Skip to content

release: v0.6.0#100

Merged
avrabe merged 1 commit into
mainfrom
release/v0.6.0
May 12, 2026
Merged

release: v0.6.0#100
avrabe merged 1 commit into
mainfrom
release/v0.6.0

Conversation

@avrabe
Copy link
Copy Markdown
Contributor

@avrabe avrabe commented May 12, 2026

Summary

Bump workspace version 0.5.0 → 0.6.0.

This release is driven entirely by real-world findings on production gale code (Verus-verified kernel-scheduler FFI wasm): closing v0.5.0's +6.3% CSE size regression, lifting two early-exit guards that made LOOM a no-op on kernel-style code, and fixing a Z3 panic that blocked the inline pass on i64-heavy modules.

Merged PRs

PR Title
#94 fix(cse): cost-aware dedup gate eliminates gale +6.3% regression
#95 feat(opt): eliminate_dead_locals — drop write-only locals (-0.86% on gale)
#96 feat(opt): eliminate_dead_stores — full backward liveness on structured wasm
#99 fix(verify): correct i64 local widths in Z3 encoder + vacuum const+drop peephole (closes #98)

Measured impact

gale_ffi (1.9 KB kernel FFI):

Build Code section Δ vs baseline
baseline 811 B
v0.5.0 (regression) 862 B +6.3%
v0.6.0 804 B -0.86%

A 6.7-point swing on real kernel-scheduler code.

calculator.wasm (2.3 MB component):

  • --passes dead-stores alone: 2,337,724 → 2,327,794 bytes (-0.4%, ~10 KB)

Test count

247 → 317 tests (+70 across all v0.6.0 PRs)

CHANGELOG

Detailed v0.6.0 entry already in CHANGELOG.md (landed via #99).

Notes

🤖 Generated with Claude Code

Bump workspace version 0.5.0 → 0.6.0.

This release is driven entirely by real-world findings on production
gale code (Verus-verified kernel-scheduler FFI wasm): closing v0.5.0's
+6.3% CSE size regression, lifting two early-exit guards that made
LOOM a no-op on kernel-style code, and fixing a Z3 panic that blocked
the inline pass on i64-heavy modules.

Merged PRs:
  #94 fix(cse):  cost-aware dedup gate eliminates gale +6.3% regression
  #95 feat(opt): eliminate_dead_locals (drop write-only locals)
  #96 feat(opt): eliminate_dead_stores (full backward liveness)
  #99 fix(verify): i64 widths in Z3 encoder + vacuum const+drop peephole (closes #98)

Measured impact on gale_ffi (1.9 KB kernel FFI):
  baseline:            code section 811 bytes
  v0.5.0 (regression): code section 862 bytes (+6.3%)
  v0.6.0 (this):       code section 804 bytes (-0.86%)

Measured impact on calculator.wasm (2.3 MB component):
  --passes dead-stores alone: -0.4% (~10 KB)

Test count: 247 → 317 (+70)
@avrabe avrabe merged commit 6fe5c1f into main May 12, 2026
12 of 18 checks passed
@avrabe avrabe deleted the release/v0.6.0 branch May 12, 2026 05:14
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Z3 SortDiffers panic in inline_functions pass on i64-heavy wasm modules (every function reverts)

1 participant