To make our FFI layer between Android and Rust safer, we adopt the
UniFFI tool from Mozilla. UniFFI allows us to create a dedicated crate
(here `client-ffi`) that contains Rust structs annotated with various
attributes. These macros then generate code at compile time that is
built into the shared object. Using a dedicated CLI from the UniFFI
project, we can then generate Kotlin bindings from this shared object.
The primary motivation for this effort is memory safety across the FFI
boundary. Most importantly, we want to ensure that:
- The session pointer is not used after it has been free'd
- Disconnecting the session frees the pointer
- Freeing the session does not happen as part of a callback as that
triggers a cyclic dependency on the Rust side (callbacks are executed on
a runtime and that runtime is dropped as part of dropping the session)
To achieve all of these goals, we move away from callbacks altogether.
UniFFI has great support for async functions. We leverage this support
to expose a `suspend fn` to Android that returns `Event`s. These events
map to the current callback functions. Internally, these events are read
from a channel with a capacity of 1000 events. It is therefore not very
time-critical that the app reads from this channel. `connlib` will
happily continue even if the channel is full. 1000 events should be more
than sufficient though in case the host app cannot immediately process
them. We don't send events very often after all.
This event-based design has major advantages: It allows us to make use
of `AutoCloseable` on the Kotlin side, meaning the `session` pointer is
only ever accessed as part of a `use` block and automatically closed
(and therefore free'd) at the end of the block.
To communicate with the session, we introduce a `TunnelCommand` which
represents all actions that the host app can send to `connlib`. These
are passed through a channel to the `suspend fn` which continuously
listens for events and commands.
Resolves: #9499
Related: #3959
---------
Signed-off-by: Thomas Eizinger <thomas@eizinger.io>
Co-authored-by: Jamil Bou Kheir <jamilbk@users.noreply.github.com>
These don't happen very often so are safe to log on INFO. That is the
default log level and it is useful to see, why we are re-connecting to
the portal.
As relict from very early designs of `connlib`, the `Callbacks` trait is
still present and defines how the host app receives events from a
running `Session`. Callbacks are not a great design pattern however
because they force the running code, i.e. `connlib`s event-loop to
execute unknown code. For example, if that code panics, all of `connlib`
is taken down. Additionally, not all consumers may want to receive
events via callbacks. The GUI and headless client for example already
have their own event-loop in which they process all kinds of things.
Having to deal with the `Callbacks` interface introduces an odd
indirection here.
To fix this, we instead return an `EventStream` when constructing a
`Session`. This essentially aligns the API of `Session` with that of a
channel. You receive two handles, one for sending in commands and one
for receiving events. A `Session` will automatically spawn itself onto
the given runtime so progress is made even if one does not poll on these
channel handles.
This greatly simplifies the code:
- We get to delete the `Callbacks` interface.
- We can delete the threaded callback adapter. This was only necessary
because we didn't want to block `connlib` with the handling of the
event. By using a channel for events, this is automatically guaranteed.
- The GUI and headless client can directly integrate the event handling
in their event-loop, without having to create an indirection with a
channel.
- It is now clear that only the Apple and Android FFI layers actually
use callbacks to communicate these events.
- We net-delete 100 LoC
The name IPC service is not very descriptive. By nature of being
separate processes, we need to use IPC to communicate between them. The
important thing is that the service process has control over the tunnel.
Therefore, we rename everything to "Tunnel service".
The only part that is not changed are historic changelog entries.
Resolves: #9048
The current `rust/` directory is a bit of a wild-west in terms of how
the crates are organised. Most of them are simply at the top-level when
in reality, they are all `connlib`-related. The Apple and Android FFI
crates - which are entrypoints in the Rust code are defined several
layers deep.
To improve the situation, we move around and rename several crates. The
end result is that all top-level crates / directories are:
- Either entrypoints into the Rust code, i.e. applications such as
Gateway, Relay or a Client
- Or crates shared across all those entrypoints, such as `telemetry` or
`logging`
