The name Stream didn't seem to be working out as intended; we kept
gravitating back to calling it Generator, which is precedented in other
languages. Also, Stream seems to beg for qualification as Input or
Output. I think we'd like to reserve Stream for things that are more
bulk-character-API-ish.
Swift SVN r13893
Previously, we would just parse vars and subscripts with no definitions,
then let getters and setters be referenced arbitrarily later. This was
problematic for a number of reasons, not least of which, the .sil file
might be invalid.
Instead, change sil to require that a protocol style definition indicate
whether a vardecl/subscript is computed or not, and whether it is both
get-able and set-able, e.g. like "var x : Int { get }". Change the
sil printer to print decls in this form, and change the SILParser to
make SILDeclRef::Func values instead of ::Getter/Setter values.
One thing that this exposed is that we weren't correctly serializing the
accessor state in modules, so accessors would get detatched from their
AbstractStorageDecls when deserialized (and in fact, their ASD never got
deserialized at all in some cases). Fix this in the serialization of
the accessors.
NFC, other than the SIL printer and parser.
Swift SVN r13884
Riding off of project_existential[_ref] was convenient, but the
resuls are used quite differently. Note that open_existential[_ref]
still don't print/parse reasonably yet.
Swift SVN r13878
From now on, /any/ changes to SIL or AST serialization must increment
VERSION_MINOR in ModuleFormat.h.
The original intent of VERSION_MAJOR/VERSION_MINOR was that VERSION_MAJOR
would only increment when backwards-incompatible changes are introduced,
and VERSION_MINOR merely indicates whether to expect additional information.
However, the module infrastructure currently isn't forgiving enough to accept
even backwards-compatible changes to the record schemas, and the SIL
serialization design might not be compatible with that at all.
So for now, treat any version number 0.x as incompatible with any other 0.y.
We can bump to 1 when we hit stability.
<rdar://problem/15494343>
Swift SVN r13841
We kinda need those. Limit the hack to pad it out with nulls to only apply in the case when the conformance list is empty but the original archetype requires conformances, which only occurs for archetypes or existentials.
At long last, we can build at -O0 again, again! Reapplying now that Jordan fixed some fallout this had on his objc printer tests. Thanks Jordan!
Swift SVN r13840
GenericSignatures with no params or requirements are a bug, so verify that they don't happen by making GenericSignature::get return null and GenericFunctionType assert that it has a nonnull signature. Hack Sema not to try to produce nongeneric GenericFunctionTypes when a function in a local type in a generic function context is type-checked; there's a deeper modeling issue that needs to be fixed here, but that's beyond the scope of 1.0. Now that GenericSignature always has at least one subtype, its factories no longer need an independent ASTContext argument.
Swift SVN r13837
We kinda need those. Limit the hack to pad it out with nulls to only apply in the case when the conformance list is empty but the original archetype requires conformances, which only occurs for archetypes or existentials.
At long last, we can build at -O0 again!
Swift SVN r13815
Change GenericFunctionType to reference a GenericSignature instead of containing its generic parameters and requirements in-line, and clean up some interface type APIs that awkwardly returned ArrayRef pairs to instead return GenericSignatures instead.
Swift SVN r13807
This is mostly useful for the standard library, whose name is going to
change to "Swift" soon. (See <rdar://problem/15972383>.) But it's good DRY.
Swift SVN r13758
Introduce a new expression kind, OpenExistentialExpr, that "opens" up
an existential value into a value of a fresh archetype type that
represents the dynamic type of the existential. That value can be
referenced (via an OpaqueValueExpr) within the within the
subexpression of OpenExistentialExpr. For example, a call to a
DynamicSelf method on an existential looks something like this:
(open_existential_expr implicit type='P'
(opaque_value_expr implicit type='opened P' @ 0x7fd95207c290
unique)
(load_expr implicit type='P'
(declref_expr type='@lvalue P' decl=t.(file).func
decl.p@t.swift:5:37 specialized=no))
(erasure_expr implicit type='P'
(call_expr type='opened P'
(archetype_member_ref_expr type='() -> opened P'
decl=t.(file).P.f@t.swift:2:8 [with Self=opened P]
(opaque_value_expr implicit type='opened P' @
0x7fd95207c290 unique))
(tuple_expr type='()')))))
Note that we're using archetype_member_ref_expr rather than
existential_member_ref_expr, because the call is operating on the
opaque_value_expr of archetype type. The outer erasure turns the
archetype value back into an existential value.
The SILGen side of this is somewhat incomplete; we're using
project_existential[_ref] to open the existential, which is almost
correct: it gives us access to the value as an archetype, but IRGen
doesn't know to treat the archetype type as a fresh archetype whose
conformances come from the existential. Additionally, the output of
the opened type is not properly parsable. I'll fix this in follow-on
commits.
Finally, the type checker very narrowly introduces support for
OpenExistentialExpr as it pertains to DynamicSelf. However, this can
generalize to support all accesses into existentials, eliminating the
need for ExistentialMemberRef and ExistentialSubscript in the AST and
protocol_method in SIL, as well as enabling more advanced existential
features should we want them later.
