Make some float methods unstable `const fn` #130568

eduardosm · 2024-09-19T16:50:08Z

Some float methods are now const fn under the const_float_methods feature gate.

I also made some unstable methods const fn, keeping their constness under their respective feature gate.

In order to support min, max, abs and copysign, the implementation of some intrinsics had to be moved from Miri to rustc_const_eval (cc @RalfJung).

Tracking issue: #130843

impl <float> {
    // #[feature(const_float_methods)]
    pub const fn recip(self) -> Self;
    pub const fn to_degrees(self) -> Self;
    pub const fn to_radians(self) -> Self;
    pub const fn max(self, other: Self) -> Self;
    pub const fn min(self, other: Self) -> Self;
    pub const fn clamp(self, min: Self, max: Self) -> Self;
    pub const fn abs(self) -> Self;
    pub const fn signum(self) -> Self;
    pub const fn copysign(self, sign: Self) -> Self;

    // #[feature(float_minimum_maximum)]
    pub const fn maximum(self, other: Self) -> Self;
    pub const fn minimum(self, other: Self) -> Self;

    // Only f16/f128 (f32/f64 already const)
    pub const fn is_sign_positive(self) -> bool;
    pub const fn is_sign_negative(self) -> bool;
    pub const fn next_up(self) -> Self;
    pub const fn next_down(self) -> Self;
}

r? libs-api

try-job: dist-s390x-linux

rustbot · 2024-09-19T16:50:18Z

The Miri subtree was changed

cc @rust-lang/miri

Some changes occurred to the CTFE / Miri interpreter

cc @rust-lang/miri

compiler/rustc_const_eval/src/interpret/intrinsics.rs

RalfJung · 2024-09-21T21:08:16Z

r=me on the interpreter and Miri parts.

eduardosm · 2024-09-24T19:45:22Z

I made clamp const too, using an approach similar to #130611 to handle the assert formatting.

tgross35 · 2024-09-24T23:28:06Z

If this is accepted, a tracking issue should be created (for now I set issue = "none").

You can go ahead and do this in advance, I don't think there is any reason these would be rejected

library/core/src/num/f128.rs

eduardosm · 2024-09-25T17:22:48Z

Created tracking issue: #130843

For f16/f32 methods, I placed commented gates under the same feature as f32/f64 (e.g., is_sign_positive under const_float_classify, next_up under float_next_up_down).

bors · 2024-10-04T21:32:39Z

☔ The latest upstream changes (presumably #130157) made this pull request unmergeable. Please resolve the merge conflicts.

eduardosm · 2024-10-04T22:05:39Z

I believe is_sign_positive / is_sign_negative do not need the second unstable gate since const_float_classify is now stable.

src/tools/miri/src/intrinsics/mod.rs

RalfJung · 2024-10-05T13:29:41Z

compiler/rustc_const_eval/src/interpret/intrinsics.rs

+    }
+}
+
+enum FloatBinOp {


Suggested change

enum FloatBinOp {

enum FloatBinIntrinsic {

We use BinOp for the MIR binops, IMO we shouldn't mix up that terminology.

Also please move this above the functions that use it.

tests/ui/consts/const-eval/float_methods.rs

RalfJung · 2024-10-05T13:53:20Z

LGTM from the interpreter side.

r? libs-api

jieyouxu · 2024-10-15T03:08:53Z

I think this failed in the rollup #131711 (comment)

2024-10-15T01:28:03.7030089Z rustc-LLVM ERROR: Cannot select: 0x7fc34458bd10: f32 = fp16_to_fp 0x7fc34458ba70
2024-10-15T01:28:03.7031272Z   0x7fc34458ba70: i32 = fp_to_fp16 0x7fc34458b0d0
2024-10-15T01:28:03.7032292Z     0x7fc34458b0d0: f32,ch = CopyFromReg 0x7fc34490cbf8, Register:f32 %1
2024-10-15T01:28:03.7033231Z       0x7fc34458b7d0: f32 = Register %1
2024-10-15T01:28:03.7034318Z In function: _ZN4core3f1621_$LT$impl$u20$f16$GT$5clamp15assert_at_const17h0a21f61a7a1319afE

