Commit Graph

303 Commits

Author SHA1 Message Date
Dario Nieuwenhuis 88256dbdcf Update to Rust 2018. Fixes #271 2020-12-27 17:54:56 +01:00
Dario Nieuwenhuis 4e11070ff4 Async/await waker support. 2020-12-27 17:31:49 +01:00
Dario Nieuwenhuis 29cc59c89d Merge pull request #395 from crawford/clippy
Various cleanup suggested by cargo clippy
2020-12-26 23:53:25 +01:00
Alex Crawford c9f57150a8 Dereference match expressions to clean up patterns
These were flagged by `cargo clippy`:

    warning: you don't need to add `&` to all patterns
2020-12-26 10:32:51 -08:00
Alex Crawford 0c7cdbb4ef Remove unnecessary returns
These were flagged by `cargo clippy`:

    warning: unneeded `return` statement
2020-12-26 10:32:51 -08:00
Alex Crawford c58711e11b Use newline variants of write macro
These were flagged by `cargo clippy`:

    warning: using `println!("")`
    warning: using `write!()` with a format string that ends in a single
             newline
2020-12-26 10:32:42 -08:00
Alex Crawford d217c81e5d Remove explicit calls to as_ref/as_mut
These were flagged by `cargo clippy`:

    warning: this call to `as_ref` does nothing
    warning: this call to `as_mut` does nothing
2020-12-26 10:27:32 -08:00
Dario Nieuwenhuis a7d3f604a3 Don't use matches! macro, for Rust 1.36 support 2020-12-26 03:12:48 +01:00
Dario Nieuwenhuis 67e03b34b6 Fix seq_to_transmit incorrectly returning true when a FIN was enqueued.
If there's data queued that doesn't fit into the remote window, we can't
send the FIN either, so seq_to_transmit should return false.
2020-12-26 03:04:17 +01:00
Dario Nieuwenhuis f89d240c29 tcp: don't send data outside the remote window 2020-12-19 03:23:44 +00:00
Dario Nieuwenhuis 42dc04852b tcp: allow sending ACKs in FinWait2 state. 2020-12-19 01:13:59 +00:00
mustermeiszer 7b4997d5c7 Taking into account TCP and IP header
The MTU consists of TCP header, IP header and the payload.
In the former fix, this was not taken into account.
2020-10-23 15:13:18 +02:00
mustermeiszer d2aa3edbb7 Fixes unused MTU settings at TcpSocket dispatch
This commit fixes a small bug, where the TcpSocket computed the maximum
size per payload in a tx_buffer without taking into account the MTU
defined by the underlying network device.
2020-10-23 14:12:20 +02:00
Ryan Summers 1ae400acaa Return RST to unexpected ACK in SYN-SENT state.
Before this commit, the socket got stuck in an unusable state.
2020-08-17 05:44:58 +00:00
Thales Fragoso 88d01234a8 Match silent_until behavior of NeighborState and Cache 2020-08-17 04:44:58 +00:00
whitequark 548a019991 Update src/socket/meta.rs 2020-08-11 16:26:53 +00:00
Thales Fragoso 24e4868761 Allow for ARP retry during egress 2020-08-10 22:06:06 -03:00
Dario Nieuwenhuis 4cfe8dafdb Do not send window updates in states that shouldn't do so.
Previously, the window update was sent anytime it changed. This is
not desirable for 2 reasons:

- It would send window updates in TimeWait or Closed states, which is incorrect.
- It would send window updates in states where we've already received the
  remote side's FIN, such as CloseWait. This is not necessarily incorrect, but
  is useless, since the remote side is not going to send us more data, therefore
  it's not interested in our window size anymore.
2020-06-25 04:06:29 +00:00
Lucas Zanela d5c44a37d5 Fix overly strict lifetime in TcpSocket. 2020-06-15 12:39:06 +00:00
Dario Nieuwenhuis d1e614516a Remove now-redundant `remote_last_ack` update. 2020-06-12 18:49:33 +00:00
Dario Nieuwenhuis 544b3789d0 Document `Error::Finished` in `recv()`. 2020-06-12 09:43:44 +00:00
Dario Nieuwenhuis a176641192 Return `Error::Finished` in `recv()` on graceful close.
This allows applications to distinguish whether the remote end has
gracefully closed the connection with a FIN (in which case it has
received all the data intact), or the connection has failed due to e.g. a
RST or a timeout (in which case the received data may be truncated).
2020-06-12 09:43:44 +00:00
Dario Nieuwenhuis 0d31e0a134 Always send updated ack number in `ack_reply()`.
Previously, `ack_reply()` was sending the ack number present in `remote_last_ack`,
which is only updated by `dispatch`. This means the ack replies could contain
a wrong, old ack number.

