Add `f16` formatting and parsing by tgross35 · Pull Request #127013 · rust-lang/rust

tgross35 · 2024-06-27T04:49:47Z

Use the same algorithms as for f32 and f64 to implement f16 parsing and printing.

try-job: x86_64-gnu-aux

tgross35 · 2024-06-27T04:59:12Z

@rustbot label +rla-silenced

tgross35 · 2024-06-27T09:16:35Z

@rustbot label +F-f16_and_f128

tgross35 · 2024-07-09T19:31:19Z

This will need #127510

bors · 2024-07-15T20:16:57Z

☔ The latest upstream changes (presumably #127020) made this pull request unmergeable. Please resolve the merge conflicts.

library/core/src/num/dec2flt/number.rs

tgross35 · 2024-07-28T19:46:30Z

Update: I'm really just waiting on #128083 to bump stage0, managing #[cfg(not(bootstrap))] and always building stage1 got too painful.

bors · 2024-07-30T00:17:37Z

☔ The latest upstream changes (presumably #128360) made this pull request unmergeable. Please resolve the merge conflicts.

Use the existing Lemire (decimal -> float) and Dragon / Grisu algorithms (float -> decimal) to add support for `f16`. This allows updating the implementation for `Display` to the expected behavior for `Display` (currently it prints the a hex bitwise representation), matching other floats, and adds a `FromStr` implementation. In order to avoid crashes when compiling with Cranelift or on targets where f16 is not well supported, a fallback is used if `cfg(target_has_reliable_f16)` is not true.

Add `f16` formatting and parsing Use the same algorithms as for `f32` and `f64` to implement `f16` parsing and printing. try-job: x86_64-gnu-aux

Extend the existing tests for `f32` and `f64` with versions that include `f16`'s new printing and parsing implementations. Co-authored-by: Speedy_Lex <alex.ciocildau@gmail.com>

This requires a fix to the subnormal test to cap the maximum allowed value within the maximum mantissa.

tgross35 · 2025-05-18T16:52:29Z

@bors try

Add `f16` formatting and parsing Use the same algorithms as for `f32` and `f64` to implement `f16` parsing and printing. try-job: x86_64-gnu-aux

bors · 2025-05-18T16:53:39Z

⌛ Trying commit 250869e with merge fc768c1...

bors · 2025-05-18T18:46:17Z

☀️ Try build successful - checks-actions
Build commit: fc768c1 (fc768c1d28c6ea7cd5c231c94c80f94ac5b7ee52)

tgross35 · 2025-05-18T19:00:04Z

Needed a config for Miri.

@bors r=Mark-Simulacrum rollup=iffy

bors · 2025-05-18T19:00:06Z

📌 Commit 250869e has been approved by Mark-Simulacrum

It is now in the queue for this repository.

…Simulacrum Add `f16` formatting and parsing Use the same algorithms as for `f32` and `f64` to implement `f16` parsing and printing. try-job: x86_64-gnu-aux

Rollup of 6 pull requests Successful merges: - rust-lang#127013 (Add `f16` formatting and parsing) - rust-lang#130887 (Safer implementation of RepeatN) - rust-lang#140154 (Cygwin support in rustc) - rust-lang#140874 (make `rustc_attr_parsing` less dominant in the rustc crate graph) - rust-lang#141094 (add regression test for rust-lang#101650) - rust-lang#141110 ([std] fix the presentation of `split_off_mut` and `split_off` documentation) r? `@ghost` `@rustbot` modify labels: rollup

bors · 2025-05-18T20:38:04Z

⌛ Testing commit 250869e with merge 4d051fb...

bors · 2025-05-18T23:51:57Z

☀️ Test successful - checks-actions
Approved by: Mark-Simulacrum
Pushing 4d051fb to master...

github-actions · 2025-05-18T23:54:46Z

What is this?

This is an experimental post-merge analysis report that shows differences in test outcomes between the merged PR and its parent PR.