jieyouxu · 2024-10-15T03:10:16Z

@bors r-

Make some float methods unstable `const fn` Some float methods are now `const fn` under the `const_float_methods` feature gate. I also made some unstable methods `const fn`, keeping their constness under their respective feature gate. In order to support `min`, `max`, `abs` and `copysign`, the implementation of some intrinsics had to be moved from Miri to rustc_const_eval (cc `@RalfJung).` Tracking issue: rust-lang#130843 ```rust impl <float> { // #[feature(const_float_methods)] pub const fn recip(self) -> Self; pub const fn to_degrees(self) -> Self; pub const fn to_radians(self) -> Self; pub const fn max(self, other: Self) -> Self; pub const fn min(self, other: Self) -> Self; pub const fn clamp(self, min: Self, max: Self) -> Self; pub const fn abs(self) -> Self; pub const fn signum(self) -> Self; pub const fn copysign(self, sign: Self) -> Self; // #[feature(float_minimum_maximum)] pub const fn maximum(self, other: Self) -> Self; pub const fn minimum(self, other: Self) -> Self; // Only f16/f128 (f32/f64 already const) pub const fn is_sign_positive(self) -> bool; pub const fn is_sign_negative(self) -> bool; pub const fn next_up(self) -> Self; pub const fn next_down(self) -> Self; } ``` r? libs-api try-job: dist-s390x-linux

compiler-errors · 2024-10-15T03:12:14Z

@bors rollup=never

tgross35 · 2024-10-15T03:13:48Z

Ah yeah yuck, every function in stdlib that takes a f16 or f128 in the signature still needs #[inline] so LLVM doesn't try to codegen and crash on platforms where that is buggy. This includes the assert_at_const and assert_at_rt functions.

Some float methods are now `const fn` under the `const_float_methods` feature gate. In order to support `min`, `max`, `abs` and `copysign`, the implementation of some intrinsics had to be moved from Miri to rustc_const_eval.

eduardosm · 2024-10-15T08:48:14Z

Fixed by adding #[inline] to assert_at_const. Tested locally with ./x build --target s390x-unknown-linux-gnu library/core

RalfJung · 2024-10-15T09:11:35Z

@bors r=RalfJung,tgross35 rollup

bors · 2024-10-15T09:11:39Z

📌 Commit c09ed3e has been approved by RalfJung,tgross35

It is now in the queue for this repository.

RalfJung · 2024-10-15T09:12:55Z

Ah yeah yuck, every function in stdlib that takes a f16 or f128 in the signature still needs #[inline] so LLVM doesn't try to codegen and crash on platforms where that is buggy. This includes the assert_at_const and assert_at_rt functions.

Wow f16 / f128 are more unstable than I thought.^^ Is there an issue tracking specifically the codegen crashes? Can we have a central workaround instead of dozens of #[inline] everywhere where we will surely forget to remove some?

tgross35 · 2024-10-15T12:54:58Z

Wow f16 / f128 are more unstable than I thought.^^ Is there an issue tracking specifically the codegen crashes? Can we have a central workaround instead of dozens of #[inline] everywhere where we will surely forget to remove some?

The list of platforms where just seeing the types is enough to make LLVM crash is fortunately pretty limited, wasm and arm64ec being some notably more popular exceptions https://github.com/rust-lang/compiler-builtins/blob/cb060052ab7e4bad408c85d44be7e60096e93e38/configure.rs#L53-L77. AIUI however, we need the inline annotations anyway since Cranelift doesn't support the types, and having anything not inline would mean it can't compile std.

I'm not sure what would be better here.

RalfJung · 2024-10-15T13:17:00Z

I'm not sure what would be better here.

We can automatically mark functions as inline if they have a f16/f128 argument. 🙈

tgross35 · 2024-10-15T13:18:18Z

That works 😄

beetrees · 2024-10-15T16:10:07Z

The codegen crashes are currently tracked as part of the big list in the tracking issue at #116909 (comment), but there isn't a separate dedicated tracking issue just for them. Adding a cfg via core/std's build.rs for whether f16/f128 codegen will succeed or not is definitely possible: compiler-builtins already has such a cfg, and std already has a different list for whether f16/f128 tests will pass or not.

