nimbus-eth1/tests/test_ledger.nim

733 lines
21 KiB
Nim
Raw Normal View History

# Nimbus
# Copyright (c) 2018-2025 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.
import
std/[strformat, strutils, importutils],
eth/common/[keys, transaction_utils],
stew/byteutils,
stew/endians2,
../nimbus/config,
../nimbus/db/storage_types,
../nimbus/common/common,
../nimbus/core/chain,
../nimbus/core/tx_pool,
../nimbus/transaction,
../nimbus/constants,
../nimbus/db/ledger {.all.}, # import all private symbols
unittest2
import results
const
genesisFile = "tests/customgenesis/cancun123.json"
hexPrivKey = "af1a9be9f1a54421cac82943820a0fe0f601bb5f4f6d0bccc81c613f0ce6ae22"
# The above privKey will generate this address
# senderAddr = hexToByteArray[20]("73cf19657412508833f618a15e8251306b3e6ee5")
type
TestEnv = object
com: CommonRef
xdb: CoreDbRef
txs: seq[Transaction]
txi: seq[int] # selected index into txs[] (crashable sender addresses)
vaultKey: PrivateKey
nonce : uint64
chainId : ChainId
xp : TxPoolRef
chain : ForkedChainRef
# ------------------------------------------------------------------------------
# Helpers
# ------------------------------------------------------------------------------
proc pp*(a: Address): string =
a.toHex[32 .. 39].toLowerAscii
proc pp*(tx: Transaction): string =
# "(" & tx.ecRecover.value.pp & "," & $tx.nonce & ")"
"(" & tx.recoverSender().value().pp & "," & $tx.nonce & ")"
proc pp*(h: Hash32): string =
h.data.toHex[52 .. 63].toLowerAscii
proc pp*(tx: Transaction; ledger: LedgerRef): string =
let address = tx.recoverSender().value()
"(" & address.pp &
"," & $tx.nonce &
";" & $ledger.getNonce(address) &
"," & $ledger.getBalance(address) &
")"
when isMainModule:
import chronicles
proc setTraceLevel =
discard
when defined(chronicles_runtime_filtering) and loggingEnabled:
setLogLevel(LogLevel.TRACE)
proc setErrorLevel =
discard
when defined(chronicles_runtime_filtering) and loggingEnabled:
setLogLevel(LogLevel.ERROR)
# ------------------------------------------------------------------------------
# Private functions
# ------------------------------------------------------------------------------
proc privKey(keyHex: string): PrivateKey =
let kRes = PrivateKey.fromHex(keyHex)
if kRes.isErr:
echo kRes.error
quit(QuitFailure)
kRes.get()
proc initEnv(): TestEnv =
let
conf = makeConfig(@[
"--custom-network:" & genesisFile
])
let
com = CommonRef.new(
newCoreDbRef DefaultDbMemory, nil,
conf.networkId,
conf.networkParams
)
aristo: fork support via layers/txframes (#2960) * aristo: fork support via layers/txframes This change reorganises how the database is accessed: instead holding a "current frame" in the database object, a dag of frames is created based on the "base frame" held in `AristoDbRef` and all database access happens through this frame, which can be thought of as a consistent point-in-time snapshot of the database based on a particular fork of the chain. In the code, "frame", "transaction" and "layer" is used to denote more or less the same thing: a dag of stacked changes backed by the on-disk database. Although this is not a requirement, in practice each frame holds the change set of a single block - as such, the frame and its ancestors leading up to the on-disk state represents the state of the database after that block has been applied. "committing" means merging the changes to its parent frame so that the difference between them is lost and only the cumulative changes remain - this facility enables frames to be combined arbitrarily wherever they are in the dag. In particular, it becomes possible to consolidate a set of changes near the base of the dag and commit those to disk without having to re-do the in-memory frames built on top of them - this is useful for "flattening" a set of changes during a base update and sending those to storage without having to perform a block replay on top. Looking at abstractions, a side effect of this change is that the KVT and Aristo are brought closer together by considering them to be part of the "same" atomic transaction set - the way the code gets organised, applying a block and saving it to the kvt happens in the same "logical" frame - therefore, discarding the frame discards both the aristo and kvt changes at the same time - likewise, they are persisted to disk together - this makes reasoning about the database somewhat easier but has the downside of increased memory usage, something that perhaps will need addressing in the future. Because the code reasons more strictly about frames and the state of the persisted database, it also makes it more visible where ForkedChain should be used and where it is still missing - in particular, frames represent a single branch of history while forkedchain manages multiple parallel forks - user-facing services such as the RPC should use the latter, ie until it has been finalized, a getBlock request should consider all forks and not just the blocks in the canonical head branch. Another advantage of this approach is that `AristoDbRef` conceptually becomes more simple - removing its tracking of the "current" transaction stack simplifies reasoning about what can go wrong since this state now has to be passed around in the form of `AristoTxRef` - as such, many of the tests and facilities in the code that were dealing with "stack inconsistency" are now structurally prevented from happening. The test suite will need significant refactoring after this change. Once this change has been merged, there are several follow-ups to do: * there's no mechanism for keeping frames up to date as they get committed or rolled back - TODO * naming is confused - many names for the same thing for legacy reason * forkedchain support is still missing in lots of code * clean up redundant logic based on previous designs - in particular the debug and introspection code no longer makes sense * the way change sets are stored will probably need revisiting - because it's a stack of changes where each frame must be interrogated to find an on-disk value, with a base distance of 128 we'll at minimum have to perform 128 frame lookups for *every* database interaction - regardless, the "dag-like" nature will stay * dispose and commit are poorly defined and perhaps redundant - in theory, one could simply let the GC collect abandoned frames etc, though it's likely an explicit mechanism will remain useful, so they stay for now More about the changes: * `AristoDbRef` gains a `txRef` field (todo: rename) that "more or less" corresponds to the old `balancer` field * `AristoDbRef.stack` is gone - instead, there's a chain of `AristoTxRef` objects that hold their respective "layer" which has the actual changes * No more reasoning about "top" and "stack" - instead, each `AristoTxRef` can be a "head" that "more or less" corresponds to the old single-history `top` notion and its stack * `level` still represents "distance to base" - it's computed from the parent chain instead of being stored * one has to be careful not to use frames where forkedchain was intended - layers are only for a single branch of history! * fix layer vtop after rollback * engine fix * Fix test_txpool * Fix test_rpc * Fix copyright year * fix simulator * Fix copyright year * Fix copyright year * Fix tracer * Fix infinite recursion bug * Remove aristo and kvt empty files * Fic copyright year * Fix fc chain_kvt * ForkedChain refactoring * Fix merge master conflict * Fix copyright year * Reparent txFrame * Fix test * Fix txFrame reparent again * Cleanup and fix test * UpdateBase bugfix and fix test * Fixe newPayload bug discovered by hive * Fix engine api fcu * Clean up call template, chain_kvt, andn txguid * Fix copyright year * work around base block loading issue * Add test * Fix updateHead bug * Fix updateBase bug * Change func commitBase to proc commitBase * Touch up and fix debug mode crash --------- Co-authored-by: jangko <jangko128@gmail.com>
2025-02-06 08:04:50 +01:00
chain = ForkedChainRef.init(com)
TestEnv(
com : com,
xdb : com.db,
vaultKey: privKey(hexPrivKey),
nonce : 0'u64,
chainId : conf.networkParams.config.chainId,
xp : TxPoolRef.new(chain),
chain : chain,
)
func makeTx(
env: var TestEnv,
recipient: Address,
amount: UInt256,
payload: openArray[byte] = []): Transaction =
const
gasLimit = 75000.GasInt
gasPrice = 30.gwei
let tx = Transaction(
txType : TxLegacy,
chainId : env.chainId,
nonce : AccountNonce(env.nonce),
gasPrice: gasPrice,
gasLimit: gasLimit,
to : Opt.some(recipient),
value : amount,
payload : @payload
)
inc env.nonce
signTransaction(tx, env.vaultKey, eip155 = true)
func initAddr(z: int): Address =
const L = sizeof(result)
result.data[L-sizeof(uint32)..^1] = toBytesBE(z.uint32)
proc importBlock(env: TestEnv; blk: Block) =
env.chain.importBlock(blk).isOkOr:
raiseAssert "persistBlocks() failed at block #" &
$blk.header.number & " msg: " & error
aristo: fork support via layers/txframes (#2960) * aristo: fork support via layers/txframes This change reorganises how the database is accessed: instead holding a "current frame" in the database object, a dag of frames is created based on the "base frame" held in `AristoDbRef` and all database access happens through this frame, which can be thought of as a consistent point-in-time snapshot of the database based on a particular fork of the chain. In the code, "frame", "transaction" and "layer" is used to denote more or less the same thing: a dag of stacked changes backed by the on-disk database. Although this is not a requirement, in practice each frame holds the change set of a single block - as such, the frame and its ancestors leading up to the on-disk state represents the state of the database after that block has been applied. "committing" means merging the changes to its parent frame so that the difference between them is lost and only the cumulative changes remain - this facility enables frames to be combined arbitrarily wherever they are in the dag. In particular, it becomes possible to consolidate a set of changes near the base of the dag and commit those to disk without having to re-do the in-memory frames built on top of them - this is useful for "flattening" a set of changes during a base update and sending those to storage without having to perform a block replay on top. Looking at abstractions, a side effect of this change is that the KVT and Aristo are brought closer together by considering them to be part of the "same" atomic transaction set - the way the code gets organised, applying a block and saving it to the kvt happens in the same "logical" frame - therefore, discarding the frame discards both the aristo and kvt changes at the same time - likewise, they are persisted to disk together - this makes reasoning about the database somewhat easier but has the downside of increased memory usage, something that perhaps will need addressing in the future. Because the code reasons more strictly about frames and the state of the persisted database, it also makes it more visible where ForkedChain should be used and where it is still missing - in particular, frames represent a single branch of history while forkedchain manages multiple parallel forks - user-facing services such as the RPC should use the latter, ie until it has been finalized, a getBlock request should consider all forks and not just the blocks in the canonical head branch. Another advantage of this approach is that `AristoDbRef` conceptually becomes more simple - removing its tracking of the "current" transaction stack simplifies reasoning about what can go wrong since this state now has to be passed around in the form of `AristoTxRef` - as such, many of the tests and facilities in the code that were dealing with "stack inconsistency" are now structurally prevented from happening. The test suite will need significant refactoring after this change. Once this change has been merged, there are several follow-ups to do: * there's no mechanism for keeping frames up to date as they get committed or rolled back - TODO * naming is confused - many names for the same thing for legacy reason * forkedchain support is still missing in lots of code * clean up redundant logic based on previous designs - in particular the debug and introspection code no longer makes sense * the way change sets are stored will probably need revisiting - because it's a stack of changes where each frame must be interrogated to find an on-disk value, with a base distance of 128 we'll at minimum have to perform 128 frame lookups for *every* database interaction - regardless, the "dag-like" nature will stay * dispose and commit are poorly defined and perhaps redundant - in theory, one could simply let the GC collect abandoned frames etc, though it's likely an explicit mechanism will remain useful, so they stay for now More about the changes: * `AristoDbRef` gains a `txRef` field (todo: rename) that "more or less" corresponds to the old `balancer` field * `AristoDbRef.stack` is gone - instead, there's a chain of `AristoTxRef` objects that hold their respective "layer" which has the actual changes * No more reasoning about "top" and "stack" - instead, each `AristoTxRef` can be a "head" that "more or less" corresponds to the old single-history `top` notion and its stack * `level` still represents "distance to base" - it's computed from the parent chain instead of being stored * one has to be careful not to use frames where forkedchain was intended - layers are only for a single branch of history! * fix layer vtop after rollback * engine fix * Fix test_txpool * Fix test_rpc * Fix copyright year * fix simulator * Fix copyright year * Fix copyright year * Fix tracer * Fix infinite recursion bug * Remove aristo and kvt empty files * Fic copyright year * Fix fc chain_kvt * ForkedChain refactoring * Fix merge master conflict * Fix copyright year * Reparent txFrame * Fix test * Fix txFrame reparent again * Cleanup and fix test * UpdateBase bugfix and fix test * Fixe newPayload bug discovered by hive * Fix engine api fcu * Clean up call template, chain_kvt, andn txguid * Fix copyright year * work around base block loading issue * Add test * Fix updateHead bug * Fix updateBase bug * Change func commitBase to proc commitBase * Touch up and fix debug mode crash --------- Co-authored-by: jangko <jangko128@gmail.com>
2025-02-06 08:04:50 +01:00
proc getLedger(txFrame: CoreDbTxRef): LedgerRef =
LedgerRef.init(txFrame)
func getRecipient(tx: Transaction): Address =
tx.to.expect("transaction have no recipient")
# ------------------------------------------------------------------------------
# Crash test function, finding out about how the transaction framework works ..
# ------------------------------------------------------------------------------
proc modBalance(ac: LedgerRef, address: Address) =
## This function is crucial for profucing the crash. If must
## modify the balance so that the database gets written.
# ac.blindBalanceSetter(address)
ac.addBalance(address, 1.u256)
proc runTrial2ok(env: TestEnv, ledger: LedgerRef; inx: int) =
## Run two blocks, the first one with *rollback*.
let eAddr = env.txs[inx].getRecipient
block:
let accTx = ledger.beginSavepoint
ledger.modBalance(eAddr)
ledger.rollback(accTx)
block:
let accTx = ledger.beginSavepoint
ledger.modBalance(eAddr)
ledger.commit(accTx)
ledger.persist()
