Member operators of concrete nominal types must declare at least
one parameter with that type, like
```
struct S {
static func +(lhs: S, rhs: Int) -> S {}
}
```
For protocol member operators, we would look for a parameter of type
`Self`, or an existential type `any P`. While the latter was
consistent with the concrete nominal type case, it was actually
wrong because then the resulting interface type does not give the
type checker any way to bind the `Self` type parameter.
There were two existing test cases that now produce errors, which I
believe is now correct. While this is technically a source break,
because these bogus operators seemingly cannot be witnessed or called,
such a protocol probably had no conforming types.
Fixes https://github.com/apple/swift/issues/73201.
The presence of Copyable/Escapable conformances doesn't affect ABI. Only
their absence in terms of suppressed requirements like `~Copyable` need
to be output for diffing by the APIDigester.
The IUO-ness of imported declarations is not actually computed by IsImplicitlyUnwrappedOptionalRequest. Instead, ClangImporter manually sets the bit to `true` after the declaration’s type is imported and expects IsImplicitlyUnwrappedOptionalRequest to always set it to `false` for all other imported declarations.
Normally, declaration types are imported greedily as soon as the declaration is created. However, a ClangImporter refactoring in apple/swift#61026 deferred the import of a VarDecl’s type, and therefore the setting of its IUO bit, until the first time InterfaceTypeRequest is evaluated.
It turns out that there is nothing to guarantee that InterfaceTypeRequest will be evaluated before IsImplicitlyUnwrappedOptionalRequest, so if isImplicitlyUnwrappedOptional() was fetched before getInterfaceType() was called, it would return an incorrect result. The only known client that accesses the information in this order is the API digester, but in theory any part of the compiler could fall into this trap.
Force the evaluation of InterfaceTypeRequest during IsImplicitlyUnwrappedOptionalRequest when necessary to compute the IUO bit for an imported VarDecl, and add a test to prove that this fixes the observed bug in the API digester.
`module.map` as a module map name has been discouraged since 2014, and
Clang will soon warn on its usage. This patch renames all instances of
`module.map` in the Swift tests to `module.modulemap` in preparation
for this change to Clang.
rdar://106123303
The first generic parameter of an `OpaqueTypeDecl` was still being used
as the "underlying" interface type of the opaque type, which is
incorrect for both structural and named opaque result types. Eliminate
this notion, because the (declared) interface type already has the
correct structure.
Only ABI checking depended on the old "underlying" type, so rework it to
instead substitute into properly for structural opaque result types as
well.
Deserialization required a small adjustment to eliminate a cycle
because the interface type of an `OpaqueTypeDecl` involves opaque
archetype types, which reference the declaration itself... so
deserialize the interface type later, now that it's correct.
Previously, we use USR as a delegate for mangled name. However, USR won't incorporate name changes made
by attributes like @_silgen_name. Instead, we should add a dedicated field for canonical mangled names.
rdar://84202064
Previously, we use USR as a delegate for mangled name. However, USR won't incorporate name changes made
by attributes like @_silgen_name. Instead, we should add a dedicated field for canonical mangled names.
rdar://84202064
Tests using a mock SDK can't load the stdlib of the normal SDK (or the
just-built one) with the recent changes. Use `-parse-stdlib` in those
tests to prevent loading the stdlib implicitly.
Removing or updating @_alwaysEmitIntoClient functions are never ABI-breaking, thus
we should exclude them from the symbol set for ABI stability checking.
rdar://67883661
Most of this is enabled by an upstream change to generalize the tests
for the api-digester to properly handle all targets. The primary change
here is to account for the Swift-level ABI differences for the standard
library between Apple Silicon macOS and Intel macOS. Specifically, anything
related to Float80 will be removed from the standard library on Apple Silicon
macOS, so account for the differences.
Previously inherited constructors would be skipped from added in member list, depending on the order of request evaluator calls.
This was a regression compared to swift 5.2
Lazy loading checked if the ClangDecl was hidden, but loading all
members did not. Let's make loadAllMembers() behave like the lazy
path, and fix some of the mock SDKs in the test suite.
Because we won’t be serializing this attribute, add custom diagnostics for the cases where:
- We add @_hasMissingDesignatedInits to an open class, which means subclasses won’t be able to inherit its inits
- We remove @_inheritsConvenienceInitializers, which means APIs are removed
If client's subclass provides an implementation of all of its superclass designated
initializers, it automatically inherits all of the superclass convenience initializers.
This means if a new designated init is added to the base class, the inherited
convenience init may be missing and cause breakage.
SR-11454
The framework baselines are installed at 'lib/swift/FrameworkABIBaseline' and the
tool is inside 'bin'. This patch teaches the executable to locate baselines
from the relative path.
In addition, this patch moves the stdlib ABI/API baselines to the canonical
location so we don't have to check the stability of the stdlib using a
different mechanism from other Swift frameworks.
ABI/API checker should check semantic differences of two modules.
Adhering too strictly to the actual ASTs could yield false positives. This
patch populates ObjC, Dynamic and Final to the attribute list if AST
APIs say so.
rdar://50217247
Framework authors usually have different schemes for different deployment
targets. We should exclude platform-unavailable ABIs from the Json file so
developers will only be warned of the breakages that are relevant to the current
scheme.
rdar://54273296
This change introduces a Json format change where we always print fully
qualified type names everywhere. This is beneficial for diagnosing purposes but
may not be ideal for generating migrator scripts. To resolve this conflict, the
patch also introduces a flag -migrator to opt-out fully qualified type names.
After the format change, we need to update the ABI and API baselines for the
Swift stdlib.
There could be cases where several modules having types of the same name.
Using fully qualified names could help us detect changing from using
one of them to another.
The commit also updates the existing ABI/API baselines to incorporate such change.
Inlinable decls from swift interface files allow us to construct the underlying types
of opaque result types, whose changes can break ABI.
rdar://52273137
Removing accessors other than getter and setter can be ABI breaking. This
patch starts to formally include all accessor decls in the tree and diagnose
their removal. This change only applies to the ABI checker since we still
exclude accessors other than getter and setter when diagnosing source
compatibility.
Including accessors formally can also allow us to check the missing
of availability attributes for newly added accessors.
rdar://52063421
With module stability from the compiler, we can now use an earlier version
of the SDK as the baseline and compare the new SDK with it. This patch
adds a new -bsdk flag to the tool to specify the baseline SDK path.
We used to use a Json format to capture the ABI/API detail of Swift modules to
walk-around module compatibility issues across different compiler versions. Since
Swift module now is in stable format, we can compare two swiftinterface files
directly without dumping the content into Json format.
