perf(decode + encode-greedy): close 3-5× donor gap on negative-level decompress; share SIMD primitives + add dedicated greedy strategy

[polaz]

Status (re-measured)

The original 3-5× negative-level decode gap is closed to 1.3-1.4×; the missing dedicated greedy strategy has landed. This issue now tracks the residual negative-level decode parity work only.

Current numbers (i9, decodecorpus-z000033, level_-1_fast):

Stream source	Rust decode	C decode	Gap (was)
c_stream	907 µs (1.05 GiB/s)	647 µs (1.47 GiB/s)	1.40× (was 3.69×)
rust_stream	1.18 ms (824 MiB/s)	895 µs (1.06 GiB/s)	1.32× (was 5.35×)

Landed since the original report

• Greedy strategy: dedicated StrategyTag::Greedy (L5) with lazy_depth = 0 on the Row finder — the reference's own greedy shape (its greedy/lazy share the row-search template with depth 0). The L4 dfast outlier was separately closed by the donor greedy double-fast port.
• XXH64 frame checksum: gated on the frame's checksum flag (−61% on flag-off frames) and hashed per-block while cache-hot on the direct path; no longer a post-decode cold walk.
• Per-kernel decoders: four ISA tiers (Scalar/BMI2/AVX2/VBMI2) with monolithic per-tier sequence loops, BMI2 bit-reader specialization, per-tier match-copy chains.
• SIMD wildcopy / overshooting copies: donor-shape overshoot-tolerant copies with bounded tails per kernel.

Residual (the 1.3-1.4×)

The remaining gap is the known sequence-decode body delta vs the reference's single bmi2.constprop monolith (HUF 4-stream burst on literal-heavy frames being the biggest single item on weak-compression fixtures). Tracked levers, in ROI order:

1. HUF burst decompress port (one inlined monolith vs our 3-fn split) — order-of-magnitude self-time delta on literal-heavy frames.
2. Sequence-loop body instruction diff vs the reference per-tier.

Kill-switch criteria

Stop pulling individual levers when both stream sources sit within ~1.1× of the reference on the negative-level corpus, or when a lever returns <2% twice in a row (record the negative result and move on).

[coderabbitai]

🔗 Related PRs

#92 - perf(decoding): SIMD HUF kernels with runtime dispatch [merged]
#179 - perf(encoder): donor-parity greedy parse at L4 — ratio + speed win [merged]
#194 - perf(decode): fused sequence executor + SIMD/FSE hot-path cleanup + DoS-safe rollback [merged]
#213 - perf(decode): SIMD-16 fast path for short offsets {1, 2, 4} [merged]
#238 - perf(decode): collapse HUF 4-stream burst-gate to single-cursor olimit [merged]
#269 - perf(decode): route short-block fallback through inline executor (z000033 −25%) [merged]
#278 - perf(encoding): HUF_flags_preferRepeat for fast strategies + small literals (#23 G6) [merged]
#331 - perf(encode): close level-5 greedy speed gap (row match-span cap) [merged]
#377 - perf(decode): cut HUF/FSE entropy-build overhead on the decode hot path [merged]

---

📝 Issue Planner

_{Check the box below or use the @coderabbitai plan command to generate an implementation plan and prompts that you can use with your favorite coding assistant.}

• Create Plan

---

🧪 Issue enrichment is currently in open beta.

You can configure auto-planning by selecting labels in the issue_enrichment configuration.

To disable automatic issue enrichment, add the following to your .coderabbit.yaml:

issue_enrichment:
  auto_enrich:
    enabled: false

💬 Have feedback or questions? Drop into our discord!