proc runTrial3(env: TestEnv, ledger: LedgerRef; inx: int; rollback: bool) =
## Run three blocks, the second one optionally with *rollback*.
let eAddr = env.txs[inx].getRecipient
block:
let accTx = ledger.beginSavepoint
ledger.modBalance(eAddr)
ledger.commit(accTx)
ledger.persist()
Update storage tree admin (#2419) * Tighten `CoreDb` API for accounts why: Apart from cruft, the way to fetch the accounts state root via a `CoreDbColRef` record was unnecessarily complicated. * Extend `CoreDb` API for accounts to cover storage tries why: In future, this will make the notion of column objects obsolete. Storage trees will then be indexed by the account address rather than the vertex ID equivalent like a `CoreDbColRef`. * Apply new/extended accounts API to ledger and tests details: This makes the `distinct_ledger` module obsolete * Remove column object constructors why: They were needed as an abstraction of MPT sub-trees including storage trees. Now, storage trees are handled by the account (e.g. via address) they belong to and all other trees can be identified by a constant well known vertex ID. So there is no need for column objects anymore. Still there are some left-over column object methods wnich will be removed next. * Remove `serialise()` and `PayloadRef` from default Aristo API why: Not needed. `PayloadRef` was used for unstructured/unknown payload formats (account or blob) and `serialise()` was used for decodng `PayloadRef`. Now it is known in advance what the payload looks like. * Added query function `hasStorageData()` whether a storage area exists why: Useful for supporting `slotStateEmpty()` of the `CoreDb` API * In the `Ledger` replace `storage.stateEmpty()` by `slotStateEmpty()` * On Aristo, hide the storage root/vertex ID in the `PayloadRef` why: The storage vertex ID is fully controlled by Aristo while the `AristoAccount` object is controlled by the application. With the storage root part of the `AristoAccount` object, there was a useless administrative burden to keep that storage root field up to date. * Remove cruft, update comments etc. * Update changed MPT access paradigms why: Fixes verified proxy tests * Fluffy cosmetics
2024-06-27 09:01:26 +00:00
block body2:
let accTx = ledger.beginSavepoint
ledger.modBalance(eAddr)
if rollback:
ledger.rollback(accTx)
Update storage tree admin (#2419) * Tighten `CoreDb` API for accounts why: Apart from cruft, the way to fetch the accounts state root via a `CoreDbColRef` record was unnecessarily complicated. * Extend `CoreDb` API for accounts to cover storage tries why: In future, this will make the notion of column objects obsolete. Storage trees will then be indexed by the account address rather than the vertex ID equivalent like a `CoreDbColRef`. * Apply new/extended accounts API to ledger and tests details: This makes the `distinct_ledger` module obsolete * Remove column object constructors why: They were needed as an abstraction of MPT sub-trees including storage trees. Now, storage trees are handled by the account (e.g. via address) they belong to and all other trees can be identified by a constant well known vertex ID. So there is no need for column objects anymore. Still there are some left-over column object methods wnich will be removed next. * Remove `serialise()` and `PayloadRef` from default Aristo API why: Not needed. `PayloadRef` was used for unstructured/unknown payload formats (account or blob) and `serialise()` was used for decodng `PayloadRef`. Now it is known in advance what the payload looks like. * Added query function `hasStorageData()` whether a storage area exists why: Useful for supporting `slotStateEmpty()` of the `CoreDb` API * In the `Ledger` replace `storage.stateEmpty()` by `slotStateEmpty()` * On Aristo, hide the storage root/vertex ID in the `PayloadRef` why: The storage vertex ID is fully controlled by Aristo while the `AristoAccount` object is controlled by the application. With the storage root part of the `AristoAccount` object, there was a useless administrative burden to keep that storage root field up to date. * Remove cruft, update comments etc. * Update changed MPT access paradigms why: Fixes verified proxy tests * Fluffy cosmetics
2024-06-27 09:01:26 +00:00
break body2
ledger.commit(accTx)
ledger.persist()
block:
let accTx = ledger.beginSavepoint
ledger.modBalance(eAddr)
ledger.commit(accTx)
ledger.persist()
proc runTrial3Survive(env: TestEnv, ledger: LedgerRef; inx: int; noisy = false) =
## Run three blocks with extra db frames and *rollback*.
let eAddr = env.txs[inx].getRecipient
block:
block:
let accTx = ledger.beginSavepoint
ledger.modBalance(eAddr)
ledger.commit(accTx)
ledger.persist()
block:
let accTx = ledger.beginSavepoint
ledger.modBalance(eAddr)
ledger.rollback(accTx)
block:
block:
let accTx = ledger.beginSavepoint
ledger.modBalance(eAddr)
ledger.commit(accTx)
ledger.persist()
ledger.persist()
proc runTrial4(env: TestEnv, ledger: LedgerRef; inx: int; rollback: bool) =
## Like `runTrial3()` but with four blocks and extra db transaction frames.
let eAddr = env.txs[inx].getRecipient
block:
block:
let accTx = ledger.beginSavepoint
ledger.modBalance(eAddr)
ledger.commit(accTx)
ledger.persist()
block:
let accTx = ledger.beginSavepoint
ledger.modBalance(eAddr)
ledger.commit(accTx)
ledger.persist()
Update storage tree admin (#2419) * Tighten `CoreDb` API for accounts why: Apart from cruft, the way to fetch the accounts state root via a `CoreDbColRef` record was unnecessarily complicated. * Extend `CoreDb` API for accounts to cover storage tries why: In future, this will make the notion of column objects obsolete. Storage trees will then be indexed by the account address rather than the vertex ID equivalent like a `CoreDbColRef`. * Apply new/extended accounts API to ledger and tests details: This makes the `distinct_ledger` module obsolete * Remove column object constructors why: They were needed as an abstraction of MPT sub-trees including storage trees. Now, storage trees are handled by the account (e.g. via address) they belong to and all other trees can be identified by a constant well known vertex ID. So there is no need for column objects anymore. Still there are some left-over column object methods wnich will be removed next. * Remove `serialise()` and `PayloadRef` from default Aristo API why: Not needed. `PayloadRef` was used for unstructured/unknown payload formats (account or blob) and `serialise()` was used for decodng `PayloadRef`. Now it is known in advance what the payload looks like. * Added query function `hasStorageData()` whether a storage area exists why: Useful for supporting `slotStateEmpty()` of the `CoreDb` API * In the `Ledger` replace `storage.stateEmpty()` by `slotStateEmpty()` * On Aristo, hide the storage root/vertex ID in the `PayloadRef` why: The storage vertex ID is fully controlled by Aristo while the `AristoAccount` object is controlled by the application. With the storage root part of the `AristoAccount` object, there was a useless administrative burden to keep that storage root field up to date. * Remove cruft, update comments etc. * Update changed MPT access paradigms why: Fixes verified proxy tests * Fluffy cosmetics
2024-06-27 09:01:26 +00:00
block body3:
let accTx = ledger.beginSavepoint
ledger.modBalance(eAddr)
if rollback:
ledger.rollback(accTx)
Update storage tree admin (#2419) * Tighten `CoreDb` API for accounts why: Apart from cruft, the way to fetch the accounts state root via a `CoreDbColRef` record was unnecessarily complicated. * Extend `CoreDb` API for accounts to cover storage tries why: In future, this will make the notion of column objects obsolete. Storage trees will then be indexed by the account address rather than the vertex ID equivalent like a `CoreDbColRef`. * Apply new/extended accounts API to ledger and tests details: This makes the `distinct_ledger` module obsolete * Remove column object constructors why: They were needed as an abstraction of MPT sub-trees including storage trees. Now, storage trees are handled by the account (e.g. via address) they belong to and all other trees can be identified by a constant well known vertex ID. So there is no need for column objects anymore. Still there are some left-over column object methods wnich will be removed next. * Remove `serialise()` and `PayloadRef` from default Aristo API why: Not needed. `PayloadRef` was used for unstructured/unknown payload formats (account or blob) and `serialise()` was used for decodng `PayloadRef`. Now it is known in advance what the payload looks like. * Added query function `hasStorageData()` whether a storage area exists why: Useful for supporting `slotStateEmpty()` of the `CoreDb` API * In the `Ledger` replace `storage.stateEmpty()` by `slotStateEmpty()` * On Aristo, hide the storage root/vertex ID in the `PayloadRef` why: The storage vertex ID is fully controlled by Aristo while the `AristoAccount` object is controlled by the application. With the storage root part of the `AristoAccount` object, there was a useless administrative burden to keep that storage root field up to date. * Remove cruft, update comments etc. * Update changed MPT access paradigms why: Fixes verified proxy tests * Fluffy cosmetics
2024-06-27 09:01:26 +00:00
break body3
ledger.commit(accTx)
ledger.persist()
block:
block:
let accTx = ledger.beginSavepoint
ledger.modBalance(eAddr)
ledger.commit(accTx)
ledger.persist()
# ------------------------------------------------------------------------------
# Test Runner
# ------------------------------------------------------------------------------
const
NumTransactions = 17
NumBlocks = 13
feeRecipient = initAddr(401)
prevRandao = Bytes32 EMPTY_UNCLE_HASH # it can be any valid hash
proc runLedgerTransactionTests(noisy = true) =
suite "Ledger nesting scenarios":
var env = initEnv()
test "Create transactions and blocks":
var
recipientSeed = 501
blockTime = EthTime.now()
for _ in 0..<NumBlocks:
for _ in 0..<NumTransactions:
let recipient = initAddr(recipientSeed)
let tx = env.makeTx(recipient, 1.u256)
check env.xp.addTx(tx).isOk
inc recipientSeed
check env.xp.len == NumTransactions
env.xp.prevRandao = prevRandao
env.xp.feeRecipient = feeRecipient
env.xp.timestamp = blockTime
blockTime = EthTime(blockTime.uint64 + 1'u64)
let r = env.xp.assembleBlock()
if r.isErr:
debugEcho r.error
check false
return
let blk = r.get.blk
env.importBlock(blk)
check blk.transactions.len == NumTransactions
env.xp.removeNewBlockTxs(blk)
for tx in blk.transactions:
env.txs.add tx
let head = env.chain.latestHeader
test &"Collect unique recipient addresses from {env.txs.len} txs," &
&" head=#{head.number}":
# since we generate our own transactions instead of replaying
# from testnet blocks, the recipients already unique.
for n,tx in env.txs:
#let a = tx.getRecipient
env.txi.add n
test &"Run {env.txi.len} two-step trials with rollback":
for n in env.txi:
aristo: fork support via layers/txframes (#2960) * aristo: fork support via layers/txframes This change reorganises how the database is accessed: instead holding a "current frame" in the database object, a dag of frames is created based on the "base frame" held in `AristoDbRef` and all database access happens through this frame, which can be thought of as a consistent point-in-time snapshot of the database based on a particular fork of the chain. In the code, "frame", "transaction" and "layer" is used to denote more or less the same thing: a dag of stacked changes backed by the on-disk database. Although this is not a requirement, in practice each frame holds the change set of a single block - as such, the frame and its ancestors leading up to the on-disk state represents the state of the database after that block has been applied. "committing" means merging the changes to its parent frame so that the difference between them is lost and only the cumulative changes remain - this facility enables frames to be combined arbitrarily wherever they are in the dag. In particular, it becomes possible to consolidate a set of changes near the base of the dag and commit those to disk without having to re-do the in-memory frames built on top of them - this is useful for "flattening" a set of changes during a base update and sending those to storage without having to perform a block replay on top. Looking at abstractions, a side effect of this change is that the KVT and Aristo are brought closer together by considering them to be part of the "same" atomic transaction set - the way the code gets organised, applying a block and saving it to the kvt happens in the same "logical" frame - therefore, discarding the frame discards both the aristo and kvt changes at the same time - likewise, they are persisted to disk together - this makes reasoning about the database somewhat easier but has the downside of increased memory usage, something that perhaps will need addressing in the future. Because the code reasons more strictly about frames and the state of the persisted database, it also makes it more visible where ForkedChain should be used and where it is still missing - in particular, frames represent a single branch of history while forkedchain manages multiple parallel forks - user-facing services such as the RPC should use the latter, ie until it has been finalized, a getBlock request should consider all forks and not just the blocks in the canonical head branch. Another advantage of this approach is that `AristoDbRef` conceptually becomes more simple - removing its tracking of the "current" transaction stack simplifies reasoning about what can go wrong since this state now has to be passed around in the form of `AristoTxRef` - as such, many of the tests and facilities in the code that were dealing with "stack inconsistency" are now structurally prevented from happening. The test suite will need significant refactoring after this change. Once this change has been merged, there are several follow-ups to do: * there's no mechanism for keeping frames up to date as they get committed or rolled back - TODO * naming is confused - many names for the same thing for legacy reason * forkedchain support is still missing in lots of code * clean up redundant logic based on previous designs - in particular the debug and introspection code no longer makes sense * the way change sets are stored will probably need revisiting - because it's a stack of changes where each frame must be interrogated to find an on-disk value, with a base distance of 128 we'll at minimum have to perform 128 frame lookups for *every* database interaction - regardless, the "dag-like" nature will stay * dispose and commit are poorly defined and perhaps redundant - in theory, one could simply let the GC collect abandoned frames etc, though it's likely an explicit mechanism will remain useful, so they stay for now More about the changes: * `AristoDbRef` gains a `txRef` field (todo: rename) that "more or less" corresponds to the old `balancer` field * `AristoDbRef.stack` is gone - instead, there's a chain of `AristoTxRef` objects that hold their respective "layer" which has the actual changes * No more reasoning about "top" and "stack" - instead, each `AristoTxRef` can be a "head" that "more or less" corresponds to the old single-history `top` notion and its stack * `level` still represents "distance to base" - it's computed from the parent chain instead of being stored * one has to be careful not to use frames where forkedchain was intended - layers are only for a single branch of history! * fix layer vtop after rollback * engine fix * Fix test_txpool * Fix test_rpc * Fix copyright year * fix simulator * Fix copyright year * Fix copyright year * Fix tracer * Fix infinite recursion bug * Remove aristo and kvt empty files * Fic copyright year * Fix fc chain_kvt * ForkedChain refactoring * Fix merge master conflict * Fix copyright year * Reparent txFrame * Fix test * Fix txFrame reparent again * Cleanup and fix test * UpdateBase bugfix and fix test * Fixe newPayload bug discovered by hive * Fix engine api fcu * Clean up call template, chain_kvt, andn txguid * Fix copyright year * work around base block loading issue * Add test * Fix updateHead bug * Fix updateBase bug * Change func commitBase to proc commitBase * Touch up and fix debug mode crash --------- Co-authored-by: jangko <jangko128@gmail.com>
2025-02-06 08:04:50 +01:00
let dbTx = env.xdb.ctx.txFrameBegin(nil)
defer: dbTx.dispose()
