We comment out a newly-added assert from the Bitcoin
side, because it's unclear how to replicate it
correctly in Elements.
Co-authored-by: Glenn Willen <gwillen@nerdnet.org>
When receiving unwanted headers from a peer, first check whether we have
recent "best known block" state for that peer. If not, accept one batch
of headers to refresh it. Otherwise, we can become stuck in a state where
we need blocks, but we don't know which peers have those blocks available.
This especially happens during startup, when we don't have this information
for any peers until we start processing header messages.
1066d10f71 scripted-diff: rename TxRelay members (John Newbery)
575bbd0dea [net processing] Move tx relay data to Peer (John Newbery)
785f55f7ee [net processing] Move m_wtxid_relay to Peer (John Newbery)
36346703f8 [net] Add CNode.m_relays_txs and CNode.m_bloom_filter_loaded (John Newbery)
Pull request description:
This continues the work of moving application layer data into net_processing, by moving all tx data into the new Peer object added in #19607.
For motivation, see #19398.
ACKs for top commit:
dergoegge:
ACK 1066d10f71 - This is a good layer separation improvement with no behavior changes.
glozow:
utACK 1066d10f71
Tree-SHA512: 0c9d6b8a0a05e2d816b6d6588b7df133842ec960ae67667813422aa7bd8eb5308599c714f3822a98ddbdf364ffab9050b055079277ba4aff24092557ff99ebcc
This makes code that uses the helper less verbose.
Moreover, this makes net_processing C++20 compliant. Otherwise, it would
lead to a compile error (see below). C++20 disables aggregate
initialization when any constructor is declared. See
http://open-std.org/JTC1/SC22/WG21/docs/papers/2018/p1008r1.pdf
net_processing.cpp:1627:42: error: no matching constructor for initialization of 'CSerializedNetMsg'
m_connman.PushMessage(pnode, CSerializedNetMsg{ser_cmpctblock.data, ser_cmpctblock.m_type});
^ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
fa61dd44f9 p2p: Serialize cmpctblock at most once in NewPoWValidBlock (MarcoFalke)
Pull request description:
Instead of serializing for each peer, serialize at most once and copy the raw data for each peer.
ACKs for top commit:
shaavan:
reACK fa61dd44f9
theStack:
Code-review ACK fa61dd44f9
Tree-SHA512: ed029aeaea67fdac8ddb865069f8166bc0dd8480418c405628e3e1a43b61161584a09a1814668bcd220602e8732e188be2bfed9242aa81bdbd92c64c702ed138
Also, remove cs_main guard from m_wtxid_relay_peers and make it atomic.
This should be fine since we don't need m_wtxid_relay_peers to be
synchronized with m_wtxid_relay exactly at all times.
After this change, RelayTransaction no longer requires cs_main.
We'll move the transaction relay data into Peer in subsequent commits,
but the inbound eviction logic needs to know if the peer is relaying
txs and if the peer has loaded a bloom filter.
This is currently redundant information with m_tx_relay->fRelayTxes,
but when m_tx_relay is moved into net_processing, then we'll need these
separate fields in CNode.
faa329fd46 refactor: Release cs_main before MaybeSendFeefilter (MarcoFalke)
Pull request description:
There is no need for any lock to be held, because net processing is single threaded. So holding the validation lock cs_main for sending a feefilter is confusing and might even degrade blockchain-related RPC performance minimally.
ACKs for top commit:
ajtowns:
ACK faa329fd46 ; code review only
vasild:
ACK faa329fd46
Tree-SHA512: 3e7f9faff1631cc64c86fc1a354ada67617ad1e7a046625cc741f4711854eb41ca8aad5a51ef0d94ff65947b68dba8345c9f786b20ee0a8b7a2e8741cfced21f
`GetListenPort()` uses a simple logic: "if `-port=P` is given, then we
must be listening on `P`, otherwise we must be listening on `8333`".
This is however not true if `-bind=` has been provided with `:port` part
or if `-whitebind=` has been provided. Thus, extend `GetListenPort()` to
return the port from `-bind=` or `-whitebind=`, if any.
Fixes https://github.com/bitcoin/bitcoin/issues/20184 (cases 1. 2. 3. 5.)
In `PeerManagerImpl::RelayAddress()` we used just the hash of the
address that is being relayed to decide where to relay it to. Include
the port in that hash, so that e.g. `1.1.1.1:5555` and `1.1.1.1:6666`
get relayed to different peers. Those are two different, distinct
services after all.
224d87855e net, refactor: Drop tautological local variables (Hennadii Stepanov)
3073a9917b scripted-diff: Rename CNetMessage::m_command with CNetMessage::m_type (Hennadii Stepanov)
Pull request description:
https://github.com/bitcoin/bitcoin/pull/18533#issue-594592488:
> a message is not a command, but simply a message of some type
Continuation of bitcoin/bitcoin#18533 and bitcoin/bitcoin#18937.
ACKs for top commit:
theStack:
Concept and code-review ACK 224d87855e
shaavan:
Code Review ACK 224d87855e
w0xlt:
crACK 224d878
Tree-SHA512: 898cafb44708dae1413fcc1533d809d75878891354f1b5edaaec1287f4921c31adc9330f4d42d82544a39689886bc17fee71ea587f9199fd5cc849d376f82176
9b8dcb25b5 [net processing] Rename PoissonNextSendInbound to NextInvToInbounds (John Newbery)
ea99f5d01e [net processing] Move PoissonNextSendInbound to PeerManager (John Newbery)
bb060746df scripted-diff: replace PoissonNextSend with GetExponentialRand (John Newbery)
03cfa1b603 [refactor] Use uint64_t and std namespace in PoissonNextSend (John Newbery)
9e64d69bf7 [move] Move PoissonNextSend to src/random and update comment (John Newbery)
Pull request description:
`PoissonNextSend` and `PoissonNextSendInbound` are used in the p2p code to obfuscate various regularly occurring processes, in order to make it harder for others to get timing-based information deterministically.
