Add optional zerocopy support #62
Draft
Conversation
This file contains 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
meithecatte
requested changes
Nov 8, 2024
| T::Numeric: Immutable, | ||
| T::Numeric: FromZeros, | ||
| { | ||
| fn only_derive_is_allowed_to_implement_this_trait() {} |
There was a problem hiding this comment.
I think this could use a comment linking to google/zerocopy#287, so that this doesn't scream "crimes! damn crimes!" at anyone who looks at this later.
src/lib.rs
Outdated
Comment on lines
1104
to
1107
| { | ||
| (my_candidate.read_unaligned::<zerocopy::pointer::BecauseImmutable>() ^ T::ALL_BITS) | ||
| == T::EMPTY | ||
| } |
There was a problem hiding this comment.
This bit manipulation looks suspicious. Please add tests that exercise this. Or, maybe something like BitFlags::from_bits_truncate(x).bits() == x would be better, in terms of code reuse.
1cda797 to
b67dbea
Compare
With the zerocopy feature activated we implement some zerocopy traits containing an `only_derive_is_allowed_to_implement_this_trait` function. That is fine as long as we know what we are doing.
meithecatte
reviewed
Nov 19, 2024
Comment on lines
+689
to
+695
| #[must_use] | ||
| #[inline(always)] | ||
| pub fn validate_bits(bits: T::Numeric) -> bool { | ||
| // SAFETY: We're truncating out all the invalid bits so it will | ||
| // only be different if there are invalid bits set. | ||
| (bits & T::ALL_BITS) == bits | ||
| } |
There was a problem hiding this comment.
Honestly, I'm not sure if this makes sense as a public API.
| // Dereference the pointer to the candidate | ||
| let candidate = | ||
| my_candidate.read_unaligned::<zerocopy::pointer::BecauseImmutable>(); | ||
| return BitFlags::<T>::validate_bits(candidate); |
There was a problem hiding this comment.
Suggested change
| return BitFlags::<T>::validate_bits(candidate); | |
| BitFlags::<T>::validate_bits(candidate) |
| @@ -1102,8 +1129,11 @@ mod impl_zerocopy { | |||
| let my_candidate = | |||
| unsafe { candidate.assume_validity::<zerocopy::pointer::invariant::Valid>() }; | |||
| { | |||
| (my_candidate.read_unaligned::<zerocopy::pointer::BecauseImmutable>() ^ T::ALL_BITS) | |||
| == T::EMPTY | |||
| // TODO: Currently this assumes that the candidate is aligned. We actually need to check this beforehand | |||
There was a problem hiding this comment.
Not sure if this is something I would've caught. Do you have any links I could reference for example zerocopy implementations for similar types?
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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 PR adds a flag that enables implementations for zerocopy traits on BitFlags.
Converting a
BitFlagsinto the underlying number is trivial and always safe, so we can just derive those traits. Converting from a number toBitFlagsis more tricky, because not every bit-pattern is valid. To assert that the bit-pattern is valid this PR provides a custom implementation of thezerocopy::TryFromBytestrait. Custom implementations for zerocopy traitsTryFromBytesseem somewhat discouraged but supported, so it should be fine to implement that.