Implement EIP-5920 - PAY opcode by lu-pinto · Pull Request #8498 · besu-eth/besu

lu-pinto · 2025-03-31T17:26:56Z

PR description

Implement new PAY opcode

Fixed Issue(s)

fixes #8438

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

jochem-brouwer

Saw this on discord, did a quick skim-through of this :)

jochem-brouwer · 2025-03-31T23:09:42Z

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

+    final boolean accountIsWarm = frame.warmUpAddress(to);
+
+    if (frame.isStatic() && hasValue && !Objects.equals(frame.getSenderAddress(), to)) {
+      return new OperationResult(cost(to, true, recipient, true), ExceptionalHaltReason.ILLEGAL_STATE_CHANGE);


EIP will need clarification for this, but if you use the CALL opcode and you send value (also to self) in a static context, this is not allowed and will throw.

Yes but isn't it what StaticCall actually is? A Call with no value that does not change any state. But EIP should be clarified for sure

One can argue that static calls don't do logs though so you be correct

Yes but isn't it what StaticCall actually is? A Call with no value that does not change any state. But EIP should be clarified for sure

Note that I'm talking about a PAY with value, but the PAY is actually to the current address. So the state does not change. The EIP will need clarification for this. But, note for the CALL opcode that if value is being sent in a STATICCALL context, the opcode will exceptionally abort (so even though no state is changed if it would be executed).

I will open a PR for this against the EIP.

jochem-brouwer · 2025-03-31T23:10:39Z

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

+
+    final Bytes toAddressBytes = frame.getStackItem(1);
+    if (toAddressBytes.size() > 20
+      && toAddressBytes.numberOfLeadingZeroBytes() < 12) {


Does the stack trim zeros by default? If not, what if I push 21 bytes to the stack with a leading 0 and an address of 20 nonzero bytes? This throws, but should be valid.

PUSH0 PUSH21 00<20 nonzero bytes> PAY maybe?

jochem-brouwer · 2025-03-31T23:15:46Z

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

+    }
+
+    final Address to = Words.toAddress(toAddressBytes);
+    final Wei value = Wei.wrap(frame.getStackItem(0));


I believe the stack is [addr, val, ...] and it seems to be implemented [val, addr, ...] here. So stack[0] should be val and stack[1] addr.

jochem-brouwer · 2025-03-31T23:17:42Z

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

+    }
+    if (accountIsWarm
+      //TODO: what about precompile accounts?
+      || gasCalculator().isPrecompile(to)) {


Precompiles are added to warm addresses by default

https://eips.ethereum.org/EIPS/eip-2929#parameters

Ok, thanks good to know. This is just the way Besu does it not sure why but in the end it's the same thing.

lu-pinto · 2025-04-02T15:10:59Z

Saw this on discord, did a quick skim-through of this :)

Thanks for the review, fixes a few things based on that

github-actions · 2025-05-04T02:14:46Z

This pr is stale because it has been open for 30 days with no activity.

macfarla

@lu-pinto what do we have to change to make this independent of EOF?

lu-pinto · 2025-05-12T15:26:32Z

@lu-pinto what do we have to change to make this independent of EOF?

Just adding it at the right hardfork configuration and aligning the return value with current EVM opcodes (*CALL).
There are also some changes happening to PAY as we speak that haven't been merge which I haven't had the time to check: ethereum/EIPs#9590

Signed-off-by: Luis Pinto <luis.pinto@consensys.net>

macfarla added the Fusaka label Mar 31, 2025

jochem-brouwer reviewed Mar 31, 2025

View reviewed changes

lu-pinto force-pushed the implement-PAY-opcode branch from cd7bd61 to 8a9e838 Compare April 2, 2025 15:10

jflo added this to Osaka May 1, 2025

jflo moved this to Todo in Osaka May 1, 2025

jflo moved this from Todo to In Progress in Osaka May 1, 2025

github-actions bot added the Stale label May 4, 2025

macfarla reviewed May 12, 2025

View reviewed changes

github-actions bot removed the Stale label May 13, 2025

lu-pinto force-pushed the implement-PAY-opcode branch 2 times, most recently from 76adc11 to 7c9b02b Compare May 15, 2025 09:31

macfarla added Osaka Osaka fork related - part of Fusaka and removed Fusaka labels May 20, 2025

lu-pinto force-pushed the implement-PAY-opcode branch from d26b61b to 42876ba Compare May 20, 2025 15:40

lu-pinto marked this pull request as ready for review May 20, 2025 15:40

lu-pinto added 8 commits May 21, 2025 17:26

Implement EIP-5920 - PAY opcode

1f345a3

Signed-off-by: Luis Pinto <luis.pinto@consensys.net>

static calls, stack positioning and address too big fixes

642d7d9

Signed-off-by: Luis Pinto <luis.pinto@consensys.net>

migrate PAY opcode out of EOF - other updates

2958ef0

Signed-off-by: Luis Pinto <luis.pinto@consensys.net>

spotless

fc70b45

Signed-off-by: Luis Pinto <luis.pinto@consensys.net>

javadoc

928f5ed

Signed-off-by: Luis Pinto <luis.pinto@consensys.net>

Add evmtool test and missing EVM definition for Osaka

2ace2f6

Signed-off-by: Luis Pinto <luis.pinto@consensys.net>

fix pay-opcode.json whitespace

e25d9c6

Signed-off-by: Luis Pinto <luis.pinto@consensys.net>

spotless

8056897

Signed-off-by: Luis Pinto <luis.pinto@consensys.net>

lu-pinto force-pushed the implement-PAY-opcode branch from 91e61a3 to 8056897 Compare May 21, 2025 16:29

jflo approved these changes May 21, 2025

View reviewed changes

lu-pinto merged commit 731e940 into besu-eth:main May 21, 2025
48 checks passed

github-project-automation bot moved this from In Progress to Done in Osaka May 21, 2025

lu-pinto self-assigned this Dec 12, 2025

Conversation

lu-pinto commented Mar 31, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

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!

jochem-brouwer left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

lu-pinto commented Apr 2, 2025

Uh oh!

github-actions bot commented May 4, 2025

Uh oh!

macfarla left a comment

Choose a reason for hiding this comment

Uh oh!

lu-pinto commented May 12, 2025

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

lu-pinto commented Mar 31, 2025 •

edited

Loading