PrintAsClang is supposed to emit declarations in the same order regardless of the compiler’s internal state, but we have repeatedly found that our current criteria are inadequate, resulting in non-functionality-affecting changes to generated header content. Add a diagnostic that’s emitted when this happens soliciting a bug report.
Since there *should* be no cases where the compiler fails to order declarations, this diagnostic is never actually emitted. Instead, we test this change by enabling `-verify` on nearly all PrintAsClang tests to make sure they are unaffected.
This did demonstrate a missing criterion that only mattered in C++ mode: extensions that varied only in their generic signature were not sorted stably. Add a sort criterion for this.
We will place tests that test the semantics and runtime of the generated
C++ code into Interop/SwiftToCxx. PrintCxx test directory will only be
used to test high level flow without concerning itself about the C++ semantics.
This is a new warning in clang when you use @import in a framework header. That’s often a mistake, but it isn’t in our generated -Swift.h headers. If we’re building with a clang that supports this warning, we’ll now emit a pragma to disable it.
Turns out libc++ already defines compatibility typedefs in this case,
and we don't care about any other C++ stdlibs on Apple platforms at
the moment.
rdar://problem/34861807
The C preprocessor rules don't short-circuit so "#if defined(__has_feature) && __has_feature(modules)" will always fail if '__has_feature' is not defined.
These tests explicitly disable the use of modules. However, the lit
configuration is such that the module cache path will be passed unconditionally
to the clang invocation. Squelch the unused option warning (error).
This is required to correctly use the mock SDK when the SDK overlay is
built and tested separately. (Otherwise, the mock SDK might not get
used, because the overlay SDK options would expand from the
%-substitution, appear first on the command line, and shadow the mock
SDK in the search path).
Swift SVN r25185
Most tests were using %swift or similar substitutions, which did not
include the target triple and SDK. The driver was defaulting to the
host OS. Thus, we could not run the tests when the standard library was
not built for OS X.
Swift SVN r24504
Doing so is safe even though we have mock SDK. The include paths for
modules with the same name in the real and mock SDKs are different, and
the module files will be distinct (because they will have a different
hash).
This reduces test runtime on OS X by 30% and brings it under a minute on
a 16-core machine.
This also uncovered some problems with some tests -- even when run for
iOS configurations, some tests would still run with macosx triple. I
fixed the tests where I noticed this issue.
rdar://problem/19125022
Swift SVN r23683
- Category names weren't unique.
- We were using an attribute to detect if something was a Swift category,
but attributes can't be used on categories.
- The test that this was all working was failing in a way that wasn't caught.
To solve these problems:
- We're using a macro to generate category names based on __LINE__ in addition
to the current module.
- The importer uses the macro to detect that the category comes from Swift
(no attribute needed).
- The test now has a deliberate error for -verify to catch.
<rdar://problem/17342287&17538553>
Swift SVN r19479
When generating a header for the Swift half of a mixed-source framework,
we can't import the framework using @import, because that means a submodule
is trying to import a parent module before the module is done being built.
This currently isn't supported in Clang, though it only recently became an
error instead of being ignored.
Instead, we now assume that the framework will have an umbrella header with
the same name as the framework module, which is the same assumption Xcode
makes when you don't provide your own module map. I'm not too concerned about
people trying to build mixed-source frameworks who /don't/ have umbrella
headers.
This doesn't affect app targets at all, which use -import-objc-header instead
of a standalone underlying module.
<rdar://problem/16879704>
Swift SVN r17984
