCM_AMDGPU_TARGETS="gfx900,gfx906,gfx908,gfx90a,gfx1030"
build:nonccl --define=no_nccl_support=true
# Windows has a relatively short command line limit, which JAX has begun to hit.
# See https://docs.bazel.build/versions/main/windows.html
build:windows --features=compiler_param_file
build:windows --features=archive_param_file
# Tensorflow uses M_* math constants that only get defined by MSVC headers if
# _USE_MATH_DEFINES is defined.
build:windows --copt=/D_USE_MATH_DEFINES
build:windows --host_copt=/D_USE_MATH_DEFINES
# Make sure to include as little of windows.h as possible
build:windows --copt=-DWIN32_LEAN_AND_MEAN
build:windows --host_copt=-DWIN32_LEAN_AND_MEAN
build:windows --copt=-DNOGDI
build:windows --host_copt=-DNOGDI
# https://devblogs.microsoft.com/cppblog/announcing-full-support-for-a-c-c-conformant-preprocessor-in-msvc/
# otherwise, there will be some compiling error due to preprocessing.
build:windows --copt=/Zc:preprocessor
build:windows --cxxopt=/std:c++17
build:windows --host_cxxopt=/std:c++17
# Generate PDB files, to generate useful PDBs, in opt compilation_mode
# --copt /Z7 is needed.
build:windows --linkopt=/DEBUG
build:windows --host_linkopt=/DEBUG
build:windows --linkopt=/OPT:REF
build:windows --host_linkopt=/OPT:REF
build:windows --linkopt=/OPT:ICF
build:windows --host_linkopt=/OPT:ICF
build:windows --incompatible_strict_action_env=true
build:linux --config=posix
build:linux --copt=-Wno-unknown-warning-option
# Workaround for gcc 10+ warnings related to upb.
# See https://github.com/tensorflow/tensorflow/issues/39467
build:linux --copt=-Wno-stringop-truncation
build:linux --copt=-Wno-array-parameter
build:macos --config=posix
# Public cache for macOS builds. The "oct2023" in the URL is just the
# date when the bucket was created and can be disregarded. It still contains the
# latest cache that is being used.
build:macos_cache --remote_cache="https://storage.googleapis.com/tensorflow-macos-bazel-cache/oct2023" --remote_upload_local_results=false
# Cache pushes are limited to Jax's CI system.
build:macos_cache_push --config=macos_cache --remote_upload_local_results=true --google_default_credentials
# Suppress all warning messages.
build:short_logs --output_filter=DONT_MATCH_ANYTHING
#########################################################################
# RBE config options below.
# Flag to enable remote config
common --experimental_repo_remote_exec
build:rbe --repo_env=BAZEL_DO_NOT_DETECT_CPP_TOOLCHAIN=1
build:rbe --google_default_credentials
build:rbe --bes_backend=buildeventservice.googleapis.com
build:rbe --bes_results_url="https://source.cloud.google.com/results/invocations"
build:rbe --bes_timeout=600s
build:rbe --define=EXECUTOR=remote
build:rbe --flaky_test_attempts=3
build:rbe --jobs=200
build:rbe --remote_executor=grpcs://remotebuildexecution.googleapis.com
build:rbe --remote_timeout=3600
build:rbe --spawn_strategy=remote,worker,standalone,local
test:rbe --test_env=USER=anon
# Attempt to minimize the amount of data transfer between bazel and the remote
# workers:
build:rbe --remote_download_toplevel
build:rbe_linux --config=rbe
build:rbe_linux --action_env=PATH="/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/local/go/bin"
build:rbe_linux --host_javabase=@bazel_toolchains//configs/ubuntu16_04_clang/1.1:jdk8
