Forming an isolated conformance to a SendableMetatype-inherting
protocol opens up a soundness hole any time the conformance is used.
Reword the recently-introduced diagnostic for this case and promote it
to an error (except when it's preconcurrency).
Fixes rdar://154808002.
An isolated conformance to a SendableMetatype-inheriting protocol
cannot actually be used in generic code, because the SendableMetatype
requirement itself prevents it. Warn about this case so folks aren't
surprised at runtime.
This is a part of issue #82550 / rdar://154437489.
The IsolatedConformances feature moves to a normal, supported feature.
Remove all of the experimental-feature flags on test cases and such.
The InferIsolatedConformances feature moves to an upcoming feature for
Swift 7. This should become an adoptable feature, adding "nonisolated"
where needed.
(cherry picked from commit 3380331e7e)
This is going to need a proper implementation in the requirement
machine. For the moment, provide a slightly-less-broken implementation
but leave a test case where we incorrectly accept racey code.
(cherry picked from commit 92774e0a3c)
While here, fix some issues around implied isolated conformances (we
could get into an inconsistent state). Also provide an educational
note discussing isolated conformances and the kinds of errors one can
see when they are used from outside of their isolation domain.
A protocol conformance can be ill-formed due to isolation mismatches
between witnesses and requirements, or with associated conformances.
Previously, such failures would be emitted as a number of separate
errors (downgraded to warnings in Swift 5), one for each witness and
potentially an extra for associated conformances. The rest was a
potential flood of diagnostics that was hard to sort through.
Collect all of the isolation-related problems for a given conformance
together and produce a single error (downgraded to a warning when
appropriate) that describes the overall issue. That error will have up
to three notes suggesting specific courses of action:
* Isolating the conformance (when the experimental feature is enabled)
* Marking the witnesses as 'nonisolated' where needed
*
The diagnostic also has notes to point out the witnesses/associated
conformances that have isolation problems. There is a new educational
note that also describes these options.
We give the same treatment to missing 'distributed' on witnesses to a
distributed protocol.
When diagnosing an isolation mismatch between a requirement and witness,
we would produce notes on the requirement itself suggesting the addition of
`async`. This is almost never what you want to do, and is often so far
away from the actual conforming type as to be useless. Remove this note,
and the non-function fallback that just points at the requirement, because
they are unhelpful.
This is staging for a rework of the way we deal with conformance-level
actor isolation problems.
With the move to explicitly specifying the global actor for an isolated
conformance, we can now have conformances whose isolation differs from
that of the type, including having actors with global-actor-isolated
conformances. Introduce this generalization to match the proposal, and
update/add tests accordingly.
Instead of using the `isolated P` syntax, switch to specifying the
global actor type directly, e.g.,
class MyClass: @MainActor MyProto { ... }
No functionality change at this point
Extend the metadata representation of protocol conformance descriptors
to include information about the global actor to which the conformance is
isolated (when there is one), as well as the conformance of that type to
the GlobalActor protocol. Emit this metadata whenever a conformance is
isolated.
When performing a conforms-to-protocol check at runtime, check whether
the conformance that was found is isolated. If so, extract the serial
executor for the global actor and check whether we are running on that
executor. If not, the conformance fails.
This is the missing check for "rule #1" in the isolated conformances proposal,
which states that an isolated conformance can only be referenced within
the same isolation domain as the conformance. For example, a
main-actor-isolated conformance can only be used within main-actor code.
Within the constraint system, introduce a new kind of conformance constraint,
a "nonisolated conforms-to" constraint, which can only be satisfied by
nonisolated conformances. Introduce this constraint instead of the normal
conforms-to constraint whenever the subject type is a type parameter that
has either a `Sendable` or `SendableMetatype` constraint, i.e., when the type
or its values can escape the current isolation domain.
When a protocol conformance somehow depends on an isolated conformance, it
must itself be isolated to the same global actor as the conformance on
which it depends.
Allow a conformance to be "isolated", meaning that it stays in the same
isolation domain as the conforming type. Only allow this for
global-actor-isolated types.
When a conformance is isolated, a nonisolated requirement can be
witnessed by a declaration with the same global actor isolation as the
enclosing type.
