feat[lang]: add blobhash() builtin (#3962)

This commit adds access to the `BLOBHASH` opcode via the new builtin
function `blobhash()`. The opcode was introduced in EIP-4844.

References:
https://eips.ethereum.org/EIPS/eip-4844#opcode-to-get-versioned-hashes

---------

Signed-off-by: Pascal Marco Caversaccio <pascal.caversaccio@hotmail.ch>
Co-authored-by: tserg <8017125+tserg@users.noreply.github.com>
Co-authored-by: Daniel Schiavini <daniel.schiavini@gmail.com>
Co-authored-by: Charles Cooper <cooper.charles.m@gmail.com>

docs/built-in-functions.rst

@@ -949,6 +949,30 @@ Utilities
         >>> ExampleContract.foo()
         0xf3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855
 
+.. py:function:: blobhash(index: uint256) -> bytes32
+
+    Return the versioned hash of the ``index``-th BLOB associated with the current transaction.
+
+    .. note::
+
+         A versioned hash consists of a single byte representing the version (currently ``0x01``), followed by the last 31 bytes of the ``SHA256`` hash of the KZG commitment (`EIP-4844 <https://eips.ethereum.org/EIPS/eip-4844>`_). For the case ``index >= len(tx.blob_versioned_hashes)``, ``blobhash(index: uint256)`` returns ``empty(bytes32)``.
+
+    .. code-block:: vyper
+
+        @external
+        @view
+        def foo(index: uint256) -> bytes32:
+            return blobhash(index)
+
+    .. code-block:: vyper
+
+        >>> ExampleContract.foo(0)
+        0xfd28610fb309939bfec12b6db7c4525446f596a5a5a66b8e2cb510b45b2bbeb5
+
+        >>> ExampleContract.foo(6)
+        0x0000000000000000000000000000000000000000000000000000000000000000
+
+
 .. py:function:: empty(typename) -> Any
 
     Return a value which is the default (zero-ed) value of its type. Useful for initializing new memory variables.

tests/evm_backends/base_env.py

@@ -150,6 +150,14 @@ def timestamp(self, value: int):
     def last_result(self) -> ExecutionResult:
         raise NotImplementedError  # must be implemented by subclasses
 
+    @property
+    def blob_hashes(self) -> list[bytes]:
+        raise NotImplementedError  # must be implemented by subclasses
+
+    @blob_hashes.setter
+    def blob_hashes(self, value: list[bytes]):
+        raise NotImplementedError  # must be implemented by subclasses
+
     def message_call(
         self,
         to: str,
@@ -159,7 +167,6 @@ def message_call(
         gas: int | None = None,
         gas_price: int = 0,
         is_modifying: bool = True,
-        blob_hashes: Optional[list[bytes]] = None,  # for blobbasefee >= Cancun
     ) -> bytes:
         raise NotImplementedError  # must be implemented by subclasses
 

tests/evm_backends/pyevm_env.py

@@ -4,16 +4,16 @@
 
 import rlp
 from cached_property import cached_property
-from eth.abc import ChainAPI, ComputationAPI
+from eth.abc import ChainAPI, ComputationAPI, VirtualMachineAPI
 from eth.chains.mainnet import MainnetChain
 from eth.constants import CREATE_CONTRACT_ADDRESS, GENESIS_DIFFICULTY
 from eth.db.atomic import AtomicDB
 from eth.exceptions import Revert, VMError
 from eth.tools.builder import chain as chain_builder
 from eth.vm.base import StateAPI
 from eth.vm.execution_context import ExecutionContext
+from eth.vm.forks.cancun.transaction_context import CancunTransactionContext
 from eth.vm.message import Message
-from eth.vm.transaction_context import BaseTransactionContext
 from eth_keys.datatypes import PrivateKey
 from eth_typing import Address
 from eth_utils import setup_DEBUG2_logging, to_canonical_address, to_checksum_address
@@ -54,13 +54,14 @@ def __init__(
         )
 
         self._last_computation: ComputationAPI = None
+        self._blob_hashes: list[bytes] = []
 
     @cached_property
     def _state(self) -> StateAPI:
         return self._vm.state
 