build:rbe_linux --javabase=@bazel_toolchains//configs/ubuntu16_04_clang/1.1:jdk8
build:rbe_linux --host_java_toolchain=@bazel_tools//tools/jdk:toolchain_hostjdk8
build:rbe_linux --java_toolchain=@bazel_tools//tools/jdk:toolchain_hostjdk8
# Non-rbe settings we should include because we do not run configure
build:rbe_linux --config=avx_linux
build:rbe_linux --linkopt=-lrt
build:rbe_linux --host_linkopt=-lrt
build:rbe_linux --linkopt=-lm
build:rbe_linux --host_linkopt=-lm
# Use the GPU toolchain until the CPU one is ready.
# https://github.com/bazelbuild/bazel/issues/13623
build:rbe_cpu_linux_base --config=rbe_linux
build:rbe_cpu_linux_base --config=cuda_clang
build:rbe_cpu_linux_base --action_env=TF_NVCC_CLANG="1"
build:rbe_cpu_linux_base --host_crosstool_top="@ubuntu20.04-clang_manylinux2014-cuda12.3-cudnn8.9_config_cuda//crosstool:toolchain"
build:rbe_cpu_linux_base --crosstool_top="@ubuntu20.04-clang_manylinux2014-cuda12.3-cudnn8.9_config_cuda//crosstool:toolchain"
build:rbe_cpu_linux_base --extra_toolchains="@ubuntu20.04-clang_manylinux2014-cuda12.3-cudnn8.9_config_cuda//crosstool:toolchain-linux-x86_64"
build:rbe_cpu_linux_base --extra_execution_platforms="@ubuntu20.04-clang_manylinux2014-cuda12.3-cudnn8.9_config_platform//:platform"
build:rbe_cpu_linux_base --host_platform="@ubuntu20.04-clang_manylinux2014-cuda12.3-cudnn8.9_config_platform//:platform"
build:rbe_cpu_linux_base --platforms="@ubuntu20.04-clang_manylinux2014-cuda12.3-cudnn8.9_config_platform//:platform"
build:rbe_cpu_linux_py3.9 --config=rbe_cpu_linux_base --repo_env=TF_PYTHON_CONFIG_REPO="@ubuntu20.04-clang_manylinux2014-cuda12.3-cudnn8.9_config_python3.9"
build:rbe_cpu_linux_py3.9 --python_path="/usr/local/bin/python3.9"
build:rbe_cpu_linux_py3.10 --config=rbe_cpu_linux_base --repo_env=TF_PYTHON_CONFIG_REPO="@ubuntu20.04-clang_manylinux2014-cuda12.3-cudnn8.9_config_python3.10"
build:rbe_cpu_linux_py3.10 --python_path="/usr/local/bin/python3.10"
build:rbe_cpu_linux_py3.11 --config=rbe_cpu_linux_base --repo_env=TF_PYTHON_CONFIG_REPO="@ubuntu20.04-clang_manylinux2014-cuda12.3-cudnn8.9_config_python3.11"
build:rbe_cpu_linux_py3.11 --python_path="/usr/local/bin/python3.11"
build:rbe_cpu_linux_py3.12 --config=rbe_cpu_linux_base --repo_env=TF_PYTHON_CONFIG_REPO="@ubuntu20.04-clang_manylinux2014-cuda12.3-cudnn8.9_config_python3.12"
build:rbe_cpu_linux_py3.12 --python_path="/usr/local/bin/python3.12"
build:rbe_linux_cuda_base --config=rbe_linux
build:rbe_linux_cuda_base --config=cuda
build:rbe_linux_cuda_base --repo_env=REMOTE_GPU_TESTING=1
build:rbe_linux_cuda12.3_nvcc_base --config=rbe_linux_cuda_base
build:rbe_linux_cuda12.3_nvcc_base --config=cuda_clang
build:rbe_linux_cuda12.3_nvcc_base --action_env=TF_NVCC_CLANG="1"
build:rbe_linux_cuda12.3_nvcc_base --action_env=TF_CUDA_VERSION=12
