Model broke NEURON CI

[JCGoran]

The model is failing in the CI with:

%model_dir%/arm64/wrap.cpp:244:5: error: no matching function for call to 'install_vector_method'
    install_vector_method(\"WrapAround\",WrapAround);
    ^~~~~~~~~~~~~~~~~~~~~
/include/oc_ansi.h:98:6: note: candidate function not viable: no known conversion from 'void (void *)' to 'double (*)(void *)' for 2nd argument
void install_vector_method(const char*, double (*)(void*));
     ^
1 error generated.

Looking at the NEURON docs, I managed to fix this by using the following diff:

diff --git a/mod/wrap.mod b/mod/wrap.mod
index ff25070..21f1e6f 100755
--- a/mod/wrap.mod
+++ b/mod/wrap.mod
@@ -26,18 +26,19 @@ double* Wrap(double* x,int n,int flen){
   return y;
 }
 
-void WrapAround(void* vv) {
+double WrapAround(void* vv) {
   double* x,*y;
   int vsz,fsz,i;
   vsz = vector_instance_px(vv,&x);
   fsz = (int) *getarg(1);
   if(fsz > vsz) {
     printf("WrapAround ERRA: invalid filter len %d > vector len %d!\n",fsz,vsz);
-    return;
+    return -1;
   }
   y = Wrap(x,vsz,fsz);
   for(i=0;i<vsz;i++) x[i]=y[i];
   free(y);
+  return 0;
 }
 
 ENDVERBATIM

Unfortunately, this isn't the only fix needed, as then we get:

-> \u001b[32mCompiling\u001b[0m %model_dir%/arm64/DynamicNetStim.cpp
%model_dir%/arm64/DynamicNetStim.cpp:416:7: error: functions that differ only in their return type cannot be overloaded
void* nrn_random_arg(int argpos);
~~~~~ ^
/include/nrnrandom.h:6:7: note: previous declaration is here
Rand* nrn_random_arg(int);
~~~~~ ^
%model_dir%/arm64/DynamicNetStim.cpp:428:13: warning: 'nrn_random_pick' is deprecated: non-void* overloads are preferred [-Wdeprecated-declarations]
                _lerand = nrn_random_pick(_p_donotuse);
                          ^
/include/nrnrandom.h:19:3: note: 'nrn_random_pick' has been explicitly marked deprecated here
[[deprecated(\"non-void* overloads are preferred\")]] double nrn_random_pick(void* r);
  ^
1 warning and 1 error generated

I think the DynamicNetStim mod file needs some additional fixes for random number generation as described here.