Clarify reasons for separate methods to declare structure functions.

Author: aardvark179

lib/truffle/truffle/cext.rb

@@ -1427,16 +1427,17 @@ def rb_data_typed_object_wrap(ruby_class, data, data_type, mark, free, size)
     object
   end
 
+  # These data function are created in separate methods to ensure they
+  # will not accidentally capture the objects they operate on, which
+  # might prevent garbage collection.
   def data_finalizer(free, data_holder)
-    # In a separate method to avoid capturing the object
     raise unless free.respond_to?(:call)
     proc {
       Truffle.invoke_primitive(:interop_call_c_with_mutex, free, [data_holder.data]) unless data_holder.data.nil?
     }
   end
 
   def data_marker(mark, data_holder)
-    # In a separate method to avoid capturing the object
     raise unless mark.respond_to?(:call)
     proc { |obj|
       create_mark_list
@@ -1446,7 +1447,6 @@ def data_marker(mark, data_holder)
   end
 
   def data_sizer(sizer, data_holder)
-    # In a separate method to avoid capturing the object
     raise unless sizer.respond_to?(:call)
     proc {
       Truffle.invoke_primitive(:interop_call_c_with_mutex, sizer, [data_holder.data])