+2002-08-25 Han-Wen Nienhuys <hanwen@cs.uu.nl>
+
+ * gc-segment.c (scm_i_get_new_heap_segment): use float in stead of
+ unsigned numbers for computing minimum heap increment. This
+ prevents weird results when a a negative minimum increment is
+ computed.
+
2002-08-24 Marius Vollmer <mvo@zagadka.ping.de>
* gc_os_dep.c: When we have __libc_stack_end, use that directly
abort ();
}
-
- /* Pick a size for the new heap segment.
- * The rule for picking the size of a segment is explained in
- * gc.h
- */
{
/* Assure that the new segment is predicted to be large enough.
*
*
* This gives dh > (f * h - y) / (1 - f)
*/
-
- /*
- where is is this explanation supposed to be? --hwn
- */
- int f = freelist->min_yield_fraction;
- unsigned long h = SCM_HEAP_SIZE;
- size_t min_cells = (f * h - 100 * (long) scm_gc_cells_collected) / (99 - f);
+ float f = freelist->min_yield_fraction / 100.0;
+ float h = SCM_HEAP_SIZE;
+ float min_cells
+ = (f * h - scm_gc_cells_collected) / (1.0 - f);
/* Make heap grow with factor 1.5 */
len = freelist->heap_size / 2;
fprintf (stderr, "(%ld < %ld)", (long) len, (long) min_cells);
#endif
- /*
- Original code adds freelist->cluster_size here.
- */
if (len < min_cells)
- len = min_cells;
+ len = (unsigned long) min_cells;
len *= sizeof (scm_t_cell);
/* force new sampling */
freelist->collected = LONG_MAX;