206 lines
5.9 KiB
Rust
Raw Normal View History

// Copyright 2018 Parity Technologies (UK) Ltd.
2017-11-22 18:01:28 +01:00
//
// Permission is hereby granted, free of charge, to any person obtaining a
// copy of this software and associated documentation files (the "Software"),
// to deal in the Software without restriction, including without limitation
// the rights to use, copy, modify, merge, publish, distribute, sublicense,
// and/or sell copies of the Software, and to permit persons to whom the
// Software is furnished to do so, subject to the following conditions:
//
// The above copyright notice and this permission notice shall be included in
// all copies or substantial portions of the Software.
//
// THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
// OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
// FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
// AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
// LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
// FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
// DEALINGS IN THE SOFTWARE.
mod codec;
[mplex] Refactoring with Patches (#1769) * Refactor Mplex. Thereby addressing the following issues: * Send a `Reset` frame when open substreams get dropped (313). * Avoid stalls caused by a read operation on one substream reading (and buffering) frames for another substream without notifying the corresponding task. I.e. the tracked read-interest must be scoped to a substream. * Remove dropped substreams from the tracked set of open substreams, to avoid artificially running into substream limits. * Update CHANGELOG. * Refine behaviour of dropping substreams. By taking the substream state into account. The refined behaviour is modeled after the behaviour of Yamux. * Tweak docs and recv buffer retention. * Further small tweaks. * Make the pending frames a FIFO queue. * Take more care to avoid keeping read-wakers around and to notify them when streams close. * Prefer wake over unregister. It is probably safer to always wake pending wakers. * Update muxers/mplex/src/codec.rs Co-authored-by: Max Inden <mail@max-inden.de> * Update muxers/mplex/src/io.rs Co-authored-by: Max Inden <mail@max-inden.de> * Some review feedback and cosmetics. * Update muxers/mplex/src/io.rs Co-authored-by: Max Inden <mail@max-inden.de> * Revise read control flow for clarity. While seemingly duplicating some control flow between `poll_next_strean` and `poll_read_stream`, the individual control flow of each read operation is easier to follow. * CI * Rename Status::Ok to Status::Open. * Rename pending_flush to pending_flush_open. * Finishing touches. * Tweak changelog. Co-authored-by: Max Inden <mail@max-inden.de>
2020-09-28 10:30:49 +02:00
mod config;
mod io;
2017-11-01 11:59:52 +01:00
pub use config::{MaxBufferBehaviour, MplexConfig};
[mplex] Refactoring with Patches (#1769) * Refactor Mplex. Thereby addressing the following issues: * Send a `Reset` frame when open substreams get dropped (313). * Avoid stalls caused by a read operation on one substream reading (and buffering) frames for another substream without notifying the corresponding task. I.e. the tracked read-interest must be scoped to a substream. * Remove dropped substreams from the tracked set of open substreams, to avoid artificially running into substream limits. * Update CHANGELOG. * Refine behaviour of dropping substreams. By taking the substream state into account. The refined behaviour is modeled after the behaviour of Yamux. * Tweak docs and recv buffer retention. * Further small tweaks. * Make the pending frames a FIFO queue. * Take more care to avoid keeping read-wakers around and to notify them when streams close. * Prefer wake over unregister. It is probably safer to always wake pending wakers. * Update muxers/mplex/src/codec.rs Co-authored-by: Max Inden <mail@max-inden.de> * Update muxers/mplex/src/io.rs Co-authored-by: Max Inden <mail@max-inden.de> * Some review feedback and cosmetics. * Update muxers/mplex/src/io.rs Co-authored-by: Max Inden <mail@max-inden.de> * Revise read control flow for clarity. While seemingly duplicating some control flow between `poll_next_strean` and `poll_read_stream`, the individual control flow of each read operation is easier to follow. * CI * Rename Status::Ok to Status::Open. * Rename pending_flush to pending_flush_open. * Finishing touches. * Tweak changelog. Co-authored-by: Max Inden <mail@max-inden.de>
