Arch Dispatching¶
xsimd provides a generic way to dispatch a function call based on the architecture the code was compiled for and the architectures available at runtime.
The :cpp:func::xsimd::dispatch function takes a functor whose call operator takes an architecture parameter as first operand, followed by any number of arguments Args...
and turn it into a
dispatching functor that takes Args...
as arguments.
-
template<class F, class ArchList = supported_architectures>
inline detail::dispatcher<F, ArchList> xsimd::dispatch(F &&f)¶
Following code showcases a usage of the :cpp:func::xsimd::dispatch function:
// functor with a call method that depends on `Arch`
struct sum {
template<class Arch, class T>
T operator()(Arch, T const* data, unsigned size)
{
using batch = xsimd::batch<T, Arch>;
batch acc(static_cast<T>(0));
const unsigned n = size / batch::size * batch::size;
for(unsigned i = 0; i != n; i += batch::size)
acc += batch::load_unaligned(data + i);
T star_acc = xsimd::hadd(acc);
for(unsigned i = n; i < size; ++i)
star_acc += data[i];
return star_acc;
}
};
// Create the dispatching function.
auto dispatched = xsimd::dispatch(sum{});
// Call the appropriate implementation based on runtime information.
float res = dispatched(data, 17);