Re: Garbage collection for system programming

"Simon" <sheath1@gmail.com> writes:




> Anyway, what I was wondering is this: How feasable is it to use a
> garbage-collected language for low-level tasks like writing operating
> systems, device drivers, memory allocation systems, and basically
> anything else you would HAVE to write in C instead of, say, Objective
> Caml.


The Fox Project at Carnegie Mellon (http://www-2.cs.cmu.edu/~fox/) has
explored using functional languages for systems software, trying to
get the imporved safety of typed functional languages without losing
sight of efficiency issues.


As John said, the real problem is not so much space overhead for
tagging etc., but real-time constraints. State-of-the-art GC
techniques are good enough for soft realtime without introducing an
unacceptable overhead. For hard real-time, you can use region
inference
(http://www.cs.cornell.edu/Nuprl/PRLSeminar/PRLSeminar99_00/Walker/nov8.html),
which replaces GC with an automatic stack-like allocation/deallocation
mechanism.


                Torben

Return to the comp.compilers page.
Search the comp.compilers archives again.