RalfJung · 2024-10-15T16:35:10Z

Specifically, the logic here, which already makes "small" functions auto-#[inline], could also check the signature.

…mpiler-errors Rollup of 7 pull requests Successful merges: - rust-lang#129794 (uefi: Implement getcwd and chdir) - rust-lang#130568 (Make some float methods unstable `const fn`) - rust-lang#131521 (rename RcBox to RcInner for consistency) - rust-lang#131701 (Don't report `on_unimplemented` message for negative traits) - rust-lang#131705 (Fix most ui tests on emscripten target) - rust-lang#131733 (Fix uninlined_format_args in stable_mir) - rust-lang#131734 (Update `arm64e-apple-tvos` maintainer) r? `@ghost` `@rustbot` modify labels: rollup

Rollup merge of rust-lang#130568 - eduardosm:const-float-methods, r=RalfJung,tgross35 Make some float methods unstable `const fn` Some float methods are now `const fn` under the `const_float_methods` feature gate. I also made some unstable methods `const fn`, keeping their constness under their respective feature gate. In order to support `min`, `max`, `abs` and `copysign`, the implementation of some intrinsics had to be moved from Miri to rustc_const_eval (cc `@RalfJung).` Tracking issue: rust-lang#130843 ```rust impl <float> { // #[feature(const_float_methods)] pub const fn recip(self) -> Self; pub const fn to_degrees(self) -> Self; pub const fn to_radians(self) -> Self; pub const fn max(self, other: Self) -> Self; pub const fn min(self, other: Self) -> Self; pub const fn clamp(self, min: Self, max: Self) -> Self; pub const fn abs(self) -> Self; pub const fn signum(self) -> Self; pub const fn copysign(self, sign: Self) -> Self; // #[feature(float_minimum_maximum)] pub const fn maximum(self, other: Self) -> Self; pub const fn minimum(self, other: Self) -> Self; // Only f16/f128 (f32/f64 already const) pub const fn is_sign_positive(self) -> bool; pub const fn is_sign_negative(self) -> bool; pub const fn next_up(self) -> Self; pub const fn next_down(self) -> Self; } ``` r? libs-api try-job: dist-s390x-linux

rustbot assigned joshtriplett Sep 19, 2024

rustbot added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. T-compiler Relevant to the compiler team, which will review and decide on the PR/issue. T-libs Relevant to the library team, which will review and decide on the PR/issue. labels Sep 19, 2024

jieyouxu added the T-libs-api Relevant to the library API team, which will review and decide on the PR/issue. label Sep 19, 2024

RalfJung reviewed Sep 21, 2024

View reviewed changes

compiler/rustc_const_eval/src/interpret/intrinsics.rs Outdated Show resolved Hide resolved

compiler/rustc_const_eval/src/interpret/intrinsics.rs Outdated Show resolved Hide resolved

eduardosm force-pushed the const-float-methods branch from 4c6f58d to e60ae9b Compare September 21, 2024 19:49

eduardosm force-pushed the const-float-methods branch from e60ae9b to bf34851 Compare September 24, 2024 19:44

This comment has been minimized.

Sign in to view

eduardosm force-pushed the const-float-methods branch from bf34851 to 10c3e6c Compare September 24, 2024 20:35

tgross35 reviewed Sep 24, 2024

View reviewed changes

library/core/src/num/f128.rs Outdated Show resolved Hide resolved

eduardosm mentioned this pull request Sep 25, 2024

Tracking Issue for #![feature(const_float_methods)] #130843

Open

3 tasks

eduardosm force-pushed the const-float-methods branch 2 times, most recently from 7caa07b to c06c255 Compare September 25, 2024 17:16

This comment has been minimized.

Sign in to view

eduardosm force-pushed the const-float-methods branch from c06c255 to a79103f Compare September 25, 2024 17:23

eduardosm mentioned this pull request Sep 25, 2024

Tracking Issue for float_minimum_maximum #91079

Open

4 tasks

eduardosm force-pushed the const-float-methods branch from a79103f to fa312ed Compare October 4, 2024 22:02

RalfJung reviewed Oct 5, 2024

View reviewed changes

src/tools/miri/src/intrinsics/mod.rs Outdated Show resolved Hide resolved

RalfJung reviewed Oct 5, 2024

View reviewed changes

eduardosm force-pushed the const-float-methods branch from a61692c to 593f644 Compare October 5, 2024 13:41

rustbot assigned BurntSushi Oct 5, 2024

bors added S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. and removed S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. labels Oct 15, 2024

This comment was marked as resolved.

