🚀 go-pugleaf

meissner@cygnus.com (Michael Meissner) writes:

>kanze@lts.sel.alcatel.de (James Kanze US/ESC 60/3/141 #40763) writes:
>
>| Also: the C++ committee considered and rejected nested functions.  I
>| believe that at least part of the reason for the rejection was that
>| there exist (or might exist) architectures on which trampolines cannot
>| be implemented, and in this case, function pointers become more
>| complex, which violates the ``you only pay for what you use''
>| principle.
>
>Trampolines are probably the buggiest feature of the GNU C compiler. 

That may be so, but that does not imply that it need be so.

>You have
>to go through incredible hair on many machines, particularly RISC machines with
>Harvard architectures.  Often times, you have to make a system call in order to
>create the trampolines (since there is no user callable instruction sequence to
>correctly flush the caches).  Sometimes, the machine + OS just gives absolutely
>no way to implement trampolines (DEC's Ultrix on MIPS based DECstations comes
>to mind).

There are definitely trampoline-based ways of implementing nested
functions which work even on mips-dec-ultrix.  Perhaps if I ever get
the time I will do so.

-- 
Fergus Henderson             |  "Australia is the richest country in the world,
fjh@cs.mu.oz.au              |   according to a new system of measuring wealth
http://www.cs.mu.oz.au/~fjh  |   announced by the World Bank yesterday."
PGP: finger fjh@128.250.37.3 |  - Melbourne newspaper "The Age", 18 Sept 1995.