Optimization for miranda

• Optimize cava and edge detection schedules (CPU schedules are multi-core now).
• Add fork-extend pass to make fork dimensions multiple of some number (closes #28 (closed)).
• Make GCM add extra dimensions to per-thread arrays in fork-joins.
• Make GCM allocate extra memory for calls in parallel, make RT backend use different memories per parallel call.
• Make not finding xdot binary crash the compiler.
• Some misc. fixes.
• Add conv-like fork-join test.

hercules_opt/src/fork_transforms.rs

@@ -916,7 +916,9 @@ pub fn chunk_all_forks_unguarded(
     };
 
     for (fork, _) in fork_join_map {
-        chunk_fork_unguarded(editor, *fork, dim_idx, dc_id, order);
+        if editor.is_mutable(*fork) {
+            chunk_fork_unguarded(editor, *fork, dim_idx, dc_id, order);
+        }
     }
 }
 // Splits a dimension of a single fork join into multiple.