Similar to the non-resilient case, except we also emit a 'relocation
function'. The class descriptor now contains this relocation function
if the class has resilient ancestry, and the relocation function
calls the runtime's swift_relocateClassMetadata() entry point.
The metadata completion function calls swift_initClassMetadata() and
does layout, just like the non-resilient case.
Fixes <rdar://problem/40810002>.
Now that we don't need the superclass before calling
swift_relocateClassMetadata(), it seems simpler to set it
here instead of doing it in various places in IRGen.
Using the superclass metadata here no longer makes sense with two-phase
init, in case the superclass metadata depends on the class being
instantiated.
It would also be nice to rework the resilient class metadata 'pattern'
to be its own data structure that's true const, instead of just the
prefix of a real class metadata, but for now let's keep the existing
crappy design.
If a class has generic ancestry or resiliently-sized fields, but is
itself not generic and does not have resilient ancestry, we must
perform runtime metadata initialization, but we can initialize
the metadata in-place.
As with generic classes or classes with resilient ancestry, we
copy generic requirements and field offset vectors from the
superclass. We also calculate the layout of the class's direct
fields at runtime.
Unlike the fully resilient case, we don't copy vtable entries
from the superclass, or install the direct class's vtable
entries from the type context descriptor. Instead, we statically
emit the vtable as with fixed-size class metadata.
Both the in-place and resilient case call the same runtime
entry point to initialize class metadata; the new HasStaticVTable
flag in ClassLayoutFlags is used to select between the two
behaviors concerning the vtable.
These functions don't accept local variable heap memory, although the names make it sound like they work on anything. When you try, they mistakenly identify such things as ObjC objects, call through to the equivalent objc_* function, and crash confusingly. This adds Object to the name of each one to make it more clear what they accept.
rdar://problem/37285743
This saves us some expensive cross-referencing and caching in the runtime, and lets us reclaim the `isReflectable` bit from the context descriptor flags (since a null field descriptor is a suitable and more accurate indicator of whether a type is reflectable).
This allows us to layout-optimize Optional<T> when T is a struct with an
extra-inhabitant-bearing field anywhere in its definition, not only at
the beginning. rdar://problem/43019427
By design, we don't want private or function-nested types to be accessible by mangled name, since they don't have stable identities, and they could inadvertently become ABI if someone serialized a mangled string and expected to deserialize it into a type. Fixes rdar://problem/39826794 .
The token contents doesn't really matter, but it can't start with a digit if it's going to show up in mangled names using identifier grammar. `s/0x/$/` for some 80s flair.
- Instead of keeping multiple flags in the type descriptor flags,
just keep a single flag indicating the presence of additional
import information after the name.
- That import information consists of a sequence of null-terminated
C strings, terminated by an empty string (i.e. by a double null
terminator), each prefixed with a character describing its purpose.
- In addition to the symbol namespace and related entity name,
include the ABI name if it differs from the user-facing name of the
type, and make the name the user-facing Swift name.
There's a remaining issue here that isn't great: we don't correctly
represent the parent relationship between error types and their codes,
and instead we just use the Clang module as the parent. But I'll
leave that for a later commit.
Previously, when a tuple type had non-fixed layout, we would compute
a layout by building the metadata for that tuple type and then
extracting the layout from the VWT. This can be quite expensive
because it involves constructing the exact metadata for types like
arrays and functions despite those types being fixed-layout across
all instantiations. It also tends to cause unnecessary recursive-type
issues, especially with enums where tuples are currently used to model
cases with mutliple payloads. Since we just need a layout, computing
it directly from element layouts instead of constructing metadata for
the formal type lets us take advantage of all the other fast paths for
layout construction, e.g. for fixed types and single-field aggregates.
This is a good improvement overall, but it also serves to alleviate
some of the problems of rdar://40810002 / SR-7876 in a way that
might be suitable for integration to 4.2.
- `swift_getForeignTypeMetadata` is now a request/response function.
- The initialization function is now a completion function, and the
pointer to it has moved into the type descriptor.
- The cache variable is no longer part of the ABI; it's an
implementation detail of the access function.
- The two points above mean that there is no special header on foreign
type metadata and therefore that they can be marked constant when
there isn't something about them that needs to be initialized.
