-
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
[PERF] Revert "Auto merge of #93670 - erikdesjardins:noundef, r=nikic" #94106
Conversation
r? @lcnr (rust-highfive has picked a reviewer for you, use r? to override) |
@bors try @rust-timer queue |
Awaiting bors try build completion. @rustbot label: +S-waiting-on-perf |
⌛ Trying commit 4346bd1 with merge e6ad3f369d4d6850bc9d7e7d0093e30692701718... |
☀️ Try build successful - checks-actions |
Queued e6ad3f369d4d6850bc9d7e7d0093e30692701718 with parent 30b3f35, future comparison URL. |
r? @ghost |
Finished benchmarking commit (e6ad3f369d4d6850bc9d7e7d0093e30692701718): comparison url. Summary: This benchmark run did not return any relevant results. 24 results were found to be statistically significant but too small to be relevant. If you disagree with this performance assessment, please file an issue in rust-lang/rustc-perf. Benchmarking this pull request likely means that it is perf-sensitive, so we're automatically marking it as not fit for rolling up. While you can manually mark this PR as fit for rollup, we strongly recommend not doing so since this PR led to changes in compiler perf. @bors rollup=never |
Hmm. It's slightly less significant, but probably still worth dealing with: Opened #94127. |
At opt-level=0, apply only ABI-affecting attributes to functions This should provide a small perf improvement for debug builds, and should more than cancel out the perf regression from adding noundef (rust-lang#93670 (comment), rust-lang#94106). r? `@nikic`
This change caused a perf regression in debug builds, due to the cost of adding attributes: #93670 (comment)
The recent LLVM 14 merge (#93577) improved the performance of attribute-related code though, so it may have made this a nonissue.
I do not expect to merge this PR as-is: depending on perf results, I'll either do nothing (if reverting shows no/minimal improvement), or open a PR that makes us only add
noundef
(and related attrs) when optimizing (since they're only useful for optimizations, and the regression is only significant for debug builds).