<#part sign=pgpmime>
On Wed, 29 Feb 2012 20:36:04 -0800, Mark Wiebe <[email protected]> wrote:
> The answer to my original question appears to be "no", something that would
> be really nice to change. In the meantime, here is my solution which allows
> the VBO to be used with an ElementwiseKernel:
<snip like a duck>
I'm slightly confused about why you weren't able to make a GPUArray from
your mapped device pointer. Calling the GPUArray constructor with the
data and base kwargs should have done what you need. Do you remember how
that failed?
> I also had to add the following line in compyte/dtype.py so that I
> could use the "double2" CUDA type:
>
> register_dtype(np.dtype([('x', np.float64), ('y', np.float64)]),
> "double2")
This is also surprising to me, since a double2 type is created as
gpuarray.vec.double2. Why did this not work for you?
Andreas
_______________________________________________
PyCUDA mailing list
[email protected]
http://lists.tiker.net/listinfo/pycuda
