Build failing for custom tensorflow inside docker

[DreamingRaven]

I believe there needs to be a patch applied to allow tensorflow to build with newer grpc versions, as tensorflow only supports a single commit of grpc.

I believe this is related to:

• Update grpc dependency for glibc 2.30 compatibility tensorflow/tensorflow#33758
• Build tensorflow error as "external/grpc/src/core/lib/gpr/log_linux.cc:43:13: error: ambiguating new declaration of 'long int gettid()'" clearlinux/distribution#1151

Error message

cc1plus: warning: command line option '-Wno-implicit-function-declaration' is valid for C/ObjC but not for C++
ERROR: /home/archie/.cache/bazel/_bazel_archie/ddef78ddb52319615715a709a4a5202c/external/grpc/BUILD:507:1: C++ compilation of rule '@grpc//:gpr_base' failed (Exit 1)
external/grpc/src/core/lib/gpr/log_linux.cc:43:13: error: ambiguating new declaration of 'long int gettid()'
   43 | static long gettid(void) { return syscall(__NR_gettid); }
      |             ^~~~~~
In file included from /usr/include/unistd.h:1170,
                 from external/grpc/src/core/lib/gpr/log_linux.cc:41:
/usr/include/bits/unistd_ext.h:34:16: note: old declaration '__pid_t gettid()'
   34 | extern __pid_t gettid (void) __THROW;
      |                ^~~~~~
external/grpc/src/core/lib/gpr/log_linux.cc:43:13: warning: 'long int gettid()' defined but not used [-Wunused-function]
   43 | static long gettid(void) { return syscall(__NR_gettid); }
      |             ^~~~~~
Target //tensorflow/tools/pip_package:build_pip_package failed to build
Use --verbose_failures to see the command lines of failed build steps.
ERROR: /home/archie/git/tf-seal/tensorflow-c-17/tensorflow/tools/pip_package/BUILD:49:1 C++ compilation of rule '@grpc//:gpr_base' failed (Exit 1)
INFO: Elapsed time: 108.783s, Critical Path: 30.85s
INFO: 682 processes: 682 local.
FAILED: Build did NOT complete successfully
make: *** [Makefile:52: tensorflow] Error 1

Reproduce

Using this arch dockerfile in current directory:
https://github.com/DreamingRaven/nemesyst/blob/37b7c546f7e9c0c85f4916a8c57b07f4e2c26f90/examples/containers/tf_seal/Dockerfile

• sudo docker build -t archer/tf_seal .

• sudo docker run --gpus all -it archer/tf_seal bash or if there is no nvidia container toolkit sudo docker run -it archer/tf_seal bash to get an interactive terminal

• cd ~/git/tf-seal/

• make tensorflow

[justin1121]

Hey @DreamingRaven thanks for the issue. We're not actively supporting arch linux right now but would happily accept a PR that fixes this issue as long as it runs on ubuntu 19.04 as well.

[DreamingRaven]

No probs @justin1121 I figured as much, I will be working on a fix in the background slowly as its not a top priority for me at this current point in time. But at some point in the next month or so I will look further into it as it would be preferable to be able to use up-to-date arch tools inside the container.

However do you happen to have a working ubuntu dockerfile/ docker-compose for tf-seal at the moment so that I can compare, and check compatibility of any of my own changes? Sorry I didn't see it on first check.

[justin1121]

Awesome 😄. Definitely able to lend a hand if you get stuck anywhere. Feel free to reach out.

[mortendahl]

Hi @DreamingRaven, just wanted to check if there are any updates here?

[DreamingRaven]

@mortendahl Appologies, due to constraints + CKKS scheme, I have gone pure pybind11 + seal v 3.4.5 so I can have more fine grain control of the encryption and serialisation for databases. but I will need to move to GPU soon for neural network speedup which will likely lead me back here to create an arch compatible build.

[CharlesKung]

@DreamingRaven Hi, did you find any framework that could support GPU speedup for encrypted data? I'm not quite sure this lib could be usble.