-
Notifications
You must be signed in to change notification settings - Fork 12.7k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Avoid creating SmallVec
s in global_llvm_features
#97579
Conversation
r? @nagisa (rust-highfive has picked a reviewer for you, use r? to override) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Nice improvement to readability! See nits inline though.
@@ -218,6 +218,9 @@ pub fn check_tied_features( | |||
sess: &Session, | |||
features: &FxHashMap<&str, bool>, | |||
) -> Option<&'static [&'static str]> { | |||
if features.len() == 0 { |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Nit: this is probably better written as something like:
if !features.is_empty() {
for tied in ... {
...
}
}
return None;
The code isn't particularly deeply nested here and avoids having to specify None
as a return twice.
@@ -485,35 +490,35 @@ pub(crate) fn global_llvm_features(sess: &Session, diagnostics: bool) -> Vec<Str | |||
} | |||
diag.emit(); | |||
} | |||
Some((enable_disable, feature)) | |||
|
|||
if diagnostics && tied_features.iter().any(|tied| tied.contains(&feature)) { |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This is doing a second linear lookup through the tied_features
list (in addition to the one that will happen later in the check_tied_features
.
Ignoring entirely the fact that what's going on here is an exercise in micro-optimization, and the overall performance of the compiler is likely going to stay pretty much the same, I feel like such double lookup is probably defeating any improvements gained here by all the other changes.
It would seem to me that we'd be better off by just not checking this here (while leaving everything else intact):
if diagnostics && tied_features.iter().any(|tied| tied.contains(&feature)) { | |
if diagnostics { |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks for pointing out! That makes sense and I think we can keep the FIXME here
c10ae82
to
b3cd892
Compare
Thanks for reviewing! I made the corresponding corrections |
@bors r+ thanks! |
📌 Commit b3cd892 has been approved by |
Rollup of 5 pull requests Successful merges: - rust-lang#97312 (Compute lifetimes in scope at diagnostic time) - rust-lang#97495 (Add E0788 for improper #[no_coverage] usage) - rust-lang#97579 (Avoid creating `SmallVec`s in `global_llvm_features`) - rust-lang#97767 (interpret: do not claim UB until we looked more into variadic functions) - rust-lang#97787 (E0432: rust 2018 -> rust 2018 or later in --explain message) Failed merges: r? `@ghost` `@rustbot` modify labels: rollup
This PR made a simple optimization to avoid creating extra
SmallVec
s by adjusting the use of iterator statements.Also, given the very small size of
tied_target_features
, there is no need to insert each feature into the FxHashMap.