7702 refactor, remove CodeDelegationAccount abstraction by daniellehrner · Pull Request #8408 · hyperledger/besu

daniellehrner · 2025-03-12T10:12:00Z

PR description

Removes the CodeDelegationAbstraction to avoid calling Account.getCode() every time that we retrieve an account.

Fixed Issue(s)

fixes #8392

Thanks for sending a pull request! Have you done the following?

Checked out our contribution guidelines?
Considered documentation and added the doc-change-required label to this PR if updates are required.
Considered the changelog and included an update if required.
For database changes (e.g. KeyValueSegmentIdentifier) considered compatibility and performed forwards and backwards compatibility tests

Locally, you can run these tests to catch failures early:

spotless: ./gradlew spotlessApply
unit tests: ./gradlew build
acceptance tests: ./gradlew acceptanceTest
integration tests: ./gradlew integrationTest
reference tests: ./gradlew ethereum:referenceTests:referenceTests

Copilot

Pull Request Overview

This PR refactors code delegation handling by removing the CodeDelegationAccount abstraction and centralizing delegation logic into helper classes and services. Key changes include:

Removal of CodeDelegationAccount, MutableCodeDelegationDelegationAccount, and related files.
Introduction of CodeDelegationHelper and a simplified CodeDelegationService.
Updates to tests, protocol spec configurations, and transaction processing logic to use the new delegation approach.

Reviewed Changes

Copilot reviewed 14 out of 14 changed files in this pull request and generated no comments.

Show a summary per file

File	Description
evm/src/main/java/org/hyperledger/besu/evm/operation/InsufficientGasException.java	New exception class for signaling insufficient gas during delegation resolution.
evm/src/main/java/org/hyperledger/besu/evm/worldstate/CodeDelegationHelper.java	Introduces helper methods for code delegation resolution.
ethereum/core/src/main/java/org/hyperledger/besu/ethereum/mainnet/CodeDelegationProcessor.java	Refactored to use WorldUpdater and a newly instantiated CodeDelegationService.
evm/src/main/java/org/hyperledger/besu/evm/operation/AbstractCallOperation.java	Updates to delegate code resolution, integrating the new helper methods and exception handling.
Other files	Removal of legacy code delegation account abstractions and adjustments in tests and transaction processing.

Comments suppressed due to low confidence (1)

evm/src/main/java/org/hyperledger/besu/evm/operation/AbstractCallOperation.java:264

Consider replacing RuntimeException with a more specific exception type to clearly indicate the missing target account condition and improve downstream error handling.

throw new RuntimeException("A delegated code account must have a target account");

Signed-off-by: Daniel Lehrner <daniel.lehrner@consensys.net>

...um/core/src/main/java/org/hyperledger/besu/ethereum/mainnet/MainnetTransactionProcessor.java

ethereum/core/src/main/java/org/hyperledger/besu/ethereum/mainnet/MainnetProtocolSpecs.java

evm/src/main/java/org/hyperledger/besu/evm/worldstate/CodeDelegationHelper.java

siladu · 2025-03-20T01:14:30Z

evm/src/main/java/org/hyperledger/besu/evm/operation/AbstractCallOperation.java

+    if (maybeTargetAccount.isEmpty()) {
+      throw new RuntimeException("A delegated code account must have a target account");
+    }


I'm struggling to reconcile this error case with with the EIP spec or EELS - can you point out the relevant part of the spec please?

Also think we are doing it twice: here and in CodeDelegationGasCostHelper.codeDelegationGasCost

That is not part of the spec, just defensive programming. So whenever we get an optional we should check if it's empty before calling get()

Is this a condition where we could diverge from other client behaviour and it could lead to consensus issues? If there are error conditions missing should we update the spec?

I don't think so . Theoretically between line 259 where we call hasCodeDelegation() and between 266 where we call maybeTargetAccount.isEmpty() a bit could flip in memory and we are unable to recover the target account. I've just added it because AFAIK it best practive to always call isPresent() / !isEMpty() before getting an optional, but I can remove it if you think it's confusing.

I don't think we should code around hardware bugs. But I think the error conditions should match across client implementations, and ideally be part of the spec.
If this can truly never happen then maybe it shouldn't be Optional?

