perf(levm): use bitmap for jumpdests

[ilitteri]

Motivation

The original is_blacklisted function in JumpTargetFilter was a performance bottleneck, consuming ~30% of execution time during mainnet blocks stateless execution.

Description

Inspired by Geth's codeBitmap approach, this PR:

1. Replaces incremental iteration with a one-time precompute method that scans the entire bytecode, building a BitVec<u8, Msb0> where bits mark valid JUMPDEST positions, skipping PUSH1..PUSH32 data bytes.
2. Updates is_blacklisted to O(1) bit lookup: address >= bytecode.len() || !jumpdests[address].

[github-actions]

Lines of code report

Total lines added: 1
Total lines removed: 0
Total lines changed: 1

Detailed view

+------------------------------------+-------+------+
| File                               | Lines | Diff |
+------------------------------------+-------+------+
| ethrex/crates/vm/levm/src/utils.rs | 427   | +1   |
+------------------------------------+-------+------+

[github-actions]

Benchmark Results Comparison

No significant difference was registered for any benchmark run.

Detailed Results

Benchmark Results: BubbleSort

Command	Mean [s]	Min [s]	Max [s]	Relative
main_revm_BubbleSort	4.822 ± 0.073	4.778	5.018	1.05 ± 0.02
main_levm_BubbleSort	4.603 ± 0.009	4.587	4.615	1.00
pr_revm_BubbleSort	4.777 ± 0.023	4.760	4.838	1.04 ± 0.01
pr_levm_BubbleSort	4.633 ± 0.049	4.588	4.727	1.01 ± 0.01

Benchmark Results: ERC20Approval

Command	Mean [s]	Min [s]	Max [s]	Relative
main_revm_ERC20Approval	1.554 ± 0.009	1.543	1.566	1.01 ± 0.01
main_levm_ERC20Approval	1.655 ± 0.020	1.638	1.706	1.08 ± 0.01
pr_revm_ERC20Approval	1.539 ± 0.005	1.533	1.550	1.00
pr_levm_ERC20Approval	1.642 ± 0.013	1.622	1.665	1.07 ± 0.01

Benchmark Results: ERC20Mint

Command	Mean [ms]	Min [ms]	Max [ms]	Relative
main_revm_ERC20Mint	185.1 ± 0.7	184.2	186.5	1.01 ± 0.01
main_levm_ERC20Mint	196.1 ± 1.4	194.4	199.0	1.07 ± 0.01
pr_revm_ERC20Mint	183.6 ± 1.1	182.6	185.7	1.00
pr_levm_ERC20Mint	197.5 ± 4.4	195.2	210.0	1.08 ± 0.02

Benchmark Results: ERC20Transfer

Command	Mean [ms]	Min [ms]	Max [ms]	Relative
main_revm_ERC20Transfer	351.6 ± 1.3	350.4	354.9	1.00
main_levm_ERC20Transfer	387.7 ± 5.7	381.3	400.8	1.10 ± 0.02
pr_revm_ERC20Transfer	352.6 ± 5.6	348.8	363.1	1.00 ± 0.02
pr_levm_ERC20Transfer	385.5 ± 3.4	381.7	393.2	1.10 ± 0.01

Benchmark Results: Factorial

Command	Mean [ms]	Min [ms]	Max [ms]	Relative
main_revm_Factorial	237.4 ± 0.7	235.8	238.3	1.00
main_levm_Factorial	277.2 ± 3.1	274.7	285.4	1.17 ± 0.01
pr_revm_Factorial	238.4 ± 4.3	235.7	250.6	1.00 ± 0.02
pr_levm_Factorial	279.5 ± 6.5	275.1	297.3	1.18 ± 0.03

Benchmark Results: FactorialRecursive

Command	Mean [s]	Min [s]	Max [s]	Relative
main_revm_FactorialRecursive	1.609 ± 0.021	1.573	1.636	1.01 ± 0.02
main_levm_FactorialRecursive	8.566 ± 0.035	8.484	8.599	5.37 ± 0.09
pr_revm_FactorialRecursive	1.596 ± 0.027	1.526	1.623	1.00
pr_levm_FactorialRecursive	8.523 ± 0.073	8.429	8.620	5.34 ± 0.10

Benchmark Results: Fibonacci

Command	Mean [ms]	Min [ms]	Max [ms]	Relative
main_revm_Fibonacci	216.2 ± 1.9	211.2	217.8	1.00
main_levm_Fibonacci	271.1 ± 2.0	269.2	275.9	1.25 ± 0.01
pr_revm_Fibonacci	216.6 ± 2.7	213.6	223.4	1.00 ± 0.02
pr_levm_Fibonacci	268.1 ± 4.4	259.1	274.0	1.24 ± 0.02