aristo: fork support via layers/txframes (#2960) * aristo: fork support via layers/txframes This change reorganises how the database is accessed: instead holding a "current frame" in the database object, a dag of frames is created based on the "base frame" held in `AristoDbRef` and all database access happens through this frame, which can be thought of as a consistent point-in-time snapshot of the database based on a particular fork of the chain. In the code, "frame", "transaction" and "layer" is used to denote more or less the same thing: a dag of stacked changes backed by the on-disk database. Although this is not a requirement, in practice each frame holds the change set of a single block - as such, the frame and its ancestors leading up to the on-disk state represents the state of the database after that block has been applied. "committing" means merging the changes to its parent frame so that the difference between them is lost and only the cumulative changes remain - this facility enables frames to be combined arbitrarily wherever they are in the dag. In particular, it becomes possible to consolidate a set of changes near the base of the dag and commit those to disk without having to re-do the in-memory frames built on top of them - this is useful for "flattening" a set of changes during a base update and sending those to storage without having to perform a block replay on top. Looking at abstractions, a side effect of this change is that the KVT and Aristo are brought closer together by considering them to be part of the "same" atomic transaction set - the way the code gets organised, applying a block and saving it to the kvt happens in the same "logical" frame - therefore, discarding the frame discards both the aristo and kvt changes at the same time - likewise, they are persisted to disk together - this makes reasoning about the database somewhat easier but has the downside of increased memory usage, something that perhaps will need addressing in the future. Because the code reasons more strictly about frames and the state of the persisted database, it also makes it more visible where ForkedChain should be used and where it is still missing - in particular, frames represent a single branch of history while forkedchain manages multiple parallel forks - user-facing services such as the RPC should use the latter, ie until it has been finalized, a getBlock request should consider all forks and not just the blocks in the canonical head branch. Another advantage of this approach is that `AristoDbRef` conceptually becomes more simple - removing its tracking of the "current" transaction stack simplifies reasoning about what can go wrong since this state now has to be passed around in the form of `AristoTxRef` - as such, many of the tests and facilities in the code that were dealing with "stack inconsistency" are now structurally prevented from happening. The test suite will need significant refactoring after this change. Once this change has been merged, there are several follow-ups to do: * there's no mechanism for keeping frames up to date as they get committed or rolled back - TODO * naming is confused - many names for the same thing for legacy reason * forkedchain support is still missing in lots of code * clean up redundant logic based on previous designs - in particular the debug and introspection code no longer makes sense * the way change sets are stored will probably need revisiting - because it's a stack of changes where each frame must be interrogated to find an on-disk value, with a base distance of 128 we'll at minimum have to perform 128 frame lookups for *every* database interaction - regardless, the "dag-like" nature will stay * dispose and commit are poorly defined and perhaps redundant - in theory, one could simply let the GC collect abandoned frames etc, though it's likely an explicit mechanism will remain useful, so they stay for now More about the changes: * `AristoDbRef` gains a `txRef` field (todo: rename) that "more or less" corresponds to the old `balancer` field * `AristoDbRef.stack` is gone - instead, there's a chain of `AristoTxRef` objects that hold their respective "layer" which has the actual changes * No more reasoning about "top" and "stack" - instead, each `AristoTxRef` can be a "head" that "more or less" corresponds to the old single-history `top` notion and its stack * `level` still represents "distance to base" - it's computed from the parent chain instead of being stored * one has to be careful not to use frames where forkedchain was intended - layers are only for a single branch of history! * fix layer vtop after rollback * engine fix * Fix test_txpool * Fix test_rpc * Fix copyright year * fix simulator * Fix copyright year * Fix copyright year * Fix tracer * Fix infinite recursion bug * Remove aristo and kvt empty files * Fic copyright year * Fix fc chain_kvt * ForkedChain refactoring * Fix merge master conflict * Fix copyright year * Reparent txFrame * Fix test * Fix txFrame reparent again * Cleanup and fix test * UpdateBase bugfix and fix test * Fixe newPayload bug discovered by hive * Fix engine api fcu * Clean up call template, chain_kvt, andn txguid * Fix copyright year * work around base block loading issue * Add test * Fix updateHead bug * Fix updateBase bug * Change func commitBase to proc commitBase * Touch up and fix debug mode crash --------- Co-authored-by: jangko <jangko128@gmail.com>
2025-02-06 08:04:50 +01:00
let ledger = dbTx.getLedger()
env.runTrial2ok(ledger, n)
test &"Run {env.txi.len} three-step trials with rollback":
for n in env.txi:
aristo: fork support via layers/txframes (#2960) * aristo: fork support via layers/txframes This change reorganises how the database is accessed: instead holding a "current frame" in the database object, a dag of frames is created based on the "base frame" held in `AristoDbRef` and all database access happens through this frame, which can be thought of as a consistent point-in-time snapshot of the database based on a particular fork of the chain. In the code, "frame", "transaction" and "layer" is used to denote more or less the same thing: a dag of stacked changes backed by the on-disk database. Although this is not a requirement, in practice each frame holds the change set of a single block - as such, the frame and its ancestors leading up to the on-disk state represents the state of the database after that block has been applied. "committing" means merging the changes to its parent frame so that the difference between them is lost and only the cumulative changes remain - this facility enables frames to be combined arbitrarily wherever they are in the dag. In particular, it becomes possible to consolidate a set of changes near the base of the dag and commit those to disk without having to re-do the in-memory frames built on top of them - this is useful for "flattening" a set of changes during a base update and sending those to storage without having to perform a block replay on top. Looking at abstractions, a side effect of this change is that the KVT and Aristo are brought closer together by considering them to be part of the "same" atomic transaction set - the way the code gets organised, applying a block and saving it to the kvt happens in the same "logical" frame - therefore, discarding the frame discards both the aristo and kvt changes at the same time - likewise, they are persisted to disk together - this makes reasoning about the database somewhat easier but has the downside of increased memory usage, something that perhaps will need addressing in the future. Because the code reasons more strictly about frames and the state of the persisted database, it also makes it more visible where ForkedChain should be used and where it is still missing - in particular, frames represent a single branch of history while forkedchain manages multiple parallel forks - user-facing services such as the RPC should use the latter, ie until it has been finalized, a getBlock request should consider all forks and not just the blocks in the canonical head branch. Another advantage of this approach is that `AristoDbRef` conceptually becomes more simple - removing its tracking of the "current" transaction stack simplifies reasoning about what can go wrong since this state now has to be passed around in the form of `AristoTxRef` - as such, many of the tests and facilities in the code that were dealing with "stack inconsistency" are now structurally prevented from happening. The test suite will need significant refactoring after this change. Once this change has been merged, there are several follow-ups to do: * there's no mechanism for keeping frames up to date as they get committed or rolled back - TODO * naming is confused - many names for the same thing for legacy reason * forkedchain support is still missing in lots of code * clean up redundant logic based on previous designs - in particular the debug and introspection code no longer makes sense * the way change sets are stored will probably need revisiting - because it's a stack of changes where each frame must be interrogated to find an on-disk value, with a base distance of 128 we'll at minimum have to perform 128 frame lookups for *every* database interaction - regardless, the "dag-like" nature will stay * dispose and commit are poorly defined and perhaps redundant - in theory, one could simply let the GC collect abandoned frames etc, though it's likely an explicit mechanism will remain useful, so they stay for now More about the changes: * `AristoDbRef` gains a `txRef` field (todo: rename) that "more or less" corresponds to the old `balancer` field * `AristoDbRef.stack` is gone - instead, there's a chain of `AristoTxRef` objects that hold their respective "layer" which has the actual changes * No more reasoning about "top" and "stack" - instead, each `AristoTxRef` can be a "head" that "more or less" corresponds to the old single-history `top` notion and its stack * `level` still represents "distance to base" - it's computed from the parent chain instead of being stored * one has to be careful not to use frames where forkedchain was intended - layers are only for a single branch of history! * fix layer vtop after rollback * engine fix * Fix test_txpool * Fix test_rpc * Fix copyright year * fix simulator * Fix copyright year * Fix copyright year * Fix tracer * Fix infinite recursion bug * Remove aristo and kvt empty files * Fic copyright year * Fix fc chain_kvt * ForkedChain refactoring * Fix merge master conflict * Fix copyright year * Reparent txFrame * Fix test * Fix txFrame reparent again * Cleanup and fix test * UpdateBase bugfix and fix test * Fixe newPayload bug discovered by hive * Fix engine api fcu * Clean up call template, chain_kvt, andn txguid * Fix copyright year * work around base block loading issue * Add test * Fix updateHead bug * Fix updateBase bug * Change func commitBase to proc commitBase * Touch up and fix debug mode crash --------- Co-authored-by: jangko <jangko128@gmail.com>
2025-02-06 08:04:50 +01:00
let dbTx = env.xdb.ctx.txFrameBegin(nil)
defer: dbTx.dispose()
aristo: fork support via layers/txframes (#2960) * aristo: fork support via layers/txframes This change reorganises how the database is accessed: instead holding a "current frame" in the database object, a dag of frames is created based on the "base frame" held in `AristoDbRef` and all database access happens through this frame, which can be thought of as a consistent point-in-time snapshot of the database based on a particular fork of the chain. In the code, "frame", "transaction" and "layer" is used to denote more or less the same thing: a dag of stacked changes backed by the on-disk database. Although this is not a requirement, in practice each frame holds the change set of a single block - as such, the frame and its ancestors leading up to the on-disk state represents the state of the database after that block has been applied. "committing" means merging the changes to its parent frame so that the difference between them is lost and only the cumulative changes remain - this facility enables frames to be combined arbitrarily wherever they are in the dag. In particular, it becomes possible to consolidate a set of changes near the base of the dag and commit those to disk without having to re-do the in-memory frames built on top of them - this is useful for "flattening" a set of changes during a base update and sending those to storage without having to perform a block replay on top. Looking at abstractions, a side effect of this change is that the KVT and Aristo are brought closer together by considering them to be part of the "same" atomic transaction set - the way the code gets organised, applying a block and saving it to the kvt happens in the same "logical" frame - therefore, discarding the frame discards both the aristo and kvt changes at the same time - likewise, they are persisted to disk together - this makes reasoning about the database somewhat easier but has the downside of increased memory usage, something that perhaps will need addressing in the future. Because the code reasons more strictly about frames and the state of the persisted database, it also makes it more visible where ForkedChain should be used and where it is still missing - in particular, frames represent a single branch of history while forkedchain manages multiple parallel forks - user-facing services such as the RPC should use the latter, ie until it has been finalized, a getBlock request should consider all forks and not just the blocks in the canonical head branch. Another advantage of this approach is that `AristoDbRef` conceptually becomes more simple - removing its tracking of the "current" transaction stack simplifies reasoning about what can go wrong since this state now has to be passed around in the form of `AristoTxRef` - as such, many of the tests and facilities in the code that were dealing with "stack inconsistency" are now structurally prevented from happening. The test suite will need significant refactoring after this change. Once this change has been merged, there are several follow-ups to do: * there's no mechanism for keeping frames up to date as they get committed or rolled back - TODO * naming is confused - many names for the same thing for legacy reason * forkedchain support is still missing in lots of code * clean up redundant logic based on previous designs - in particular the debug and introspection code no longer makes sense * the way change sets are stored will probably need revisiting - because it's a stack of changes where each frame must be interrogated to find an on-disk value, with a base distance of 128 we'll at minimum have to perform 128 frame lookups for *every* database interaction - regardless, the "dag-like" nature will stay * dispose and commit are poorly defined and perhaps redundant - in theory, one could simply let the GC collect abandoned frames etc, though it's likely an explicit mechanism will remain useful, so they stay for now More about the changes: * `AristoDbRef` gains a `txRef` field (todo: rename) that "more or less" corresponds to the old `balancer` field * `AristoDbRef.stack` is gone - instead, there's a chain of `AristoTxRef` objects that hold their respective "layer" which has the actual changes * No more reasoning about "top" and "stack" - instead, each `AristoTxRef` can be a "head" that "more or less" corresponds to the old single-history `top` notion and its stack * `level` still represents "distance to base" - it's computed from the parent chain instead of being stored * one has to be careful not to use frames where forkedchain was intended - layers are only for a single branch of history! * fix layer vtop after rollback * engine fix * Fix test_txpool * Fix test_rpc * Fix copyright year * fix simulator * Fix copyright year * Fix copyright year * Fix tracer * Fix infinite recursion bug * Remove aristo and kvt empty files * Fic copyright year * Fix fc chain_kvt * ForkedChain refactoring * Fix merge master conflict * Fix copyright year * Reparent txFrame * Fix test * Fix txFrame reparent again * Cleanup and fix test * UpdateBase bugfix and fix test * Fixe newPayload bug discovered by hive * Fix engine api fcu * Clean up call template, chain_kvt, andn txguid * Fix copyright year * work around base block loading issue * Add test * Fix updateHead bug * Fix updateBase bug * Change func commitBase to proc commitBase * Touch up and fix debug mode crash --------- Co-authored-by: jangko <jangko128@gmail.com>
2025-02-06 08:04:50 +01:00
let ledger = dbTx.getLedger()
env.runTrial3(ledger, n, rollback = true)
test &"Run {env.txi.len} three-step trials with extra db frame rollback" &
" throwing Exceptions":
for n in env.txi:
aristo: fork support via layers/txframes (#2960) * aristo: fork support via layers/txframes This change reorganises how the database is accessed: instead holding a "current frame" in the database object, a dag of frames is created based on the "base frame" held in `AristoDbRef` and all database access happens through this frame, which can be thought of as a consistent point-in-time snapshot of the database based on a particular fork of the chain. In the code, "frame", "transaction" and "layer" is used to denote more or less the same thing: a dag of stacked changes backed by the on-disk database. Although this is not a requirement, in practice each frame holds the change set of a single block - as such, the frame and its ancestors leading up to the on-disk state represents the state of the database after that block has been applied. "committing" means merging the changes to its parent frame so that the difference between them is lost and only the cumulative changes remain - this facility enables frames to be combined arbitrarily wherever they are in the dag. In particular, it becomes possible to consolidate a set of changes near the base of the dag and commit those to disk without having to re-do the in-memory frames built on top of them - this is useful for "flattening" a set of changes during a base update and sending those to storage without having to perform a block replay on top. Looking at abstractions, a side effect of this change is that the KVT and Aristo are brought closer together by considering them to be part of the "same" atomic transaction set - the way the code gets organised, applying a block and saving it to the kvt happens in the same "logical" frame - therefore, discarding the frame discards both the aristo and kvt changes at the same time - likewise, they are persisted to disk together - this makes reasoning about the database somewhat easier but has the downside of increased memory usage, something that perhaps will need addressing in the future. Because the code reasons more strictly about frames and the state of the persisted database, it also makes it more visible where ForkedChain should be used and where it is still missing - in particular, frames represent a single branch of history while forkedchain manages multiple parallel forks - user-facing services such as the RPC should use the latter, ie until it has been finalized, a getBlock request should consider all forks and not just the blocks in the canonical head branch. Another advantage of this approach is that `AristoDbRef` conceptually becomes more simple - removing its tracking of the "current" transaction stack simplifies reasoning about what can go wrong since this state now has to be passed around in the form of `AristoTxRef` - as such, many of the tests and facilities in the code that were dealing with "stack inconsistency" are now structurally prevented from happening. The test suite will need significant refactoring after this change. Once this change has been merged, there are several follow-ups to do: * there's no mechanism for keeping frames up to date as they get committed or rolled back - TODO * naming is confused - many names for the same thing for legacy reason * forkedchain support is still missing in lots of code * clean up redundant logic based on previous designs - in particular the debug and introspection code no longer makes sense * the way change sets are stored will probably need revisiting - because it's a stack of changes where each frame must be interrogated to find an on-disk value, with a base distance of 128 we'll at minimum have to perform 128 frame lookups for *every* database interaction - regardless, the "dag-like" nature will stay * dispose and commit are poorly defined and perhaps redundant - in theory, one could simply let the GC collect abandoned frames etc, though it's likely an explicit mechanism will remain useful, so they stay for now More about the changes: * `AristoDbRef` gains a `txRef` field (todo: rename) that "more or less" corresponds to the old `balancer` field * `AristoDbRef.stack` is gone - instead, there's a chain of `AristoTxRef` objects that hold their respective "layer" which has the actual changes * No more reasoning about "top" and "stack" - instead, each `AristoTxRef` can be a "head" that "more or less" corresponds to the old single-history `top` notion and its stack * `level` still represents "distance to base" - it's computed from the parent chain instead of being stored * one has to be careful not to use frames where forkedchain was intended - layers are only for a single branch of history! * fix layer vtop after rollback * engine fix * Fix test_txpool * Fix test_rpc * Fix copyright year * fix simulator * Fix copyright year * Fix copyright year * Fix tracer * Fix infinite recursion bug * Remove aristo and kvt empty files * Fic copyright year * Fix fc chain_kvt * ForkedChain refactoring * Fix merge master conflict * Fix copyright year * Reparent txFrame * Fix test * Fix txFrame reparent again * Cleanup and fix test * UpdateBase bugfix and fix test * Fixe newPayload bug discovered by hive * Fix engine api fcu * Clean up call template, chain_kvt, andn txguid * Fix copyright year * work around base block loading issue * Add test * Fix updateHead bug * Fix updateBase bug * Change func commitBase to proc commitBase * Touch up and fix debug mode crash --------- Co-authored-by: jangko <jangko128@gmail.com>