     @cached_property
-    def _vm(self):
+    def _vm(self) -> VirtualMachineAPI:
         return self._chain.get_vm()
 
     @cached_property
@@ -109,6 +110,14 @@ def last_result(self) -> ExecutionResult:
             gas_used=result.get_gas_used(),
         )
 
+    @property
+    def blob_hashes(self) -> list[bytes]:
+        return self._blob_hashes
+
+    @blob_hashes.setter
+    def blob_hashes(self, value: list[bytes]):
+        self._blob_hashes = value
+
     def message_call(
         self,
         to: str,
@@ -118,7 +127,6 @@ def message_call(
         gas: int | None = None,
         gas_price: int = 0,
         is_modifying: bool = True,
-        blob_hashes: Optional[list[bytes]] = None,  # for blobbasefee >= Cancun
     ):
         if isinstance(data, str):
             data = bytes.fromhex(data.removeprefix("0x"))
@@ -135,7 +143,7 @@ def message_call(
                     gas=self.gas_limit if gas is None else gas,
                     is_static=not is_modifying,
                 ),
-                transaction_context=BaseTransactionContext(origin=sender, gas_price=gas_price),
+                transaction_context=self._make_tx_context(sender, gas_price),
             )
         except VMError as e:
             # py-evm raises when user is out-of-funds instead of returning a failed computation
@@ -144,6 +152,14 @@ def message_call(
         self._check_computation(computation)
         return computation.output
 
+    def _make_tx_context(self, sender, gas_price):
+        context_class = self._state.transaction_context_class
+        context = context_class(origin=sender, gas_price=gas_price)
+        if self._blob_hashes:
+            assert isinstance(context, CancunTransactionContext)
+            context._blob_versioned_hashes = self._blob_hashes
+        return context
+
     def clear_transient_storage(self) -> None:
         try:
             self._state.clear_transient_storage()
@@ -185,7 +201,7 @@ def _deploy(self, code: bytes, value: int, gas: int = None) -> str:
                     gas=gas or self.gas_limit,
                     create_address=target_address,
                 ),
-                transaction_context=BaseTransactionContext(origin=sender, gas_price=0),
+                transaction_context=self._make_tx_context(sender, gas_price=0),
             )
         except VMError as e:
             # py-evm raises when user is out-of-funds instead of returning a failed computation

tests/evm_backends/revm_env.py

@@ -76,6 +76,16 @@ def last_result(self) -> ExecutionResult:
             logs=result.logs,
         )
 
+    @property
+    def blob_hashes(self):
+        return self._evm.env.tx.blob_hashes
+
+    @blob_hashes.setter
+    def blob_hashes(self, value):
+        tx = self._evm.env.tx
+        tx.blob_hashes = value
+        self._evm.set_tx_env(tx)
+
     def message_call(
         self,
         to: str,
@@ -89,10 +99,6 @@ def message_call(
     ):
         if isinstance(data, str):
             data = bytes.fromhex(data.removeprefix("0x"))
-        if blob_hashes is not None:
-            tx = self._evm.env.tx
-            tx.blob_hashes = blob_hashes
-            self._evm.set_tx_env(tx)
 
         try:
             return self._evm.message_call(

tests/functional/builtins/codegen/test_blobhash.py

@@ -0,0 +1,65 @@
+import random
+
+import pytest
+
+from vyper.compiler import compile_code
+
+valid_list = [
+    """
+@external
+@view
+def foo() -> bytes32:
+    return blobhash(0)
+    """,
+    """
+@external
+@view
+def foo() -> bytes32:
+    a: bytes32 = 0x0000000000000000000000000000000000000000000000000000000000000005
+    a = blobhash(2)
+    return a
+    """,
+    """
+@external
+@view
+def foo() -> bytes32:
+    a: bytes32 = blobhash(0)
+    assert a != empty(bytes32)
+    return a
+    """,
+    """
+@external
+@view
+def foo() -> bytes32:
+    a: bytes32 = blobhash(1337)
+    assert a == empty(bytes32)
+    return a
+    """,
+]
+
+
+@pytest.mark.requires_evm_version("cancun")
+@pytest.mark.parametrize("good_code", valid_list)
+def test_blobhash_success(good_code):
+    assert compile_code(good_code) is not None
+    out = compile_code(good_code, output_formats=["opcodes_runtime"])
+    assembly = out["opcodes_runtime"].split(" ")
+    assert "BLOBHASH" in assembly
+
+
+@pytest.mark.requires_evm_version("cancun")
+def test_get_blobhashes(env, get_contract, tx_failed):
+    code = """
+@external
+def get_blobhash(i: uint256) -> bytes32:
+    return blobhash(i)
+"""
+    c = get_contract(code)
+
+    # mock the evm blobhash attribute
+    env.blob_hashes = [random.randbytes(32) for _ in range(6)]
+
+    for i in range(6):
+        assert c.get_blobhash(i) == env.blob_hashes[i]
+
+    assert c.get_blobhash(len(env.blob_hashes)) == b"\0" * 32

