26311028863
88%

Ran 22 May 2026 08:51PM UTC

Jobs 1

Files 70

Run time 1min

Badge

Embed ▾

Committed 22 May 2026 08:43PM UTC coverage: 87.985%. Remained the same

Build # 26311028863

Build Type

push

github

Committed by

web-flow

Commit Message

Fix benchmarks to exclude allocator-reset setup time from measurements (#789)

* Fix benchmarks to exclude allocator-reset setup time from measurements

The deserialize and deserialize_2026 benchmarks were calling
a.restore_checkpoint() inside b.iter(), so each measurement included
the checkpoint-reset overhead.  All four benchmark files also used a
Instant::now()/start.elapsed() pattern inside b.iter() that criterion
ignores entirely (it measures wall-clock time of the whole closure).

Fixes:
- Use b.iter_custom() for benchmarks that require per-iteration allocator
  reset: restore_checkpoint() runs before each Instant::now(), so only
  the actual deserialization is timed
- Use b.iter(|| black_box(...)) for benchmarks with no per-iter setup
  (serialize_2026, intern, serialized_length, tree_hash_from_stream)
- Remove the now-redundant `use std::time::Instant` imports from files
  that no longer need manual timing

Co-authored-by: Cursor <cursoragent@cursor.com>

Coverage Stats

8231 of 9355 relevant lines covered (87.99%)

21894018.8 hits per line

Jobs

ID	Job ID	Ran	Files	Coverage
1	26311028863.1	22 May 2026 08:51PM UTC	70	87.99	GitHub Action Run

Chia-Network / clvm_rs / 26311028863
88%

README BADGES
x

Markdown

Textile

RDoc

HTML

Rst

Jobs

Source Files on build 26311028863

Chia-Network / clvm_rs / 26311028863 88%

README BADGES x

Markdown

Textile

RDoc

HTML

Rst

Jobs

Source Files on build 26311028863

Chia-Network / clvm_rs / 26311028863
88%

README BADGES
x