2025-02-06 08:04:50 +01:00
let dbTx = env.xdb.ctx.txFrameBegin(nil)
defer: dbTx.dispose()
aristo: fork support via layers/txframes (#2960) * aristo: fork support via layers/txframes This change reorganises how the database is accessed: instead holding a "current frame" in the database object, a dag of frames is created based on the "base frame" held in `AristoDbRef` and all database access happens through this frame, which can be thought of as a consistent point-in-time snapshot of the database based on a particular fork of the chain. In the code, "frame", "transaction" and "layer" is used to denote more or less the same thing: a dag of stacked changes backed by the on-disk database. Although this is not a requirement, in practice each frame holds the change set of a single block - as such, the frame and its ancestors leading up to the on-disk state represents the state of the database after that block has been applied. "committing" means merging the changes to its parent frame so that the difference between them is lost and only the cumulative changes remain - this facility enables frames to be combined arbitrarily wherever they are in the dag. In particular, it becomes possible to consolidate a set of changes near the base of the dag and commit those to disk without having to re-do the in-memory frames built on top of them - this is useful for "flattening" a set of changes during a base update and sending those to storage without having to perform a block replay on top. Looking at abstractions, a side effect of this change is that the KVT and Aristo are brought closer together by considering them to be part of the "same" atomic transaction set - the way the code gets organised, applying a block and saving it to the kvt happens in the same "logical" frame - therefore, discarding the frame discards both the aristo and kvt changes at the same time - likewise, they are persisted to disk together - this makes reasoning about the database somewhat easier but has the downside of increased memory usage, something that perhaps will need addressing in the future. Because the code reasons more strictly about frames and the state of the persisted database, it also makes it more visible where ForkedChain should be used and where it is still missing - in particular, frames represent a single branch of history while forkedchain manages multiple parallel forks - user-facing services such as the RPC should use the latter, ie until it has been finalized, a getBlock request should consider all forks and not just the blocks in the canonical head branch. Another advantage of this approach is that `AristoDbRef` conceptually becomes more simple - removing its tracking of the "current" transaction stack simplifies reasoning about what can go wrong since this state now has to be passed around in the form of `AristoTxRef` - as such, many of the tests and facilities in the code that were dealing with "stack inconsistency" are now structurally prevented from happening. The test suite will need significant refactoring after this change. Once this change has been merged, there are several follow-ups to do: * there's no mechanism for keeping frames up to date as they get committed or rolled back - TODO * naming is confused - many names for the same thing for legacy reason * forkedchain support is still missing in lots of code * clean up redundant logic based on previous designs - in particular the debug and introspection code no longer makes sense * the way change sets are stored will probably need revisiting - because it's a stack of changes where each frame must be interrogated to find an on-disk value, with a base distance of 128 we'll at minimum have to perform 128 frame lookups for *every* database interaction - regardless, the "dag-like" nature will stay * dispose and commit are poorly defined and perhaps redundant - in theory, one could simply let the GC collect abandoned frames etc, though it's likely an explicit mechanism will remain useful, so they stay for now More about the changes: * `AristoDbRef` gains a `txRef` field (todo: rename) that "more or less" corresponds to the old `balancer` field * `AristoDbRef.stack` is gone - instead, there's a chain of `AristoTxRef` objects that hold their respective "layer" which has the actual changes * No more reasoning about "top" and "stack" - instead, each `AristoTxRef` can be a "head" that "more or less" corresponds to the old single-history `top` notion and its stack * `level` still represents "distance to base" - it's computed from the parent chain instead of being stored * one has to be careful not to use frames where forkedchain was intended - layers are only for a single branch of history! * fix layer vtop after rollback * engine fix * Fix test_txpool * Fix test_rpc * Fix copyright year * fix simulator * Fix copyright year * Fix copyright year * Fix tracer * Fix infinite recursion bug * Remove aristo and kvt empty files * Fic copyright year * Fix fc chain_kvt * ForkedChain refactoring * Fix merge master conflict * Fix copyright year * Reparent txFrame * Fix test * Fix txFrame reparent again * Cleanup and fix test * UpdateBase bugfix and fix test * Fixe newPayload bug discovered by hive * Fix engine api fcu * Clean up call template, chain_kvt, andn txguid * Fix copyright year * work around base block loading issue * Add test * Fix updateHead bug * Fix updateBase bug * Change func commitBase to proc commitBase * Touch up and fix debug mode crash --------- Co-authored-by: jangko <jangko128@gmail.com>
2025-02-06 08:04:50 +01:00
let ledger = dbTx.getLedger()
env.runTrial3Survive(ledger, n, noisy)
test &"Run {env.txi.len} tree-step trials without rollback":
for n in env.txi:
aristo: fork support via layers/txframes (#2960) * aristo: fork support via layers/txframes This change reorganises how the database is accessed: instead holding a "current frame" in the database object, a dag of frames is created based on the "base frame" held in `AristoDbRef` and all database access happens through this frame, which can be thought of as a consistent point-in-time snapshot of the database based on a particular fork of the chain. In the code, "frame", "transaction" and "layer" is used to denote more or less the same thing: a dag of stacked changes backed by the on-disk database. Although this is not a requirement, in practice each frame holds the change set of a single block - as such, the frame and its ancestors leading up to the on-disk state represents the state of the database after that block has been applied. "committing" means merging the changes to its parent frame so that the difference between them is lost and only the cumulative changes remain - this facility enables frames to be combined arbitrarily wherever they are in the dag. In particular, it becomes possible to consolidate a set of changes near the base of the dag and commit those to disk without having to re-do the in-memory frames built on top of them - this is useful for "flattening" a set of changes during a base update and sending those to storage without having to perform a block replay on top. Looking at abstractions, a side effect of this change is that the KVT and Aristo are brought closer together by considering them to be part of the "same" atomic transaction set - the way the code gets organised, applying a block and saving it to the kvt happens in the same "logical" frame - therefore, discarding the frame discards both the aristo and kvt changes at the same time - likewise, they are persisted to disk together - this makes reasoning about the database somewhat easier but has the downside of increased memory usage, something that perhaps will need addressing in the future. Because the code reasons more strictly about frames and the state of the persisted database, it also makes it more visible where ForkedChain should be used and where it is still missing - in particular, frames represent a single branch of history while forkedchain manages multiple parallel forks - user-facing services such as the RPC should use the latter, ie until it has been finalized, a getBlock request should consider all forks and not just the blocks in the canonical head branch. Another advantage of this approach is that `AristoDbRef` conceptually becomes more simple - removing its tracking of the "current" transaction stack simplifies reasoning about what can go wrong since this state now has to be passed around in the form of `AristoTxRef` - as such, many of the tests and facilities in the code that were dealing with "stack inconsistency" are now structurally prevented from happening. The test suite will need significant refactoring after this change. Once this change has been merged, there are several follow-ups to do: * there's no mechanism for keeping frames up to date as they get committed or rolled back - TODO * naming is confused - many names for the same thing for legacy reason * forkedchain support is still missing in lots of code * clean up redundant logic based on previous designs - in particular the debug and introspection code no longer makes sense * the way change sets are stored will probably need revisiting - because it's a stack of changes where each frame must be interrogated to find an on-disk value, with a base distance of 128 we'll at minimum have to perform 128 frame lookups for *every* database interaction - regardless, the "dag-like" nature will stay * dispose and commit are poorly defined and perhaps redundant - in theory, one could simply let the GC collect abandoned frames etc, though it's likely an explicit mechanism will remain useful, so they stay for now More about the changes: * `AristoDbRef` gains a `txRef` field (todo: rename) that "more or less" corresponds to the old `balancer` field * `AristoDbRef.stack` is gone - instead, there's a chain of `AristoTxRef` objects that hold their respective "layer" which has the actual changes * No more reasoning about "top" and "stack" - instead, each `AristoTxRef` can be a "head" that "more or less" corresponds to the old single-history `top` notion and its stack * `level` still represents "distance to base" - it's computed from the parent chain instead of being stored * one has to be careful not to use frames where forkedchain was intended - layers are only for a single branch of history! * fix layer vtop after rollback * engine fix * Fix test_txpool * Fix test_rpc * Fix copyright year * fix simulator * Fix copyright year * Fix copyright year * Fix tracer * Fix infinite recursion bug * Remove aristo and kvt empty files * Fic copyright year * Fix fc chain_kvt * ForkedChain refactoring * Fix merge master conflict * Fix copyright year * Reparent txFrame * Fix test * Fix txFrame reparent again * Cleanup and fix test * UpdateBase bugfix and fix test * Fixe newPayload bug discovered by hive * Fix engine api fcu * Clean up call template, chain_kvt, andn txguid * Fix copyright year * work around base block loading issue * Add test * Fix updateHead bug * Fix updateBase bug * Change func commitBase to proc commitBase * Touch up and fix debug mode crash --------- Co-authored-by: jangko <jangko128@gmail.com>
2025-02-06 08:04:50 +01:00
let dbTx = env.xdb.ctx.txFrameBegin(nil)
defer: dbTx.dispose()
aristo: fork support via layers/txframes (#2960) * aristo: fork support via layers/txframes This change reorganises how the database is accessed: instead holding a "current frame" in the database object, a dag of frames is created based on the "base frame" held in `AristoDbRef` and all database access happens through this frame, which can be thought of as a consistent point-in-time snapshot of the database based on a particular fork of the chain. In the code, "frame", "transaction" and "layer" is used to denote more or less the same thing: a dag of stacked changes backed by the on-disk database. Although this is not a requirement, in practice each frame holds the change set of a single block - as such, the frame and its ancestors leading up to the on-disk state represents the state of the database after that block has been applied. "committing" means merging the changes to its parent frame so that the difference between them is lost and only the cumulative changes remain - this facility enables frames to be combined arbitrarily wherever they are in the dag. In particular, it becomes possible to consolidate a set of changes near the base of the dag and commit those to disk without having to re-do the in-memory frames built on top of them - this is useful for "flattening" a set of changes during a base update and sending those to storage without having to perform a block replay on top. Looking at abstractions, a side effect of this change is that the KVT and Aristo are brought closer together by considering them to be part of the "same" atomic transaction set - the way the code gets organised, applying a block and saving it to the kvt happens in the same "logical" frame - therefore, discarding the frame discards both the aristo and kvt changes at the same time - likewise, they are persisted to disk together - this makes reasoning about the database somewhat easier but has the downside of increased memory usage, something that perhaps will need addressing in the future. Because the code reasons more strictly about frames and the state of the persisted database, it also makes it more visible where ForkedChain should be used and where it is still missing - in particular, frames represent a single branch of history while forkedchain manages multiple parallel forks - user-facing services such as the RPC should use the latter, ie until it has been finalized, a getBlock request should consider all forks and not just the blocks in the canonical head branch. Another advantage of this approach is that `AristoDbRef` conceptually becomes more simple - removing its tracking of the "current" transaction stack simplifies reasoning about what can go wrong since this state now has to be passed around in the form of `AristoTxRef` - as such, many of the tests and facilities in the code that were dealing with "stack inconsistency" are now structurally prevented from happening. The test suite will need significant refactoring after this change. Once this change has been merged, there are several follow-ups to do: * there's no mechanism for keeping frames up to date as they get committed or rolled back - TODO * naming is confused - many names for the same thing for legacy reason * forkedchain support is still missing in lots of code * clean up redundant logic based on previous designs - in particular the debug and introspection code no longer makes sense * the way change sets are stored will probably need revisiting - because it's a stack of changes where each frame must be interrogated to find an on-disk value, with a base distance of 128 we'll at minimum have to perform 128 frame lookups for *every* database interaction - regardless, the "dag-like" nature will stay * dispose and commit are poorly defined and perhaps redundant - in theory, one could simply let the GC collect abandoned frames etc, though it's likely an explicit mechanism will remain useful, so they stay for now More about the changes: * `AristoDbRef` gains a `txRef` field (todo: rename) that "more or less" corresponds to the old `balancer` field * `AristoDbRef.stack` is gone - instead, there's a chain of `AristoTxRef` objects that hold their respective "layer" which has the actual changes * No more reasoning about "top" and "stack" - instead, each `AristoTxRef` can be a "head" that "more or less" corresponds to the old single-history `top` notion and its stack * `level` still represents "distance to base" - it's computed from the parent chain instead of being stored * one has to be careful not to use frames where forkedchain was intended - layers are only for a single branch of history! * fix layer vtop after rollback * engine fix * Fix test_txpool * Fix test_rpc * Fix copyright year * fix simulator * Fix copyright year * Fix copyright year * Fix tracer * Fix infinite recursion bug * Remove aristo and kvt empty files * Fic copyright year * Fix fc chain_kvt * ForkedChain refactoring * Fix merge master conflict * Fix copyright year * Reparent txFrame * Fix test * Fix txFrame reparent again * Cleanup and fix test * UpdateBase bugfix and fix test * Fixe newPayload bug discovered by hive * Fix engine api fcu * Clean up call template, chain_kvt, andn txguid * Fix copyright year * work around base block loading issue * Add test * Fix updateHead bug * Fix updateBase bug * Change func commitBase to proc commitBase * Touch up and fix debug mode crash --------- Co-authored-by: jangko <jangko128@gmail.com>
2025-02-06 08:04:50 +01:00
let ledger = dbTx.getLedger()
env.runTrial3(ledger, n, rollback = false)
test &"Run {env.txi.len} four-step trials with rollback and db frames":
for n in env.txi:
