I want to use OpenMP target construct to run a loop on a GPU device. After many trial, I discovered that my gcc compiler fails to compile a recursive function.

After many trials, I isolated my problem in this small test.c function:

int func(int par){
 if(par<0) return par;
 return func(par--);
}

int main(void)
{
  int N = 100;
      #pragma omp target teams distribute parallel for 
      for (int i = 0; i<N; i++) {
          func(i);
  }
}

This is the way I compile it

gcc-12  -foffload=nvptx-none -fcf-protection=none -fno-stack-protector -foffload=-misa=sm_35  -fopenmp test.c -o testcpp.exe

This is the error I get:

test.c:5:5: error: alias definitions not supported in this configuration
    5 | int func(int par){
      |     ^
mkoffload: fatal error: x86_64-linux-gnu-accel-nvptx-none-gcc-12 returned 1 exit status
compilation terminated.
lto-wrapper: fatal error: /usr/lib/gcc/x86_64-linux-gnu/12//accel/nvptx-none/mkoffload returned 1 exit status
compilation terminated.
/usr/bin/ld: error: lto-wrapper failed
collect2: error: ld returned 1 exit status

The error disappears if I do not call func recursively.

Note: I am aware that this test case is useless, however, consider that this is just a small example to reproduce the behaviour I have in a more complicated set-up.

How to reproduce the issue: I set up this small google-colab notebook where you can reproduce the issue: https://colab.research.google.com/drive/1zHVwKBt1hrkrDvw5IxqPwxovpC8gkVwA

My question is: how can I compile a kernel with recursion using OpenMP target construct?