Please check the type of change your PR introduces: - [ ] Bugfix - [ ] Feature - [ ] Code style update (formatting, renaming) - [ ] Refactoring (no functional changes, no api changes) - [ ] Build related changes - [x] Documentation content changes - [ ] Other (please describe):  Resolves #<Issue number>  - - -  Time spent on this PR:  feat: add ecAdd precompile (kkrt-labs#405) Please check the type of change your PR introduces: - [ ] Bugfix - [x] Feature - [ ] Code style update (formatting, renaming) - [ ] Refactoring (no functional changes, no api changes) - [ ] Build related changes - [ ] Documentation content changes - [ ] Other (please describe): There is no ecAdd implementation. Resolves kkrt-labs#362 - Implements ecAdd precompile and a unit test. Time spent on this PR: 1 day fix ci and readme (kkrt-labs#406) Please check the type of change your PR introduces: - [ ] Bugfix - [ ] Feature - [ ] Code style update (formatting, renaming) - [ ] Refactoring (no functional changes, no api changes) - [x] Build related changes - [ ] Documentation content changes - [ ] Other (please describe): The CI fails because the auto format cannot commit to main. There is an open discussion about how this should be done (allowing the bot to push to a protected branch is not the right way, see https://github.com/orgs/community/discussions/25305). This is a quick fix. Just applied `make-format` to fix the CI. I've also fixed the build link in the CI and replaced the project owner in all-contributors Feat staticcall disallows subcontext statechanges (kkrt-labs#260) (kkrt-labs#398) Please check the type of change your PR introduces: - [ ] Bugfix - [X] Feature - [ ] Code style update (formatting, renaming) - [ ] Refactoring (no functional changes, no api changes) - [ ] Build related changes - [ ] Documentation content changes - [ ] Other (please describe): Opcodes invoked in staticcall calling context can change state. Resolves kkrt-labs#260 Stateful opcodes check against a read_only field in the calling context. When in read only mode, it returns a context with an updated stack and bypasses state changing behavior. Time spent on this PR: .75 days Co-authored-by: kakarot CI <sayajin-labs@users.noreply.github.com> Remove unsupported opcode section (kkrt-labs#408) Please check the type of change your PR introduces: - [ ] Bugfix - [ ] Feature - [ ] Code style update (formatting, renaming) - [ ] Refactoring (no functional changes, no api changes) - [ ] Build related changes - [x] Documentation content changes - [ ] Other (please describe): Diagram in README is not up to date Diagram shows 100% opcodes. Use foundry output files in for tests and remove old docker build (kkrt-labs#410) Please check the type of change your PR introduces: - [ ] Bugfix - [ ] Feature - [ ] Code style update (formatting, renaming) - [x] Refactoring (no functional changes, no api changes) - [ ] Build related changes - [ ] Documentation content changes - [ ] Other (please describe): Solidity files are duplicated and foundry is not used to compile files used for testing. Use foundry for building all the solidity files in the projet and make these builds available for integration tests. I removed the scripts using docker that became useless since it's a single command with forge, so added it directly to the Makefile Time spent on this PR: 0.5 [Precompiles] ecrecover (kkrt-labs#407) Please check the type of change your PR introduces: - [ ] Bugfix - [x] Feature - [ ] Code style update (formatting, renaming) - [ ] Refactoring (no functional changes, no api changes) - [ ] Build related changes - [ ] Documentation content changes - [ ] Other (please describe): Not implemented Resolves kkrt-labs#357 Implemented Fix some bugs related to Precompiles when used in integration test. We need to add integration test for all opcodes! Time spent on this PR: 1 day Fix state update is persisted across tests (kkrt-labs#412) Please check the type of change your PR introduces: - [x] Bugfix - [x] Feature - [ ] Code style update (formatting, renaming) - [ ] Refactoring (no functional changes, no api changes) - [ ] Build related changes - [ ] Documentation content changes - [ ] Other (please describe): When using a Solidity contracts in tests, state modifications are persisted across different