802 lines
30 KiB
Nim
802 lines
30 KiB
Nim
# beacon_chain
|
|
# Copyright (c) 2018-2023 Status Research & Development GmbH
|
|
# Licensed and distributed under either of
|
|
# * MIT license (license terms in the root directory or at https://opensource.org/licenses/MIT).
|
|
# * Apache v2 license (license terms in the root directory or at https://www.apache.org/licenses/LICENSE-2.0).
|
|
# at your option. This file may not be copied, modified, or distributed except according to those terms.
|
|
|
|
{.push raises: [].}
|
|
|
|
import
|
|
std/[options, tables, sets, macros],
|
|
chronicles, chronos, snappy, snappy/codec,
|
|
libp2p/switch,
|
|
../spec/datatypes/[phase0, altair, bellatrix, capella, eip4844],
|
|
../spec/[helpers, forks, network],
|
|
".."/[beacon_clock],
|
|
../networking/eth2_network,
|
|
../consensus_object_pools/blockchain_dag,
|
|
../rpc/rest_constants
|
|
|
|
logScope:
|
|
topics = "sync"
|
|
|
|
const
|
|
MAX_REQUEST_BLOCKS* = 1024
|
|
MAX_REQUEST_BLOBS_SIDECARS* = 128
|
|
|
|
blockResponseCost = allowedOpsPerSecondCost(64) # Allow syncing ~64 blocks/sec (minus request costs)
|
|
|
|
# https://github.com/ethereum/consensus-specs/blob/v1.3.0-alpha.0/specs/altair/light-client/p2p-interface.md#configuration
|
|
MAX_REQUEST_LIGHT_CLIENT_UPDATES* = 128
|
|
lightClientBootstrapResponseCost = allowedOpsPerSecondCost(1)
|
|
## Only one bootstrap per peer should ever be needed - no need to allow more
|
|
lightClientUpdateResponseCost = allowedOpsPerSecondCost(1000)
|
|
## Updates are tiny - we can allow lots of them
|
|
lightClientFinalityUpdateResponseCost = allowedOpsPerSecondCost(100)
|
|
lightClientOptimisticUpdateResponseCost = allowedOpsPerSecondCost(100)
|
|
|
|
type
|
|
StatusMsg* = object
|
|
forkDigest*: ForkDigest
|
|
finalizedRoot*: Eth2Digest
|
|
finalizedEpoch*: Epoch
|
|
headRoot*: Eth2Digest
|
|
headSlot*: Slot
|
|
|
|
ValidatorSetDeltaFlags {.pure.} = enum
|
|
Activation = 0
|
|
Exit = 1
|
|
|
|
ValidatorChangeLogEntry* = object
|
|
case kind*: ValidatorSetDeltaFlags
|
|
of Activation:
|
|
pubkey: ValidatorPubKey
|
|
else:
|
|
index: uint32
|
|
|
|
BeaconSyncNetworkState = ref object
|
|
dag: ChainDAGRef
|
|
cfg: RuntimeConfig
|
|
forkDigests: ref ForkDigests
|
|
genesisBlockRoot: Eth2Digest
|
|
getBeaconTime: GetBeaconTimeFn
|
|
|
|
BeaconSyncPeerState* = ref object
|
|
statusLastTime*: chronos.Moment
|
|
statusMsg*: StatusMsg
|
|
|
|
BlockRootSlot* = object
|
|
blockRoot: Eth2Digest
|
|
slot: Slot
|
|
|
|
BlockRootsList* = List[Eth2Digest, Limit MAX_REQUEST_BLOCKS]
|
|
|
|
template readChunkPayload*(
|
|
conn: Connection, peer: Peer, MsgType: type ForkySignedBeaconBlock):
|
|
Future[NetRes[MsgType]] =
|
|
readChunkPayload(conn, peer, MsgType)
|
|
|
|
proc readChunkPayload*(
|
|
conn: Connection, peer: Peer, MsgType: type (ref ForkedSignedBeaconBlock)):
|
|
Future[NetRes[MsgType]] {.async.} =
|
|
var contextBytes: ForkDigest
|
|
try:
|
|
await conn.readExactly(addr contextBytes, sizeof contextBytes)
|
|
except CatchableError:
|
|
return neterr UnexpectedEOF
|
|
|
|
if contextBytes == peer.network.forkDigests.phase0:
|
|
let res = await readChunkPayload(conn, peer, phase0.SignedBeaconBlock)
|
|
if res.isOk:
|
|
return ok newClone(ForkedSignedBeaconBlock.init(res.get))
|
|
else:
|
|
return err(res.error)
|
|
elif contextBytes == peer.network.forkDigests.altair:
|
|
let res = await readChunkPayload(conn, peer, altair.SignedBeaconBlock)