2020-09-28 10:30:49 +02:00
use bytes::Bytes;
use codec::LocalStreamId;
use futures::{future, prelude::*, ready};
use libp2p_core::{
upgrade::{InboundUpgrade, OutboundUpgrade, UpgradeInfo},
Multiaddr, StreamMuxer,
};
use parking_lot::Mutex;
use std::{cmp, iter, pin::Pin, sync::Arc, task::Context, task::Poll};
2017-11-01 11:59:52 +01:00
impl UpgradeInfo for MplexConfig {
type Info = &'static [u8];
type InfoIter = iter::Once<Self::Info>;
fn protocol_info(&self) -> Self::InfoIter {
iter::once(self.protocol_name)
}
}
impl<C> InboundUpgrade<C> for MplexConfig
where
C: AsyncRead + AsyncWrite + Unpin,
{
type Output = Multiplex<C>;
[mplex] Refactoring with Patches (#1769) * Refactor Mplex. Thereby addressing the following issues: * Send a `Reset` frame when open substreams get dropped (313). * Avoid stalls caused by a read operation on one substream reading (and buffering) frames for another substream without notifying the corresponding task. I.e. the tracked read-interest must be scoped to a substream. * Remove dropped substreams from the tracked set of open substreams, to avoid artificially running into substream limits. * Update CHANGELOG. * Refine behaviour of dropping substreams. By taking the substream state into account. The refined behaviour is modeled after the behaviour of Yamux. * Tweak docs and recv buffer retention. * Further small tweaks. * Make the pending frames a FIFO queue. * Take more care to avoid keeping read-wakers around and to notify them when streams close. * Prefer wake over unregister. It is probably safer to always wake pending wakers. * Update muxers/mplex/src/codec.rs Co-authored-by: Max Inden <mail@max-inden.de> * Update muxers/mplex/src/io.rs Co-authored-by: Max Inden <mail@max-inden.de> * Some review feedback and cosmetics. * Update muxers/mplex/src/io.rs Co-authored-by: Max Inden <mail@max-inden.de> * Revise read control flow for clarity. While seemingly duplicating some control flow between `poll_next_strean` and `poll_read_stream`, the individual control flow of each read operation is easier to follow. * CI * Rename Status::Ok to Status::Open. * Rename pending_flush to pending_flush_open. * Finishing touches. * Tweak changelog. Co-authored-by: Max Inden <mail@max-inden.de>
2020-09-28 10:30:49 +02:00
type Error = io::Error;
type Future = future::Ready<Result<Self::Output, io::Error>>;
fn upgrade_inbound(self, socket: C, _: Self::Info) -> Self::Future {
[mplex] Refactoring with Patches (#1769) * Refactor Mplex. Thereby addressing the following issues: * Send a `Reset` frame when open substreams get dropped (313). * Avoid stalls caused by a read operation on one substream reading (and buffering) frames for another substream without notifying the corresponding task. I.e. the tracked read-interest must be scoped to a substream. * Remove dropped substreams from the tracked set of open substreams, to avoid artificially running into substream limits. * Update CHANGELOG. * Refine behaviour of dropping substreams. By taking the substream state into account. The refined behaviour is modeled after the behaviour of Yamux. * Tweak docs and recv buffer retention. * Further small tweaks. * Make the pending frames a FIFO queue. * Take more care to avoid keeping read-wakers around and to notify them when streams close. * Prefer wake over unregister. It is probably safer to always wake pending wakers. * Update