Swift SVN r13740
There are some straggling references to the context generic param list, but nothing uses the non-interface param or result types anymore!
Swift SVN r13725
SubscriptDecl is created, then the accessors are installed on it.
This allows us to create the subscript decl before the accessors
have been parsed, allowing us to build the subscript even in invalid
cases (better for later error recovery).
More importantly, this allows us to add it to Decls before calling
parseGetSet, so we can now make parseGetSet add accessors to Decls
without breaking source order (something that deeply upsets the IDE
features).
With all this untangled, we can now remove the 'addAccessorsInOrder'
hack where we parsed the accessors and then later tried to figure out
which order they came for the purpose of linking up the AST: accessors
now work just like everything else.
Swift SVN r13708
now that they are implicitly updated. This exposes two things:
1) we're unncessarily serializing selfdecls in ctors and dtors.
2) The index pattern of a SubscriptDecl has no sensible DeclContext that
owns variables in it.
I'll deal with the first tomorrow, I'm not sure what to do with
the second one.
Swift SVN r13703
Also, disallow creating Modules and FileUnits on the stack. They must always
live as long as the ASTContext.
<rdar://problem/15596964>
Swift SVN r13671
Edge SILFunction one step closer to independence from SILFunctionType context by taking the generic param list as a separate constructor parameter, and serializing those params alongside the function record. For now we still pass in the context params from the SILFunctionType in most cases, because the logic for finding the generic params tends to be entangled in type lowering, but this pushes the problem up a step.
Thanks Jordan for helping work out the serialization changes needed.
Compared to r13036, this version of the patch includes the decls_block RecordKind enumerators for the GENERIC_PARAM_LIST layouts in the sil_block RecordKind enumerator, as Jordan had suggested before. r13036 caused buildbot failures when building for iOS, but I am unable to reproduce those failures locally now.
Swift SVN r13485
This re-applies r13401, reverted in r13404. This wasn't actually causing
problems, but got pulled along with r13400 (reverted in r13405).
Swift SVN r13452
It's OK to clone shared definitions, and we need to in order to carry non-inlined specializations with the transparent functions that reference them.
Swift SVN r13288
Also, don't allow file-relative imports (i.e. don't consider a /source file's/
location as an include path). Both of these should be requested by the user,
at least at this point.
Swift SVN r13249
with FuncDecls. This allows us to eliminate special case code for handling
self in various parts of the compiler.
This also improves loc info (debug info and AST info) because 'self' now
has a location instead of being invalid.
I also took the opportunity to factor a bunch of places creating self decls
to use similar patterns and less copy and paste code.
Swift SVN r13196
Body of closure needs to be serialized if its caller is serialized.
To fix this, we serialize private functions that are referenced by a serialized
function.
rdar://15884245
Swift SVN r13162
Allow IfStmts and WhileStmts to have as their condition either an expression, as usual, or a pattern binding introduced by 'var' or 'let', which will conditionally bind to the value inside an optional. Unlike normal pattern bindings, these bindings require an in-line initializer, which will be required to be Optional type. Parse variable bindings in this position, and type-check them by requiring an Optional on the right-hand side and unwrapping it to form the pattern type. Extend SILGen's lowering of if and while statements to handle conditionally binding variables.
Swift SVN r13146
Making DynamicSelf its own special type node makes it easier to opt-in
to the behavior we want rather than opting out of the behavior we
don't want. Some things already work better with this representation,
such as mangling and overriding; others are more broken, such as the
handling of DynamicSelf within generic classes and the lookup of the
DynamicSelf type.
Swift SVN r13141
This keeps us from having to deal with fat swiftmodules for now.
In the long run we're hoping to solve this problem with build configurations,
so that a single module file can support multiple architectures.
(See <rdar://problem/15056323>)
<rdar://problem/15204953>
Swift SVN r13135
Edge SILFunction one step closer to independence from SILFunctionType context by taking the generic param list as a separate constructor parameter, and serializing those params alongside the function record. For now we still pass in the context params from the SILFunctionType in most cases, because the logic for finding the generic params tends to be entangled in type lowering, but this pushes the problem up a step.
Thanks Jordan for helping work out the serialization changes needed.
Swift SVN r13036
We add two records in sil_block to specify a witness table record and a method
entry record. Out of the four entry types, only "Method" is handled in this
commit.
Two records are also added to sil_index_block to search for a specific witness
table given a unique identifier. The interface lookupWitnessTable is not
implemented yet.
Right now, we serialize a witness table only when sil-serialize-all is on and
deserialize all witness tables in the module when sil-link-all is on.
rdar://15722175
Swift SVN r13000
Instead of taking the scenic route of building a Swift function type from the parameter types and re-lowering it to a SIL type, which loses the parameter convention information, derive the type of the partial_apply directly from the type of its callee. Addresses part of the problem with <rdar://problem/15884245>, but does not fix it, because it looks like we still don't serialize closures referenced inside transparent functions.
Swift SVN r12970
