nimbus-eth1/nimbus/transaction/host_services.nim

313 lines
12 KiB
Nim
Raw Normal View History

# Nimbus - Services available to EVM code that is run for a transaction
#
# Copyright (c) 2019-2021 Status Research & Development GmbH
# Licensed under either of
# * Apache License, version 2.0, ([LICENSE-APACHE](LICENSE-APACHE) or http://www.apache.org/licenses/LICENSE-2.0)
# * MIT license ([LICENSE-MIT](LICENSE-MIT) or http://opensource.org/licenses/MIT)
# at your option. This file may not be copied, modified, or distributed except according to those terms.
#{.push raises: [].}
import
times, stint, chronicles,
2022-12-02 04:39:12 +00:00
eth/common/eth_types, ../db/accounts_cache,
../common/[evmforks, common],
".."/[vm_state, vm_computation, vm_internals, vm_gas_costs],
./host_types, ./host_trace, ./host_call_nested
proc setupTxContext(host: TransactionHost) =
# Conversion issues:
#
# `txContext.tx_gas_price` is 256-bit, but `vmState.txGasPrice` is 64-bit
# signed (`GasInt`), and in reality it tends to be a fairly small integer,
# usually < 16 bits. Our EVM truncates whatever it gets blindly to 64-bit
# anyway. Largest ever so far may be 100,000,000.
# https://medium.com/amberdata/most-expensive-transaction-in-ethereum-blockchain-history-99d9a30d8e02
#
# `txContext.block_number` is 64-bit signed. This is actually too small for
# the Nimbus `BlockNumber` type which is 256-bit (for now), so we truncate
# the other way.
#
# `txContext.chain_id` is 256-bit, but `vmState.chaindb.config.chainId` is
# 64-bit or 32-bit depending on the target CPU architecture (Nim `uint`).
# Our EVM truncates whatever it gets blindly to 64-bit or 32-bit.
#
# No conversion required with the other fields:
#
# `txContext.tx_origin` and `txContext.block_coinbase` are 20-byte Ethereum
# addresses, no issues with these.
#
# `txContext.block_timestamp` is 64-bit signed. `vmState.timestamp.toUnix`
# is from Nim `std/times` and returns `int64` so this matches. (It's
# overkill that we store a full seconds and nanoseconds object in
# `vmState.timestamp` though.)
#
# `txContext.block_gas_limit` is 64-bit signed (EVMC assumes
# [EIP-1985](https://eips.ethereum.org/EIPS/eip-1985) although it's not
# officially accepted), and `vmState.gasLimit` is too (`GasInt`).
#
# `txContext.block_prev_randao` is 256-bit, and this one can genuinely take
# values over much of the 256-bit range.
let vmState = host.vmState
host.txContext.tx_gas_price = vmState.txCtx.gasPrice.u256.toEvmc
host.txContext.tx_origin = vmState.txCtx.origin.toEvmc
# vmState.coinbase now unused
host.txContext.block_coinbase = vmState.blockCtx.coinbase.toEvmc
# vmState.blockNumber now unused
Redesign of BaseVMState descriptor (#923) * Redesign of BaseVMState descriptor why: BaseVMState provides an environment for executing transactions. The current descriptor also provides data that cannot generally be known within the execution environment, e.g. the total gasUsed which is available not before after all transactions have finished. Also, the BaseVMState constructor has been replaced by a constructor that does not need pre-initialised input of the account database. also: Previous constructor and some fields are provided with a deprecated annotation (producing a lot of noise.) * Replace legacy directives in production sources * Replace legacy directives in unit test sources * fix CI (missing premix update) * Remove legacy directives * chase CI problem * rebased * Re-introduce 'AccountsCache' constructor optimisation for 'BaseVmState' re-initialisation why: Constructing a new 'AccountsCache' descriptor can be avoided sometimes when the current state root is properly positioned already. Such a feature existed already as the update function 'initStateDB()' for the 'BaseChanDB' where the accounts cache was linked into this desctiptor. The function 'initStateDB()' was removed and re-implemented into the 'BaseVmState' constructor without optimisation. The old version was of restricted use as a wrong accounts cache state would unconditionally throw an exception rather than conceptually ask for a remedy. The optimised 'BaseVmState' re-initialisation has been implemented for the 'persistBlocks()' function. also: moved some test helpers to 'test/replay' folder * Remove unused & undocumented fields from Chain descriptor why: Reduces attack surface in general & improves reading the code.
