bsc/core/tx_pool.go
Péter Szilágyi 6793ffa12b
Merge pull request #21300 from rjl493456442/txpool-fix-queued-evictions
core: fix queued transaction eviction
2020-07-24 11:14:42 +03:00

1600 lines
54 KiB
Go

// Copyright 2014 The go-ethereum Authors
// This file is part of the go-ethereum library.
//
// The go-ethereum library is free software: you can redistribute it and/or modify
// it under the terms of the GNU Lesser General Public License as published by
// the Free Software Foundation, either version 3 of the License, or
// (at your option) any later version.
//
// The go-ethereum library is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
// GNU Lesser General Public License for more details.
//
// You should have received a copy of the GNU Lesser General Public License
// along with the go-ethereum library. If not, see <http://www.gnu.org/licenses/>.
package core
import (
"errors"
"math"
"math/big"
"sort"
"sync"
"time"
"github.com/ethereum/go-ethereum/common"
"github.com/ethereum/go-ethereum/common/prque"
"github.com/ethereum/go-ethereum/core/state"
"github.com/ethereum/go-ethereum/core/types"
"github.com/ethereum/go-ethereum/event"
"github.com/ethereum/go-ethereum/log"
"github.com/ethereum/go-ethereum/metrics"
"github.com/ethereum/go-ethereum/params"
)
const (
// chainHeadChanSize is the size of channel listening to ChainHeadEvent.
chainHeadChanSize = 10
// txSlotSize is used to calculate how many data slots a single transaction
// takes up based on its size. The slots are used as DoS protection, ensuring
// that validating a new transaction remains a constant operation (in reality
// O(maxslots), where max slots are 4 currently).
txSlotSize = 32 * 1024
// txMaxSize is the maximum size a single transaction can have. This field has
// non-trivial consequences: larger transactions are significantly harder and
// more expensive to propagate; larger transactions also take more resources
// to validate whether they fit into the pool or not.
txMaxSize = 4 * txSlotSize // 128KB
)
var (
// ErrAlreadyKnown is returned if the transactions is already contained
// within the pool.
ErrAlreadyKnown = errors.New("already known")
// ErrInvalidSender is returned if the transaction contains an invalid signature.
ErrInvalidSender = errors.New("invalid sender")
// ErrUnderpriced is returned if a transaction's gas price is below the minimum
// configured for the transaction pool.
ErrUnderpriced = errors.New("transaction underpriced")
// ErrReplaceUnderpriced is returned if a transaction is attempted to be replaced
// with a different one without the required price bump.
ErrReplaceUnderpriced = errors.New("replacement transaction underpriced")
// ErrGasLimit is returned if a transaction's requested gas limit exceeds the
// maximum allowance of the current block.
ErrGasLimit = errors.New("exceeds block gas limit")
// ErrNegativeValue is a sanity error to ensure no one is able to specify a
// transaction with a negative value.
ErrNegativeValue = errors.New("negative value")
// ErrOversizedData is returned if the input data of a transaction is greater
// than some meaningful limit a user might use. This is not a consensus error
// making the transaction invalid, rather a DOS protection.
ErrOversizedData = errors.New("oversized data")
)
var (
evictionInterval = time.Minute // Time interval to check for evictable transactions
statsReportInterval = 8 * time.Second // Time interval to report transaction pool stats
)
var (
// Metrics for the pending pool
pendingDiscardMeter = metrics.NewRegisteredMeter("txpool/pending/discard", nil)
pendingReplaceMeter = metrics.NewRegisteredMeter("txpool/pending/replace", nil)
pendingRateLimitMeter = metrics.NewRegisteredMeter("txpool/pending/ratelimit", nil) // Dropped due to rate limiting
pendingNofundsMeter = metrics.NewRegisteredMeter("txpool/pending/nofunds", nil) // Dropped due to out-of-funds
// Metrics for the queued pool
queuedDiscardMeter = metrics.NewRegisteredMeter("txpool/queued/discard", nil)
queuedReplaceMeter = metrics.NewRegisteredMeter("txpool/queued/replace", nil)
queuedRateLimitMeter = metrics.NewRegisteredMeter("txpool/queued/ratelimit", nil) // Dropped due to rate limiting
queuedNofundsMeter = metrics.NewRegisteredMeter("txpool/queued/nofunds", nil) // Dropped due to out-of-funds
queuedEvictionMeter = metrics.NewRegisteredMeter("txpool/queued/eviction", nil) // Dropped due to lifetime
// General tx metrics
knownTxMeter = metrics.NewRegisteredMeter("txpool/known", nil)
validTxMeter = metrics.NewRegisteredMeter("txpool/valid", nil)
invalidTxMeter = metrics.NewRegisteredMeter("txpool/invalid", nil)
underpricedTxMeter = metrics.NewRegisteredMeter("txpool/underpriced", nil)
pendingGauge = metrics.NewRegisteredGauge("txpool/pending", nil)
queuedGauge = metrics.NewRegisteredGauge("txpool/queued", nil)
localGauge = metrics.NewRegisteredGauge("txpool/local", nil)
slotsGauge = metrics.NewRegisteredGauge("txpool/slots", nil)
)
// TxStatus is the current status of a transaction as seen by the pool.
type TxStatus uint
const (
TxStatusUnknown TxStatus = iota
TxStatusQueued
TxStatusPending
TxStatusIncluded
)
// blockChain provides the state of blockchain and current gas limit to do
// some pre checks in tx pool and event subscribers.
type blockChain interface {
CurrentBlock() *types.Block
GetBlock(hash common.Hash, number uint64) *types.Block
StateAt(root common.Hash) (*state.StateDB, error)
SubscribeChainHeadEvent(ch chan<- ChainHeadEvent) event.Subscription
}
// TxPoolConfig are the configuration parameters of the transaction pool.
type TxPoolConfig struct {
Locals []common.Address // Addresses that should be treated by default as local
NoLocals bool // Whether local transaction handling should be disabled
Journal string // Journal of local transactions to survive node restarts
Rejournal time.Duration // Time interval to regenerate the local transaction journal
PriceLimit uint64 // Minimum gas price to enforce for acceptance into the pool
PriceBump uint64 // Minimum price bump percentage to replace an already existing transaction (nonce)
AccountSlots uint64 // Number of executable transaction slots guaranteed per account
GlobalSlots uint64 // Maximum number of executable transaction slots for all accounts
AccountQueue uint64 // Maximum number of non-executable transaction slots permitted per account
GlobalQueue uint64 // Maximum number of non-executable transaction slots for all accounts
Lifetime time.Duration // Maximum amount of time non-executable transaction are queued
}
// DefaultTxPoolConfig contains the default configurations for the transaction
// pool.
var DefaultTxPoolConfig = TxPoolConfig{
Journal: "transactions.rlp",
Rejournal: time.Hour,
PriceLimit: 1,
PriceBump: 10,
AccountSlots: 16,
GlobalSlots: 4096,
AccountQueue: 64,
GlobalQueue: 1024,
Lifetime: 3 * time.Hour,
}
// sanitize checks the provided user configurations and changes anything that's
// unreasonable or unworkable.
