The main point of this change is to make sure that a shared function always has a body: both, in the optimizer pipeline and in the swiftmodule file.
This is important because the compiler always needs to emit code for a shared function. Shared functions cannot be referenced from outside the module.
In several corner cases we missed to maintain this invariant which resulted in unresolved-symbol linker errors.
As side-effect of this change we can drop the shared_external SIL linkage and the IsSerializable flag, which simplifies the serialization and linkage concept.
This ensures that opened archetypes always inherit any outer generic parameters from the context in which they reside. This matters because class bounds may bind generic parameters from these outer contexts, and without the outer context you can wind up with ill-formed generic environments like
<τ_0_0, where τ_0_0 : C<T>, τ_0_0 : P>
Where T is otherwise unbound because there is no entry for it among the generic parameters of the environment's associated generic signature.
We now schedule conformance emissions in basically the same way
we do for types and declarations, which means that we'll emit them
uniquely in the module file instead of redundantly at every use.
This should produce substantially smaller module files overall,
especially for modules that heavily use generics. It also means
that we can remove all the unfortunate code to support using
different abbrev codes for them in different bitcode blocks.
Requirement lists are now emitted inline in the records that need
them instead of as trailing records. I think this will improve
space usage, but mostly it assists in eliminating the problem
where abbrev codes are shared between blocks.
The main effect of this will be that in IRGen we will use llvm.dbg.addr instead
of llvm.dbg.declare. We must do this since llvm.dbg.declare implies that the
given address is valid throughout the program.
This just adds the instructions/printing/parsing/serialization/deserialization.
rdar://85020571
This is an instruction that I am going to use to drive some of the ownership
based dataflow optimizations that I am writing now. The instruction contains a
kind that allows one to know what type of checking is required and allows the
need to add a bunch of independent instructions for independent checkers. Each
checker is responsible for removing all of its own mark instructions. NOTE:
MarkMustCheckInst is only allowed in Raw SIL since once we are in Canonical SIL
we want to ensure that all such checking has already occurred.
Store a list of argument effects in a function, which specify if and how arguments escape.
Such effects can be specified in the Swift source code (for details see docs/ReferenceGuides/UnderscoredAttributes.md) or derived in an optimization pass.
For details see the documentation in SwiftCompilerSources/Sources/SIL/Effects.swift.
In addition to the predefined cases, like "readnone", "readonly", etc. support providing a custom string, which will be parsed later.
Also, allow multiple effects attributes to be put onto a function.
The `@exclusivity(unchecked)` attribute can be used on variables to selectively disable exclusivity checking.
For completeness, also the `@exclusivity(checked)` variant is supported: it turns on exclusivity checking for specific variables if exclusivity enforcement is disabled by the command line option.
This new attribute is a missing implementation part of SE-0176 (https://github.com/apple/swift-evolution/blob/main/proposals/0176-enforce-exclusive-access-to-memory.md).
rdar://31121356
Nested archetypes are represented by their base archetype kinds (primary,
opened, or opaque type) with an interface type that is a nested type,
as represented by a DependentMemberType. This provides a more uniform
representation of archetypes throughout the frontend.
Form opened archetype types based on an interface type and existential
type, rather than assuming all OpenedArchetypeType instances only
represent the root. Sink the UUID, existential type, and actual creation
of the opened archetype into the opened generic environment, so we
consistently only create new archetype instances from the generic
environment. This slims down OpenedArchetypeType and makes it work
similarly to the other archetype kinds, as well as generalizing it
to support nested types.
Sink the existential type and UUID of an
As another step toward eliminating NestedArchetypeType, generalize the
representation, construction, and serialization of primary and sequence
archetypes to interface types, rather than generic parameter types.
Introduce a new instruction `dealloc_stack_ref ` and remove the `stack` flag from `dealloc_ref`.
The `dealloc_ref [stack]` was confusing, because all it does is to mark the deallocation of the stack space for a stack promoted object.
