Commit 5e8bc3d5 authored by mergify[bot]'s avatar mergify[bot] Committed by GitHub

Merge pull request #4757 from ethereum-optimism/sequencer-engine-control

op-node: refactor sequencer to use EngineControl, test sequencing
parents c0aaf9b1 755fef5d
package actions
import (
"context"
"github.com/ethereum/go-ethereum/log"
"github.com/stretchr/testify/require"
"github.com/ethereum-optimism/optimism/op-node/eth"
"github.com/ethereum-optimism/optimism/op-node/metrics"
"github.com/ethereum-optimism/optimism/op-node/rollup"
"github.com/ethereum-optimism/optimism/op-node/rollup/derive"
"github.com/ethereum-optimism/optimism/op-node/rollup/driver"
)
// MockL1OriginSelector is a shim to override the origin as sequencer, so we can force it to stay on an older origin.
type MockL1OriginSelector struct {
actual *driver.L1OriginSelector
originOverride eth.L1BlockRef // override which origin gets picked
}
func (m *MockL1OriginSelector) FindL1Origin(ctx context.Context, l1Head eth.L1BlockRef, l2Head eth.L2BlockRef) (eth.L1BlockRef, error) {
if m.originOverride != (eth.L1BlockRef{}) {
return m.originOverride, nil
}
return m.actual.FindL1Origin(ctx, l1Head, l2Head)
}
// L2Sequencer is an actor that functions like a rollup node,
// without the full P2P/API/Node stack, but just the derivation state, and simplified driver with sequencing ability.
type L2Sequencer struct {
L2Verifier
sequencer *driver.Sequencer
l1OriginSelector *driver.L1OriginSelector
seqOldOrigin bool // stay on current L1 origin when sequencing a block, unless forced to adopt the next origin
sequencer *driver.Sequencer
failL2GossipUnsafeBlock error // mock error
mockL1OriginSelector *MockL1OriginSelector
}
func NewL2Sequencer(t Testing, log log.Logger, l1 derive.L1Fetcher, eng L2API, cfg *rollup.Config, seqConfDepth uint64) *L2Sequencer {
ver := NewL2Verifier(t, log, l1, eng, cfg)
attrBuilder := derive.NewFetchingAttributesBuilder(cfg, l1, eng)
l1OriginSelector := &MockL1OriginSelector{
actual: driver.NewL1OriginSelector(log, cfg, l1, seqConfDepth),
}
return &L2Sequencer{
L2Verifier: *ver,
sequencer: driver.NewSequencer(log, cfg, eng, ver.derivation, attrBuilder, metrics.NoopMetrics),
l1OriginSelector: driver.NewL1OriginSelector(log, cfg, l1, seqConfDepth),
seqOldOrigin: false,
sequencer: driver.NewSequencer(log, cfg, ver.derivation, attrBuilder, l1OriginSelector),
mockL1OriginSelector: l1OriginSelector,
failL2GossipUnsafeBlock: nil,
}
}
......@@ -47,22 +62,7 @@ func (s *L2Sequencer) ActL2StartBlock(t Testing) {
return
}
parent := s.derivation.UnsafeL2Head()
var origin eth.L1BlockRef
if s.seqOldOrigin {
// force old origin, for testing purposes
oldOrigin, err := s.l1.L1BlockRefByHash(t.Ctx(), parent.L1Origin.Hash)
require.NoError(t, err, "failed to get current origin: %s", parent.L1Origin)
origin = oldOrigin
s.seqOldOrigin = false // don't repeat this
} else {
// select origin the real way
l1Origin, err := s.l1OriginSelector.FindL1Origin(t.Ctx(), s.l1State.L1Head(), parent)
require.NoError(t, err)
origin = l1Origin
}
err := s.sequencer.StartBuildingBlock(t.Ctx(), origin)
err := s.sequencer.StartBuildingBlock(t.Ctx(), s.l1State.L1Head())
require.NoError(t, err, "failed to start block building")
s.l2Building = true
......@@ -76,24 +76,21 @@ func (s *L2Sequencer) ActL2EndBlock(t Testing) {
}
s.l2Building = false
payload, err := s.sequencer.CompleteBuildingBlock(t.Ctx())
_, err := s.sequencer.CompleteBuildingBlock(t.Ctx())
// TODO: there may be legitimate temporary errors here, if we mock engine API RPC-failure.
// For advanced tests we can catch those and print a warning instead.
require.NoError(t, err)
ref, err := derive.PayloadToBlockRef(payload, &s.rollupCfg.Genesis)
require.NoError(t, err, "payload must convert to block ref")
s.derivation.SetUnsafeHead(ref)
// TODO: action-test publishing of payload on p2p
}
// ActL2KeepL1Origin makes the sequencer use the current L1 origin, even if the next origin is available.
func (s *L2Sequencer) ActL2KeepL1Origin(t Testing) {
if s.seqOldOrigin { // don't do this twice
t.InvalidAction("already decided to keep old L1 origin")
return
}
s.seqOldOrigin = true
parent := s.derivation.UnsafeL2Head()
// force old origin, for testing purposes
oldOrigin, err := s.l1.L1BlockRefByHash(t.Ctx(), parent.L1Origin.Hash)
require.NoError(t, err, "failed to get current origin: %s", parent.L1Origin)
s.mockL1OriginSelector.originOverride = oldOrigin
}
// ActBuildToL1Head builds empty blocks until (incl.) the L1 head becomes the L2 origin
......@@ -109,7 +106,7 @@ func (s *L2Sequencer) ActBuildToL1Head(t Testing) {
func (s *L2Sequencer) ActBuildToL1HeadExcl(t Testing) {
for {
s.ActL2PipelineFull(t)
nextOrigin, err := s.l1OriginSelector.FindL1Origin(t.Ctx(), s.l1State.L1Head(), s.derivation.UnsafeL2Head())
nextOrigin, err := s.mockL1OriginSelector.FindL1Origin(t.Ctx(), s.l1State.L1Head(), s.derivation.UnsafeL2Head())
require.NoError(t, err)
if nextOrigin.Number >= s.l1State.L1Head().Number {
break
......
......@@ -148,13 +148,6 @@ func (dp *DerivationPipeline) BuildingPayload() (onto eth.L2BlockRef, id eth.Pay
return dp.eng.BuildingPayload()
}
// SetUnsafeHead changes the forkchoice state unsafe head, without changing the engine.
//
// deprecated: use the EngineControl interface instead.
func (dp *DerivationPipeline) SetUnsafeHead(head eth.L2BlockRef) {
dp.eng.SetUnsafeHead(head)
}
// AddUnsafePayload schedules an execution payload to be processed, ahead of deriving it from L1
func (dp *DerivationPipeline) AddUnsafePayload(payload *eth.ExecutionPayload) {
dp.eng.AddUnsafePayload(payload)
......
......@@ -14,7 +14,6 @@ import (
type Metrics interface {
RecordPipelineReset()
RecordSequencingError()
RecordPublishingError()
RecordDerivationError()
......@@ -28,9 +27,8 @@ type Metrics interface {
SetDerivationIdle(idle bool)
RecordL1ReorgDepth(d uint64)
CountSequencedTxs(count int)
SequencerMetrics
EngineMetrics
}
type L1Chain interface {
......@@ -48,7 +46,6 @@ type L2Chain interface {
type DerivationPipeline interface {
Reset()
Step(ctx context.Context) error
SetUnsafeHead(head eth.L2BlockRef)
AddUnsafePayload(payload *eth.ExecutionPayload)
Finalize(ref eth.L1BlockRef)
FinalizedL1() eth.L1BlockRef
......@@ -68,14 +65,12 @@ type L1StateIface interface {
L1Finalized() eth.L1BlockRef
}
type L1OriginSelectorIface interface {
FindL1Origin(ctx context.Context, l1Head eth.L1BlockRef, l2Head eth.L2BlockRef) (eth.L1BlockRef, error)
}
type SequencerIface interface {
StartBuildingBlock(ctx context.Context, l1Origin eth.L1BlockRef) error
StartBuildingBlock(ctx context.Context, l1Head eth.L1BlockRef) error
CompleteBuildingBlock(ctx context.Context) (*eth.ExecutionPayload, error)
PlanNextSequencerAction(sequenceErr error) (delay time.Duration, seal bool, onto eth.BlockID)
PlanNextSequencerAction() time.Duration
RunNextSequencerAction(ctx context.Context, l1Head eth.L1BlockRef) *eth.ExecutionPayload
BuildingOnto() eth.L2BlockRef
}
type Network interface {
......@@ -90,7 +85,10 @@ func NewDriver(driverCfg *Config, cfg *rollup.Config, l2 L2Chain, l1 L1Chain, ne
verifConfDepth := NewConfDepth(driverCfg.VerifierConfDepth, l1State.L1Head, l1)
derivationPipeline := derive.NewDerivationPipeline(log, cfg, verifConfDepth, l2, metrics)
attrBuilder := derive.NewFetchingAttributesBuilder(cfg, l1, l2)
sequencer := NewSequencer(log, cfg, l2, derivationPipeline, attrBuilder, metrics)
engine := derivationPipeline
meteredEngine := NewMeteredEngine(cfg, engine, metrics, log)
sequencer := NewSequencer(log, cfg, meteredEngine, attrBuilder, findL1Origin)
return &Driver{
l1State: l1State,
derivation: derivationPipeline,
......@@ -106,7 +104,6 @@ func NewDriver(driverCfg *Config, cfg *rollup.Config, l2 L2Chain, l1 L1Chain, ne
snapshotLog: snapshotLog,
l1: l1,
l2: l2,
l1OriginSelector: findL1Origin,
sequencer: sequencer,
network: network,
metrics: metrics,
......
......@@ -2,7 +2,6 @@ package driver
import (
"context"
"errors"
"fmt"
"time"
......@@ -20,16 +19,8 @@ type Downloader interface {
FetchReceipts(ctx context.Context, blockHash common.Hash) (eth.BlockInfo, types.Receipts, error)
}
type SequencerMetrics interface {
RecordSequencerBuildingDiffTime(duration time.Duration)
RecordSequencerSealingTime(duration time.Duration)
}
type EngineState interface {
Finalized() eth.L2BlockRef
UnsafeL2Head() eth.L2BlockRef
SafeL2Head() eth.L2BlockRef
Origin() eth.L1BlockRef
type L1OriginSelectorIface interface {
FindL1Origin(ctx context.Context, l1Head eth.L1BlockRef, l2Head eth.L2BlockRef) (eth.L1BlockRef, error)
}
// Sequencer implements the sequencing interface of the driver: it starts and completes block building jobs.
......@@ -37,40 +28,44 @@ type Sequencer struct {
log log.Logger
config *rollup.Config
l2 derive.Engine
engineState EngineState
engine derive.EngineControl
attrBuilder derive.AttributesBuilder
buildingOnto eth.L2BlockRef
buildingID eth.PayloadID
buildingStartTime time.Time
attrBuilder derive.AttributesBuilder
l1OriginSelector L1OriginSelectorIface
metrics SequencerMetrics
// timeNow enables sequencer testing to mock the time
timeNow func() time.Time
nextAction time.Time
}
func NewSequencer(log log.Logger, cfg *rollup.Config, l2 derive.Engine, engineState EngineState, attributesBuilder derive.AttributesBuilder, metrics SequencerMetrics) *Sequencer {
func NewSequencer(log log.Logger, cfg *rollup.Config, engine derive.EngineControl, attributesBuilder derive.AttributesBuilder, l1OriginSelector L1OriginSelectorIface) *Sequencer {
return &Sequencer{
log: log,
config: cfg,
l2: l2,
metrics: metrics,
engineState: engineState,
attrBuilder: attributesBuilder,
log: log,
config: cfg,
engine: engine,
timeNow: time.Now,
attrBuilder: attributesBuilder,
l1OriginSelector: l1OriginSelector,
}
}
// StartBuildingBlock initiates a block building job on top of the given L2 head, safe and finalized blocks, and using the provided l1Origin.
func (d *Sequencer) StartBuildingBlock(ctx context.Context, l1Origin eth.L1BlockRef) error {
l2Head := d.engineState.UnsafeL2Head()
func (d *Sequencer) StartBuildingBlock(ctx context.Context, l1Head eth.L1BlockRef) error {
l2Head := d.engine.UnsafeL2Head()
// Figure out which L1 origin block we're going to be building on top of.
l1Origin, err := d.l1OriginSelector.FindL1Origin(ctx, l1Head, l2Head)
if err != nil {
d.log.Error("Error finding next L1 Origin", "err", err)
return err
}
if !(l2Head.L1Origin.Hash == l1Origin.ParentHash || l2Head.L1Origin.Hash == l1Origin.Hash) {
return fmt.Errorf("cannot build new L2 block with L1 origin %s (parent L1 %s) on current L2 head %s with L1 origin %s", l1Origin, l1Origin.ParentHash, l2Head, l2Head.L1Origin)
}
d.log.Info("creating new block", "parent", l2Head, "l1Origin", l1Origin)
if d.buildingID != (eth.PayloadID{}) { // This may happen when we decide to build a different block in response to a reorg. Or when previous block building failed.
d.log.Warn("did not finish previous block building, starting new building now", "prev_onto", d.buildingOnto, "prev_payload_id", d.buildingID, "new_onto", l2Head)
}
d.buildingStartTime = time.Now()
fetchCtx, cancel := context.WithTimeout(ctx, time.Second*20)
defer cancel()
......@@ -90,20 +85,11 @@ func (d *Sequencer) StartBuildingBlock(ctx context.Context, l1Origin eth.L1Block
"num", l2Head.Number+1, "time", uint64(attrs.Timestamp),
"origin", l1Origin, "origin_time", l1Origin.Time, "noTxPool", attrs.NoTxPool)
// And construct our fork choice state. This is our current fork choice state and will be
// updated as a result of executing the block based on the attributes described above.
fc := eth.ForkchoiceState{
HeadBlockHash: l2Head.Hash,
SafeBlockHash: d.engineState.SafeL2Head().Hash,
FinalizedBlockHash: d.engineState.Finalized().Hash,
}
// Start a payload building process.
id, errTyp, err := derive.StartPayload(ctx, d.l2, fc, attrs)
errTyp, err := d.engine.StartPayload(ctx, l2Head, attrs, false)
if err != nil {
return fmt.Errorf("failed to start building on top of L2 chain %s, error (%d): %w", l2Head, errTyp, err)
}
d.buildingOnto = l2Head
d.buildingID = id
return nil
}
......@@ -111,75 +97,90 @@ func (d *Sequencer) StartBuildingBlock(ctx context.Context, l1Origin eth.L1Block
// Warning: the safe and finalized L2 blocks as viewed during the initiation of the block building are reused for completion of the block building.
// The Execution engine should not change the safe and finalized blocks between start and completion of block building.
func (d *Sequencer) CompleteBuildingBlock(ctx context.Context) (*eth.ExecutionPayload, error) {
if d.buildingID == (eth.PayloadID{}) {
return nil, fmt.Errorf("cannot complete payload building: not currently building a payload")
}
sealingStart := time.Now()
l2Head := d.engineState.UnsafeL2Head()
if d.buildingOnto.Hash != l2Head.Hash {
return nil, fmt.Errorf("engine reorged from %s to %s while building block", d.buildingOnto, l2Head)
}
fc := eth.ForkchoiceState{
HeadBlockHash: l2Head.Hash,
SafeBlockHash: d.engineState.SafeL2Head().Hash,
FinalizedBlockHash: d.engineState.Finalized().Hash,
}
// Actually execute the block and add it to the head of the chain.
payload, errTyp, err := derive.ConfirmPayload(ctx, d.log, d.l2, fc, d.buildingID, false)
payload, errTyp, err := d.engine.ConfirmPayload(ctx)
if err != nil {
return nil, fmt.Errorf("failed to complete building on top of L2 chain %s, id: %s, error (%d): %w", d.buildingOnto, d.buildingID, errTyp, err)
return nil, fmt.Errorf("failed to complete building block: error (%d): %w", errTyp, err)
}
now := time.Now()
sealTime := now.Sub(sealingStart)
buildTime := now.Sub(d.buildingStartTime)
d.metrics.RecordSequencerSealingTime(sealTime)
d.metrics.RecordSequencerBuildingDiffTime(buildTime - time.Duration(d.config.BlockTime)*time.Second)
d.log.Debug("sequenced block", "seal_time", sealTime, "build_time", buildTime)
d.buildingID = eth.PayloadID{}
return payload, nil
}
// PlanNextSequencerAction returns a desired delay till the next action, and if we should seal the block:
// - true whenever we need to complete a block
// - false whenever we need to start a block
func (d *Sequencer) PlanNextSequencerAction(sequenceErr error) (delay time.Duration, seal bool, onto eth.BlockID) {
blockTime := time.Duration(d.config.BlockTime) * time.Second
head := d.engineState.UnsafeL2Head()
// CancelBuildingBlock cancels the current open block building job.
// This sequencer only maintains one block building job at a time.
func (d *Sequencer) CancelBuildingBlock(ctx context.Context) {
// force-cancel, we can always continue block building, and any error is logged by the engine state
_ = d.engine.CancelPayload(ctx, true)
}
// based on the build error, delay and start over again
if sequenceErr != nil {
if errors.Is(sequenceErr, UninitializedL1StateErr) {
// temporary errors are not so bad, just retry in 500ms
return 500 * time.Millisecond, false, head.ID()
} else {
// we just hit an unknown type of error, delay a re-attempt by as much as a block
return blockTime, false, head.ID()
}
// PlanNextSequencerAction returns a desired delay till the RunNextSequencerAction call.
func (d *Sequencer) PlanNextSequencerAction() time.Duration {
head := d.engine.UnsafeL2Head()
now := d.timeNow()
buildingOnto, buildingID, _ := d.engine.BuildingPayload()
// We may have to wait till the next sequencing action, e.g. upon an error.
// If the head changed we need to respond and will not delay the sequencing.
if delay := d.nextAction.Sub(now); delay > 0 && buildingOnto.Hash == head.Hash {
return delay
}
blockTime := time.Duration(d.config.BlockTime) * time.Second
payloadTime := time.Unix(int64(head.Time+d.config.BlockTime), 0)
remainingTime := time.Until(payloadTime)
remainingTime := payloadTime.Sub(now)
// If we started building a block already, and if that work is still consistent,
// then we would like to finish it by sealing the block.
if d.buildingID != (eth.PayloadID{}) && d.buildingOnto.Hash == head.Hash {
if buildingID != (eth.PayloadID{}) && buildingOnto.Hash == head.Hash {
// if we started building already, then we will schedule the sealing.
if remainingTime < sealingDuration {
return 0, true, head.ID() // if there's not enough time for sealing, don't wait.
return 0 // if there's not enough time for sealing, don't wait.
} else {
// finish with margin of sealing duration before payloadTime
return remainingTime - sealingDuration, true, head.ID()
return remainingTime - sealingDuration
}
} else {
// if we did not yet start building, then we will schedule the start.
if remainingTime > blockTime {
// if we have too much time, then wait before starting the build
return remainingTime - blockTime, false, head.ID()
return remainingTime - blockTime
} else {
// otherwise start instantly
return 0, false, head.ID()
return 0
}
}
}
// BuildingOnto returns the L2 head reference that the latest block is or was being built on top of.
func (d *Sequencer) BuildingOnto() eth.L2BlockRef {
ref, _, _ := d.engine.BuildingPayload()
return ref
}
// RunNextSequencerAction starts new block building work, or seals existing work,
// and is best timed by first awaiting the delay returned by PlanNextSequencerAction.
// If a new block is successfully sealed, it will be returned for publishing, nil otherwise.
func (d *Sequencer) RunNextSequencerAction(ctx context.Context, l1Head eth.L1BlockRef) *eth.ExecutionPayload {
if _, buildingID, _ := d.engine.BuildingPayload(); buildingID != (eth.PayloadID{}) {
payload, err := d.CompleteBuildingBlock(ctx)
if err != nil {
d.log.Error("sequencer failed to seal new block", "err", err)
d.nextAction = d.timeNow().Add(time.Second)
if buildingID != (eth.PayloadID{}) { // don't keep stale block building jobs around, try to cancel them
d.CancelBuildingBlock(ctx)
}
return nil
} else {
d.log.Info("sequencer successfully built a new block", "block", payload.ID(), "time", uint64(payload.Timestamp), "txs", len(payload.Transactions))
return payload
}
} else {
err := d.StartBuildingBlock(ctx, l1Head)
if err != nil {
d.log.Error("sequencer failed to start building new block", "err", err)
d.nextAction = d.timeNow().Add(time.Second)
} else {
d.log.Info("sequencer started building new block", "payload_id", buildingID)
}
return nil
}
}
package driver
import (
"context"
crand "crypto/rand"
"encoding/binary"
"errors"
"fmt"
"math/big"
"math/rand"
"testing"
"time"
"github.com/ethereum/go-ethereum/common"
"github.com/ethereum/go-ethereum/core/types"
"github.com/ethereum/go-ethereum/log"
"github.com/stretchr/testify/require"
"github.com/ethereum-optimism/optimism/op-node/eth"
"github.com/ethereum-optimism/optimism/op-node/rollup"
"github.com/ethereum-optimism/optimism/op-node/rollup/derive"
"github.com/ethereum-optimism/optimism/op-node/testlog"
"github.com/ethereum-optimism/optimism/op-node/testutils"
)
type FakeEngineControl struct {
finalized eth.L2BlockRef
safe eth.L2BlockRef
unsafe eth.L2BlockRef
buildingOnto eth.L2BlockRef
buildingID eth.PayloadID
buildingSafe bool
buildingAttrs *eth.PayloadAttributes
buildingStart time.Time
cfg *rollup.Config
timeNow func() time.Time
makePayload func(onto eth.L2BlockRef, attrs *eth.PayloadAttributes) *eth.ExecutionPayload
errTyp derive.BlockInsertionErrType
err error
totalBuildingTime time.Duration
totalBuiltBlocks int
totalTxs int
}
func (m *FakeEngineControl) avgBuildingTime() time.Duration {
return m.totalBuildingTime / time.Duration(m.totalBuiltBlocks)
}
func (m *FakeEngineControl) avgTxsPerBlock() float64 {
return float64(m.totalTxs) / float64(m.totalBuiltBlocks)
}
func (m *FakeEngineControl) StartPayload(ctx context.Context, parent eth.L2BlockRef, attrs *eth.PayloadAttributes, updateSafe bool) (errType derive.BlockInsertionErrType, err error) {
if m.err != nil {
return m.errTyp, m.err
}
m.buildingID = eth.PayloadID{}
_, _ = crand.Read(m.buildingID[:])
m.buildingOnto = parent
m.buildingSafe = updateSafe
m.buildingAttrs = attrs
m.buildingStart = m.timeNow()
return derive.BlockInsertOK, nil
}
func (m *FakeEngineControl) ConfirmPayload(ctx context.Context) (out *eth.ExecutionPayload, errTyp derive.BlockInsertionErrType, err error) {
if m.err != nil {
return nil, m.errTyp, m.err
}
buildTime := m.timeNow().Sub(m.buildingStart)
m.totalBuildingTime += buildTime
m.totalBuiltBlocks += 1
payload := m.makePayload(m.buildingOnto, m.buildingAttrs)
ref, err := derive.PayloadToBlockRef(payload, &m.cfg.Genesis)
if err != nil {
panic(err)
}
m.unsafe = ref
if m.buildingSafe {
m.safe = ref
}
m.resetBuildingState()
m.totalTxs += len(payload.Transactions)
return payload, derive.BlockInsertOK, nil
}
func (m *FakeEngineControl) CancelPayload(ctx context.Context, force bool) error {
if force {
m.resetBuildingState()
}
return m.err
}
func (m *FakeEngineControl) BuildingPayload() (onto eth.L2BlockRef, id eth.PayloadID, safe bool) {
return m.buildingOnto, m.buildingID, m.buildingSafe
}
func (m *FakeEngineControl) Finalized() eth.L2BlockRef {
return m.finalized
}
func (m *FakeEngineControl) UnsafeL2Head() eth.L2BlockRef {
return m.unsafe
}
func (m *FakeEngineControl) SafeL2Head() eth.L2BlockRef {
return m.safe
}
func (m *FakeEngineControl) resetBuildingState() {
m.buildingID = eth.PayloadID{}
m.buildingOnto = eth.L2BlockRef{}
m.buildingSafe = false
m.buildingAttrs = nil
}
var _ derive.EngineControl = (*FakeEngineControl)(nil)
type testAttrBuilderFn func(ctx context.Context, l2Parent eth.L2BlockRef, epoch eth.BlockID) (attrs *eth.PayloadAttributes, err error)
func (fn testAttrBuilderFn) PreparePayloadAttributes(ctx context.Context, l2Parent eth.L2BlockRef, epoch eth.BlockID) (attrs *eth.PayloadAttributes, err error) {
return fn(ctx, l2Parent, epoch)
}
var _ derive.AttributesBuilder = (testAttrBuilderFn)(nil)
type testOriginSelectorFn func(ctx context.Context, l1Head eth.L1BlockRef, l2Head eth.L2BlockRef) (eth.L1BlockRef, error)
func (fn testOriginSelectorFn) FindL1Origin(ctx context.Context, l1Head eth.L1BlockRef, l2Head eth.L2BlockRef) (eth.L1BlockRef, error) {
return fn(ctx, l1Head, l2Head)
}
var _ L1OriginSelectorIface = (testOriginSelectorFn)(nil)
// TestSequencerChaosMonkey runs the sequencer in a mocked adversarial environment with
// repeated random errors in dependencies and poor clock timing.
// At the end the health of the chain is checked to show that the sequencer kept the chain in shape.
func TestSequencerChaosMonkey(t *testing.T) {
mockL1Hash := func(num uint64) (out common.Hash) {
out[31] = 1
binary.BigEndian.PutUint64(out[:], num)
return
}
mockL2Hash := func(num uint64) (out common.Hash) {
out[31] = 2
binary.BigEndian.PutUint64(out[:], num)
return
}
mockL1ID := func(num uint64) eth.BlockID {
return eth.BlockID{Hash: mockL1Hash(num), Number: num}
}
mockL2ID := func(num uint64) eth.BlockID {
return eth.BlockID{Hash: mockL2Hash(num), Number: num}
}
rng := rand.New(rand.NewSource(12345))
l1Time := uint64(100000)
// mute errors. We expect a lot of the mocked errors to cause error-logs. We check chain health at the end of the test.
log := testlog.Logger(t, log.LvlCrit)
cfg := &rollup.Config{
Genesis: rollup.Genesis{
L1: mockL1ID(100000),
L2: mockL2ID(200000),
L2Time: l1Time + 300, // L2 may start with a relative old L1 origin and will have to catch it up
SystemConfig: eth.SystemConfig{},
},
BlockTime: 2,
MaxSequencerDrift: 30,
}
// keep track of the L1 timestamps we mock because sometimes we only have the L1 hash/num handy
l1Times := map[eth.BlockID]uint64{cfg.Genesis.L1: l1Time}
genesisL2 := eth.L2BlockRef{
Hash: cfg.Genesis.L2.Hash,
Number: cfg.Genesis.L2.Number,
ParentHash: mockL2Hash(cfg.Genesis.L2.Number - 1),
Time: cfg.Genesis.L2Time,
L1Origin: cfg.Genesis.L1,
SequenceNumber: 0,
}
// initialize our engine state
engControl := &FakeEngineControl{
finalized: genesisL2,
safe: genesisL2,
unsafe: genesisL2,
cfg: cfg,
}
// start wallclock at 5 minutes after the current L2 head. The sequencer has some catching up to do!
clockTime := time.Unix(int64(engControl.unsafe.Time)+5*60, 0)
clockFn := func() time.Time {
return clockTime
}
engControl.timeNow = clockFn
// mock payload building, we don't need to process any real txs.
engControl.makePayload = func(onto eth.L2BlockRef, attrs *eth.PayloadAttributes) *eth.ExecutionPayload {
txs := make([]eth.Data, 0)
txs = append(txs, attrs.Transactions...) // include deposits
if !attrs.NoTxPool { // if we are allowed to sequence from tx pool, mock some txs
n := rng.Intn(20)
for i := 0; i < n; i++ {
txs = append(txs, []byte(fmt.Sprintf("mock sequenced tx %d", i)))
}
}
return &eth.ExecutionPayload{
ParentHash: onto.Hash,
BlockNumber: eth.Uint64Quantity(onto.Number) + 1,
Timestamp: attrs.Timestamp,
BlockHash: mockL2Hash(onto.Number),
Transactions: txs,
}
}
// We keep attribute building simple, we don't talk to a real execution engine in this test.
// Sometimes we fake an error in the attributes preparation.
var attrsErr error
attrBuilder := testAttrBuilderFn(func(ctx context.Context, l2Parent eth.L2BlockRef, epoch eth.BlockID) (attrs *eth.PayloadAttributes, err error) {
if attrsErr != nil {
return nil, attrsErr
}
seqNr := l2Parent.SequenceNumber + 1
if epoch != l2Parent.L1Origin {
seqNr = 0
}
l1Info := &testutils.MockBlockInfo{
InfoHash: epoch.Hash,
InfoParentHash: mockL1Hash(epoch.Number - 1),
InfoCoinbase: common.Address{},
InfoRoot: common.Hash{},
InfoNum: epoch.Number,
InfoTime: l1Times[epoch],
InfoMixDigest: [32]byte{},
InfoBaseFee: big.NewInt(1234),
InfoReceiptRoot: common.Hash{},
}
infoDep, err := derive.L1InfoDepositBytes(seqNr, l1Info, cfg.Genesis.SystemConfig)
require.NoError(t, err)
testGasLimit := eth.Uint64Quantity(10_000_000)
return &eth.PayloadAttributes{
Timestamp: eth.Uint64Quantity(l2Parent.Time + cfg.BlockTime),
PrevRandao: eth.Bytes32{},
SuggestedFeeRecipient: common.Address{},
Transactions: []eth.Data{infoDep},
NoTxPool: false,
GasLimit: &testGasLimit,
}, nil
})
maxL1BlockTimeGap := uint64(100)
// The origin selector just generates random L1 blocks based on RNG
var originErr error
originSelector := testOriginSelectorFn(func(ctx context.Context, l1Head eth.L1BlockRef, l2Head eth.L2BlockRef) (eth.L1BlockRef, error) {
if originErr != nil {
return eth.L1BlockRef{}, originErr
}
origin := eth.L1BlockRef{
Hash: mockL1Hash(l2Head.L1Origin.Number),
Number: l2Head.L1Origin.Number,
ParentHash: mockL1Hash(l2Head.L1Origin.Number),
Time: l1Times[l2Head.L1Origin],
}
// randomly make a L1 origin appear, if we can even select it
nextL2Time := l2Head.Time + cfg.BlockTime
if nextL2Time <= origin.Time {
return origin, nil
}
maxTimeIncrement := nextL2Time - origin.Time
if maxTimeIncrement > maxL1BlockTimeGap {
maxTimeIncrement = maxL1BlockTimeGap
}
if rng.Intn(10) == 0 {
nextOrigin := eth.L1BlockRef{
Hash: mockL1Hash(origin.Number + 1),
Number: origin.Number + 1,
ParentHash: origin.Hash,
Time: origin.Time + 1 + uint64(rng.Int63n(int64(maxTimeIncrement))),
}
l1Times[nextOrigin.ID()] = nextOrigin.Time
return nextOrigin, nil
} else {
return origin, nil
}
})
seq := NewSequencer(log, cfg, engControl, attrBuilder, originSelector)
seq.timeNow = clockFn
l1Head := eth.L1BlockRef{} // TODO this is getting removed
// try to build 1000 blocks, with 5x as many planning attempts, to handle errors and clock problems
desiredBlocks := 1000
for i := 0; i < 5*desiredBlocks && engControl.totalBuiltBlocks < desiredBlocks; i++ {
delta := seq.PlanNextSequencerAction()
x := rng.Float32()
if x < 0.01 { // 1%: mess a lot with the clock: simulate a hang of up to 30 seconds
if i < desiredBlocks/2 { // only in first 50% of blocks to let it heal, hangs take time
delta = time.Duration(rng.Float64() * float64(time.Second*30))
}
} else if x < 0.1 { // 9%: mess with the timing, -50% to 50% off
delta = time.Duration((0.5 + rng.Float64()) * float64(delta))
} else if x < 0.5 {
// 40%: mess slightly with the timing, -10% to 10% off
delta = time.Duration((0.9 + rng.Float64()*0.2) * float64(delta))
}
clockTime = clockTime.Add(delta)
// reset errors
originErr = nil
attrsErr = nil
engControl.err = nil
engControl.errTyp = derive.BlockInsertOK
// maybe make something maybe fail, or try a new L1 origin
switch rng.Intn(10) { // 40% chance to fail sequencer action (!!!)
case 0:
originErr = errors.New("mock origin error")
case 1:
attrsErr = errors.New("mock attributes error")
case 2:
engControl.err = errors.New("mock temporary engine error")
engControl.errTyp = derive.BlockInsertTemporaryErr
case 3:
engControl.err = errors.New("mock prestate engine error")
engControl.errTyp = derive.BlockInsertPrestateErr
default:
// no error
}
payload := seq.RunNextSequencerAction(context.Background(), l1Head)
if payload != nil {
require.Equal(t, engControl.UnsafeL2Head().ID(), payload.ID(), "head must stay in sync with emitted payloads")
var tx types.Transaction
require.NoError(t, tx.UnmarshalBinary(payload.Transactions[0]))
info, err := derive.L1InfoDepositTxData(tx.Data())
require.NoError(t, err)
require.GreaterOrEqual(t, uint64(payload.Timestamp), info.Time, "ensure L2 time >= L1 time")
}
}
// Now, even though:
// - the start state was behind the wallclock
// - the L1 origin was far behind the L2
// - we made all components fail at random
// - messed with the clock
// the L2 chain was still built and stats are healthy on average!
l2Head := engControl.UnsafeL2Head()
t.Logf("avg build time: %s, clock timestamp: %d, L2 head time: %d, L1 origin time: %d, avg txs per block: %f", engControl.avgBuildingTime(), clockFn().Unix(), l2Head.Time, l1Times[l2Head.L1Origin], engControl.avgTxsPerBlock())
require.Equal(t, engControl.totalBuiltBlocks, desiredBlocks, "persist through random errors and build the desired blocks")
require.Equal(t, l2Head.Time, cfg.Genesis.L2Time+uint64(desiredBlocks)*cfg.BlockTime, "reached desired L2 block timestamp")
require.GreaterOrEqual(t, l2Head.Time, l1Times[l2Head.L1Origin], "the L2 time >= the L1 time")
require.Less(t, l2Head.Time-l1Times[l2Head.L1Origin], uint64(100), "The L1 origin time is close to the L2 time")
require.Less(t, clockTime.Sub(time.Unix(int64(l2Head.Time), 0)).Abs(), 2*time.Second, "L2 time is accurate, within 2 seconds of wallclock")
require.Greater(t, engControl.avgBuildingTime(), time.Second, "With 2 second block time and 1 second error backoff and healthy-on-average errors, building time should at least be a second")
require.Greater(t, engControl.avgTxsPerBlock(), 3.0, "We expect at least 1 system tx per block, but with a mocked 0-10 txs we expect an higher avg")
}
......@@ -25,8 +25,6 @@ type SyncStatus = eth.SyncStatus
// sealingDuration defines the expected time it takes to seal the block
const sealingDuration = time.Millisecond * 50
var UninitializedL1StateErr = errors.New("the L1 Head in L1 State is not initialized yet")
type Driver struct {
l1State L1StateIface
......@@ -71,11 +69,10 @@ type Driver struct {
// L2 Signals:
unsafeL2Payloads chan *eth.ExecutionPayload
l1 L1Chain
l2 L2Chain
l1OriginSelector L1OriginSelectorIface
sequencer SequencerIface
network Network // may be nil, network for is optional
l1 L1Chain
l2 L2Chain
sequencer SequencerIface
network Network // may be nil, network for is optional
metrics Metrics
log log.Logger
......@@ -142,75 +139,6 @@ func (s *Driver) OnUnsafeL2Payload(ctx context.Context, payload *eth.ExecutionPa
}
}
// startNewL2Block starts sequencing a new L2 block on top of the unsafe L2 Head.
func (s *Driver) startNewL2Block(ctx context.Context) error {
l2Head := s.derivation.UnsafeL2Head()
l1Head := s.l1State.L1Head()
if l1Head == (eth.L1BlockRef{}) {
return UninitializedL1StateErr
}
// Figure out which L1 origin block we're going to be building on top of.
l1Origin, err := s.l1OriginSelector.FindL1Origin(ctx, l1Head, l2Head)
if err != nil {
s.log.Error("Error finding next L1 Origin", "err", err)
return err
}
// Rollup is configured to not start producing blocks until a specific L1 block has been
// reached. Don't produce any blocks until we're at that genesis block.
if l1Origin.Number < s.config.Genesis.L1.Number {
s.log.Info("Skipping block production because the next L1 Origin is behind the L1 genesis", "next", l1Origin.ID(), "genesis", s.config.Genesis.L1)
return fmt.Errorf("the L1 origin %s cannot be before genesis at %s", l1Origin, s.config.Genesis.L1)
}
// Should never happen. Sequencer will halt if we get into this situation somehow.
nextL2Time := l2Head.Time + s.config.BlockTime
if nextL2Time < l1Origin.Time {
s.log.Error("Cannot build L2 block for time before L1 origin",
"l2Unsafe", l2Head, "nextL2Time", nextL2Time, "l1Origin", l1Origin, "l1OriginTime", l1Origin.Time)
return fmt.Errorf("cannot build L2 block on top %s for time %d before L1 origin %s at time %d",
l2Head, nextL2Time, l1Origin, l1Origin.Time)
}
// Start creating the new block.
return s.sequencer.StartBuildingBlock(ctx, l1Origin)
}
// completeNewBlock completes a previously started L2 block sequencing job.
func (s *Driver) completeNewBlock(ctx context.Context) error {
payload, err := s.sequencer.CompleteBuildingBlock(ctx)
if err != nil {
s.metrics.RecordSequencingError()
s.log.Error("Failed to seal block as sequencer", "err", err)
return err
}
// Generate an L2 block ref from the payload.
newUnsafeL2Head, err := derive.PayloadToBlockRef(payload, &s.config.Genesis)
if err != nil {
s.metrics.RecordSequencingError()
s.log.Error("Sequenced payload cannot be transformed into valid L2 block reference", "err", err)
return fmt.Errorf("sequenced payload cannot be transformed into valid L2 block reference: %w", err)
}
// Update our L2 head block based on the new unsafe block we just generated.
s.derivation.SetUnsafeHead(newUnsafeL2Head)
s.log.Info("Sequenced new l2 block", "l2_unsafe", newUnsafeL2Head, "l1_origin", newUnsafeL2Head.L1Origin, "txs", len(payload.Transactions), "time", newUnsafeL2Head.Time)
s.metrics.CountSequencedTxs(len(payload.Transactions))
if s.network != nil {
if err := s.network.PublishL2Payload(ctx, payload); err != nil {
s.log.Warn("failed to publish newly created block", "id", payload.ID(), "err", err)
s.metrics.RecordPublishingError()
// publishing of unsafe data via p2p is optional. Errors are not severe enough to change/halt sequencing but should be logged and metered.
}
}
return nil
}
// the eventLoop responds to L1 changes and internal timers to produce L2 blocks.
func (s *Driver) eventLoop() {
defer s.wg.Done()
......@@ -259,34 +187,23 @@ func (s *Driver) eventLoop() {
// L1 chain that we need to handle.
reqStep()
blockTime := time.Duration(s.config.BlockTime) * time.Second
var sequenceErr error
var sequenceErrTime time.Time
sequencerTimer := time.NewTimer(0)
var sequencerCh <-chan time.Time
var sequencingPlannedOnto eth.BlockID
var sequencerSealNext bool
planSequencerAction := func() {
delay, seal, onto := s.sequencer.PlanNextSequencerAction(sequenceErr)
if sequenceErr != nil && time.Since(sequenceErrTime) > delay {
sequenceErr = nil
}
delay := s.sequencer.PlanNextSequencerAction()
sequencerCh = sequencerTimer.C
if len(sequencerCh) > 0 { // empty if not already drained before resetting
<-sequencerCh
}
sequencerTimer.Reset(delay)
sequencingPlannedOnto = onto
sequencerSealNext = seal
}
for {
// If we are sequencing, update the trigger for the next sequencer action.
// If we are sequencing, and the L1 state is ready, update the trigger for the next sequencer action.
// This may adjust at any time based on fork-choice changes or previous errors.
if s.driverConfig.SequencerEnabled && !s.driverConfig.SequencerStopped {
if s.driverConfig.SequencerEnabled && !s.driverConfig.SequencerStopped && s.l1State.L1Head() != (eth.L1BlockRef{}) {
// update sequencer time if the head changed
if sequencingPlannedOnto != s.derivation.UnsafeL2Head().ID() {
if s.sequencer.BuildingOnto().ID() != s.derivation.UnsafeL2Head().ID() {
planSequencerAction()
}
} else {
......@@ -295,22 +212,14 @@ func (s *Driver) eventLoop() {
select {
case <-sequencerCh:
s.log.Info("sequencing now!", "seal", sequencerSealNext, "idle_derivation", s.idleDerivation)
if sequencerSealNext {
// try to seal the current block task, and allow it to take up to 3 block times.
// If this fails we will simply start a new block building job.
ctx, cancel := context.WithTimeout(ctx, 3*blockTime)
sequenceErr = s.completeNewBlock(ctx)
cancel()
} else {
// Start the block building, don't allow the starting of sequencing to get stuck for more the time of 1 block.
ctx, cancel := context.WithTimeout(ctx, blockTime)
sequenceErr = s.startNewL2Block(ctx)
cancel()
}
if sequenceErr != nil {
s.log.Error("sequencing error", "err", sequenceErr)
sequenceErrTime = time.Now()
payload := s.sequencer.RunNextSequencerAction(ctx, s.l1State.L1Head())
if s.network != nil && payload != nil {
// Publishing of unsafe data via p2p is optional.
// Errors are not severe enough to change/halt sequencing but should be logged and metered.
if err := s.network.PublishL2Payload(ctx, payload); err != nil {
s.log.Warn("failed to publish newly created block", "id", payload.ID(), "err", err)
s.metrics.RecordPublishingError()
}
}
planSequencerAction() // schedule the next sequencer action to keep the sequencing looping
case payload := <-s.unsafeL2Payloads:
......@@ -386,8 +295,8 @@ func (s *Driver) eventLoop() {
} else {
s.log.Info("Sequencer has been started")
s.driverConfig.SequencerStopped = false
sequencingPlannedOnto = eth.BlockID{}
close(resp.err)
planSequencerAction() // resume sequencing
}
case respCh := <-s.stopSequencer:
if s.driverConfig.SequencerStopped {
......
// On develop
package driver
import (
"context"
"errors"
"math/big"
"math/rand"
"testing"
"time"
"github.com/ethereum/go-ethereum/common"
"github.com/ethereum/go-ethereum/log"
"github.com/stretchr/testify/require"
"github.com/ethereum-optimism/optimism/op-node/eth"
"github.com/ethereum-optimism/optimism/op-node/metrics"
"github.com/ethereum-optimism/optimism/op-node/rollup"
"github.com/ethereum-optimism/optimism/op-node/rollup/derive"
"github.com/ethereum-optimism/optimism/op-node/testutils"
)
type TestDummyOutputImpl struct {
willError bool
cfg *rollup.Config
l1Origin eth.L1BlockRef
l2Head eth.L2BlockRef
}
func (d *TestDummyOutputImpl) PlanNextSequencerAction(sequenceErr error) (delay time.Duration, seal bool, onto eth.BlockID) {
return 0, d.l1Origin != (eth.L1BlockRef{}), d.l2Head.ParentID()
}
func (d *TestDummyOutputImpl) StartBuildingBlock(ctx context.Context, l1Origin eth.L1BlockRef) error {
d.l1Origin = l1Origin
return nil
}
func (d *TestDummyOutputImpl) CompleteBuildingBlock(ctx context.Context) (*eth.ExecutionPayload, error) {
// If we're meant to error, return one
if d.willError {
return nil, errors.New("the TestDummyOutputImpl.createNewBlock operation failed")
}
info := &testutils.MockBlockInfo{
InfoHash: d.l1Origin.Hash,
InfoParentHash: d.l1Origin.ParentHash,
InfoCoinbase: common.Address{},
InfoRoot: common.Hash{},
InfoNum: d.l1Origin.Number,
InfoTime: d.l1Origin.Time,
InfoMixDigest: [32]byte{},
InfoBaseFee: big.NewInt(123),
InfoReceiptRoot: common.Hash{},
}
infoTx, err := derive.L1InfoDepositBytes(d.l2Head.SequenceNumber, info, eth.SystemConfig{})
if err != nil {
panic(err)
}
payload := eth.ExecutionPayload{
ParentHash: d.l2Head.Hash,
FeeRecipient: common.Address{},
StateRoot: eth.Bytes32{},
ReceiptsRoot: eth.Bytes32{},
LogsBloom: eth.Bytes256{},
PrevRandao: eth.Bytes32{},
BlockNumber: eth.Uint64Quantity(d.l2Head.Number + 1),
GasLimit: 0,
GasUsed: 0,
Timestamp: eth.Uint64Quantity(d.l2Head.Time + d.cfg.BlockTime),
ExtraData: nil,
BaseFeePerGas: eth.Uint256Quantity{},
BlockHash: common.Hash{123},
Transactions: []eth.Data{infoTx},
}
return &payload, nil
}
var _ SequencerIface = (*TestDummyOutputImpl)(nil)
type TestDummyDerivationPipeline struct {
DerivationPipeline
l2Head eth.L2BlockRef
l2SafeHead eth.L2BlockRef
l2Finalized eth.L2BlockRef
}
func (d TestDummyDerivationPipeline) Reset() {}
func (d TestDummyDerivationPipeline) Step(ctx context.Context) error { return nil }
func (d TestDummyDerivationPipeline) SetUnsafeHead(head eth.L2BlockRef) {}
func (d TestDummyDerivationPipeline) AddUnsafePayload(payload *eth.ExecutionPayload) {}
func (d TestDummyDerivationPipeline) Finalized() eth.L2BlockRef { return d.l2Head }
func (d TestDummyDerivationPipeline) SafeL2Head() eth.L2BlockRef { return d.l2SafeHead }
func (d TestDummyDerivationPipeline) UnsafeL2Head() eth.L2BlockRef { return d.l2Finalized }
type TestDummyL1OriginSelector struct {
retval eth.L1BlockRef
}
func (l TestDummyL1OriginSelector) FindL1Origin(ctx context.Context, l1Head eth.L1BlockRef, l2Head eth.L2BlockRef) (eth.L1BlockRef, error) {
return l.retval, nil
}
// TestRejectCreateBlockBadTimestamp tests that a block creation with invalid timestamps will be caught.