Most of the time this simply caused wasting an extra roundtrip until the socket state
settles down, but in extreme cases it could lead to an infinite loop of packets, such
as in https://github.com/smoltcp-rs/smoltcp/issues/338

Fixes #338
2020-06-11 23:24:24 +00:00
Dario Nieuwenhuis ffcaa83502 Accept data packets in FIN_WAIT_2 state.
Previously, data packets were never accepted in FIN_WAIT_2. The missing
match case was dropping them as "unexpected packet"s.
2020-06-11 20:19:44 +00:00
fdb-hiroshima 09f4a8f551 use provided ip for TcpSocket::connect instead of 0.0.0.0 2020-04-16 07:55:17 +00:00
Pierre Krieger 9c5f77dd7e Implement Hash for SocketHandle
Closes: #320
Approved by: whitequark
2019-12-23 03:53:53 +08:00
Scott Mabin d7916c9a92 Add capacity methods to all sockets. 2019-11-05 23:07:12 +00:00
Richard Meadows 186bc05d77 Allow TCPSockets in a SocketSet without 'static storage lifetime
`TCPSocket` buffers are type `RingBuffer<'a, _>` whilst other
socket types have buffers with type `PacketBuffer<'a, 'b,
_>`.

There is an enum over socket types `pub enum Socket<'a, 'b: 'a>`,
 which is in turn used by `set::Item` and `SocketSet`. The lifetime
 `'b` is not required for `TcpSocket`, but the `Into<Socket>`
 implementation for `TCPSocket` specifies the `'static` lifetime for
 `'b`.

Therefore using `TCPSocket` in a `SocketSet` currently requires the
lifetime `'b` for other sockets to be `'static` which is unnecessarily
restrictive. This patch fixes this.

The lifetime signature of socket specifies `'b: 'a`, that is `'b`
outlives `'a`. Instead of `'static`, this is also satisfied by
`impl<'a> Into<Socket<'a, 'a>> for TcpSocket<'a>` as `'a` "outlives"
`'a` by definition.

The example
[here](https://gist.github.com/richardeoin/b562e79d0d499e00a8bf4944f00594d9)
fails to compile without this patch.

Closes: #304
Approved by: whitequark
2019-08-15 08:31:48 +08:00
Andreas Molzer 8e1fa540c3 Restore Session::finish call in socket ref
Using move semantics allows an Option to keep track of the
initialization state while satisfying the borrow checker. This replaces
the functionality that the 'consumed' flag had before. It also retains
the smaller object size since the Option of a reference can use the
representation of the null pointer, which is an invalid reference, as a
niche for the None variant.

Closes: #301
Approved by: whitequark
2019-07-23 20:13:07 +08:00
whitequark 2902ca00c0 Fix potential unsoundness in SocketRef::into_inner()/drop().
This didn't ever actually cause UB because the drop implementation
should have compiled to a no-op under all conditions; `finish` is
a part of a patchset that never got in. (Even if it didn't, the code
should have been dynamically sound, under all but the most strict
interpretations of Rust's aliasing requirements.)

But it's still a problem, at least because there is a warning for
this potential unsoundness that will become an error in the future.
So everything else aside, we can't use this pattern going forward.