func (config *TxPoolConfig) sanitize() TxPoolConfig {
conf := *config
if conf.Rejournal < time.Second {
log.Warn("Sanitizing invalid txpool journal time", "provided", conf.Rejournal, "updated", time.Second)
conf.Rejournal = time.Second
}
if conf.PriceLimit < 1 {
log.Warn("Sanitizing invalid txpool price limit", "provided", conf.PriceLimit, "updated", DefaultTxPoolConfig.PriceLimit)
conf.PriceLimit = DefaultTxPoolConfig.PriceLimit
}
if conf.PriceBump < 1 {
log.Warn("Sanitizing invalid txpool price bump", "provided", conf.PriceBump, "updated", DefaultTxPoolConfig.PriceBump)
conf.PriceBump = DefaultTxPoolConfig.PriceBump
}
if conf.AccountSlots < 1 {
log.Warn("Sanitizing invalid txpool account slots", "provided", conf.AccountSlots, "updated", DefaultTxPoolConfig.AccountSlots)
conf.AccountSlots = DefaultTxPoolConfig.AccountSlots
}
if conf.GlobalSlots < 1 {
log.Warn("Sanitizing invalid txpool global slots", "provided", conf.GlobalSlots, "updated", DefaultTxPoolConfig.GlobalSlots)
conf.GlobalSlots = DefaultTxPoolConfig.GlobalSlots
}
if conf.AccountQueue < 1 {
log.Warn("Sanitizing invalid txpool account queue", "provided", conf.AccountQueue, "updated", DefaultTxPoolConfig.AccountQueue)
conf.AccountQueue = DefaultTxPoolConfig.AccountQueue
}
if conf.GlobalQueue < 1 {
log.Warn("Sanitizing invalid txpool global queue", "provided", conf.GlobalQueue, "updated", DefaultTxPoolConfig.GlobalQueue)
conf.GlobalQueue = DefaultTxPoolConfig.GlobalQueue
}
if conf.Lifetime < 1 {
log.Warn("Sanitizing invalid txpool lifetime", "provided", conf.Lifetime, "updated", DefaultTxPoolConfig.Lifetime)
conf.Lifetime = DefaultTxPoolConfig.Lifetime
}
return conf
}
// TxPool contains all currently known transactions. Transactions
// enter the pool when they are received from the network or submitted
// locally. They exit the pool when they are included in the blockchain.
//
// The pool separates processable transactions (which can be applied to the
// current state) and future transactions. Transactions move between those
// two states over time as they are received and processed.
type TxPool struct {
config TxPoolConfig
chainconfig *params.ChainConfig
chain blockChain
gasPrice *big.Int
txFeed event.Feed
scope event.SubscriptionScope
signer types.Signer
mu sync.RWMutex
istanbul bool // Fork indicator whether we are in the istanbul stage.
currentState *state.StateDB // Current state in the blockchain head
pendingNonces *txNoncer // Pending state tracking virtual nonces
currentMaxGas uint64 // Current gas limit for transaction caps
locals *accountSet // Set of local transaction to exempt from eviction rules
journal *txJournal // Journal of local transaction to back up to disk
pending map[common.Address]*txList // All currently processable transactions
queue map[common.Address]*txList // Queued but non-processable transactions
beats map[common.Address]time.Time // Last heartbeat from each known account
all *txLookup // All transactions to allow lookups
priced *txPricedList // All transactions sorted by price
chainHeadCh chan ChainHeadEvent
chainHeadSub event.Subscription
reqResetCh chan *txpoolResetRequest
reqPromoteCh chan *accountSet
queueTxEventCh chan *types.Transaction
reorgDoneCh chan chan struct{}
reorgShutdownCh chan struct{} // requests shutdown of scheduleReorgLoop
wg sync.WaitGroup // tracks loop, scheduleReorgLoop
}
type txpoolResetRequest struct {
oldHead, newHead *types.Header
}
// NewTxPool creates a new transaction pool to gather, sort and filter inbound
// transactions from the network.
func NewTxPool(config TxPoolConfig, chainconfig *params.ChainConfig, chain blockChain) *TxPool {
// Sanitize the input to ensure no vulnerable gas prices are set
config = (&config).sanitize()
// Create the transaction pool with its initial settings
pool := &TxPool{
config: config,
chainconfig: chainconfig,
chain: chain,
signer: types.NewEIP155Signer(chainconfig.ChainID),
pending: make(map[common.Address]*txList),
queue: make(map[common.Address]*txList),
beats: make(map[common.Address]time.Time),
all: newTxLookup(),
chainHeadCh: make(chan ChainHeadEvent, chainHeadChanSize),
reqResetCh: make(chan *txpoolResetRequest),
reqPromoteCh: make(chan *accountSet),
queueTxEventCh: make(chan *types.Transaction),
reorgDoneCh: make(chan chan struct{}),
reorgShutdownCh: make(chan struct{}),
gasPrice: new(big.Int).SetUint64(config.PriceLimit),
}
pool.locals = newAccountSet(pool.signer)
for _, addr := range config.Locals {
log.Info("Setting new local account", "address", addr)
pool.locals.add(addr)
}
pool.priced = newTxPricedList(pool.all)
pool.reset(nil, chain.CurrentBlock().Header())
// Start the reorg loop early so it can handle requests generated during journal loading.
pool.wg.Add(1)
go pool.scheduleReorgLoop()
// If local transactions and journaling is enabled, load from disk
if !config.NoLocals && config.Journal != "" {
pool.journal = newTxJournal(config.Journal)
if err := pool.journal.load(pool.AddLocals); err != nil {
log.Warn("Failed to load transaction journal", "err", err)
}
if err := pool.journal.rotate(pool.local()); err != nil {
log.Warn("Failed to rotate transaction journal", "err", err)
}
}
// Subscribe events from blockchain and start the main event loop.
pool.chainHeadSub = pool.chain.SubscribeChainHeadEvent(pool.chainHeadCh)
pool.wg.Add(1)
go pool.loop()
return pool
}
// loop is the transaction pool's main event loop, waiting for and reacting to
// outside blockchain events as well as for various reporting and transaction
// eviction events.
func (pool *TxPool) loop() {
defer pool.wg.Done()
var (
prevPending, prevQueued, prevStales int
// Start the stats reporting and transaction eviction tickers
report = time.NewTicker(statsReportInterval)
evict = time.NewTicker(evictionInterval)
journal = time.NewTicker(pool.config.Rejournal)
// Track the previous head headers for transaction reorgs
head = pool.chain.CurrentBlock()
)
defer report.Stop()
defer evict.Stop()
defer journal.Stop()
for {
select {
// Handle ChainHeadEvent
case ev := <-pool.chainHeadCh:
if ev.Block != nil {
pool.requestReset(head.Header(), ev.Block.Header())
head = ev.Block
}
// System shutdown.
