RFC: Keyword unreservations (pure, sizeof, alignof, offsetof) #2421

Centril · 2018-04-26T18:22:42Z

🖼️ Rendered

⏭ Tracking issue

📝 Summary

We unreserve:

pure
~~unsized~~
sizeof
alignof
offsetof
~~priv~~

🎵 Note

If you have a particular objection and use case for some keyword this RFC wants to unreserve, please speak up and leave a comment.

Centril · 2018-04-26T18:24:03Z

text/0000-unreservations-2018.md

+*"go look at `std::mem::size_of` instead"*. However, we believe it is better
+to allow users the freedom to use these keywords instead.
+
+## Rationale for `priv`


cc @withoutboats - You wanted to keep this?

I wanted to keep this!

@petrochenkov Could you elaborate a lil bit on possible use cases you foresee?

Considering how much activity we have around module proposals, including visibility, I think holding onto this one seems reasonable for now.

Aight. As we discussed on the lang team meeting; I'll scratch this one of the list in this round of unreservations. We can always revisit later.

Centril · 2018-04-26T18:24:35Z

text/0000-unreservations-2018.md

+In both 1. and 2., `pure` can be contextual.
+We also don't think that the drawbacks are significant for `pure`.
+
+## Rationale for `unsize`


cc @eddyb - Double checking: This is fine?

cc @nikomatsakis @mikeyhew

unsized may be useful for custom DST, so I think it's worth holding onto for now

@mikeyhew Can you elaborate a bit? (possibly with a code example?)

I'd use dyn instead of unsize for DST.

@mikeyhew No specific use case; your comment helps me refine the RFC tho :) But @kennytm's idea seems nice? If this discussion doesn't develop I'll strike unsized in a bit with your motivation.

I'd use dyn for dynamic existentials, not dynamic sizes, e.g. you could imagine [T] expanding to dyn<n: usize> [T; n] and dyn Trait to dyn<T: Trait> T.

@eddyb But what about let vla = [T; dyn size];?

I've now scratched unsized given @mikeyhew's motivation and that unreserving it violates:

[..] we are sure that we have no intention of using the keyword in the future [..]

@Centril That's not a size, that's a length. Dynamic existential shorthands (when each bound type/value name is used only once, in an unambiguous way) could be dyn Trait and [T; dyn usize] (or just dyn, since the type can be inferred or specified elsewhere).

joshtriplett · 2018-04-26T18:34:19Z

I don't want to keep offsetof as a keyword, but I do want us to have a macro offsetof! (or offset_of!) someday. If anyone is interested in an RFC for that, I'd be happy to work with them.

eddyb · 2018-04-26T19:38:01Z

@joshtriplett I dislike having macros that expand to builtins that cannot be written with other syntax, so I'm curious which approach you're looking for. I guess using an union and taking addresses into it via let $StructName { ref $field_name, .. } = uninit.data; could work without being builtin?

EDIT: I've marked the remaining offsetof discussion as offtopic as per #2421 (comment) - there's a link to an internals post in #2421 (comment) if you wish to continue the discussion there.

hanna-kruppe · 2018-04-26T20:16:55Z

text/0000-unreservations-2018.md

+## Rationale for `sizeof`, `alignof`, and `offsetof`
+
+We already have [`std::mem::size_of`](https://doc.rust-lang.org/nightly/std/mem/fn.size_of.html) and similar which
+are `const fn`s or can be.


offsetof can't be a const fn, at least not without some sort of static reflection being added to the language – which is a huge feature and AFAIK nothing specific in this direction has ever been proposed. I agree it could be unreserved (and implemented as a macro, as @eddyb showed), but this rationale here only applies to sizeof and alignof, not offsetof.

Note that I would prefer if the macro generated a const fn, so you could use it anywhere a constant value was expected. Most of it will soon become possible, with the exception of the assert, although that's mostly a sanity check and shouldn't be required for soundness.

@rkruppe In theory we could know the delta between fields of the same structure (or from the start of the structure to one field) statically at compile time, without needing full static reflection. And many uses of offsetof in C want it in a compile-time-constant context..