For now, remove the Drop impl, exploiting the fact that `finish`
is a no-op. Going forward this will likely require unsafe code, or
some fundamental change in the way SocketRef is implemented, but
I could not come up with any way to do it. :(
2019-06-22 10:52:39 +00:00
whitequark 563434997c Fix spurious crash in commit bd3912e5. 2019-06-22 09:04:30 +00:00
whitequark be7750486f Bump log dependency to version 0.4. NFC.
This doesn't affect downstream code because log 0.3.9 is a facade
crate implemented in terms of log 0.4, and so log 0.3 and log 0.4
APIs can be used together.
2019-06-22 08:31:11 +00:00
whitequark 59f76f1868 Fix docstring for SocketRef::into_inner(). NFC. 2019-06-22 07:49:27 +00:00
whitequark f32522bef7 Fix docstrings for TcpSocket::{send,recv}_slice(). NFC. 2019-06-22 07:49:24 +00:00
whitequark 4e77a29c91 Clarify TcpSocket::recv_queue() semantics. NFC. 2019-06-22 07:03:15 +00:00
whitequark efdf3a2fb3 Fix typo in docstring. NFC. 2019-06-22 06:50:58 +00:00
Andreas Molzer c9fbd56977 Remove an unused mut qualifier on a variable
Fixes a denied lint uncovered on a recent nightly compiler version,
apparently through improved analysis. The code in question is only
compiled with the proto-ipv6 feature and older compiler versions did not
detect the unused qualifier, sometime around 2019-04-23.

Closes: #291
Approved by: whitequark
2019-04-26 20:01:48 +08:00
Kai Lüke 49e985a66e Implement peek functions for UDP sockets
This needed a peek function for the packet buffer.

Closes: #278
Approved by: whitequark
2019-02-28 00:58:33 +08:00
jhwgh1968 e867832214 Minimal Implementation of TCP Selective Acknowledgement
Closes: #266
Approved by: whitequark
2019-01-01 13:51:02 +08:00
jhwgh1968 88016e675f Fully implement TCP Window Scaling
Closes: #253
Approved by: whitequark
2018-08-20 06:57:44 +08:00
Astro a8f2725784 Implement IGMPv1/v2 processing.
Closes: #178
Approved by: whitequark
2018-08-01 10:26:51 +08:00
whitequark 5bcb358eb5 Rename `new` method on Packet types to `new_checked`.
Fixes #195.

Closes: #254
Approved by: dlrobertson
2018-07-11 12:59:05 +08:00
jhwgh1968 1745df2110 Implement TCP Window Scaling, Phase 1 (Issue #106)
Closes: #232
Approved by: whitequark
2018-06-25 03:08:44 +08:00
Kai Lüke b23df24a2a Only trigger fast retransmit for data to send
This fixes entering a loop when both endpoints
are stuck in retransmission.

https://github.com/m-labs/smoltcp/issues/224

Closes: #233
Approved by: whitequark
2018-06-18 22:19:42 +08:00
Kai Lüke 272baab3ef Reset retrasmission delay after fast retrasmit
Closes: #233
Approved by: whitequark
2018-06-18 22:19:42 +08:00
Dan Robertson 126f144295 Update docs and fix warnings
Closes: #238
Approved by: dlrobertson
2018-06-18 10:49:16 +08:00
Ole Martin Ruud 022cad8b11 Fix implementation of fast retransmit (TcpSocket)
Improve detection of duplicate ACKs by checking that the segment does
not contain data.

Move implementation from the 'reject unacceptable acknowledgements'
to later in the process function, because a duplicate ACK is not an
'unacceptable' acknowledgement but rather a condition to update state.

Implement more tests to verify the operation of the fast retransmit
implementation.

Related issue: #224

Closes: #225
Approved by: whitequark
2018-06-10 02:53:35 +08:00
Astro e5a3de210c Optimize RingBuffer::enqueue_many_with to an empty buffer.
This optimization makes sure that enqueueing to an empty ring buffer
would never wrap around, in turn ensuring that enqueueing to an empty
packet buffer would never use more than one metadata entry.
Right now, pushing buffer-sized packets into a packet buffer requires
at least two metadata entries, which is surprising and wasteful.
2018-05-24 22:00:00 +00:00
Ole Martin Ruud f501198a68 Implement fast retransmit for TcpSocket
Implement a simple way to discover and respond to three duplicate ACKs
from the reciver of the data. This leads to a fast retransmit of the
last unacknowledged seqment. This is feature is described in [RFC 5681].

Closes: #104

Closes: #214
Approved by: whitequark
2018-05-18 00:40:47 +08:00
Ole Martin Ruud 9533e3ac17 Update the docstrings of PollAt
Closes: #214
Approved by: whitequark
2018-05-18 00:40:47 +08:00