case <-pool.chainHeadSub.Err():
close(pool.reorgShutdownCh)
return
// Handle stats reporting ticks
case <-report.C:
pool.mu.RLock()
pending, queued := pool.stats()
stales := pool.priced.stales
pool.mu.RUnlock()
if pending != prevPending || queued != prevQueued || stales != prevStales {
log.Debug("Transaction pool status report", "executable", pending, "queued", queued, "stales", stales)
prevPending, prevQueued, prevStales = pending, queued, stales
}
// Handle inactive account transaction eviction
case <-evict.C:
pool.mu.Lock()
for addr := range pool.queue {
// Skip local transactions from the eviction mechanism
if pool.locals.contains(addr) {
continue
}
// Any non-locals old enough should be removed
if time.Since(pool.beats[addr]) > pool.config.Lifetime {
list := pool.queue[addr].Flatten()
for _, tx := range list {
pool.removeTx(tx.Hash(), true)
}
queuedEvictionMeter.Mark(int64(len(list)))
}
}
pool.mu.Unlock()
// Handle local transaction journal rotation
case <-journal.C:
if pool.journal != nil {
pool.mu.Lock()
if err := pool.journal.rotate(pool.local()); err != nil {
log.Warn("Failed to rotate local tx journal", "err", err)
}
pool.mu.Unlock()
}
}
}
}
// Stop terminates the transaction pool.
func (pool *TxPool) Stop() {
// Unsubscribe all subscriptions registered from txpool
pool.scope.Close()
// Unsubscribe subscriptions registered from blockchain
pool.chainHeadSub.Unsubscribe()
pool.wg.Wait()
if pool.journal != nil {
pool.journal.close()
}
log.Info("Transaction pool stopped")
}
// SubscribeNewTxsEvent registers a subscription of NewTxsEvent and
// starts sending event to the given channel.
func (pool *TxPool) SubscribeNewTxsEvent(ch chan<- NewTxsEvent) event.Subscription {
return pool.scope.Track(pool.txFeed.Subscribe(ch))
}
// GasPrice returns the current gas price enforced by the transaction pool.
func (pool *TxPool) GasPrice() *big.Int {
pool.mu.RLock()
defer pool.mu.RUnlock()
return new(big.Int).Set(pool.gasPrice)
}
// SetGasPrice updates the minimum price required by the transaction pool for a
// new transaction, and drops all transactions below this threshold.
func (pool *TxPool) SetGasPrice(price *big.Int) {
pool.mu.Lock()
defer pool.mu.Unlock()
pool.gasPrice = price
for _, tx := range pool.priced.Cap(price, pool.locals) {
pool.removeTx(tx.Hash(), false)
}
log.Info("Transaction pool price threshold updated", "price", price)
}
// Nonce returns the next nonce of an account, with all transactions executable
// by the pool already applied on top.
func (pool *TxPool) Nonce(addr common.Address) uint64 {
pool.mu.RLock()
defer pool.mu.RUnlock()
return pool.pendingNonces.get(addr)
}
// Stats retrieves the current pool stats, namely the number of pending and the
// number of queued (non-executable) transactions.
func (pool *TxPool) Stats() (int, int) {
pool.mu.RLock()
defer pool.mu.RUnlock()
return pool.stats()
}
// stats retrieves the current pool stats, namely the number of pending and the
// number of queued (non-executable) transactions.
func (pool *TxPool) stats() (int, int) {
pending := 0
for _, list := range pool.pending {
pending += list.Len()
}
queued := 0
for _, list := range pool.queue {
queued += list.Len()
}
return pending, queued
}
// Content retrieves the data content of the transaction pool, returning all the
// pending as well as queued transactions, grouped by account and sorted by nonce.
func (pool *TxPool) Content() (map[common.Address]types.Transactions, map[common.Address]types.Transactions) {
pool.mu.Lock()
defer pool.mu.Unlock()
pending := make(map[common.Address]types.Transactions)
for addr, list := range pool.pending {
pending[addr] = list.Flatten()
}
queued := make(map[common.Address]types.Transactions)
for addr, list := range pool.queue {
queued[addr] = list.Flatten()
}
return pending, queued
}
// Pending retrieves all currently processable transactions, grouped by origin
// account and sorted by nonce. The returned transaction set is a copy and can be
// freely modified by calling code.
func (pool *TxPool) Pending() (map[common.Address]types.Transactions, error) {
pool.mu.Lock()
defer pool.mu.Unlock()
pending := make(map[common.Address]types.Transactions)
for addr, list := range pool.pending {
pending[addr] = list.Flatten()
}
return pending, nil
}
// Locals retrieves the accounts currently considered local by the pool.
func (pool *TxPool) Locals() []common.Address {
pool.mu.Lock()
defer pool.mu.Unlock()
return pool.locals.flatten()
}
// local retrieves all currently known local transactions, grouped by origin
// account and sorted by nonce. The returned transaction set is a copy and can be
// freely modified by calling code.
func (pool *TxPool) local() map[common.Address]types.Transactions {
txs := make(map[common.Address]types.Transactions)
for addr := range pool.locals.accounts {
if pending := pool.pending[addr]; pending != nil {
txs[addr] = append(txs[addr], pending.Flatten()...)
}
if queued := pool.queue[addr]; queued != nil {
txs[addr] = append(txs[addr], queued.Flatten()...)
}
}
return txs
}
// validateTx checks whether a transaction is valid according to the consensus
// rules and adheres to some heuristic limits of the local node (price and size).
func (pool *TxPool) validateTx(tx *types.Transaction, local bool) error {
// Reject transactions over defined size to prevent DOS attacks
if uint64(tx.Size()) > txMaxSize {
return ErrOversizedData
}
// Transactions can't be negative. This may never happen using RLP decoded
// transactions but may occur if you create a transaction using the RPC.
if tx.Value().Sign() < 0 {
return ErrNegativeValue
}
// Ensure the transaction doesn't exceed the current block limit gas.
if pool.currentMaxGas < tx.Gas() {
return ErrGasLimit
}
// Make sure the transaction is signed properly
from, err := types.Sender(pool.signer, tx)
if err != nil {
return ErrInvalidSender
}
// Drop non-local transactions under our own minimal accepted gas price
local = local || pool.locals.contains(from) // account may be local even if the transaction arrived from the network
if !local && tx.GasPriceIntCmp(pool.gasPrice) < 0 {
return ErrUnderpriced
}
// Ensure the transaction adheres to nonce ordering
if pool.currentState.GetNonce(from) > tx.Nonce() {
return ErrNonceTooLow
}
// Transactor should have enough funds to cover the costs
// cost == V + GP * GL
if pool.currentState.GetBalance(from).Cmp(tx.Cost()) < 0 {
return ErrInsufficientFunds
}
// Ensure the transaction has more gas than the basic tx fee.
intrGas, err := IntrinsicGas(tx.Data(), tx.To() == nil, true, pool.istanbul)
if err != nil {
return err
}
if tx.Gas() < intrGas {
return ErrIntrinsicGas
}
return nil
}
// add validates a transaction and inserts it into the non-executable queue for later
// pending promotion and execution. If the transaction is a replacement for an already
// pending or queued one, it overwrites the previous transaction if its price is higher.
