The last sed command used for adding the import statement and the '\n' was
working as intended on Linux, but was failing on macOS, due to the difference in
sed behavior between GNU and BSD.
Fixed the issue by using an actual newline character with $'\n', which correctly
works on both platforms.
Fixes https://github.com/tensorflow/tensorflow/issues/39756
PiperOrigin-RevId: 313691192
Change-Id: I96e5544c03641bae05753e8f7d1346c8aa1c0f6e
Rather than defaulting to public, default to private and explicitly mark the
targets that should be externally visible. This causes the visibility to line up
with what is available in the BUILD.system file.
PiperOrigin-RevId: 292064746
Change-Id: I4964983457b3e5f31200531fe35f2aa316c73833
This commit:
adf6e22e4a (diff-4bdae93a605044a27f9df49a92bcf398L129)
Removed the linkopts section from the cc_binary 'flatc' from
third_party/flatbuffers/BUILD.bazel. That caused builds to fail with:
/usr/bin/ld: bazel-out/host/bin/external/flatbuffers/src/libflatbuffers.a(idl_parser.o): undefined reference to symbol 'cos@@GLIBC_2.17'
/dt7/lib/powerpc64le-linux-gnu/libm.so.6: error adding symbols: DSO missing from command line
collect2: error: ld returned 1 exit status
(This is GPU builds on ppc64le)
I tried passing `--linkopt=-lm` on the bazel build command but it still failed.
Restoring the original linkopts section does resolve this problem. (and it
seems better to have the linkopt here than globally for the entire build)
A simple `linkopts = ["-lm"],` works for ppc64le, but I figured for windows and
other platforms it was better to restore the original values.
Current the visualization script for TensorFlow Lite files only works if you build through Bazel. To make it more accessible, this change uses Flatbuffer's new Python interface to extract the information from the file, rather than calling out to an external tool. It also adds some tests.
A lot of the changes here are related to upgrading to the latest version of Flatbuffers, which has an impact on other parts of the code.
PiperOrigin-RevId: 291255530
Change-Id: I28a64a182e6197a58926d02c41988e4791c60fce
Current the visualization script for TensorFlow Lite files only works if you build through Bazel. To make it more accessible, this change uses Flatbuffer's new Python interface to extract the information from the file, rather than calling out to an external tool. It also adds some tests.
A lot of the changes here are related to upgrading to the latest version of Flatbuffers, whi...
PiperOrigin-RevId: 290887928
Change-Id: I90403c1f4233f472ee5d7c4d21eed8233cb7d008
Current the visualization script for TensorFlow Lite files only works if you build through Bazel. To make it more accessible, this change uses Flatbuffer's new Python interface to extract the information from the file, rather than calling out to an external tool. It also adds some tests.
A lot of the changes here are related to upgrading to the latest version of Flatbuffers, which has an impact on other parts of the code.
PiperOrigin-RevId: 290788860
Change-Id: I4db3442110c48e01da2b5696b693562c806735fd
- Update the flatbuffer download URL to use a pinned version.
- Had to provide a mirror url that doesn't exist, since it is required
by the validation flow, the flatbuffer version will be added later.
PiperOrigin-RevId: 214487576
* Update flatbuffers to 1.9.0
This fix updates flatbuffers to 1.9.0. The previous version
used (971a681) in tf was released last year, and is not
a versioned release. This fix updates to the latest versioned
release of 1.9.0.
Signed-off-by: Yong Tang <yong.tang.github@outlook.com>
* Add missing files of java_generator.cc to fix build error.
Signed-off-by: Yong Tang <yong.tang.github@outlook.com>