muxers/mplex/src/codec.rs Co-authored-by: Max Inden <mail@max-inden.de> * Update muxers/mplex/src/io.rs Co-authored-by: Max Inden <mail@max-inden.de> * Some review feedback and cosmetics. * Update muxers/mplex/src/io.rs Co-authored-by: Max Inden <mail@max-inden.de> * Revise read control flow for clarity. While seemingly duplicating some control flow between `poll_next_strean` and `poll_read_stream`, the individual control flow of each read operation is easier to follow. * CI * Rename Status::Ok to Status::Open. * Rename pending_flush to pending_flush_open. * Finishing touches. * Tweak changelog. Co-authored-by: Max Inden <mail@max-inden.de>
2020-09-28 10:30:49 +02:00
future::ready(Ok(Multiplex {
io: Arc::new(Mutex::new(io::Multiplexed::new(socket, self))),
[mplex] Refactoring with Patches (#1769) * Refactor Mplex. Thereby addressing the following issues: * Send a `Reset` frame when open substreams get dropped (313). * Avoid stalls caused by a read operation on one substream reading (and buffering) frames for another substream without notifying the corresponding task. I.e. the tracked read-interest must be scoped to a substream. * Remove dropped substreams from the tracked set of open substreams, to avoid artificially running into substream limits. * Update CHANGELOG. * Refine behaviour of dropping substreams. By taking the substream state into account. The refined behaviour is modeled after the behaviour of Yamux. * Tweak docs and recv buffer retention. * Further small tweaks. * Make the pending frames a FIFO queue. * Take more care to avoid keeping read-wakers around and to notify them when streams close. * Prefer wake over unregister. It is probably safer to always wake pending wakers. * Update muxers/mplex/src/codec.rs Co-authored-by: Max Inden <mail@max-inden.de> * Update muxers/mplex/src/io.rs Co-authored-by: Max Inden <mail@max-inden.de> * Some review feedback and cosmetics. * Update muxers/mplex/src/io.rs Co-authored-by: Max Inden <mail@max-inden.de> * Revise read control flow for clarity. While seemingly duplicating some control flow between `poll_next_strean` and `poll_read_stream`, the individual control flow of each read operation is easier to follow. * CI * Rename Status::Ok to Status::Open. * Rename pending_flush to pending_flush_open. * Finishing touches. * Tweak changelog. Co-authored-by: Max Inden <mail@max-inden.de>
2020-09-28 10:30:49 +02:00
}))
2017-11-01 11:59:52 +01:00
}
}
2017-11-01 11:59:52 +01:00
impl<C> OutboundUpgrade<C> for MplexConfig
where
C: AsyncRead + AsyncWrite + Unpin,
{
type Output = Multiplex<C>;
[mplex] Refactoring with Patches (#1769) * Refactor Mplex. Thereby addressing the following issues: * Send a `Reset` frame when open substreams get dropped (313). * Avoid stalls caused by a read operation on one substream reading (and buffering) frames for another substream without notifying the corresponding task. I.e. the tracked read-interest must be scoped to a substream. * Remove dropped substreams from the tracked set of open substreams, to avoid artificially running into substream limits. * Update CHANGELOG. * Refine behaviour of dropping substreams. By taking the substream state into account. The refined behaviour is modeled after the behaviour of Yamux. * Tweak docs and recv buffer retention. * Further small tweaks. * Make the pending frames a FIFO queue. * Take more care to avoid keeping read-wakers around and to notify them when streams close. * Prefer wake over unregister. It is probably safer to always wake pending wakers. * Update muxers/mplex/src/codec.rs Co-authored-by: Max Inden <mail@max-inden.de> * Update muxers/mplex/src/io.rs Co-authored-by: Max Inden <mail@max-inden.de> * Some review feedback and cosmetics. * Update muxers/mplex/src/io.rs Co-authored-by: Max Inden <mail@max-inden.de> * Revise read control flow for clarity. While seemingly duplicating some control flow between `poll_next_strean` and `poll_read_stream`, the individual control flow of each read operation is easier to follow. * CI * Rename Status::Ok to Status::Open. * Rename pending_flush to pending_flush_open. * Finishing touches. * Tweak changelog. Co-authored-by: Max Inden <mail@max-inden.de>
