Failing tests that do not test mandatory combine are updated to skip
the mandatory combine pass. Othere tests are updated to use otherwise
removed values.
SIL type lowering erases DynamicSelfType, so we generate
incorrect code when casting to DynamicSelfType. Fixing this
requires a fair amount of plumbing, but most of the
changes are mechanical.
Note that the textual SIL syntax for casts has changed
slightly; the target type is now a formal type without a '$',
not a SIL type.
Also, the unconditional_checked_cast_value and
checked_cast_value_br instructions now take the _source_
formal type as well, just like the *_addr forms they are
intended to replace.
Specifically, we were preferring the always correct ownership kind specified by
the FunctionType and ignoring what we parsed from the argument. This PR changes
ossa to give a nice error when this is detected and fixes the places where this
tests were written incorrectly.
When deserializing an opaque type xref inside an extension context, we were looking
incorrectly in the base module of the type being extended, rather than in the module
of the extension, where the opaque type would really be. Fixes rdar://problem/51775500.
This includes a small refactoring of OpaqueTypeDecl deserialization to break the inevitable
cycle between deserializing the namingDecl, and the namingDecl turning around and re-
deserializing its opaque return type. This is NFC but avoids some unnecessary work.
Note that SILModule::lookUpWitnessTable() only attempts to deserialize
the witness table if we already have a declaration. In the
SILLinkerVisitor, we would call lookUpWitnessTable(), see if it returned
null, create a declaration, and if deserializeLazily is true, we would
call lookUpWitnessTable() again; this time, there was a declaration, and
we would deserialize the witness table.
However, if there was already a witness table declaration when we first
called lookUpWitnessTable(), we would trigger deserialization, even if
deserializeLazily is false.
Change the first call to pass false.
I had to update a couple of tests; I believe they were already somewhat
nonsensical.
In a previous commit, I banned in the verifier any SILValue from producing
ValueOwnershipKind::Any in preparation for this.
This change arises out of discussions in between John, Andy, and I around
ValueOwnershipKind::Trivial. The specific realization was that this ownership
kind was an unnecessary conflation of the a type system idea (triviality) with
an ownership idea (@any, an ownership kind that is compatible with any other
ownership kind at value merge points and can only create). This caused the
ownership model to have to contort to handle the non-payloaded or trivial cases
of non-trivial enums. This is unnecessary if we just eliminate the any case and
in the verifier separately verify that trivial => @any (notice that we do not
verify that @any => trivial).
NOTE: This is technically an NFC intended change since I am just replacing
Trivial with Any. That is why if you look at the tests you will see that I
actually did not need to update anything except removing some @trivial ownership
since @any ownership is represented without writing @any in the parsed sil.
rdar://46294760
This is how we originally controlled whether or not we printed out ownership
annotations when we printed SIL. Since then, I have changed (a few months ago I
believe) the ownership model eliminator to know how to eliminate these
annotations from the SIL itself. So this hack can be removed.
As an additional benefit, this will let me rename -enable-sil-ownership to
-enable-sil-ownership-verifier. This will I hope eliminate confusion around this
option in the short term while I am preparing to work on semantic sil again.
rdar://42509812
LinkNormal mode is used by the mandatory pipeline. It only needs
to deserialize what is necessary for code generation, that is
functions with shared linkage that must be emitted into the client.
Mandatory inlining performs devirtualization, which is able
to deserialize vtables on its own. And since class methods
cannot be @_transparent, we don't have to deserialize any
vtable entries unless we're in LinkAll mode, which indicates
we're running the performance pipeline.
Code may end up indirectly using a witness table for a Clang-imported type by inlining code that used the conformance from another module, in which case we need to ensure we have a local definition at hand in the inlining module so we can have something to link against independently. This needs to be fixed from both sides:
- During serialization, serialize not only witness tables from the current module, but from Clang-imported modules too
- During deserialization, when the SILLinker walks a loaded module, ensure that all shared conformances get deserialized, including those from ApplyInsts and inherited/associated type protocol requirements.
Fixes rdar://problem/38687726.
An over-eager assertion in ModuleFile::maybeReadGenericParams() rejected
deserialization into the context of a generic subscript for a generic
parameter within one of its accessors. Weaken the assertion; the
DeclContext of the generic parameter will be overwritten with the
correct context later.
