Merge Transfer and TransferWithGas gadget #1777

ChihChengLiang · 2024-02-21T19:14:39Z

Description

Transfer and TransferWithGas gadgets are mostly the same. We can merge them and duplicate them.
There are duplicated logics for creating the receiver account (writing empty hash). We remove the duplicated one.
The update balance gadget is vague, it is hard to understand whether we want to add or subtract balance. We extract a helpful method for that.
deduplicate the transfer_to logic in busmapping.
Fix the receiver creation condition on busmapping.
Fix the order of the sender account read-write in begin tx assignment.

Thank @curryrasul for initializing this conversation.

Issue Link

Type of change

Refactor (no updates to logic)

Content

Test

cargo test -p zkevm-circuits create
cargo test -p zkevm-circuits begin_tx
cargo test -p zkevm-circuits end_tx

ChihChengLiang · 2024-02-22T17:48:36Z

bus-mapping/src/circuit_input_builder/input_state_ref.rs

        value: Word,
        reversible: bool,
    ) -> Result<(), Error> {
-        // If receiver doesn't exist, create it
-        if (!receiver_exists && !value.is_zero()) || must_create {
+        if !receiver_exists && (!value.is_zero() || opcode_is_create) {


This condition slips through the old PR that fixes all creation conditions.
https://github.com/privacy-scaling-explorations/zkevm-circuits/pull/1079/files/c0e6909d5135aa74e3ce415879d993a34050a823#r1336618956

We have to think carefully about this condition. I would love input from reviewers.

What is receiver_exists?

We can check if code_hash is 0. It is an artificial rule we added before.

Nonce is 0. Should we add this? address_collision in create has this check.

opcode_is_create

originally called must_create. I think this covers 3 cases:

EOA transfer to create an account. Can we call this situtation "opcode_is_create"?

Create receiver with CREATE opcode

Create receiver with CREATE2 opcode

From geth impl., an account which doesn't exist will be created only when transfer value is non-zero.

But, no matter this account exsists or not, it'll be created again in CREATE/2 opcode. You can see createAccount is called in create and then will find this

// createObject creates a new state object. If there is an existing account with // the given address, it is overwritten and returned as the second return value.

in createObject

The conclusion is the original condition is correct. In your fix, the account won't be re-created if the receiver exists.

For the naming of opcode_is_create. I would vote is_create_account or force_create_account as you mentioned "EOA transfer to create an account", it doesn't fix the name.

an account which doesn't exist will be created only when transfer value is non-zero.

I think this is correct and it matches our behavior.

no matter this account exists or not, it'll be created again in CREATE/2 opcode.

This is incorrect. Three lines above your quoted createAccount locates a nonzero nonce and non-empty hash check that early exits the function and returns an ErrContractAddressCollision error.

The EIP-1014: Skinny CREATE2 also states "if nonce or code (of the create2 destination address) is nonzero, then the create-operation fails."

So for this PR, the conclusion is that a CREATEX opcode does not guarantee an overwrite of an account.

But you might wonder, why the create opcode only checks non empty code hash and non zero nonce. What about storage root and balance? Can we have an account that has code="" and nonce=0, but storage root !=0 or balance !=0? The tl;dr is we can't, for non-trivial reason.

I fell into a rabbit hole of a historical study and how the EIP was defined. Fortunately, Gary wrote the best summary of this issue. See ethereum/go-ethereum#28666.

Sorry for the confusion. Maybe we could put the link ethereum/go-ethereum#28666 in our comment.

I added to the comment in the constraint part, which might be the place people will check more frequently 77cfbcb.

EOA transfer to create an account. Can we call this situtation "opcode_is_create"?

My opinion for opcode_is_create to represents create an account with tx.to = null is a bit confusing. How about just name it is_create to align with other place e.g. call.is_create?

Sorry I missed the renaming part before. I fixed the renaming here 7449289

zkevm-circuits/src/evm_circuit/execution/begin_tx.rs

bus-mapping/src/circuit_input_builder/input_state_ref.rs

KimiWu123

Not finished the review yet, but seems some changes needed and the light testing is falied.

ChihChengLiang

Fixing the read-write inconsistency. I'll this back to draft

bus-mapping/src/circuit_input_builder/input_state_ref.rs

ChihChengLiang

I collected all the reads and writes from assignment functions everywhere back to the transfer gadget. I hope this reduces the code complexity.

ChihChengLiang · 2024-02-23T11:34:45Z

bus-mapping/src/circuit_input_builder/input_state_ref.rs

        value: Word,
        reversible: bool,
    ) -> Result<(), Error> {
-        // If receiver doesn't exist, create it
-        if (!receiver_exists && !value.is_zero()) || must_create {
+        if !receiver_exists && (!value.is_zero() || opcode_is_create) {


We have to think carefully about this condition. I would love input from reviewers.

What is receiver_exists?