The only foreign-metadata initialization we actually do right now is
of the superclass field of a foreign class, and since that relationship
is a proper DAG, it's not actually possible to have recursive
initialization problems. But this is the right long-term thing to do,
and it removes one of the last two clients of once-based initialization.
The prefab'ed value witness tables for reference storage types are a
premature optimization. Not all scenarios are covered, and those that
are "look suspect" according to John McCall.
As part of this, rename TypeMetadataRecordKind to TypeReferenceKind
and consistently give it three bits of storage.
The better modelling of these type references appears to have been
sufficient to make dynamic conformance checks succeed, which is good
but unexpected.
From what I see the only fields are DATA_VALUE_WITNESS which
all have type size_t. I converted them to use the target-dependent
`StoredSize`. While I was around I fixed also isValueInline()
to do the right thing (it was using ValueBuffer instead of
TargetValueBuffer) and all the getters for the data value witnesses.
<rdar://problem/41546568>
Rather than storing a mangled name in a Swift protocol descriptor,
which encodes information that is redundant with the context of the
protocol, store an unmangled name as in nominal type descriptors. Update
the various places where this name is used to extract the demangle
tree from the context descriptors.
Reimplement protocol descriptors for Swift protocols as a kind of
context descriptor, dropping the Objective-C protocol compatibility
layout. The new protocol descriptors have several advantages over the
current implementation:
* They drop all of the unused fields required for layout-compatibility
with Objective-C protocols.
* They encode the full requirement signature of the protocol. This
maintains more information about the protocol itself, including
(e.g.) correctly encoding superclass requirements.
* They fit within the general scheme of context descriptors, rather than
being their own thing, which allows us to share more code with
nominal type descriptors.
* They only use relative pointers, so they’re smaller and can be placed
in read-only memory
Implements rdar://problem/38815359.
In a generic requirement, distinguish between Swift and
Objective-C protocols using a spare bit within the relative
(indirectable) reference to the protocol.
Switch one entry point in the runtime (swift_getExistentialTypeMetadata)
to use ProtocolDescriptorRef rather than a protocol descriptor. Update
IRGen to produce ProtocolDescriptorRef instances for its calls, setting
the discriminator bit appropriately.
Within the runtime, verify that all instances of ProtocolDescriptorRef have
the right layout, i.e., the discriminator bit is set for @objc protocols
but not Swift protocols.
Use ProtocolDescriptorRefs within the runtime representation of
existential type metadata (TargetExistentialTypeMetadata) instead of
bare protocol descriptor pointers. Start rolling out the use of
ProtocolDescriptorRef in a few places in the runtime that touch this
code. Note that we’re not yet establishing any strong invariants on
the TargetProtocolDescriptorRef instances.
While here, replace TargetExistentialTypeMetadata’s hand-rolled pointer
arithmetic with swift::ABI::TrailingObjects and centralize knowledge of
its layout better.
'const T *' isn't compatible with a function pointer, so upstream
Clang complained about the 'patch_t' convenience constructor we were
using. It's not like we need general functionality or convenience
here, so just pass the members of the patch_t type separately and
without any templating, and drop it entirely.
No functionality change.
Follow-up to 3ed3774e07. On Apple OSs that don't have the new
Objective-C runtime function 'objc_setHook_getImageName', override the
system definition of 'class_getImageName' by literally rewriting
symbol tables at run time.
Yes, you read that correctly.
The low-level part of this patch was written by Greg Parker, then
simplified and tweaked by me to fit the Swift coding style. Don't try
this at home; it comes with all sorts of caveats and won't actually
work on this year's iOS. (Fortunately we don't need it there, because
that will have the new ObjC entry point.)
The rest of the patch is pretty straightforward: the replacement
implementation calls the code that supports Swift objects (the same
code we use on newer OSs), which then chains back to the original
system implementation of class_getImageName. May we never have to
touch this again.
rdar://problem/41535552
This not only restores the correct behavior for classes with generic
ancestry, but also handles actual generic classes as well. (This is
the function that backs Foundation's Bundle.init(for: AnyClass)
initializer.)
https://bugs.swift.org/browse/SR-1917
rdar://problem/33450609&40367300
