What is the operations or treatments we can do on xvalues and not on prvalues and vice versa

Question

I'm wondering why rvalue expressions are subdivided in two groups (xvalues and prvalues).

Saying differently is there a lot of place where the treatment on rvalues is different according to the fact the expression is either an xvalue or a prvalue? (I think for instance at copy-elision are feature like that). Too much to be enumerated here maybe?

Answer 1

I am somewhat unclear on what is asked, but some of the differences between xvalues and prvalues are:

• non-class xvalue can be cv-qualified, prvalue can't be

• class xvalue can be polymorphic, class prvalue is always its most derived type