2022-01-18 16:19:32 +00:00
host.txContext.block_number = (vmState.blockNumber
.truncate(typeof(host.txContext.block_number)))
# vmState.timestamp now unused
host.txContext.block_timestamp = cast[int64](vmState.blockCtx.timestamp)
# vmState.gasLimit now unused
host.txContext.block_gas_limit = vmState.blockCtx.gasLimit
# vmState.difficulty now unused
2022-12-02 04:39:12 +00:00
host.txContext.chain_id = vmState.com.chainId.uint.u256.toEvmc
host.txContext.block_base_fee = vmState.blockCtx.fee.get(0.u256).toEvmc
if vmState.txCtx.versionedHashes.len > 0:
type
BlobHashPtr = typeof host.txContext.blob_hashes
host.txContext.blob_hashes = cast[BlobHashPtr](vmState.txCtx.versionedHashes[0].addr)
else:
host.txContext.blob_hashes = nil
host.txContext.blob_hashes_count= vmState.txCtx.versionedHashes.len.csize_t
host.txContext.blob_base_fee = vmState.txCtx.blobBaseFee.toEvmc
EVMC: Byte-endian conversions for 256-bit numeric values Perform byte-endian conversion for 256-bit numeric values, but not 256-bit hashes. These conversions are necessary for EVMC binary compatibility. In new EVMC, all host-side conversions are explicit, calling `flip256`. These conversions are performed in the EVMC "glue" code, which deals with the binary interface, so the host services aren't aware of conversions. We intend to skip these conversions when Nimbus host calls Nimbus EVM, even when it's a shared library, using a negotiated EVMC extension. But for now we're focused on correctness and cross-validation with third party EVMs. The overhead of endian conversion is not too high because most EVMC host calls access the database anyway. `getTxContext` does not, so the conversions from that are cached here. Also, well-optimised EVMs don't call it often. It is arguable whether endian conversion should occur for storage slots (`key`). In favour of no conversion: Slot keys are 32-byte blobs, and this is clear in the EVMC definition where slot keys are `evmc_bytes32` (not `evmc_uint256be`), meaning treating as a number is _not_ expected by EVMC. Although they are often small numbers, sometimes they are a hash from the contract code plus a number. Slot keys are hashed on the host side with Keccak256 before any database calls, so the host side does not look at them numerically. In favour of conversion: They are often small numbers and it is helpful to log them as such, rather than a long string of zero digits with 1-2 non-zero. The representation in JSON has leading zeros removed, like a number rather than a 32-byte blob. There is also an interesting space optimisation when the keys are used unhashed in storage. Nimbus currently treats slot keys on the host side as numbers, and the tests pass when endian conversion is done. So to remain consistent with other parts of Nimbus we convert slot keys. Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-11-24 15:54:59 +00:00