@joshtriplett What you describe sounds more like a special keyword than a const fn! I don't see a nice, generic (i.e., which enables more than just const fn offsetof) extension to current Rust that allows one to write const fn offsetof in a library. Well, nothing that's smaller/simpler than the beginnings of a reflection system. Once you start introducing things like "reifying a field into a value and passing it to a function" (which seems necessary to even define the interface of const fn offsetof, let alone its implementation) you're already in reflection land.

(I agree that offset calculations need to work in constant contexts, but as @eddyb said, the macro is perfectly compatible with that!)

@rkruppe This is true; I'll clarify.

Centril · 2018-04-26T20:35:09Z

Update: I've removed priv from the list of unreservations and moved it to the "future work" section.

text/0000-unreservations-2018.md

Centril · 2018-05-02T19:40:39Z

I made a topic for it: https://internals.rust-lang.org/t/discussion-on-offset-of/7440

Just to double check: Everyone is fine with unreserving offsetof?
Please mark this comment with 👍 if yes, or 👎 if not, and then explain why (if 👎).

joshtriplett · 2018-05-03T22:15:46Z

Based on discussion in the lang team meeting, we're ready to unreserve the remaining keywords in this RFC.

@rfcbot fcp merge

rfcbot · 2018-05-03T22:15:46Z

Team member @joshtriplett has proposed to merge this. The next step is review by the rest of the tagged teams:

No concerns currently listed.

Once a majority of reviewers approve (and none object), this will enter its final comment period. If you spot a major issue that hasn't been raised at any point in this process, please speak up!

See this document for info about what commands tagged team members can give me.

rfcbot · 2018-05-17T16:32:37Z

🔔 This is now entering its final comment period, as per the review above. 🔔

text/0000-unreservations-2018.md

rfcbot · 2018-05-27T16:36:41Z

The final comment period, with a disposition to merge, as per the review above, is now complete.

Centril · 2018-05-27T22:33:00Z

Huzzah! This RFC is merged!

Tracking issue: rust-lang/rust#51115

- According to [rust-lang/rfcs#2421](rust-lang/rfcs#2421)

(See rust-lang/rfcs#2421)

rfc, unreservations: initial version.

609031e

Centril added T-lang Relevant to the language team, which will review and decide on the RFC. I-nominated labels Apr 26, 2018

Centril commented Apr 26, 2018

View reviewed changes

Centril mentioned this pull request Apr 26, 2018

RFC: Unreserve proc #2420

Merged

This comment has been minimized.

Sign in to view

Centril removed the I-nominated label Apr 26, 2018

This comment has been minimized.

Sign in to view

hanna-kruppe reviewed Apr 26, 2018

View reviewed changes

This comment has been minimized.

Sign in to view

rfc, unreservations: scratch priv from unreservations.

47408ea

Centril added 2 commits April 26, 2018 22:39

rfc, unreservations: clarify offsetof wrt. const fn & macro.

2bddc3c

rfc, unreservations: fix typo.

c44febd

This comment has been minimized.

Sign in to view

scottmcm reviewed Apr 27, 2018

View reviewed changes

text/0000-unreservations-2018.md Outdated Show resolved Hide resolved

scottmcm reviewed Apr 27, 2018

View reviewed changes

text/0000-unreservations-2018.md Outdated Show resolved Hide resolved

Centril added 2 commits April 27, 2018 22:31

rfc, unreservations: fix unsize -> unsized typos.

8c4f63f

rfc, unreservations: scratch unsized from unreservations.

d797c44

Centril changed the title ~~RFC: Keyword unreservations (pure, unsize, sizeof, alignof, offsetof, priv)~~ RFC: Keyword unreservations (pure, sizeof, alignof, offsetof) Apr 30, 2018

This comment has been minimized.

Sign in to view

bjorn3 reviewed May 2, 2018

View reviewed changes

text/0000-unreservations-2018.md Outdated Show resolved Hide resolved

This comment has been minimized.

