Re: [Gcl-devel] bug in debug.{c,lsp}

[Camm Maguire]

Greetings!  And thanks for the report!  I've fixed this.  Actually I
introduced the bug earlier when cleaning up lisp compiler warnings.
These were reported as unused variables.  I erroneously assumed from
my C experience that the compiler would have checked as well that all
calls to the function pass acceptable argument lists, but this is
(apparently) not the case.  This would be nice to add at some point.  

Beyond this, there is still something in the ansi code that is
degrading the nice emacs lisp source debugging that Dr. Schelter
had set up.  :resume and :continue don't seem to work at breakpoints,
and the debug stack level appears incorrectly set.  Any insights here
most appreciated.

Take care,

Peter Wood <address@hidden> writes:

> Hi
> 
> In a recently downloaded CVS GCL (about 45 minutes ago), in
> lsp/debug.lsp there's a bug in the lisp definition of
> #'print-stack-frame, where the last paramter 'env' has been left out.
> This results in not being able to print a backtrace, since although
> 'env' doesn't appear to be used in the defun, #'print-stack-frame is
> called from a number of places with more than six (ie 7) arguments.
> 
> The corresponding bug in debug.c is c function 'L50':
> 
>       if(vs_top-vs_base>6) too_many_arguments();
> 
> The '6' should be 7.
> 
> I would guess 'env' is included as a paramter because it might be
> useful in future improvements to the debugger.  
> Perhaps '(declare (ignore env))' ?
> 
> Regards,
> Peter
> 
> 
> 
> 
> _______________________________________________
> 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