# Most host functions do `flip256` in `evmc_host_glue`, but due to this
# result being cached, it's better to do `flip256` when filling the cache.
host.txContext.tx_gas_price = flip256(host.txContext.tx_gas_price)
host.txContext.chain_id = flip256(host.txContext.chain_id)
host.txContext.block_base_fee = flip256(host.txContext.block_base_fee)
host.txContext.blob_base_fee = flip256(host.txContext.blob_base_fee)
EVMC: Byte-endian conversions for 256-bit numeric values Perform byte-endian conversion for 256-bit numeric values, but not 256-bit hashes. These conversions are necessary for EVMC binary compatibility. In new EVMC, all host-side conversions are explicit, calling `flip256`. These conversions are performed in the EVMC "glue" code, which deals with the binary interface, so the host services aren't aware of conversions. We intend to skip these conversions when Nimbus host calls Nimbus EVM, even when it's a shared library, using a negotiated EVMC extension. But for now we're focused on correctness and cross-validation with third party EVMs. The overhead of endian conversion is not too high because most EVMC host calls access the database anyway. `getTxContext` does not, so the conversions from that are cached here. Also, well-optimised EVMs don't call it often. It is arguable whether endian conversion should occur for storage slots (`key`). In favour of no conversion: Slot keys are 32-byte blobs, and this is clear in the EVMC definition where slot keys are `evmc_bytes32` (not `evmc_uint256be`), meaning treating as a number is _not_ expected by EVMC. Although they are often small numbers, sometimes they are a hash from the contract code plus a number. Slot keys are hashed on the host side with Keccak256 before any database calls, so the host side does not look at them numerically. In favour of conversion: They are often small numbers and it is helpful to log them as such, rather than a long string of zero digits with 1-2 non-zero. The representation in JSON has leading zeros removed, like a number rather than a 32-byte blob. There is also an interesting space optimisation when the keys are used unhashed in storage. Nimbus currently treats slot keys on the host side as numbers, and the tests pass when endian conversion is done. So to remain consistent with other parts of Nimbus we convert slot keys. Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-11-24 15:54:59 +00:00
# EIP-4399
# Transfer block randomness to difficulty OPCODE
let difficulty = vmState.difficultyOrPrevRandao.toEvmc
host.txContext.block_prev_randao = flip256(difficulty)
EVMC: Byte-endian conversions for 256-bit numeric values Perform byte-endian conversion for 256-bit numeric values, but not 256-bit hashes. These conversions are necessary for EVMC binary compatibility. In new EVMC, all host-side conversions are explicit, calling `flip256`. These conversions are performed in the EVMC "glue" code, which deals with the binary interface, so the host services aren't aware of conversions. We intend to skip these conversions when Nimbus host calls Nimbus EVM, even when it's a shared library, using a negotiated EVMC extension. But for now we're focused on correctness and cross-validation with third party EVMs. The overhead of endian conversion is not too high because most EVMC host calls access the database anyway. `getTxContext` does not, so the conversions from that are cached here. Also, well-optimised EVMs don't call it often. It is arguable whether endian conversion should occur for storage slots (`key`). In favour of no conversion: Slot keys are 32-byte blobs, and this is clear in the EVMC definition where slot keys are `evmc_bytes32` (not `evmc_uint256be`), meaning treating as a number is _not_ expected by EVMC. Although they are often small numbers, sometimes they are a hash from the contract code plus a number. Slot keys are hashed on the host side with Keccak256 before any database calls, so the host side does not look at them numerically. In favour of conversion: They are often small numbers and it is helpful to log them as such, rather than a long string of zero digits with 1-2 non-zero. The representation in JSON has leading zeros removed, like a number rather than a 32-byte blob. There is also an interesting space optimisation when the keys are used unhashed in storage. Nimbus currently treats slot keys on the host side as numbers, and the tests pass when endian conversion is done. So to remain consistent with other parts of Nimbus we convert slot keys. Signed-off-by: Jamie Lokier <jamie@shareable.org>
2021-11-24 15:54:59 +00:00
host.cachedTxContext = true
const use_evmc_glue = defined(evmc_enabled)
# When using the EVMC binary interface, each of the functions below is wrapped
# in another function that converts types to be compatible with the binary
# interface, and the functions below are not called directly. The conversions
# mostly just cast between byte-compatible types, so to avoid a redundant call
# layer, make the functions below `{.inline.}` when wrapped in this way.
when use_evmc_glue:
{.push inline.}
proc accountExists(host: TransactionHost, address: HostAddress): bool {.show.} =
if host.vmState.fork >= FkSpurious:
not host.vmState.readOnlyStateDB.isDeadAccount(address)
else:
host.vmState.readOnlyStateDB.accountExists(address)
