Bogdan, David, On Tue, 14 Dec 2010 08:55:51 +0100, David Libault <[email protected]> wrote: > After further checking, it does not look like a design decision, but a > strict implementation of the standard ! See opencl1.1 page 163 : > > "6.1.5 Alignment of Types > > For 3-component vector data types, the size of the data type is 4 * > sizeof(component). This means that a 3-component vector data type will > be aligned to a 4 * sizeof(component) boundary. The vload3 and vstore3 > built-in functions can be used to read and write, respectively, > 3-component vector data types from an array of packed scalar data > type." > > So Apple's implementation looks correct... > > Andreas, > > What implementation of opencl are you using on your 64-bit Linux, as > it might have an alignment issue ? Could you please try the code > proposed by Bogdan to check float3 size ?
I think it might've been pure luck that things worked without crashing for me. I can say right now that AMD CPU CL 2.2 and 2.3 (both implement CL 1.1) conforms to the above, i.e. sizeof(float3) == 16. The CUDA 3.2 CL implementation doesn't implement CL 1.1, so I can't check right now, but I suspect their CL 1.1 conformance candidate is fine, too. In other words, nothing to see. Andreas
pgpgc3465Vyo5.pgp
Description: PGP signature
_______________________________________________ PyOpenCL mailing list [email protected] http://lists.tiker.net/listinfo/pyopencl