I've only just now realized that maybeTargetAccount isn't needed any further down the code and have completely removed it. It seems it was a leftover from a refactoring.

In any case I've removed the Optional from getTargetAccount to make it clear that it should only be used with accounts that have code delegation.

evm/src/main/java/org/hyperledger/besu/evm/operation/AbstractCallOperation.java

evm/src/main/java/org/hyperledger/besu/evm/operation/InsufficientGasException.java

...um/core/src/main/java/org/hyperledger/besu/ethereum/mainnet/MainnetTransactionProcessor.java

…the constructor Signed-off-by: Daniel Lehrner <daniel.lehrner@consensys.net>

…t of gas during code delegation resolution Signed-off-by: Daniel Lehrner <daniel.lehrner@consensys.net>

…action

Signed-off-by: daniellehrner <daniel.lehrner@consensys.net>

...um/core/src/main/java/org/hyperledger/besu/ethereum/mainnet/MainnetTransactionProcessor.java

matkt

Added comment but I will take more time tomorrow to check the code

evm/src/main/java/org/hyperledger/besu/evm/operation/AbstractCallOperation.java

evm/src/main/java/org/hyperledger/besu/evm/operation/AbstractExtCallOperation.java

ethereum/core/src/main/java/org/hyperledger/besu/ethereum/core/CodeDelegation.java

…rdered balance check for code delegation resolution, reordered recId check Signed-off-by: daniellehrner <daniel.lehrner@consensys.net>

…action

…too deep to be spec compliant Signed-off-by: daniellehrner <daniel.lehrner@consensys.net>

…action

Signed-off-by: daniellehrner <daniel.lehrner@consensys.net>

macfarla · 2025-03-28T06:08:03Z

running hive tests on this PR - docker image 25.3-develop-07da3dc
7702 all passing

locally ran some other randomly chosen EIPs, no failures

EIP7702
./hive --sim "ethereum/eest/consume-engine" --client "besu" --client-file="clients_pectra.yaml" --sim.buildarg "fixtures=https://github.com/ethereum/execution-spec-tests/releases/download/v4.1.0/fixtures_develop.tar.gz" --client.checktimelimit=60s --sim.parallelism=4 -sim.limit ".*tests/prague/eip7702.*"
...
Mar 28 13:41:31.457 INF simulation ethereum/eest/consume-engine finished suites=1 tests=476 failed=0
EIP6110
Mar 28 13:50:38.078 INF simulation ethereum/eest/consume-engine finished suites=1 tests=38 failed=0
EIP2935
Mar 28 15:54:20.039 INF simulation ethereum/eest/consume-engine finished suites=1 tests=19 failed=0 
EIP7516
➜  hive git:(master) ✗ ./hive --sim "ethereum/eest/consume-engine" --client "besu" --client-file="clients_pectra.yaml" --sim.buildarg "fixtures=https://github.com/ethereum/execution-spec-tests/releases/download/v4.1.0/fixtures_develop.tar.gz" --client.checktimelimit=60s --sim.parallelism=4 -sim.limit ".*tests/prague/eip7516.*"

macfarla · 2025-03-28T09:43:04Z

full run all passed

Mar 28 16:59:51.881 INF simulation ethereum/eest/consume-engine finished suites=1 tests=1742 failed=0
./hive --sim "ethereum/eest/consume-engine" --client "besu"  --sim.buildarg    7.53s user 20.81s system 1% cpu 34:33.76 total

…s with empty code hashes, remove unused methods from AccountState, reorder processCodeFromAccount to make it easier to read Signed-off-by: daniellehrner <daniel.lehrner@consensys.net>

…action

Signed-off-by: daniellehrner <daniel.lehrner@consensys.net>

…action

daniellehrner · 2025-03-30T08:23:23Z

I ran the Hive tests again after the latest changes, all are still passing:

