bench: Add benchmark for block and Epoch Processing by guha-rahul · Pull Request #121 · ChainSafe/lodestar-z

guha-rahul · 2025-12-09T11:55:36Z

tracks #102

Adds benchmark for block and epoch transition
fix a bug in getAttestationDataSigningRoot
Unlock spec tests to test for proposer and signatures in state transition which was probably failing because of the above issue

gemini-code-assist · 2025-12-09T11:55:55Z

Summary of Changes

Hello @guha-rahul, I'm Gemini Code Assist¹! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!

This pull request significantly enhances the performance testing suite by introducing detailed benchmarks for both block and epoch processing within the state transition logic. These new benchmarks will help in identifying performance bottlenecks and optimizing critical operations. Additionally, it includes a crucial bug fix in the attestation data signing root calculation, which in turn allowed for the re-enabling of proposer and signature verification in the state transition spec tests, improving test coverage and reliability. The Fulu fork epoch configuration was also updated.

Highlights

Block Processing Benchmarks: Introduced new benchmarks for individual process_block operations, such as header, withdrawals, execution payload, randao, eth1 data, operations, and sync aggregate, to measure their performance.
Epoch Processing Benchmarks: Added comprehensive benchmarks for all process_epoch operations, including justification, inactivity updates, rewards, registry updates, slashings, and more, to evaluate epoch transition performance.
Bug Fix in Attestation Data Signing Root: Corrected an issue in getAttestationDataSigningRoot by changing computeEpochAtSlot to computeStartSlotAtEpoch for calculating the slot, ensuring accurate attestation data signing roots.
Enabled Spec Test Verifications: Unlocked and enabled verification for proposer and signatures in state transition spec tests, which were previously disabled due to the bug now fixed, improving test coverage and reliability.
Fulu Fork Epoch Update: Updated the FULU_FORK_EPOCH configuration for the mainnet chain to its correct value of 411392.

Using Gemini Code Assist

The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.

Invoking Gemini

You can request assistance from Gemini at any point by creating a comment using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands on the current page.

Feature	Command	Description
Code Review	`/gemini review`	Performs a code review for the current pull request in its current state.
Pull Request Summary	`/gemini summary`	Provides a summary of the current pull request in its current state.
Comment	@gemini-code-assist	Responds in comments when explicitly tagged, both in pull request comments and review comments.
Help	`/gemini help`	Displays a list of available commands.

Customization

To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.

Limitations & Feedback

Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for Github and other Google products, sign up here.

You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.

Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩

gemini-code-assist

Code Review

This pull request introduces new benchmarks for process_block and process_epoch operations within the state transition logic, specifically for the Fulu fork. These benchmarks, located in bench/state_transition/process_block.zig and bench/state_transition/process_epoch.zig, load real mainnet state and block data from SSZ files to measure the performance of individual block and epoch processing steps. A critical configuration update was made in src/config/chain/networks/mainnet.zig, setting FULU_FORK_EPOCH to 411392 to enable the Fulu fork on the mainnet, which is essential for the new benchmarks and Fulu-related logic. A critical bug fix was also implemented in src/state_transition/signature_sets/indexed_attestation.zig, changing computeEpochAtSlot to computeStartSlotAtEpoch to ensure the getDomain function receives the correct Slot type for attestation data signing root calculation. To enhance test coverage and correctness, verify_proposer and verify_signatures were enabled in the spec tests (test/spec/runner/transition.zig), validating the signature verification mechanisms, especially after the bug fix. The build system (build.zig and zbuild.zon) was updated to include these new benchmark executables and their respective test targets.

gemini-code-assist · 2025-12-09T11:58:57Z

    .ELECTRA_FORK_EPOCH = 364032,
    .FULU_FORK_VERSION = b(4, "0x06000000"),
-    .FULU_FORK_EPOCH = std.math.maxInt(u64),
+    .FULU_FORK_EPOCH = 411392,


Changing FULU_FORK_EPOCH from std.math.maxInt(u64) to a concrete value 411392 is a critical configuration update. This enables the Fulu fork on the mainnet configuration, which is essential for the new benchmarks and any Fulu-related logic to function correctly. This is a functional change, not a bug, and is necessary for the Fulu benchmarks to run correctly.