|
|
if res.isOk:
|
|
return ok newClone(ForkedSignedBeaconBlock.init(res.get))
|
|
else:
|
|
return err(res.error)
|
|
elif contextBytes == peer.network.forkDigests.bellatrix:
|
|
let res = await readChunkPayload(conn, peer, bellatrix.SignedBeaconBlock)
|
|
if res.isOk:
|
|
return ok newClone(ForkedSignedBeaconBlock.init(res.get))
|
|
else:
|
|
return err(res.error)
|
|
elif contextBytes == peer.network.forkDigests.capella:
|
|
let res = await readChunkPayload(conn, peer, capella.SignedBeaconBlock)
|
|
if res.isOk:
|
|
return ok newClone(ForkedSignedBeaconBlock.init(res.get))
|
|
else:
|
|
return err(res.error)
|
|
elif contextBytes == peer.network.forkDigests.eip4844:
|
|
let res = await readChunkPayload(conn, peer, eip4844.SignedBeaconBlock)
|
|
if res.isOk:
|
|
return ok newClone(ForkedSignedBeaconBlock.init(res.get))
|
|
else:
|
|
return err(res.error)
|
|
else:
|
|
return neterr InvalidContextBytes
|
|
|
|
proc readChunkPayload*(
|
|
conn: Connection, peer: Peer, MsgType: type (ref SignedBeaconBlockAndBlobsSidecar)):
|
|
Future[NetRes[MsgType]] {.async.} =
|
|
var contextBytes: ForkDigest
|
|
try:
|
|
await conn.readExactly(addr contextBytes, sizeof contextBytes)
|
|
except CatchableError:
|
|
return neterr UnexpectedEOF
|
|
|
|
if contextBytes == peer.network.forkDigests.eip4844:
|
|
let res = await readChunkPayload(conn, peer, SignedBeaconBlockAndBlobsSidecar)
|
|
if res.isOk:
|
|
return ok newClone(res.get)
|
|
else:
|
|
return err(res.error)
|
|
else:
|
|
return neterr InvalidContextBytes
|
|
|
|
proc readChunkPayload*(
|
|
conn: Connection, peer: Peer, MsgType: type (ref BlobsSidecar)):
|
|
Future[NetRes[MsgType]] {.async.} =
|
|
var contextBytes: ForkDigest
|
|
try:
|
|
await conn.readExactly(addr contextBytes, sizeof contextBytes)
|
|
except CatchableError:
|
|
return neterr UnexpectedEOF
|
|
|
|
if contextBytes == peer.network.forkDigests.eip4844:
|
|
let res = await readChunkPayload(conn, peer, BlobsSidecar)
|
|
if res.isOk:
|
|
return ok newClone(res.get)
|
|
else:
|
|
return err(res.error)
|
|
else:
|
|
return neterr InvalidContextBytes
|
|
|
|
proc readChunkPayload*(
|
|
conn: Connection, peer: Peer, MsgType: type SomeForkedLightClientObject):
|
|
Future[NetRes[MsgType]] {.async.} =
|
|
var contextBytes: ForkDigest
|
|
try:
|
|
await conn.readExactly(addr contextBytes, sizeof contextBytes)
|
|
except CatchableError:
|
|
return neterr UnexpectedEOF
|
|
let contextFork =
|
|
peer.network.forkDigests[].stateForkForDigest(contextBytes).valueOr:
|
|
return neterr InvalidContextBytes
|
|
|
|
withLcDataFork(lcDataForkAtStateFork(contextFork)):
|
|
when lcDataFork > LightClientDataFork.None:
|
|
let res = await eth2_network.readChunkPayload(
|
|
conn, peer, MsgType.Forky(lcDataFork))
|
|
if res.isOk:
|
|
if contextFork !=
|
|
peer.network.cfg.stateForkAtEpoch(res.get.contextEpoch):
|
|
return neterr InvalidContextBytes
|
|
var obj = ok MsgType(kind: lcDataFork)
|
|
obj.get.forky(lcDataFork) = res.get
|
|
return obj
|
|
else:
|
|
return err(res.error)
|
|
else:
|
|
return neterr InvalidContextBytes
|
|
|
|
func shortLog*(s: StatusMsg): auto =
|
|
(
|
|
forkDigest: s.forkDigest,
|
|
finalizedRoot: shortLog(s.finalizedRoot),
|
|
finalizedEpoch: shortLog(s.finalizedEpoch),
|
|
headRoot: shortLog(s.headRoot),
|
|
headSlot: shortLog(s.headSlot)
|
|
)
|
|
chronicles.formatIt(StatusMsg): shortLog(it)