//
// If a newly added transaction is marked as local, its sending account will be
// whitelisted, preventing any associated transaction from being dropped out of the pool
// due to pricing constraints.
func (pool *TxPool) add(tx *types.Transaction, local bool) (replaced bool, err error) {
// If the transaction is already known, discard it
hash := tx.Hash()
if pool.all.Get(hash) != nil {
log.Trace("Discarding already known transaction", "hash", hash)
knownTxMeter.Mark(1)
return false, ErrAlreadyKnown
}
// If the transaction fails basic validation, discard it
if err := pool.validateTx(tx, local); err != nil {
log.Trace("Discarding invalid transaction", "hash", hash, "err", err)
invalidTxMeter.Mark(1)
return false, err
}
// If the transaction pool is full, discard underpriced transactions
if uint64(pool.all.Count()) >= pool.config.GlobalSlots+pool.config.GlobalQueue {
// If the new transaction is underpriced, don't accept it
if !local && pool.priced.Underpriced(tx, pool.locals) {
log.Trace("Discarding underpriced transaction", "hash", hash, "price", tx.GasPrice())
underpricedTxMeter.Mark(1)
return false, ErrUnderpriced
}
// New transaction is better than our worse ones, make room for it
drop := pool.priced.Discard(pool.all.Slots()-int(pool.config.GlobalSlots+pool.config.GlobalQueue)+numSlots(tx), pool.locals)
for _, tx := range drop {
log.Trace("Discarding freshly underpriced transaction", "hash", tx.Hash(), "price", tx.GasPrice())
underpricedTxMeter.Mark(1)
pool.removeTx(tx.Hash(), false)
}
}
// Try to replace an existing transaction in the pending pool
from, _ := types.Sender(pool.signer, tx) // already validated
if list := pool.pending[from]; list != nil && list.Overlaps(tx) {
// Nonce already pending, check if required price bump is met
inserted, old := list.Add(tx, pool.config.PriceBump)
if !inserted {
pendingDiscardMeter.Mark(1)
return false, ErrReplaceUnderpriced
}
// New transaction is better, replace old one
if old != nil {
pool.all.Remove(old.Hash())
pool.priced.Removed(1)
pendingReplaceMeter.Mark(1)
}
pool.all.Add(tx)
pool.priced.Put(tx)
pool.journalTx(from, tx)
pool.queueTxEvent(tx)
log.Trace("Pooled new executable transaction", "hash", hash, "from", from, "to", tx.To())
// Successful promotion, bump the heartbeat
pool.beats[from] = time.Now()
return old != nil, nil
}
// New transaction isn't replacing a pending one, push into queue
replaced, err = pool.enqueueTx(hash, tx)
if err != nil {
return false, err
}
// Mark local addresses and journal local transactions
if local {
if !pool.locals.contains(from) {
log.Info("Setting new local account", "address", from)
pool.locals.add(from)
}
}
if local || pool.locals.contains(from) {
localGauge.Inc(1)
}
pool.journalTx(from, tx)
log.Trace("Pooled new future transaction", "hash", hash, "from", from, "to", tx.To())
return replaced, nil
}
// enqueueTx inserts a new transaction into the non-executable transaction queue.
//
// Note, this method assumes the pool lock is held!
func (pool *TxPool) enqueueTx(hash common.Hash, tx *types.Transaction) (bool, error) {
// Try to insert the transaction into the future queue
from, _ := types.Sender(pool.signer, tx) // already validated
if pool.queue[from] == nil {
pool.queue[from] = newTxList(false)
}
inserted, old := pool.queue[from].Add(tx, pool.config.PriceBump)
if !inserted {
// An older transaction was better, discard this
queuedDiscardMeter.Mark(1)
return false, ErrReplaceUnderpriced
}
// Discard any previous transaction and mark this
if old != nil {
pool.all.Remove(old.Hash())
pool.priced.Removed(1)
queuedReplaceMeter.Mark(1)
} else {
// Nothing was replaced, bump the queued counter
queuedGauge.Inc(1)
}
if pool.all.Get(hash) == nil {
pool.all.Add(tx)
pool.priced.Put(tx)
}
// If we never record the heartbeat, do it right now.
if _, exist := pool.beats[from]; !exist {
pool.beats[from] = time.Now()
}
return old != nil, nil
}
// journalTx adds the specified transaction to the local disk journal if it is
// deemed to have been sent from a local account.
func (pool *TxPool) journalTx(from common.Address, tx *types.Transaction) {
// Only journal if it's enabled and the transaction is local
if pool.journal == nil || !pool.locals.contains(from) {
return
}
if err := pool.journal.insert(tx); err != nil {
log.Warn("Failed to journal local transaction", "err", err)
}
}
// promoteTx adds a transaction to the pending (processable) list of transactions
// and returns whether it was inserted or an older was better.
//
// Note, this method assumes the pool lock is held!
func (pool *TxPool) promoteTx(addr common.Address, hash common.Hash, tx *types.Transaction) bool {
// Try to insert the transaction into the pending queue
if pool.pending[addr] == nil {
pool.pending[addr] = newTxList(true)
}
list := pool.pending[addr]
inserted, old := list.Add(tx, pool.config.PriceBump)
if !inserted {
// An older transaction was better, discard this
pool.all.Remove(hash)
pool.priced.Removed(1)
pendingDiscardMeter.Mark(1)
return false
}
// Otherwise discard any previous transaction and mark this
if old != nil {
pool.all.Remove(old.Hash())
pool.priced.Removed(1)
pendingReplaceMeter.Mark(1)
} else {
// Nothing was replaced, bump the pending counter
pendingGauge.Inc(1)
}
// Failsafe to work around direct pending inserts (tests)
if pool.all.Get(hash) == nil {
pool.all.Add(tx)
pool.priced.Put(tx)
}
// Set the potentially new pending nonce and notify any subsystems of the new tx
pool.pendingNonces.set(addr, tx.Nonce()+1)
// Successful promotion, bump the heartbeat
pool.beats[addr] = time.Now()
return true
}
// AddLocals enqueues a batch of transactions into the pool if they are valid, marking the
// senders as a local ones, ensuring they go around the local pricing constraints.
//
// This method is used to add transactions from the RPC API and performs synchronous pool
// reorganization and event propagation.
func (pool *TxPool) AddLocals(txs []*types.Transaction) []error {
return pool.addTxs(txs, !pool.config.NoLocals, true)
}
// AddLocal enqueues a single local transaction into the pool if it is valid. This is
// a convenience wrapper aroundd AddLocals.
func (pool *TxPool) AddLocal(tx *types.Transaction) error {
errs := pool.AddLocals([]*types.Transaction{tx})
return errs[0]
}
// AddRemotes enqueues a batch of transactions into the pool if they are valid. If the
// senders are not among the locally tracked ones, full pricing constraints will apply.