Benchmark Results: FibonacciRecursive

Command	Mean [ms]	Min [ms]	Max [ms]	Relative
main_revm_FibonacciRecursive	856.5 ± 12.6	832.2	872.1	1.00
main_levm_FibonacciRecursive	939.3 ± 8.1	932.9	956.1	1.10 ± 0.02
pr_revm_FibonacciRecursive	860.9 ± 6.5	847.3	869.7	1.01 ± 0.02
pr_levm_FibonacciRecursive	928.1 ± 10.6	918.3	950.8	1.08 ± 0.02

Benchmark Results: ManyHashes

Command	Mean [ms]	Min [ms]	Max [ms]	Relative
main_revm_ManyHashes	12.3 ± 0.0	12.2	12.4	1.00
main_levm_ManyHashes	13.6 ± 0.1	13.5	13.9	1.11 ± 0.01
pr_revm_ManyHashes	12.3 ± 0.1	12.2	12.6	1.00 ± 0.01
pr_levm_ManyHashes	13.6 ± 0.0	13.5	13.7	1.11 ± 0.01

Benchmark Results: MstoreBench

Command	Mean [ms]	Min [ms]	Max [ms]	Relative
main_revm_MstoreBench	273.2 ± 2.2	270.8	277.5	1.00
main_levm_MstoreBench	329.1 ± 0.8	328.0	330.4	1.20 ± 0.01
pr_revm_MstoreBench	273.8 ± 3.1	270.0	279.3	1.00 ± 0.01
pr_levm_MstoreBench	331.1 ± 4.4	327.6	342.9	1.21 ± 0.02

Benchmark Results: Push

Command	Mean [ms]	Min [ms]	Max [ms]	Relative
main_revm_Push	301.2 ± 5.1	297.9	315.7	1.00 ± 0.02
main_levm_Push	381.7 ± 4.6	373.6	389.8	1.27 ± 0.02
pr_revm_Push	299.9 ± 0.9	298.5	301.4	1.00
pr_levm_Push	380.0 ± 3.7	375.9	386.1	1.27 ± 0.01

Benchmark Results: SstoreBench_no_opt

Command	Mean [ms]	Min [ms]	Max [ms]	Relative
main_revm_SstoreBench_no_opt	217.1 ± 0.8	216.4	219.0	2.43 ± 0.08
main_levm_SstoreBench_no_opt	90.5 ± 3.5	86.6	97.8	1.01 ± 0.05
pr_revm_SstoreBench_no_opt	217.2 ± 0.4	216.6	217.7	2.43 ± 0.08
pr_levm_SstoreBench_no_opt	89.5 ± 2.8	86.6	96.3	1.00

[github-actions]

Benchmark Block Execution Results Comparison Against Main

Command	Mean [s]	Min [s]	Max [s]	Relative
base	87.399 ± 0.232	86.977	87.748	1.05 ± 0.01
head	82.894 ± 0.371	82.445	83.472	1.00

[Copilot]

Pull Request Overview

This PR optimizes the JumpTargetFilter implementation to address a performance bottleneck where the original is_blacklisted function consumed ~30% of execution time during mainnet block execution.

• Replaces incremental iteration with precomputed bitmap approach using BitVec<u8, Msb0>
• Changes is_blacklisted from potentially O(n) to O(1) lookup operation
• Adopts Geth's codeBitmap strategy for identifying valid JUMPDEST positions

Reviewed Changes

Copilot reviewed 2 out of 3 changed files in this pull request and generated 2 comments.

File	Description
crates/vm/levm/src/utils.rs	Refactors JumpTargetFilter to use bitmap-based jumpdest tracking instead of incremental filtering
crates/vm/levm/Cargo.toml	Adds bitvec dependency for bitmap functionality

---

_{Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.}

[Copilot]

The skip calculation is incorrect. For PUSH1 through PUSH32, the number of data bytes to skip should be opcode as usize - Opcode::PUSH1 as usize + 1. Currently, using PUSH0 as the base means PUSH1 skips 1 byte, PUSH2 skips 2 bytes, etc., but it should skip 1, 2, 3... up to 32 bytes respectively.

Suggested change

	let skip = opcode as usize - Opcode::PUSH0 as usize;
	let skip = opcode as usize - Opcode::PUSH1 as usize + 1;

[JereSalo]

Why not just check if it's a jumpdest rather than checking that it's not blacklisted now that we have a bitvec with jumpdests? It feels more straightforward and less complex but I don't know if the other approach has more benefits

[edg-l]

old gas bench looks good