When trying to run unit tests com docker, when the build ended I got this message. There were about 5 errors regarding gcc. Namely:
gcc: error: unrecognized command line option ‘-mamx-int8’
gcc: error: unrecognized command line option ‘-mavxvnni’; did you mean ‘-mavx512vnni’?
While the others just said:
gcc: fatal error: Killed signal terminated program cc1plus
compilation terminated.
Could this be the reason for the build to fail? If so, how do I fix this?
Regarding my setup, I ran the command on docker ( docker run -it --rm -v $PWD:/tmp -w /tmp tensorflow/build:2.17-python3.12) and
gcc --version
gcc (Ubuntu 9.4.0-1ubuntu1~20.04.2) 9.4.0
On the same subject, when I re-ran the docker command followed by the bazel command, it restarted downloading and compiling about 15k files. Do I need to do this every time I want to run unit tests?
I have the same issue. I tested with the 2.12-python3.10, 2.15-python3.10, 2.15-python3.11, 2.16-python3.10 images and had the same or similar errors every time.
I also was facing the same. A possible workaround is to enable llvm+ clang support while running ./configure as is shown in https://www.tensorflow.org/install/source
With clang support enabled the compilation seems to take more time though.
Please refer to this link Docker Linux builds and follow the commands to build Linux packages from source. These images already contain the source code and dependencies required to build TensorFlow.
Let us know if the issue still persists. Thank you.