//
// This method is used to add transactions from the p2p network and does not wait for pool
// reorganization and internal event propagation.
func (pool *TxPool) AddRemotes(txs []*types.Transaction) []error {
return pool.addTxs(txs, false, false)
}
// This is like AddRemotes, but waits for pool reorganization. Tests use this method.
func (pool *TxPool) AddRemotesSync(txs []*types.Transaction) []error {
return pool.addTxs(txs, false, true)
}
// This is like AddRemotes with a single transaction, but waits for pool reorganization. Tests use this method.
func (pool *TxPool) addRemoteSync(tx *types.Transaction) error {
errs := pool.AddRemotesSync([]*types.Transaction{tx})
return errs[0]
}
// AddRemote enqueues a single transaction into the pool if it is valid. This is a convenience
// wrapper around AddRemotes.
//
// Deprecated: use AddRemotes
func (pool *TxPool) AddRemote(tx *types.Transaction) error {
errs := pool.AddRemotes([]*types.Transaction{tx})
return errs[0]
}
// addTxs attempts to queue a batch of transactions if they are valid.
func (pool *TxPool) addTxs(txs []*types.Transaction, local, sync bool) []error {
// Filter out known ones without obtaining the pool lock or recovering signatures
var (
errs = make([]error, len(txs))
news = make([]*types.Transaction, 0, len(txs))
)
for i, tx := range txs {
// If the transaction is known, pre-set the error slot
if pool.all.Get(tx.Hash()) != nil {
errs[i] = ErrAlreadyKnown
knownTxMeter.Mark(1)
continue
}
// Exclude transactions with invalid signatures as soon as
// possible and cache senders in transactions before
// obtaining lock
_, err := types.Sender(pool.signer, tx)
if err != nil {
errs[i] = ErrInvalidSender
invalidTxMeter.Mark(1)
continue
}
// Accumulate all unknown transactions for deeper processing
news = append(news, tx)
}
if len(news) == 0 {
return errs
}
// Process all the new transaction and merge any errors into the original slice
pool.mu.Lock()
newErrs, dirtyAddrs := pool.addTxsLocked(news, local)
pool.mu.Unlock()
var nilSlot = 0
for _, err := range newErrs {
for errs[nilSlot] != nil {
nilSlot++
}
errs[nilSlot] = err
}
// Reorg the pool internals if needed and return
done := pool.requestPromoteExecutables(dirtyAddrs)
if sync {
<-done
}
return errs
}
// addTxsLocked attempts to queue a batch of transactions if they are valid.
// The transaction pool lock must be held.
func (pool *TxPool) addTxsLocked(txs []*types.Transaction, local bool) ([]error, *accountSet) {
dirty := newAccountSet(pool.signer)
errs := make([]error, len(txs))
for i, tx := range txs {
replaced, err := pool.add(tx, local)
errs[i] = err
if err == nil && !replaced {
dirty.addTx(tx)
}
}
validTxMeter.Mark(int64(len(dirty.accounts)))
return errs, dirty
}
// Status returns the status (unknown/pending/queued) of a batch of transactions
// identified by their hashes.
func (pool *TxPool) Status(hashes []common.Hash) []TxStatus {
status := make([]TxStatus, len(hashes))
for i, hash := range hashes {
tx := pool.Get(hash)
if tx == nil {
continue
}
from, _ := types.Sender(pool.signer, tx) // already validated
pool.mu.RLock()
if txList := pool.pending[from]; txList != nil && txList.txs.items[tx.Nonce()] != nil {
status[i] = TxStatusPending
} else if txList := pool.queue[from]; txList != nil && txList.txs.items[tx.Nonce()] != nil {
status[i] = TxStatusQueued
}
// implicit else: the tx may have been included into a block between
// checking pool.Get and obtaining the lock. In that case, TxStatusUnknown is correct
pool.mu.RUnlock()
}
return status
}
// Get returns a transaction if it is contained in the pool and nil otherwise.
func (pool *TxPool) Get(hash common.Hash) *types.Transaction {
return pool.all.Get(hash)
}
// Has returns an indicator whether txpool has a transaction cached with the
// given hash.
func (pool *TxPool) Has(hash common.Hash) bool {
return pool.all.Get(hash) != nil
}
// removeTx removes a single transaction from the queue, moving all subsequent
// transactions back to the future queue.
func (pool *TxPool) removeTx(hash common.Hash, outofbound bool) {
// Fetch the transaction we wish to delete
tx := pool.all.Get(hash)
if tx == nil {
return
}
addr, _ := types.Sender(pool.signer, tx) // already validated during insertion
// Remove it from the list of known transactions
pool.all.Remove(hash)
if outofbound {
pool.priced.Removed(1)
}
if pool.locals.contains(addr) {
localGauge.Dec(1)
}
// Remove the transaction from the pending lists and reset the account nonce
if pending := pool.pending[addr]; pending != nil {
if removed, invalids := pending.Remove(tx); removed {
// If no more pending transactions are left, remove the list
if pending.Empty() {
delete(pool.pending, addr)
}
// Postpone any invalidated transactions
for _, tx := range invalids {
pool.enqueueTx(tx.Hash(), tx)
}
// Update the account nonce if needed
pool.pendingNonces.setIfLower(addr, tx.Nonce())
// Reduce the pending counter
pendingGauge.Dec(int64(1 + len(invalids)))
return
}
}
// Transaction is in the future queue
if future := pool.queue[addr]; future != nil {
if removed, _ := future.Remove(tx); removed {
// Reduce the queued counter
queuedGauge.Dec(1)
}
if future.Empty() {
delete(pool.queue, addr)
delete(pool.beats, addr)
}
}
}
// requestPromoteExecutables requests a pool reset to the new head block.
// The returned channel is closed when the reset has occurred.
func (pool *TxPool) requestReset(oldHead *types.Header, newHead *types.Header) chan struct{} {
select {
case pool.reqResetCh <- &txpoolResetRequest{oldHead, newHead}:
return <-pool.reorgDoneCh
case <-pool.reorgShutdownCh:
return pool.reorgShutdownCh
}
}
// requestPromoteExecutables requests transaction promotion checks for the given addresses.
