The following line of code executes successfully with PyCUDA 0.94.2
but fails when I try to run it with PyCUDA 2011.1.2.
x_gpu = gpuarray.zeros(3, np.complex128)
The error message is
/usr/lib64/python2.6/site-packages/pycuda/gpuarray.pyc in zeros(shape,
dtype, allocator, order)
748 result = GPUArray(shape, dtype, allocator, order=order)
749 zero = np.zeros((), dtype)
--> 750 result.fill(zero)
751 return result
752
/usr/lib64/python2.6/site-packages/pycuda/gpuarray.pyc in fill(self,
value, stream)
496 func = elementwise.get_fill_kernel(self.dtype)
497 func.prepared_async_call(self._grid, self._block,
stream,
--> 498 value, self.gpudata, self.mem_size)
499
500 return self
/usr/lib64/python2.6/site-packages/pycuda/driver.pyc in
function_prepared_async_call(func, grid, block, stream, *args,
**kwargs)
290
291 from pycuda._pvt_struct import pack
--> 292 func._param_setv(0, pack(func.arg_format, *args))
293
294 for texref in func.texrefs:
error: required argument is not a complex
L.G.
_______________________________________________
PyCUDA mailing list
[email protected]
http://lists.tiker.net/listinfo/pycuda
