SILInstruction::clone doesn't know how to clone instructions that produce
the archetype uuid. SILCloner is equipped to handle such instructions.
Optimizations like LoopRotate use SILInstruction::clone and will be
incorrect for such instructions.
rdar://130047619
Now that the underlying issue (PrunedLiveness' merging of summaries for
branch instructions) has been fixed, reinstate lifetime completion and
add a test to verify that it behaves correctly.
This reverts commit c552b90b61
("Temporarily turn off completing lifetimes of block arguments").
rdar://130427564
When we have a non trivial enum lexical stack location, and we store a
none value, after mem2reg we can end up with an ownership error
because of the leaking move_value [lexical].
This change avoids creating move_value [lexical] for such stores.
Do the same when we have store_borrow of none values, even though there
is no ownership error in this case, handle it like store for consistency
A guaranteed function argument is live for the duration of the function.
It should be safe to allow TempRValueOpt when it is the base of the copy source of a lexical alloc_stack.
For a mark_dependence user of the stack, look through the mark_dependence chain
to match the partial_apply.
Since this is a superset of the `markDepChain` logic, remove that.
So that we don't have 2 ways to check for the same.
This corresponds to the parameter-passing convention of the Itanium C++
ABI, in which the argument is passed indirectly and possibly modified,
but not destroyed, by the callee.
@in_cxx is handled the same way as @in in callers and @in_guaranteed in
callees. OwnershipModelEliminator emits the call to destroy_addr that is
needed to destroy the argument in the caller.
rdar://122707697
The de-virtualizer utility didn't handle indirect error results when de-virtualizing class or actor methods.
This resulted in a missing argument for the indirect error result in the new try_apply instruction.
rdar://130545338
Currently not all types are visited in canSerialize* calls, sometimes
resulting in an internal type getting @usableFromInline, which is
incorrect.
For example, for `let q = P() as? Q`, where Q is an internal class
inherting a public class P, Q is not visited in the canSerialize*
checks, thus resulting in `@usableFromInline class Q`; this is not
the intended behavior in the conservative mode used by PackageCMO
as it modifies AST.
To properly fix, instruction visitor needs to be refactored to do
both the "canSerialize" check (that visits all types) and serialize
or update visibility (modify AST in non-conservative modes).
This PR provides a short-term fix that prevents modifying AST, and
also ensures that the generated interfaces with PackageCMO flags
are not affected by the optimization or contain modified AST.
rdar://130292190
Otherwise, we will have differing isolation from other parameters since
the isolations will look different since one will have the .none value
as an instance and the other will not have one and instead will rely on
the AST isolation info. That is the correct behavior here since we do
not actually have an actor here.
I also removed some undefined behavior in the merging code. The way the
code should work is that we should check if the merge fails and in such
a case emit an unknown pattern error... instead of not checking
appropriately on the next iteration and hitting undefined behavior.
rdar://130396399
- While an opaque borrow access occurs to part of a value, the entire scope of
the access needs to be treated as a liveness range, so add the `EndAccess`es
to the liveness range.
- The SIL verifier may crash the compiler on SILGen-generated code when the
developer's source contains consume-during-borrow code patterns. Allow
`load_borrow` instructions to be marked `[unchecked]`, which suppresses
verifier checks until the move checker runs and gets a chance to properly
diagnose these errors.
Fixes rdar://124360175.
From the perspective of the IR, we are changing SILIsolationInfo such that
inferring an actor instance means looking at equivalence classes of values where
we consider operands to look through instructions to be equivalent to their dest
value. The result is that cases where the IR maybe puts in a copy_value or the
like, we consider the copy_value to have the same isolation info as using the
actor directly. This prevents a class of crashes due to merge failings. Example:
```swift
actor MyActor {
init() async {
init(ns: NonSendableKlass) async {
self.k = NonSendableKlass()
self.helper(ns)
}
func helper(_ newK: NonSendableKlass) {}
}
```
Incidently, we already had a failing test case from this behavior rather than
the one that was the original genesis. Specifically:
1. If a function's SILIsolationInfo is the same as the isolation info of a
SILValue, we assume that no transfer actually occurs.
2. Since we were taking too static of a view of actor instances when comparing,
we would think that a SILIsolationInfo of a #isolation parameter to as an
argument would be different than the ambient's function isolation which is also
that same one. So we would emit a transfer non transferrable error if we pass in
any parameters of the ambient function into another isolated function. Example:
```swift
actor Test {
@TaskLocal static var local: Int?
func withTaskLocal(isolation: isolated (any Actor)? = #isolation,
_ body: (consuming NonSendableValue, isolated (any Actor)?) -> Void) async {
Self.$local.withValue(12) {
// We used to get these errors here since we thought that body's isolation
// was different than the body's isolation.
//
// warning: sending 'body' risks causing data races
// note: actor-isolated 'body' is captured by a actor-isolated closure...
body(NonSendableValue(), isolation)
}
}
}
```
rdar://129400019
When devirtualizing witness method calls, it can happen that we need a cast between ABI compatible return types.
We were missing supporting type casts between nominal types which are ABI compatible.
This comes from whole-module reasoning of protocol conformances.
If a protocol only has a single conformance where the associated type (`ID`) is some concrete type (e.g. `Int`), then the devirtualizer knows that `p.get()` can only return an `Int`:
```
public struct X2<ID> {
let p: any P2<ID>
public func testit(i: ID, x: ID) -> S2<ID> {
return p.get(x: x)
}
}
```
and after devirtualizing the `get` function, its result must be cast from `Int` to `ID`.
The `layoutIsTypeDependent` utility is basically only used here to assert that this cast can only happen between layout compatible types.
rdar://129004015
This is usually the case. Some examples, where they layout is _not_ dependent:
```
struct S<T> {
var x: Int // no members which depend on T
}
struct S<T> {
var c: SomeClass<T> // a class reference does not depend on the layout of the class
}
```
Create two versions of the following functions:
isConsumedParameter
isGuaranteedParameter
SILParameterInfo::isConsumed
SILParameterInfo::isGuaranteed
SILArgumentConvention::isOwnedConvention
SILArgumentConvention::isGuaranteedConvention
These changes will be needed when we add a new convention for
non-trivial C++ types as the functions will return different answers
depending on whether they are called for the caller or the callee. This
commit doesn't change any functionality.
- Keep witness thunk linkage private for a package protocol member in SILGen.
- Optimize private/hidden functions during Package CMO; if they don't contain
references that have private/hidden symbols, serialize them and set the linkage
to shared. For unserialized witness thunks, set the linkage to package, so the
witness table itself can be serialized.
- Update witness table and vtable serialization.
Resolves rdar://129976582
