Use direct blst Pippenger for BLS MSM

[decofe]

Summary

• replace blst's Rust MultiPoint::mult wrapper with direct blst_p1s/blst_p2s_mult_pippenger calls for BLS12-381 MSM
• keep the existing single-point scalar multiplication fast path
• allocate scratch with blst's scratch sizing helpers and pass null-terminated pointer batches expected by the C API

Tests

• cargo fmt -p revm-precompile
• cargo test -p revm-precompile bls12_381 --features blst
• cargo test -p revm-precompile --features blst
• cargo bench -p revm-precompile --features blst --bench bench --no-run

Prompted by: @Rjected

[Rjected]

only marking ready to see if codspeed runs

[codspeed-hq]

Merging this PR will not alter performance

⚠️ Different runtime environments detected

Some benchmarks with significant performance changes were compared across different runtime environments,
which may affect the accuracy of the results.

Open the report in CodSpeed to investigate

⚡ 1 improved benchmark
❌ 1 regressed benchmark
✅ 179 untouched benchmarks
⏩ 1 skipped benchmark¹

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
❌	Simulation	g2_msm (size 2)	2.7 ms	2.8 ms	-5.37%
⚡	Simulation	g1_msm (size 2)	1.4 ms	1.3 ms	+7.15%

Tip

Investigate this regression by commenting @codspeedbot fix this regression on this PR, or directly use the CodSpeed MCP with your agent.

---

_{Comparing dan/direct-blst-pippenger-g1msm (af36d41) with main (b31a3d5)}

Footnotes

1. 1 benchmark was skipped, so the baseline result was used instead. If it was deleted from the codebase, click here and archive it to remove it from the performance reports. ↩