./hive --sim "ethereum/eest/consume-engine" --client "besu" --client-file="./besu_local.yaml" --sim.buildarg fixtures=https://github.com/ethereum/execution-spec-tests/releases/download/v4.1.0/fixtures_develop.tar.gz --client.checktimelimit=120s --sim.parallelism=4 --sim.limit ".*fork_Prague.*"

ahamlat

From a performance perspective, it looks good to me.
@matkt Could you take a last look

ahamlat · 2025-03-31T09:57:42Z

...ore/src/test/java/org/hyperledger/besu/ethereum/mainnet/MainnetTransactionProcessorTest.java

        .thenReturn(ValidationResult.valid());
    when(worldState.getOrCreateSenderAccount(senderAddress)).thenReturn(senderAccount);
    when(worldState.get(toAddress.get())).thenReturn(receiverAccount);
+    when(receiverAccount.getCodeHash()).thenReturn(Hash.fromHexStringLenient("0x1"));


Why is this needed ?

The test assumes that messageCallProcessor.getCodeFromEVM() is called and prepares the mock for it in line 186. Therefor we need to return an non-empty code hash when receiverAccount.getCodeHash() is called.

evm/src/main/java/org/hyperledger/besu/evm/operation/AbstractCallOperation.java

evm/src/main/java/org/hyperledger/besu/evm/worldstate/CodeDelegationGasCostHelper.java

…action

Signed-off-by: daniellehrner <daniel.lehrner@consensys.net>

matkt

LGTM on my side

daniellehrner requested a review from Copilot March 12, 2025 10:12

Copilot AI reviewed Mar 12, 2025

View reviewed changes

daniellehrner force-pushed the feat/issue-8392/remove_code_delegation_abstraction branch 2 times, most recently from f8b0f59 to b7304b2 Compare March 14, 2025 12:40

daniellehrner marked this pull request as ready for review March 19, 2025 13:19

daniellehrner added 6 commits March 19, 2025 14:20

7702 refactor, remove CodeDelegationAccount abstraction

b34ae92

Signed-off-by: Daniel Lehrner <daniel.lehrner@consensys.net>

fix check if account has a code delegation

9f9c6ba

Signed-off-by: Daniel Lehrner <daniel.lehrner@consensys.net>

fix unit test

35d1b91

Signed-off-by: Daniel Lehrner <daniel.lehrner@consensys.net>

fix CodeDelegationProcessorTest

bf436bd

Signed-off-by: Daniel Lehrner <daniel.lehrner@consensys.net>

fix warming of target address, fix account retrieval

c5b34f8

Signed-off-by: Daniel Lehrner <daniel.lehrner@consensys.net>

bugfix: use target address for calculation of code resolution

1897206

Signed-off-by: Daniel Lehrner <daniel.lehrner@consensys.net>

daniellehrner force-pushed the feat/issue-8392/remove_code_delegation_abstraction branch from 2888729 to 1897206 Compare March 19, 2025 13:20

daniellehrner requested review from Copilot and removed request for Copilot March 19, 2025 13:24

siladu reviewed Mar 20, 2025

View reviewed changes

daniellehrner added 5 commits March 20, 2025 16:51

move CodeDelegationService from CodeDelegationProcessor.process() to …

5e8f614

…the constructor Signed-off-by: Daniel Lehrner <daniel.lehrner@consensys.net>

added test for CodeDelegationHelper, added debug logs when running ou…

06514c4

…t of gas during code delegation resolution Signed-off-by: Daniel Lehrner <daniel.lehrner@consensys.net>

Merge branch 'main' into feat/issue-8392/remove_code_delegation_abstr…

08fad2a

…action

Merge branch 'main' into feat/issue-8392/remove_code_delegation_abstr…

b37abd0

…action

remove Optional from getTargetAccount()

99e53e2

Signed-off-by: daniellehrner <daniel.lehrner@consensys.net>

ahamlat reviewed Mar 25, 2025

View reviewed changes

...um/core/src/main/java/org/hyperledger/besu/ethereum/mainnet/MainnetTransactionProcessor.java Show resolved Hide resolved

matkt reviewed Mar 25, 2025

View reviewed changes

evm/src/main/java/org/hyperledger/besu/evm/operation/AbstractCallOperation.java Show resolved Hide resolved

evm/src/main/java/org/hyperledger/besu/evm/operation/AbstractExtCallOperation.java Outdated Show resolved Hide resolved

ahamlat reviewed Mar 25, 2025

View reviewed changes

ethereum/core/src/main/java/org/hyperledger/besu/ethereum/core/CodeDelegation.java Outdated Show resolved Hide resolved

daniellehrner added 5 commits March 26, 2025 11:10

changed check for code delegation in MainnetTransactionProcessor, reo…

e069fe5

…rdered balance check for code delegation resolution, reordered recId check Signed-off-by: daniellehrner <daniel.lehrner@consensys.net>

