When a class has an unavailable conformance to a protocol, do not
inherit that unavailable conformance, because it can get in the way of
subclasses defining their own (properly-available) conformance.
Fixes rdar://89992569.
When determining whether a superclass conforms to a particular protocol,
skip unavailable conformances. This way, we don't minimize away a
constraint that might only apply to subclasses of the specified
superclass.
Fixes rdar://91853658.
The three options are now:
* `explicit`: Enforce Sendable constraints where it has been explicitly adopted and perform actor-isolation checking wherever code has adopted concurrency. (This is the default)
* `targeted`: Enforce Sendable constraints and perform actor-isolation checking wherever code has adopted concurrency, including code that has explicitly adopted Sendable.
* `complete`: Enforce Sendable constraints and actor-isolation checking throughout the entire module.
when a variable with type `<<error-type>>` was captured in a
sendable function's environment, we use to emit a diagnostic
about it not being sendable, but it's for a bogus type.
at some point this issue disappeared as the sendable implementation
evolved, but this commit adds regression coverage.
resolves rdar://82452688
Use the explicit check for a Sendable conformance (even an unavailable
one) as the mechanism for determining whether to diagnose a
missing/unavailable Sendable conformance in a particular context.
When in "existing" Swift code that is Swift 5.x and has not adopted
concurrency, allow mismatches in function types that only involve
ABI-neutral concurrency changes (e.g., adding `@Sendable` or removing
a global actor) by downgrading the diagnostic to a warning. This
improves the story for incremental adoption of concurrency in an
existing code base.
As part of this, generalize the facility for downgrading an error to a
warning when performing diagnostics in the constraint solver, using the
new diagnostic behavior limits rather than duplicating diagnostics.
