create_mirror_view_and_copy

[ibaned]

Lets say that users are porting a code using unmanaged views:

auto h_data = Kokkos::View<T*, Kokkos::HostSpace, Kokkos::MemoryUnmanaged>(my_ptr, 1000);
auto d_data = Kokkos::create_mirror_view(Kokkos::DefaultExecutionSpace(), h_data);
Kokkos::deep_copy(d_data, h_data);

This will do three things when compiled with CUDA:

1. Allocate d_data. This is fine.
2. Initialize d_data with zeros! This is what I want to avoid. The data is about to be overwritten.
3. Deep copy from host to device. This is fine.

By combining these three actions under one API, we can optimize out initialization in the case when data is about to be overwritten. More generally, we should offer a create_uninitialized_mirror_view.

Second case: The above code doesn't even compile when T=const double. The reason being that deep_copy doesn't accept a const left hand side. This is the fundamental complaint in #728 I think. By combining these three lines into one API, we could specialize it for the const case such that in CUDA it creates a mirror, does a deep copy, and then returns a const View, while in other cases it simply returns the input View.

So, there are at least two reasons why the proposed single API would improve performance. Also, I think this is basically the first and most common thing that users do when they start porting.

[dholladay00]

I too have to do this. I receive inputs from the host code and they have to be transferred to the default execution space's memory space. It is nice that this can be done in a way that avoids copies when possible, but making them const is a little clunky. Something like this would be very nice.

// const input data
View<input*, MemoryUnmanaged, HostSpace> h_a(ptr, N);
View<input*> d_a_nc = create_mirror_view(DefaultExecutionSpace(), h_a);
deep_copy(d_a_nc, h_a);
View<const input*, RandomAccess> d_a(d_a_nc);
// output data
View<output*, MemoryUnmanaged, HostSpace> h_b(other_ptr, N);
View<output*> d_b = create_mirror_view(DefaultExecutionSpace(), h_b);
deep_copy(d_b, h_b);

[ibaned]

Yea, thats basically what I meant. There should be one API that chooses the right option from your example automatically.

[ibaned]

Just for book-keeping, the VPIC team would benefit from this.

[stanmoore1]

Would be nice to have something similar for Kokkos::DualView as well.

[ibaned]

First draft implementation in PR #1164