bazel-genfiles/ and *.pb.h not been generated with bazel-building tensorflow2.4.1 from source

[no-define]

System information

• OS Platform and Distribution ( Linux Ubuntu 18.04):
• TensorFlow installed from (source)
• TensorFlow version: Tags v2.4.1
• Python version: Python 3.7.6
• Installed using virtualenv? pip? conda?: conda
• Bazel version (if compiling from source): have tried bazel 3.4.0 、 bazel 3.1.0
• GCC/Compiler version (if compiling from source): gcc (GCC) 7.4.0
• CUDA/cuDNN version: cuda_11.1 , cudnn 8.0.5
• GPU model and memory: rtx3060
• nvidia-driver version: 460.91
• eigen version : eigen-3.3.90

I want to build tensorflow c++ api with bazel, no errors were reported during bazel-build process .
the command I used: bazel build --config=opt --config=cuda //tensorflow:libtensorflow_cc.so

however,while compiling a C++ program, I got the following error:
fatal error: tensorflow/core/framework/device_attributes.pb.h: No such file or directory #include "tensorflow/core/framework/device_attributes.pb.h" ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ compilation terminated.

After that I checked the tensorflow/core/framework folder and found no *.pb.h files in it ,only some .proto files,(etc. device_attributes.proto ) . Not even the bazel-genfiles folder.
The *.pb.h files should be generated by protobuf , what's wrong with my protobuf ?
the protobuf version and url refer to workspace.bzl

I don't know how to solve it. Thanks for your help.

[tilakrayal]

@no-define ,
Could you please update TensorFlow to the latest stable version v2.7 and let us know if you are facing the same error. Also please take a look at this issue 1 and 2 with the similar error.It helps.Thanks!

[no-define]

@tilakrayal hi,I've tried manually generating pb.h by command sudo protoc <protofile> --cpp_out=./
, now the original error has been fixed.

Now the error is tensorflow/third_party/eigen3/unsupported/Eigen/CXX11/Tensor:1:42: error: #include nested too deeply #include "unsupported/Eigen/CXX11/Tensor"

I'm going to find a fix for this problem.
Thank you anyway!

[google-ml-butler]

Are you satisfied with the resolution of your issue?
Yes
No

[tangjie77wd]

unsupported/Eigen/CXX11/Tensor

You could see the solution in https://blog.csdn.net/wd1603926823/article/details/92843830 because the author solved the similar problem like the following shows:
./unsupported/Eigen/CXX11/Tensor:1:42: error: #include nested too deeply tensorflow/core/lib/core/threadpool.cc:88:49: error: expected template-name before ‘<’ token struct ThreadPool::Impl : Eigen::ThreadPoolTempl<EigenEnvironment> { ^ tensorflow/core/lib/core/threadpool.cc:88:49: error: expected ‘{’ before ‘<’ token tensorflow/core/lib/core/threadpool.cc:88:49: error: expected unqualified-id before ‘<’ token tensorflow/core/lib/core/threadpool.cc:217:1: error: expected ‘}’ at end of input } // namespace tensorflow ^ tensorflow/core/lib/core/threadpool.cc:217:1: error: expected ‘}’ at end of input

[no-define]

tensorflow/third_party/eigen3/unsupported/Eigen/CXX11/Tensor:1:42: error: #include nested too deeply #include "unsupported/Eigen/CXX11/Tensor"

way to solve this problem : https://stackoverflow.com/questions/54779775/issues-including-eigen-for-simple-c-tensorflow-lite-test-program

find . -name "eigen*" -type d

then edit your CMakeList.txt :
include_directories( ${CMAKE_CURRENT_SOURCE_DIR}/tf/eigen )

try replace eigen path listed one by one.