tests. Furthermore it's not possible to deploy to times the same contract (with different constructor args for example). The `StarknetState` is cached in an `autouse` fixture; contracts can be deployed at module level, cached while state is reset after each test to its initial value. I have moved some tests according to TODOs that were written herein (test concerning EXTCODECOPY inside test_erc20 and test_counter. `Counter.sol` actually had no test so I wrote a bunch of them. Time spent on this PR: 1 day Parse events data for solidity contracts (kkrt-labs#414) Please check the type of change your PR introduces: - [ ] Bugfix - [x] Feature - [ ] Code style update (formatting, renaming) - [ ] Refactoring (no functional changes, no api changes) - [ ] Build related changes - [ ] Documentation content changes - [ ] Other (please describe): There is no way to test events emitted by a Solidity contract Contract's events can be accessed with `contract.events.{event_name}` (one attribute per event). This returns a list of emitted events in the last transaction; events are described as dict, eg.: ``` await token.approve(...) assert token.events.Approval == [{ "owner": "0x..", "spender": "0x...", "value": 10, }] ``` `web3.Contract` class already had `events.{event_name}` attributes. I've overridden this attribute to ease manipulation in our context. Time spent on this PR: 1 day add unchecked and in place cases for decrementing (kkrt-labs#415)   Please check the type of change your PR introduces: - [ ] Bugfix - [ ] Feature - [ ] Code style update (formatting, renaming) - [ ] Refactoring (no functional changes, no api changes) - [ ] Build related changes - [ ] Documentation content changes - [X] Other (please describe): Minimal working example of error cases that came up when testing the solmates erc721 impl. The decrementing comes up in `burn` and `transferFrom` and tests involving these had failure. See: https://github.com/transmissions11/solmate/blob/main/src/tokens/ERC721.sol#L178  Resolves #<Issue number>  - - -  Time spent on this PR: 1/4 of a day  Co-authored-by: kakarot CI <sayajin-labs@users.noreply.github.com> fix test env kethaa format cairo code fix ci fix tests test clarification, typos, comments and fixtures fix: comments & TODO: address fixture && interface id fix: comments TODO: addresses fixture with new Clement's PR Feat/unsupported precompile event (kkrt-labs#404) In order to allow for granular knowledge of what precompile users need most, we include the address of the unimplemented precompile in the `NotImplementedPrecompile` error ## Pull request type Please check the type of change your PR introduces: * [ ] Bugfix * [x] Feature * [ ] Code style update (formatting, renaming) * [ ] Refactoring (no functional changes, no api changes) * [ ] Build related changes * [ ] Documentation content changes * [ ] Other (please describe): Resolves kkrt-labs#372 Time spent on this PR: 0.1d Co-authored-by: kakarot CI <sayajin-labs@users.noreply.github.com> docs/Deployment Process (kkrt-labs#409)   Please check the type of change your PR introduces: - [ ] Bugfix - [ ] Feature - [ ] Code style update (formatting, renaming) - [ ] Refactoring (no functional changes, no api changes) - [ ] Build related changes - [X] Documentation content changes - [ ] Other (please describe):  Resolves kkrt-labs#382  - Adding starknet.py lib to poetry dependencies. - Documenting Process of deploying Kakarot using the .py deployment script + Some general documentation regarding the different components and their functionality. - Time spent on this PR: 0.5d Creating a github action to automatically generate the contract's call graph + readme (kkrt-labs#413) This PR contains: - A new github action - Updating the readme with a small decription Please check the type of change your PR introduces: - [ ] Bugfix - [X] Feature - [ ] Code style update (formatting, renaming) - [ ] Refactoring (no functional changes, no api changes) - [ ] Build related changes - [X] Documentation content changes - [ ] Other (please describe): Time spent on this PR: 1 day Co-authored-by: gaetbout <gaetant.ansotte@hotmail.com> Co-authored-by: kakarot CI <sayajin-labs@users.noreply.github.com> Univ2 token test (kkrt-labs#425) Please check the type of change your PR