aristo: fork support via layers/txframes (#2960) * aristo: fork support via layers/txframes This change reorganises how the database is accessed: instead holding a "current frame" in the database object, a dag of frames is created based on the "base frame" held in `AristoDbRef` and all database access happens through this frame, which can be thought of as a consistent point-in-time snapshot of the database based on a particular fork of the chain. In the code, "frame", "transaction" and "layer" is used to denote more or less the same thing: a dag of stacked changes backed by the on-disk database. Although this is not a requirement, in practice each frame holds the change set of a single block - as such, the frame and its ancestors leading up to the on-disk state represents the state of the database after that block has been applied. "committing" means merging the changes to its parent frame so that the difference between them is lost and only the cumulative changes remain - this facility enables frames to be combined arbitrarily wherever they are in the dag. In particular, it becomes possible to consolidate a set of changes near the base of the dag and commit those to disk without having to re-do the in-memory frames built on top of them - this is useful for "flattening" a set of changes during a base update and sending those to storage without having to perform a block replay on top. Looking at abstractions, a side effect of this change is that the KVT and Aristo are brought closer together by considering them to be part of the "same" atomic transaction set - the way the code gets organised, applying a block and saving it to the kvt happens in the same "logical" frame - therefore, discarding the frame discards both the aristo and kvt changes at the same time - likewise, they are persisted to disk together - this makes reasoning about the database somewhat easier but has the downside of increased memory usage, something that perhaps will need addressing in the future. Because the code reasons more strictly about frames and the state of the persisted database, it also makes it more visible where ForkedChain should be used and where it is still missing - in particular, frames represent a single branch of history while forkedchain manages multiple parallel forks - user-facing services such as the RPC should use the latter, ie until it has been finalized, a getBlock request should consider all forks and not just the blocks in the canonical head branch. Another advantage of this approach is that `AristoDbRef` conceptually becomes more simple - removing its tracking of the "current" transaction stack simplifies reasoning about what can go wrong since this state now has to be passed around in the form of `AristoTxRef` - as such, many of the tests and facilities in the code that were dealing with "stack inconsistency" are now structurally prevented from happening. The test suite will need significant refactoring after this change. Once this change has been merged, there are several follow-ups to do: * there's no mechanism for keeping frames up to date as they get committed or rolled back - TODO * naming is confused - many names for the same thing for legacy reason * forkedchain support is still missing in lots of code * clean up redundant logic based on previous designs - in particular the debug and introspection code no longer makes sense * the way change sets are stored will probably need revisiting - because it's a stack of changes where each frame must be interrogated to find an on-disk value, with a base distance of 128 we'll at minimum have to perform 128 frame lookups for *every* database interaction - regardless, the "dag-like" nature will stay * dispose and commit are poorly defined and perhaps redundant - in theory, one could simply let the GC collect abandoned frames etc, though it's likely an explicit mechanism will remain useful, so they stay for now More about the changes: * `AristoDbRef` gains a `txRef` field (todo: rename) that "more or less" corresponds to the old `balancer` field * `AristoDbRef.stack` is gone - instead, there's a chain of `AristoTxRef` objects that hold their respective "layer" which has the actual changes * No more reasoning about "top" and "stack" - instead, each `AristoTxRef` can be a "head" that "more or less" corresponds to the old single-history `top` notion and its stack * `level` still represents "distance to base" - it's computed from the parent chain instead of being stored * one has to be careful not to use frames where forkedchain was intended - layers are only for a single branch of history! * fix layer vtop after rollback * engine fix * Fix test_txpool * Fix test_rpc * Fix copyright year * fix simulator * Fix copyright year * Fix copyright year * Fix tracer * Fix infinite recursion bug * Remove aristo and kvt empty files * Fic copyright year * Fix fc chain_kvt * ForkedChain refactoring * Fix merge master conflict * Fix copyright year * Reparent txFrame * Fix test * Fix txFrame reparent again * Cleanup and fix test * UpdateBase bugfix and fix test * Fixe newPayload bug discovered by hive * Fix engine api fcu * Clean up call template, chain_kvt, andn txguid * Fix copyright year * work around base block loading issue * Add test * Fix updateHead bug * Fix updateBase bug * Change func commitBase to proc commitBase * Touch up and fix debug mode crash --------- Co-authored-by: jangko <jangko128@gmail.com>
2025-02-06 08:04:50 +01:00
let dbTx = env.xdb.ctx.txFrameBegin(nil)
defer: dbTx.dispose()
aristo: fork support via layers/txframes (#2960) * aristo: fork support via layers/txframes This change reorganises how the database is accessed: instead holding a "current frame" in the database object, a dag of frames is created based on the "base frame" held in `AristoDbRef` and all database access happens through this frame, which can be thought of as a consistent point-in-time snapshot of the database based on a particular fork of the chain. In the code, "frame", "transaction" and "layer" is used to denote more or less the same thing: a dag of stacked changes backed by the on-disk database. Although this is not a requirement, in practice each frame holds the change set of a single block - as such, the frame and its ancestors leading up to the on-disk state represents the state of the database after that block has been applied. "committing" means merging the changes to its parent frame so that the difference between them is lost and only the cumulative changes remain - this facility enables frames to be combined arbitrarily wherever they are in the dag. In particular, it becomes possible to consolidate a set of changes near the base of the dag and commit those to disk without having to re-do the in-memory frames built on top of them - this is useful for "flattening" a set of changes during a base update and sending those to storage without having to perform a block replay on top. Looking at abstractions, a side effect of this change is that the KVT and Aristo are brought closer together by considering them to be part of the "same" atomic transaction set - the way the code gets organised, applying a block and saving it to the kvt happens in the same "logical" frame - therefore, discarding the frame discards both the aristo and kvt changes at the same time - likewise, they are persisted to disk together - this makes reasoning about the database somewhat easier but has the downside of increased memory usage, something that perhaps will need addressing in the future. Because the code reasons more strictly about frames and the state of the persisted database, it also makes it more visible where ForkedChain should be used and where it is still missing - in particular, frames represent a single branch of history while forkedchain manages multiple parallel forks - user-facing services such as the RPC should use the latter, ie until it has been finalized, a getBlock request should consider all forks and not just the blocks in the canonical head branch. Another advantage of this approach is that `AristoDbRef` conceptually becomes more simple - removing its tracking of the "current" transaction stack simplifies reasoning about what can go wrong since this state now has to be passed around in the form of `AristoTxRef` - as such, many of the tests and facilities in the code that were dealing with "stack inconsistency" are now structurally prevented from happening. The test suite will need significant refactoring after this change. Once this change has been merged, there are several follow-ups to do: * there's no mechanism for keeping frames up to date as they get committed or rolled back - TODO * naming is confused - many names for the same thing for legacy reason * forkedchain support is still missing in lots of code * clean up redundant logic based on previous designs - in particular the debug and introspection code no longer makes sense * the way change sets are stored will probably need revisiting - because it's a stack of changes where each frame must be interrogated to find an on-disk value, with a base distance of 128 we'll at minimum have to perform 128 frame lookups for *every* database interaction - regardless, the "dag-like" nature will stay * dispose and commit are poorly defined and perhaps redundant - in theory, one could simply let the GC collect abandoned frames etc, though it's likely an explicit mechanism will remain useful, so they stay for now More about the changes: * `AristoDbRef` gains a `txRef` field (todo: rename) that "more or less" corresponds to the old `balancer` field * `AristoDbRef.stack` is gone - instead, there's a chain of `AristoTxRef` objects that hold their respective "layer" which has the actual changes * No more reasoning about "top" and "stack" - instead, each `AristoTxRef` can be a "head" that "more or less" corresponds to the old single-history `top` notion and its stack * `level` still represents "distance to base" - it's computed from the parent chain instead of being stored * one has to be careful not to use frames where forkedchain was intended - layers are only for a single branch of history! * fix layer vtop after rollback * engine fix * Fix test_txpool * Fix test_rpc * Fix copyright year * fix simulator * Fix copyright year * Fix copyright year * Fix tracer * Fix infinite recursion bug * Remove aristo and kvt empty files * Fic copyright year * Fix fc chain_kvt * ForkedChain refactoring * Fix merge master conflict * Fix copyright year * Reparent txFrame * Fix test * Fix txFrame reparent again * Cleanup and fix test * UpdateBase bugfix and fix test * Fixe newPayload bug discovered by hive * Fix engine api fcu * Clean up call template, chain_kvt, andn txguid * Fix copyright year * work around base block loading issue * Add test * Fix updateHead bug * Fix updateBase bug * Change func commitBase to proc commitBase * Touch up and fix debug mode crash --------- Co-authored-by: jangko <jangko128@gmail.com>
2025-02-06 08:04:50 +01:00
let ledger = dbTx.getLedger()
env.runTrial4(ledger, n, rollback = true)
proc runLedgerBasicOperationsTests() =
suite "Ledger basic operations tests":
setup:
const emptyAcc {.used.} = Account.init()
var
memDB = newCoreDbRef DefaultDbMemory
aristo: fork support via layers/txframes (#2960) * aristo: fork support via layers/txframes This change reorganises how the database is accessed: instead holding a "current frame" in the database object, a dag of frames is created based on the "base frame" held in `AristoDbRef` and all database access happens through this frame, which can be thought of as a consistent point-in-time snapshot of the database based on a particular fork of the chain. In the code, "frame", "transaction" and "layer" is used to denote more or less the same thing: a dag of stacked changes backed by the on-disk database. Although this is not a requirement, in practice each frame holds the change set of a single block - as such, the frame and its ancestors leading up to the on-disk state represents the state of the database after that block has been applied. "committing" means merging the changes to its parent frame so that the difference between them is lost and only the cumulative changes remain - this facility enables frames to be combined arbitrarily wherever they are in the dag. In particular, it becomes possible to consolidate a set of changes near the base of the dag and commit those to disk without having to re-do the in-memory frames built on top of them - this is useful for "flattening" a set of changes during a base update and sending those to storage without having to perform a block replay on top. Looking at abstractions, a side effect of this change is that the KVT and Aristo are brought closer together by considering them to be part of the "same" atomic transaction set - the way the code gets organised, applying a block and saving it to the kvt happens in the same "logical" frame - therefore, discarding the frame discards both the aristo and kvt changes at the same time - likewise, they are persisted to disk together - this makes reasoning about the database somewhat easier but has the downside of increased memory usage, something that perhaps will need addressing in the future. Because the code reasons more strictly about frames and the state of the persisted database, it also makes it more visible where ForkedChain should be used and where it is still missing - in particular, frames represent a single branch of history while forkedchain manages multiple parallel forks - user-facing services such as the RPC should use the latter, ie until it has been finalized, a getBlock request should consider all forks and not just the blocks in the canonical head branch. Another advantage of this approach is that `AristoDbRef` conceptually becomes more simple - removing its tracking of the "current" transaction stack simplifies reasoning about what can go wrong since this state now has to be passed around in the form of `AristoTxRef` - as such, many of the tests and facilities in the code that were dealing with "stack inconsistency" are now structurally prevented from happening. The test suite will need significant refactoring after this change. Once this change has been merged, there are several follow-ups to do: * there's no mechanism for keeping frames up to date as they get committed or rolled back - TODO * naming is confused - many names for the same thing for legacy reason * forkedchain support is still missing in lots of code * clean up redundant logic based on previous designs - in particular the debug and introspection code no longer makes sense * the way change sets are stored will probably need revisiting - because it's a stack of changes where each frame must be interrogated to find an on-disk value, with a base distance of 128 we'll at minimum have to perform 128 frame lookups for *every* database interaction - regardless, the "dag-like" nature will stay * dispose and commit are poorly defined and perhaps redundant - in theory, one could simply let the GC collect abandoned frames etc, though it's likely an explicit mechanism will remain useful, so they stay for now More about the changes: * `AristoDbRef` gains a `txRef` field (todo: rename) that "more or less" corresponds to the old `balancer` field * `AristoDbRef.stack` is gone - instead, there's a chain of `AristoTxRef` objects that hold their respective "layer" which has the actual changes * No more reasoning about "top" and "stack" - instead, each `AristoTxRef` can be a "head" that "more or less" corresponds to the old single-history `top` notion and its stack * `level` still represents "distance to base" - it's computed from the parent chain instead of being stored * one has to be careful not to use frames where forkedchain was intended - layers are only for a single branch of history! * fix layer vtop after rollback * engine fix * Fix test_txpool * Fix test_rpc * Fix copyright year * fix simulator * Fix copyright year * Fix copyright year * Fix tracer * Fix infinite recursion bug * Remove aristo and kvt empty files * Fic copyright year * Fix fc chain_kvt * ForkedChain refactoring * Fix merge master conflict * Fix copyright year * Reparent txFrame * Fix test * Fix txFrame reparent again * Cleanup and fix test * UpdateBase bugfix and fix test * Fixe newPayload bug discovered by hive * Fix engine api fcu * Clean up call template, chain_kvt, andn txguid * Fix copyright year * work around base block loading issue * Add test * Fix updateHead bug * Fix updateBase bug * Change func commitBase to proc commitBase * Touch up and fix debug mode crash --------- Co-authored-by: jangko <jangko128@gmail.com>
2025-02-06 08:04:50 +01:00
ledger {.used.} = LedgerRef.init(memDB.baseTxFrame())
address {.used.} = address"0x0f572e5295c57f15886f9b263e2f6d2d6c7b5ec6"
code {.used.} = hexToSeqByte("0x0f572e5295c57f15886f9b263e2f6d2d6c7b5ec6")
stateRoot {.used.} : Hash32
test "accountExists and isDeadAccount":
check ledger.accountExists(address) == false
check ledger.isDeadAccount(address) == true
ledger.setBalance(address, 1000.u256)
check ledger.accountExists(address) == true
check ledger.isDeadAccount(address) == false
ledger.setBalance(address, 0.u256)
ledger.setNonce(address, 1)
check ledger.isDeadAccount(address) == false
ledger.setCode(address, code)
ledger.setNonce(address, 0)
check ledger.isDeadAccount(address) == false
ledger.setCode(address, newSeq[byte]())
check ledger.isDeadAccount(address) == true
check ledger.accountExists(address) == true
test "clone storage":
# give access to private fields of AccountRef
privateAccess(AccountRef)
var x = AccountRef(
overlayStorage: Table[UInt256, UInt256](),
originalStorage: newTable[UInt256, UInt256]()
)
x.overlayStorage[10.u256] = 11.u256
x.overlayStorage[11.u256] = 12.u256
x.originalStorage[10.u256] = 11.u256
x.originalStorage[11.u256] = 12.u256
var y = x.clone(cloneStorage = true)
y.overlayStorage[12.u256] = 13.u256
y.originalStorage[12.u256] = 13.u256
check 12.u256 notin x.overlayStorage
check 12.u256 in y.overlayStorage
check x.overlayStorage.len == 2
check y.overlayStorage.len == 3
check 12.u256 in x.originalStorage
check 12.u256 in y.originalStorage
check x.originalStorage.len == 3
check y.originalStorage.len == 3
test "Ledger various operations":