Merge branch 'main' into feat/issue-8392/remove_code_delegation_abstr…

0f9f630

…action

code resolution gas must be deducted before we check for frame depth …

af7fe7e

…too deep to be spec compliant Signed-off-by: daniellehrner <daniel.lehrner@consensys.net>

Merge branch 'main' into feat/issue-8392/remove_code_delegation_abstr…

484d0bc

…action

add getTransactionType to mocks of transactions

8b5e8f1

Signed-off-by: daniellehrner <daniel.lehrner@consensys.net>

daniellehrner added 2 commits March 28, 2025 21:26

check code hash before getting code to return empty bytes for account…

02ce0bd

…s with empty code hashes, remove unused methods from AccountState, reorder processCodeFromAccount to make it easier to read Signed-off-by: daniellehrner <daniel.lehrner@consensys.net>

Merge branch 'main' into feat/issue-8392/remove_code_delegation_abstr…

2a3f946

…action

daniellehrner marked this pull request as draft March 28, 2025 20:28

daniellehrner added 2 commits March 29, 2025 13:23

check code hash for null before equals, fix test

007c8c6

Signed-off-by: daniellehrner <daniel.lehrner@consensys.net>

Merge branch 'main' into feat/issue-8392/remove_code_delegation_abstr…

42794f3

…action

daniellehrner marked this pull request as ready for review March 30, 2025 08:25

ahamlat reviewed Mar 31, 2025

View reviewed changes

matkt reviewed Apr 1, 2025

View reviewed changes

daniellehrner added 2 commits April 1, 2025 12:08

Merge branch 'main' into feat/issue-8392/remove_code_delegation_abstr…

eb7e452

…action

optimizations

f0ef3c6

Signed-off-by: daniellehrner <daniel.lehrner@consensys.net>

matkt approved these changes Apr 1, 2025

View reviewed changes

ahamlat approved these changes Apr 1, 2025

View reviewed changes

daniellehrner merged commit 5bb4e9e into hyperledger:main Apr 1, 2025
43 checks passed

daniellehrner deleted the feat/issue-8392/remove_code_delegation_abstraction branch April 1, 2025 15:12

Conversation

daniellehrner commented Mar 12, 2025

PR description

Fixed Issue(s)

Thanks for sending a pull request! Have you done the following?

Locally, you can run these tests to catch failures early:

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull Request Overview

Reviewed Changes

Uh oh!

Uh oh!

Uh oh!

Uh oh!

siladu Mar 20, 2025

Choose a reason for hiding this comment

Uh oh!

siladu Mar 20, 2025

Choose a reason for hiding this comment

Uh oh!

daniellehrner Mar 21, 2025

Choose a reason for hiding this comment

Uh oh!

siladu Mar 22, 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!

daniellehrner Mar 24, 2025

Choose a reason for hiding this comment

Uh oh!

siladu Mar 25, 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!

daniellehrner Mar 25, 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!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

matkt left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

macfarla commented Mar 28, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

macfarla commented Mar 28, 2025

Uh oh!

daniellehrner commented Mar 30, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

ahamlat left a comment • 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!

ahamlat Mar 31, 2025

Choose a reason for hiding this comment

Uh oh!

daniellehrner Apr 1, 2025

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

matkt left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

siladu Mar 22, 2025 •

edited

Loading

siladu Mar 25, 2025 •

edited

Loading

daniellehrner Mar 25, 2025 •

edited

Loading

macfarla commented Mar 28, 2025 •

edited

Loading

daniellehrner commented Mar 30, 2025 •

edited

Loading

ahamlat left a comment •

edited

Loading