We can check if code_hash is 0. It is an artificial rule we added before.

Nonce is 0. Should we add this? address_collision in create has this check.

opcode_is_create

originally called must_create. I think this covers 3 cases:

EOA transfer to create an account. Can we call this situtation "opcode_is_create"?

Create receiver with CREATE opcode

Create receiver with CREATE2 opcode

KimiWu123

Not finished yet, but would like to discuss the callee existence earlier.

zkevm-circuits/src/evm_circuit/util/common_gadget.rs

ChihChengLiang

Replied some comments that I have quick answers

zkevm-circuits/src/evm_circuit/util/common_gadget.rs

KimiWu123

LGTM. It looks nice and clean now. Great work! The remaining thing is this dicussion thread

zkevm-circuits/src/evm_circuit/execution/create.rs

ChihChengLiang

Hi @KimiWu123, I replied to the feedback to the receiver creation check.

ChihChengLiang · 2024-03-04T17:31:39Z

bus-mapping/src/circuit_input_builder/input_state_ref.rs

        value: Word,
        reversible: bool,
    ) -> Result<(), Error> {
-        // If receiver doesn't exist, create it
-        if (!receiver_exists && !value.is_zero()) || must_create {
+        if !receiver_exists && (!value.is_zero() || opcode_is_create) {


an account which doesn't exist will be created only when transfer value is non-zero.

I think this is correct and it matches our behavior.

no matter this account exists or not, it'll be created again in CREATE/2 opcode.

This is incorrect. Three lines above your quoted createAccount locates a nonzero nonce and non-empty hash check that early exits the function and returns an ErrContractAddressCollision error.

The EIP-1014: Skinny CREATE2 also states "if nonce or code (of the create2 destination address) is nonzero, then the create-operation fails."

So for this PR, the conclusion is that a CREATEX opcode does not guarantee an overwrite of an account.

But you might wonder, why the create opcode only checks non empty code hash and non zero nonce. What about storage root and balance? Can we have an account that has code="" and nonce=0, but storage root !=0 or balance !=0? The tl;dr is we can't, for non-trivial reason.

I fell into a rabbit hole of a historical study and how the EIP was defined. Fortunately, Gary wrote the best summary of this issue. See ethereum/go-ethereum#28666.

hero78119

Thanks for the code cleanup!
LGTM with few nitpicks

zkevm-circuits/src/evm_circuit/util/common_gadget.rs

hero78119 · 2024-03-05T08:57:17Z

bus-mapping/src/circuit_input_builder/input_state_ref.rs

        value: Word,
        reversible: bool,
    ) -> Result<(), Error> {
-        // If receiver doesn't exist, create it
-        if (!receiver_exists && !value.is_zero()) || must_create {
+        if !receiver_exists && (!value.is_zero() || opcode_is_create) {


EOA transfer to create an account. Can we call this situtation "opcode_is_create"?

My opinion for opcode_is_create to represents create an account with tx.to = null is a bit confusing. How about just name it is_create to align with other place e.g. call.is_create?

hero78119

👍 👍

github-actions bot added the crate-zkevm-circuits Issues related to the zkevm-circuits workspace member label Feb 21, 2024

ChihChengLiang force-pushed the merge-transfer-gadget branch 2 times, most recently from 788cb4b to 8d4d596 Compare February 22, 2024 14:03

github-actions bot added the crate-bus-mapping Issues related to the bus-mapping workspace member label Feb 22, 2024

ChihChengLiang force-pushed the merge-transfer-gadget branch from ee2917e to dacf577 Compare February 22, 2024 16:24

ChihChengLiang marked this pull request as ready for review February 22, 2024 17:34

ChihChengLiang commented Feb 22, 2024

View reviewed changes

ChihChengLiang force-pushed the merge-transfer-gadget branch from 42c3aa0 to af62f15 Compare February 22, 2024 18:03

KimiWu123 self-requested a review February 23, 2024 02:32

KimiWu123 reviewed Feb 23, 2024

View reviewed changes

bus-mapping/src/circuit_input_builder/input_state_ref.rs Show resolved Hide resolved

KimiWu123 suggested changes Feb 23, 2024

View reviewed changes

ChihChengLiang commented Feb 23, 2024

View reviewed changes

bus-mapping/src/circuit_input_builder/input_state_ref.rs Show resolved Hide resolved

ChihChengLiang marked this pull request as draft February 23, 2024 09:56

ChihChengLiang force-pushed the merge-transfer-gadget branch from cf4e4d2 to 4a400a3 Compare February 23, 2024 11:25

ChihChengLiang marked this pull request as ready for review February 23, 2024 11:25