aristo: fork support via layers/txframes (#2960) * aristo: fork support via layers/txframes This change reorganises how the database is accessed: instead holding a "current frame" in the database object, a dag of frames is created based on the "base frame" held in `AristoDbRef` and all database access happens through this frame, which can be thought of as a consistent point-in-time snapshot of the database based on a particular fork of the chain. In the code, "frame", "transaction" and "layer" is used to denote more or less the same thing: a dag of stacked changes backed by the on-disk database. Although this is not a requirement, in practice each frame holds the change set of a single block - as such, the frame and its ancestors leading up to the on-disk state represents the state of the database after that block has been applied. "committing" means merging the changes to its parent frame so that the difference between them is lost and only the cumulative changes remain - this facility enables frames to be combined arbitrarily wherever they are in the dag. In particular, it becomes possible to consolidate a set of changes near the base of the dag and commit those to disk without having to re-do the in-memory frames built on top of them - this is useful for "flattening" a set of changes during a base update and sending those to storage without having to perform a block replay on top. Looking at abstractions, a side effect of this change is that the KVT and Aristo are brought closer together by considering them to be part of the "same" atomic transaction set - the way the code gets organised, applying a block and saving it to the kvt happens in the same "logical" frame - therefore, discarding the frame discards both the aristo and kvt changes at the same time - likewise, they are persisted to disk together - this makes reasoning about the database somewhat easier but has the downside of increased memory usage, something that perhaps will need addressing in the future. Because the code reasons more strictly about frames and the state of the persisted database, it also makes it more visible where ForkedChain should be used and where it is still missing - in particular, frames represent a single branch of history while forkedchain manages multiple parallel forks - user-facing services such as the RPC should use the latter, ie until it has been finalized, a getBlock request should consider all forks and not just the blocks in the canonical head branch. Another advantage of this approach is that `AristoDbRef` conceptually becomes more simple - removing its tracking of the "current" transaction stack simplifies reasoning about what can go wrong since this state now has to be passed around in the form of `AristoTxRef` - as such, many of the tests and facilities in the code that were dealing with "stack inconsistency" are now structurally prevented from happening. The test suite will need significant refactoring after this change. Once this change has been merged, there are several follow-ups to do: * there's no mechanism for keeping frames up to date as they get committed or rolled back - TODO * naming is confused - many names for the same thing for legacy reason * forkedchain support is still missing in lots of code * clean up redundant logic based on previous designs - in particular the debug and introspection code no longer makes sense * the way change sets are stored will probably need revisiting - because it's a stack of changes where each frame must be interrogated to find an on-disk value, with a base distance of 128 we'll at minimum have to perform 128 frame lookups for *every* database interaction - regardless, the "dag-like" nature will stay * dispose and commit are poorly defined and perhaps redundant - in theory, one could simply let the GC collect abandoned frames etc, though it's likely an explicit mechanism will remain useful, so they stay for now More about the changes: * `AristoDbRef` gains a `txRef` field (todo: rename) that "more or less" corresponds to the old `balancer` field * `AristoDbRef.stack` is gone - instead, there's a chain of `AristoTxRef` objects that hold their respective "layer" which has the actual changes * No more reasoning about "top" and "stack" - instead, each `AristoTxRef` can be a "head" that "more or less" corresponds to the old single-history `top` notion and its stack * `level` still represents "distance to base" - it's computed from the parent chain instead of being stored * one has to be careful not to use frames where forkedchain was intended - layers are only for a single branch of history! * fix layer vtop after rollback * engine fix * Fix test_txpool * Fix test_rpc * Fix copyright year * fix simulator * Fix copyright year * Fix copyright year * Fix tracer * Fix infinite recursion bug * Remove aristo and kvt empty files * Fic copyright year * Fix fc chain_kvt * ForkedChain refactoring * Fix merge master conflict * Fix copyright year * Reparent txFrame * Fix test * Fix txFrame reparent again * Cleanup and fix test * UpdateBase bugfix and fix test * Fixe newPayload bug discovered by hive * Fix engine api fcu * Clean up call template, chain_kvt, andn txguid * Fix copyright year * work around base block loading issue * Add test * Fix updateHead bug * Fix updateBase bug * Change func commitBase to proc commitBase * Touch up and fix debug mode crash --------- Co-authored-by: jangko <jangko128@gmail.com>
2025-02-06 08:04:50 +01:00
var ac = LedgerRef.init(memDB.baseTxFrame())
var addr1 = initAddr(1)
check ac.isDeadAccount(addr1) == true
check ac.accountExists(addr1) == false
check ac.contractCollision(addr1) == false
ac.setBalance(addr1, 1000.u256)
check ac.getBalance(addr1) == 1000.u256
ac.subBalance(addr1, 100.u256)
check ac.getBalance(addr1) == 900.u256
ac.addBalance(addr1, 200.u256)
check ac.getBalance(addr1) == 1100.u256
ac.setNonce(addr1, 1)
check ac.getNonce(addr1) == 1
ac.incNonce(addr1)
check ac.getNonce(addr1) == 2
ac.setCode(addr1, code)
check ac.getCode(addr1) == code
ac.setStorage(addr1, 1.u256, 10.u256)
check ac.getStorage(addr1, 1.u256) == 10.u256
check ac.getCommittedStorage(addr1, 1.u256) == 0.u256
check ac.contractCollision(addr1) == true
check ac.getCodeSize(addr1) == code.len
ac.persist()
stateRoot = ac.getStateRoot()
aristo: fork support via layers/txframes (#2960) * aristo: fork support via layers/txframes This change reorganises how the database is accessed: instead holding a "current frame" in the database object, a dag of frames is created based on the "base frame" held in `AristoDbRef` and all database access happens through this frame, which can be thought of as a consistent point-in-time snapshot of the database based on a particular fork of the chain. In the code, "frame", "transaction" and "layer" is used to denote more or less the same thing: a dag of stacked changes backed by the on-disk database. Although this is not a requirement, in practice each frame holds the change set of a single block - as such, the frame and its ancestors leading up to the on-disk state represents the state of the database after that block has been applied. "committing" means merging the changes to its parent frame so that the difference between them is lost and only the cumulative changes remain - this facility enables frames to be combined arbitrarily wherever they are in the dag. In particular, it becomes possible to consolidate a set of changes near the base of the dag and commit those to disk without having to re-do the in-memory frames built on top of them - this is useful for "flattening" a set of changes during a base update and sending those to storage without having to perform a block replay on top. Looking at abstractions, a side effect of this change is that the KVT and Aristo are brought closer together by considering them to be part of the "same" atomic transaction set - the way the code gets organised, applying a block and saving it to the kvt happens in the same "logical" frame - therefore, discarding the frame discards both the aristo and kvt changes at the same time - likewise, they are persisted to disk together - this makes reasoning about the database somewhat easier but has the downside of increased memory usage, something that perhaps will need addressing in the future. Because the code reasons more strictly about frames and the state of the persisted database, it also makes it more visible where ForkedChain should be used and where it is still missing - in particular, frames represent a single branch of history while forkedchain manages multiple parallel forks - user-facing services such as the RPC should use the latter, ie until it has been finalized, a getBlock request should consider all forks and not just the blocks in the canonical head branch. Another advantage of this approach is that `AristoDbRef` conceptually becomes more simple - removing its tracking of the "current" transaction stack simplifies reasoning about what can go wrong since this state now has to be passed around in the form of `AristoTxRef` - as such, many of the tests and facilities in the code that were dealing with "stack inconsistency" are now structurally prevented from happening. The test suite will need significant refactoring after this change. Once this change has been merged, there are several follow-ups to do: * there's no mechanism for keeping frames up to date as they get committed or rolled back - TODO * naming is confused - many names for the same thing for legacy reason * forkedchain support is still missing in lots of code * clean up redundant logic based on previous designs - in particular the debug and introspection code no longer makes sense * the way change sets are stored will probably need revisiting - because it's a stack of changes where each frame must be interrogated to find an on-disk value, with a base distance of 128 we'll at minimum have to perform 128 frame lookups for *every* database interaction - regardless, the "dag-like" nature will stay * dispose and commit are poorly defined and perhaps redundant - in theory, one could simply let the GC collect abandoned frames etc, though it's likely an explicit mechanism will remain useful, so they stay for now More about the changes: * `AristoDbRef` gains a `txRef` field (todo: rename) that "more or less" corresponds to the old `balancer` field * `AristoDbRef.stack` is gone - instead, there's a chain of `AristoTxRef` objects that hold their respective "layer" which has the actual changes * No more reasoning about "top" and "stack" - instead, each `AristoTxRef` can be a "head" that "more or less" corresponds to the old single-history `top` notion and its stack * `level` still represents "distance to base" - it's computed from the parent chain instead of being stored * one has to be careful not to use frames where forkedchain was intended - layers are only for a single branch of history! * fix layer vtop after rollback * engine fix * Fix test_txpool * Fix test_rpc * Fix copyright year * fix simulator * Fix copyright year * Fix copyright year * Fix tracer * Fix infinite recursion bug * Remove aristo and kvt empty files * Fic copyright year * Fix fc chain_kvt * ForkedChain refactoring * Fix merge master conflict * Fix copyright year * Reparent txFrame * Fix test * Fix txFrame reparent again * Cleanup and fix test * UpdateBase bugfix and fix test * Fixe newPayload bug discovered by hive * Fix engine api fcu * Clean up call template, chain_kvt, andn txguid * Fix copyright year * work around base block loading issue * Add test * Fix updateHead bug * Fix updateBase bug * Change func commitBase to proc commitBase * Touch up and fix debug mode crash --------- Co-authored-by: jangko <jangko128@gmail.com>
2025-02-06 08:04:50 +01:00
var db = LedgerRef.init(memDB.baseTxFrame())
db.setBalance(addr1, 1100.u256)
db.setNonce(addr1, 2)
db.setCode(addr1, code)
db.setStorage(addr1, 1.u256, 10.u256)
check stateRoot == db.getStateRoot()
# Ledger readonly operations using previous hash
aristo: fork support via layers/txframes (#2960) * aristo: fork support via layers/txframes This change reorganises how the database is accessed: instead holding a "current frame" in the database object, a dag of frames is created based on the "base frame" held in `AristoDbRef` and all database access happens through this frame, which can be thought of as a consistent point-in-time snapshot of the database based on a particular fork of the chain. In the code, "frame", "transaction" and "layer" is used to denote more or less the same thing: a dag of stacked changes backed by the on-disk database. Although this is not a requirement, in practice each frame holds the change set of a single block - as such, the frame and its ancestors leading up to the on-disk state represents the state of the database after that block has been applied. "committing" means merging the changes to its parent frame so that the difference between them is lost and only the cumulative changes remain - this facility enables frames to be combined arbitrarily wherever they are in the dag. In particular, it becomes possible to consolidate a set of changes near the base of the dag and commit those to disk without having to re-do the in-memory frames built on top of them - this is useful for "flattening" a set of changes during a base update and sending those to storage without having to perform a block replay on top. Looking at abstractions, a side effect of this change is that the KVT and Aristo are brought closer together by considering them to be part of the "same" atomic transaction set - the way the code gets organised, applying a block and saving it to the kvt happens in the same "logical" frame - therefore, discarding the frame discards both the aristo and kvt changes at the same time - likewise, they are persisted to disk together - this makes reasoning about the database somewhat easier but has the downside of increased memory usage, something that perhaps will need addressing in the future. Because the code reasons more strictly about frames and the state of the persisted database, it also makes it more visible where ForkedChain should be used and where it is still missing - in particular, frames represent a single branch of history while forkedchain manages multiple parallel forks - user-facing services such as the RPC should use the latter, ie until it has been finalized, a getBlock request should consider all forks and not just the blocks in the canonical head branch. Another advantage of this approach is that `AristoDbRef` conceptually becomes more simple - removing its tracking of the "current" transaction stack simplifies reasoning about what can go wrong since this state now has to be passed around in the form of `AristoTxRef` - as such, many of the tests and facilities in the code that were dealing with "stack inconsistency" are now structurally prevented from happening. The test suite will need significant refactoring after this change. Once this change has been merged, there are several follow-ups to do: * there's no mechanism for keeping frames up to date as they get committed or rolled back - TODO * naming is confused - many names for the same thing for legacy reason * forkedchain support is still missing in lots of code * clean up redundant logic based on previous designs - in particular the debug and introspection code no longer makes sense * the way change sets are stored will probably need revisiting - because it's a stack of changes where each frame must be interrogated to find an on-disk value, with a base distance of 128 we'll at minimum have to perform 128 frame lookups for *every* database interaction - regardless, the "dag-like" nature will stay * dispose and commit are poorly defined and perhaps redundant - in theory, one could simply let the GC collect abandoned frames etc, though it's likely an explicit mechanism will remain useful, so they stay for now More about the changes: * `AristoDbRef` gains a `txRef` field (todo: rename) that "more or less" corresponds to the old `balancer` field * `AristoDbRef.stack` is gone - instead, there's a chain of `AristoTxRef` objects that hold their respective "layer" which has the actual changes * No more reasoning about "top" and "stack" - instead, each `AristoTxRef` can be a "head" that "more or less" corresponds to the old single-history `top` notion and its stack * `level` still represents "distance to base" - it's computed from the parent chain instead of being stored * one has to be careful not to use frames where forkedchain was intended - layers are only for a single branch of history! * fix layer vtop after rollback * engine fix * Fix test_txpool * Fix test_rpc * Fix copyright year * fix simulator * Fix copyright year * Fix copyright year * Fix tracer * Fix infinite recursion bug * Remove aristo and kvt empty files * Fic copyright year * Fix fc chain_kvt * ForkedChain refactoring * Fix merge master conflict * Fix copyright year * Reparent txFrame * Fix test * Fix txFrame reparent again * Cleanup and fix test * UpdateBase bugfix and fix test * Fixe newPayload bug discovered by hive * Fix engine api fcu * Clean up call template, chain_kvt, andn txguid * Fix copyright year * work around base block loading issue * Add test * Fix updateHead bug * Fix updateBase bug * Change func commitBase to proc commitBase * Touch up and fix debug mode crash --------- Co-authored-by: jangko <jangko128@gmail.com>
2025-02-06 08:04:50 +01:00
var ac2 = LedgerRef.init(memDB.baseTxFrame())
var addr2 = initAddr(2)
check ac2.getCodeHash(addr2) == emptyAcc.codeHash
check ac2.getBalance(addr2) == emptyAcc.balance
check ac2.getNonce(addr2) == emptyAcc.nonce
check ac2.getCode(addr2) == []
check ac2.getCodeSize(addr2) == 0
check ac2.getCommittedStorage(addr2, 1.u256) == 0.u256
check ac2.getStorage(addr2, 1.u256) == 0.u256
check ac2.contractCollision(addr2) == false
check ac2.accountExists(addr2) == false
check ac2.isDeadAccount(addr2) == true
ac2.persist()
# readonly operations should not modify
# state trie at all
check ac2.getStateRoot() == stateRoot
test "Ledger code retrieval after persist called":
