Migrate `ServiceWorkerMsg` and `SWManagerMsg` to GenericChannel
Testing: Covered by service worker wpt tests
Part of #38912
Signed-off-by: Jonathan Schwender <schwenderjonathan@gmail.com>
One commonly encountered mechanism in servo is using ipc channels
together with the router, to register a custom callback to run in the
current process, when receiving a reply.
The new `GenericCallback` abstracts over this, and allows executing an
arbitrary callback in the process of the `GenericCallback` creator. In
multiprocess mode, this internally uses ipc channels and follows the
existing pattern.
In single process mode, we execute the callback directly, which avoids
one call to the router.
Executing the callback still incurs synchronization, since we need to
support cloning the abstraction, and the callback closure may be
`FnMut`. Future work could provide more optimized abstractions for
callbacks that don't have these requirements.
This PR allows applying #38782 again, which was previously reverted in
#38940 due to the lack of custom callback support.
See also the module documentation in `generic_channel/callback.rs`.
Testing: This PR adds unit tests. Also passes the manual testcase from
#38939
Part of #38912
---------
Signed-off-by: Jonathan Schwender <schwenderjonathan@gmail.com>
Signed-off-by: Jonathan Schwender <55576758+jschwe@users.noreply.github.com>
Co-authored-by: Josh Matthews <josh@joshmatthews.net>
This change includes the following additions to GenericChannel:
- Add a GenericSend trait which is meant to replace the `IpcSend` trait
over time, as channels are migrated. For the time being this means, that
we often need to use `GenericSend::send()` to disambiguate from the
`IpcSend::send` function, until all usages of `IpcSend` have been
replaced.
- Add an OpaqueSender impl for GenericSender
- Add a profiled version of GenericChannel. The profiling is 1:1 the
same as for the existing profiled IPC channel, namely that only the
blocked time during `recv` is measured.
Testing: No functional changes, covered by existing tests
Part of #38912
---------
Signed-off-by: Jonathan Schwender <schwenderjonathan@gmail.com>
This reverts commit fb1c0a4c48.
Previously in `create_compositor_channel`, the [routing callback][1] was
setup so that a message received on the Compositor's IPC receiver will
be
forwarded to the local receiver using the `CompositorProxy` which also
takes care of waking up the event loop. In #38782, this was changed so
that the routing callbacks simply forwards the message directly without
going via the `CompositorProxy`. This breaks behaviours that rely on the
event loop being woken up on message sending, e.g. updating image frames
for animated gifs.
Since the GenericChannel API doesn't allow custom routing callbacks,
revert this change until we figure out a better solution.
[1]:
d2ccce6052/components/servo/lib.rs (L1114)
Signed-off-by: Mukilan Thiyagarajan <mukilan@igalia.com>
Signed-off-by: Mukilan Thiyagarajan <mukilan@igalia.com>
Besides migrating the channel to GenericChannel, this PR adds
`routed_channel_with_local_sender()` to `generic_channel`. This is for
existing use-cases, where we want to provide both an IPC capable
GenericSender, as well as a crossbeam Sender, for efficient sending if
the sender is in the same process.
Testing: All of our channels should send / receive at least some
messages during WPT tests.
Signed-off-by: Jonathan Schwender <schwenderjonathan@gmail.com>
In Single-process mode, sending crossbeam channels over an ipc channel
is perfectly safe to do (since everything is in the same process) and
allows us to more easily incrementally port channels.
In Multi-process mode, GenericChannels will always be IPC channels, so
we won't hit the "serialize crossbeam channels" branch (since the only
way to construct channels, checks the process mode). This property is
ensured by `channel()` being the only way to construct a `GenericSender`
and Receiver pair. To achieve this, we make the previously `pub` enum
private, and wrap it in a newtype, so that the type can't be constructed
from outside the module.
To be extra safe, we still check if we are in multiprocess mode or not
during (de-)serialization and emit an error.
Testing: Add a new unit-test to ensure sending GenericSender / Receivers
over an ipc_channel works.
---------
Signed-off-by: Jonathan Schwender <schwenderjonathan@gmail.com>
We should not be using `route_ipc_receiver_to_new_crossbeam_receiver` or
similar methods, that `unwrap()` on the ROUTER thread if they encounter
IPC errors. Instead, we now propagate the error to the crossbeam
receiver.
In the GenericChannel::Crossbeam case this means, that we need to use a
`Result<T>` as the data type, even though the Result variant is always
okay, so that the receiver type is the same regardless of `IPC` or not.
This is required, so we have the same channel type, and can pass the
inner crossbeam channel into e.g. `select!`, without having to wrap or
re-implement select.
This also means, that as we switch towards GenericChannel, we will
gradually improve our error handling and eventually remove the existing
panics on IPC errors.
These changes were extracted out of
https://github.com/servo/servo/pull/38782
Testing: Covered by existing tests. No new panics were introduced.
Signed-off-by: Jonathan Schwender <schwenderjonathan@gmail.com>
Motivation:
Using our GenericChannel abstraction allows us to optimize IPC in
single-process mode to just use cross-beam channel.
To keep the diff low, and get early feedback, this PR only tackles a
single channel, but the intention is to port all ipc channels to the
generic channel, which allows us to skip serializing and deserializing
messages in single process mode.
Based on:
- https://github.com/servo/servo/pull/38638
- https://github.com/servo/servo/pull/38636
Testing: Covered by existing tests
---------
Signed-off-by: Jonathan Schwender <schwenderjonathan@gmail.com>
Implement Serialize and Deserialize for GenericReceiver to also allow
the Receiver to be sent across
ipc channels.
This is necessary to allow using the GenericChannel in more places.
Testing: Manually tested on follow-up feature branch. Does not require
new tests.
---------
Signed-off-by: Jonathan Schwender <schwenderjonathan@gmail.com>
Signed-off-by: Jonathan Schwender <55576758+jschwe@users.noreply.github.com>
Co-authored-by: Martin Robinson <mrobinson@igalia.com>