build:rbe_linux_cuda12.3_nvcc_base --action_env=TF_CUDNN_VERSION=8
build:rbe_linux_cuda12.3_nvcc_base --action_env=CUDA_TOOLKIT_PATH="/usr/local/cuda-12"
build:rbe_linux_cuda12.3_nvcc_base --action_env=LD_LIBRARY_PATH="/usr/local/cuda:/usr/local/cuda/lib64:/usr/local/cuda/extras/CUPTI/lib64:/usr/local/tensorrt/lib"
build:rbe_linux_cuda12.3_nvcc_base --host_crosstool_top="@ubuntu20.04-clang_manylinux2014-cuda12.3-cudnn8.9_config_cuda//crosstool:toolchain"
build:rbe_linux_cuda12.3_nvcc_base --crosstool_top="@ubuntu20.04-clang_manylinux2014-cuda12.3-cudnn8.9_config_cuda//crosstool:toolchain"
build:rbe_linux_cuda12.3_nvcc_base --extra_toolchains="@ubuntu20.04-clang_manylinux2014-cuda12.3-cudnn8.9_config_cuda//crosstool:toolchain-linux-x86_64"
build:rbe_linux_cuda12.3_nvcc_base --extra_execution_platforms="@ubuntu20.04-clang_manylinux2014-cuda12.3-cudnn8.9_config_platform//:platform"
build:rbe_linux_cuda12.3_nvcc_base --host_platform="@ubuntu20.04-clang_manylinux2014-cuda12.3-cudnn8.9_config_platform//:platform"
build:rbe_linux_cuda12.3_nvcc_base --platforms="@ubuntu20.04-clang_manylinux2014-cuda12.3-cudnn8.9_config_platform//:platform"
build:rbe_linux_cuda12.3_nvcc_base --repo_env=TF_CUDA_CONFIG_REPO="@ubuntu20.04-clang_manylinux2014-cuda12.3-cudnn8.9_config_cuda"
build:rbe_linux_cuda12.3_nvcc_base --repo_env=TF_NCCL_CONFIG_REPO="@ubuntu20.04-clang_manylinux2014-cuda12.3-cudnn8.9_config_nccl"
# RBE machines have an older CUDA driver version, so we have to enable driver forward compatibility
build:rbe_linux_cuda12.3_nvcc_base --test_env=LD_LIBRARY_PATH=/usr/local/cuda/compat
build:rbe_linux_cuda12.3_nvcc_py3.9 --config=rbe_linux_cuda12.3_nvcc_base --repo_env=TF_PYTHON_CONFIG_REPO="@ubuntu20.04-clang_manylinux2014-cuda12.3-cudnn8.9_config_python3.9"
build:rbe_linux_cuda12.3_nvcc_py3.9 --python_path="/usr/local/bin/python3.9"
build:rbe_linux_cuda12.3_nvcc_py3.10 --config=rbe_linux_cuda12.3_nvcc_base --repo_env=TF_PYTHON_CONFIG_REPO="@ubuntu20.04-clang_manylinux2014-cuda12.3-cudnn8.9_config_python3.10"
build:rbe_linux_cuda12.3_nvcc_py3.10 --python_path="/usr/local/bin/python3.10"
build:rbe_linux_cuda12.3_nvcc_py3.11 --config=rbe_linux_cuda12.3_nvcc_base --repo_env=TF_PYTHON_CONFIG_REPO="@ubuntu20.04-clang_manylinux2014-cuda12.3-cudnn8.9_config_python3.11"
build:rbe_linux_cuda12.3_nvcc_py3.11 --python_path="/usr/local/bin/python3.11"
build:rbe_linux_cuda12.3_nvcc_py3.12 --config=rbe_linux_cuda12.3_nvcc_base --repo_env=TF_PYTHON_CONFIG_REPO="@ubuntu20.04-clang_manylinux2014-cuda12.3-cudnn8.9_config_python3.12"
build:rbe_linux_cuda12.3_nvcc_py3.12 --python_path="/usr/local/bin/python3.12"
# These you may need to change for your own GCP project.
build:tensorflow_testing_rbe --project_id=tensorflow-testing
common:tensorflow_testing_rbe_linux --remote_instance_name=projects/tensorflow-testing/instances/default_instance