// The returned channel is closed when the promotion checks have occurred.
func (pool *TxPool) requestPromoteExecutables(set *accountSet) chan struct{} {
select {
case pool.reqPromoteCh <- set:
return <-pool.reorgDoneCh
case <-pool.reorgShutdownCh:
return pool.reorgShutdownCh
}
}
// queueTxEvent enqueues a transaction event to be sent in the next reorg run.
func (pool *TxPool) queueTxEvent(tx *types.Transaction) {
select {
case pool.queueTxEventCh <- tx:
case <-pool.reorgShutdownCh:
}
}
// scheduleReorgLoop schedules runs of reset and promoteExecutables. Code above should not
// call those methods directly, but request them being run using requestReset and
// requestPromoteExecutables instead.
func (pool *TxPool) scheduleReorgLoop() {
defer pool.wg.Done()
var (
curDone chan struct{} // non-nil while runReorg is active
nextDone = make(chan struct{})
launchNextRun bool
reset *txpoolResetRequest
dirtyAccounts *accountSet
queuedEvents = make(map[common.Address]*txSortedMap)
)
for {
// Launch next background reorg if needed
if curDone == nil && launchNextRun {
// Run the background reorg and announcements
go pool.runReorg(nextDone, reset, dirtyAccounts, queuedEvents)
// Prepare everything for the next round of reorg
curDone, nextDone = nextDone, make(chan struct{})
launchNextRun = false
reset, dirtyAccounts = nil, nil
queuedEvents = make(map[common.Address]*txSortedMap)
}
select {
case req := <-pool.reqResetCh:
// Reset request: update head if request is already pending.
if reset == nil {
reset = req
} else {
reset.newHead = req.newHead
}
launchNextRun = true
pool.reorgDoneCh <- nextDone
case req := <-pool.reqPromoteCh:
// Promote request: update address set if request is already pending.
if dirtyAccounts == nil {
dirtyAccounts = req
} else {
dirtyAccounts.merge(req)
}
launchNextRun = true
pool.reorgDoneCh <- nextDone
case tx := <-pool.queueTxEventCh:
// Queue up the event, but don't schedule a reorg. It's up to the caller to
// request one later if they want the events sent.
addr, _ := types.Sender(pool.signer, tx)
if _, ok := queuedEvents[addr]; !ok {
queuedEvents[addr] = newTxSortedMap()
}
queuedEvents[addr].Put(tx)
case <-curDone:
curDone = nil
case <-pool.reorgShutdownCh:
// Wait for current run to finish.
if curDone != nil {
<-curDone
}
close(nextDone)
return
}
}
}
// runReorg runs reset and promoteExecutables on behalf of scheduleReorgLoop.
func (pool *TxPool) runReorg(done chan struct{}, reset *txpoolResetRequest, dirtyAccounts *accountSet, events map[common.Address]*txSortedMap) {
defer close(done)
var promoteAddrs []common.Address
if dirtyAccounts != nil {
promoteAddrs = dirtyAccounts.flatten()
}
pool.mu.Lock()
if reset != nil {
// Reset from the old head to the new, rescheduling any reorged transactions
pool.reset(reset.oldHead, reset.newHead)
// Nonces were reset, discard any events that became stale
for addr := range events {
events[addr].Forward(pool.pendingNonces.get(addr))
if events[addr].Len() == 0 {
delete(events, addr)
}
}
// Reset needs promote for all addresses
promoteAddrs = promoteAddrs[:0]
for addr := range pool.queue {
promoteAddrs = append(promoteAddrs, addr)
}
}
// Check for pending transactions for every account that sent new ones
promoted := pool.promoteExecutables(promoteAddrs)
// If a new block appeared, validate the pool of pending transactions. This will
// remove any transaction that has been included in the block or was invalidated
// because of another transaction (e.g. higher gas price).
if reset != nil {
pool.demoteUnexecutables()
}
// Ensure pool.queue and pool.pending sizes stay within the configured limits.
pool.truncatePending()
pool.truncateQueue()
// Update all accounts to the latest known pending nonce
for addr, list := range pool.pending {
highestPending := list.LastElement()
pool.pendingNonces.set(addr, highestPending.Nonce()+1)
}
pool.mu.Unlock()
// Notify subsystems for newly added transactions
for _, tx := range promoted {
addr, _ := types.Sender(pool.signer, tx)
if _, ok := events[addr]; !ok {
events[addr] = newTxSortedMap()
}
events[addr].Put(tx)
}
if len(events) > 0 {
var txs []*types.Transaction
for _, set := range events {
txs = append(txs, set.Flatten()...)
}
pool.txFeed.Send(NewTxsEvent{txs})
}
}
// reset retrieves the current state of the blockchain and ensures the content
// of the transaction pool is valid with regard to the chain state.
func (pool *TxPool) reset(oldHead, newHead *types.Header) {
// If we're reorging an old state, reinject all dropped transactions
var reinject types.Transactions
if oldHead != nil && oldHead.Hash() != newHead.ParentHash {
// If the reorg is too deep, avoid doing it (will happen during fast sync)
oldNum := oldHead.Number.Uint64()
newNum := newHead.Number.Uint64()
if depth := uint64(math.Abs(float64(oldNum) - float64(newNum))); depth > 64 {
log.Debug("Skipping deep transaction reorg", "depth", depth)
} else {
// Reorg seems shallow enough to pull in all transactions into memory
var discarded, included types.Transactions
var (
rem = pool.chain.GetBlock(oldHead.Hash(), oldHead.Number.Uint64())
add = pool.chain.GetBlock(newHead.Hash(), newHead.Number.Uint64())
)
if rem == nil {
// This can happen if a setHead is performed, where we simply discard the old
// head from the chain.
// If that is the case, we don't have the lost transactions any more, and
// there's nothing to add
if newNum < oldNum {
// If the reorg ended up on a lower number, it's indicative of setHead being the cause
log.Debug("Skipping transaction reset caused by setHead",
"old", oldHead.Hash(), "oldnum", oldNum, "new", newHead.Hash(), "newnum", newNum)
} else {
// If we reorged to a same or higher number, then it's not a case of setHead
log.Warn("Transaction pool reset with missing oldhead",
"old", oldHead.Hash(), "oldnum", oldNum, "new", newHead.Hash(), "newnum", newNum)
}
return
}
for rem.NumberU64() > add.NumberU64() {
discarded = append(discarded, rem.Transactions()...)
if rem = pool.chain.GetBlock(rem.ParentHash(), rem.NumberU64()-1); rem == nil {
log.Error("Unrooted old chain seen by tx pool", "block", oldHead.Number, "hash", oldHead.Hash())
return
}
}
for add.NumberU64() > rem.NumberU64() {
included = append(included, add.Transactions()...)
if add = pool.chain.GetBlock(add.ParentHash(), add.NumberU64()-1); add == nil {
log.Error("Unrooted new chain seen by tx pool", "block", newHead.Number, "hash", newHead.Hash())
return
}
}
for rem.Hash() != add.Hash() {
discarded = append(discarded, rem.Transactions()...)
if rem = pool.chain.GetBlock(rem.ParentHash(), rem.NumberU64()-1); rem == nil {
log.Error("Unrooted old chain seen by tx pool", "block", oldHead.Number, "hash", oldHead.Hash())
return
}
included = append(included, add.Transactions()...)
if add = pool.chain.GetBlock(add.ParentHash(), add.NumberU64()-1); add == nil {
log.Error("Unrooted new chain seen by tx pool", "block", newHead.Number, "hash", newHead.Hash())
return
}
}
reinject = types.TxDifference(discarded, included)
}
}
// Initialize the internal state to the current head
if newHead == nil {
newHead = pool.chain.CurrentBlock().Header() // Special case during testing
}
statedb, err := pool.chain.StateAt(newHead.Root)
if err != nil {
log.Error("Failed to reset txpool state", "err", err)
return
}
pool.currentState = statedb
pool.pendingNonces = newTxNoncer(statedb)
pool.currentMaxGas = newHead.GasLimit
// Inject any transactions discarded due to reorgs
log.Debug("Reinjecting stale transactions", "count", len(reinject))
senderCacher.recover(pool.signer, reinject)
pool.addTxsLocked(reinject, false)
// Update all fork indicator by next pending block number.