Sign in to view

rfcbot added proposed-final-comment-period Currently awaiting signoff of all team members in order to enter the final comment period. disposition-merge This RFC is in PFCP or FCP with a disposition to merge it. labels May 3, 2018

RalfJung mentioned this pull request May 8, 2018

const fn tracking issue (RFC 911) rust-lang/rust#24111

Closed

3 tasks

rfcbot added final-comment-period Will be merged/postponed/closed in ~10 calendar days unless new substational objections are raised. and removed proposed-final-comment-period Currently awaiting signoff of all team members in order to enter the final comment period. labels May 17, 2018

brennie suggested changes May 23, 2018

View reviewed changes

text/0000-unreservations-2018.md Outdated Show resolved Hide resolved

Centril added 2 commits May 23, 2018 12:33

rfc, unreservations: s/hold of/hold off

ec70270

rfc, unreservations: fix introduced typo.

42d255a

brennie approved these changes May 24, 2018

View reviewed changes

rfcbot added finished-final-comment-period The final comment period is finished for this RFC. and removed final-comment-period Will be merged/postponed/closed in ~10 calendar days unless new substational objections are raised. labels May 27, 2018

Centril mentioned this pull request May 27, 2018

Tracking issue for RFC 2421, "Keyword unreservations (pure, sizeof, alignof, offsetof)" rust-lang/rust#51115

Closed

2 tasks

RFC 2421

17bc1e7

Centril merged commit 7a2ce51 into rust-lang:master May 27, 2018

Centril deleted the rfc/unreservations-2018 branch May 27, 2018 22:33

Centril added A-syntax Syntax related proposals & ideas A-keyword Proposals relating to keywords. labels Nov 23, 2018

gnzlbg mentioned this pull request Mar 15, 2019

Interaction of isize / usize with value layout rust-lang/unsafe-code-guidelines#102

Closed

nwtnni added a commit to nwtnni/rust.vim that referenced this pull request Jun 5, 2019

Remove unreserved keywords from highlighting group

9ddd7df

- According to [rust-lang/rfcs#2421](rust-lang/rfcs#2421)

nwtnni added a commit to nwtnni/rust.vim that referenced this pull request Jun 5, 2019

Remove unreserved keywords from highlighting group

efd93fd

(See rust-lang/rfcs#2421)

nwtnni mentioned this pull request Jun 5, 2019

Remove unreserved keywords from highlighting group rust-lang/rust.vim#329

Merged

da-x pushed a commit to rust-lang/rust.vim that referenced this pull request Jun 6, 2019

Remove unreserved keywords from highlighting group

6d6c410

(See rust-lang/rfcs#2421)

scottmcm mentioned this pull request Apr 1, 2021

Raw Keywords #3098

Open

scottmcm mentioned this pull request Apr 12, 2024

Add size_of and size_of_val and align_of and align_of_val to the prelude rust-lang/rust#123168

Merged

RFC: Keyword unreservations (pure, sizeof, alignof, offsetof) #2421

RFC: Keyword unreservations (pure, sizeof, alignof, offsetof) #2421

Conversation

Centril commented Apr 26, 2018 • edited Loading

📝 Summary

🎵 Note

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Centril Apr 26, 2018 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Centril Apr 26, 2018 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

This comment has been minimized.

joshtriplett commented Apr 26, 2018

eddyb commented Apr 26, 2018 • edited Loading

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

hanna-kruppe Apr 26, 2018 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

Centril commented Apr 26, 2018

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

Centril commented May 2, 2018

joshtriplett commented May 3, 2018

rfcbot commented May 3, 2018 • edited by aturon Loading

rfcbot commented May 17, 2018

rfcbot commented May 27, 2018

Centril commented May 27, 2018

Centril commented Apr 26, 2018 •

edited

Loading

Centril Apr 26, 2018 •

edited

Loading

Centril Apr 26, 2018 •

edited

Loading

eddyb commented Apr 26, 2018 •

edited

Loading

hanna-kruppe Apr 26, 2018 •

edited

Loading

rfcbot commented May 3, 2018 •

edited by aturon

Loading