Initialize Fiber with an explicit stack

Doesn't change the public API (the stack is still taken from the current
scheduler's stack pool), but introduces an undocumented initializer that
takes the stack and stack_bottom pointers.

This will allow a few scenarios, mostly for RFC 2:

- start fibers with a fake stack when we don't need to run the fibers,
  for example during specs that only need fiber objects (and would leak
  memory since we only release stacks after the fiber has run);
- during a cross execution context spawn, we can pick a stack from the
  destination context instead of the current context (so we can recycle
  stacks from fibers that terminated in the desination context).

Author: ysbaddaden

src/fiber.cr

@@ -97,14 +97,19 @@ class Fiber
   # When the fiber is executed, it runs *proc* in its context.
   #
   # *name* is an optional and used only as an internal reference.
-  def initialize(@name : String? = nil, &@proc : ->)
-    @context = Context.new
-    @stack, @stack_bottom =
+  def self.new(name : String? = nil, &proc : ->)
+    stack, stack_bottom =
       {% if flag?(:interpreted) %}
         {Pointer(Void).null, Pointer(Void).null}
       {% else %}
         Crystal::Scheduler.stack_pool.checkout
       {% end %}
+    new(name, stack, stack_bottom, &proc)
+  end
+
+  # :nodoc:
+  def initialize(@name : String?, @stack : Void*, @stack_bottom : Void*, &@proc : ->)
+    @context = Context.new
 
     fiber_main = ->(f : Fiber) { f.run }