gemini-code-assist · 2025-12-09T11:58:57Z


 pub fn getAttestationDataSigningRoot(cached_state: *const CachedBeaconStateAllForks, data: *const AttestationData, out: *[32]u8) !void {
-    const slot = computeEpochAtSlot(data.target.epoch);
+    const slot = computeStartSlotAtEpoch(data.target.epoch);


This change from computeEpochAtSlot to computeStartSlotAtEpoch is a critical bug fix. The getDomain function expects a Slot type for its message_slot parameter, but data.target.epoch is an Epoch type. computeStartSlotAtEpoch correctly converts the epoch number to its starting slot number, ensuring the getDomain function receives the expected input. This directly addresses the bug mentioned in the PR description.

gemini-code-assist · 2025-12-09T11:58:57Z

+                        .verify_proposer = true,
+                        .verify_signatures = true,


Enabling verify_proposer and verify_signatures in the spec tests is a significant improvement for test coverage and correctness. This change ensures that the state transition logic is thoroughly validated, especially in light of the bug fix in getAttestationDataSigningRoot mentioned in the PR description. It indicates increased confidence in the correctness of the signature verification mechanisms.

guha-rahul · 2025-12-10T14:44:02Z

@spiral-ladder zbench has a character limit for benchmark names, So i was abbreviating some of the names of the benchmark names.

spiral-ladder · 2025-12-11T02:33:14Z

@spiral-ladder zbench has a character limit for benchmark names, So i was abbreviating some of the names of the benchmark names.

I see! Wasn't aware of the 22-character limit, thanks for letting me know.

spiral-ladder

Looking great - gave a pass through

spiral-ladder · 2025-12-12T02:56:54Z

+pub fn slotFromStateBytes(state_bytes: []const u8) ?Slot {
+    if (state_bytes.len < 48) return null;
+    return std.mem.readInt(u64, state_bytes[40..48], .little);
+}
+
+/// Read slot from raw SignedBeaconBlock SSZ bytes (offset 100)
+pub fn slotFromBlockBytes(block_bytes: []const u8) ?Slot {
+    if (block_bytes.len < 108) return null;
+    return std.mem.readInt(u64, block_bytes[100..108], .little);
+}


2 points here:

I think given that these are only utils used in benchmarking we should just keep them in bench/state_transition rather than put them in config/fork.zig. We can move these functions if we ever use them outside of benchmarking.

I think we can assume we're always going to be loading valid bytes here and if not we should be error-ing instead anyway:

Suggested change

pub fn slotFromStateBytes(state_bytes: []const u8) ?Slot {

if (state_bytes.len < 48) return null;

return std.mem.readInt(u64, state_bytes[40..48], .little);

}

/// Read slot from raw SignedBeaconBlock SSZ bytes (offset 100)

pub fn slotFromBlockBytes(block_bytes: []const u8) ?Slot {

if (block_bytes.len < 108) return null;

return std.mem.readInt(u64, block_bytes[100..108], .little);

}

/// Read slot from raw BeaconState SSZ bytes (offset 40)

pub fn slotFromStateBytes(state_bytes: []const u8) Slot {

std.debug.assert(state_bytes.len >= 48);

return std.mem.readInt(u64, state_bytes[40..48], .little);

}

/// Read slot from raw SignedBeaconBlock SSZ bytes (offset 100)

pub fn slotFromBlockBytes(block_bytes: []const u8) Slot {

std.debug.assert(block_bytes.len >= 108);

return std.mem.readInt(u64, block_bytes[100..108], .little);

}

moved these and the loadState and loadBlock to a utils file inside bench

spiral-ladder · 2025-12-12T07:38:21Z

+    // Actual processEpoch function
+    try bench.addParam("process_epoch", &ProcessEpochBench{ .cached_state = cached_state }, .{});


What does "Actual processEpoch function" mean?

I meant the non segmented one. Is "continuous" a better name or something more descriptive?