// This does not test:
// - The findL1Origin call (it is hardcoded to be the head)
// - The outputInterface used to create a new block from a given payload.
// - The DerivationPipeline setting unsafe head (a mock provider is used to pretend to set it)
// - Metrics (only mocked enough to let the method proceed)
// - Publishing (network is set to nil so publishing won't occur)
func TestRejectCreateBlockBadTimestamp(t *testing.T) {
// Create our random provider
rng := rand.New(rand.NewSource(rand.Int63()))
// Create our context for methods to execute under
ctx := context.Background()
// Create our fake L1/L2 heads and link them accordingly
l1HeadRef := testutils.RandomBlockRef(rng)
l2HeadRef := testutils.RandomL2BlockRef(rng)
l2l1OriginBlock := l1HeadRef
l2HeadRef.L1Origin = l2l1OriginBlock.ID()
// Create a rollup config
cfg := rollup.Config{
BlockTime: uint64(60),
Genesis: rollup.Genesis{
L1: l1HeadRef.ID(),
L2: l2HeadRef.ID(),
L2Time: 0x7000, // dummy value
},
}
// Patch our timestamp so we fail
l2HeadRef.Time = l2l1OriginBlock.Time - (cfg.BlockTime * 2)
// Create our outputter
outputProvider := &TestDummyOutputImpl{cfg: &cfg, l2Head: l2HeadRef, willError: false}
// Create our state
s := Driver{
l1State: &L1State{
l1Head: l1HeadRef,
log: log.New(),
metrics: metrics.NoopMetrics,
},
log: log.New(),
l1OriginSelector: TestDummyL1OriginSelector{retval: l1HeadRef},
config: &cfg,
sequencer: outputProvider,
derivation: TestDummyDerivationPipeline{},
metrics: metrics.NoopMetrics,
}
// Create a new block
// - L2Head's L1Origin, its timestamp should be greater than L1 genesis.
// - L2Head timestamp + BlockTime should be greater than or equal to the L1 Time.
err := s.startNewL2Block(ctx)
if err == nil {
err = s.completeNewBlock(ctx)
}
// Verify the L1Origin's block number is greater than L1 genesis in our config.
if l2l1OriginBlock.Number < s.config.Genesis.L1.Number {
require.NoError(t, err, "L1Origin block number should be greater than the L1 genesis block number")
}
// Verify the new L2 block to create will have a time stamp equal or newer than our L1 origin block we derive from.
if l2HeadRef.Time+cfg.BlockTime < l2l1OriginBlock.Time {
// If not, we expect a specific error.
// TODO: This isn't the cleanest, we should construct + compare the whole error message.
require.NotNil(t, err)
require.Contains(t, err.Error(), "cannot build L2 block on top")
require.Contains(t, err.Error(), "for time")
require.Contains(t, err.Error(), "before L1 origin")
return
}
// If we expected the outputter to error, capture that here
if outputProvider.willError {
require.NotNil(t, err, "outputInterface failed to createNewBlock, so createNewL2Block should also have failed")
return
}
// Otherwise we should have no error.
require.NoError(t, err, "error raised in TestRejectCreateBlockBadTimestamp")
}
// FuzzRejectCreateBlockBadTimestamp is a property test derived from the TestRejectCreateBlockBadTimestamp unit test.
// It fuzzes timestamps and block times to find a configuration to violate error checking.
func FuzzRejectCreateBlockBadTimestamp(f *testing.F) {
f.Fuzz(func(t *testing.T, randSeed int64, l2Time uint64, blockTime uint64, forceOutputFail bool, currentL2HeadTime uint64) {
// Create our random provider
rng := rand.New(rand.NewSource(randSeed))
// Create our context for methods to execute under
ctx := context.Background()
// Create our fake L1/L2 heads and link them accordingly
l1HeadRef := testutils.RandomBlockRef(rng)
l2HeadRef := testutils.RandomL2BlockRef(rng)
l2l1OriginBlock := l1HeadRef
l2HeadRef.L1Origin = l2l1OriginBlock.ID()
// TODO: Cap our block time so it doesn't overflow
if blockTime > 0x100000 {
blockTime = 0x100000
}
// Create a rollup config
cfg := rollup.Config{
BlockTime: blockTime,
Genesis: rollup.Genesis{
L1: l1HeadRef.ID(),
L2: l2HeadRef.ID(),
L2Time: l2Time, // dummy value
},
}
// Patch our timestamp so we fail
l2HeadRef.Time = currentL2HeadTime
// Create our outputter
outputProvider := &TestDummyOutputImpl{cfg: &cfg, l2Head: l2HeadRef, willError: forceOutputFail}
// Create our state
s := Driver{
l1State: &L1State{
l1Head: l1HeadRef,
log: log.New(),
metrics: metrics.NoopMetrics,
},
log: log.New(),
l1OriginSelector: TestDummyL1OriginSelector{retval: l1HeadRef},
config: &cfg,
sequencer: outputProvider,
derivation: TestDummyDerivationPipeline{},
metrics: metrics.NoopMetrics,
}
// Create a new block
// - L2Head's L1Origin, its timestamp should be greater than L1 genesis.
// - L2Head timestamp + BlockTime should be greater than or equal to the L1 Time.
err := s.startNewL2Block(ctx)
if err == nil {
err = s.completeNewBlock(ctx)
}
// Verify the L1Origin's timestamp is greater than L1 genesis in our config.
if l2l1OriginBlock.Number < s.config.Genesis.L1.Number {
require.NoError(t, err)
return
}
// Verify the new L2 block to create will have a time stamp equal or newer than our L1 origin block we derive from.
if l2HeadRef.Time+cfg.BlockTime < l2l1OriginBlock.Time {
// If not, we expect a specific error.
// TODO: This isn't the cleanest, we should construct + compare the whole error message.
require.NotNil(t, err)
require.Contains(t, err.Error(), "cannot build L2 block on top")
require.Contains(t, err.Error(), "for time")
require.Contains(t, err.Error(), "before L1 origin")
return
}
// Otherwise we should have no error.
require.Nil(t, err)
// If we expected the outputter to error, capture that here
if outputProvider.willError {
require.NotNil(t, err, "outputInterface failed to createNewBlock, so createNewL2Block should also have failed")
return
}
// Otherwise we should have no error.
require.NoError(t, err, "L1Origin block number should be greater than the L1 genesis block number")
})
}
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment