* drop support for v4 (obsolete, doesn't work with all clients since
they use chunking and other obsolete v4 features that we're missing or
don't support it at all)
* rework asyncraises
* always store generated p2p macro code (similar to eth2)
* preparation for chronos cancellation support (more to be done here)
* when peer is disconnected, ensure pending handshakes and requests are
notified (instead of waiting for timeout)
* disallow raising from `onPeerDisconnected` - this simplifies
disconnection coordination among async tasks
* introduce several warning logs for protocol breaches - these should be
removed eventually, pending q/a on the rlpx layer in general
* fix snappy compression - the payload without msgId should be
compressed
* remove strict checks on unused fields in RLPx message header (this
matches geth behavior and the spirit of EIP-8)
* add snappy dep
* disconnect peers that send non-hello messages during initial hello
step
* fix devp2p protocol version - 4 because we don't implement snappy
(yet) - this is cosmetic since this particular version field is not
actually being used
* fix ack message length checking
* move RLPx transport code to separate module, annotate with asyncraises
* increase max RLPx message size to 16mb, per EIP-706
* make sure both accept/connect timeout after 10s
* aim to log every connection attempt once at debug level
* make capability-id/context-id check more accurate
* disallow random messages before hello
Since these types were written, we've gained an executable spec:
https://github.com/ethereum/execution-specs
This PR aligns some of the types we use with this spec to simplify
comparisons and cross-referencing.
Using a `distinct` type is a tradeoff between nim ergonomics, type
safety and the ability to work around nim quirks and stdlib weaknesses.
In particular, it allows us to overload common functions such as `hash`
with correct and performant versions as well as maintain control over
string conversions etc at the cost of a little bit of ceremony when
instantiating them.
Apart from distinct byte types, `Hash32`, is introduced in lieu of the
existing `Hash256`, again aligning this commonly used type with the spec
which picks bytes rather than bits in the name.
* Using unsigned types for message type and requst IDs
why:
Negative values are neither defined for RLP nor in the protocol specs
which refer to the RLPs (see yellow paper app B clause (199).
* Fix `int` argument (must be `uint`) in fuzzing tests
why:
Not part of all tests so it slipped through.
* Refactor p2pProtocol internals
* Attempt to fix rlp crash with Nim v2 (#658)
* Attempt to fix rlp crash with Nim v2
* Fix test_ecies for nim v2
* Reduce compiler warnings
* Resolve ambiquity in testutils
* Disable nim devel continue-on-error
* Add metrics related to devp2p peer connections
* Avoid reconnecting to peers that just failed connection
- Add SeenTable to avoid reconnecting to peers immediately after
a failed connect. Depending on the failure, the amount of time is
different. This is similar to what is done in nimbus-eth2.
- Attempt to rework rlpxConnect at the same time, in order to
make sure that errors are properly handled. The current structure
is far from ideal, but it is hopefully a small step in the right
direction. To many oddities in there right now to really rework
rlpxConnect properply.
* Fix rlpx thunk fuzzer
The old code, which did it in sequence, made it take a very long
time to accumulate peers in the peer pool, particularly because
occasionally a connection-attempt would hang for a couple of minutes
before timing out.
RLP Enum deserialization would currently not check if "hole values"
were attempted to be converted to the enum type. Now use
checkedEnumAssign and fail with RlpTypeMismatch on invalid values.
There is at least one occurance of an enum with holes in rlpx p2p:
DisconnectionReason. For this enum the issue could occur.
Also:
- Added enum RLP tests and rlpx p2p disconnect message tests to
test the DisconnectionReason with enum hole value.
- Fixed worse custom DisconnectionReason decoding occurance
in rlpx in waitSingleMsg proc where this issue could occur.
Nim devel brach(1.7.1) introduce gc=orc as default mode.
Because the p2p protocol using unsafe pointer operations
for it's ProtocolInfo and using global variables scattered
around, the orc mistakenly(or maybe correctly) crash the protocol.
* Move NetworkId type to common eth_types
NetworkId is after all a common type and this way it avoid an
application that requires it to also import all devp2p related
network types.
RLP related calls are moved to eth_types_rlp, this means that p2p
protocols that send NetworkId over the wire need to import this
too. Or just common in general.
* Remove # in front of multiline comment end bracket
These seem to be interpreted wrongly by the GitHub code browser.
`eth_types` is being imported from many projects and ends up causing
long build times due to its extensive import lists - this PR starts
cleaning some of that up by moving the chain DB and RLP to their own
modules.
this PR also moves `keccakHash` to its own module and uses it in many
places.
* Provide dedicated `DisconnectionReason` enum type RLP reader
why:
Without this reader, the program communicating via RLPX will crash when
receiving out of bound reason codes disconnect message.
Out of bound value assignments to an enum causes a `RangeError`defect
and consequently the program to terminate. This `RangeError` is avoided
here and a `MalformedRlpError` catchable error raised.
* Using default exception type in bespoke `read(DisconnectionReason)`
why:
This should not differ from the default enum parser. The particular
message is different and more targeted, here.
Note: The default RLP parser was not used because `
`array[1,DisconnectionReason]` is currently not properly handled and
should give a siliar error message as a `DisconnectionReason` error.
* De-clutter, custom read() was not needed
Co-authored-by: jordan <jordan@curd.mjh-it.com>
* Add build_dcli target and add it to CI
* Fix local imports for dcli
* And use local imports for all other files too
* Use local imports in tests and rlpx protocols
* use bearssl rng throughout
* seeder can fail
* imports and exports
* modules, sigh
* one more try
* move var
* even fewer thread vars
* remove out-of-date genrated files
* simplify some modules
* mark several modules with raises
* fix clearing of keys in auth.nim
* fix keyfile case dropping off
* fix keyfile stream storage
* uuid should be output in lowercase
* enode: simplify API
* Make messages with invalid ID a catchable error as we should not disconnect on this
* Add unimplemented message IDs used by Status
* Make whisper log less verbose + add comment on duplicate messages
* Cleanup + add documentation
Lib2P2 handles RPC requests and responses with separate streams
while DEV2P2 is relying on tagged messages transmitted over a
single stream. To cover both models through the same application
code, we introduce a new `response` variable in the request handlers.
The user is supposed to issue a call to `response.send` in order to
reply to the request. Please note that the `response.send` signature
is strongly typed and depends on the current message.
Jacek Sieka
Jordan Hrycaj
andri lim
Kim De Mey
Adam Spitz
Jordan Hrycaj
Zahary Karadjov
Jacek Sieka
Mamy Ratsimbazafy
Yuriy Glukhov