Add `right_shift_arithmetic` op

[kevinstephano]

We are missing the operation right_shift_arithmetic that differs from bitwise_right_shift because it preserves the sign bit.

https://en.wikipedia.org/wiki/Arithmetic_shift

[rdspring1]

Apparently, the right shift operator in CPP implements right shift arithmetic. I'm assuming CUDA has the same behavior.

For negative a, the value of a >> b is implementation-defined (in most implementations, this performs arithmetic right shift, so that the result remains negative).
https://en.cppreference.com/w/cpp/language/operator_arithmetic