Fixes rdar://problem/37408205.
This is going to be used for "always emit into client" functions,
such as default argument generators and stored property
initializers.
- In dead function elimination, these functions behave identically to
public functions, serving as "anchors" for the mark-and-sweep
analysis.
- There is no external variant of this linkage, because external
declarations can use HiddenExternal linkage -- the definition should
always be emitted by another translation unit in the same Swift
module.
- When deserialized, they receive shared linkage, because we want the
linker to coalesce multiple copies of the same deserialized
definition if it was deserialized from multiple translation units
in the same Swift module.
- When IRGen emits a definition with this linkage, it receives the
same LLVM-level linkage as a hidden definition, ensuring it does not
have a public entry point.
This is technically a source break, but the @_fixed_layout attribute
is not official yet. If anyone really cares, we can make this
conditional on -swift-version 5 later, but I'd rather not.
This change is necessary so that we can give property initializers
non-public linkage. Currently they are public, because they can be
referenced from inlinable initializers.
Now that property initializers inside a @_fixed_layout type can
only reference public symbols, they no longer have to be public,
but making that change requires a bit more work.
Also, add a third [serializable] state for functions whose bodies we
*can* serialize, but only do so if they're referenced from another
serialized function.
This will be used for bodies synthesized for imported definitions,
such as init(rawValue:), etc, and various thunks, but for now this
change is NFC.
This is dead code and can be re-added if it is needed. Right now though there
really isnt a ValueOwnershipKind that corresponds to deallocating and I do not
want to add a new ValueOwnershipKind for dead code.
What I've implemented here deviates from the current proposal text
in the following ways:
- I had to introduce a FunctionArrowPrecedence to capture the parsing
of -> in expression contexts.
- I found it convenient to continue to model the assignment property
explicitly.
- The comparison and casting operators have historically been
non-associative; I have chosen to preserve that, since I don't
think this proposal intended to change it.
- This uses the precedence group names and higherThan/lowerThan
as agreed in discussion.
Similarly to how we've always handled parameter types, we
now recursively expand tuples in result types and separately
determine a result convention for each result.
The most important code-generation change here is that
indirect results are now returned separately from each
other and from any direct results. It is generally far
better, when receiving an indirect result, to receive it
as an independent result; the caller is much more likely
to be able to directly receive the result in the address
they want to initialize, rather than having to receive it
in temporary memory and then copy parts of it into the
target.
The most important conceptual change here that clients and
producers of SIL must be aware of is the new distinction
between a SILFunctionType's *parameters* and its *argument
list*. The former is just the formal parameters, derived
purely from the parameter types of the original function;
indirect results are no longer in this list. The latter
includes the indirect result arguments; as always, all
the indirect results strictly precede the parameters.
Apply instructions and entry block arguments follow the
argument list, not the parameter list.
A relatively minor change is that there can now be multiple
direct results, each with its own result convention.
This is a minor change because I've chosen to leave
return instructions as taking a single operand and
apply instructions as producing a single result; when
the type describes multiple results, they are implicitly
bound up in a tuple. It might make sense to split these
up and allow e.g. return instructions to take a list
of operands; however, it's not clear what to do on the
caller side, and this would be a major change that can
be separated out from this already over-large patch.
Unsurprisingly, the most invasive changes here are in
SILGen; this requires substantial reworking of both call
emission and reabstraction. It also proved important
to switch several SILGen operations over to work with
RValue instead of ManagedValue, since otherwise they
would be forced to spuriously "implode" buffers.
While making some changes that resulted in materializeForSet getting
synthesized more often, I noticed that some tests started failing
because they could not find the Optional type, which appears in the
type of materializeForSet. Fix these tests by defining an Optional
and passing in '-module-name Swift' so that we can find it.
This feels a little bit gross, but Dmitri says it is preferrable to
importing Swift from a test.
Having a separate address and container value returned from alloc_stack is not really needed in SIL.
Even if they differ we have both addresses available during IRGen, because a dealloc_stack is always dominated by the corresponding alloc_stack in the same function.
Although this commit quite large, most changes are trivial. The largest non-trivial change is in IRGenSIL.
This commit is a NFC regarding the generated code. Even the generated SIL is the same (except removed #0, #1 and @local_storage).