build:tensorflow_testing_rbe_linux --config=tensorflow_testing_rbe
# START CROSS-COMPILE CONFIGS
# Set execution platform to Linux x86
# Note: Lot of the "host_" flags such as "host_cpu" and "host_crosstool_top"
# flags seem to be actually used to specify the execution platform details. It
# seems it is this way because these flags are old and predate the distinction
# between host and execution platform.
build:cross_compile_base --host_cpu=k8
build:cross_compile_base --host_crosstool_top=@xla//tools/toolchains/cross_compile/cc:cross_compile_toolchain_suite
build:cross_compile_base --extra_execution_platforms=@xla//tools/toolchains/cross_compile/config:linux_x86_64
# START LINUX AARCH64 CROSS-COMPILE CONFIGS
build:cross_compile_linux_arm64 --config=cross_compile_base
# Set the target CPU to Aarch64
build:cross_compile_linux_arm64 --platforms=@xla//tools/toolchains/cross_compile/config:linux_aarch64
build:cross_compile_linux_arm64 --cpu=aarch64
build:cross_compile_linux_arm64 --crosstool_top=@xla//tools/toolchains/cross_compile/cc:cross_compile_toolchain_suite
build:rbe_cross_compile_base --config=rbe
# JAX depends on some local Python headers that are configured as Genrule. They
# are present on the local host machine but not on the remote execution machine,
# leading to build failures. To resolve the issue, the following line is added
# to make sure all Genrule targets are excuted locally.
build:rbe_cross_compile_base --strategy=Genrule=standalone
# Due to the above strategy, all Genrule commands are executed locally, but the
# following actions invoke tools (E.g `flatc`, `llvm-tblgen`, etc.) that are
# only executabe on the RBE (x86) machine, so the strategy_regexp options are
# added to override and run the actions using remote strategy.
build:rbe_cross_compile_base --strategy_regexp='Generating code from table.*=remote'
build:rbe_cross_compile_base --strategy_regexp='Generating flatbuffer files.*=remote'
build:rbe_cross_compile_base --strategy_regexp='Executing genrule @llvm-project.*=remote'
# RBE cross-compile configs for Linux Aarch64
build:rbe_cross_compile_linux_arm64 --config=cross_compile_linux_arm64
build:rbe_cross_compile_linux_arm64 --config=rbe_cross_compile_base
# END LINUX AARCH64 CROSS-COMPILE CONFIGS
# START MACOS CROSS-COMPILE CONFIGS
build:cross_compile_macos_x86 --config=cross_compile_base
build:cross_compile_macos_x86 --config=nonccl
# Target Catalina (10.15) as the minimum supported OS
build:cross_compile_macos_x86 --action_env MACOSX_DEPLOYMENT_TARGET=10.15
# Set the target CPU to Darwin x86
build:cross_compile_macos_x86 --platforms=@xla//tools/toolchains/cross_compile/config:darwin_x86_64
build:cross_compile_macos_x86 --cpu=darwin
build:cross_compile_macos_x86 --crosstool_top=@xla//tools/toolchains/cross_compile/cc:cross_compile_toolchain_suite
# When RBE cross-compiling for macOS, we need to explicitly register the
# toolchain. Otherwise, oddly, RBE complains that a "docker container must be
# specified".
build:cross_compile_macos_x86 --extra_toolchains=@xla//tools/toolchains/cross_compile/config:macos-x86-cross-compile-cc-toolchain
# Map --platforms=darwin_x86_64 to --cpu=darwin and vice-versa to make selects()
# and transistions that use these flags work. The flag --platform_mappings needs
# to be set to a file that exists relative to the package path roots.
build:cross_compile_macos_x86 --platform_mappings=platform_mappings
# RBE cross-compile configs for Darwin x86
build:rbe_cross_compile_macos_x86 --config=cross_compile_macos_x86
build:rbe_cross_compile_macos_x86 --config=rbe_cross_compile_base
# END MACOS CROSS-COMPILE CONFIGS
# END CROSS-COMPILE CONFIGS
#############################################################################
#############################################################################
# Some configs to make getting some forms of debug builds. In general, the
# codebase is only regularly built with optimizations. Use 'debug_symbols' to
# just get symbols for the parts of XLA/PJRT that jaxlib uses.
# Or try 'debug' to get a build with assertions enabled and minimal
# optimizations.
# Include these in a local .bazelrc.user file as:
# build --config=debug_symbols
# Or:
# build --config=debug
#
# Additional files can be opted in for debug symbols by adding patterns
# to a per_file_copt similar to below.
#############################################################################
build:debug_symbols --strip=never --per_file_copt="xla/pjrt|xla/python@-g3"
build:debug --config debug_symbols -c fastbuild
# Load `.jax_configure.bazelrc` file written by build.py
try-import %workspace%/.jax_configure.bazelrc
# Load rc file with user-specific options.
try-import %workspace%/.bazelrc.user