aristo: fork support via layers/txframes (#2960) * aristo: fork support via layers/txframes This change reorganises how the database is accessed: instead holding a "current frame" in the database object, a dag of frames is created based on the "base frame" held in `AristoDbRef` and all database access happens through this frame, which can be thought of as a consistent point-in-time snapshot of the database based on a particular fork of the chain. In the code, "frame", "transaction" and "layer" is used to denote more or less the same thing: a dag of stacked changes backed by the on-disk database. Although this is not a requirement, in practice each frame holds the change set of a single block - as such, the frame and its ancestors leading up to the on-disk state represents the state of the database after that block has been applied. "committing" means merging the changes to its parent frame so that the difference between them is lost and only the cumulative changes remain - this facility enables frames to be combined arbitrarily wherever they are in the dag. In particular, it becomes possible to consolidate a set of changes near the base of the dag and commit those to disk without having to re-do the in-memory frames built on top of them - this is useful for "flattening" a set of changes during a base update and sending those to storage without having to perform a block replay on top. Looking at abstractions, a side effect of this change is that the KVT and Aristo are brought closer together by considering them to be part of the "same" atomic transaction set - the way the code gets organised, applying a block and saving it to the kvt happens in the same "logical" frame - therefore, discarding the frame discards both the aristo and kvt changes at the same time - likewise, they are persisted to disk together - this makes reasoning about the database somewhat easier but has the downside of increased memory usage, something that perhaps will need addressing in the future. Because the code reasons more strictly about frames and the state of the persisted database, it also makes it more visible where ForkedChain should be used and where it is still missing - in particular, frames represent a single branch of history while forkedchain manages multiple parallel forks - user-facing services such as the RPC should use the latter, ie until it has been finalized, a getBlock request should consider all forks and not just the blocks in the canonical head branch. Another advantage of this approach is that `AristoDbRef` conceptually becomes more simple - removing its tracking of the "current" transaction stack simplifies reasoning about what can go wrong since this state now has to be passed around in the form of `AristoTxRef` - as such, many of the tests and facilities in the code that were dealing with "stack inconsistency" are now structurally prevented from happening. The test suite will need significant refactoring after this change. Once this change has been merged, there are several follow-ups to do: * there's no mechanism for keeping frames up to date as they get committed or rolled back - TODO * naming is confused - many names for the same thing for legacy reason * forkedchain support is still missing in lots of code * clean up redundant logic based on previous designs - in particular the debug and introspection code no longer makes sense * the way change sets are stored will probably need revisiting - because it's a stack of changes where each frame must be interrogated to find an on-disk value, with a base distance of 128 we'll at minimum have to perform 128 frame lookups for *every* database interaction - regardless, the "dag-like" nature will stay * dispose and commit are poorly defined and perhaps redundant - in theory, one could simply let the GC collect abandoned frames etc, though it's likely an explicit mechanism will remain useful, so they stay for now More about the changes: * `AristoDbRef` gains a `txRef` field (todo: rename) that "more or less" corresponds to the old `balancer` field * `AristoDbRef.stack` is gone - instead, there's a chain of `AristoTxRef` objects that hold their respective "layer" which has the actual changes * No more reasoning about "top" and "stack" - instead, each `AristoTxRef` can be a "head" that "more or less" corresponds to the old single-history `top` notion and its stack * `level` still represents "distance to base" - it's computed from the parent chain instead of being stored * one has to be careful not to use frames where forkedchain was intended - layers are only for a single branch of history! * fix layer vtop after rollback * engine fix * Fix test_txpool * Fix test_rpc * Fix copyright year * fix simulator * Fix copyright year * Fix copyright year * Fix tracer * Fix infinite recursion bug * Remove aristo and kvt empty files * Fic copyright year * Fix fc chain_kvt * ForkedChain refactoring * Fix merge master conflict * Fix copyright year * Reparent txFrame * Fix test * Fix txFrame reparent again * Cleanup and fix test * UpdateBase bugfix and fix test * Fixe newPayload bug discovered by hive * Fix engine api fcu * Clean up call template, chain_kvt, andn txguid * Fix copyright year * work around base block loading issue * Add test * Fix updateHead bug * Fix updateBase bug * Change func commitBase to proc commitBase * Touch up and fix debug mode crash --------- Co-authored-by: jangko <jangko128@gmail.com>
2025-02-06 08:04:50 +01:00
var ac = LedgerRef.init(memDB.baseTxFrame())
var addr2 = initAddr(2)
ac.setCode(addr2, code)
ac.persist()
check ac.getCode(addr2) == code
let
key = contractHashKey(keccak256(code))
aristo: fork support via layers/txframes (#2960) * aristo: fork support via layers/txframes This change reorganises how the database is accessed: instead holding a "current frame" in the database object, a dag of frames is created based on the "base frame" held in `AristoDbRef` and all database access happens through this frame, which can be thought of as a consistent point-in-time snapshot of the database based on a particular fork of the chain. In the code, "frame", "transaction" and "layer" is used to denote more or less the same thing: a dag of stacked changes backed by the on-disk database. Although this is not a requirement, in practice each frame holds the change set of a single block - as such, the frame and its ancestors leading up to the on-disk state represents the state of the database after that block has been applied. "committing" means merging the changes to its parent frame so that the difference between them is lost and only the cumulative changes remain - this facility enables frames to be combined arbitrarily wherever they are in the dag. In particular, it becomes possible to consolidate a set of changes near the base of the dag and commit those to disk without having to re-do the in-memory frames built on top of them - this is useful for "flattening" a set of changes during a base update and sending those to storage without having to perform a block replay on top. Looking at abstractions, a side effect of this change is that the KVT and Aristo are brought closer together by considering them to be part of the "same" atomic transaction set - the way the code gets organised, applying a block and saving it to the kvt happens in the same "logical" frame - therefore, discarding the frame discards both the aristo and kvt changes at the same time - likewise, they are persisted to disk together - this makes reasoning about the database somewhat easier but has the downside of increased memory usage, something that perhaps will need addressing in the future. Because the code reasons more strictly about frames and the state of the persisted database, it also makes it more visible where ForkedChain should be used and where it is still missing - in particular, frames represent a single branch of history while forkedchain manages multiple parallel forks - user-facing services such as the RPC should use the latter, ie until it has been finalized, a getBlock request should consider all forks and not just the blocks in the canonical head branch. Another advantage of this approach is that `AristoDbRef` conceptually becomes more simple - removing its tracking of the "current" transaction stack simplifies reasoning about what can go wrong since this state now has to be passed around in the form of `AristoTxRef` - as such, many of the tests and facilities in the code that were dealing with "stack inconsistency" are now structurally prevented from happening. The test suite will need significant refactoring after this change. Once this change has been merged, there are several follow-ups to do: * there's no mechanism for keeping frames up to date as they get committed or rolled back - TODO * naming is confused - many names for the same thing for legacy reason * forkedchain support is still missing in lots of code * clean up redundant logic based on previous designs - in particular the debug and introspection code no longer makes sense * the way change sets are stored will probably need revisiting - because it's a stack of changes where each frame must be interrogated to find an on-disk value, with a base distance of 128 we'll at minimum have to perform 128 frame lookups for *every* database interaction - regardless, the "dag-like" nature will stay * dispose and commit are poorly defined and perhaps redundant - in theory, one could simply let the GC collect abandoned frames etc, though it's likely an explicit mechanism will remain useful, so they stay for now More about the changes: * `AristoDbRef` gains a `txRef` field (todo: rename) that "more or less" corresponds to the old `balancer` field * `AristoDbRef.stack` is gone - instead, there's a chain of `AristoTxRef` objects that hold their respective "layer" which has the actual changes * No more reasoning about "top" and "stack" - instead, each `AristoTxRef` can be a "head" that "more or less" corresponds to the old single-history `top` notion and its stack * `level` still represents "distance to base" - it's computed from the parent chain instead of being stored * one has to be careful not to use frames where forkedchain was intended - layers are only for a single branch of history! * fix layer vtop after rollback * engine fix * Fix test_txpool * Fix test_rpc * Fix copyright year * fix simulator * Fix copyright year * Fix copyright year * Fix tracer * Fix infinite recursion bug * Remove aristo and kvt empty files * Fic copyright year * Fix fc chain_kvt * ForkedChain refactoring * Fix merge master conflict * Fix copyright year * Reparent txFrame * Fix test * Fix txFrame reparent again * Cleanup and fix test * UpdateBase bugfix and fix test * Fixe newPayload bug discovered by hive * Fix engine api fcu * Clean up call template, chain_kvt, andn txguid * Fix copyright year * work around base block loading issue * Add test * Fix updateHead bug * Fix updateBase bug * Change func commitBase to proc commitBase * Touch up and fix debug mode crash --------- Co-authored-by: jangko <jangko128@gmail.com>
2025-02-06 08:04:50 +01:00
val = memDB.baseTxFrame().get(key.toOpenArray).valueOr: EmptyBlob
check val == code
test "accessList operations":
proc verifyAddrs(ac: LedgerRef, addrs: varargs[int]): bool =
for c in addrs:
if not ac.inAccessList(c.initAddr):
return false
true
proc verifySlots(ac: LedgerRef, address: int, slots: varargs[int]): bool =
let a = address.initAddr
if not ac.inAccessList(a):
return false
for c in slots:
if not ac.inAccessList(a, c.u256):
return false
true
proc accessList(ac: LedgerRef, address: int) {.inline.} =
ac.accessList(address.initAddr)
proc accessList(ac: LedgerRef, address, slot: int) {.inline.} =
ac.accessList(address.initAddr, slot.u256)
aristo: fork support via layers/txframes (#2960) * aristo: fork support via layers/txframes This change reorganises how the database is accessed: instead holding a "current frame" in the database object, a dag of frames is created based on the "base frame" held in `AristoDbRef` and all database access happens through this frame, which can be thought of as a consistent point-in-time snapshot of the database based on a particular fork of the chain. In the code, "frame", "transaction" and "layer" is used to denote more or less the same thing: a dag of stacked changes backed by the on-disk database. Although this is not a requirement, in practice each frame holds the change set of a single block - as such, the frame and its ancestors leading up to the on-disk state represents the state of the database after that block has been applied. "committing" means merging the changes to its parent frame so that the difference between them is lost and only the cumulative changes remain - this facility enables frames to be combined arbitrarily wherever they are in the dag. In particular, it becomes possible to consolidate a set of changes near the base of the dag and commit those to disk without having to re-do the in-memory frames built on top of them - this is useful for "flattening" a set of changes during a base update and sending those to storage without having to perform a block replay on top. Looking at abstractions, a side effect of this change is that the KVT and Aristo are brought closer together by considering them to be part of the "same" atomic transaction set - the way the code gets organised, applying a block and saving it to the kvt happens in the same "logical" frame - therefore, discarding the frame discards both the aristo and kvt changes at the same time - likewise, they are persisted to disk together - this makes reasoning about the database somewhat easier but has the downside of increased memory usage, something that perhaps will need addressing in the future. Because the code reasons more strictly about frames and the state of the persisted database, it also makes it more visible where ForkedChain should be used and where it is still missing - in particular, frames represent a single branch of history while forkedchain manages multiple parallel forks - user-facing services such as the RPC should use the latter, ie until it has been finalized, a getBlock request should consider all forks and not just the blocks in the canonical head branch. Another advantage of this approach is that `AristoDbRef` conceptually becomes more simple - removing its tracking of the "current" transaction stack simplifies reasoning about what can go wrong since this state now has to be passed around in the form of `AristoTxRef` - as such, many of the tests and facilities in the code that were dealing with "stack inconsistency" are now structurally prevented from happening. The test suite will need significant refactoring after this change. Once this change has been merged, there are several follow-ups to do: * there's no mechanism for keeping frames up to date as they get committed or rolled back - TODO * naming is confused - many names for the same thing for legacy reason * forkedchain support is still missing in lots of code * clean up redundant logic based on previous designs - in particular the debug and introspection code no longer makes sense * the way change sets are stored will probably need revisiting - because it's a stack of changes where each frame must be interrogated to find an on-disk value, with a base distance of 128 we'll at minimum have to perform 128 frame lookups for *every* database interaction - regardless, the "dag-like" nature will stay * dispose and commit are poorly defined and perhaps redundant - in theory, one could simply let the GC collect abandoned frames etc, though it's likely an explicit mechanism will remain useful, so they stay for now More about the changes: * `AristoDbRef` gains a `txRef` field (todo: rename) that "more or less" corresponds to the old `balancer` field * `AristoDbRef.stack` is gone - instead, there's a chain of `AristoTxRef` objects that hold their respective "layer" which has the actual changes * No more reasoning about "top" and "stack" - instead, each `AristoTxRef` can be a "head" that "more or less" corresponds to the old single-history `top` notion and its stack * `level` still represents "distance to base" - it's computed from the parent chain instead of being stored * one has to be careful not to use frames where forkedchain was intended - layers are only for a single branch of history! * fix layer vtop after rollback * engine fix * Fix test_txpool * Fix test_rpc * Fix copyright year * fix simulator * Fix copyright year * Fix copyright year * Fix tracer * Fix infinite recursion bug * Remove aristo and kvt empty files * Fic copyright year * Fix fc chain_kvt * ForkedChain refactoring * Fix merge master conflict * Fix copyright year * Reparent txFrame * Fix test * Fix txFrame reparent again * Cleanup and fix test * UpdateBase bugfix and fix test * Fixe newPayload bug discovered by hive * Fix engine api fcu * Clean up call template, chain_kvt, andn txguid * Fix copyright year * work around base block loading issue * Add test * Fix updateHead bug * Fix updateBase bug * Change func commitBase to proc commitBase * Touch up and fix debug mode crash --------- Co-authored-by: jangko <jangko128@gmail.com>
2025-02-06 08:04:50 +01:00
var ac = LedgerRef.init(memDB.baseTxFrame())
ac.accessList(0xaa)
ac.accessList(0xbb, 0x01)
ac.accessList(0xbb, 0x02)
check ac.verifyAddrs(0xaa, 0xbb)
check ac.verifySlots(0xbb, 0x01, 0x02)
check ac.verifySlots(0xaa, 0x01) == false
check ac.verifySlots(0xaa, 0x02) == false
var sp = ac.beginSavepoint
# some new ones
ac.accessList(0xbb, 0x03)
ac.accessList(0xaa, 0x01)
ac.accessList(0xcc, 0x01)
ac.accessList(0xcc)
check ac.verifyAddrs(0xaa, 0xbb, 0xcc)
check ac.verifySlots(0xaa, 0x01)
check ac.verifySlots(0xbb, 0x01, 0x02, 0x03)
check ac.verifySlots(0xcc, 0x01)
ac.rollback(sp)
check ac.verifyAddrs(0xaa, 0xbb)
check ac.verifyAddrs(0xcc) == false
check ac.verifySlots(0xcc, 0x01) == false
sp = ac.beginSavepoint
ac.accessList(0xbb, 0x03)
ac.accessList(0xaa, 0x01)
ac.accessList(0xcc, 0x01)
ac.accessList(0xcc)
ac.accessList(0xdd, 0x04)
ac.commit(sp)
check ac.verifyAddrs(0xaa, 0xbb, 0xcc)
check ac.verifySlots(0xaa, 0x01)
check ac.verifySlots(0xbb, 0x01, 0x02, 0x03)
check ac.verifySlots(0xcc, 0x01)
check ac.verifySlots(0xdd, 0x04)
test "transient storage operations":
