gmake[3]: *** No rule to make target benchmark/lib/libbenchmark.a', needed by bench/sg_benchmark'. Stop.

[Demokritus]

The description
By the end of building process after launching the command
$ ./build.sh bench
and after getting reached 100% in building
I got the following error message:
gmake[3]: *** No rule to make target benchmark/lib/libbenchmark.a', needed by bench/sg_benchmark'. Stop.
gmake[2]: *** [bench/CMakeFiles/sg_benchmark.dir/all] Error 2
gmake[1]: *** [bench/CMakeFiles/sg_benchmark.dir/rule] Error 2
gmake: *** [sg_benchmark] Error 2

Steps/Code to reproduce bug

1. As system versions of GCC, G++ didn't meet my needs due to the fact that they were obsolete
   I had to install them locally and specified paths to them altering options of cmake
   in build.sh:195 like that:

cmake -D CMAKE_C_COMPILER=/scratch/gsergei/pkgs/bin/gcc \
             -D CMAKE_CXX_COMPILER=/scratch/gsergei/pkgs/bin/g++ \

Other options were left as they were originally.

2. Along with that I had to reassign some environment variables.
   How they were altered:

export CUDACXX=/scratch/gsergei/cuda-10.2/bin/nvcc
export CUDA_HOME=/scratch/gsergei/cuda-10.2
export LD_LIBRARY_PATH=/scratch/gsergei/cuda-10.2/lib64:/usr/local/lib::/home/gsergei/.local/lib:/home/gsergei/.local/lib:/home/gsergei/.local/lib:/home/gsergei/.local/lib:/home/gsergei/.local/lib

3) Then compilation of "bench" has been started by the command:
$ ./build.sh bench 2>&1 | tee -a compile.log

Expected behavior
Successful completion of building was expected.

Environment details (please complete the following information):

• Environment location: [Bare-metal, Docker, Cloud(specify cloud provider)]
• Linux Distro/Architecture: [Ubuntu 16.04 amd64]
• GPU Model/Driver: [V100 and driver 396.44]
• CUDA: [10.2]
• Method of cuDF & cuML install: from source
  Cmake version - 3.18.5, gcc/g++ - 9.3.0

Additional context

Besides the info above you can find below the file with the output
of "print_env.sh" and the log with stdout of "build.sh".

Also I experienced another error before:
"CUDA 10.2 incompatible with GCC 9.3 and Clang 9.0"
It was the same as the issue on espressomd/espresso#3654
Their solution helped me though I adopted it for my case.
Due to inability to modify files in "/usr/local/cuda-10.2" I had to create a local copy
of cuda and edited a bit conditions in "crt/host_config.h" there.
As the result it accepted the version 9.3.0 of GCC.
compile.log
my_env.txt

[divyegala]

@teju85 any idea on this one?

[github-actions]

This issue has been labeled inactive-30d due to no recent activity in the past 30 days. Please close this issue if no further response or action is needed. Otherwise, please respond with a comment indicating any updates or changes to the original issue and/or confirm this issue still needs to be addressed. This issue will be labeled inactive-90d if there is no activity in the next 60 days.

[github-actions]

This issue has been labeled inactive-90d due to no recent activity in the past 90 days. Please close this issue if no further response or action is needed. Otherwise, please respond with a comment indicating any updates or changes to the original issue and/or confirm this issue still needs to be addressed.

[dantegd]

This should be working currently, but if not we will open a new issue with more recent versions