sandbox/bugs/memory_leak_solve.c

Memory leak in solve.h

We use the kuramoto.c example to test for memory leaks in solve.h. At each iteration, the memory used by realloc scalar increases. To see this, we run with

CFLAGS="-O2 -DMTRACE=1" make memory_leak_solve.tst

and plot memory use as usual. We can see that realloc_scalar() increases at each step. A possible fix is to add delete ({_res, _da}); at the end of the solve() macro.

diff -rN -u old-basilisk/src/solve.h new-basilisk/src/solve.h
--- old-basilisk/src/solve.h	2026-02-20 10:09:11.808248041 +0100
+++ new-basilisk/src/solve.h	2026-02-20 10:09:11.808421631 +0100
@@ -172,6 +172,7 @@
              "  res: %g nrelax: %d\n", LINENO, a.name,
              _s.i, _s.resa, _s.nrelax),
       fflush (ferr);
+  delete ({_res, _da});
   return _s;
 }}

#include "grid/multigrid1D.h"
#include "solve.h"

int main()
{  
  init_grid (128);
  L0 = 32.*pi;
  periodic (right);
  scalar u[];
  foreach()
    u[] = cos(x/16.)*(1. + sin(x/16.));
  
  double dt = 1e-1;
  int i = 0;
  TOLERANCE = 1e-6;
  for (double t = 0; t <= 10; t += dt, i++) {
    if (i % 1 == 0) {
      foreach()
        fprintf (stdout, "%g %g %g\n", t, x, u[]);
      fputs ("\n", stdout);
    }
    scalar b[];
    foreach()
      b[] = u[] - dt*u[]*(u[1] - u[-1])/(2.*Delta);
    mgstats stats = solve (u,
			   u[] + dt*(u[-1] - 2.*u[] + u[1])/sq(Delta)
			   + dt*(u[-2] - 4.*u[-1] + 6.*u[] - 4.*u[1] + u[2])/sq(sq(Delta)),
			   b[]);
    fprintf (stderr, "%g %d\n", t, stats.i);
  }
}