Landing: cc5179f3a36a
Project / Subsystem
gcc / gcc/tree-optimization/vect
Date
2026-05-12
Author
Lili Cui
Commit
cc5179f3a36a087891beecd1b3cb23d172dc76c7
Source
github
Perf win
Yes
Breaking
No
All attributes
- project
- gcc
- subsystem
- gcc/tree-optimization/vect
- patch_id
- —
- commit_hash
- cc5179f3a36a087891beecd1b3cb23d172dc76c7
- source_type
- github
- headline
- Allow Single-Lane SLP Fallback When Limit Is Exhausted
- tldr
- GCC now allows single-lane SLP vectorization to proceed even when the multi-lane SLP discovery limit is exhausted, fixing missed optimizations in loops with co…
- author
- Lili Cui
- outcome
- committed
- performance_win
- true
- breaking_change
- false
- series_id
- gcc:tree-optimization/vect: allow single-lane slp fallback when limit is exhausted
- series_parts
- []
- tags
-
- • optimization
- • vectorization
- • slp
- discussion_id_link
- —
- bugzilla_pr
- —
- date
- 2026-05-12T00:00:00.000Z
GCC’s SLP vectorizer now attempts single-lane vectorization even after the multi-lane vectorization limit is reached. Previously, the limit check blocked all SLP discovery, including single-lane fallbacks, leading to missed optimizations in code with multiple independent conditional reductions. This change improves performance (3.8% on EMR, 1.4% on Znver5 for 731.astcenc_r with -Ofast).