introduces: - [ ] Bugfix - [x] Feature - [ ] Code style update (formatting, renaming) - [ ] Refactoring (no functional changes, no api changes) - [ ] Build related changes - [ ] Documentation content changes - [ ] Other (please describe): The uniswap contracts are not tested The tests from https://github.com/Uniswap/v2-core/blob/master/test/UniswapV2ERC20.spec.ts have been ported into the kakarot framework. Still need to port factory and pair tests, see https://github.com/Uniswap/v2-core/tree/master/test Time spent on this PR: 2 day addresses fixture addresses fixture CI Generating contract's call graphs remove helper clone functions CI Generating contract's call graphs remove unused bytes_to_words CI Generating contract's call graphs parametized tests CI Generating contract's call graphs precompile RIPEMD-160 + Tests (kkrt-labs#421) Please check the type of change your PR introduces: - [ ] Bugfix - [X] Feature - [ ] Code style update (formatting, renaming) - [ ] Refactoring (no functional changes, no api changes) - [ ] Build related changes - [ ] Documentation content changes - [ ] Other (please describe): no precompile RIPEMD-160 Resolves kkrt-labs#359 precompile RIPEMD-160 🚀  Time spent on this PR: 1 day  Co-authored-by: kakarot CI <sayajin-labs@users.noreply.github.com> Complete solmate Erc20 integration testing (kkrt-labs#352) (kkrt-labs#430) Please check the type of change your PR introduces: - [ ] Bugfix - [X] Feature - [ ] Code style update (formatting, renaming) - [ ] Refactoring (no functional changes, no api changes) - [ ] Build related changes - [ ] Documentation content changes - [ ] Other (please describe): Incomplete integration test coverage for the solmates erc20 contract. Resolves kkrt-labs#352 Richer integration testing ported from solmates [here](https://github.com/transmissions11/solmate/blob/main/src/test/ERC20.t.sol)  - - -  Time spent on this PR: 3/4s of a day  Co-authored-by: Clément Walter <clement0walter@gmail.com> Co-authored-by: kakarot CI <sayajin-labs@users.noreply.github.com> Feat/erc721 testing (kkrt-labs#353) (kkrt-labs#418) Please check the type of change your PR introduces: - [ ] Bugfix - [X] Feature - [ ] Code style update (formatting, renaming) - [ ] Refactoring (no functional changes, no api changes) - [ ] Build related changes - [ ] Documentation content changes - [ ] Other (please describe): Minimal integration testing for the solmates erc721 contract Resolves kkrt-labs#353 Richer tests based on those found [here](https://github.com/transmissions11/solmate/blob/main/src/test/ERC721.t.sol) - All tests involving burning or transferring will fail, because they involve unchecked arithmetic ops. - I reproduced a MWE here kkrt-labs#415 and in the telegram danilo suggested a fix, which I added to the example and resolved it - With those changes in kkrt-labs#415, tests should pass 100% (or atleast they do locally) - I made two changes to the solmates logic to adapt to our test runner, will comment in line but briefly - safeMint / safeTransfer were methods that had dispatching based on arguments. I renamed the variants off each method that take a data argument safeMint2 and safeTransfer2. - Justification: when I tried to call the methods as is, I was seeing the following error: ```python def get_function_by_identifier( fns: Sequence[ContractFunction], identifier: str ) -> ContractFunction: if len(fns) > 1: raise ValueError( 'Found multiple functions with matching {0}. ' 'Found: {1!r}'.format(identifier, fns) ValueError: Found multiple functions with matching name. Found: [<Function safeMint(address,uint256,bytes)>, <Function safeMint(address,uint256)> ``` - In the `ERC721Recipient` contract, there is a public address field named 'from.' I renamed this to from_. - Justification: when trying to call the getter 'from' in python I was getting this error, which I presumed to mean there was some overlap in the kakarot wrapping: ```shell File "./starknet/kakarot/tests/integration/solidity_contracts/Solmate/test_erc721.py", line 399 recipient_from = await erc_721_recipient.from() SyntaxError: invalid syntax ```  Time spent on this PR: 1 day  Co-authored-by: kakarot CI <sayajin-labs@users.noreply.github.com> CI Generating contract's call graphs fix: typo and native functions CI Generating contract's