next := new(big.Int).Add(newHead.Number, big.NewInt(1))
pool.istanbul = pool.chainconfig.IsIstanbul(next)
}
// promoteExecutables moves transactions that have become processable from the
// future queue to the set of pending transactions. During this process, all
// invalidated transactions (low nonce, low balance) are deleted.
func (pool *TxPool) promoteExecutables(accounts []common.Address) []*types.Transaction {
// Track the promoted transactions to broadcast them at once
var promoted []*types.Transaction
// Iterate over all accounts and promote any executable transactions
for _, addr := range accounts {
list := pool.queue[addr]
if list == nil {
continue // Just in case someone calls with a non existing account
}
// Drop all transactions that are deemed too old (low nonce)
forwards := list.Forward(pool.currentState.GetNonce(addr))
for _, tx := range forwards {
hash := tx.Hash()
pool.all.Remove(hash)
}
log.Trace("Removed old queued transactions", "count", len(forwards))
// Drop all transactions that are too costly (low balance or out of gas)
drops, _ := list.Filter(pool.currentState.GetBalance(addr), pool.currentMaxGas)
for _, tx := range drops {
hash := tx.Hash()
pool.all.Remove(hash)
}
log.Trace("Removed unpayable queued transactions", "count", len(drops))
queuedNofundsMeter.Mark(int64(len(drops)))
// Gather all executable transactions and promote them
readies := list.Ready(pool.pendingNonces.get(addr))
for _, tx := range readies {
hash := tx.Hash()
if pool.promoteTx(addr, hash, tx) {
promoted = append(promoted, tx)
}
}
log.Trace("Promoted queued transactions", "count", len(promoted))
queuedGauge.Dec(int64(len(readies)))
// Drop all transactions over the allowed limit
var caps types.Transactions
if !pool.locals.contains(addr) {
caps = list.Cap(int(pool.config.AccountQueue))
for _, tx := range caps {
hash := tx.Hash()
pool.all.Remove(hash)
log.Trace("Removed cap-exceeding queued transaction", "hash", hash)
}
queuedRateLimitMeter.Mark(int64(len(caps)))
}
// Mark all the items dropped as removed
pool.priced.Removed(len(forwards) + len(drops) + len(caps))
queuedGauge.Dec(int64(len(forwards) + len(drops) + len(caps)))
if pool.locals.contains(addr) {
localGauge.Dec(int64(len(forwards) + len(drops) + len(caps)))
}
// Delete the entire queue entry if it became empty.
if list.Empty() {
delete(pool.queue, addr)
delete(pool.beats, addr)
}
}
return promoted
}
// truncatePending removes transactions from the pending queue if the pool is above the
// pending limit. The algorithm tries to reduce transaction counts by an approximately
// equal number for all for accounts with many pending transactions.
func (pool *TxPool) truncatePending() {
pending := uint64(0)
for _, list := range pool.pending {
pending += uint64(list.Len())
}
if pending <= pool.config.GlobalSlots {
return
}
pendingBeforeCap := pending
// Assemble a spam order to penalize large transactors first
spammers := prque.New(nil)
for addr, list := range pool.pending {
// Only evict transactions from high rollers
if !pool.locals.contains(addr) && uint64(list.Len()) > pool.config.AccountSlots {
spammers.Push(addr, int64(list.Len()))
}
}
// Gradually drop transactions from offenders
offenders := []common.Address{}
for pending > pool.config.GlobalSlots && !spammers.Empty() {
// Retrieve the next offender if not local address
offender, _ := spammers.Pop()
offenders = append(offenders, offender.(common.Address))
// Equalize balances until all the same or below threshold
if len(offenders) > 1 {
// Calculate the equalization threshold for all current offenders
threshold := pool.pending[offender.(common.Address)].Len()
// Iteratively reduce all offenders until below limit or threshold reached
for pending > pool.config.GlobalSlots && pool.pending[offenders[len(offenders)-2]].Len() > threshold {
for i := 0; i < len(offenders)-1; i++ {
list := pool.pending[offenders[i]]
caps := list.Cap(list.Len() - 1)
for _, tx := range caps {
// Drop the transaction from the global pools too
hash := tx.Hash()
pool.all.Remove(hash)
// Update the account nonce to the dropped transaction
pool.pendingNonces.setIfLower(offenders[i], tx.Nonce())
log.Trace("Removed fairness-exceeding pending transaction", "hash", hash)
}
pool.priced.Removed(len(caps))
pendingGauge.Dec(int64(len(caps)))
if pool.locals.contains(offenders[i]) {
localGauge.Dec(int64(len(caps)))
}
pending--
}
}
}
}
// If still above threshold, reduce to limit or min allowance
if pending > pool.config.GlobalSlots && len(offenders) > 0 {
for pending > pool.config.GlobalSlots && uint64(pool.pending[offenders[len(offenders)-1]].Len()) > pool.config.AccountSlots {
for _, addr := range offenders {
list := pool.pending[addr]
caps := list.Cap(list.Len() - 1)
for _, tx := range caps {
// Drop the transaction from the global pools too
hash := tx.Hash()
pool.all.Remove(hash)
// Update the account nonce to the dropped transaction
pool.pendingNonces.setIfLower(addr, tx.Nonce())
log.Trace("Removed fairness-exceeding pending transaction", "hash", hash)
}
pool.priced.Removed(len(caps))
pendingGauge.Dec(int64(len(caps)))
if pool.locals.contains(addr) {
localGauge.Dec(int64(len(caps)))
}
pending--
}
}
}
pendingRateLimitMeter.Mark(int64(pendingBeforeCap - pending))
}
// truncateQueue drops the oldes transactions in the queue if the pool is above the global queue limit.
func (pool *TxPool) truncateQueue() {
queued := uint64(0)
for _, list := range pool.queue {
queued += uint64(list.Len())
}
if queued <= pool.config.GlobalQueue {
return
}
// Sort all accounts with queued transactions by heartbeat
addresses := make(addressesByHeartbeat, 0, len(pool.queue))
for addr := range pool.queue {
if !pool.locals.contains(addr) { // don't drop locals
addresses = append(addresses, addressByHeartbeat{addr, pool.beats[addr]})
}
}
sort.Sort(addresses)
// Drop transactions until the total is below the limit or only locals remain
for drop := queued - pool.config.GlobalQueue; drop > 0 && len(addresses) > 0; {
addr := addresses[len(addresses)-1]
list := pool.queue[addr.address]
addresses = addresses[:len(addresses)-1]
// Drop all transactions if they are less than the overflow
if size := uint64(list.Len()); size <= drop {
for _, tx := range list.Flatten() {
pool.removeTx(tx.Hash(), true)
}
drop -= size
queuedRateLimitMeter.Mark(int64(size))
continue
}
// Otherwise drop only last few transactions
txs := list.Flatten()
for i := len(txs) - 1; i >= 0 && drop > 0; i-- {
pool.removeTx(txs[i].Hash(), true)
drop--
queuedRateLimitMeter.Mark(1)
}
}
}
// demoteUnexecutables removes invalid and processed transactions from the pools
// executable/pending queue and any subsequent transactions that become unexecutable
// are moved back into the future queue.