Comparing 4455c89 (parent) -> 4d051fb (this PR)

Test differences

Show 17 test diffs

Stage 0

num::dec2flt::decimal::check_fast_path_f16: [missing] -> pass (J1)
num::dec2flt::float::test_f16_consts: [missing] -> pass (J1)
num::dec2flt::float::test_f16_integer_decode: [missing] -> pass (J1)
num::dec2flt::lemire::compute_float_f16_rounding: [missing] -> pass (J1)
num::flt2dec::random::exact_f16_random_equivalence_test: [missing] -> pass (J1)
num::flt2dec::random::shortest_f16_exhaustive_equivalence_test: [missing] -> pass (J1)

Stage 1

num::dec2flt::decimal::check_fast_path_f16: [missing] -> pass (J0)
num::dec2flt::float::test_f16_consts: [missing] -> pass (J0)
num::dec2flt::float::test_f16_integer_decode: [missing] -> pass (J0)
num::dec2flt::lemire::compute_float_f16_rounding: [missing] -> pass (J0)
num::flt2dec::random::exact_f16_random_equivalence_test: [missing] -> pass (J0)
num::flt2dec::random::shortest_f16_exhaustive_equivalence_test: [missing] -> pass (J2)
num::flt2dec::random::shortest_f16_exhaustive_equivalence_test: [missing] -> ignore (J3)

Additionally, 4 doctest diffs were found. These are ignored, as they are noisy.

Job group index

Test dashboard

Run

cargo run --manifest-path src/ci/citool/Cargo.toml -- \
    test-dashboard 4d051fb306e661654d088892e02e69b8c0c39d43 --output-dir test-dashboard

And then open test-dashboard/index.html in your browser to see an overview of all executed tests.

Job duration changes

dist-aarch64-linux: 5276.3s -> 7698.3s (45.9%)
dist-x86_64-apple: 10179.7s -> 7488.0s (-26.4%)
x86_64-apple-2: 4891.2s -> 4440.8s (-9.2%)
aarch64-apple: 3815.6s -> 4158.7s (9.0%)
x86_64-apple-1: 7586.7s -> 6936.2s (-8.6%)
dist-apple-various: 5946.6s -> 6435.1s (8.2%)
x86_64-msvc-ext1: 7033.4s -> 7521.5s (6.9%)
i686-gnu-2: 6282.8s -> 6683.9s (6.4%)
x86_64-gnu-aux: 5838.3s -> 6206.0s (6.3%)
x86_64-gnu-llvm-20-2: 6240.9s -> 5852.2s (-6.2%)

How to interpret the job duration changes?

Job durations can vary a lot, based on the actual runner instance
that executed the job, system noise, invalidated caches, etc. The table above is provided
mostly for t-infra members, for simpler debugging of potential CI slow-downs.

tgross35 · 2025-05-18T23:56:35Z

🎉🎉🎉 just a month short of this PR’s one year anniversary, I’m very glad to have this in! Thanks @speedy-lex for all the help.

rust-timer · 2025-05-19T01:24:19Z

Finished benchmarking commit (4d051fb): comparison URL.

Overall result: ❌ regressions - no action needed

@rustbot label: -perf-regression

Instruction count

This is the most reliable metric that we have; it was used to determine the overall result at the top of this comment. However, even this metric can sometimes exhibit noise.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	0.8%	[0.4%, 1.0%]	4
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-	-	0

Max RSS (memory usage)

Results (secondary 0.2%)

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	0.7%	[0.6%, 0.9%]	3
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-0.7%	[-0.7%, -0.6%]	2
All ❌✅ (primary)	-	-	0

Cycles

Results (secondary 1.3%)

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	2.9%	[0.4%, 5.4%]	6
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-1.9%	[-2.0%, -1.9%]	3
All ❌✅ (primary)	-	-	0

Binary size

Results (primary 0.2%, secondary 0.3%)

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	0.2%	[0.0%, 0.5%]	16
Regressions ❌ (secondary)	0.3%	[0.0%, 0.5%]	71
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	0.2%	[0.0%, 0.5%]	16

Bootstrap: 775.321s -> 775.275s (-0.01%)
Artifact size: 365.53 MiB -> 365.57 MiB (0.01%)

…mulacrum Add `f16` formatting and parsing Use the same algorithms as for `f32` and `f64` to implement `f16` parsing and printing. try-job: x86_64-gnu-aux

Upstream Rust currently does not support printing f16s in decimal. For the intrinsics tests to work, make C++ print float16_t in the same format. Can be droppen once #127013 is merged.

