[PPL-devel] drop_some_non_ineger_points raises " Invalid memory access of location 0x0 rip=0x1011a50ca "

[Enea Zaffanella]

On 08/30/2012 04:46 AM, Zell wrote:
> Hello,
>
> I may probably wrongly use the PPL's Java interface, but here are my
> naive questions among others,
>
> 1. Is it normal that the following codes raise this error " Invalid
> memory access of location 0x0 rip=0x1011a50ca " ?
> (The JAVA's interface for 'drop_some_non_integer_points" requies an
> integer or variable-set plus  an integers as its parameters, rather than
> C's
>                             " void drop_some_non_integer_points
> (Complexity_Class complexity=ANY_COMPLEXITY)"  )
>   .
> .....cod
>          Constraint c = new Constraint(le_5Z,
> Relation_Symbol.GREATER_THAN,           le_X);
>          Constraint_System cs = new Constraint_System();
>          cs .add(c);
>          NNC_Polyhedron ph2=new NNC_Polyhedron(cs);
>          ph2.drop_some_non_integer_points(new Integer(1));
> ......


We have a Java interface issue.

We modified the Java interface to accept values from Complexity_Class 
instead of plain integers, but it seems that the Java method declaration 
for drop_some_... was not updated accordingly.
The (native) implementation of the method has been updated though, hence 
the exception you observe, since there is a type mismatch.

Another minor problem is that we are needlessly adding an "_2" subscript 
to the method name taking 2 parameters (rather than exploiting overloading).

So, thanks for reporting, we will correct this issue.


> 2. I don't find ways to specify a constraint with explicite rational or
> real numbers, like x>=2.5, because the 'Coefficient' only accepts
> integers. Does than mean that I am obliged to
> use 2 *x >=5 rather than x >2.5?

Yes, you should specify integer coefficients (2*x >= 5).

Enea.

>
> Thanks.
> zell.
>
>
> _______________________________________________
> PPL-devel mailing list
> PPL-devel at cs.unipr.it
> http://www.cs.unipr.it/mailman/listinfo/ppl-devel
>