func (pool *TxPool) demoteUnexecutables() {
// Iterate over all accounts and demote any non-executable transactions
for addr, list := range pool.pending {
nonce := pool.currentState.GetNonce(addr)
// Drop all transactions that are deemed too old (low nonce)
olds := list.Forward(nonce)
for _, tx := range olds {
hash := tx.Hash()
pool.all.Remove(hash)
log.Trace("Removed old pending transaction", "hash", hash)
}
// Drop all transactions that are too costly (low balance or out of gas), and queue any invalids back for later
drops, invalids := list.Filter(pool.currentState.GetBalance(addr), pool.currentMaxGas)
for _, tx := range drops {
hash := tx.Hash()
log.Trace("Removed unpayable pending transaction", "hash", hash)
pool.all.Remove(hash)
}
pool.priced.Removed(len(olds) + len(drops))
pendingNofundsMeter.Mark(int64(len(drops)))
for _, tx := range invalids {
hash := tx.Hash()
log.Trace("Demoting pending transaction", "hash", hash)
pool.enqueueTx(hash, tx)
}
pendingGauge.Dec(int64(len(olds) + len(drops) + len(invalids)))
if pool.locals.contains(addr) {
localGauge.Dec(int64(len(olds) + len(drops) + len(invalids)))
}
// If there's a gap in front, alert (should never happen) and postpone all transactions
if list.Len() > 0 && list.txs.Get(nonce) == nil {
gapped := list.Cap(0)
for _, tx := range gapped {
hash := tx.Hash()
log.Error("Demoting invalidated transaction", "hash", hash)
pool.enqueueTx(hash, tx)
}
pendingGauge.Dec(int64(len(gapped)))
}
// Delete the entire pending entry if it became empty.
if list.Empty() {
delete(pool.pending, addr)
}
}
}
// addressByHeartbeat is an account address tagged with its last activity timestamp.
type addressByHeartbeat struct {
address common.Address
heartbeat time.Time
}
type addressesByHeartbeat []addressByHeartbeat
func (a addressesByHeartbeat) Len() int { return len(a) }
func (a addressesByHeartbeat) Less(i, j int) bool { return a[i].heartbeat.Before(a[j].heartbeat) }
func (a addressesByHeartbeat) Swap(i, j int) { a[i], a[j] = a[j], a[i] }
// accountSet is simply a set of addresses to check for existence, and a signer
// capable of deriving addresses from transactions.
type accountSet struct {
accounts map[common.Address]struct{}
signer types.Signer
cache *[]common.Address
}
// newAccountSet creates a new address set with an associated signer for sender
// derivations.
func newAccountSet(signer types.Signer, addrs ...common.Address) *accountSet {
as := &accountSet{
accounts: make(map[common.Address]struct{}),
signer: signer,
}
for _, addr := range addrs {
as.add(addr)
}
return as
}
// contains checks if a given address is contained within the set.
func (as *accountSet) contains(addr common.Address) bool {
_, exist := as.accounts[addr]
return exist
}
func (as *accountSet) empty() bool {
return len(as.accounts) == 0
}
// containsTx checks if the sender of a given tx is within the set. If the sender
// cannot be derived, this method returns false.
func (as *accountSet) containsTx(tx *types.Transaction) bool {
if addr, err := types.Sender(as.signer, tx); err == nil {
return as.contains(addr)
}
return false
}
// add inserts a new address into the set to track.
func (as *accountSet) add(addr common.Address) {
as.accounts[addr] = struct{}{}
as.cache = nil
}
// addTx adds the sender of tx into the set.
func (as *accountSet) addTx(tx *types.Transaction) {
if addr, err := types.Sender(as.signer, tx); err == nil {
as.add(addr)
}
}
// flatten returns the list of addresses within this set, also caching it for later
// reuse. The returned slice should not be changed!
func (as *accountSet) flatten() []common.Address {
if as.cache == nil {
accounts := make([]common.Address, 0, len(as.accounts))
for account := range as.accounts {
accounts = append(accounts, account)
}
as.cache = &accounts
}
return *as.cache
}
// merge adds all addresses from the 'other' set into 'as'.
func (as *accountSet) merge(other *accountSet) {
for addr := range other.accounts {
as.accounts[addr] = struct{}{}
}
as.cache = nil
}
// txLookup is used internally by TxPool to track transactions while allowing lookup without
// mutex contention.
//
// Note, although this type is properly protected against concurrent access, it
// is **not** a type that should ever be mutated or even exposed outside of the
// transaction pool, since its internal state is tightly coupled with the pools
// internal mechanisms. The sole purpose of the type is to permit out-of-bound
// peeking into the pool in TxPool.Get without having to acquire the widely scoped
// TxPool.mu mutex.
type txLookup struct {
all map[common.Hash]*types.Transaction
slots int
lock sync.RWMutex
}
// newTxLookup returns a new txLookup structure.
func newTxLookup() *txLookup {
return &txLookup{
all: make(map[common.Hash]*types.Transaction),
}
}
// Range calls f on each key and value present in the map.
func (t *txLookup) Range(f func(hash common.Hash, tx *types.Transaction) bool) {
t.lock.RLock()
defer t.lock.RUnlock()
for key, value := range t.all {
if !f(key, value) {
break
}
}
}
// Get returns a transaction if it exists in the lookup, or nil if not found.
func (t *txLookup) Get(hash common.Hash) *types.Transaction {
t.lock.RLock()
defer t.lock.RUnlock()
return t.all[hash]
}
// Count returns the current number of items in the lookup.
func (t *txLookup) Count() int {
t.lock.RLock()
defer t.lock.RUnlock()
return len(t.all)
}
// Slots returns the current number of slots used in the lookup.
func (t *txLookup) Slots() int {
t.lock.RLock()
defer t.lock.RUnlock()
return t.slots
}
// Add adds a transaction to the lookup.
func (t *txLookup) Add(tx *types.Transaction) {
t.lock.Lock()
defer t.lock.Unlock()
t.slots += numSlots(tx)
slotsGauge.Update(int64(t.slots))
t.all[tx.Hash()] = tx
}
// Remove removes a transaction from the lookup.
func (t *txLookup) Remove(hash common.Hash) {
t.lock.Lock()
defer t.lock.Unlock()
t.slots -= numSlots(t.all[hash])
slotsGauge.Update(int64(t.slots))
delete(t.all, hash)
}
// numSlots calculates the number of slots needed for a single transaction.
func numSlots(tx *types.Transaction) int {
return int((tx.Size() + txSlotSize - 1) / txSlotSize)
}