Commit Graph

54 Commits

Author SHA1 Message Date
Sean McArthur
3cbc158210 send reset CANCEL when SendStream is dropped with no end-of-stream sent (#210) 2018-01-04 11:06:06 -08:00
Sean McArthur
1ea9a8fc7e ignore received frames on a stream locally reset for some time (#174)
- Adds config duration for how long to ignore frames on a reset stream
- Adds config for how many reset streams can be held at a time
2017-12-18 11:09:38 -08:00
Carl Lerche
9378846da8 Client should validate request URI. (#181)
This patch adds checks for the request URI and rejects invalid URIs. In
the case of forwarding an HTTP 1.1 request with a path, an "http" pseudo
header is added to satisfy the HTTP/2.0 spec.

Closes #179
2017-12-11 13:42:00 -06:00
Sean McArthur
79003d0d45 reject connection-specific headers (#173)
- When receiving, return a PROTOCOL_ERROR.
- When sending, return a user error about malformed headers.

Closes #36
2017-11-14 11:16:29 -08:00
Carl Lerche
c4fc2928fe API cleanup (#155)
* Change send_reset to take &mut self.

While calling this function is the last thing that should be done with
the instance, the intent of the h2 library is not to be used directly by
users, but to be used as an implementation detail by other libraries.

Requiring `self` on `send_reset` is pretty annoying when calling the
function from inside a `Future` implementation. Also, all the other fns
on the type take `&mut self`.

* Remove the P: Peer generic from internals

* Split out `Respond` from `server::Stream`

This new type is used to send HTTP responses to the client as well as
reserve streams for push promises.

* Remove unused `Send` helper.

This could be brought back later when the API becomes stable.

* Unite `client` and `server` types

* Remove `B` generic from internal proto structs

This is a first step in removing the `B` generic from public API types
that do not strictly require it.

Currently, all public API types must be generic over `B` even if they do
not actually interact with the send data frame type. The first step in
removing this is to remove `B` as a generic on all internal types.

* Remove `Buffer<B>` from inner stream state

This is the next step in removing the `B` generic from all public API
types. The send buffer is the only type that requires `B`. It has now
been extracted from the rest of the stream state.

The strategy used in this PR requires an additional `Arc` and `Mutex`,
but this is not a fundamental requirement. The additional overhead can
be avoided with a little bit of unsafe code. However, this optimization
should not be made until it is proven that it is required.

* Remove `B` generic from `Body` + `ReleaseCapacity`

This commit actually removes the generic from these two public API
types. Also note, that removing the generic requires that `B: 'static`.
This is because there is no more generic on `Body` and `ReleaseCapacity`
and the compiler must be able to ensure that `B` outlives all `Body` and
`ReleaseCapacity` handles.

In practice, in an async world, passing a non 'static `B` is never going
to happen.

* Remove generic from `ResponseFuture`

This change also makes generic free types `Send`. The original strategy
of using a trait object meant that those handles could not be `Send`.
The solution was to avoid using the send buffer when canceling a stream.
This is done by transitioning the stream state to `Canceled`, a new
`Cause` variant.

* Simplify Send::send_reset

Now that implicit cancelation goes through a separate path, the
send_reset function can be simplified.

* Export types common to client & server at root

* Rename Stream -> SendStream, Body -> RecvStream

* Implement send_reset on server::Respond
2017-10-19 20:02:08 -07:00
Carl Lerche
7c287af0d0 Fix some flow control bugs. (#152)
* Release stream capacity back to the connection to avoid capacity
leaks.
* Actually notify waiting tasks when capacity becomes available.
2017-10-13 14:15:20 -07:00
Sean McArthur
5c1bde7d62 add set_target_window_size methods to Server and Client (#149)
Closes #101
2017-10-13 11:19:56 -07:00
Holt Chesley
2aee78c7d7 Issue 128: Convert frame::Reason to struct (#142)
Alter frame::Reason to a struct with a single u32 member.
Introduce Constants to the impl for existing Reasons. Change all usage
in the library and its tests to adopt this change,
using the new constants.
2017-10-08 13:13:07 -07:00
Eliza Weisman
2e3dcf602c StreamRef sends RST_STREAM on drop (#109)
This PR modifies the `Drop` implementation for `StreamRef` to reset the underlying stream if it is the last reference to that stream. Since both `Stream` and `Body` are internally just a `StreamRef`, this means they will both reset the stream on drop; thus, this closes #100.