ChihChengLiang commented Feb 23, 2024

View reviewed changes

ChihChengLiang requested a review from KimiWu123 February 26, 2024 08:20

KimiWu123 reviewed Feb 26, 2024

View reviewed changes

ChihChengLiang commented Feb 26, 2024

View reviewed changes

zkevm-circuits/src/evm_circuit/util/common_gadget.rs Outdated Show resolved Hide resolved

zkevm-circuits/src/evm_circuit/util/common_gadget.rs Show resolved Hide resolved

KimiWu123 self-requested a review February 27, 2024 02:04

KimiWu123 reviewed Feb 27, 2024

View reviewed changes

zkevm-circuits/src/evm_circuit/execution/create.rs Outdated Show resolved Hide resolved

ChihChengLiang added 9 commits February 27, 2024 13:05

make balance update readable

11be6ee

visibility control

2811228

unused comment

7f73a7f

must create is known

45e19c6

a sure place to dedup create account

f62b1cc

account_write is always true

b919f09

the abstraction is no longer needed

0241779

merge transfer gadgets ugly

618f033

WIP

37c3fd1

ChihChengLiang added 8 commits February 27, 2024 13:05

deduplicate and fix busmapping constraint

f82c3a7

fix the read write order

79d2c3f

doc fix

52e80c4

dedup busmapping transfer

0ee4aa4

rm must_create

e3e1b1c

move all the transfer rws handling to transfer gadget

56ccc66

the early return skipped the receiver creation

7a13f05

feedback

598ad11

ChihChengLiang force-pushed the merge-transfer-gadget branch from c9f89d4 to 598ad11 Compare February 27, 2024 09:31

nitpick

0522b5c

ChihChengLiang commented Mar 4, 2024

View reviewed changes

KimiWu123 approved these changes Mar 5, 2024

View reviewed changes

hero78119 self-requested a review March 5, 2024 06:22

point to Geth\#28666 in the comment

77cfbcb

hero78119 reviewed Mar 5, 2024

View reviewed changes

ChihChengLiang added 3 commits March 5, 2024 18:54

rename receiver -> receiver_balance

990afe1

rename opcode_is_create -> is_create

7449289

expand receiver creation args

f8773e2

hero78119 approved these changes Mar 5, 2024

View reviewed changes

hero78119 added this pull request to the merge queue Mar 5, 2024

Merged via the queue into main with commit 41e3408 Mar 5, 2024
15 checks passed

hero78119 deleted the merge-transfer-gadget branch March 5, 2024 12:44

ed255 added this to the Feature Completeness milestone Mar 7, 2024

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Merge Transfer and TransferWithGas gadget #1777

Merge Transfer and TransferWithGas gadget #1777

ChihChengLiang commented Feb 21, 2024 •

edited

Loading

ChihChengLiang Feb 22, 2024

ChihChengLiang Feb 23, 2024

KimiWu123 Feb 26, 2024 •

edited

Loading

KimiWu123 Feb 26, 2024 •

edited

Loading

ChihChengLiang Mar 4, 2024 •

edited

Loading

KimiWu123 Mar 5, 2024

ChihChengLiang Mar 5, 2024

hero78119 Mar 5, 2024

ChihChengLiang Mar 5, 2024

KimiWu123 left a comment

ChihChengLiang left a comment

ChihChengLiang left a comment

ChihChengLiang Feb 23, 2024

KimiWu123 left a comment

ChihChengLiang left a comment

KimiWu123 left a comment

ChihChengLiang left a comment

ChihChengLiang Mar 4, 2024 •

edited

Loading

hero78119 left a comment

hero78119 Mar 5, 2024

hero78119 left a comment

Merge Transfer and TransferWithGas gadget #1777

Merge Transfer and TransferWithGas gadget #1777

Conversation

ChihChengLiang commented Feb 21, 2024 • edited Loading

Description

Issue Link

Type of change

Content

Test

Choose a reason for hiding this comment

Choose a reason for hiding this comment

KimiWu123 Feb 26, 2024 • edited Loading

Choose a reason for hiding this comment

KimiWu123 Feb 26, 2024 • edited Loading

Choose a reason for hiding this comment

ChihChengLiang Mar 4, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

KimiWu123 left a comment

Choose a reason for hiding this comment

ChihChengLiang left a comment

Choose a reason for hiding this comment

ChihChengLiang left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

KimiWu123 left a comment

Choose a reason for hiding this comment

ChihChengLiang left a comment

Choose a reason for hiding this comment

KimiWu123 left a comment

Choose a reason for hiding this comment

ChihChengLiang left a comment

Choose a reason for hiding this comment

ChihChengLiang Mar 4, 2024 • edited Loading

Choose a reason for hiding this comment

hero78119 left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

hero78119 left a comment

Choose a reason for hiding this comment

ChihChengLiang commented Feb 21, 2024 •

edited

Loading

KimiWu123 Feb 26, 2024 •

edited

Loading

KimiWu123 Feb 26, 2024 •

edited

Loading

ChihChengLiang Mar 4, 2024 •

edited

Loading

ChihChengLiang Mar 4, 2024 •

edited

Loading