This reflects the fact that the attribute's only for compiler-internal use, and isn't really equivalent to C's asm attribute, since it doesn't change the calling convention to be C-compatible.
All refutable patterns and function parameters marked with 'var'
is now an error.
- Using explicit 'let' keyword on function parameters causes a warning.
- Don't suggest making function parameters mutable
- Remove uses in the standard library
- Update tests
rdar://problem/23378003
'Ss' appears in manglings tens of thousands of times in the standard library and is also incredibly frequent in other modules. This alone is enough to shrink the standard library by 59KB.
Swift SVN r32409
Replace the use of -sil-inline-threshold with @inline(never) on specific
functions. This is required in some cases in order to remove the generic
specialization pass and allow the inliner to do the specialization
instead (specifically the cases where the parameter is 0, which causes
us to exit the inliner immediately).
Also add public in some places to ensure that we do not toss out
functions before we even make it to the inliner (once the standalone
generic specialization pass is removed).
Swift SVN r31126
The rule changes are as follows:
* All functions (introduced with the 'func' keyword) have argument
labels for arguments beyond the first, by default. Methods are no
longer special in this regard.
* The presence of a default argument no longer implies an argument
label.
The actual changes to the parser and printer are fairly simple; the
rest of the noise is updating the standard library, overlays, tests,
etc.
With the standard library, this change is intended to be API neutral:
I've added/removed #'s and _'s as appropriate to keep the user
interface the same. If we want to separately consider using argument
labels for more free functions now that the defaults in the language
have shifted, we can tackle that separately.
Fixes rdar://problem/17218256.
Swift SVN r27704
The deallocating parameter convention is a new convention put on a
non-trivial parameter if the caller function guarantees to the callee
that the parameter has the deallocating bit set in its object header.
This means that retains and releases do not need to be emitted on these
parameters even though they are non-trivial. This helps to solve a bug
in +0 self and makes it trivial for the optimizer to perform
optimizations based on this property.
It is not emitted yet by SILGen and will only be put on the self
argument of Deallocator functions.
Swift SVN r26179
Some AST nodes and SIL instructions need to reference conformances for a
particular type. If that type was imported from Clang, however, the
conformance may not exist when the AST node or SIL function gets deserialized
later. The SIL case is the problem case: fragile SIL code may contain a
reference to a conformance never mentioned in the AST of the code being
compiled, and since conformances are synthesized on demand during type-checking,
this will lead to a crash. SIL deserialization isn't supposed to be doing
work on its own (though it ~can~ import new Clang decls at the moment), so
the best answer is to serialize the conformances directly, like we would with
specialized or inherited conformances.
We can probably do better here in the long run (we don't even unique
conformances like this within a module), but this should at least handle the
immediately known problem cases.
rdar://problem/18669402
Swift SVN r22857
SIL functions use AST GenericParamTypeDecls, but they don't have a useful
DeclContext, so they just use the AST module associated with the current
SILModule. However, when it comes time to reserialize referenced functions
with shared_external linkage (such as closures defined in fragile public
functions), the serializer was trying to cross-reference those generic
parameters rather than reserialize them, because they aren't part of the
current source file. And because these decls aren't attached to a specific
AST DeclContext, we can't properly cross-reference them---nor should we.
This commit introduces a targeted case in the cross-reference logic to force
re-serializing these declarations. In the long run we may want to reconsider
using AST GenericParamLists for SILFunctions.
rdar://problem/18673024
Swift SVN r22800
This is needed for tests which define internal functions which should not be eliminated.
So far this was not needed because of a hack which prevented whole-module-optimizations for tests.
Swift SVN r22658
generates them.
Modify getAsCanonicalGenericSignature to dump same-type requirements last.
Also mix the conformance requirements on assocaited archetypes with the witness
markers.
SILParser used to put witness markers for all assocaited archetypes, then
add same-type requirements, and finally the conformance requirements on
associated archetypes. This causes mismatch types between deserialized
SILFunctionTypes and parsed SILFunctionTypes.
rdar://17998988
Swift SVN r21423
The second type of WitnessMarker for deserialized GenericSignature is null.
But the second type for parsed GenericSignature is not null, causing type
mismatch error when linking a SILFunction.
This commit ignores the second type of WitnessMarker when profiling the
GenericSignature.
rdar://17998988
Swift SVN r21178
