Now we can discern the types of values in heap boxes at runtime!
Closure reference captures are a common way of creating reference
cycles, so this provides some basic infrastructure for detecting those
someday.
A closure capture descriptor has the following:
- The number of captures.
- The number of sources of metadata reachable from the closure.
This is important for substituting generics at runtime since we
can't know precisely what will get captured until we observe a
closure.
- The number of types in the NecessaryBindings structure.
This is a holding tank in a closure for sources of metadata that
can't be gotten from the captured values themselves.
- The metadata source map, a list of pairs, for each
source of metadata for every generic argument needed to perform
substitution at runtime.
Key: The typeref for the generic parameter visible from the closure
in the Swift source.
Value: The metadata source, which describes how to crawl the heap from
the closure to get to the metadata for that generic argument.
- A list of typerefs for the captured values themselves.
Follow-up: IRGen tests for various capture scenarios, which will include
MetadataSource encoding tests.
rdar://problem/24989531
This reorganization allows adding attributes that refer to types.
I need this for a @_specialize attribute with a type list.
PrintOptions.h and other headers depend on these enums. But Attr.h
defines a lot of classes that almost never need to be included.
Each runtime function definition in RuntimeFunctions.def states which calling convention
should be used for this runtime function. But IRGen and LLVMPasses were not always
properly propagating this declared calling convention all the way down to llvm's Call instructions.
In many cases, the standard C convention was set for the call irrespective of the actual calling
convention defined for a given runtime function. As a result, incorrect code was generated.
This commit tries to fix all those places, where such a mismatch was found. In many cases this is
achieved by defining a helper function CreateCall in such a way that makes sure that the call instruction
gets the same calling convention as the one used by its callee operand.
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.
This prevents the linker from trying to emit relative relocations to locally-defined public symbols into dynamic libraries, which gives ld.so heartache.
Make sure to set the linkage correctly, treat the selector data as
non-constant, note that it is externally-initialized, and add it to
llvm.compiler.used rather than llvm.used.
This should cover most temporary buffers, except for those used by indirected value arguments, which need some cooperation with CallEmission to properly mark lifetime end after the call's completed.
Modeling nonescaping captures as @inout parameters is wrong, because captures are allowed to share state, unlike 'inout' parameters, which are allowed to assume to some degree that there are no aliases during the parameter's scope. To model this, introduce a new @inout_aliasable parameter convention to indicate an indirect parameter that can be written to, not only by the current function, but by well-typed, well-synchronized aliasing accesses too. (This is unrelated to our discussions of adding a "type-unsafe-aliasable" annotation to pointer_to_address to allow for safe pointer punning.)
This is a bit of a hodge-podge of related changes that I decided
weren't quite worth teasing apart:
First, rename the weak{Retain,Release} entrypoints to
unowned{Retain,Release} to better reflect their actual use
from generated code.
Second, standardize the names of the rest of the entrypoints around
unowned{operation}.
Third, standardize IRGen's internal naming scheme and API for
reference-counting so that (1) there are generic functions for
emitting operations using a given reference-counting style and
(2) all operations explicitly call out the kind and style of
reference counting.
Finally, implement a number of new entrypoints for unknown unowned
reference-counting. These entrypoints use a completely different
and incompatible scheme for working with ObjC references. The
primary difference is that the new scheme abandons the flawed idea
(which I take responsibility for) that we can simulate an unowned
reference count for ObjC references, and instead moves towards an
address-only scheme when the reference might store an ObjC reference.
(The current implementation is still trivially takable, but that is
not something we should be relying on.) These will be tested in a
follow-up commit. For now, we still rely on the bad assumption of
reference-countability.
SILFunctionType of the method instead of its formal type.
Gives more accurate information to the @encoding, makes
foreign error conventions work implicitly, and allows
IRGen's Swift-to-Clang to avoid duplicating arbitrary
amounts of the bridging logic from SILGen.
Some finagling was required in order to avoid calling
getConstantFunctionType from within other kinds of
lowering, which might have re-entered a generic context.
Also required fixing a bug with the type lowering of
optional DynamicSelfTypes where we would end up with
a substituted type in the lowered type.
Also, for some reason, our @encoding for -dealloc
methods was pretending that there was a formal parameter.
There didn't seem to be any justification for this,
and it's not like Clang does that. Fixed.
