Bump edition to 2024; MSRV 1.85+; version to v3.0.0-pre
#149
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This commit is intended as the base commit for other proposed breaking
changes, almost all of which will involve bumping the MSRV close to this
(in fact if we want `const fn` support for `black_box` we need to go all
the way to 1.86, this doesn't quite do that yet).
It includes a few different changes, and I'm happy to split these up
into separate PRs if that's helpful:
- Updates CI config
- sets `permissions` to read-only
- sets `RUSTFLAGS` to `-D warnings` to deny warnings
- uses `dtolnay/rust-toolchain` instead of the unmaintained
`actions-rs/toolchain@v1`
- uses `run` directly instead of the unmaintained
`actions-rs/cargo@v1`
- removes feature-specific tests because features have been removed(!)
- adds additional `unsafe` code testing using `cargo careful` and
`cargo miri`
- tests both `dev` and `release` builds
- Bumps version to `3.0.0-pre` to denote breaking changes
- Removes obsolete `html_root_url`
- Bumps edition to 2024; note this changed rustfmt rules and so rustfmt
has also been applied with the new rules
- Bumps MSRV to 1.85.0 and adds `rust-version` setting to `Cargo.toml`
- Removes obsolete settings related to Travis CI (which has since been
replaced by GitHub Actions)
- Adds `.github` to `exclude` in `Cargo.toml`
- Bumps `rand` from `0.8` to `0.9` (though `0.10` will be released soon)
- Removes all features: with the MSRV bump, we no longer need any of
these feature gates as all features will be available (the `std`
feature didn't actually use anything from `std` and is thus pointless)
- Makes `core::hint::black_box` the only value barrier/fence strategy
for `Choice`: despite the scary documentation, it should be superior
to the previous `core::ptr::read_volatile` approach, at least based
on questions I have posed to various Rust compiler engineers, and
is also `const fn` stable in Rust 1.86.
tarcieri
added a commit
to tarcieri/subtle
that referenced
this pull request
Jan 4, 2026
Depends on: dalek-cryptography#149 At the time this provided an interesting way to use the `subtle` implementation of `black_box`, but now that `core::hint::black_box` is stable someone who wants equivalent functionality can implement it themselves using that.
tarcieri
commented
Jan 4, 2026
|
|
||
| macro_rules! generate_integer_conditional_select_tests { | ||
| ($($t:ty)*) => ($( | ||
| ($($t:ty),*) => ($( |
Contributor
Author
There was a problem hiding this comment.
Having a comma between type names seemed more consistent with the generate_integer_equal_tests! macro
tarcieri
commented
Jan 4, 2026
tarcieri
force-pushed
the
v3.x-pre/2024-edition-and-cleaups
branch
from
20946de to
4776266
Compare
tarcieri
added a commit
to tarcieri/subtle
that referenced
this pull request
Jan 4, 2026
Depends on: dalek-cryptography#149 At the time this provided an interesting way to use the `subtle` implementation of `black_box`, but now that `core::hint::black_box` is stable someone who wants equivalent functionality can implement it themselves using that.
tarcieri
added a commit
to tarcieri/subtle
that referenced
this pull request
Jan 4, 2026
Depends on: dalek-cryptography#149 At the time this provided an interesting way to use the `subtle` implementation of `black_box`, but now that `core::hint::black_box` is stable someone who wants equivalent functionality can implement it themselves using that.
Contributor
Author
|
I now notice I typoed "cleanups" In the branch name 🤦 |
tarcieri
force-pushed
the
v3.x-pre/2024-edition-and-cleaups
branch
from
4776266 to
fb03fa4
Compare
tarcieri
force-pushed
the
v3.x-pre/2024-edition-and-cleaups
branch
from
fb03fa4 to
39e1376
Compare
tarcieri
added a commit
to tarcieri/subtle
that referenced
this pull request
Jan 4, 2026
Depends on: dalek-cryptography#149 At the time this provided an interesting way to use the `subtle` implementation of `black_box`, but now that `core::hint::black_box` is stable someone who wants equivalent functionality can implement it themselves using that.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This commit is intended as the base commit for other proposed breaking changes, almost all of which will involve bumping the MSRV close to this (in fact if we want
const fnsupport forblack_boxwe need to go all the way to 1.86, this doesn't quite do that yet).It includes a few different changes, and I'm happy to split these up into separate PRs if that's helpful:
permissionsto read-onlyRUSTFLAGSto-D warningsto deny warningsdtolnay/rust-toolchaininstead of the unmaintainedactions-rs/toolchain@v1rundirectly instead of the unmaintainedactions-rs/cargo@v1unsafecode testing usingcargo carefulandcargo miridevandreleasebuilds3.0.0-preto denote breaking changeshtml_root_urlrust-versionsetting toCargo.toml.githubtoexcludeinCargo.tomlrandfrom0.8to0.9(though0.10will be released soon)stdfeature didn't actually use anything fromstdand is thus pointless)core::hint::black_boxthe only value barrier/fence strategy forChoice: despite the scary documentation, it should be superior to the previouscore::ptr::read_volatileapproach, at least based on questions I have posed to various Rust compiler engineers, and is alsoconst fnstable in Rust 1.86. In subsequent commits, I plan on attempting to add thecmovcrate which providesasm!-backed equivalents ofConditionalSelectandConstantTimeEq