2020-09-28 10:30:49 +02:00
type Error = io::Error;
type Future = future::Ready<Result<Self::Output, io::Error>>;
fn upgrade_outbound(self, socket: C, _: Self::Info) -> Self::Future {
[mplex] Refactoring with Patches (#1769) * Refactor Mplex. Thereby addressing the following issues: * Send a `Reset` frame when open substreams get dropped (313). * Avoid stalls caused by a read operation on one substream reading (and buffering) frames for another substream without notifying the corresponding task. I.e. the tracked read-interest must be scoped to a substream. * Remove dropped substreams from the tracked set of open substreams, to avoid artificially running into substream limits. * Update CHANGELOG. * Refine behaviour of dropping substreams. By taking the substream state into account. The refined behaviour is modeled after the behaviour of Yamux. * Tweak docs and recv buffer retention. * Further small tweaks. * Make the pending frames a FIFO queue. * Take more care to avoid keeping read-wakers around and to notify them when streams close. * Prefer wake over unregister. It is probably safer to always wake pending wakers. * Update muxers/mplex/src/codec.rs Co-authored-by: Max Inden <mail@max-inden.de> * Update muxers/mplex/src/io.rs Co-authored-by: Max Inden <mail@max-inden.de> * Some review feedback and cosmetics. * Update muxers/mplex/src/io.rs Co-authored-by: Max Inden <mail@max-inden.de> * Revise read control flow for clarity. While seemingly duplicating some control flow between `poll_next_strean` and `poll_read_stream`, the individual control flow of each read operation is easier to follow. * CI * Rename Status::Ok to Status::Open. * Rename pending_flush to pending_flush_open. * Finishing touches. * Tweak changelog. Co-authored-by: Max Inden <mail@max-inden.de>
2020-09-28 10:30:49 +02:00
future::ready(Ok(Multiplex {
io: Arc::new(Mutex::new(io::Multiplexed::new(socket, self))),
[mplex] Refactoring with Patches (#1769) * Refactor Mplex. Thereby addressing the following issues: * Send a `Reset` frame when open substreams get dropped (313). * Avoid stalls caused by a read operation on one substream reading (and buffering) frames for another substream without notifying the corresponding task. I.e. the tracked read-interest must be scoped to a substream. * Remove dropped substreams from the tracked set of open substreams, to avoid artificially running into substream limits. * Update CHANGELOG. * Refine behaviour of dropping substreams. By taking the substream state into account. The refined behaviour is modeled after the behaviour of Yamux. * Tweak docs and recv buffer retention. * Further small tweaks. * Make the pending frames a FIFO queue. * Take more care to avoid keeping read-wakers around and to notify them when streams close. * Prefer wake over unregister. It is probably safer to always wake pending wakers. * Update muxers/mplex/src/codec.rs Co-authored-by: Max Inden <mail@max-inden.de> * Update muxers/mplex/src/io.rs Co-authored-by: Max Inden <mail@max-inden.de> * Some review feedback and cosmetics. * Update muxers/mplex/src/io.rs Co-authored-by: Max Inden <mail@max-inden.de> * Revise read control flow for clarity. While seemingly duplicating some control flow between `poll_next_strean` and `poll_read_stream`, the individual control flow of each read operation is easier to follow. * CI * Rename Status::Ok to Status::Open. * Rename pending_flush to pending_flush_open. * Finishing touches. * Tweak changelog. Co-authored-by: Max Inden <mail@max-inden.de>
2020-09-28 10:30:49 +02:00
}))
2017-11-01 11:59:52 +01:00
}
}
/// Multiplexer. Implements the `StreamMuxer` trait.
