osx - Aparapi add sample -

i'm studing aparapi (https://code.google.com/p/aparapi/) , have strange behaviour of 1 of sample included. sample first, "add". building , executing it, ok. put following code testing if gpu used

if(!kernel.getexecutionmode().equals(kernel.execution_mode.gpu)){     system.out.println("kernel did not execute on gpu!"); } 

and works fine. but, if try change size of array 512 number greater 999 (for example 1000), have following output:

!!!!!!! clenqueuendrangekernel() failed invalid work group size after clenqueuendrangekernel, globalsize[0] = 1000, localsize[0] = 128 apr 18, 2013 1:31:01 pm com.amd.aparapi.kernelrunner executeopencl warning: ### cl exec seems have failed. trying revert java ### jtp  kernel did not execute on gpu! 

here's code:

  final int size = 1000;    final float[] = new float[size];   final float[] b = new float[size];    (int = 0; < size; i++) {      a[i] = (float)(math.random()*100);      b[i] = (float)(math.random()*100);   }    final float[] sum = new float[size];    kernel kernel = new kernel(){      @override public void run() {         int gid = getglobalid();         sum[gid] = a[gid] + b[gid];      }   };    range range = range.create(size);   kernel.execute(range);    system.out.println(kernel.getexecutionmode());   if (!kernel.getexecutionmode().equals(kernel.execution_mode.gpu)){      system.out.println("kernel did not execute on gpu!");   }    kernel.dispose(); 

}

i tried specifying size using

range range = range.create(size, 128); 

as suggested in google group, nothing changed.

i'm running on mac os x 10.8 java 1.6.0_43. aparapi version latest (2012-01-23).

am missing something? ideas?

thanks in advance

aparapi inherits 'grid style' of implementation opencl. when specify range of execution (say 1024), opencl break 'range' groups of equal size. possibly 4 groups of 256, or 8 groups of 128.

the group size must factor of range (so assert(range%groupsize==0)).

by default aparapi internally selects group size.

but choosing specify range , group size using

range r= range.range(n,128)

you responsible ensuring n%128==0.

from error, looks chose range.range(1000,128).

sadly 1000 % 128 != 0 range fail.

if specifiy

range r = range.range(n)

aparapi choose valid group size, finding highest common factor of n.

try dropping 128 the second arg.

gary