Landing: 39eac69cd3af
Project / Subsystem
gcc / gcc
Date
2026-03-10
Author
Richard Biener
Commit
39eac69cd3afabaeaf70fed19722bf049eac9819
Source
github
Perf win
Yes
Breaking
No
All attributes
- project
- gcc
- subsystem
- gcc
- patch_id
- —
- commit_hash
- 39eac69cd3afabaeaf70fed19722bf049eac9819
- source_type
- github
- headline
- Use scalar costs for peeling cost estimation in vectorizer.
- tldr
- GCC's vectorizer now uses pre-calculated scalar costs when estimating loop peeling costs, improving accuracy and compile time.
- author
- Richard Biener
- outcome
- committed
- performance_win
- true
- breaking_change
- false
- series_id
- —
- series_parts
- []
- tags
-
- • vectorization
- • optimization
- • performance
- discussion_id_link
- —
- bugzilla_pr
- —
- date
- 2026-03-10T00:00:00.000Z
GCC’s vectorizer now uses the scalar_costs summary when computing the peeling cost, avoiding redundant computations. This change replaces repeated walks of LOOP_VINFO_SCALAR_ITERATION_COST and fallbacks to builtin_vectorization_cost with a direct usage of the scalar costs, which should also reduce compile time.