RC6, Linux 2.4.10, SuSE 7.3, 128MB, P3 600

Rendering of the following scene causes a segmentation fault:

// BEGIN
#include "param.inc"

#declare Fx = function(U,V){sin(U)*cos(V)}
#declare Fy = function(U,V){cos(U)}
#declare Fz = function(U,V){sin(U)*sin(V)}

#declare Umin= 0;
#declare Umax= 2*pi;
#declare Vmin= 0;
#declare Vmax= 2*pi;

#declare Iter_U= 5;
#declare Iter_V= 5;

Parametric()
// END

Output:

// BEGIN
Parsing....
 Calculating 36 vertices for 50 triangles


Segmentation fault
// END

I'm sorry I haven't been able to reproduce this error without Ingo's 
param.inc, which can be found at
http://news.povray.org/povray.binaries.scene-files/18291/
-- 
Felix Wiemann

Post a reply to this message

Felix Wiemann wrote:
> RC6, Linux 2.4.10, SuSE 7.3, 128MB, P3 600
> 
> Rendering of the following scene causes a segmentation fault:
> 
> // BEGIN
> #include "param.inc"
> 
> #declare Fx = function(U,V){sin(U)*cos(V)}
> 
> Segmentation fault
> // END


I had a similar problem just now.
Before, for the builtin functions, the functions.inc did
not have to be included. This seems to be a MUST now.
To segfault in this case is a bug of course. But now,
that I found a workarround I am happy.

Patrick

Post a reply to this message

in news:3d148dcb@news.povray.org Felix Wiemann wrote:

in news:3d148dcb@news.povray.org Felix Wiemann wrote:

> RC6, Linux 2.4.10, SuSE 7.3, 128MB, P3 600
> 
> Rendering of the following scene causes a segmentation fault:
> 
> // BEGIN
> #include "param.inc"
> [...]
> #declare Umin= 0;

try:

#declare Umin= FromU(0);

else it will do a vnormalize on a <0,0,0> vector.
Should give a parse error though (as it does under windows), not a 
segmentation fault.

Ingo

p.s. there is a more recent version at:
    	http://members.home.nl/seedseven/
the syntax looks more like POV-Ray's now.

Post a reply to this message

ingo wrote:

> #declare Umin= FromU(0);
> 
> else it will do a vnormalize on a <0,0,0> vector.

#local A = vnormalize(<0, 0, 0>);

gives the expected output:

Parse Error: Cannot normalize zero-length vector.
-- 
Felix Wiemann

Post a reply to this message

Patrick Fromberg wrote:

> Before, for the builtin functions, the functions.inc did
> not have to be included. This seems to be a MUST now.

In fact,
#include "functions.inc"
solves the problem. Very strange...
-- 
Felix Wiemann

Post a reply to this message

Felix Wiemann wrote:

> In fact,
> #include "functions.inc"
> solves the problem. Very strange...
  ^^^^^^^^^^^^^^^^^^
I.e. gives a parse error.
-- 
Felix Wiemann

Post a reply to this message