Re: [Gcl-devel] Segmentaton fault

[Camm Maguire]

Greetigns, and thanks for your report!

I cannot reproduce this with 2.6.1, cltl1 or ansi, on Linux.  What
version and platform are you using?  Interpreted I get an invocation
history stack overflow, compiled I appear to go on forever.  I'm
guessing this means you are using windows, in which case you may have
stumbled upon a very simple example which may be tickling some long
outstanding windows specific compiler bugs.  If so, I'd suggest
building your gcl with --enable-debug, loading your function
definitions into foo.lisp, (compile-file "foo.lisp" :system-p t)
(compiler::link (list "foo.o") "new_gcl"), gdb new_gcl, (test), and
report the backtrace of the crash in gdb (i.e. 'bt').

Take care,

Eric Merritt <address@hidden> writes:

> Guys,
> 
>  I was fiddling around trying to figure out how well
> gcl handled recursion (not well by functional
> standards, but find by lisp standards, it seems). In
> any case I came up with the following bit of code
> invariably causes a segmentation fault when compiled
> and run. However, when interpreted it simply causes a
> condition (as it should).
> 
> (defun test ()
>   (print 'hello)
>   (test1))
> (defun test1 ()
>   (print 'World)
>   (test))
> 
> (test)
> 
>  I suspect this is becuase the default compilation
> options remove all checks, but a segfault still
> surprised me. I was wondaring if one of you more
> experienced people might offer some insight.
> 
> 
>       
>               
> __________________________________
> Do you Yahoo!?
> Yahoo! Photos: High-quality 4x6 digital prints for 25¢
> http://photos.yahoo.com/ph/print_splash
> 
> 
> _______________________________________________
> Gcl-devel mailing list
> address@hidden
> http://mail.gnu.org/mailman/listinfo/gcl-devel
> 
> 

-- 
Camm Maguire                                            address@hidden
==========================================================================
"The earth is but one country, and mankind its citizens."  --  Baha'u'llah