Re: Q: access to subprogram

[dewar@cs.nyu.edu (Robert Dewar)]

Thomas Wolff said

Why should displays be a menace where static links are not? Who would
actually prefer to really follow static links at all, effectively almost
*searching* for variables at run-time? Until now I thought this would
only be a theoretical concept and perhaps one of very early compilers.

  Static links are by far the more common way of hanling uplevel
  references, and are far from being "only a theoretical concept".
  Most modern compilers use static links. There is no searching
  involved, just loading the entries that are needed, and if they
  are used frequently, they will stay in registers anyway.

  In fact the idea of static links came a little later on, the early
  Algol-60 compilers used displays.

Why not just push the display on the stack and pass its address??
(If you really wanted to pass the display itself for some reason, you
could still use the program's maximal display size, wasting a few
bytes in these occasional cases.)

  That's the original Algol-60 approach but it is never used these
  days that I know of, since it is in practice definitely inferior
  to both static links and the use of global displays. It introduces
  unconditionally a large overhead of copying the display at every
  call whether or not it is useful to do so.

  (Yes, I know the Intel x86 hardware uses that approach. I use this
   as one of the examples of very bad CISC design, not only is an
   over-complex instruction built-in, but it is a bad algorithm
   that is built-in!)