Sign in to view

Make some float methods unstable const fn

c09ed3e

Some float methods are now `const fn` under the `const_float_methods` feature gate. In order to support `min`, `max`, `abs` and `copysign`, the implementation of some intrinsics had to be moved from Miri to rustc_const_eval.

eduardosm force-pushed the const-float-methods branch from 029a881 to c09ed3e Compare October 15, 2024 08:46

bors added S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. and removed S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. labels Oct 15, 2024

compiler-errors mentioned this pull request Oct 15, 2024

Rollup of 7 pull requests #131747

Merged

bors merged commit 2f3f001 into rust-lang:master Oct 15, 2024
6 checks passed

rustbot added this to the 1.84.0 milestone Oct 15, 2024

eduardosm deleted the const-float-methods branch October 16, 2024 10:07

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Make some float methods unstable `const fn` #130568

Make some float methods unstable `const fn` #130568

eduardosm commented Sep 19, 2024 •

edited by jieyouxu

Loading

rustbot commented Sep 19, 2024

RalfJung commented Sep 21, 2024

eduardosm commented Sep 24, 2024

This comment has been minimized.

tgross35 commented Sep 24, 2024

This comment has been minimized.

eduardosm commented Sep 25, 2024

bors commented Oct 4, 2024

eduardosm commented Oct 4, 2024

RalfJung Oct 5, 2024

RalfJung commented Oct 5, 2024

jieyouxu commented Oct 15, 2024 •

edited

Loading

jieyouxu commented Oct 15, 2024

This comment was marked as resolved.

This comment was marked as resolved.

compiler-errors commented Oct 15, 2024

tgross35 commented Oct 15, 2024 •

edited

Loading

This comment was marked as resolved.

This comment was marked as resolved.

eduardosm commented Oct 15, 2024

RalfJung commented Oct 15, 2024

bors commented Oct 15, 2024

RalfJung commented Oct 15, 2024

tgross35 commented Oct 15, 2024

RalfJung commented Oct 15, 2024

tgross35 commented Oct 15, 2024

beetrees commented Oct 15, 2024

RalfJung commented Oct 15, 2024

Make some float methods unstable const fn #130568

Make some float methods unstable const fn #130568

Conversation

eduardosm commented Sep 19, 2024 • edited by jieyouxu Loading

rustbot commented Sep 19, 2024

RalfJung commented Sep 21, 2024

eduardosm commented Sep 24, 2024

This comment has been minimized.

tgross35 commented Sep 24, 2024

This comment has been minimized.

eduardosm commented Sep 25, 2024

bors commented Oct 4, 2024

eduardosm commented Oct 4, 2024

RalfJung Oct 5, 2024

Choose a reason for hiding this comment

RalfJung commented Oct 5, 2024

jieyouxu commented Oct 15, 2024 • edited Loading

jieyouxu commented Oct 15, 2024

This comment was marked as resolved.

This comment was marked as resolved.

compiler-errors commented Oct 15, 2024

tgross35 commented Oct 15, 2024 • edited Loading

This comment was marked as resolved.

This comment was marked as resolved.

eduardosm commented Oct 15, 2024

RalfJung commented Oct 15, 2024

bors commented Oct 15, 2024

RalfJung commented Oct 15, 2024

tgross35 commented Oct 15, 2024

RalfJung commented Oct 15, 2024

tgross35 commented Oct 15, 2024

beetrees commented Oct 15, 2024

RalfJung commented Oct 15, 2024

Make some float methods unstable `const fn` #130568

Make some float methods unstable `const fn` #130568

eduardosm commented Sep 19, 2024 •

edited by jieyouxu

Loading

jieyouxu commented Oct 15, 2024 •

edited

Loading

tgross35 commented Oct 15, 2024 •

edited

Loading