# TODO: Why is `address` an argument in `getStorage`, `setStorage` and
# `selfDestruct`, if an EVM is only allowed to do these things to its own
# contract account and the host always knows which account?
proc getStorage(host: TransactionHost, address: HostAddress, key: HostKey): HostValue {.show.} =
host.vmState.readOnlyStateDB.getStorage(address, key)
proc setStorageStatus(host: TransactionHost, address: HostAddress,
key: HostKey, newVal: HostValue): EvmcStorageStatus {.show.} =
let
db = host.vmState.readOnlyStateDB
currentVal = db.getStorage(address, key)
if currentVal == newVal:
return EVMC_STORAGE_ASSIGNED
host.vmState.mutateStateDB:
db.setStorage(address, key, newVal)
# https://eips.ethereum.org/EIPS/eip-1283
let originalVal = db.getCommittedStorage(address, key)
if originalVal == currentVal:
if originalVal.isZero:
return EVMC_STORAGE_ADDED
# !is_zero(original_val)
if newVal.isZero:
return EVMC_STORAGE_DELETED
else:
return EVMC_STORAGE_MODIFIED
# originalVal != currentVal
if originalVal.isZero.not:
if currentVal.isZero:
if originalVal == newVal:
return EVMC_STORAGE_DELETED_RESTORED
else:
return EVMC_STORAGE_DELETED_ADDED
# !is_zero(current_val)
if newVal.isZero:
return EVMC_STORAGE_MODIFIED_DELETED
# !is_zero(new_val)
if originalVal == newVal:
return EVMC_STORAGE_MODIFIED_RESTORED
else:
return EVMC_STORAGE_ASSIGNED
# is_zero(original_val)
if originalVal == newVal:
return EVMC_STORAGE_ADDED_DELETED
else:
return EVMC_STORAGE_ASSIGNED
proc setStorage(host: TransactionHost, address: HostAddress,
key: HostKey, newVal: HostValue): EvmcStorageStatus {.show.} =
let
status = setStorageStatus(host, address, key, newVal)
fork = host.vmState.fork
refund = SstoreCost[fork][status].gasRefund
if refund != 0:
# TODO: Refund depends on `Computation` at the moment.
host.computation.gasMeter.refundGas(refund)
status
proc getBalance(host: TransactionHost, address: HostAddress): HostBalance {.show.} =
host.vmState.readOnlyStateDB.getBalance(address)
proc getCodeSize(host: TransactionHost, address: HostAddress): HostSize {.show.} =
# TODO: Check this `HostSize`, it was copied as `uint` from other code.
# Note: Old `evmc_host` uses `getCode(address).len` instead.
host.vmState.readOnlyStateDB.getCodeSize(address).HostSize
proc getCodeHash(host: TransactionHost, address: HostAddress): HostHash {.show.} =
let db = host.vmState.readOnlyStateDB
# TODO: Copied from `Computation`, but check if that code is wrong with
# `FkSpurious`, as it has different calls from `accountExists` above.
if not db.accountExists(address) or db.isEmptyAccount(address):
default(HostHash)
else:
db.getCodeHash(address)
proc copyCode(host: TransactionHost, address: HostAddress,
code_offset: HostSize, buffer_data: ptr byte,
buffer_size: HostSize): HostSize {.show.} =
# We must handle edge cases carefully to prevent overflows. `len` is signed
# type `int`, but `code_offset` and `buffer_size` are _unsigned_, and may
# have large values (deliberately if attacked) that exceed the range of `int`.
#
# Comparing signed and unsigned types is _unsafe_: A type-conversion will
# take place which breaks the comparison for some values. So here we use
# explicit type-conversions, always compare the same types, and always
# convert towards the type that cannot truncate because preceding checks have
# been used to reduce the possible value range.
#
# Note, when there is no code, `getCode` result is empty `seq`. It was `nil`
# when the DB was first implemented, due to Nim language changes since then.
var code: seq[byte] = host.vmState.readOnlyStateDB.getCode(address)
var safe_len: int = code.len # It's safe to assume >= 0.