tests/functional/codegen/environment_variables/test_blobbasefee.py

@@ -19,15 +19,14 @@ def get_blobbasefee() -> uint256:
     env.set_excess_blob_gas(10**6)
 
     # kzg_hash(b"Vyper is the language of the sneks")
-    blob_hashes = [
+    env.blob_hashes = [
         (bytes.fromhex("015a5c97e3cc516f22a95faf7eefff00eb2fee7a65037fde07ac5446fc93f2a0"))
     ] * 6
 
     env.message_call(
         "0xb45BEc6eeCA2a09f4689Dd308F550Ad7855051B5",  # random address
         gas=21000,
         gas_price=10**10,
-        blob_hashes=blob_hashes,
     )
 
     excess_blob_gas = env.get_excess_blob_gas()

vyper/builtins/functions.py

@@ -38,9 +38,11 @@
 from vyper.codegen.ir_node import Encoding, scope_multi
 from vyper.codegen.keccak256_helper import keccak256_helper
 from vyper.evm.address_space import MEMORY
+from vyper.evm.opcodes import version_check
 from vyper.exceptions import (
     ArgumentException,
     CompilerPanic,
+    EvmVersionException,
     InvalidLiteral,
     InvalidType,
     StateAccessViolation,
@@ -1213,6 +1215,18 @@ def build_IR(self, expr, args, kwargs, contact):
         )
 
 
+class BlobHash(BuiltinFunctionT):
+    _id = "blobhash"
+    _inputs = [("index", UINT256_T)]
+    _return_type = BYTES32_T
+
+    @process_inputs
+    def build_IR(self, expr, args, kwargs, contact):
+        if not version_check(begin="cancun"):
+            raise EvmVersionException("`blobhash` is not available pre-cancun", expr)
+        return IRnode.from_list(["blobhash", args[0]], typ=BYTES32_T)
+
+
 class RawRevert(BuiltinFunctionT):
     _id = "raw_revert"
     _inputs = [("data", BytesT.any())]
@@ -2594,6 +2608,7 @@ def _try_fold(self, node):
     "as_wei_value": AsWeiValue(),
     "raw_call": RawCall(),
     "blockhash": BlockHash(),
+    "blobhash": BlobHash(),
     "bitwise_and": BitwiseAnd(),
     "bitwise_or": BitwiseOr(),
     "bitwise_xor": BitwiseXor(),

vyper/evm/opcodes.py

@@ -75,6 +75,7 @@
     "CHAINID": (0x46, 0, 1, 2),
     "SELFBALANCE": (0x47, 0, 1, 5),
     "BASEFEE": (0x48, 0, 1, 2),
+    "BLOBHASH": (0x49, 1, 1, (None, None, None, 3)),
     "BLOBBASEFEE": (0x4A, 0, 1, (None, None, None, 2)),
     "POP": (0x50, 1, 0, 2),
     "MLOAD": (0x51, 1, 1, 3),

vyper/venom/ir_node_to_venom.py

@@ -48,6 +48,7 @@
         "signextend",
         "chainid",
         "basefee",
+        "blobhash",
         "blobbasefee",
         "timestamp",
         "blockhash",

vyper/venom/venom_to_assembly.py

@@ -97,6 +97,7 @@
         "delegatecall",
         "codesize",
         "basefee",
+        "blobhash",
         "blobbasefee",
         "prevrandao",
         "difficulty",