update relocation definitions to track p2786r13 standard

[guptapratykshh]

Fixes #6625

Proposed Changes

• added is_replaceable trait to align with p2786r13
• updated is_trivially_relocatable to use the standard version if the c++26 feature macro is present
• wired up relocate_at to call std::relocate_at when available

Any background context you want to provide?

Checklist

Not all points below apply to all pull requests.

• I have added a new feature and have added tests to go along with it.
• I have fixed a bug and have added a regression test.
• I have added a test using random numbers; I have made sure it uses a seed, and that random numbers generated are valid inputs for the tests.

[StellarBot]

Can one of the admins verify this patch?

[hkaiser]

@isidorostsa Would you please have a look?

[isidorostsa]

Hey @guptapratykshh thank you for taking a look at this!

Unfortunately, P2786 was pulled last minute from C++26 in the Kona meeting, so the issue you tackled is outdated, I apologize for that.

That being said, it wouldn't hurt to implement the replaceable trait.

Great code!

[isidorostsa]

this should also be added to the replacabillity definition

[isidorostsa]

I don't think std::relocate_at is being proposed, but I'm not sure.

If it is, it would be better to avoid defining all the helpers and replace everything with a using relocate_at.

[isidorostsa]

A type is replaceable if what you get from move constructing is the same you get from move assigning.
In code:

T move_assigned = some_t();

T move_constructed(std::move(existing));    // 1
move_assigned = std::move(existing);        // 2

You want move_constructed to be the same as move_assigned, and the conditions you've put forth are not enough to understand if this is true.

Replaceable is an opt-in trait, you might want to take a look at how we deal with such traits in the is_trivially_relocatable definition.

Take a look at this paper to learn more about how to detect replaceability.

[isidorostsa]

you should replace this with your own copyright here

[isidorostsa]

Those tests should be updated in accordance with the changes in the implementation of the replacabillity trait.

[isidorostsa]

Why is this checking for trivial relocatabillity?
How would a user opt in?

I suggest reviewing P2786 to get a tighter grip on the specifics of this type trait.

Thanks!

[guptapratykshh]

thanks for the feedback. i have updated is_replaceable to strictly follow p2786r13, remove the trivial relocatability check, and allow users to opt in via specialization.