ci: updates for Rust 1.84

[Owen-CH-Leung]

See this

ruff has released 0.9.0 few hours ago and the CI didn't pin the ruff version, hence ruff checks that noxfile.py is malformed, and the CI returns this error. Since the fmt CI errors out, the remaining CI pipelines are marked as cancelled and not running

This MR is created by running ruff format . which reformats the noxfile.py

[LilyFoote]

Given this is purely internal formatting we don't need the newsfragment for the changelog.

[Owen-CH-Leung]

Given this is purely internal formatting we don't need the newsfragment for the changelog.

Thanks. Removed changelog

[ngoldbaum]

Well that's fun, rustup update stable is broken right now. I can trigger it on my local dev setup.

[Owen-CH-Leung]

Well that's fun, rustup update stable is broken right now. I can trigger it on my local dev setup.

Same here on my mac m1. I got :

info: syncing channel updates for 'stable-aarch64-apple-darwin'
info: latest update on 2025-01-09, rust version 1.84.0 (9fc6b4312 2025-01-07)
error: component 'rust-std' for target 'wasm32-wasi' is unavailable for download for channel 'stable'

[ngoldbaum]

You can fix it locally with

rustup target remove wasm32-wasi
rustup update
rustup target add wasm32-wasip1

I'll open a separate PR to (hopefully) fix CI.

[ngoldbaum]

I'll open a separate PR to (hopefully) fix CI.

Actually it looks like both fixes need to go in simultaneously. Can you apply the change from my other PR to this PR?

[codspeed-hq]

CodSpeed Performance Report

Merging #4846 will improve performances by 15.87%

_{Comparing Owen-CH-Leung:Fix-ruff-fmt-error (408d14b) with main (c0f08c2)}

Summary

⚡ 1 improvements
✅ 83 untouched benchmarks

Benchmarks breakdown

	Benchmark	main	Owen-CH-Leung:Fix-ruff-fmt-error	Change
⚡	extract_float_downcast_fail	425.8 ns	367.5 ns	+15.87%

[Icxolu]

We will also need to bless the ui tests for CI to pass again. (I've done that locally, I may push those here as well)

[ngoldbaum]

I've done that locally, I may push those here as well

please do so, I just pushed the fix the failing target

sorry @Owen-CH-Leung - didn't want to trample on your branch but you happened to have unlucky timing coinciding with a rust release and I'd like to fix this quickly.

[ngoldbaum]

Looks like a UI test change is missing:

EXPECTED:
  ┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈
  error[E0277]: the trait bound `PyErr: From<MyError>` is not satisfied
    --> tests/ui/invalid_result_conversion.rs:22:25
     |
  22 | fn should_not_work() -> Result<(), MyError> {
     |                         ^^^^^^ the trait `From<MyError>` is not implemented for `PyErr`, which is required by `MyError: Into<PyErr>`
     |
     = help: the following other types implement trait `From<T>`:
               `PyErr` implements `From<AddrParseError>`
               `PyErr` implements `From<DecodeUtf16Error>`
               `PyErr` implements `From<DowncastError<'_, '_>>`
               `PyErr` implements `From<DowncastIntoError<'_>>`
               `PyErr` implements `From<FromUtf16Error>`
               `PyErr` implements `From<FromUtf8Error>`
               `PyErr` implements `From<Infallible>`
               `PyErr` implements `From<IntoInnerError<W>>`
             and $N others
     = note: required for `MyError` to implement `Into<PyErr>`
  ┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈
  
  ACTUAL OUTPUT:
  ┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈
  error[E0277]: the trait bound `PyErr: From<MyError>` is not satisfied
    --> tests/ui/invalid_result_conversion.rs:22:25
     |
  22 | fn should_not_work() -> Result<(), MyError> {
     |                         ^^^^^^ the trait `From<MyError>` is not implemented for `PyErr`
     |
     = help: the following other types implement trait `From<T>`:
               `PyErr` implements `From<AddrParseError>`
               `PyErr` implements `From<DecodeUtf16Error>`
               `PyErr` implements `From<DowncastError<'_, '_>>`
               `PyErr` implements `From<DowncastIntoError<'_>>`
               `PyErr` implements `From<FromUtf16Error>`
               `PyErr` implements `From<FromUtf8Error>`
               `PyErr` implements `From<Infallible>`
               `PyErr` implements `From<IntoInnerError<W>>`
             and $N others
     = note: required for `MyError` to implement `Into<PyErr>`
  ┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈┈
  note: If the actual output is the correct output you can bless it by rerunning
        your test with the environment variable TRYBUILD=overwrite

[Icxolu]

Yeah, that one does not run on windows, that's why it slipped through 😓 . Will fix it shortly.

[ngoldbaum]

Maybe we should temporarily allow the WASM emscripten CI to fail?

[Icxolu]

It's probably this one I would guess...

The wasm32-unknown-emscripten target’s binary release of the standard library is now built with the latest emsdk 3.1.68, which fixes an ABI-incompatibility with Emscripten >= 3.1.42. If you are locally using a version of emsdk with an incompatible ABI (e.g. before 3.1.42 or a future one), you should build your code with -Zbuild-std to ensure that std uses the correct ABI.

It looks like we build with 3.1.13, but I'm not sure whats decides which version gets used.

[Icxolu]

Doesn't look like that worked...

[Icxolu]

Looks like that worked 🎉