Upstream Rust currently does not support printing f16s in decimal. For the intrinsics tests to work, make C++ print float16_t in the same format. Can be droppen once rust-lang#127013 is merged.

rustbot added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. T-libs Relevant to the library team, which will review and decide on the PR/issue. labels Jun 27, 2024

This comment has been minimized.

Sign in to view

rustbot added the rla-silenced Silences rust-log-analyzer postings to the PR it's added on. label Jun 27, 2024

rustbot added the F-f16_and_f128 `#![feature(f16)]`, `#![feature(f128)]` label Jun 27, 2024

fbstj reviewed Jul 15, 2024

View reviewed changes

library/core/src/num/dec2flt/number.rs Outdated Show resolved Hide resolved

tgross35 mentioned this pull request Jul 18, 2024

Tracking Issue for f16 and f128 float types #116909

Open

92 tasks

tgross35 force-pushed the f16-format-parse branch 2 times, most recently from 7c3f9c1 to f3ebeb3 Compare July 23, 2024 09:52

tgross35 force-pushed the f16-format-parse branch from 2eaa479 to 422c52e Compare August 6, 2024 08:47

Dylan-DPC 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-review Status: Awaiting review from the assignee but also interested parties. labels Aug 20, 2024

tgross35 force-pushed the f16-format-parse branch 6 times, most recently from 21ffabc to 2098f01 Compare August 25, 2024 07:53

rustbot added the T-bootstrap Relevant to the bootstrap subteam: Rust's build system (x.py and src/bootstrap) label Aug 25, 2024

tgross35 force-pushed the f16-format-parse branch 5 times, most recently from 404089f to 3636530 Compare August 30, 2024 09:04

This comment was marked as outdated.

Sign in to view

tgross35 and others added 2 commits May 18, 2025 16:51

float: Add tests for f16 conversions to and from decimal

977d841

Extend the existing tests for `f32` and `f64` with versions that include `f16`'s new printing and parsing implementations. Co-authored-by: Speedy_Lex <alex.ciocildau@gmail.com>

float: Add f16 to test-float-parse

250869e

This requires a fix to the subnormal test to cap the maximum allowed value within the maximum mantissa.

tgross35 force-pushed the f16-format-parse branch from 1c38c28 to 250869e Compare May 18, 2025 16:51

This comment was marked as outdated.

Sign in to view

fmease mentioned this pull request May 18, 2025

Rollup of 6 pull requests #141233

Closed

bors added the merged-by-bors This PR was explicitly merged by bors. label May 18, 2025

bors merged commit 4d051fb into rust-lang:master May 18, 2025
7 checks passed

rustbot added this to the 1.89.0 milestone May 18, 2025

tgross35 deleted the f16-format-parse branch May 18, 2025 23:56

folkertdev mentioned this pull request May 19, 2025

avx512_target_feature is now stable on nightly rust-lang/stdarch#1802

Merged

Uh oh!

Conversation

tgross35 commented Jun 27, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

This comment has been minimized.

tgross35 commented Jun 27, 2024

Uh oh!

tgross35 commented Jun 27, 2024

Uh oh!

tgross35 commented Jul 9, 2024

Uh oh!

bors commented Jul 15, 2024

Uh oh!

Uh oh!

tgross35 commented Jul 28, 2024

Uh oh!

bors commented Jul 30, 2024

Uh oh!

This comment was marked as outdated.

This comment was marked as outdated.

tgross35 commented May 18, 2025

Uh oh!

bors commented May 18, 2025

Uh oh!

This comment was marked as outdated.

bors commented May 18, 2025

Uh oh!

tgross35 commented May 18, 2025

Uh oh!

bors commented May 18, 2025

Uh oh!

bors commented May 18, 2025

Uh oh!

bors commented May 18, 2025

Uh oh!

Uh oh!

github-actions bot commented May 18, 2025

Test differences

Stage 0

Stage 1

Job duration changes

Uh oh!

tgross35 commented May 18, 2025

Uh oh!

rust-timer commented May 19, 2025

Overall result: ❌ regressions - no action needed

Instruction count

Max RSS (memory usage)

Cycles

Binary size

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

11 participants

tgross35 commented Jun 27, 2024 •

edited

Loading