system/serenity
system/serenity
1
0
Fork 0
mirror of https://github.com/SerenityOS/serenity synced 2024-09-06 17:06:31 +00:00
Code Issues Packages Projects Releases Wiki Activity

Kernel: Implement Nagle’s Algorithm

Browse source 
This is an initial implementation, about as basic as intended by the
RFC, and not configurable from userspace at the moment. It should reduce
the amount of low-sized packets sent, reducing overhead and thereby
network traffic.
This commit is contained in:
kleines Filmröllchen 2023-08-17 19:53:49 +02:00 committed by Tim Schumacher
parent ed966a80e2
commit 12e534c8c6
1 changed files with 12 additions and 0 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

12
Kernel/Net/TCPSocket.cpp
View file

@ -210,6 +210,18 @@ ErrorOr<size_t> TCPSocket::protocol_send(UserOrKernelBuffer const& data, size_t
if (routing_decision.is_zero())
return set_so_error(EHOSTUNREACH);
size_t mss = routing_decision.adapter->mtu() - sizeof(IPv4Packet) - sizeof(TCPPacket);
// RFC 896 (Nagle’s algorithm): https://www.ietf.org/rfc/rfc0896
// "The solution is to inhibit the sending of new TCP segments when
// new outgoing data arrives from the user if any previously
// transmitted data on the connection remains unacknowledged. This
// inhibition is to be unconditional; no timers, tests for size of
// data received, or other conditions are required."
// FIXME: Make this configurable via TCP_NODELAY.
auto has_unacked_data = m_unacked_packets.with_shared([&](auto const& packets) { return packets.size > 0; });
if (has_unacked_data && data_length < mss)
return 0;
data_length = min(data_length, mss);
TRY(send_tcp_packet(TCPFlags::PSH | TCPFlags::ACK, &data, data_length, &routing_decision));
return data_length;