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Outdated suggestions cannot be applied.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.
Using
vectorize()as opposed to directly calling functions annotated#[inline(always)]is meant to introduce an inlining barrier and make the compiler make inlining decisions. Therefore putting#[inline]on it is rather pointless: if I wanted aggressive inlining, I wouldn't be callingvectorize()in the first place.#[inline]also enables cross-crate inlining, but sincevectorize()is a generic function it will be instantiated in the caller crate, so this is not a concern.More info about inlining: https://matklad.github.io/2021/07/09/inline-in-rust.html
Motivation
I'm seeing a large performance regression from switching from
widetofearless_simdin QuState/PhastFT#58 due to overly aggressive inlining. This alone doesn't fix it, but seems like a sensible first step.