if code_offset >= safe_len.HostSize:
return 0
let safe_offset = code_offset.int
safe_len = safe_len - safe_offset
if buffer_size < safe_len.HostSize:
safe_len = buffer_size.int
if safe_len > 0:
copyMem(buffer_data, code[safe_offset].addr, safe_len)
return safe_len.HostSize
proc selfDestruct(host: TransactionHost, address, beneficiary: HostAddress) {.show.} =
host.vmState.mutateStateDB:
let localBalance = db.getBalance(address)
if host.vmState.fork >= FkCancun:
# Zeroing contract balance except beneficiary
# is the same address
db.subBalance(address, localBalance)
# Transfer to beneficiary
db.addBalance(beneficiary, localBalance)
db.selfDestruct6780(address)
else:
# Transfer to beneficiary
db.addBalance(beneficiary, localBalance)
db.selfDestruct(address)
template call(host: TransactionHost, msg: EvmcMessage): EvmcResult =
# `call` is special. The C stack usage must be kept small for deeply nested
# EVM calls. To ensure small stack, `{.show.}` must be handled at
# `host_call_nested`, not here, and this function must use `template` to
# inline at Nim level (same for `callEvmcNested`). `{.inline.}` is not good
# enough. Due to object return it ends up using a lot more stack.
host.callEvmcNested(msg)
proc getTxContext(host: TransactionHost): EvmcTxContext {.show.} =
if not host.cachedTxContext:
host.setupTxContext()
return host.txContext
proc getBlockHash(host: TransactionHost, number: HostBlockNumber): HostHash {.show.} =
# TODO: Clean up the different messy block number types.
host.vmState.getAncestorHash(number.toBlockNumber)
proc emitLog(host: TransactionHost, address: HostAddress,
data: ptr byte, data_size: HostSize,
topics: ptr HostTopic, topics_count: HostSize) {.show.} =
var log: Log
# Note, this assumes the EVM ensures `data_size` and `topics_count` cannot be
# unreasonably large values. Largest `topics_count` should be 4 according to
# EVMC documentation, but we won't restrict it here.
if topics_count > 0:
let topicsArray = cast[ptr UncheckedArray[HostTopic]](topics)
let count = topics_count.int
log.topics = newSeq[Topic](count)
for i in 0 ..< count:
log.topics[i] = topicsArray[i]
if (data_size > 0):
log.data = newSeq[byte](data_size.int)
copyMem(log.data[0].addr, data, data_size.int)
log.address = address
host.vmState.stateDB.addlogEntry(log)
proc accessAccount(host: TransactionHost, address: HostAddress): EvmcAccessStatus {.show.} =
host.vmState.mutateStateDB:
if not db.inAccessList(address):
db.accessList(address)
return EVMC_ACCESS_COLD
else:
return EVMC_ACCESS_WARM
proc accessStorage(host: TransactionHost, address: HostAddress,
key: HostKey): EvmcAccessStatus {.show.} =
host.vmState.mutateStateDB:
if not db.inAccessList(address, key):
db.accessList(address, key)
return EVMC_ACCESS_COLD
else:
return EVMC_ACCESS_WARM
proc getTransientStorage(host: TransactionHost,
address: HostAddress, key: HostKey): HostValue {.show.} =
host.vmState.readOnlyStateDB.getTransientStorage(address, key)
proc setTransientStorage(host: TransactionHost, address: HostAddress,
key: HostKey, newVal: HostValue) {.show.} =
host.vmState.mutateStateDB:
db.setTransientStorage(address, key, newVal)
when use_evmc_glue:
{.pop: inline.}
const included_from_host_services = true
include ./evmc_host_glue
else:
export
accountExists, getStorage, storage, getBalance, getCodeSize, getCodeHash,
copyCode, selfDestruct, getTxContext, call, getBlockHash, emitLog