This commit reapplies r29266 with a conservative build fix
that disables ObjC property descriptors for @objc properties
that lack a getter. That should only be possible in SIL
files, because @objc should force accessors to be synthesized.
Arguably, Sema shouldn't be marking things implicitly @objc
in SIL files, but I'll leave that decision open for now.
Swift SVN r29272
SILFunctionType of the method instead of its formal type.
Gives more accurate information to the @encoding, makes
foreign error conventions work implicitly, and allows
IRGen's Swift-to-Clang to avoid duplicating arbitrary
amounts of the bridging logic from SILGen.
Some finagling was required in order to avoid calling
getConstantFunctionType from within other kinds of
lowering, which might have re-entered a generic context.
Also required fixing a bug with the type lowering of
optional DynamicSelfTypes where we would end up with
a substituted type in the lowered type.
Also, for some reason, our @encoding for -dealloc
methods was pretending that there was a formal parameter.
There didn't seem to be any justification for this,
and it's not like Clang does that. Fixed.
Swift SVN r29266
After latest set of fixes I only ran the tests on iPhone, which disables
the IRGen test. Ugh...
- r29261 removed an IBAction check in requiresObjCMethodDescriptor() --
turns out the IBAction attribute does not imply @objc, so this was
wrong.
- r29263 needed a test update for @nonobjc override semantics.
Swift SVN r29265
Now that generic subclasses of @objc classes are supported, dust off
Doug Gregor's fix for <rdar://problem/20385288>. It is now an error
to override an @objc declaration with something that cannot be
represented as @objc.
For example, the override of foo() here will not compile unless
it is explicitly marked @nonobjc:
func foo(i: Int) {}
...
override func foo(i: Int?) {}
Note that I updated IRGen to delete some logic for figuring out when
to emit @objc metadata. We can now rely on Sema to correctly set
isObjC(), instead of checking overrides ourselves. This was wrong
anyway, now that we can have @nonobjc overrides of @objc methods,
and vice versa.
Swift SVN r29263
It's not worth burning more than three registers on a parameter, and doing so causes code size issues for large structs and enums. Make it so that values with more than three explosion members get passed indirectly, just like they get returned indirectly.
This time, modify emitPartialApplyForwarder not to attempt to 'tail' call the original function when indirect arguments get alloca'ed on the stack, which is UB, and don't use "byval", as suggested by John.
Swift SVN r29032
It's not worth burning more than three registers on a parameter, and doing so causes code size issues for large structs and enums. Make it so that values with more than three explosion members get passed indirectly, just like they get returned indirectly.
Swift SVN r29016
Using LLVM large integers to represent enum payloads has been causing compiler performance and code size problems with large types, and has also exposed a long tail of backend bugs. Replace them with an "EnumPayload" abstraction that manages breaking a large opaque binary value into chunks, along with masking, testing, and extracting typed data from the binary blob. For now, use a word-sized chunking schema always, though the architecture here is set up to eventually allow the use of an arbitrary explosion schema, which would benefit single-payload enums by allowing the payload to follow the explosion schema of the contained value.
This time, adjust the assertion in emitCompare not to perform a check before we've established that the payload is empty, since APInt doesn't have a 0-bit state and the default-constructed form is nondeterminisitic. (We should probably use a more-tailored representation for enum payload bit patterns than APInt or ClusteredBitVector.)
Swift SVN r28985
Using LLVM large integers to represent enum payloads has been causing compiler performance and code size problems with large types, and has also exposed a long tail of backend bugs. Replace them with an "EnumPayload" abstraction that manages breaking a large opaque binary value into chunks, along with masking, testing, and extracting typed data from the binary blob. For now, use a word-sized chunking schema always, though the architecture here is set up to eventually allow the use of an arbitrary explosion schema, which would benefit single-payload enums by allowing the payload to follow the explosion schema of the contained value.
Swift SVN r28982
All llvm::Functions created during IRGen will have target-cpu and target-features
attributes if they are non-null.
Update testing cases to expect the attribute in function definition.
Add testing case function-target-features.swift to verify target-cpu and
target-features.
rdar://20772331
Swift SVN r28186
These aren't really orthogonal concerns--you'll never have a @thick @cc(objc_method), or an @objc_block @cc(witness_method)--and we have gross decision trees all over the codebase that try to hopscotch between the subset of combinations that make sense. Stop the madness by eliminating AbstractCC and folding its states into SILFunctionTypeRepresentation. This cleans up a ton of code across the compiler.
