Landing: e5912e7dc1e3
Project / Subsystem
gcc / gcc/sched-deps
Date
2026-07-27
Author
Kyrylo Tkachov
Commit
e5912e7dc1e3edb72f43624355a7e363cd76f0b3
Source
github
Perf win
Yes
Breaking
No
All attributes
- project
- gcc
- subsystem
- gcc/sched-deps
- patch_id
- —
- commit_hash
- e5912e7dc1e3edb72f43624355a7e363cd76f0b3
- source_type
- github
- headline
- Improves scheduler performance by recording barriers once.
- tldr
- Reduces scheduler analysis time by avoiding redundant barrier recording.
- author
- Kyrylo Tkachov
- outcome
- committed
- performance_win
- true
- breaking_change
- false
- series_id
- —
- series_parts
- []
- tags
-
- • optimization
- • scheduler
- • compile-time
- • gcc
- discussion_id_link
- —
- bugzilla_pr
- —
- date
- 2026-07-27T00:00:00.000Z
The GCC scheduler’s dependency analysis, specifically how it records instruction barriers before register reload, has been optimized. Previously, barriers were recorded for every register entry, leading to potentially quadratic time complexity in large functions. This change modifies the analysis to record barriers only once into a pending list. These pending barriers are then materialized when a register is first used, significantly speeding up the analysis phase. This is a compile-time optimization with no impact on generated code.