call graphs set solidity files to main branch last commit CI Generating contract's call graphs fix web3 CI Generating contract's call graphs fix web3 CI Generating contract's call graphs Kakarot Ethereum Account Abstraction (kkrt-labs#422) The Kakarot Ethereum Account Abstraction (KETHAA) is a set of two contracts, a deployer and the account. The deployer is in charge of deploying the accounts and the account manages the EVM transaction verification and execution. For the moment only EIP-1559 transactions are considered valid and no tx is executed. Few more things will be added (see related issues). Not final state. Please check the type of change your PR introduces: - [ ] Bugfix - [x] Feature - [ ] Code style update (formatting, renaming) - [ ] Refactoring (no functional changes, no api changes) - [ ] Build related changes - [ ] Documentation content changes - [ ] Other (please describe):  Resolves kkrt-labs#388 Kakarot itself is not altered. Time spent on this PR: 10 days Co-authored-by: kakarot CI <sayajin-labs@users.noreply.github.com> docs: add evm summary (kkrt-labs#429)   Please check the type of change your PR introduces: - [ ] Bugfix - [ ] Feature - [ ] Code style update (formatting, renaming) - [ ] Refactoring (no functional changes, no api changes) - [ ] Build related changes - [ ] Documentation content changes - [ ] Other (please describe):  Resolves #<Issue number>  - - -  Time spent on this PR:  CI Generating contract's call graphs tests:rlp_list_encode CI Formatting Auto Commit fix conflicts CI Formatting Auto Commit docs: put time spent on PR field at top of PR template (kkrt-labs#435)   Please check the type of change your PR introduces: - [ ] Bugfix - [ ] Feature - [ ] Code style update (formatting, renaming) - [ ] Refactoring (no functional changes, no api changes) - [ ] Build related changes - [x] Documentation content changes - [ ] Other (please describe):  Resolves #<Issue number>  - - -  Time spent on this PR:  Load revert reason short string (kkrt-labs#431) Please check the type of change your PR introduces: - [x] Bugfix - [ ] Feature - [ ] Code style update (formatting, renaming) - [ ] Refactoring (no functional changes, no api changes) - [ ] Build related changes - [ ] Documentation content changes - [ ] Other (please describe): The revert opcode does not load the revert reason but only returns a random number. Resolves kkrt-labs#416 The reason is loaded as a bytes array. Only the short string is propagated with the `error_attr`. Time spent on this PR: 1 day We definitely need to fix REVERT (see kkrt-labs#432) because this is close to a nightmare in terms of hard fixing to have something ok for testing. Fix: typos (kkrt-labs#437)  Please check the type of change your PR introduces: - [x] Documentation content changes CI Generating contract's call graphs remove unused file CI Generating contract's call graphs pending fix env (source main branch) & review CI Formatting Auto Commit clean CI Generating contract's call graphs Add autoflake to format and poetry check to build (kkrt-labs#436) Please check the type of change your PR introduces: - [ ] Bugfix - [ ] Feature - [ ] Code style update (formatting, renaming) - [ ] Refactoring (no functional changes, no api changes) - [x] Build related changes - [ ] Documentation content changes - [ ] Other (please describe): Some files have unused imports and variables Added the tool `autoflake` to format-check and format to check and automatically fix these issues. Time spent on this PR: 0.5 Add tests for CREATE2 and require in PlainOpcodes (kkrt-labs#440) Time spent on this PR: 1 day Please check the type of change your PR introduces: - [ ] Bugfix - [x] Feature - [ ] Code style update (formatting, renaming) - [ ] Refactoring (no functional changes, no api changes) - [ ] Build related changes - [ ] Documentation content changes - [ ] Other (please describe): Facing an error when testing Univ2Factory, I have eventually manage to isolate the error case in a `PlainOpcode.sol` test. The error is illustrated in `class TestRequire` that is currently skipped (see also create issue kkrt-labs#439) UniswapV2Factory test (kkrt-labs#433) Please check the type of change your PR introduces: - [ ] Bugfix - [x] Feature - [ ] Code style update (formatting, renaming) - [ ] Refactoring (no functional changes, no api changes) - [ ] Build related changes - [ ] Documentation content changes - [ ] Other (please describe): The `UniswapV2Factory.sol` contract is not tested. Tests from https://github.com/Uniswap/v2-core/blob/master/test/UniswapV2Factory.spec.ts have been ported to kakarot test framework. Still missing UniswapV2Pair. Time spent on this PR: 2 days Co-authored-by: kakarot CI <sayajin-labs@users.noreply.github.com> Test [Fix] Generate 20 bytes instead of a uint160 in Univ2Factory test (kkrt-labs#444) Time spent on this PR: 0.5 Please check the type of change your PR introduces: - [x] Bugfix - [ ] Feature - [ ] Code style update (formatting, renaming) - [ ] Refactoring (no functional changes, no api changes) - [ ] Build related changes - [ ] Documentation content changes - [ ] Other (please describe): Random addresses are generated without setting the seed These addresses may have a leading `0` nibble in bytes representation which breaks the `to_checksum_address` Set seed Generate 20 random bytes instead of a uint160. fix: change RNG to allow seeding (kkrt-labs#447) Time spent on this PR: 0.2 days Please check the type of change your PR introduces: - [x] Bugfix - [ ] Feature - [ ] Code style update (formatting, renaming) - [ ] Refactoring (no functional changes, no api changes) - [ ] Build related changes - [ ] Documentation content changes - [ ] Other (please describe): Random bytes drawing for TEST_ADDRESSES is not seedable. The TEST_ADDRESSES are deterministically generated. Add devcontainers for codespace Use Python3 microsoft image and setup after build Add test for ORIGIN and CALLER (+some refacto) (kkrt-labs#448) Time spent on this PR: 1 day Please check the type of change your PR introduces: - [ ] Bugfix - [x] Feature - [ ] Code style update (formatting, renaming) - [ ] Refactoring (no functional changes, no api changes) - [ ] Build related changes - [ ] Documentation content changes - [ ] Other (please describe): Resolves kkrt-labs#342 No test to illustrate the bug of the ORIGIN opcode Added a `originAndSender` function in `PlainOpcodes.sol` to check that `tx.origin` and `msg.sender` are correctly returned when called both from an EAO and from a contract. Some minor refacto along the way, especially I added a forge test to check my .sol function to be tested in kakarot (!). We may want to start diff testing one day or another CI Generating contract's call graphs simplify tests CI Formatting Auto Commit Bugfix: create2 opcode should use deterministically derived evm contract address (kkrt-labs#443)   Time spent on this PR: 1.75 days  Please check the type of change your PR introduces: - [X] Bugfix - [ ] Feature - [ ] Code style update (formatting, renaming) - [ ] Refactoring (no functional changes, no api changes) - [ ] Build related changes - [ ] Documentation content changes - [ ] Other (please describe): The EVM address of a contract account is computed as follows: ``` // Generate EVM_contract address from the new cairo contract // TODO: Use RLP to compute proper EVM address, see https://www.evm.codes/#f0 let (_, low) = split_felt(starknet_contract_address); local evm_contract_address = 0xAbdE100700000000000000000000000000000000 + low; ```  Resolves kkrt-labs#427 also resolves kkrt-labs#426 wrt create2 opcode  Create2's address is defined via ``` address = keccak256(0xff + sender_address + salt + keccak256(initialisation_code))[12:] ``` - - - There is a current temporary split in deploy code, where create2 logic is handled separately from create. When both construct their address deterministically, the temporary refactor will be removed.  Co-authored-by: Clément Walter <clement0walter@gmail.com> Co-authored-by: kakarot CI <sayajin-labs@users.noreply.github.com> CI Generating contract's call graphs fix randomness in tests CI Formatting Auto Commit fix randomness in tests CI Generating contract's call graphs test randomness & compare list to implementation CI Generating contract's call graphs fix tests CI Generating contract's call graphs Refacto tests CI Generating contract's call graphs CI Generating contract's call graphs

Complete integration tests coverage (solmate tests) for ERC20 #352

Description

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions