LNK2019 unresolved external symbol __std_reverse_trivially_swappable_8 when compiling proto_text.vcxproj

[MennoK]

System information

• Have I written custom code (as opposed to using a stock example script provided in TensorFlow): N/A
• OS Platform and Distribution (e.g., Linux Ubuntu 16.04): Windows 10
• TensorFlow installed from (source or binary): source
• TensorFlow version (use command below): 1.4
• Python version: 3.5
• Bazel version (if compiling from source): N/A
• GCC/Compiler version (if compiling from source): Cmake 3.10.2, swigwin 3.0.12, Visual studio 2017, but toolset 2015 v140
• CUDA/cuDNN version: N/A
• GPU model and memory: N/A
• Exact command to reproduce:

I'm following the CMake guide as described here: https://github.com/tensorflow/tensorflow/blob/fbddebee0bf07dadfb2b15ec678291dd5730ca99/tensorflow/contrib/cmake/README.md.

Error appears after following command:
MSBuild /p:Configuration=Release tf_tutorials_example_trainer.vcxproj

Describe the problem

Error when compiling the proto_text.vcxproj

Source code / logs

"C:\tensorflow\tensorflow\contrib\cmake\build\tf_tutorials_example_trainer.vcxproj" (default target) (1) ->
"C:\tensorflow\tensorflow\contrib\cmake\build\tf_cc_framework.vcxproj" (default target) (3) ->
"C:\tensorflow\tensorflow\contrib\cmake\build\tf_core_framework.vcxproj" (default target) (4) ->
"C:\tensorflow\tensorflow\contrib\cmake\build\proto_text.vcxproj" (default target) (5) ->
(Link target) ->
  libprotobuf.lib(text_format.obj) : error LNK2019: unresolved external symbol __std_reverse_trivially_swappable_8 referenced in function "void __cdecl std::_Buffered_inplace_merge_d
ivide_and_conquer2<class google::protobuf::Message const * *,__int64,class google::protobuf::Message const *,class google::protobuf::MapEntryMessageComparator>(class google::protobuf
::Message const * *,class google::protobuf::Message const * *,class google::protobuf::Message const * *,__int64,__int64,struct std::_Temporary_buffer<class google::protobuf::Message
const *> &,class google::protobuf::MapEntryMessageComparator,class google::protobuf::Message const * *,class google::protobuf::Message const * *,__int64,__int64)" (??$_Buffered_inpla
ce_merge_divide_and_conquer2@PEAPEBVMessage@protobuf@google@@_JPEBV123@VMapEntryMessageComparator@23@@std@@YAXPEAPEBVMessage@protobuf@google@@00_J1AEAU?$_Temporary_buffer@PEBVMessage
@protobuf@google@@@0@VMapEntryMessageComparator@23@0011@Z) [C:\tensorflow\tensorflow\contrib\cmake\build\proto_text.vcxproj]
  C:\tensorflow\tensorflow\contrib\cmake\build\Release\proto_text.exe : fatal error LNK1120: 1 unresolved externals [C:\tensorflow\tensorflow\contrib\cmake\build\proto_text.vcxproj]

[cy89]

@MennoK I'm not sure that you're running any of the "known-good" configurations in that README.md file (please correct me if that's wrong). I think we'll need to ask for help from the wider community.

[MennoK]

I'm using:

Microsoft Visual Studio 2017 with Visual C++ toolset 2015 (v140)
Anaconda 4.1.1 (Python 3.5 64-bit)
swigwin-3.0.12
Cuda and Cudnn aren't applicable in this case
CMake 3.10.2
Windows 10, but the target platform is set to 8.1

[truokola]

I have both VS2015 and VS2017 installed but I build TF with VS2015 by using cmake -G "Visual Studio 14 2015 Win64". But this setting is not transmitted to the child builds and therefore external projects, including protobuf, are built with VS2017. And in this case the build fails with exactly the error reported here.

However, if I rename the VS2017 installation directory to some nonsense so that CMake cannot find it, then also the externals use VS2015 and the build succeeds. So apparently this problem arises from mixing two different toolsets.

Similarly, if you are using the CMake -T switch to select the toolset, this selection does not propagate to the external builds.

[kenji0x02]

In my case, the error has been resolved by using CMake 3.6.0 known as good configurations version described here:
https://github.com/tensorflow/tensorflow/blob/fbddebee0bf07dadfb2b15ec678291dd5730ca99/tensorflow/contrib/cmake/README.md

[sushreebarsa]

@MennoK You are using an older version of tensorflow(1.x) which is not actively supported now, please verify it on the latest TF version 2.9.0 and let us know in case you face any errors.Thank you!

[google-ml-butler]

This issue has been automatically marked as stale because it has no recent activity. It will be closed if no further activity occurs. Thank you.

[google-ml-butler]

Closing as stale. Please reopen if you'd like to work on this further.