The assertion that the store no longer contains the dropped stream ID at the end of the `Drop` method  had to be removed, as the stream has to be reset from inside of a `transition` block (which now manages releasing that ID for us), and the `transition` closure moves the value of `stream`, making the assertion no longer possible.

Modifications to some of the tests in `flow_control.rs` were also necessary, in order to prevent `StreamRef`s from being dropped too early.
2017-10-05 18:05:18 -05:00
Sean McArthur
ecd2764f4b when receiving a GOAWAY, allow earlier streams to still process (#133)
Once all active streams have finished, send a GOAWAY back and close the
connection.
2017-10-05 15:32:13 -07:00
Sean McArthur
f8efb053b9 split Client into (Client, Connection) (#107)
The Connection type is a `Future` that drives all of the IO of the
client connection.

The Client type is separate, and is used to send requests into the
connection.
2017-09-28 16:55:12 -07:00
Sean McArthur
db8c109817 Client::poll_ready() returns an Error if next stream ID would overflow (#103)
Closes #102
2017-09-19 14:16:32 -07:00
Sean McArthur
0c8bd75224 check for StreamId overflow (#68) 2017-09-19 13:10:48 -07:00
Sean McArthur
a8a4cd2be1 add Client config to disable server push
- Adds `Client::builder().enable_push(false)` to disable push
- Client sends a GO_AWAY if receiving a push when it's disabled
2017-09-18 10:49:35 -07:00
Sean McArthur
f7d14861e5 rustfmt: add trailing commas in match arms, set fn call to block stle (#85) 2017-09-12 19:29:06 -07:00
Oliver Gould
897bf84163 Use rustfmt to enforce consistent formatting
This change adds a .rustfmt.toml that includes ALL supported settings,
12 of which we have overridden to attempt to cater to our own
proclivities.

rustfmt is checked in the rust-nightly CI job.
2017-09-12 22:29:35 +00:00
Carl Lerche
5c0efcf8c4 Ref count stream state and release when final (#73)
Previously, stream state was never released so that long-lived connections
leaked memory.

Now, stream states are reference-counted and freed from the stream slab
when complete.  Locally reset streams are retained so that received frames
may be ignored.
2017-09-10 16:01:19 -07:00
Carl Lerche
ae6bad6cef Fix some flow control bugs
When a stream is closed, connection level capacity should be released
back to the connection and then assigned to other streams waiting for
capacity.
2017-09-07 16:45:20 -07:00
Carl Lerche
c122e97127 Refactor errors (#46)
This patch does a bunch of refactoring, mostly around error types, but it also
paves the way to allow `Codec` to be used standalone.

* `Codec` (and `FramedRead` / `FramedWrite`) is broken out into a codec module.
* An h2-codec crate is created that re-exports the frame and codec modules.
* New error types are introduced in the internals:
  * `RecvError` represents errors caused by trying to receive a frame.
  * `SendError` represents errors caused by trying to send a frame.
  * `UserError` is an enum of potential errors caused by invalid usage
    by the user of the lib.
  * `ProtoError` is either a `Reason` or an `io::Error`. However it doesn't
    specify connection or stream level.
  * `h2::Error` is an opaque error type and is the only error type exposed
    by the public API (used to be `ConnectionError`).

