Repository navigation
Skip registration for empty unreachable tails - #1328
Merged
charliermarsh merged 1 commit intoOct 18, 2026
Merged
Conversation
charliermarsh
deleted the
charlie/codex-skip-empty-unreachable-tails-20261018
branch
October 18, 2026 08:05
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
We avoid creating a discarded compiler when a terminating statement has no remaining statements. Nonempty unreachable tails retain the existing constant-cache and private-name handling.
Matched project profiles use 0.57-0.88% fewer instructions for Mypy across the sampled Python 3.12-3.15 configurations, 0.57% fewer for SymPy and 0.93% fewer for Django on Python 3.15 O2. The 177-case instruction geometric mean improves by 0.17%, with individual changes from -3.37% to +0.54%. All measured project outputs are byte-identical.
Native CPU time decreases 1.19% for Mypy and 1.17% for SymPy, improving in both measurement orders. Mypy peak RSS rises 0.75 MiB (2.53%); SymPy peak RSS falls 0.28%. Each study retains one timing warning and all 120 measured samples, with no exclusions.
Django native CPU timing is order-sensitive: +0.21% in base-first order and -2.98% in head-first order, with flat peak RSS. We do not claim a native CPU gain for Django. All compatibility configurations pass with no new differences.