|
|
|
|
func disconnectReasonName(reason: uint64): string =
|
|
# haha, nim doesn't support uint64 in `case`!
|
|
if reason == uint64(ClientShutDown): "Client shutdown"
|
|
elif reason == uint64(IrrelevantNetwork): "Irrelevant network"
|
|
elif reason == uint64(FaultOrError): "Fault or error"
|
|
else: "Disconnected (" & $reason & ")"
|
|
|
|
func forkDigestAtEpoch(state: BeaconSyncNetworkState,
|
|
epoch: Epoch): ForkDigest =
|
|
state.forkDigests[].atEpoch(epoch, state.cfg)
|
|
|
|
proc getCurrentStatus(state: BeaconSyncNetworkState): StatusMsg =
|
|
let
|
|
dag = state.dag
|
|
wallSlot = state.getBeaconTime().slotOrZero
|
|
|
|
if dag != nil:
|
|
StatusMsg(
|
|
forkDigest: state.forkDigestAtEpoch(wallSlot.epoch),
|
|
finalizedRoot: dag.finalizedHead.blck.root,
|
|
finalizedEpoch: dag.finalizedHead.slot.epoch,
|
|
headRoot: dag.head.root,
|
|
headSlot: dag.head.slot)
|
|
else:
|
|
StatusMsg(
|
|
forkDigest: state.forkDigestAtEpoch(wallSlot.epoch),
|
|
finalizedRoot: state.genesisBlockRoot,
|
|
finalizedEpoch: GENESIS_EPOCH,
|
|
headRoot: state.genesisBlockRoot,
|
|
headSlot: GENESIS_SLOT)
|
|
|
|
proc checkStatusMsg(state: BeaconSyncNetworkState, status: StatusMsg):
|
|
Result[void, cstring] =
|
|
let
|
|
dag = state.dag
|
|
wallSlot = (state.getBeaconTime() + MAXIMUM_GOSSIP_CLOCK_DISPARITY).slotOrZero
|
|
|
|
if status.finalizedEpoch > status.headSlot.epoch:
|
|
# Can be equal during genesis or checkpoint start
|
|
return err("finalized epoch newer than head")
|
|
|
|
if status.headSlot > wallSlot:
|
|
return err("head more recent than wall clock")
|
|
|
|
if state.forkDigestAtEpoch(wallSlot.epoch) != status.forkDigest:
|
|
return err("fork digests differ")
|
|
|
|
if dag != nil:
|
|
if status.finalizedEpoch <= dag.finalizedHead.slot.epoch:
|
|
let blockId = dag.getBlockIdAtSlot(status.finalizedEpoch.start_slot())
|
|
if blockId.isSome and
|
|
(not status.finalizedRoot.isZero) and
|
|
status.finalizedRoot != blockId.get().bid.root:
|
|
return err("peer following different finality")
|
|
else:
|
|
if status.finalizedEpoch == GENESIS_EPOCH:
|
|
if status.finalizedRoot != state.genesisBlockRoot:
|
|
return err("peer following different finality")
|
|
|
|
ok()
|
|
|
|
proc handleStatus(peer: Peer,
|
|
state: BeaconSyncNetworkState,
|
|
theirStatus: StatusMsg): Future[bool] {.gcsafe.}
|
|
|
|
proc setStatusMsg(peer: Peer, statusMsg: StatusMsg) {.gcsafe.}
|
|
|
|
{.pop.} # TODO fix p2p macro for raises
|
|
|
|
p2pProtocol BeaconSync(version = 1,
|
|
networkState = BeaconSyncNetworkState,
|
|
peerState = BeaconSyncPeerState):
|
|
|
|
onPeerConnected do (peer: Peer, incoming: bool) {.async.}:
|
|
debug "Peer connected",
|
|
peer, peerId = shortLog(peer.peerId), incoming
|
|
# Per the eth2 protocol, whoever dials must send a status message when
|
|
# connected for the first time, but because of how libp2p works, there may
|
|
# be a race between incoming and outgoing connections and disconnects that
|
|
# makes the incoming flag unreliable / obsolete by the time we get to
|
|
# this point - instead of making assumptions, we'll just send a status
|
|
# message redundantly.
|
|
# TODO(zah)
|
|
# the spec does not prohibit sending the extra status message on
|
|
# incoming connections, but it should not be necessary - this would
|
|
# need a dedicated flow in libp2p that resolves the race conditions -
|
|
# this needs more thinking around the ordering of events and the
|
|
# given incoming flag
|
|
let
|
|
ourStatus = peer.networkState.getCurrentStatus()
|
|
theirStatus = await peer.status(ourStatus, timeout = RESP_TIMEOUT)
|
|
|
|
if theirStatus.isOk:
|
|
discard await peer.handleStatus(peer.networkState, theirStatus.get())
|
|
else:
|
|
debug "Status response not received in time",
|
|
peer, errorKind = theirStatus.error.kind
|
|
await peer.disconnect(FaultOrError)
|
|
|
|
proc status(peer: Peer,
|
|
theirStatus: StatusMsg,
|
|
response: SingleChunkResponse[StatusMsg])
|
|
{.async, libp2pProtocol("status", 1, isRequired = true).} =
|
|
let ourStatus = peer.networkState.getCurrentStatus()
|
|
trace "Sending status message", peer = peer, status = ourStatus
|
|
await response.send(ourStatus)
|
|
discard await peer.handleStatus(peer.networkState, theirStatus)
|
|
|
|
proc ping(peer: Peer, value: uint64): uint64
|
|
{.libp2pProtocol("ping", 1, isRequired = true).} =
|
|
return peer.network.metadata.seq_number
|
|
|
|
# https://github.com/ethereum/consensus-specs/blob/v1.3.0-rc.2/specs/altair/p2p-interface.md#transitioning-from-v1-to-v2
|
|
proc getMetaData(peer: Peer): uint64
|
|
{.libp2pProtocol("metadata", 1, isRequired = true).} =
|
|
raise newException(InvalidInputsError, "GetMetaData v1 unsupported")
|
|
|
|
proc getMetadata_v2(peer: Peer): altair.MetaData
|
|
{.libp2pProtocol("metadata", 2, isRequired = true).} =
|
|
return peer.network.metadata
|
|
|
|
proc beaconBlocksByRange_v2(
|
|
peer: Peer,
|
|
startSlot: Slot,
|
|
reqCount: uint64,
|
|
reqStep: uint64,
|
|
response: MultipleChunksResponse[
|
|
ref ForkedSignedBeaconBlock, MAX_REQUEST_BLOCKS])
|
|
{.async, libp2pProtocol("beacon_blocks_by_range", 2).} =
|
|
# TODO Semantically, this request should return a non-ref, but doing so
|
|
# runs into extreme inefficiency due to the compiler introducing
|
|
# hidden copies - in future nim versions with move support, this should
|
|
# be revisited
|
|
# TODO This code is more complicated than it needs to be, since the type
|
|
# of the multiple chunks response is not actually used in this server
|
|
# implementation (it's used to derive the signature of the client
|
|
# function, not in the code below!)
|
|
# TODO although you can't tell from this function definition, a magic
|
|
# client call that returns `seq[ref ForkedSignedBeaconBlock]` will
|
|
# will be generated by the libp2p macro - we guarantee that seq items
|
|
# are `not-nil` in the implementation
|
|
# TODO reqStep is deprecated - future versions can remove support for
|
|
# values != 1: https://github.com/ethereum/consensus-specs/pull/2856
|
|
|
|
trace "got range request", peer, startSlot,
|
|
count = reqCount, step = reqStep
|
|
if reqCount == 0 or reqStep == 0:
|
|
raise newException(InvalidInputsError, "Empty range requested")
|
|
|
|
var blocks: array[MAX_REQUEST_BLOCKS, BlockId]
|
|
let
|
|
dag = peer.networkState.dag
|
|
# Limit number of blocks in response
|
|
count = int min(reqCount, blocks.lenu64)
|
|
endIndex = count - 1
|
|
startIndex =
|
|
dag.getBlockRange(startSlot, reqStep,
|
|
blocks.toOpenArray(0, endIndex))
|
|
|
|
var
|
|
found = 0
|
|
bytes: seq[byte]
|
|
|
|
for i in startIndex..endIndex:
|
|
if dag.getBlockSZ(blocks[i], bytes):
|
|
# In general, there is not much intermediate time between post-merge
|
|
# blocks all being optimistic and none of them being optimistic. The
|
|
# EL catches up, tells the CL the head is verified, and that's it.
|
|
if blocks[i].slot.epoch >= dag.cfg.BELLATRIX_FORK_EPOCH and
|
|
dag.is_optimistic(dag.head.root):
|
|
continue
|
|
|
|
let uncompressedLen = uncompressedLenFramed(bytes).valueOr:
|
|
warn "Cannot read block size, database corrupt?",
|
|
bytes = bytes.len(), blck = shortLog(blocks[i])
|
|
continue
|
|
|
|
# TODO extract from libp2pProtocol
|
|
peer.awaitQuota(blockResponseCost, "beacon_blocks_by_range/2")
|
|
peer.network.awaitQuota(blockResponseCost, "beacon_blocks_by_range/2")
|
|
|
|
await response.writeBytesSZ(
|
|
uncompressedLen, bytes,
|
|
peer.networkState.forkDigestAtEpoch(blocks[i].slot.epoch).data)
|
|
|
|
inc found
|
|
|
|
debug "Block range request done",
|
|
peer, startSlot, count, reqStep
|
|
|
|
proc beaconBlocksByRoot_v2(
|
|
peer: Peer,
|
|
# Please note that the SSZ list here ensures that the
|
|
# spec constant MAX_REQUEST_BLOCKS is enforced:
|
|
blockRoots: BlockRootsList,
|
|
response: MultipleChunksResponse[
|
|
ref ForkedSignedBeaconBlock, MAX_REQUEST_BLOCKS])
|
|
{.async, libp2pProtocol("beacon_blocks_by_root", 2).} =
|
|
# TODO Semantically, this request should return a non-ref, but doing so
|
|
# runs into extreme inefficiency due to the compiler introducing
|
|
# hidden copies - in future nim versions with move support, this should
|
|
# be revisited
|
|
# TODO This code is more complicated than it needs to be, since the type
|
|
# of the multiple chunks response is not actually used in this server
|
|
# implementation (it's used to derive the signature of the client
|
|
# function, not in the code below!)
|
|
# TODO although you can't tell from this function definition, a magic
|
|
# client call that returns `seq[ref ForkedSignedBeaconBlock]` will
|
|
# will be generated by the libp2p macro - we guarantee that seq items
|
|
# are `not-nil` in the implementation
|
|
if blockRoots.len == 0:
|
|
raise newException(InvalidInputsError, "No blocks requested")
|
|
|
|
let
|
|
dag = peer.networkState.dag
|
|
count = blockRoots.len
|
|
|
|
var
|
|
found = 0
|
|
bytes: seq[byte]
|
|
|
|
for i in 0..<count:
|
|
let
|
|
blockRef = dag.getBlockRef(blockRoots[i]).valueOr:
|
|
continue
|
|
|
|
if dag.getBlockSZ(blockRef.bid, bytes):
|
|
# In general, there is not much intermediate time between post-merge
|
|
# blocks all being optimistic and none of them being optimistic. The
|
|
# EL catches up, tells the CL the head is verified, and that's it.
|
|
if blockRef.slot.epoch >= dag.cfg.BELLATRIX_FORK_EPOCH and
|
|
dag.is_optimistic(dag.head.root):
|
|
continue
|
|
|
|
let uncompressedLen = uncompressedLenFramed(bytes).valueOr:
|
|
warn "Cannot read block size, database corrupt?",
|
|
bytes = bytes.len(), blck = shortLog(blockRef)
|
|
continue
|
|
|
|
# TODO extract from libp2pProtocol
|
|
peer.awaitQuota(blockResponseCost, "beacon_blocks_by_root/2")
|
|
peer.network.awaitQuota(blockResponseCost, "beacon_blocks_by_root/2")
|
|
|
|
await response.writeBytesSZ(
|
|
uncompressedLen, bytes,
|
|
peer.networkState.forkDigestAtEpoch(blockRef.slot.epoch).data)
|
|
|
|
inc found
|
|
|
|
debug "Block root request done",
|
|
peer, roots = blockRoots.len, count, found
|
|
|
|
# https://github.com/ethereum/consensus-specs/blob/v1.3.0-rc.1/specs/eip4844/p2p-interface.md#beaconblockandblobssidecarbyroot-v1
|
|
proc beaconBlockAndBlobsSidecarByRoot_v1(
|
|
peer: Peer,
|
|
# Please note that the SSZ list here ensures that the
|
|
# spec constant MAX_REQUEST_BLOCKS is enforced:
|
|
blockRoots: BlockRootsList,
|
|
response: MultipleChunksResponse[
|
|
ref SignedBeaconBlockAndBlobsSidecar, MAX_REQUEST_BLOCKS])
|
|
{.async, libp2pProtocol("beacon_block_and_blobs_sidecar_by_root", 1).} =
|
|
# unlike for beaconBlocksByRoot_v2, we don't need to
|
|
# dynamically decode the correct fork here. so returning a ref
|
|
# is solely for performance sake
|
|
|
|
if blockRoots.len == 0:
|
|
raise newException(InvalidInputsError, "No blocks requested")
|
|
|
|
let
|
|
dag = peer.networkState.dag
|
|
count = blockRoots.len
|
|
# https://github.com/ethereum/consensus-specs/blob/v1.3.0-rc.1/specs/eip4844/p2p-interface.md#beaconblockandblobssidecarbyroot-v1
|
|
min_epochs =
|
|
if MIN_EPOCHS_FOR_BLOBS_SIDECARS_REQUESTS > dag.head.slot.epoch:
|
|
dag.head.slot.epoch
|
|
else:
|
|
Epoch(MIN_EPOCHS_FOR_BLOBS_SIDECARS_REQUESTS)
|
|
|
|
epochBoundary =
|
|
if dag.head.slot.epoch - min_epochs < dag.cfg.EIP4844_FORK_EPOCH:
|
|
dag.cfg.EIP4844_FORK_EPOCH
|
|
else:
|
|
dag.head.slot.epoch - MIN_EPOCHS_FOR_BLOBS_SIDECARS_REQUESTS
|
|
minimum_request_epoch = max(dag.finalizedHead.slot.epoch, epochBoundary)
|
|
|
|
var
|
|
found = 0
|
|
bytes: seq[byte]
|
|
blck: Opt[eip4844.TrustedSignedBeaconBlock]
|
|
blobsSidecar: Opt[BlobsSidecar]
|
|
|
|
for i in 0..<count:
|
|
let
|
|
blockRef = dag.getBlockRef(blockRoots[i]).valueOr:
|
|
continue
|
|
|
|
blck = dag.getBlock(blockRef.bid, eip4844.TrustedSignedBeaconBlock)
|
|
if blck.isNone():
|
|
continue
|
|
|
|
if blockRef.bid.slot.epoch < minimum_request_epoch:
|
|
raise newException(ResourceUnavailableError, BlobsOutOfRange)
|
|
|
|
# In general, there is not much intermediate time between post-merge
|
|
# blocks all being optimistic and none of them being optimistic. The
|
|
# EL catches up, tells the CL the head is verified, and that's it.
|
|
if blockRef.slot.epoch >= dag.cfg.BELLATRIX_FORK_EPOCH and
|
|
dag.is_optimistic(dag.head.root):
|
|
continue
|
|
|
|
blobsSidecar = dag.db.getBlobsSidecar(blockRef.bid.root)
|
|
if blobsSidecar.isNone():
|
|
continue
|
|
|
|
peer.awaitQuota(blockResponseCost, "beacon_block_and_blobs_sidecar_by_root/1")
|
|
peer.network.awaitQuota(blockResponseCost, "beacon_block_and_blobs_sidecar_by_root/1")
|
|
|
|
let sbbabs = SignedBeaconBlockAndBlobsSidecar(
|
|
beacon_block: asSigned(blck.get()),
|
|
blobs_sidecar: blobsSidecar.get())
|
|
bytes = snappy.encodeFramed(SSZ.encode(sbbabs))
|
|
|
|
let uncompressedLen = uncompressedLenFramed(bytes).valueOr:
|
|
warn "Cannot read block and blobs size", length = bytes.len()
|
|
continue
|
|
|
|
await response.writeBytesSZ(
|
|
uncompressedLen, bytes,
|
|
peer.networkState.forkDigestAtEpoch(blockRef.slot.epoch).data)
|
|
|
|
inc found
|
|
|
|
debug "Block and blobs sidecar root request done",
|
|
peer, roots = blockRoots.len, count, found
|
|
|
|
# https://github.com/ethereum/consensus-specs/blob/v1.3.0-rc.0/specs/eip4844/p2p-interface.md#blobssidecarsbyrange-v1
|
|
proc blobsSidecarsByRange(
|
|
peer: Peer,
|
|
startSlot: Slot,
|
|
reqCount: uint64,
|
|
response: MultipleChunksResponse[
|
|
ref BlobsSidecar, MAX_REQUEST_BLOBS_SIDECARS])
|
|
{.async, libp2pProtocol("blobs_sidecars_by_range", 1).} =
|
|
# TODO This code is more complicated than it needs to be, since the type
|
|
# of the multiple chunks response is not actually used in this server
|
|
# implementation (it's used to derive the signature of the client
|
|
# function, not in the code below!)
|
|
# TODO although you can't tell from this function definition, a magic
|
|
# client call that returns `seq[ref BlobsSidecar]` will
|
|
# will be generated by the libp2p macro - we guarantee that seq items
|
|
# are `not-nil` in the implementation
|
|
|
|
trace "got blobs range request", peer, startSlot, count = reqCount
|
|
if reqCount == 0:
|
|
raise newException(InvalidInputsError, "Empty range requested")
|
|
|
|
let
|
|
dag = peer.networkState.dag
|
|
epochBoundary =
|
|
if MIN_EPOCHS_FOR_BLOBS_SIDECARS_REQUESTS >= dag.head.slot.epoch:
|
|
GENESIS_EPOCH
|
|
else:
|
|
dag.head.slot.epoch - MIN_EPOCHS_FOR_BLOBS_SIDECARS_REQUESTS
|
|
|
|
if startSlot.epoch < epochBoundary:
|
|
raise newException(ResourceUnavailableError, BlobsOutOfRange)
|
|
|
|
var blockIds: array[MAX_REQUEST_BLOBS_SIDECARS, BlockId]
|
|
let
|
|
# Limit number of blocks in response
|
|
count = int min(reqCount, blockIds.lenu64)
|
|
endIndex = count - 1
|
|
startIndex =
|
|
dag.getBlockRange(startSlot, 1, blockIds.toOpenArray(0, endIndex))
|
|
|
|
var
|
|
found = 0
|
|
bytes: seq[byte]
|
|
|
|
for i in startIndex..endIndex:
|
|
if dag.db.getBlobsSidecarSZ(blockIds[i].root, bytes):
|
|
# In general, there is not much intermediate time between post-merge
|
|
# blocks all being optimistic and none of them being optimistic. The
|
|
# EL catches up, tells the CL the head is verified, and that's it.
|
|
if blockIds[i].slot.epoch >= dag.cfg.BELLATRIX_FORK_EPOCH and
|
|
dag.is_optimistic(dag.head.root):
|
|
continue
|
|
|
|
let uncompressedLen = uncompressedLenFramed(bytes).valueOr:
|
|
warn "Cannot read blobs sidecar size, database corrupt?",
|
|
bytes = bytes.len(), blck = shortLog(blockIds[i])
|
|
continue
|
|
|
|
# TODO extract from libp2pProtocol
|
|
peer.awaitQuota(blockResponseCost, "blobs_sidecars_by_range/1")
|
|
peer.network.awaitQuota(blockResponseCost, "blobs_sidecars_by_range/1")
|
|
|
|
await response.writeBytesSZ(
|
|
uncompressedLen, bytes,
|
|
peer.networkState.forkDigestAtEpoch(blockIds[i].slot.epoch).data)
|
|
|
|
inc found
|
|
|
|
debug "BlobsSidecar range request done",
|
|
peer, startSlot, count
|
|
|
|
# https://github.com/ethereum/consensus-specs/blob/v1.3.0-rc.0/specs/altair/light-client/p2p-interface.md#getlightclientbootstrap
|
|
proc lightClientBootstrap(
|
|
peer: Peer,
|
|
blockRoot: Eth2Digest,
|
|
response: SingleChunkResponse[ForkedLightClientBootstrap])
|
|
{.async, libp2pProtocol("light_client_bootstrap", 1,
|
|
isLightClientRequest = true).} =
|
|
trace "Received LC bootstrap request", peer, blockRoot
|
|
let dag = peer.networkState.dag
|
|
doAssert dag.lcDataStore.serve
|
|
|
|
let bootstrap = dag.getLightClientBootstrap(blockRoot)
|
|
withForkyBootstrap(bootstrap):
|
|
when lcDataFork > LightClientDataFork.None:
|
|
let
|
|
contextEpoch = forkyBootstrap.contextEpoch
|
|
contextBytes = peer.networkState.forkDigestAtEpoch(contextEpoch).data
|
|
|
|
# TODO extract from libp2pProtocol
|
|
peer.awaitQuota(
|
|
lightClientBootstrapResponseCost,
|
|
"light_client_bootstrap/1")
|
|
await response.sendSSZ(forkyBootstrap, contextBytes)
|
|
else:
|
|
raise newException(ResourceUnavailableError, LCBootstrapUnavailable)
|
|
|
|
debug "LC bootstrap request done", peer, blockRoot
|
|
|
|
# https://github.com/ethereum/consensus-specs/blob/v1.3.0-alpha.0/specs/altair/light-client/p2p-interface.md#lightclientupdatesbyrange
|
|
proc lightClientUpdatesByRange(
|
|
peer: Peer,
|
|
startPeriod: SyncCommitteePeriod,
|
|
reqCount: uint64,
|
|
response: MultipleChunksResponse[
|
|
ForkedLightClientUpdate, MAX_REQUEST_LIGHT_CLIENT_UPDATES])
|
|
{.async, libp2pProtocol("light_client_updates_by_range", 1,
|
|
isLightClientRequest = true).} =
|
|
trace "Received LC updates by range request", peer, startPeriod, reqCount
|
|
let dag = peer.networkState.dag
|
|
doAssert dag.lcDataStore.serve
|
|
|
|
let
|
|
headPeriod = dag.head.slot.sync_committee_period
|
|
# Limit number of updates in response
|
|
maxSupportedCount =
|
|
if startPeriod > headPeriod:
|
|
0'u64
|
|
else:
|
|
min(headPeriod + 1 - startPeriod, MAX_REQUEST_LIGHT_CLIENT_UPDATES)
|
|
count = min(reqCount, maxSupportedCount)
|
|
onePastPeriod = startPeriod + count
|
|
|
|
var found = 0
|
|
for period in startPeriod..<onePastPeriod:
|
|
let update = dag.getLightClientUpdateForPeriod(period)
|
|
withForkyUpdate(update):
|
|
when lcDataFork > LightClientDataFork.None:
|
|
let
|
|
contextEpoch = forkyUpdate.contextEpoch
|
|
contextBytes =
|
|
peer.networkState.forkDigestAtEpoch(contextEpoch).data
|
|
|
|
# TODO extract from libp2pProtocol
|
|
peer.awaitQuota(
|
|
lightClientUpdateResponseCost,
|
|
"light_client_updates_by_range/1")
|
|
await response.writeSSZ(forkyUpdate, contextBytes)
|
|
inc found
|
|
else:
|
|
discard
|
|
|
|
debug "LC updates by range request done", peer, startPeriod, count, found
|
|
|
|
# https://github.com/ethereum/consensus-specs/blob/v1.3.0-rc.0/specs/altair/light-client/p2p-interface.md#getlightclientfinalityupdate
|
|
proc lightClientFinalityUpdate(
|
|
peer: Peer,
|
|
response: SingleChunkResponse[ForkedLightClientFinalityUpdate])
|
|
{.async, libp2pProtocol("light_client_finality_update", 1,
|
|
isLightClientRequest = true).} =
|
|
trace "Received LC finality update request", peer
|
|
let dag = peer.networkState.dag
|
|
doAssert dag.lcDataStore.serve
|
|
|
|
let finality_update = dag.getLightClientFinalityUpdate()
|
|
withForkyFinalityUpdate(finality_update):
|
|
when lcDataFork > LightClientDataFork.None:
|
|
let
|
|
contextEpoch = forkyFinalityUpdate.contextEpoch
|
|
contextBytes = peer.networkState.forkDigestAtEpoch(contextEpoch).data
|
|
|
|
# TODO extract from libp2pProtocol
|
|
peer.awaitQuota(
|
|
lightClientFinalityUpdateResponseCost,
|
|
"light_client_finality_update/1")
|
|
await response.sendSSZ(forkyFinalityUpdate, contextBytes)
|
|
else:
|
|
raise newException(ResourceUnavailableError, LCFinUpdateUnavailable)
|
|
|
|
debug "LC finality update request done", peer
|
|
|
|
# https://github.com/ethereum/consensus-specs/blob/v1.3.0-alpha.0/specs/altair/light-client/p2p-interface.md#getlightclientoptimisticupdate
|
|
proc lightClientOptimisticUpdate(
|
|
peer: Peer,
|
|
response: SingleChunkResponse[ForkedLightClientOptimisticUpdate])
|
|
{.async, libp2pProtocol("light_client_optimistic_update", 1,
|
|
isLightClientRequest = true).} =
|
|
trace "Received LC optimistic update request", peer
|
|
let dag = peer.networkState.dag
|
|
doAssert dag.lcDataStore.serve
|
|
|
|
let optimistic_update = dag.getLightClientOptimisticUpdate()
|
|
withForkyOptimisticUpdate(optimistic_update):
|
|
when lcDataFork > LightClientDataFork.None:
|
|
let
|
|
contextEpoch = forkyOptimisticUpdate.contextEpoch
|
|
contextBytes = peer.networkState.forkDigestAtEpoch(contextEpoch).data
|
|
|
|
# TODO extract from libp2pProtocol
|
|
peer.awaitQuota(
|
|
lightClientOptimisticUpdateResponseCost,
|
|
"light_client_optimistic_update/1")
|
|
await response.sendSSZ(forkyOptimisticUpdate, contextBytes)
|
|
else:
|
|
raise newException(ResourceUnavailableError, LCOptUpdateUnavailable)
|
|
|
|
debug "LC optimistic update request done", peer
|
|
|
|
proc goodbye(peer: Peer,
|
|
reason: uint64)
|
|
{.async, libp2pProtocol("goodbye", 1, isRequired = true).} =
|
|
debug "Received Goodbye message", reason = disconnectReasonName(reason), peer
|
|
|
|
proc setStatusMsg(peer: Peer, statusMsg: StatusMsg) =
|
|
debug "Peer status", peer, statusMsg
|
|
peer.state(BeaconSync).statusMsg = statusMsg
|
|
peer.state(BeaconSync).statusLastTime = Moment.now()
|
|
|
|
proc handleStatus(peer: Peer,
|
|
state: BeaconSyncNetworkState,
|
|
theirStatus: StatusMsg): Future[bool] {.async, gcsafe.} =
|
|
let
|
|
res = checkStatusMsg(state, theirStatus)
|
|
|
|
return if res.isErr():
|
|
debug "Irrelevant peer", peer, theirStatus, err = res.error()
|
|
await peer.disconnect(IrrelevantNetwork)
|
|
false
|
|
else:
|
|
peer.setStatusMsg(theirStatus)
|
|
|
|
if peer.connectionState == Connecting:
|
|
# As soon as we get here it means that we passed handshake succesfully. So
|
|
# we can add this peer to PeerPool.
|
|
await peer.handlePeer()
|
|
true
|
|
|
|
proc updateStatus*(peer: Peer): Future[bool] {.async.} =
|
|
## Request `status` of remote peer ``peer``.
|
|
let
|
|
nstate = peer.networkState(BeaconSync)
|
|
ourStatus = getCurrentStatus(nstate)
|
|
|
|
let theirFut = awaitne peer.status(ourStatus, timeout = RESP_TIMEOUT)
|
|
if theirFut.failed():
|
|
return false
|
|
else:
|
|
let theirStatus = theirFut.read()
|
|
if theirStatus.isOk:
|
|
return await peer.handleStatus(nstate, theirStatus.get())
|
|
else:
|
|
return false
|
|
|
|
proc getHeadSlot*(peer: Peer): Slot =
|
|
## Returns head slot for specific peer ``peer``.
|
|
peer.state(BeaconSync).statusMsg.headSlot
|
|
|
|
proc getFinalizedEpoch*(peer: Peer): Epoch =
|
|
## Returns head slot for specific peer ``peer``.
|
|
peer.state(BeaconSync).statusMsg.finalizedEpoch
|
|
|
|
proc initBeaconSync*(network: Eth2Node, dag: ChainDAGRef,
|
|
getBeaconTime: GetBeaconTimeFn) =
|
|
var networkState = network.protocolState(BeaconSync)
|
|
networkState.dag = dag
|
|
networkState.cfg = dag.cfg
|
|
networkState.forkDigests = dag.forkDigests
|
|
networkState.genesisBlockRoot = dag.genesisBlockRoot
|
|
networkState.getBeaconTime = getBeaconTime
|
|
|
|
proc initBeaconSync*(network: Eth2Node,
|
|
cfg: RuntimeConfig,
|
|
forkDigests: ref ForkDigests,
|
|
genesisBlockRoot: Eth2Digest,
|
|
getBeaconTime: GetBeaconTimeFn) =
|
|
var networkState = network.protocolState(BeaconSync)
|
|
networkState.dag = nil
|
|
networkState.cfg = cfg
|
|
networkState.forkDigests = forkDigests
|
|
networkState.genesisBlockRoot = genesisBlockRoot
|
|
networkState.getBeaconTime = getBeaconTime
|