The naming of these functions has been confusing to several people (including myself, see also #23347) because the resulting random timestamps don't follow a Poisson distribution but an exponential distribution (related to events in a Poisson process, hence the name). This PR
- moves `PoissonNextSend()` out of `net` to `random` and renames it to `GetExponentialRand()`
- moves `PoissonNextSendInbound()` out of `CConnman` to `PeerManager` and renames it to `NextInvToInbounds()`
- adds documentation for these functions
This is work by jnewbery - due to him being less active currently, I opened the PR and will address feedback.
ACKs for top commit:
jnewbery:
ACK 9b8dcb25b5
hebasto:
ACK 9b8dcb25b5, I have reviewed the code and it looks OK, I agree it can be merged.
theStack:
ACK 9b8dcb25b5📊
Tree-SHA512: 85c366c994e7147f9981fe863fb9838502643fa61ffd32d55a43feef96a38b79a5daa2c4d38ce01074897cc95fa40c76779816edad53f5265b81b05c3a1f4f50
-BEGIN VERIFY SCRIPT-
sed -i 's/std::string m_command;/std::string m_type;/g' ./src/net.h
sed -i 's/* command and size./* type and size./g' ./src/net.h
sed -i 's/msg.m_command/msg.m_type/g' ./src/net.cpp ./src/net_processing.cpp ./src/test/fuzz/p2p_transport_serialization.cpp
-END VERIFY SCRIPT-
This distribution is used for more than just the next inv send, so make
the name more generic.
Also rename to "exponential" to avoid the confusion that this is a
poisson distribution.
-BEGIN VERIFY SCRIPT-
ren() { sed -i "s/\<$1\>/$2/g" $(git grep -l "$1" ./src) ; }
ren PoissonNextSend GetExponentialRand
ren "a poisson timer" "an exponential timer"
-END VERIFY SCRIPT-
fe86eb50c9 Refactor: Uses c++ init convention for time variables (Shashwat)
6111b0d7fa Refactor: Changes remaining time variable type from int to chrono (Shashwat)
Pull request description:
This PR is a follow-up to #23801.
This PR aims to make the following changes to all the time variables in **net_processing.cpp** wherever possible.
- Convert all time variables to `std::chrono.`
- Use `chorno::literals` wherever possible.
- Use `auto` keywords wherever possible.
- Use `var{val}` convention of initialization.
This PR also minimizes the number of times, serialization of time `count_seconds(..)` occurs.
ACKs for top commit:
MarcoFalke:
re-ACK fe86eb50c9 🏕
Tree-SHA512: c8684c0c60a11140027e36b6e9706a45ecdeae6b5ba0bf267e50655835daee5e5410e34096a8c4eca005f327caae1ac258cc7b8ba663eab58abf131f6d2f4d42
This commit does following changes to time variables in net_processing.h:
- Used {} initialization.
- Uses universal initializer auto.
- Uses chrono::literals
The reason for these changes is to make code simpler, and easier to
understand and rationalize.
92082ea0bb Change time variable type to std::chrono::seconds in src/net_processing.cpp (Shashwat)
Pull request description:
- This is a follow-up to PR #23758
- This changes the remaining time variable in `net_processing.cpp` from **int64_t** to **std::chrono::seconds**
ACKs for top commit:
naumenkogs:
ACK 92082ea0bb
hebasto:
re-ACK 92082ea0bb
Tree-SHA512: 559e351d9046d4ba2b842ae38da13b4befc7feee71f0762f97907812471e2840b0d43c90c92222d15619fe40cc21f11d40900500ca07b470f7ac8b0046cc1d68
fa1dc9b36a p2p: Always serialize local timestamp for version msg (MarcoFalke)
Pull request description:
Currently we serialize the local time when connecting to outbound connections and the "adjusted network" time when someone connects to us.
I presume the reason is to avoid a fingerprint in case the local time is misconfigured. However, the fingerprint still exits when:
* The local time goes out-of-sync after timedata is filled up, in which case the adjusted time is *not* adjusted. See comment in `src/timedata.cpp`. (In practise I expect no adjustment to happen after timedata is filled up by one entry more than half its size).
* The local time is off by more than 70 minutes. See `DEFAULT_MAX_TIME_ADJUSTMENT`. While there is a warning in this case, the warning might be missed by the node operator.
* The adjusted time is poisoned by an attacker. This is only a theoretical concern after commit e457513eb1.
Using the adjusted time does help in a the case where the local time is off by a constant less than 70 minutes and the node quickly connects to 5 outbound peers to retrieve the adjusted time.
Still, I think using `GetAdjustedTime` here gives a false sense of security. It will be better for node operators to instead set the correct time.
ACKs for top commit:
naumenkogs:
ACK fa1dc9b36a
laanwj:
Code review ACK fa1dc9b36a
w0xlt:
crACK fa1dc9b
Tree-SHA512: 70a0f4ab3500e6ddcde291620e35273018cefd1d9e94b91ad333e360139ed18862718bb1a9854af2bf79990bf74b05d95492f77d0747c7b9bdd276c020116dcb
