tensorflow, from source, compilation error, about 'class google::protobuf::Any' class Any; -

i have cloned latest version of latest tensorflow github , trying compile it. unfortunately getting errors.

• have written custom code (as opposed using stock example script provided in tensorflow): no
• os platform , distribution (e.g., linux ubuntu 16.04): linux ubuntu 16.04
• tensorflow installed (source or binary): trying install source
• python version: 3.5
• bazel version (if compiling source): 0.5.3
• cuda/cudnn version: cuda 8.0 cudnn 5
• gpu model , memory: gtx 1080 ti
• exact command reproduce: bazel build --config=opt --config=cuda //tensorflow/tools/pip_package:build_pip_package

i getting following compilation error:

error: tensorflow/tensorflow/contrib/gdr/build:52:1: c++ compilation of rule '//tensorflow/contrib/gdr:gdr_memory_manager' failed (exit 1) in file included tensorflow/contrib/gdr/gdr_memory_manager.cc:33:0: ./tensorflow/core/common_runtime/bfc_allocator.h: in member function 'std::__cxx11::string tensorflow::bfcallocator::chunk::debugstring(tensorflow::bfcallocator*, bool)': ./tensorflow/core/common_runtime/bfc_allocator.h:131:27: warning: comparison between signed , unsigned integer expressions [-wsign-compare]        if (recurse && prev != bfcallocator::kinvalidchunkhandle) {                            ^ ./tensorflow/core/common_runtime/bfc_allocator.h:135:27: warning: comparison between signed , unsigned integer expressions [-wsign-compare]        if (recurse && next != bfcallocator::kinvalidchunkhandle) {                            ^ tensorflow/contrib/gdr/gdr_memory_manager.cc: in member function 'virtual tensorflow::status tensorflow::{anonymous}::gdrmemorymanager::transportoptionsfromtensor(google::protobuf::any*, const tensorflow::tensor&, tensorflow::device*, tensorflow::devicecontext*, bool)': tensorflow/contrib/gdr/gdr_memory_manager.cc:488:28: error: invalid use of incomplete type 'class google::protobuf::any'    mutable_transport_options->packfrom(remote_mr);                             ^ in file included tensorflow/contrib/gdr/gdr_memory_manager.cc:18:0: ./tensorflow/contrib/gdr/gdr_memory_manager.h:23:7: note: forward declaration of 'class google::protobuf::any'  class any;        ^ tensorflow/contrib/gdr/gdr_memory_manager.cc: in member function 'virtual tensorflow::status tensorflow::{anonymous}::gdrmemorymanager::tensorfromtransportoptions(tensorflow::tensor*, const google::protobuf::any&, tensorflow::device*, tensorflow::devicecontext*, bool)': tensorflow/contrib/gdr/gdr_memory_manager.cc:497:25: error: invalid use of incomplete type 'const class google::protobuf::any'    if (!transport_options.unpackto(&remote_mr)) {                          ^ in file included tensorflow/contrib/gdr/gdr_memory_manager.cc:18:0: ./tensorflow/contrib/gdr/gdr_memory_manager.h:23:7: note: forward declaration of 'class google::protobuf::any'  class any;        ^ target //tensorflow/tools/pip_package:build_pip_package failed build use --verbose_failures see command lines of failed build steps. info: elapsed time: 1288.498s, critical path: 166.51s failed: build did not complete 

and configuration follow:

./configure have bazel 0.5.3 installed. please specify location of python. [default /usr/bin/python]: /usr/bin/python3 found possible python library paths: /usr/local/lib/python3.5/dist-packages /usr/lib/python3/dist-packages please input desired python library path use.  default /usr/local/lib/python3.5/dist-packages wish build tensorflow jemalloc malloc support? [y/n]:  jemalloc malloc support enabled tensorflow.  wish build tensorflow google cloud platform support? [y/n]:  no google cloud platform support enabled tensorflow.  wish build tensorflow hadoop file system support? [y/n]: y hadoop file system support enabled tensorflow.  wish build tensorflow xla jit support? [y/n]: y xla jit support enabled tensorflow.  wish build tensorflow gdr support? [y/n]: y gdr support enabled tensorflow.  wish build tensorflow verbs support? [y/n]: y verbs support enabled tensorflow.  wish build tensorflow opencl support? [y/n]: y opencl support enabled tensorflow.  please specify c++ compiler should used host c++ compiler. [default /usr/bin/g++]:  please specify c compiler should used host c compiler. [default /usr/bin/gcc]:  please specify location computecpp sycl 1.2 installed. [default /usr/local/computecpp]:  wish build tensorflow cuda support? [y/n]: y cuda support enabled tensorflow.  please specify cuda sdk version want use, e.g. 7.0. [leave empty default cuda 8.0]:  please specify location cuda 8.0 toolkit installed. refer readme.md more details. [default /usr/local/cuda]:  "please specify cudnn version want use. [leave empty default cudnn 6.0]: 5 please specify location cudnn 5 library installed. refer readme.md more details. [default /usr/local/cuda]: please specify list of comma-separated cuda compute capabilities want build with. can find compute capability of device at: https://developer.nvidia.com/cuda-gpus. please note each additional compute capability increases build time , binary size. [default is: 6.1] want use clang cuda compiler? [y/n]:  nvcc used cuda compiler.  please specify gcc should used nvcc host compiler. [default /usr/bin/gcc]:  wish build tensorflow mpi support? [y/n]:  no mpi support enabled tensorflow.  please specify optimization flags use during compilation when bazel option "--config=opt" specified [default -march=native]:  add "--config=mkl" bazel command build mkl support. please note mkl on macos or windows still not supported. if use local mkl instead of downloading, please set environment variable "tf_mkl_root" every time before build. configuration finished 

how can rid of error??

i have managed compile few weeks ago (version 1.3) before latest release. cannot solve error now.

thanks in advance