pub struct Multiplex<C> {
io: Arc<Mutex<io::Multiplexed<C>>>,
2017-11-01 11:59:52 +01:00
}
[mplex] Refactoring with Patches (#1769) * Refactor Mplex. Thereby addressing the following issues: * Send a `Reset` frame when open substreams get dropped (313). * Avoid stalls caused by a read operation on one substream reading (and buffering) frames for another substream without notifying the corresponding task. I.e. the tracked read-interest must be scoped to a substream. * Remove dropped substreams from the tracked set of open substreams, to avoid artificially running into substream limits. * Update CHANGELOG. * Refine behaviour of dropping substreams. By taking the substream state into account. The refined behaviour is modeled after the behaviour of Yamux. * Tweak docs and recv buffer retention. * Further small tweaks. * Make the pending frames a FIFO queue. * Take more care to avoid keeping read-wakers around and to notify them when streams close. * Prefer wake over unregister. It is probably safer to always wake pending wakers. * Update muxers/mplex/src/codec.rs Co-authored-by: Max Inden <mail@max-inden.de> * Update muxers/mplex/src/io.rs Co-authored-by: Max Inden <mail@max-inden.de> * Some review feedback and cosmetics. * Update muxers/mplex/src/io.rs Co-authored-by: Max Inden <mail@max-inden.de> * Revise read control flow for clarity. While seemingly duplicating some control flow between `poll_next_strean` and `poll_read_stream`, the individual control flow of each read operation is easier to follow. * CI * Rename Status::Ok to Status::Open. * Rename pending_flush to pending_flush_open. * Finishing touches. * Tweak changelog. Co-authored-by: Max Inden <mail@max-inden.de>
2020-09-28 10:30:49 +02:00
impl<C> StreamMuxer for Multiplex<C>
where
C: AsyncRead + AsyncWrite + Unpin,
2017-11-22 18:01:28 +01:00
{
type Substream = Substream<C>;
[mplex] Refactoring with Patches (#1769) * Refactor Mplex. Thereby addressing the following issues: * Send a `Reset` frame when open substreams get dropped (313). * Avoid stalls caused by a read operation on one substream reading (and buffering) frames for another substream without notifying the corresponding task. I.e. the tracked read-interest must be scoped to a substream. * Remove dropped substreams from the tracked set of open substreams, to avoid artificially running into substream limits. * Update CHANGELOG. * Refine behaviour of dropping substreams. By taking the substream state into account. The refined behaviour is modeled after the behaviour of Yamux. * Tweak docs and recv buffer retention. * Further small tweaks. * Make the pending frames a FIFO queue. * Take more care to avoid keeping read-wakers around and to notify them when streams close. * Prefer wake over unregister. It is probably safer to always wake pending wakers. * Update muxers/mplex/src/codec.rs Co-authored-by: Max Inden <mail@max-inden.de> * Update muxers/mplex/src/io.rs Co-authored-by: Max Inden <mail@max-inden.de> * Some review feedback and cosmetics. * Update muxers/mplex/src/io.rs Co-authored-by: Max Inden <mail@max-inden.de> * Revise read control flow for clarity. While seemingly duplicating some control flow between `poll_next_strean` and `poll_read_stream`, the individual control flow of each read operation is easier to follow. * CI * Rename Status::Ok to Status::Open. * Rename pending_flush to pending_flush_open. * Finishing touches. * Tweak changelog. Co-authored-by: Max Inden <mail@max-inden.de>
2020-09-28 10:30:49 +02:00
type Error = io::Error;
fn poll_inbound(&self, cx: &mut Context<'_>) -> Poll<Result<Self::Substream, Self::Error>> {
self.io
.lock()
.poll_next_stream(cx)
.map_ok(|stream_id| Substream::new(stream_id, self.io.clone()))
2017-11-22 18:01:28 +01:00
}
fn poll_outbound(&self, cx: &mut Context<'_>) -> Poll<Result<Self::Substream, Self::Error>> {
self.io
.lock()
.poll_open_stream(cx)
.map_ok(|stream_id| Substream::new(stream_id, self.io.clone()))
}
fn poll_address_change(&self, _: &mut Context<'_>) -> Poll<Result<Multiaddr, Self::Error>> {
Poll::Pending
}
fn poll_close(&self, cx: &mut Context<'_>) -> Poll<Result<(), io::Error>> {
[mplex] Refactoring with Patches (#1769) * Refactor Mplex. Thereby addressing the following issues: * Send a `Reset` frame when open substreams get dropped (313). * Avoid stalls caused by a read operation on one substream reading (and buffering) frames for another substream without notifying the corresponding task. I.e. the tracked read-interest must be scoped to a substream. * Remove dropped substreams from the tracked set of open substreams, to avoid artificially running into substream limits. * Update CHANGELOG. * Refine behaviour of dropping substreams. By taking the substream state into account. The refined behaviour is modeled after the behaviour of Yamux. * Tweak docs and recv buffer retention. * Further small tweaks. * Make the pending frames a FIFO queue. * Take more care to avoid keeping read-wakers around and to notify them when streams close. * Prefer wake over unregister. It is probably safer to always wake pending wakers. * Update muxers/mplex/src/codec.rs Co-authored-by: Max Inden <mail@max-inden.de> * Update muxers/mplex/src/io.rs Co-authored-by: Max Inden <mail@max-inden.de> * Some review feedback and cosmetics. * Update muxers/mplex/src/io.rs Co-authored-by: Max Inden <mail@max-inden.de> * Revise read control flow for clarity. While seemingly duplicating some control flow between `poll_next_strean` and `poll_read_stream`, the individual control flow of each read operation is easier to follow. * CI * Rename Status::Ok to Status::Open. * Rename pending_flush to pending_flush_open. * Finishing touches. * Tweak changelog. Co-authored-by: Max Inden <mail@max-inden.de>
2020-09-28 10:30:49 +02:00
self.io.lock().poll_close(cx)
}
2017-11-01 11:59:52 +01:00
}
impl<C> AsyncRead for Substream<C>
where
C: AsyncRead + AsyncWrite + Unpin,
{
fn poll_read(
self: Pin<&mut Self>,
cx: &mut Context<'_>,
buf: &mut [u8],
) -> Poll<io::Result<usize>> {
let this = self.get_mut();
loop {
// Try to read from the current (i.e. last received) frame.
