Re: beta version for test

[Lindsey Spratt]

Daniel,
The stack overflow test works as specified using my slightly modified  
version of 1.2.18 on Mac OS X 10.3.7. (The modifications are part of  
the support of XGP and not related, I think, to the stack overflow  
handling.)
What's the status of 1.2.18--is it going to be dubbed "stable" soon?

Lindsey

On Mar 18, 2004, at 7:50 AM, Daniel Diaz wrote:

> Hi,
>
> I'd like to release 1.2.18 as new stable version. Before this I'd like  
> to have feedbacks from people using other plateforms than ix86/linux,  
> ix86/cygwin/XP, ix86/msvc/XP. This new version also includes a port to  
> ix86/mingw (tested under XP). For instance tests on MacOS, 64-bit  
> arch, solaris, *BSD are welcome. So if you have some test programs  
> (e.g. the ones you used with earlier version of gprolog) and some free  
> time I'd be happy to receive your feedback (on this mailing list or  
> directly to my email). You can get a source distrib or a precompiled  
> rpm (compiled on ix86/redhat 9.0):
> ftp://ftp.inria.fr/INRIA/Projects/contraintes/gprolog/unstable/ 
> gprolog-20040317.tgz
> ftp://ftp.inria.fr/INRIA/Projects/contraintes/gprolog/unstable/ 
> gprolog-20040317.i386.rpm
> BTW: I'd like to summarise the architectures where the stack overflow  
> mechanism does not work. For this I attach a test program called  
> stack_ovfl.pl. So if you have an architecture where the test program  
> does not work, please inform me precising the OS (e.g. 'uname -a') and  
> the version of gcc used (gcc -v).
> Thank you for your help.
>
> PS: to test stack overflow detection:
>
> compile stack_ovfl.pl with:
>    gplc stack_ovfl.pl --min-size
>
> set variables (e.g. bash):
>
>    export GLOBALSZ=100
>    export LOCALSZ=2000
>    export TRAILSZ=52
>
> test with:
>
>    stack_ovfl local
>    stack_ovfl global
>    stack_ovfl trail
>
> it should provide a msg:
>    Fatal Error: XXX stack overflow (size: YY Kb, environment variable
> used: XXXSZ)
>
> where XXX is the name of the stack, YY the current size.
>
>
> --
> ===============================================
>                  Daniel Diaz
> University of Paris 1      INRIA Rocquencourt
> 75013 Paris FRANCE      78153 Le Chesnay FRANCE
>         web: http://pauillac.inria.fr/~diaz
>         email: address@hidden
>
> --
> Ce message a subi une analyse antivirus par MailScanner ; il est  
> vraisemblablement
> sans danger.
>
> /* to test stack overflow detection:
> compile with:
>    gplc stack_ovfl.pl --min-size
> set variables:
>    export GLOBALSZ=100
>    export LOCALSZ=2000
>    export TRAILSZ=52
>
> test with:
>    stack_ovfl local|global|trail
> should provide a msg:
>    Fatal Error: XXX stack overflow (size: YY Kb, environment variable  
> used: XXXSZ)
> */
>
> :-initialization(go).
>
> go :-
>         argument_value(1, T),
>         test(T), !,
>         write('finish... strange decrease correspondig stack size'), nl.
>
> go :-
>         write('invalid argument (should be one of: local/global/trail)'), nl.
>
>         
>
> test(local):-
>         t_local.
>
> test(global) :-
>         t_global(a).
>
> test(trail) :-
>         t_trail(_).
>
>
>
> t_local:-
>         t_local,
>         dummy.
>
> dummy.
>
> t_global(X):-
>         t_global(f(X,X)).
>
>
>
> t_trail(a):-
>         t_trail(_).
>
> t_trail(_).
> _______________________________________________
> Users-prolog mailing list
> address@hidden
> http://mail.gnu.org/mailman/listinfo/users-prolog