I don't think we need a special name, just non-segmented vs segmented will do.

spiral-ladder · 2026-01-06T04:32:59Z

+    // Actual processEpoch function
+    try bench.addParam("process_epoch", &ProcessEpochBench{ .cached_state = cached_state }, .{});


I don't think we need a special name, just non-segmented vs segmented will do.

spiral-ladder

Thanks again for the cleanup! Just a few more loose ends. There's still some things to cleanup but we can leave those for a followup PR.

spiral-ladder · 2026-01-08T06:20:41Z

+        const cloned = self.cached_state.clone(allocator) catch unreachable;
+        defer {
+            cloned.deinit();
+            allocator.destroy(cloned);
+        }


We can save this for a future PR, but we should probably move this clone (and other clones) out of the run definition since cloning should not be part of the cost of benchmarking processXXX

Created #164

spiral-ladder

last nits

spiral-ladder

LGTM, thanks for the PR!

guha-rahul added 4 commits December 9, 2025 16:30

fix slot calculation

fe250d5

add block processing benchmark

205aba2

unblock state transition spec test

a44f0e0

add fulu epoch processing benchmark

698a6e4

gemini-code-assist Bot reviewed Dec 9, 2025

View reviewed changes

spiral-ladder reviewed Dec 10, 2025

View reviewed changes

guha-rahul added 3 commits December 11, 2025 16:08

add fork awareness

5324760

add segment bench and chores

e09582c

add segmented epoch bench and chores

22b3ca9

guha-rahul requested a review from spiral-ladder December 11, 2025 11:09

spiral-ladder reviewed Dec 12, 2025

View reviewed changes

guha-rahul added 2 commits December 12, 2025 15:23

chores: fix comments

b7dda0e

refactor: move slot calculation and loading functions to utils

c5486df

wemeetagain mentioned this pull request Jan 5, 2026

feat: add era reader/writer #149

Merged

spiral-ladder reviewed Jan 6, 2026

View reviewed changes

Comment thread bench/state_transition/utils.zig Outdated

Comment thread bench/state_transition/utils.zig Outdated

Comment thread bench/state_transition/process_block.zig Outdated

guha-rahul added 3 commits January 7, 2026 18:22

add BenchOpts for cleaner branching

b30b230

minor nits

1e84d09

Merge remote-tracking branch 'origin/main' into bench_block

dc3afac

guha-rahul requested a review from a team as a code owner January 7, 2026 13:22

Resolve conflicts from config module refactoring

3eae3bc

guha-rahul requested a review from spiral-ladder January 7, 2026 13:42

spiral-ladder requested changes Jan 8, 2026

View reviewed changes

spiral-ladder reviewed Jan 8, 2026

View reviewed changes

Comment thread src/config/fork_seq.zig Outdated

guha-rahul added 3 commits January 8, 2026 21:13

nits

ff2b542

zbuild sync

c459ec7

zbuild sync

768f6b2

spiral-ladder reviewed Jan 9, 2026

View reviewed changes

Comment thread bench/state_transition/process_block.zig Outdated

Comment thread bench/state_transition/process_epoch.zig Outdated

Comment thread bench/state_transition/process_epoch.zig Outdated

small nits

e3133a1

spiral-ladder mentioned this pull request Jan 9, 2026

move clones out of benchmarking fns #164

Closed

wemeetagain approved these changes Jan 9, 2026

View reviewed changes

spiral-ladder self-requested a review January 10, 2026 04:44

spiral-ladder approved these changes Jan 10, 2026

View reviewed changes

spiral-ladder merged commit f7e1f21 into ChainSafe:main Jan 10, 2026
8 checks passed

		// Actual processEpoch function
		try bench.addParam("process_epoch", &ProcessEpochBench{ .cached_state = cached_state }, .{});

Uh oh!

Conversation

guha-rahul commented Dec 9, 2025

Uh oh!

gemini-code-assist Bot commented Dec 9, 2025

Summary of Changes

Highlights

Footnotes

Uh oh!

gemini-code-assist Bot left a comment

Choose a reason for hiding this comment

Code Review

Uh oh!

gemini-code-assist Bot Dec 9, 2025

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist Bot Dec 9, 2025

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist Bot Dec 9, 2025

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

guha-rahul commented Dec 10, 2025

Uh oh!

spiral-ladder commented Dec 11, 2025

Uh oh!

spiral-ladder left a comment

Choose a reason for hiding this comment

Uh oh!

spiral-ladder Dec 12, 2025

Choose a reason for hiding this comment

Uh oh!

guha-rahul Dec 12, 2025

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

spiral-ladder Dec 12, 2025

Choose a reason for hiding this comment

Uh oh!

guha-rahul Dec 12, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

spiral-ladder Jan 6, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

spiral-ladder Jan 6, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

spiral-ladder left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

spiral-ladder Jan 8, 2026

Choose a reason for hiding this comment

Uh oh!

spiral-ladder Jan 9, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

guha-rahul Dec 12, 2025 •

edited

Loading