py37-tensorflow@2.2.0 Upgrade fails.

[AndyWharton]

Upgrading py37-tensorflow@2.0.0_2+mkl+native fails during compilation with:

:info:build   external/local_config_cc/wrapped_clang '-D_FORTIFY_SOURCE=1' -fstack-protector -fcolor-diagnostics -Wall -Wthread-safety -Wself-assign -fno-omit-frame-pointer -g0 -O2 '-D_FORTIFY_SOURCE=1' -DNDEBUG '-std=c++11' -iquote external/mkl_dnn_v1 -iquote bazel-out/host/bin/external/mkl_dnn_v1 -iquote external/mkl_darwin -iquote bazel-out/host/bin/external/mkl_darwin -isystem external/mkl_dnn_v1/include -isystem bazel-out/host/bin/external/mkl_dnn_v1/include -isystem external/mkl_dnn_v1/src -isystem bazel-out/host/bin/external/mkl_dnn_v1/src -isystem external/mkl_dnn_v1/src/common -isystem bazel-out/host/bin/external/mkl_dnn_v1/src/common -isystem external/mkl_dnn_v1/src/cpu -isystem bazel-out/host/bin/external/mkl_dnn_v1/src/cpu -isystem external/mkl_dnn_v1/src/cpu/gemm -isystem bazel-out/host/bin/external/mkl_dnn_v1/src/cpu/gemm -isystem external/mkl_dnn_v1/src/cpu/xbyak -isystem bazel-out/host/bin/external/mkl_dnn_v1/src/cpu/xbyak -isystem external/mkl_darwin/include -isystem bazel-out/host/bin/external/mkl_darwin/include -MD -MF bazel-out/host/bin/external/mkl_dnn_v1/_objs/mkl_dnn/convolution_pd.d '-frandom-seed=bazel-out/host/bin/external/mkl_dnn_v1/_objs/mkl_dnn/convolution_pd.o' -isysroot __BAZEL_XCODE_SDKROOT__ -F__BAZEL_XCODE_SDKROOT__/System/Library/Frameworks -F__BAZEL_XCODE_DEVELOPER_DIR__/Platforms/MacOSX.platform/Developer/Library/Frameworks '-mmacosx-version-min=10.15' -g0 '-march=native' -g0 '-std=c++14' -fexceptions -DUSE_MKL -DUSE_CBLAS -UUSE_MKL -UUSE_CBLAS -no-canonical-prefixes -Wno-builtin-macro-redefined '-D__DATE__="redacted"' '-D__TIMESTAMP__="redacted"' '-D__TIME__="redacted"' -c external/mkl_dnn_v1/src/common/convolution_pd.cpp -o bazel-out/host/bin/external/mkl_dnn_v1/_objs/mkl_dnn/convolution_pd.o)
:info:build ERROR: /opt/local/var/macports/build/_opt_local_var_macports_sources_rsync.macports.org_macports_release_tarballs_ports_python_py-tensorflow/py37-tensorflow/work/d4cdc6a3d739c71cb25b01f8ca4c8702/external/mkl_dnn_v1/BUILD.bazel:52:1: C++ compilation of rule '@mkl_dnn_v1//:mkl_dnn' failed (Exit 1)
:info:build In file included from external/mkl_dnn_v1/src/cpu/ref_batch_normalization.cpp:22:
:info:build external/mkl_dnn_v1/src/common/dnnl_thread.hpp:42:10: fatal error: 'omp.h' file not found
:info:build #include <omp.h>
:info:build          ^~~~~~~
:info:build 1 error generated.

Should upgrades force MKL off until the build has been fixed?

[cjones051073 (Chris Jones)]

Ahh yes, good point. I neglected the fact users upgrading would retain the MKL variant....

[Chris Jones <jonesc@…>]

In 5a1632ed5897ac1a2917f94a091b53a9895242e7/macports-ports (master):

py-tensorflow: Force disable MKL support until build issues addressed
Closes: #60496