P2477 Allow programmer to control and detect coroutine elision by static constexpr bool should_elide() and

[wg21bot]

P2477R0 Allow programmer to control and detect coroutine elision by static constexpr bool should_elide() and (Chuanqi Xu)

[wg21bot]

P2477R1 Allow programmer to control and detect coroutine elision by static constexpr bool must_elide() and (Chuanqi Xu)

[wg21bot]

P2477R2 Allow programmer to control and detect coroutine elision (Chuanqi Xu)

[brycelelbach]

@jfbastien and @erichkeane: Let's have Evolution look at this first. If Evolution believes Library Evolution needs to see this, please send it our way and re-tag it LEWG.

[wg21bot]

P2477R3 Allow programmers to control coroutine elision (Chuanqi Xu)

[erichkeane]

EWG discussed this in the February 8th, 2023 morning session in Issaquah. The following polls were taken.

EWG is interested in a solution to the problem raised by P2477R3 of coroutine elision, in some form.	SF	F	N	A	SA
14	13	2	1	0

Result: Consensus

EWG prefers the approach suggested by Gor, where the programmer has an option to prohibit elision, or to have it be an error when the optimizer fails to elide a coroutine.	SF	F	N	A	SA
17	7	3	2	1

Result: Consensus

[noita-player]

what revision is needed? this is necessary to even consider use of coroutines...

[jensmaurer]

@noita-player, the latest comment by @erichkeane shows the current EWG consensus. It seems the paper does not yet reflect that consensus, so a revision of the paper is needed.

[cor3ntin]

@erichkeane I'm removing lewg related tags, send it back to us when you are done with it :)

[erichkeane]

@cor3ntin : I believe from the votes/my notes that we intended this to go to LEWG once it is revised.