I couldn't quite eliminate AbstractCC's information from AST function types, since SIL type lowering transiently created AnyFunctionTypes with AbstractCCs set, even though these never occur at the source level. To accommodate type lowering, allow AnyFunctionType::ExtInfo to carry a SILFunctionTypeRepresentation, and arrange for the overlapping representations to share raw values.
In order to avoid disturbing test output, AST and SILFunctionTypes are still printed and parsed using the existing @thin/@thick/@objc_block and @cc() attributes, which is kind of gross, but lets me stage in the real source-breaking change separately.
Swift SVN r27095
"Autoclosure" is uninteresting to SIL. "noescape" isn't currently used by SIL and we shouldn't have it until it has a meaningful effect on SIL. "throws" should be adequately represented by a SIL function type having an error result.
Swift SVN r27023
The set of attributes that make sense at the AST level is increasingly divergent from those at the SIL level, so it doesn't really make sense for these to be the same. It'll also help prevent us from accidental unwanted propagation of attributes from the AST to SIL, which has caused bugs in the past. For staging purposes, start off with SILFunctionType's versions exactly the same as the FunctionType versions, which necessitates some ugly glue code but minimizes the potential disruption.
Swift SVN r27022
Previously some parts of the compiler referred to them as "fields",
and most referred to them as "elements". Use the more generic 'elements'
nomenclature because that's what we refer to other things in the compiler
(e.g. the elements of a bracestmt).
At the same time, make the API better by providing "getElement" consistently
and using it, instead of getElements()[i].
NFC.
Swift SVN r26894
As part of this, re-arrange the argument order so that
generic arguments come before the context, which comes
before the error result. Be more consistent about always
adding a context parameter on thick functions, even
when it's unused. Pull out the witness-method Self
argument so that it appears last after the error
argument.
Swift SVN r26667
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
ObjC metatypes have the same extra inhabitants and spare bits as other unmanaged untagged object references. Fixes rdar://problem/19318533.
Swift SVN r24623
IRGen uses a typedef, SpareBitVector, for its principal
purpose of tracking spare bits. Other uses should not
use this typedef, and I've tried to follow that, but I
did this rewrite mostly with sed and may have missed
some fixups.
This should be almost completely NFC. There may be
some subtle changes in spare bits for witness tables
and other off-beat pointer types. I also fixed a bug
where IRGen thought that thin functions were two
pointers wide, but this wouldn't have affected anything
because we never store thin functions anyway, since
they're not a valid AST type.
This commit repplies r24305 with two fixes:
- It fixes the computation of spare bits for unusual
integer types to use the already-agreed-upon type
size instead of recomputing it. This fixes the
i386 stdlib build. Joe and I agreed that we should
also change the size to use the LLVM alloc size
instead of the next power of 2, but this patch
does not do that yet.
- It changes the spare bits in function types back
to the empty set. I'll be changing this in a
follow-up, but it needs to be tied to runtime
changes. This fixes the regression test failures.
Swift SVN r24324
IRGen uses a typedef, SpareBitVector, for its principal
purpose of tracking spare bits. Other uses should not
use this typedef, and I've tried to follow that, but I
did this rewrite mostly with sed and may have missed
some fixups.
This should be almost completely NFC. There may be
some subtle changes in spare bits for witness tables
and other off-beat pointer types. I also fixed a bug
where IRGen thought that thin functions were two
pointers wide, but this wouldn't have affected anything
because we never store thin functions anyway, since
they're not a valid AST type.
Swift SVN r24305
isn't used yet, but will be for modeling the self argument passed to an
address-only witness implementation. NFC since all this code is dead :-)
Swift SVN r23857
It's not always correct to map a Swift Bool back to ObjCBool in C land, since Bool could have originally been a proper _Bool. Pass the clang::Decl down to type lowering so we can recognize this. We still don't have a great solution for block types, because there's no decl to refer to, and Swift's user-level type system erases the distinction between void(^)(_Bool) and void(^)(BOOL). However, this is enough to let us start using C APIs that traffic in _Bool.
Swift SVN r23546