if !this.current_data.is_empty() {
let len = cmp::min(this.current_data.len(), buf.len());
buf[..len].copy_from_slice(&this.current_data.split_to(len));
return Poll::Ready(Ok(len));
}
// Read the next data frame from the multiplexed stream.
match ready!(this.io.lock().poll_read_stream(cx, this.id))? {
Some(data) => {
this.current_data = data;
}
None => return Poll::Ready(Ok(0)),
}
}
}
}
impl<C> AsyncWrite for Substream<C>
where
C: AsyncRead + AsyncWrite + Unpin,
{
fn poll_write(
self: Pin<&mut Self>,
cx: &mut Context<'_>,
buf: &[u8],
) -> Poll<io::Result<usize>> {
let this = self.get_mut();
this.io.lock().poll_write_stream(cx, this.id, buf)
}
fn poll_flush(self: Pin<&mut Self>, cx: &mut Context<'_>) -> Poll<io::Result<()>> {
let this = self.get_mut();
this.io.lock().poll_flush_stream(cx, this.id)
}
fn poll_close(self: Pin<&mut Self>, cx: &mut Context<'_>) -> Poll<io::Result<()>> {
let this = self.get_mut();
let mut io = this.io.lock();
ready!(io.poll_close_stream(cx, this.id))?;
ready!(io.poll_flush_stream(cx, this.id))?;
Poll::Ready(Ok(()))
}
}
/// Active substream to the remote.
pub struct Substream<C>
where
C: AsyncRead + AsyncWrite + Unpin,
{
[mplex] Refactoring with Patches (#1769) * Refactor Mplex. Thereby addressing the following issues: * Send a `Reset` frame when open substreams get dropped (313). * Avoid stalls caused by a read operation on one substream reading (and buffering) frames for another substream without notifying the corresponding task. I.e. the tracked read-interest must be scoped to a substream. * Remove dropped substreams from the tracked set of open substreams, to avoid artificially running into substream limits. * Update CHANGELOG. * Refine behaviour of dropping substreams. By taking the substream state into account. The refined behaviour is modeled after the behaviour of Yamux. * Tweak docs and recv buffer retention. * Further small tweaks. * Make the pending frames a FIFO queue. * Take more care to avoid keeping read-wakers around and to notify them when streams close. * Prefer wake over unregister. It is probably safer to always wake pending wakers. * Update muxers/mplex/src/codec.rs Co-authored-by: Max Inden <mail@max-inden.de> * Update muxers/mplex/src/io.rs Co-authored-by: Max Inden <mail@max-inden.de> * Some review feedback and cosmetics. * Update muxers/mplex/src/io.rs Co-authored-by: Max Inden <mail@max-inden.de> * Revise read control flow for clarity. While seemingly duplicating some control flow between `poll_next_strean` and `poll_read_stream`, the individual control flow of each read operation is easier to follow. * CI * Rename Status::Ok to Status::Open. * Rename pending_flush to pending_flush_open. * Finishing touches. * Tweak changelog. Co-authored-by: Max Inden <mail@max-inden.de>
2020-09-28 10:30:49 +02:00
/// The unique, local identifier of the substream.
