Re: Detecting endless recursion?

witness@t-online.de (Uli Kusterer) wrote:


> The best suggestion I've been able to get from some mailing lists was
> to just put a limit on the size of my pseudo-stack and when that is
> exceeded raise an error. This is very straightforward and would be
> easy to implement (and I feel stupid for not thinking of this), but I
> have no idea how to determine this limit.


> ... My impression is that other than artifical examples like
> Ackerman's function, real code doesn't nest very deeply so an arbitrary
> limit like 100 deep should do the trick. -John]


Agreed, but making this configurable at interpreter initialisation
time is worthwhile, especially if someone might use the language to
write an interpreter for some other language. The case I hit last year
was admittedly a bit wacky:


GDB seems to be partly written in LISP. I was using it to debug a pretty
complicated library, which does some quite serious internal recursion -
maybe 100 deep - and whose test harness is a LISP interpreter. GDB, when
its LISP interpreter was faced with interpreting the recursion of the
test-harness LISP interpreter, which was calling the recursive code,
understandably barfed on some of the more complex test cases, and asked me
to run it with a command-line parameter that told it to set up a
larger-than-default stack of the GDB LISP interpreter.


---
John Dallman jgd@cix.co.uk

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