There are three major changes here:
1. The addition of "SILFunctionTypeRepresentation::CXXMethod".
2. C++ methods are imported with their members *last*. Then the arguments are switched when emitting the IR for an application of the function.
3. Clang decls are now marked as foreign witnesses.
These are all steps towards being able to have C++ protocol conformance.
The new type, called ExistentialType, is not yet used in type resolution.
Later, existential types written with `any` will resolve to this type, and
bare protocol names will resolve to this type depending on context.
Adding the ability to add an optional message to the unavailable from
async attribute. This can be used to indicate other possible API to use,
or help explain why it's unavailable.
This instruction is similar to a copy_addr except that it marks a move of an
address that has to be checked. In order to keep the memory lifetime verifier
happy, the semantics before the checker runs are the mark_unresolved_move_addr is
equivalent to copy_addr [init] (not copy_addr [take][init]).
The use of this instruction is that Mandatory Inlining converts builtin "move"
to a mark_unresolved_move_addr when inlining the function "_move" (the only
place said builtin is invoked).
This is then run through a special checker (that is later in this PR) that
either proves that the mark_unresolved_move_addr can actually be a move in which
case it converts it to copy_addr [take][init] or if it can not be a move, emit
an error and convert the instruction to a copy_addr [init]. After this is done
for all instructions, we loop back through again and emit an error on any
mark_unresolved_move_addr that were not processed earlier allowing for us to
know that we have completeness.
NOTE: The move kills checker for addresses is going to run after Mandatory
Inlining, but before predictable memory opts and friends.
We've recently added the -experimental-hermetic-seal-at-link compiler flag,
which turns on aggressive dead-stripping optimizations and assumes that library
code can be optimized against client code because all users of the library
code/types are present at link/LTO time. This means that any module that's
built with -experimental-hermetic-seal-at-link requires all clients of this
module to also use -experimental-hermetic-seal-at-link. This PR enforces that
by storing a bit in the serialized module, and checking the bit when importing
modules.
This cleans up 90 instances of this warning and reduces the build spew
when building on Linux. This helps identify actual issues when
building which can get lost in the stream of warning messages. It also
helps restore the ability to build the compiler with gcc.
Required for UnsafeRawPointer.withMemoryRebound(to:)
%token = bind_memory %0 : $Builtin.RawPointer, %1 : $Builtin.Word to $T
%0 must be of $Builtin.RawPointer type
%1 must be of $Builtin.Word type
%token is an opaque $Builtin.Word representing the previously bound types
for this memory region.
This is a signal to the move value kill analysis that this is a move that should
have diagnostics emitted for it. It is a temporary addition until we add
MoveOnly to the SIL type system.
Stage in the parsing for this attribute, nothing else.
Motivated by two important reasons:
1) The pitch for variadic generics does not lay down a concrete syntax
for variadic generic parameters.
2) Paring T... and T* needlessly complicate the lexer as we must now
disambiguate them with respect to other internal operator characters
(e.g. `T...>` must lex as `(T...)>` and not `T ...>`
Which itself adds another motivation
3) We need to start parsing this attribute *now* to avoid condfail'ing
ourselves later.
I am purposely doing this in SILGen rather than at the type system level to
avoid having to have to add a bunch of boilerplate to the type system. Instead
of doing that, I am in SILGen checking for the isNoImplicitCopy bit on the
ParamDecl when we emit arguments. At that point, I set on the specific
SILArgument being emitted the bit that it is no implicit copy. In terms of
printing at the SIL level, I just printed it in front of the function argument
type like @owned, e.x.:
func myFunc(_ x: @_noImplicitCopy T) -> T {
...
}
becomes:
bb0(%0 : @noImplicitCopy @owned $T):
Some notes:
* Just to be explicit, I am making it so that no implicit copy parameters by
default are always passed at +1. The reason why I think this makes sense is
that this is the natural way of working with a move only value.
* As always, one can not write no implicit copy the attribute without passing
the flag -enable-experimental-move-only so this is NFC.
rdar://83957088