id: LocalStreamId,
/// The current data frame the substream is reading from.
current_data: Bytes,
/// Shared reference to the actual muxer.
io: Arc<Mutex<io::Multiplexed<C>>>,
[mplex] Refactoring with Patches (#1769) * Refactor Mplex. Thereby addressing the following issues: * Send a `Reset` frame when open substreams get dropped (313). * Avoid stalls caused by a read operation on one substream reading (and buffering) frames for another substream without notifying the corresponding task. I.e. the tracked read-interest must be scoped to a substream. * Remove dropped substreams from the tracked set of open substreams, to avoid artificially running into substream limits. * Update CHANGELOG. * Refine behaviour of dropping substreams. By taking the substream state into account. The refined behaviour is modeled after the behaviour of Yamux. * Tweak docs and recv buffer retention. * Further small tweaks. * Make the pending frames a FIFO queue. * Take more care to avoid keeping read-wakers around and to notify them when streams close. * Prefer wake over unregister. It is probably safer to always wake pending wakers. * Update muxers/mplex/src/codec.rs Co-authored-by: Max Inden <mail@max-inden.de> * Update muxers/mplex/src/io.rs Co-authored-by: Max Inden <mail@max-inden.de> * Some review feedback and cosmetics. * Update muxers/mplex/src/io.rs Co-authored-by: Max Inden <mail@max-inden.de> * Revise read control flow for clarity. While seemingly duplicating some control flow between `poll_next_strean` and `poll_read_stream`, the individual control flow of each read operation is easier to follow. * CI * Rename Status::Ok to Status::Open. * Rename pending_flush to pending_flush_open. * Finishing touches. * Tweak changelog. Co-authored-by: Max Inden <mail@max-inden.de>
2020-09-28 10:30:49 +02:00
}
impl<C> Substream<C>
where
C: AsyncRead + AsyncWrite + Unpin,
{
fn new(id: LocalStreamId, io: Arc<Mutex<io::Multiplexed<C>>>) -> Self {
Self {
id,
current_data: Bytes::new(),
io,
}
[mplex] Refactoring with Patches (#1769) * Refactor Mplex. Thereby addressing the following issues: * Send a `Reset` frame when open substreams get dropped (313). * Avoid stalls caused by a read operation on one substream reading (and buffering) frames for another substream without notifying the corresponding task. I.e. the tracked read-interest must be scoped to a substream. * Remove dropped substreams from the tracked set of open substreams, to avoid artificially running into substream limits. * Update CHANGELOG. * Refine behaviour of dropping substreams. By taking the substream state into account. The refined behaviour is modeled after the behaviour of Yamux. * Tweak docs and recv buffer retention. * Further small tweaks. * Make the pending frames a FIFO queue. * Take more care to avoid keeping read-wakers around and to notify them when streams close. * Prefer wake over unregister. It is probably safer to always wake pending wakers. * Update muxers/mplex/src/codec.rs Co-authored-by: Max Inden <mail@max-inden.de> * Update muxers/mplex/src/io.rs Co-authored-by: Max Inden <mail@max-inden.de> * Some review feedback and cosmetics. * Update muxers/mplex/src/io.rs Co-authored-by: Max Inden <mail@max-inden.de> * Revise read control flow for clarity. While seemingly duplicating some control flow between `poll_next_strean` and `poll_read_stream`, the individual control flow of each read operation is easier to follow. * CI * Rename Status::Ok to Status::Open. * Rename pending_flush to pending_flush_open. * Finishing touches. * Tweak changelog. Co-authored-by: Max Inden <mail@max-inden.de>
2020-09-28 10:30:49 +02:00
}
}
impl<C> Drop for Substream<C>
where
C: AsyncRead + AsyncWrite + Unpin,
{
fn drop(&mut self) {
self.io.lock().drop_stream(self.id);
}
}