There are misc code changes to enable this as well. The biggest is a new "sink"
API for `Codec`. It provides buffer which queues up a frame followed by flush
which writes everything that is queued. This departs from the `Sink` trait in
order to provide more accurate error values. For example, buffer can never fail
(but it will panic if `poll_ready` is not called first).
2017-09-02 11:12:50 -07:00
Carl Lerche
9bb34d907a Thread P generic through all 2017-08-30 18:16:21 -04:00
Carl Lerche
14f35f1be6 Handle malformed HEADERS 2017-08-30 18:16:21 -04:00
Carl Lerche
91aa1db2ff Misc protocol fixes
* Verify contiuation frame stream ID
* Fix sending RST_STREAM in certain cases.
2017-08-30 18:16:21 -04:00
Carl Lerche
0c8a94aa11 Fix send flow control bug
The send stream state is transitioned before data is buffered. As such,
the stream state could be closed while there is still data to be sent.
2017-08-30 18:16:21 -04:00
Carl Lerche
11d5f95236 Wire in trailers (#34)
Add send and receive trailer support.
2017-08-25 10:20:47 -07:00
Carl Lerche
b0e6867877 Fix warnings 2017-08-24 15:52:01 -07:00
Carl Lerche
6a6c9665cd Immediately apply initial window size to streams
The initial window size should be applied to streams once they leave the
IDLE state.
2017-08-24 11:03:33 -07:00
Carl Lerche
66dbde92ef Do not reuse next ptr for multiple linked lists
Because, you might think that each linked list has exclusive access to
the next pointer, but then there is an edge case that proves otherwise.
Also, debugging this kind of thing is annoying.
2017-08-23 20:35:53 -07:00
Oliver Gould
f839443ece implement h2::server::Stream::send_reset(Reason) and Body::is_empty() (#22) 2017-08-23 12:48:00 -07:00
Carl Lerche
807d2b7317 Wire in recv flow control (#26) 2017-08-23 11:22:24 -07:00
Carl Lerche
a623ab68b5 New send flow control (#25)
Restructure send flow control such that sending data is always accepted by `Stream`. Data frames will be buffered until there is available window to send them. Producers can monitor the available window capacity to decide if data should be produced.
2017-08-21 13:52:58 -07:00
Oliver Gould
e015d7bfba Implement Client::poll_ready (#21)
Client::poll_ready ensures that the connection is ale to to initiate a new request stream to the remote server.  When the server is at capacity, a task is stored to be notified when capacity is available.
2017-08-15 13:46:55 -07:00
Carl Lerche
8a15663ed2 Progress towards allowing large writes 2017-08-11 16:57:51 -07:00
Carl Lerche
32d4c2d5a9 Many more changes 2017-08-11 12:00:22 -07:00
Carl Lerche
c439232ed2 Handle more H2 details 2017-08-10 21:08:57 -07:00
Carl Lerche
e7c07b0b12 More debug output 2017-08-10 13:34:04 -07:00
Carl Lerche
a1b03b7724 Trying to get something working 2017-08-10 13:15:40 -07:00
Carl Lerche
7107e9fc06 More work on window updates 2017-08-09 22:34:22 -07:00
Carl Lerche
c118f86517 More flow control work 2017-08-09 16:42:55 -07:00
Carl Lerche
95bb95af01 More send flow control 2017-08-09 14:37:41 -07:00
Carl Lerche
a8c8cdb8e9 Fix build 2017-08-09 14:17:55 -07:00
Carl Lerche
dfec401fdf WIP: send flow control 2017-08-09 14:16:32 -07:00
Carl Lerche
26df3a3698 Remove P generic from type 2017-08-08 22:11:11 -07:00
Carl Lerche
314b7a1848 Wire in PushPromise 2017-08-08 13:32:36 -07:00
Carl Lerche
71da8d121f Start hooking up sending data 2017-08-07 21:01:15 -07:00
Carl Lerche
fc0a7eb898 More work 2017-08-04 12:12:22 -07:00
Carl Lerche
74b3852a58 Start working on prioritization 2017-08-03 22:44:19 -07:00
Carl Lerche
dd8412d660 Much work 2017-08-03 15:50:13 -07:00
Carl Lerche
7a804601c5 A lot of structural work 2017-08-03 11:40:50 -07:00
Carl Lerche
9f9bf85168 More restructuring 2017-08-02 14:48:10 -07:00
Carl Lerche
9d7221e6cf Misc streams cleanup 2017-08-02 13:09:14 -07:00