aristo: fork support via layers/txframes (#2960) * aristo: fork support via layers/txframes This change reorganises how the database is accessed: instead holding a "current frame" in the database object, a dag of frames is created based on the "base frame" held in `AristoDbRef` and all database access happens through this frame, which can be thought of as a consistent point-in-time snapshot of the database based on a particular fork of the chain. In the code, "frame", "transaction" and "layer" is used to denote more or less the same thing: a dag of stacked changes backed by the on-disk database. Although this is not a requirement, in practice each frame holds the change set of a single block - as such, the frame and its ancestors leading up to the on-disk state represents the state of the database after that block has been applied. "committing" means merging the changes to its parent frame so that the difference between them is lost and only the cumulative changes remain - this facility enables frames to be combined arbitrarily wherever they are in the dag. In particular, it becomes possible to consolidate a set of changes near the base of the dag and commit those to disk without having to re-do the in-memory frames built on top of them - this is useful for "flattening" a set of changes during a base update and sending those to storage without having to perform a block replay on top. Looking at abstractions, a side effect of this change is that the KVT and Aristo are brought closer together by considering them to be part of the "same" atomic transaction set - the way the code gets organised, applying a block and saving it to the kvt happens in the same "logical" frame - therefore, discarding the frame discards both the aristo and kvt changes at the same time - likewise, they are persisted to disk together - this makes reasoning about the database somewhat easier but has the downside of increased memory usage, something that perhaps will need addressing in the future. Because the code reasons more strictly about frames and the state of the persisted database, it also makes it more visible where ForkedChain should be used and where it is still missing - in particular, frames represent a single branch of history while forkedchain manages multiple parallel forks - user-facing services such as the RPC should use the latter, ie until it has been finalized, a getBlock request should consider all forks and not just the blocks in the canonical head branch. Another advantage of this approach is that `AristoDbRef` conceptually becomes more simple - removing its tracking of the "current" transaction stack simplifies reasoning about what can go wrong since this state now has to be passed around in the form of `AristoTxRef` - as such, many of the tests and facilities in the code that were dealing with "stack inconsistency" are now structurally prevented from happening. The test suite will need significant refactoring after this change. Once this change has been merged, there are several follow-ups to do: * there's no mechanism for keeping frames up to date as they get committed or rolled back - TODO * naming is confused - many names for the same thing for legacy reason * forkedchain support is still missing in lots of code * clean up redundant logic based on previous designs - in particular the debug and introspection code no longer makes sense * the way change sets are stored will probably need revisiting - because it's a stack of changes where each frame must be interrogated to find an on-disk value, with a base distance of 128 we'll at minimum have to perform 128 frame lookups for *every* database interaction - regardless, the "dag-like" nature will stay * dispose and commit are poorly defined and perhaps redundant - in theory, one could simply let the GC collect abandoned frames etc, though it's likely an explicit mechanism will remain useful, so they stay for now More about the changes: * `AristoDbRef` gains a `txRef` field (todo: rename) that "more or less" corresponds to the old `balancer` field * `AristoDbRef.stack` is gone - instead, there's a chain of `AristoTxRef` objects that hold their respective "layer" which has the actual changes * No more reasoning about "top" and "stack" - instead, each `AristoTxRef` can be a "head" that "more or less" corresponds to the old single-history `top` notion and its stack * `level` still represents "distance to base" - it's computed from the parent chain instead of being stored * one has to be careful not to use frames where forkedchain was intended - layers are only for a single branch of history! * fix layer vtop after rollback * engine fix * Fix test_txpool * Fix test_rpc * Fix copyright year * fix simulator * Fix copyright year * Fix copyright year * Fix tracer * Fix infinite recursion bug * Remove aristo and kvt empty files * Fic copyright year * Fix fc chain_kvt * ForkedChain refactoring * Fix merge master conflict * Fix copyright year * Reparent txFrame * Fix test * Fix txFrame reparent again * Cleanup and fix test * UpdateBase bugfix and fix test * Fixe newPayload bug discovered by hive * Fix engine api fcu * Clean up call template, chain_kvt, andn txguid * Fix copyright year * work around base block loading issue * Add test * Fix updateHead bug * Fix updateBase bug * Change func commitBase to proc commitBase * Touch up and fix debug mode crash --------- Co-authored-by: jangko <jangko128@gmail.com>
2025-02-06 08:04:50 +01:00
var ac = LedgerRef.init(memDB.baseTxFrame())
proc tStore(ac: LedgerRef, address, slot, val: int) =
ac.setTransientStorage(address.initAddr, slot.u256, val.u256)
proc tLoad(ac: LedgerRef, address, slot: int): UInt256 =
ac.getTransientStorage(address.initAddr, slot.u256)
proc vts(ac: LedgerRef, address, slot, val: int): bool =
ac.tLoad(address, slot) == val.u256
ac.tStore(0xaa, 3, 66)
ac.tStore(0xbb, 1, 33)
ac.tStore(0xbb, 2, 99)
check ac.vts(0xaa, 3, 66)
check ac.vts(0xbb, 1, 33)
check ac.vts(0xbb, 2, 99)
check ac.vts(0xaa, 1, 33) == false
check ac.vts(0xbb, 1, 66) == false
var sp = ac.beginSavepoint
# some new ones
ac.tStore(0xaa, 3, 77)
ac.tStore(0xbb, 1, 55)
ac.tStore(0xcc, 7, 88)
check ac.vts(0xaa, 3, 77)
check ac.vts(0xbb, 1, 55)
check ac.vts(0xcc, 7, 88)
check ac.vts(0xaa, 3, 66) == false
check ac.vts(0xbb, 1, 33) == false
check ac.vts(0xbb, 2, 99)
ac.rollback(sp)
check ac.vts(0xaa, 3, 66)
check ac.vts(0xbb, 1, 33)
check ac.vts(0xbb, 2, 99)
check ac.vts(0xcc, 7, 88) == false
sp = ac.beginSavepoint
ac.tStore(0xaa, 3, 44)
ac.tStore(0xaa, 4, 55)
ac.tStore(0xbb, 1, 22)
ac.tStore(0xdd, 2, 66)
ac.commit(sp)
check ac.vts(0xaa, 3, 44)
check ac.vts(0xaa, 4, 55)
check ac.vts(0xbb, 1, 22)
check ac.vts(0xbb, 1, 55) == false
check ac.vts(0xbb, 2, 99)
check ac.vts(0xcc, 7, 88) == false
check ac.vts(0xdd, 2, 66)
ac.clearTransientStorage()
check ac.vts(0xaa, 3, 44) == false
check ac.vts(0xaa, 4, 55) == false
check ac.vts(0xbb, 1, 22) == false
check ac.vts(0xbb, 1, 55) == false
check ac.vts(0xbb, 2, 99) == false
check ac.vts(0xcc, 7, 88) == false
check ac.vts(0xdd, 2, 66) == false
test "ledger contractCollision":
# use previous hash
aristo: fork support via layers/txframes (#2960) * aristo: fork support via layers/txframes This change reorganises how the database is accessed: instead holding a "current frame" in the database object, a dag of frames is created based on the "base frame" held in `AristoDbRef` and all database access happens through this frame, which can be thought of as a consistent point-in-time snapshot of the database based on a particular fork of the chain. In the code, "frame", "transaction" and "layer" is used to denote more or less the same thing: a dag of stacked changes backed by the on-disk database. Although this is not a requirement, in practice each frame holds the change set of a single block - as such, the frame and its ancestors leading up to the on-disk state represents the state of the database after that block has been applied. "committing" means merging the changes to its parent frame so that the difference between them is lost and only the cumulative changes remain - this facility enables frames to be combined arbitrarily wherever they are in the dag. In particular, it becomes possible to consolidate a set of changes near the base of the dag and commit those to disk without having to re-do the in-memory frames built on top of them - this is useful for "flattening" a set of changes during a base update and sending those to storage without having to perform a block replay on top. Looking at abstractions, a side effect of this change is that the KVT and Aristo are brought closer together by considering them to be part of the "same" atomic transaction set - the way the code gets organised, applying a block and saving it to the kvt happens in the same "logical" frame - therefore, discarding the frame discards both the aristo and kvt changes at the same time - likewise, they are persisted to disk together - this makes reasoning about the database somewhat easier but has the downside of increased memory usage, something that perhaps will need addressing in the future. Because the code reasons more strictly about frames and the state of the persisted database, it also makes it more visible where ForkedChain should be used and where it is still missing - in particular, frames represent a single branch of history while forkedchain manages multiple parallel forks - user-facing services such as the RPC should use the latter, ie until it has been finalized, a getBlock request should consider all forks and not just the blocks in the canonical head branch. Another advantage of this approach is that `AristoDbRef` conceptually becomes more simple - removing its tracking of the "current" transaction stack simplifies reasoning about what can go wrong since this state now has to be passed around in the form of `AristoTxRef` - as such, many of the tests and facilities in the code that were dealing with "stack inconsistency" are now structurally prevented from happening. The test suite will need significant refactoring after this change. Once this change has been merged, there are several follow-ups to do: * there's no mechanism for keeping frames up to date as they get committed or rolled back - TODO * naming is confused - many names for the same thing for legacy reason * forkedchain support is still missing in lots of code * clean up redundant logic based on previous designs - in particular the debug and introspection code no longer makes sense * the way change sets are stored will probably need revisiting - because it's a stack of changes where each frame must be interrogated to find an on-disk value, with a base distance of 128 we'll at minimum have to perform 128 frame lookups for *every* database interaction - regardless, the "dag-like" nature will stay * dispose and commit are poorly defined and perhaps redundant - in theory, one could simply let the GC collect abandoned frames etc, though it's likely an explicit mechanism will remain useful, so they stay for now More about the changes: * `AristoDbRef` gains a `txRef` field (todo: rename) that "more or less" corresponds to the old `balancer` field * `AristoDbRef.stack` is gone - instead, there's a chain of `AristoTxRef` objects that hold their respective "layer" which has the actual changes * No more reasoning about "top" and "stack" - instead, each `AristoTxRef` can be a "head" that "more or less" corresponds to the old single-history `top` notion and its stack * `level` still represents "distance to base" - it's computed from the parent chain instead of being stored * one has to be careful not to use frames where forkedchain was intended - layers are only for a single branch of history! * fix layer vtop after rollback * engine fix * Fix test_txpool * Fix test_rpc * Fix copyright year * fix simulator * Fix copyright year * Fix copyright year * Fix tracer * Fix infinite recursion bug * Remove aristo and kvt empty files * Fic copyright year * Fix fc chain_kvt * ForkedChain refactoring * Fix merge master conflict * Fix copyright year * Reparent txFrame * Fix test * Fix txFrame reparent again * Cleanup and fix test * UpdateBase bugfix and fix test * Fixe newPayload bug discovered by hive * Fix engine api fcu * Clean up call template, chain_kvt, andn txguid * Fix copyright year * work around base block loading issue * Add test * Fix updateHead bug * Fix updateBase bug * Change func commitBase to proc commitBase * Touch up and fix debug mode crash --------- Co-authored-by: jangko <jangko128@gmail.com>
2025-02-06 08:04:50 +01:00
var ac = LedgerRef.init(memDB.baseTxFrame())
let addr2 = initAddr(2)
check ac.contractCollision(addr2) == false
ac.setStorage(addr2, 1.u256, 1.u256)
check ac.contractCollision(addr2) == false
ac.persist()
check ac.contractCollision(addr2) == true
let addr3 = initAddr(3)
check ac.contractCollision(addr3) == false
ac.setCode(addr3, @[0xaa.byte, 0xbb])
check ac.contractCollision(addr3) == true
let addr4 = initAddr(4)
check ac.contractCollision(addr4) == false
ac.setNonce(addr4, 1)
check ac.contractCollision(addr4) == true
test "Ledger storage iterator":
aristo: fork support via layers/txframes (#2960) * aristo: fork support via layers/txframes This change reorganises how the database is accessed: instead holding a "current frame" in the database object, a dag of frames is created based on the "base frame" held in `AristoDbRef` and all database access happens through this frame, which can be thought of as a consistent point-in-time snapshot of the database based on a particular fork of the chain. In the code, "frame", "transaction" and "layer" is used to denote more or less the same thing: a dag of stacked changes backed by the on-disk database. Although this is not a requirement, in practice each frame holds the change set of a single block - as such, the frame and its ancestors leading up to the on-disk state represents the state of the database after that block has been applied. "committing" means merging the changes to its parent frame so that the difference between them is lost and only the cumulative changes remain - this facility enables frames to be combined arbitrarily wherever they are in the dag. In particular, it becomes possible to consolidate a set of changes near the base of the dag and commit those to disk without having to re-do the in-memory frames built on top of them - this is useful for "flattening" a set of changes during a base update and sending those to storage without having to perform a block replay on top. Looking at abstractions, a side effect of this change is that the KVT and Aristo are brought closer together by considering them to be part of the "same" atomic transaction set - the way the code gets organised, applying a block and saving it to the kvt happens in the same "logical" frame - therefore, discarding the frame discards both the aristo and kvt changes at the same time - likewise, they are persisted to disk together - this makes reasoning about the database somewhat easier but has the downside of increased memory usage, something that perhaps will need addressing in the future. Because the code reasons more strictly about frames and the state of the persisted database, it also makes it more visible where ForkedChain should be used and where it is still missing - in particular, frames represent a single branch of history while forkedchain manages multiple parallel forks - user-facing services such as the RPC should use the latter, ie until it has been finalized, a getBlock request should consider all forks and not just the blocks in the canonical head branch. Another advantage of this approach is that `AristoDbRef` conceptually becomes more simple - removing its tracking of the "current" transaction stack simplifies reasoning about what can go wrong since this state now has to be passed around in the form of `AristoTxRef` - as such, many of the tests and facilities in the code that were dealing with "stack inconsistency" are now structurally prevented from happening. The test suite will need significant refactoring after this change. Once this change has been merged, there are several follow-ups to do: * there's no mechanism for keeping frames up to date as they get committed or rolled back - TODO * naming is confused - many names for the same thing for legacy reason * forkedchain support is still missing in lots of code * clean up redundant logic based on previous designs - in particular the debug and introspection code no longer makes sense * the way change sets are stored will probably need revisiting - because it's a stack of changes where each frame must be interrogated to find an on-disk value, with a base distance of 128 we'll at minimum have to perform 128 frame lookups for *every* database interaction - regardless, the "dag-like" nature will stay * dispose and commit are poorly defined and perhaps redundant - in theory, one could simply let the GC collect abandoned frames etc, though it's likely an explicit mechanism will remain useful, so they stay for now More about the changes: * `AristoDbRef` gains a `txRef` field (todo: rename) that "more or less" corresponds to the old `balancer` field * `AristoDbRef.stack` is gone - instead, there's a chain of `AristoTxRef` objects that hold their respective "layer" which has the actual changes * No more reasoning about "top" and "stack" - instead, each `AristoTxRef` can be a "head" that "more or less" corresponds to the old single-history `top` notion and its stack * `level` still represents "distance to base" - it's computed from the parent chain instead of being stored * one has to be careful not to use frames where forkedchain was intended - layers are only for a single branch of history! * fix layer vtop after rollback * engine fix * Fix test_txpool * Fix test_rpc * Fix copyright year * fix simulator * Fix copyright year * Fix copyright year * Fix tracer * Fix infinite recursion bug * Remove aristo and kvt empty files * Fic copyright year * Fix fc chain_kvt * ForkedChain refactoring * Fix merge master conflict * Fix copyright year * Reparent txFrame * Fix test * Fix txFrame reparent again * Cleanup and fix test * UpdateBase bugfix and fix test * Fixe newPayload bug discovered by hive * Fix engine api fcu * Clean up call template, chain_kvt, andn txguid * Fix copyright year * work around base block loading issue * Add test * Fix updateHead bug * Fix updateBase bug * Change func commitBase to proc commitBase * Touch up and fix debug mode crash --------- Co-authored-by: jangko <jangko128@gmail.com>
2025-02-06 08:04:50 +01:00
var ac = LedgerRef.init(memDB.baseTxFrame(), storeSlotHash = true)
let addr2 = initAddr(2)
ac.setStorage(addr2, 1.u256, 2.u256)
ac.setStorage(addr2, 2.u256, 3.u256)
var keys: seq[UInt256]
var vals: seq[UInt256]
for k, v in ac.cachedStorage(addr2):
keys.add k
vals.add v
# before persist, there are storages in cache
check keys.len == 2
check vals.len == 2
check 1.u256 in keys
check 2.u256 in keys
# before persist, the values are all original values
check vals == @[0.u256, 0.u256]
keys.reset
vals.reset
for k, v in ac.storage(addr2):
keys.add k
vals.add k
# before persist, there are no storages in db
check keys.len == 0
check vals.len == 0
ac.persist()
for k, v in ac.cachedStorage(addr2):
keys.add k
vals.add v
# after persist, there are storages in cache
check keys.len == 2
check vals.len == 2
check 1.u256 in keys
check 2.u256 in keys
# after persist, the values are what we put into
check 2.u256 in vals
check 3.u256 in vals
keys.reset
vals.reset
for k, v in ac.storage(addr2):
keys.add k
vals.add v
# after persist, there are storages in db
check keys.len == 2
check vals.len == 2
check 1.u256 in keys
check 2.u256 in keys
check 2.u256 in vals
check 3.u256 in vals
# ------------------------------------------------------------------------------
# Main function(s)
# ------------------------------------------------------------------------------
proc ledgerMain*(noisy = defined(debug)) =
noisy.runLedgerTransactionTests
runLedgerBasicOperationsTests()
when isMainModule:
var noisy = defined(debug)
setErrorLevel()
noisy.ledgerMain
# ------------------------------------------------------------------------------
# End
# ------------------------------------------------------------------------------